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 |
|---|---|---|---|---|---|
Can deliberate interaction between the public and persons affected by leprosy reduce stigmatization ? The study described in this paper hypothesises that it can and assesses the effectiveness of a ‘contact intervention’ . This cluster-randomized controlled intervention study is part of the Stigma Assessment and Reduction of Impact ( SARI ) project conducted in Cirebon District , Indonesia . Testimonies , participatory videos and comics given or made by people affected by leprosy were used as methods to facilitate a dialogue during so-called ‘contact events’ . A mix of seven quantitative and qualitative methods , including two scales to assess aspects of stigma named the SDS and EMIC-CSS , were used to establish a baseline regarding stigma and knowledge of leprosy , monitor the implementation and assess the impact of the contact events . The study sample were community members selected using different sampling methods . The baseline shows a lack of knowledge about leprosy , a high level of stigma and contrasting examples of support . In total , 91 contact events were organised in 62 villages , directly reaching 4 , 443 community members ( mean 49 per event ) . The interview data showed that knowledge about leprosy increased and that negative attitudes reduced . The adjusted mean total score of the EMIC-CSS reduced by 4 . 95 points among respondents who had attended a contact event ( n = 58; p <0 . 001 , effect size = 0 . 75 ) compared to the score at baseline ( n = 213 ) ; for the SDS this was 3 . 56 ( p <0 . 001 , effect size = 0 . 81 ) . About 75% of those attending a contact event said they shared the information with others ( median 10 persons ) . The contact intervention was effective in increasing knowledge and improving public attitudes regarding leprosy . It is relatively easy to replicate elsewhere and does not require expensive technology . More research is needed to improve scalability . The effectiveness of a contact intervention to reduce stigma against other neglected tropical diseases and conditions should be evaluated .
The body of theory and research on stigma has been developed over fifty years since its original conceptualization by Goffman [33] . Over time , new conceptualizations of health-related stigma have been developed to improve the relevance of the concept , by scholars from different scientific disciplines ( e . g . psychology , psychiatry , public health , sociology , anthropology ) and sometimes for specific health conditions [34–38] . Not surprisingly , these conceptualizations of stigma vary [34] and a major critique is that the concept of stigma has not become any clearer . Recently , Staples argued that stigma “can become a lazy shortcut for multiple ‘social aspects’ of leprosy” [39] . He underlines the importance of a more reflexive approach in the light of the critiques . For this paper it is important to distinguish between the interacting categories or levels of stigma that are articulated in the literature . Weiss [3] identified two categories ( stigmatized and stigmatizers ) , Bos et al . [38] three levels ( individual , social and structural ) and Heijnders & van der Meij [14] five levels ( intrapersonal , interpersonal , community , institutional and structural ) . The contact intervention primarily aims to reduce stigma from so-called ‘stigmatizers’ at the social or community level . Link & Phelan’s work on stigma is interesting in this context , as it gives more in-depth insights in the processes operating at this level . In their definition , stigma exists when “elements of labelling , stereotyping , separating , status loss and discrimination co-occur in a power situation that allows these processes to unfold” . The following five components need to converge: In the first component , people distinguish and label human differences . In the second , dominant cultural beliefs link labelled persons to undesirable characteristics—to negative stereotypes . In the third , labelled persons are placed in distinct categories so as to accomplish some degree of separation of “us” from “them . ” In the fourth , labelled persons experience status loss and discrimination that lead to unequal outcomes . Finally , stigmatization is entirely contingent on access to social , economic , and political power that allows the identification of differentness , the construction of stereotypes , the separation of labelled persons into distinct categories , and the full execution of disapproval , rejection , exclusion , and discrimination [24] . Link & Phelan’s work focuses on the nature and consequences of stigma rather than the sources . These are also important for this study . Sermrittirong & van Brakel categorized the causes of leprosy-related stigma in external manifestations of the disease , cultural and religious beliefs , fear of transmission , association with people considered inferior and public health-related interventions [40] . In conclusion , this paper focuses on the impact of a contact-based intervention on public stigma , targeting the processes of labelling , stereotyping and separation . The interventions incorporate the underlying causes of stigma by addressing the knowledge , beliefs , fears and questions people have regarding the stigmatised condition . The contact intervention was developed over the course of the first one and a half years of the SARI project . The SARI project was carried out by a research team including ten research assistants ( RAs ) from around Cirebon who spoke the local languages and researchers from VU University Amsterdam and Universitas Indonesia . Three RAs had either a physical or visual disability and one was affected by leprosy . One of the PhD students and the principal investigator ( I ) are also persons with a disabilities . Conditions under which the contact intervention seems to have the greatest impact articulated in the literature were considered during the design of the intervention . These include an equal status between participants , one-on-one contact , frequent contact , contact with individuals who mildly deviate from the stereotype , an element of education , high levels of intimacy , real-world opportunities to interact and institutional support [14 , 19] . In addition , the SARI teams’ experience and understanding of the local context ( see also [8 , 9 , 41] ) at that time were important considerations for the development of the interventions . We decided to organise so called ‘contact events’ at a local level , for instance , in schools , village halls and mosques . The local RAs together with persons affected by leprosy became responsible for the organisation of the events and had to make sure the events fitted into and made optimal use of the local social structures and context . Each contact event had two core elements . The first core element is contact between affected persons and the public . An exploratory study was conducted in Cirebon , which resulted in a range of possible methods to create direct and indirect contact . In the end , testimonies were chosen as the direct method and participatory videos and comics made by people affected by leprosy as the in-direct method . The rationale for this choice was: ( i ) participants of the exploratory study thought these methods would be effective , ( ii ) the stories and messages would come from people affected by leprosy themselves , ( iii ) the testimonies and the development of participatory videos and comics on their own were expected to be valuable and empowering experiences for the participants , ( iv ) the development process would be relatively inexpensive , which is important for scalability , ( v ) material was expected to be attractive , distributable and easy to understand , and ( vi ) the SARI team could build upon previous knowledge and experience from organisations like InsightShare ( see http://www . insightshare . org/ ) and World Comics ( see http://www . worldcomics . fi/ ) . During the course of the project , two participatory videos were developed telling stories that the makers wanted to tell to the community . They are titled ‘Pastikan badai sirna’ ( Surely the storm has vanished ) and ‘Empat sahabat yang selalu berbagi’ ( Four friends who always share ) ( paper in progress ) . In addition , 32 comics ( 4 panels each ) in black and white were developed by young people affected by leprosy depicting their life experiences and again bringing across a message they wanted to share with the community . Not every contact event incorporated all three methods; a selection was made based on the available time , venue and interest of the audience . The second core element is education . From the exploratory study we concluded that there are misconceptions about the causes of leprosy and the mode of transmission , and a lack of understanding regarding the social consequences of leprosy in Cirebon District ( see also Peters et al . [8] ) . In our opinion , the common view that stigma is caused by a lack of knowledge and that “education is therefore a panacea for stigma” [32] is wrong . This opinion is supported by others authors [5 , 15 , 39 , 42] . But that is not to say that education is not an essential component [15] . Wong states the importance of talking about local beliefs and changing these gradually through exploration and clarification [42] . The messages spread during the contact events using an interactive presentation are displayed in Box 1 . During the event the audience was encouraged to share the information they gained with family members , friends and others .
A variety of both qualitative and quantitative research methods have been applied to assess stigma in community members in Cirebon District ( see Table 1 for an overview ) . Each will be explained in more depth . The interviews and FGDs were recorded , transcribed and translated to English together with the summaries of informal interviews . The quantitative data was entered in an Epi Info for Windows database ( version 3 . 5 . 3 ) and analysed using Stata 12 . 1 and SPSS 21 . The qualitative data was analysed using N-Vivo and WeftQDA . Maps were made with QGIS 2 . 1 . 0 . Demographic variables included sex , age ( in years ) , marital status ( yes/no ) , education , profession , household income per day ( in Indonesian Rupiah ( IDR ) ) , key person ( yes/no ) and knowing a person affected by leprosy ( yes/no ) . The respondents were asked for either income per day or income per month; the latter was converted into one variable ‘household income per day’ by dividing the income per month by 30 . 5 . The differences between baseline and final survey respondents were tested using a Chi-square test for categorical variables and t-test for continuous variables . To investigate the effect of the interventions we calculated means and performed univariate and multivariate regression analyses . We used a backward elimination procedure considering the P-values and the model fit ( R2 ) . P values less than . 05 were taken as significant . In addition , the effect size ( ES ) was calculated , which is the difference between the mean ( M ) total score pre ( 1 ) and post ( 2 ) intervention divided by the pooled standard deviation ( SD ) [47] . The formula is: ES = M1-M2 ( ( SD12+SD22 ) /2 ) Following Cohen , an ES of 0 . 2 is considered as small , 0 . 5 as moderate , and 0 . 8 as large [47] . For the maps , the contact event in villages with 1 contact event are visualised at the centroid of the administrative boundary of the village . The contact events in villages with more than 1 contact event are visualised with the random point generator available in QGIS . The study was approved by the Ethics Committee of Atma Jaya University , the Sub-Directorate for Leprosy and Yaws , Ministry of Health , the Provincial Public Health Office , West Java , and the District Health Office , Cirebon District . Written informed consent was obtained from the participants . The control area in this study was a “care-as-usual” area . No incentives were offered to interviewees other than a small token of appreciation , such as a drinking mug or t-shirt , in particular when participants were interviewed more than once .
In a 14 month time-span the SARI project organised 91 contact events in 62 villages , which were located in 16 different sub-districts ( see Fig 1 ) . More than 4 , 400 community members attended a contact event ( see Fig 2 ) , including 803 key persons ( 18% ) . On average 49 community members attended one event . Contact events were most frequently held at the village hall , family house and school . Testimonies were provided in 55% of the contact events . See Table 4 for more detailed information on the contact events . Providing a testimony was both difficult and rewarding for the participants affected by leprosy . Sometimes they attended a contact event a few times without giving a testimony before they felt confident and ready to do so themselves . Two responses: I felt very emotional when I told them about my experience with leprosy . Maybe because it is the first time I ever shared my feelings . I am very happy that I can let it go . I am calm and happy that the audience can appreciate this event ( Event B44 ) I enjoyed this event , I even joked around with my current condition of being still single , I never thought the audience would respond like they did and prayed for me to have a spouse ( Event B13 ) Members in the audience: i ) were actively involved ( e . g . made notes , asked questions , shared personal experiences and gave an unplanned testimony ) , ii ) were emotionally touched by the testimonies , iii ) asked for or motivated others to get a health check-up due to a concerns being affected by leprosy , iv ) requested more comics to distribute in the community , v ) asked to start a participatory video process in their own sub-district , and finally vi ) requested more contact events . There were also challenges . These related to convincing key persons ( e . g . head of village ) of the value of a contact event , practical/logistical challenges ( e . g . weak audio system , inappropriate venue , too many people and limited time ) , the audience ( e . g . tired , less involved ) or to the SARI team ( e . g . cancelations , delayed ) . In this section of the results , the focus lies on the impact of the contact events on knowledge , attitudes and the coverage . During the informal interviews people in the audience stated the importance of and willingness to share this information with others . Respondents of the final survey who attended a contact event were asked whether they had shared the information of the event with others and if so with how many others . Of the 58 respondents , 44 ( 75 . 9% ) said they shared the information with others . On average , they shared this information with 10 others ( mean 16 . 5 , median 10 , minimum 2 and maximum 99 ) . In particular , religious leaders had shared the information to large groups during religious teaching ( shared with ~25 ) , during the praying for a person who passed away ( with ~60 ) and while reading Quran ( shared with ~100 ) . In the FGDs , the village leaders , women of the religious gatherings and students indicated they passed on the information about cause , symptoms and treatment to family , neighbours and friends . One woman said she had shared the information to students and instructed them to go to the HCs for medication if they have symptoms . Village leaders said they informed the public during religious gatherings and community meetings . In addition , village leaders and women of the religious gatherings also provided information about medication and the importance of visiting the HCs to the persons affected with leprosy .
The contact intervention implemented in this study was effective in increasing knowledge and improving reported attitudes in Cirebon District , Indonesia . The contact intervention is relatively easy to replicate elsewhere and does not require expensive technology , but more research is needed to improve scalability . Video and written materials that can guide practitioners on how they could design , implement and assess a contact intervention will follow . The material will be made available on Infolep , the international knowledge centre for access to information resources on leprosy and related subjects ( see http://www . leprosy-information . org/keytopic/sari-project ) . Findings of this study support evaluation of a contact intervention to reduce stigma against other neglected tropical diseases and conditions . | Stigma plays an important role in several neglected tropical diseases such as leprosy , Buruli ulcer , lymphatic filariasis , onchocerciasis and leishmaniasis . It negatively impacts individuals affected , and often also their families , and even communities . There are different ways to address stigma . One promising intervention is called ‘contact’ . The principle is that personal contact between persons affected by a stigmatized condition and the public will demystify incorrect information , break the stereotypes and generate empathy . This in turn is believed to reduce stigma . In this paper the authors report a study that investigated the effect of a contact intervention that aimed to reduce leprosy-related public stigma in Cirebon District , Indonesia . During this intervention 91 so called ‘contact events’ were organized at a local level , for instance , in schools , villages , halls and mosques . The results show that through education , testimonies ( direct contact ) , videos and comics made by people affected by leprosy ( in-direct contact ) knowledge about leprosy increased and personal attitudes improved substantially . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | A Cluster-Randomized Controlled Intervention Study to Assess the Effect of a Contact Intervention in Reducing Leprosy-Related Stigma in Indonesia |
Polyamines are essential for cell growth of eukaryotes including the etiologic agent of human African trypanosomiasis ( HAT ) , Trypanosoma brucei . In trypanosomatids , a key enzyme in the polyamine biosynthetic pathway , S-adenosylmethionine decarboxylase ( TbAdoMetDC ) heterodimerizes with a unique catalytically-dead paralog called prozyme to form the active enzyme complex . In higher eukaryotes , polyamine metabolism is subject to tight feedback regulation by spermidine-dependent mechanisms that are absent in trypanosomatids . Instead , in T . brucei an alternative regulatory strategy based on TbAdoMetDC prozyme has evolved . We previously demonstrated that prozyme protein levels increase in response to loss of TbAdoMetDC activity . Herein , we show that prozyme levels are under translational control by monitoring incorporation of deuterated leucine into nascent prozyme protein . We furthermore identify pathway factors that regulate prozyme mRNA translation . We find evidence for a regulatory feedback mechanism in which TbAdoMetDC protein and decarboxylated AdoMet ( dcAdoMet ) act as suppressors of prozyme translation . In TbAdoMetDC null cells expressing the human AdoMetDC enzyme , prozyme levels are constitutively upregulated . Wild-type prozyme levels are restored by complementation with either TbAdoMetDC or an active site mutant , suggesting that TbAdoMetDC possesses an enzyme activity-independent function that inhibits prozyme translation . Depletion of dcAdoMet pools by three independent strategies: inhibition/knockdown of TbAdoMetDC , knockdown of AdoMet synthase , or methionine starvation , each cause prozyme upregulation , providing independent evidence that dcAdoMet functions as a metabolic signal for regulation of the polyamine pathway in T . brucei . These findings highlight a potential regulatory paradigm employing enzymes and pseudoenzymes that may have broad implications in biology .
The single-celled eukaryotic parasite Trypanosoma brucei is the causative agent of human African trypanosomiasis ( HAT ) , also known as sleeping sickness , and of nagana in cattle . According to the World Health Organization , approximately 65 million people in sub-Saharan Africa are at risk for HAT [1] . Both human infective T . brucei species ( Trypanosoma brucei gambiense and Trypanosoma brucei rhodesiense ) cause a typically fatal disease [2 , 3] , though the identification of asymptomatic individuals and of parasite reservoirs in the skin suggests individual outcomes are more complicated than previously understood [4 , 5] . While vector control and current therapies have contributed to reduced parasite burden over the past 20 years ( current cases are <5000 per year ) , the available drugs are species- and stage-dependent , toxic , and/or difficult to administer [1] . Eflornithine , which is used in combination for the treatment of late stage T . b . gambiense [6 , 7] , is an irreversible inhibitor of ornithine decarboxylase ( ODC ) , identifying the polyamine biosynthetic pathway ( Fig 1 ) as a validated pathway for the treatment of HAT [8] . In this same pathway , S-adenosylmethionine decarboxylase ( TbAdoMetDC ) , which generates the decarboxylated AdoMet ( dcAdoMet ) necessary for biosynthesis of the polyamine spermidine , was shown to be essential in T . brucei by genetic studies [9] . Inhibitors of TbAdoMetDC with in vivo anti-trypanosomal activity have also been described [10–14] . Polyamines play important cellular roles in transcription and translation [15–17] . Spermidine is essential in all eukaryotes as a substrate for the hypusine modification of translational elongation factor eIF5a , which has a global role in translational elongation and termination [18 , 19] . Furthermore , in the trypanosomatids , spermidine plays a specialized role and is conjugated to glutathione to form trypanothione , an essential redox cofactor [7] . In higher organisms , polyamine biosynthesis is tightly regulated and spermidine has been shown to feedback regulate AdoMetDC and ODC at the levels of transcription , translation , and protein stability [15–17] . In mammals and plants , AdoMetDC translation is controlled by an mRNA upstream open reading frame ( uORF ) that leads to ribosome stalling when spermidine levels are high , and ODC levels are controlled by protein turnover mediated by an inhibitory binding protein termed antizyme [20] . These regulatory mechanisms are absent in T . brucei [8] . In trypanosomatids , genes are transcribed constitutively in polycistronic units and undergo trans-splicing reactions simultaneously with polyadenylation to generate mature , monocistronic mRNAs [21] . Transcriptional regulation is generally lacking and gene expression is controlled post-transcriptionally by mRNA stability , translational regulation , and protein stability . We previously reported that T . brucei AdoMetDC is regulated by a novel allosteric mechanism . In mammals , AdoMetDC is active as a homodimer [22] , whereas , in the trypanosomatids , we demonstrated that TbAdoMetDC is functional only as a heterodimer formed between a catalytically impaired AdoMetDC subunit and inactive paralog ( pseudoenzyme ) , we termed prozyme [23–25] . Both TbAdoMetDC and prozyme are essential for enzyme activity and T . brucei cell viability [9] . Monomeric TbAdoMetDC is inactive due to autoinhibition by its N-terminus [24] . Upon heterodimerization with prozyme , the N-terminal α-helix repositions to the heterodimer interface , relieving the autoinhibition and generating the active enzyme . Furthermore , prozyme also appears to be involved in regulating the polyamine biosynthetic pathway in T . brucei [9 , 26] . Either knockdown or chemical inhibition of TbAdoMetDC led to an increase in prozyme proteins levels suggesting T . brucei regulates prozyme to compensate for reduced TbAdoMetDC activity [9 , 14 , 26 , 27] . However , the mechanistic basis for how T . brucei regulates prozyme expression has not been fully elucidated . Our previous studies suggested that prozyme expression was post-transcriptionally controlled most likely at the level of translation . Levels of prozyme mRNA were not changed in response to TbAdoMetDC knockdown or inhibition , but we identified alternatively spliced variants of prozyme mRNA showing that the longest mRNA contained a putative secondary structure suggestive of a potential regulatory role in translation [9 , 26] . In mammalian cells , spermidine is a key metabolic signal that regulates expression and activity of the polyamine pathway biosynthetic enzymes [15 , 17] . However , in T . brucei , knockdown or inhibition of other pathway enzymes ( e . g . TbODC or spermidine synthase ) did not affect prozyme protein levels despite causing cellular concentrations of spermidine and/or putrescine to decrease [28] . Thus , neither putrescine nor spermidine are likely to be involved in regulating prozyme expression . Instead , we found correlative evidence that dcAdoMet might be a regulatory metabolite . Herein , we expand on these findings by demonstrating that prozyme protein levels are controlled translationally , that the TbAdoMetDC protein itself acts as a suppressor of prozyme expression in an enzyme activity-independent manner , and we provide additional evidence that dcAdoMet , the product of AdoMetDC , acts as the key signal in a feed-back regulatory mechanism .
As noted above prozyme protein levels increase in response to inhibition of TbAdoMetDC with the mechanism-based irreversible inhibitors MDL-73811 or Genz-644131 [9 , 26 , 27] . In the presence of cycloheximide ( CHX ) , prozyme upregulation was abolished , and endogenous prozyme protein levels were stable for > 6 h , suggesting a translational mechanism [9] . To extend these findings we monitored prozyme translation directly by labeling nascent prozyme protein with deuterated leucine ( 2H7-leucine ) . The abundance of both labeled ( heavy 2H-leucine ) and unlabeled ( light 1H-leucine ) prozyme peptides was then simultaneously determined by mass spectrometry using selected reaction monitoring ( SRM ) . Leucine was chosen as the labeling reagent because preliminary studies showed prozyme protein levels were not affected by changes in leucine concentration ( 10–25 μM ) in the media ( S1A Fig ) , whereas they were impacted by changes in methionine ( discussed below ) . HMI-19 cell culture media contains >800 μM leucine but we established that 10 μM leucine was sufficient to support cell growth ( S1B Fig ) while maintaining the prozyme regulatory response in cells treated with Genz-644131 ( S1A Fig ) ( leucine concentrations in human blood and cerebral spinal fluid are reported to be 150 and 14 μM , respectively [29] ) . Leucine concentrations below 10 μM led to reduced cell growth and to reduced ability of cells to upregulate prozyme expression , likely due to the effects of starvation on overall protein synthesis . To prevent the complication of detecting peptides with combinations of heavy and light isotopic leucine , we monitored initially two peptides ( SAFPTGHPYLAGPVDR ( residues 157–172 ) and LEGFTVVHR ( residues 297–305 ) ) both of which contained only a single leucine , leading to a 7 Da shift in molecular mass per incorporated 2H7-leucine . Peptide LEGFTVVHR showed a lower limit of detection and was used to monitor 2H7-leucine incorporation in all subsequent studies . Bloodstream form ( BSF ) T . brucei Lister 427 cells were cultured in the presence of 1H-leucine ( light ) , washed in PBS , and then transferred to leucine-free medium supplemented with 10 μM 2H7-leucine ( containing dialyzed serum so that 1H-leucine was not introduced from the serum ) . Simultaneously vehicle control ( water ) , Genz-644131 , or both Genz-644131 and CHX were also added . Prozyme expression levels were then monitored over 12 h by both Western blot and SRM . Western blot analysis indicated total prozyme protein levels increased in a time-dependent manner for cells treated with Genz-644131 ( Fig 2A and S1C Fig ) , recapitulating our previous results . Prozyme upregulation was abolished , as expected , in cells simultaneously treated with Genz-644131 and CHX . SRM analysis of the unlabeled prozyme peptide ( pre-existing 1H-leucine prozyme ) showed prozyme concentrations were stable over the 12 h time course when treated with Genz-644141 , while we observed some turnover of the protein in the absence of Genz-644141 over the 3–12 h time period ( Fig 2B and S1 Table ) . In contrast , we observed a time-dependent increase in 2H7-leucine-labeled prozyme peptides by SRM analysis in samples treated with Genz-644131 , and the rate of this increase was significantly greater than observed for the untreated ( minus Genz-644131 ) control ( Fig 2C and S1 Table ) . Addition of CHX prevented incorporation of 2H7-leucine into the prozyme peptide , confirming that 2H7-leucine incorporation was dependent on translation . An increased rate of incorporation was observed in the untreated control at the first time-point ( 3 h ) that can be attributed to a feeding effect as cells were transferred into rich media after the wash step ( see below effects of methionine starvation ) . The rate of incorporation into untreated controls returned to low levels by 6 h , while cells treated with Genz-644131 continued to incorporate 2H7-leucine at an increased rate throughout the 12 h study . These data show that prozyme translation rates increase when TbAdoMetDC is inhibited with Genz-644131 , consistent with a translational regulatory mechanism . They also suggest that stabilization of prozyme from degradation occurs upon treatment with Genz-644131 and may also contribute to the increased pools of prozyme under this condition . In our previous work , we sought to determine whether TbAdoMetDC might function to regulate prozyme expression levels by either directly binding to the prozyme mRNA or by interacting with other RNA binding proteins [26] . In that study , we generated a TbAdoMetDC RNAi cell line that expressed human AdoMetDC ( HsAdoMetDC ) under the control of a tetracycline ( Tet ) promoter , such that the addition of Tet led to the simultaneous expression of HsAdoMetDC and knockdown of TbAdoMetDC . Expression of HsAdoMetDC led to restoration of WT prozyme levels upon knockdown of TbAdoMetDC . However , based on western blot analysis , we estimated that about 20% of TbAdoMetDC protein remained after RNAi knockdown , and thus this study did not rule out the possibility that TbAdoMetDC protein itself was a negative regulator of prozyme expression . To further address this question , herein we generated a TbAdoMetDC null cell line in the presence of conditionally expressed HsAdoMetDC under control of the Tet promoter ( TbAdoMetDC null+Hs ) ( Fig 3A ) . T . brucei contains two identical amd genes ( encoding AdoMetDC ) in the genome ( Tb427 . 06 . 4410 and Tb427 . 06 . 4460 in TriTrypDB ) [30] and thus as a diploid organism , contains four copies of the gene . To generate the amd null cell line , we used the Cre-loxP system [31] to remove the selectable markers after knockout of the first two alleles so that the markers could be reused in the subsequent knockout of the final gene copies ( Experimental Procedures ) . The Hsamd gene was inserted into the rRNA locus to complement the loss of TbAdoMetDC prior to removal of the final two Tbamd gene copies . In the absence of Tet , TbAdoMetDC null+Hs cells undergo a severe growth defect rescued by expression of HsAdoMetDC ( Fig 3B and S2A and S2B Fig ) . As in our previous studies , addition of the AdoMetDC inhibitor Genz-644131 to wild-type ( WT ) cells led to induction of prozyme levels detected by western blot analysis 24 h after addition of compound ( Fig 3C and 3D and S2C and S2D Fig ) [27] . In contrast , in the TbAdoMetDC null+Hs cell line , we observed constitutively high levels of prozyme in the presence or absence of Genz-644131 ( Fig 3C and 3D and S2D Fig ) . To confirm that the effects on prozyme expression were caused by changes in protein levels and not in mRNA levels , prozyme mRNA levels were evaluated by quantitative reverse transcription PCR ( RT-qPCR ) ( Fig 3E and 3F ) , demonstrating that prozyme mRNA levels remained constant in both WT and null+Hs cell lines with and without Genz-644131 . Of note , Genz-644131 is an equally effective inhibitor of both TbAdoMetDC and HsAdoMetDC [11] , thus demonstrating that loss of AdoMetDC activity in cells expressing either TbAdoMetDC and HsAdoMetDC is not sufficient on its own to lead to prozyme expression changes . Our data support a mechanism whereby TbAdoMetDC is a negative regulator of prozyme translation . In its absence prozyme is constitutively expressed at higher levels , and prozyme expression is no longer sensitive to inhibition of AdoMetDC activity . To confirm that the TbAdoMetDC null+Hs line remained capable of regulated prozyme expression we rescued the cell line by transfecting it with a constitutively expressed copy of WT TbAdoMetDC ( AdoMetDC null+Hs+Tb ) ( Fig 3A and S2A Fig ) . To differentiate between potential regulatory elements in Tbamd genetic sequence and amino acid sequence , the DNA sequence of this complement construct was from a construct codon-optimized for E . coli expression; thus the mRNA was altered while maintaining the amino acid sequence . Cells harboring the WT Tbamd complement construct expressed TbAdoMetDC constitutively , leading to restoration of WT growth even in the absence of Tet when human AdoMetDC is no longer expressed ( Fig 3B and S2A Fig ) . WT prozyme protein expression levels were restored in TbAdoMetDC null+Hs+Tb cells . Moreover , prozyme could be again upregulated with Genz-644131 treatment ( Fig 3C and 3D and S2D Fig ) . Taken together , these results suggest that the TbAdoMetDC protein , and not TbAdoMetDC gene or mRNA was responsible for the regulatory effect on prozyme protein levels . We next sought to determine whether the TbAdoMetDC protein or its enzymatic function was necessary for suppression of prozyme expression . Using a parallel approach to above we transfected TbAdoMetDC null+Hs cells with a catalytically-dead copy of TbAdoMetDC containing a mutation of the catalytic cysteine ( C100 ) to alanine ( AdoMetDC null+Hs+TbC100A ) . The C100A mutant of both human and T . cruzi AdoMetDC were previously shown to have >100-fold reduced activity over the WT enzyme [32 , 33] . Consistent with the lack of activity , TbAdoMetDC-C100A was unable to rescue growth in the absence of HsAdoMetDC expression ( –Tet ) ( Fig 3A and 3B and S2A Fig ) . However , WT prozyme protein levels were restored in this cell line , and this line was capable of Genz-644131-dependent prozyme upregulation , similarly to the TbAdoMetDC null+Hs+Tb cell line ( Fig 3C and 3D and S2D Fig ) , suggesting that TbAdoMetDC enzymatic function was dispensable for the regulatory effect . Again , prozyme mRNA levels did not significantly vary among lines or stimulatory conditions and the prozyme protein:mRNA ratios mirrored the changes in protein levels ( Fig 3E and 3F ) . Altogether , we conclude TbAdoMetDC protein suppresses translation of prozyme by an enzyme activity-independent mechanism . In mammalian cells , spermidine plays an important role as a negative feedback regulator affecting translation of AdoMetDC mRNA and turnover of ODC [8] . However , in T . brucei , prozyme protein levels are unaffected by changes in spermidine concentration [28] . Instead , previous studies suggested that dcAdoMet concentration correlated inversely to prozyme expression levels [26] . To provide further support for the hypothesis that dcAdoMet concentration is involved in controlling prozyme protein levels , we analyzed the effects of three independent mechanisms to reduce cellular dcAdoMet levels . These included our previously described use of AdoMetDC inhibitors , plus two new approaches , knockdown of AdoMet synthetase ( AdoMetSyn ) and methionine starvation . We quantitated the relative levels of AdoMet and dcAdoMet in parasites before and after treatment with the AdoMetDC inhibitor Genz-644131 for 6 h . An early time point was chosen so that the results would be independent of effects on cell growth that occur upon more extended incubation . Under conditions where prozyme was upregulated ( +Genz ) ( Fig 4A and 4B and S3 Fig ) dcAdoMet pools were depleted by 82% ( p value <0 . 0005 ) ( Fig 4C ) , while AdoMet pools were slightly elevated though this latter change was not statistically significant ( Fig 4C ) . Using a broader targeted metabolomics analysis ( 112 soluble metabolites ) of these same cell extracts we did not identify any other metabolite that significantly changed or correlated with prozyme upregulation ( Fig 4D , S2 Table and S4 Fig ) . AdoMet showed no significant change in this data set ( dcAdoMet was not measured ) . We previously showed that AdoMet levels are ~200-fold higher than dcAdoMet levels in T . brucei , [28] thus a loss of flux into dcAdoMet would not be expected to impact the AdoMet pools . Polyamine levels were unchanged with the exception of a modest ( 3 . 1 ± 1 . 6 ) -fold increase in N-acetylputrescine in the presence of Genz that was not statistically significant . In previous studies we did observe an increase in putrescine and a decrease in spermidine after a longer time of incubation with MDL 72811 ( 72 h ) , but through co-treatment with eflornithine we were able to show that the elevated putrescine levels were not linked to prozyme expression [26] . These current data suggest that N-acetylputrescine may be formed to buffer against the accumulation of putrescine . N-acetylputrescine has been observed in other published metabolomic studies in T . brucei [34 , 35] and its levels were shown to be affected similarly to putrescine after treatment with eflornithine [35] . The enzyme responsible for its formation and its role in parasite biology are unknown . As a second independent approach to manipulate the cellular dcAdoMet levels we assessed the influence of upstream pathway enzymes and metabolites on prozyme expression . Both AdoMet and dcAdoMet pools are controlled by their biosynthesis . The trypanosomatids including T . brucei encode a putative AdoMet synthetase ( TbAdoMetSyn ) , which utilizes ATP and methionine to catalyze formation of AdoMet . To validate its function , we expressed and purified recombinant TbAdoMetSyn using affinity His6-tag and characterized its activity at varying ATP and methionine concentrations . These studies demonstrated that the adometsyn gene indeed encodes an active AdoMetSyn with kinetic parameters similar to those reported for the Leishmania infantum enzyme [36] . ( Table 1 and S5A Fig ) . We next evaluated the effects of TbAdoMetSyn knockdown on regulation of the polyamine pathway via prozyme protein levels . T . brucei has 18 identical copies of TbAdoMetSyn arising from a 9-copy tandem array ( Tb427 . 6 . 4840-Tb427 . 6 . 4920 ) and the diploid genome . Targeting the full array to generate a conditional knockout line would be technically challenging . Instead we generated an RNAi line to study the effects of TbAdoMetSyn knockdown on cell growth , AdoMet and dcAdoMet pools , and prozyme expression . This TbAdoMetSyn RNAi cell line was engineered by inserting a TbAdoMetSyn hairpin sequence ( nt 602–1039 ) under control of the Tet promoter into the ribosomal gene cluster of BSF cells . Addition of Tet to the TbAdoMetSyn RNAi cell line led to a significant growth effect starting at 48 h ( Fig 5A ) . This growth arrest corresponded to an 80% decrease in TbAdoMetSyn protein and mRNA levels at 48 h ( Fig 5B–5D and S5B Fig ) as evaluated by western blot and RT-qPCR respectively . Concomitant with TbAdoMetSyn knockdown , prozyme protein levels were significantly upregulated 48 h after the addition of Tet ( Fig 5B and 5C and S5B Fig ) . To demonstrate that the observed effects were caused by TbAdoMetSyn knockdown we transformed the TbAdoMetSyn RNAi line with a RNAi-resistant T . brucei TbAdoMetSyn ( S6 Fig ) expression construct to provide genetic complementation of the knockdown ( TbAdoMetSyn RNAi+Comp ) ( Fig 5A–5D ) . In this line , TbAdoMetSyn protein levels were similar to WT levels , and WT growth rates were restored ( Fig 5A–5C ) . While we could not distinguish between endogenous TbAdoMetSyn and enzyme expressed from the scrambled complement construct by western blot analysis , RT-qPCR analysis showed that endogenous mRNA was similarly reduced in both the TbAdoMetSyn RNAi and TbAdoMetSyn RNAi+Comp lines ( Fig 5D ) . Prozyme protein levels were also restored to WT levels by genetic complementation of the RNAi line ( Fig 5B and 5C ) . Prozyme mRNA levels do not change significantly under TbAdoMetSyn RNAi or complementation conditions , thus the upregulation of prozyme protein levels upon TbAdoMetSyn knockdown occurs post-transcriptionally ( Fig 5D and 5E ) , similarly to our previous observations upon AdoMetDC knockdown or inhibition . Finally , we analyzed the effects of TbAdoMetSyn knockdown on AdoMet and dcAdoMet intracellular pools using LC-MS/MS . Knockdown of TbAdoMetSyn led to an 80–90% depletion of both AdoMet and dcAdoMet pools 48 h after the addition of Tet ( Fig 5F ) . Complementation of the TbAdoMetSyn RNAi by the scrambled TbAdoMetSyn rescued construct restored levels of both AdoMet and dcAdoMet to WT levels . These data provide the first evidence of prozyme upregulation without direct manipulation of AdoMetDC . Furthermore , dcAdoMet levels again correlate inversely with prozyme levels . While AdoMet pools were also reduced after TbAdoMetSyn knockdown , they were not affected by AdoMetDC knockdown or inhibition , and thus are unlikely to play a role in prozyme regulation . Since AdoMet is synthesized from methionine and ATP as third independent approach to reduce dcAdoMet levels we used methionine starvation to manipulate the pathway without directly perturbing the enzyme activity levels . We examined the effects of methionine starvation on cell growth , prozyme expression levels , and AdoMet and dcAdoMet levels . To determine the concentration range that would be appropriate for the methionine starvation study we first measured methionine concentration in FBS ( S3 Table ) by LC-MS/MS ( methionine = 30 ± 1 . 9 μM , which is similar to that reported for human serum [29] ) . Thus , in T . brucei medium ( HMI-19 ) supplemented with 10% FBS , the minimum methionine concentration will be 3 μM , whereas the concentration in standard HMI-19 medium is 200 μM . To determine the minimum methionine levels necessary for BSF 427 T . brucei cell viability , we performed a methionine dose response study ( 3–200 μM ) where cells were grown for 48 h to determine the minimum methionine levels necessary T . brucei cell viability . Cells exhibited an increasingly severe growth defect as methionine levels fell below 30 μM ( Fig 6A ) . Relative half-maximal growth rate was at 9 . 0 ( 8 . 2–9 . 8 ) μM ( 95% confidence interval in parenthesis ) . Finally , we also performed long-term growth rate analysis of cells grown at select methionine levels , and found that although cells grown at 3 μM methionine grow slower , they remain viable ( S7A Fig ) . The effects of varying methionine concentration on prozyme expression were then assessed to provide orthogonal support for the role of dcAdoMet in prozyme regulation . We observed a methionine dose-dependent upregulation of prozyme for methionine medium concentrations below 4 μM . Prozyme protein levels could be further increased by the addition of Genz-644131 for 24 h at all levels of methionine ( Fig 6B , 6C and S7B Fig ) . RT-qPCR analysis indicated that prozyme mRNA does not change significantly in any of these conditions , demonstrating prozyme expression is regulated post-transcriptionally ( Fig 6D ) . The effects of methionine depletion on prozyme expression do not result simply from nutrient starvation , as depletion of leucine from the growth media did not impact prozyme levels ( S1 Fig ) . Interestingly , we also observed some upregulation of AdoMetDC at low methionine concentrations ( Fig 6B ) . AdoMet and dcAdoMet measurements were made by LC-MS/MS analysis , which revealed that both AdoMet and dcAdoMet pools decreased as methionine was reduced; the concentration of dcAdoMet in cells grown at 3 μM methionine was 95-fold lower than for cells grown in 200 μM ( Fig 6F ) . dcAdoMet levels were further decreased after treatment ( 24 h ) with Genz-644131 ( Fig 6F ) . Thus we have shown by three independent genetic or chemical methods that depletion of dcAdoMet pools correlates with an upregulation of prozyme , providing further evidence for causal link between prozyme levels and dcAdoMet concentration .
Polyamine biosynthesis is tightly regulated in many eukaryotes , however the mechanism by which this regulation is achieved is very different in T . brucei [8] . In contrast to mammalian cells , in T . brucei , the polyamines spermidine and putrescine do not play significant roles in regulating polyamine biosynthesis in general , or in regulating TbAdoMetDC activity or prozyme expression , specifically [28] . Instead , prozyme regulates TbAdoMetDC at the enzyme level while at the cellular level prozyme protein levels are responsive to perturbations that effect pathway flux ( e . g . TbAdoMetDC RNAi or chemical inhibition ) [9 , 25] . Herein , we have shown that the increase in prozyme protein in the presence of Genz-644131 occurs at the level of translation by directly measuring the rate of prozyme synthesis with stable isotopes . We then expanded on our mechanistic understanding of this regulation by using a TbAdoMetDC null cell line to show that TbAdoMetDC is a suppressor of prozyme translation in an enzyme activity-independent manner . This is the first demonstration of a non-enzymatic regulatory function for AdoMetDC . We also showed strong correlative evidence using three independent methods that low levels of dcAdoMet trigger a relief of this suppression leading to increased prozyme protein levels , thus associating dcAdoMet with a regulatory function . Together , these data suggest that a two-component regulatory system controls prozyme expression; TbAdoMetDC serves as a negative regulator of translation while the cell also senses dcAdoMet levels , such that translational repression is relieved when dcAdoMet levels are low ( Fig 7 ) . Enzymatic activity is not required for the TbAdoMetDC regulatory roles . It is mechanistically unclear how TbAdoMetDC functions as a suppressor of prozyme expression . Because transcription initiation control is absent in kinetoplastids at the level of individual genes , most regulation occurs post-transcriptionally [21] . Several ribosomal profiling studies suggest there is extensive translational regulation [37–39] . In mammalian cells , dihydrofolate reductase ( DHFR ) and thymidylate synthase ( TS ) have been shown to autoregulate translation by binding their own mRNA [40–42] . The same mechanism of regulation was shown for Plasmodium falciparum DHFR [43] . Analogously , TbAdoMetDC may interact directly with prozyme mRNA to control its translation . Alternatively , this interaction may be mediated by a RNA-binding protein ( RBP ) . In T . brucei , there are over a hundred predicted RBPs , most of which are uncharacterized [44] . In either case , we hypothesize that TbAdoMetDC would form a complex with prozyme mRNA and prevent its translation . The work described herein also provides the first evidence of prozyme regulation independent of changes to TbAdoMetDC and supports a role for the reaction product dcAdoMet in this regulation . Previous studies depended on genetic knockdown or irreversible chemical inhibition of TbAdoMetDC by Genz-644131 . Through knockdown of TbAdoMetSyn or methionine starvation , we were able to deplete downstream metabolite pools and upregulate prozyme independently of manipulation of TbAdoMetDC . dcAdoMet levels were substantially decreased after all three perturbations . A broader metabolite analysis was also undertaken at an early time point ( 6 h ) after initiation of Genz-644131 treatment so that the effects on the metabolome could be separated from cell growth changes that occur after longer incubations . The only significantly altered pathway metabolite was dcAdoMet . These data thus support the hypothesis that dcAdoMet acts as a metabolic signal and its depletion triggers increased prozyme protein expression . AdoMet is the methyl donor for most cellular methylation reactions , including DNA , RNA , and proteins . The ratio of AdoMet to S-adenosylhomocysteine has been used as an index for the methylation ability of the cell [45 , 46] . We have shown that inhibition of TbAdoMetDC by Genz-644131 did not significantly affect levels of AdoMet , thus alteration of these ratios is unlikely to control prozyme expression . Furthermore , our studies have shown that we can target either downstream or upstream pathways affecting AdoMet to initiate increased prozyme translation . Based on these findings , we conclude that the control of prozyme expression is methylation independent . We hypothesize that depletion of dcAdoMet pools alleviates suppression of prozyme translation by TbAdoMetDC , but how dcAdoMet pools are sensed remains an open question . We have previously hypothesized that the putative secondary structure located in the 3’UTR of prozyme mRNA may contain a riboswitch-like function by binding dcAdoMet [26] . In bacteria , AdoMet and S-adenosylhomocysteine-binding riboswitches have been characterized that regulate translation of methionine and cysteine metabolism [47–51] . While these mechanisms acted through the 5’UTR , in T . brucei , 3’UTRs of only a few hundred nucleotides can play major roles in regulating mRNA translation and decay [21] . In one potential model , dcAdoMet could be bound by a riboswitch in the prozyme 3’UTR , and this secondary structure can then be bound by TbAdoMetDC or an associated RBP mediator to inhibit translation of prozyme mRNA . Alternatively , dcAdoMet may be bound directly by TbAdoMetDC , which then serves as the sensor to promote binding of itself or another binding partner to prozyme mRNA . The C100A-TbAdoMetDC mutation employed in our studies reduces activity but not AdoMet binding , so there remains the possibility that the active site is involved in sensing dcAdoMet levels . Whether there is a difference in translation rates between the larger and shorter prozyme ORF-containing mRNA transcripts also remains to be determined . An intriguing possibility highlighted by our studies is that the paradigm of pseudoenzymes as metabolic regulators will be found in other trypanosomatid pathways , with evolution of these mechanisms perhaps driven by their reliance on post-transcriptional control mechanisms . In recent years , two other enzyme-prozyme complexes have been identified in T . brucei; deoxyhypusine synthase ( DHS ) and protein arginine methyltransferase ( PRMT1 ) , both of which also require formation of a complex between enzymatically impaired and inactive paralogs ( pseudoenzymes ) to generate the active enzyme [52 , 53] . However , while these pseudoenzyme/enzyme complexes are required for enzyme activity , a regulatory role for these pseudoenzymes in controlling metabolism similar to the prozyme regulatory mechanism seems unlikely as both DHS [52] and PRMT1 [53] exhibit dependent expression such that knockdown of one subunit led to loss of both subunits . This does not preclude other activities by these pseudoenzymes to regulate metabolism in their respective pathways . A growing literature describes diverse roles for paralogous pseudoenzymes functioning as regulators of their respective enzymes in metazoan genomes [54–56] . Given that T . brucei relies heavily on post-transcriptional mechanisms for gene regulation , the use of pseudoenzymes in regulatory roles may be enriched relative to other organisms .
T . brucei genomic sequences were obtained from TriTrypDB and gene accession numbers are as follows: amd ( encodes AdoMetDC ) Tb927 . 6 . 4460/Tb927 . 6 . 4410; prozyme Tb927 . 6 . 4470; TbAdoMetSyn Tb927 . 6 . 4840-Tb927 . 6 . 4920 ( 9-copy array ) ; tert Tb927 . 11 . 10190 . The accession number for human amd is NM_001634 . 5 . Experiments were performed using either Trypanosoma brucei bloodstream-form ( BSF ) 427 or single marker ( SM ) cells that constitutively express the Tet repressor and T7 RNA polymerase ( maintained in the presence of Geneticin ( G418 ) ) [57] . T . brucei cells were cultured in HMI-19 medium [58] with 10% fetal bovine serum ( FBS ) ( Tet-free , heat-inactivated; Gemini Bio-Products ) or dialyzed FBS ( Tet-free , heat-inactivated; Gemini Bio-Products ) at 37°C and 5% CO2 [9 , 59] . Parasite transfections were performed as previously described [60] . Antibiotics were used at the following concentrations: G418 ( 2 . 5 μg/mL; Life Technologies ) , phleomycin ( 2 . 5 μg/mL; InvivoGen ) , hygromycin ( 1–2 . 5 μg/mL; Gemini Bio-Products ) , puromycin ( 1 μg/mL; Sigma ) , tetracycline ( Tet ) ( 1 μg/mL; Sigma ) , and ganciclovir ( GCV ) ( InvivoGen ) was used at 40 μg/mL . Genz-644131 ( a generous gift from Genzyme , presently Sanofi ) was used at 15 nM ( 10X EC50 ) [27] . Cycloheximide ( Sigma ) was used at 50 μg/mL [9] . For 2H7-leucine labeling conditions , cell lines were plated into leucine-free HMI-19 medium ( prepared with custom-ordered leucine-free IMDM ( Invitrogen ) ) with 10% dialyzed FBS and supplemented with either 10 μM 2H7-leucine for SRM analysis or for growth studies supplemented with sterile-filtered leucine ( Sigma ) dissolved in ddH2O to the desired concentrations . For methionine-limiting conditions , cell lines were cultured in methionine-free HMI-19 medium ( prepared from custom-ordered methionine-free IMDM ( Invitrogen ) ) supplemented with 10% FBS . Sterile-filtered methionine ( Sigma ) dissolved in ddH2O was then used to supplement medium at desired concentrations . Normal HMI-19 medium contains 200 μM methionine and 800 μM leucine [61] , both in about 10-fold excess of concentrations observed in human serum [29] . PCR reactions were performed with Phusion high-fidelity DNA polymerase ( NEB ) . Plasmids propagated using Stellar ( Clontech Laboratories ) or Invitrogen One Shot TOP10 ( Thermo Fisher Scientific ) cells . T . brucei contains two identical amd genes ( Tb427 . 06 . 4410 and Tb427 . 06 . 4460 in TriTrypDB ) that encode AdoMetDC [30] , thus as a diploid organism four amd genes are present in the genome . Due to limiting availability of resistance markers , knockout ( KO ) of the four gene copies was performed in two rounds using selectable marker cassettes flanked by loxP sites . Marker cassettes were removed with Cre-recombinase after each round and subsequently reused in following steps as described [31] . The human amd Tet-regulated complement construct was inserted into the rRNA locus after removal of the first two alleles . Cloning primers are provided in S4 Table . Starting from SM cells , the first two Tbamd loci were replaced with resistance marker cassettes ( hygromycin-resistance gene hyg and puromycin-resistance gene pac ) fused to the Herpes simplex virus thymidine kinase gene ( HSVtk ) , flanked by loxP sites . Resistance markers fused to HSVtk were amplified from pHJ17 ( hyg ) and pHJ18 ( pac ) [31] ( Addgene ) with primers p1/p2 ( hyg and pac ) . 5’ and 3’ flanking regions of Tbamd were amplified from SM genomic DNA with primers p3/p4 ( 5’UTR–1 ) and p5/p6 ( 3’UTR–1 ) . The first pair of KO constructs were generated by fusion PCR of hyg or pac and 5’UTR–1 and 3’UTR–1 flanking amplicons with primers p7/p8 as described [62] and the resulting PCR fragments cloned into pCR-Blunt II-TOPO vector using Zero Blunt TOPO PCR cloning kit ( Thermo Fisher Scientific ) ( KO1-hyg-TOPO and KO1-pac-TOPO ) . Knockouts were performed by concurrent transfection of SM cells with NsiI-excised KO1-hyg-TOPO and KO1-pac-TOPO under hygromycin and puromycin selection ( TbAdoMetDC KO1-hyg/pac ) . To recycle the selection markers , a TbAdoMetDC KO1-hyg/Pac line was transiently transfected with pLew100cre-del-tetO ( Addgene ) derived from the construct pLEW100cre by deleting the Tet operator [63] to express Cre recombinase . Transfectants were subjected to negative selection with ganciclovir ( 40 μg/mL ) in the absence of hygromycin and puromycin to select for lines with HSVtk excised by Cre recombinase ( KO1 ) . The resulting TbAdoMetDC KO1 cell line lacks two of four amd alleles . Hsamd was cloned from the previously described pET28b-derived plasmid [64] with primers p9/p10 into pLew100v5 [57] under control of a Tet-regulatable promoter ( p100-HsAdoMetDC ) . The sequence was confirmed with primers p33/p34 . TbAdoMetDC KO1 cells were transfected with NotI–linearized p100-HsAdoMetDC under selection with phleomycin ( TbAdoMetDC null1+Hs ) . The second set of 5’ and 3’ flanking regions of Tb amd ( internal to the first KO ) were amplified from SM genomic DNA with primers p11/p12 ( 5’UTR–2 ) and p13/p14 ( 3’UTR–2 ) and the KO constructs were generated by fusion PCR with hyg or pac resistance markers and 5’UTR–2 and 3’UTR–2 flanking amplicons with primers p15/p16 ( KO2-hyg and KO2-p ) as described above . The Tbamd null line was then generated by sequential transfection of a TbAdoMetDC null1+Hs with NsiI-excised KO2-hyg-TOPO under hygromycin selection ( TbAdoMetDC null2+Hs-hyg ) and then NsiI-excised KO2-pac-TOPO under hygromycin and puromycin selection in the presence of Tet ( TbAdoMetDC null2+Hs-hyg/pac ) and other maintenance antibiotics ( G418 and phleomycin ) . The hyg and pac resistance genes were then removed under negative selection with GCV in the presence of Tet yielding the final Tbamd null cell line that expressed human AdoMetDC under the control of the Tet promoter ( TbAdoMetDC null+Hs ) . The absence of Tbamd was verified by RT-qPCR ( S2B Fig ) with primers p45/p46 relative to α–Tubulin with primers p47/p48 . TbAdoMetDC null+Hs was complemented with WT or catalytically-dead Tbamd . The DNA sequence that was used had been codon-optimized for E . coli expression thus this allowed us to introduce a different mRNA sequence while maintaining the amino acid sequence . The catalytically-dead TbAdoMetDC mutant was generated by site-directed mutagenesis of our previously described E . coli TbAdoMetDC expression construct [24] subcloned into the pCR-Blunt II-TOPO vector . Primers p21/22 were used to convert the catalytic C100A with PfuTurbo DNA polymerase ( Agilent Technologies ) ( TOPO-AdoMetDCscrm-C100A ) . The reaction was digested with DpnI ( NEB ) and transfected into TOP10 cells . WT TbAdoMetDCscrm and catalytically-dead TbAdoMetDCscrm-C100A were amplified with primers p19/20 using pET28bSmt3-TbAdoMetDC [24] and TOPO-AdoMetDCscrm-C100A plasmids , respectively , as templates . PCR products were cloned into the HindIII/BamHI sites of pLew90 for constitutive expression in T . brucei [57] . Sequences were confirmed with primers p31/p32 . TbAdoMetDC null+Hs was transfected with NotI-linearized p90-TbAdoMetDCscrm or p90-TbAdoMetDCscrm-C100A under selection with hygromycin ( TbAdoMetDC null+Hs+Tb and TbAdoMetDC null+Hs+TbC100A ) . Incorporation of the Tbamd genes were validated by western blot analysis ( S2A Fig ) . Cell growth analyses were performed as previously described using the CellTiter-Glo reagent ( Promega ) [27] . Determination of the 50% growth inhibitory concentration ( EC50 ) of Genz-644131 in TbAdoMetDC null lines was made after 24 h of incubation with drug from a starting inoculum of 1 × 105 cells/mL in HMI-19 with 10% FBS using serial dilutions of Genz-644131 at 0 . 1% ( v/v ) DMSO . The leucine and methionine concentration required for 50% maximal growth ( EC50 ) was measured in BSF 427 cells after 24 h and 48 h , respectively , from a starting inoculum of 1 × 105 and 3 × 103 cells/mL in HMI-19 prepared as described above . Cloning primers are described in Table S4 . RNAi target sequences were chosen based on RNAit [65] and primers ( Sigma ) were designed to amplify Tbadometsyn nucleotides 602–1039 ( TbAdoMetSyn-RNAi-insert ) . The insert was amplified by Platinum Taq DNA Polymerase ( Invitrogen ) and cloned into the pCR8/GW/TOPO vector . Sequencing with primer M13-21 was used to identify a clone with the ORF integrated in the forward direction ( TOPO-TbAdoMetSyn-RNAi ) and this clone was then inserted into the pTrypRNAiGateway [66] vector by recombination using Gateway LR Clonase II Enzyme mix ( Invitrogen ) , generating a Tet-regulated short-hairpin with the ( pTRG-TbAdoMetSyn ) . The integrity of the insert in the resultant clone was confirmed by sequencing using primers p29/30 . SM cells ( maintained in G418 1 μg/mL ) were transfected with NotI–linearized pTRG-Tbadometsyn under selection with phleomycin ( TbAdoMetSyn RNAi ) . To generate an RNAi-resistant complement gene , the Tbadometsyn ORF ( nucleotides 601–1035 ) was synthesized by GenScript to contain scrambled codons ( different RNA sequence that maintained the correct amino acid sequence ( pUC57-TbAdoMetSyn; full sequence in S6 Fig ) . The Tbadometsyn ORF was amplified from this vector using primers p25/26 and inserted into HindIII/BamHI-digested pLew100v5-hyg , a modified pLew100v5 vector ( gift of George Cross ) that contains the hyg resistance cassette [57] , using the InFusion cloning kit ( Takara ) ( p100H-TbAdoMetSyncomp ) . Sequences were confirmed with primers p33/p34 . TbAdoMetSyn RNAi cells were then transfected with NotI–linearized p100H-TbAdoMetSyncomp under hygromycin selection ( TbAdoMetSyn RNAi + comp ) . To generate an E . coli expression construct for TbAdoMetSyn , pUC57-TbAdoMetSyn was used as a template , the ORF was amplified using primers p27/28 and inserted into a BamHI/XhoI-digested pET28bTEV plasmid ( pET28b ( Novagen ) plasmid with Tobacco Etch Virus ( TEV ) protease site substituted for the thrombin site , described in [67] with InFusion cloning kit ( Takara ) allowing for expression of a His6-tagged TbAdoMetSyn . The sequence was verified with primers p35/36 ( pET28bTEV-TbAdoMetSyn ) . For protein expression , pET28bTEV-TbAdoMetSyn was transformed into Novagen BL21 ( DE3 ) cells under kanamycin selection ( 50 μg/ml ) ( NEB ) . Cells were grown at 37°C for 2 h until OD600 = 0 . 4 and cooled to 16°C . After 0 . 5 h at OD600 = 0 . 6 , His6-TbAdoMetSyn expression was induced with IPTG ( 0 . 2 mM ) for 22 h . Cells were pelleted by centrifugation at 3 , 500 × g for 20 min and resuspended in lysis Buffer A ( 100 mM HEPES , pH 8 . 0 , 300 mM KCl , 5 mM MgSO4 , 5 mM imidazole , 10% glycerol ( v/v ) , 0 . 1% ( v/v ) triton X-100 and supplemented with 1 mM β-mercaptoethanol , 2 mM phenylmethylsulfonyl fluoride ( PMSF ) , 1 μg/mL leupeptin , 2 μg/mL antipain , 10 μg/mL benzamidine , 1 μg/mL pepstain , and 1 μg/mL chymostatin . Cells were lysed by cell disruption using an EmulsiFlex-C5 ( Avestin ) at 5 , 000–10 , 000 psi , and cell debris was pelleted by centrifugation ( 50 , 000 × g for 90 min ) . Soluble protein was purified from lysate by Ni2+-affinity chromatography ( HisTrap FF column , GE Healthcare ) on an ÄKTA purifier system ( GE Healthcare ) with Buffer A and Buffer B ( Buffer A with 500 mM imidazole ) . Contaminants were washed off the column with 8% Buffer B and TbAdoMetSyn was eluted with a linear gradient 8–50% Buffer B . TbAdoMetSyn-containing fractions were pooled and imidazole content reduced by 100-fold through serial concentration ( Amicon Ultra-15 Ultracell 30K centrifugal filters ( Merck Millipore ) ) and dilution with Buffer A . The His6-tag was removed by incubation with 50 μg TurboTEV protease ( BioVision ) for 16 h . Untagged TbAdoMetSyn was purified by passage through a Ni2+-affinity HisTrap FF column and collected in the flow-through . TbAdoMetSyn-containing fractions were pooled and concentrated as above . Protein concentration was determined using Bio-Rad Protein Assay Dye reagent and protein was >95% pure based on SDS-PAGE analysis . Activity was measured using a previously described spectroscopic assay [68] . Pyrophosphate ( PPi ) release by TbAdoMetSyn was measured with a coupled enzyme system in Pyrophophate Reagent ( Sigma ) containing a PPi-dependent fructose-6-phosphate kinase , aldolase , triosephosphate isomerase , glycerophosphate dehydrogenase . The assay was performed in Assay buffer ( 50 mM HEPES , pH 8 . 0 , 100 mM KCl , 5 mM MgSO4 , 2 mM 1 , 4-dithiothreitol ( DTT , Sigma ) , 0 . 05% ( v/v ) Triton X-100 reduced ) in in 96-well half-area UV-Star plates ( Phenix ) with 50 μL Assay buffer , 35 μL Pyrophosphate Reagent , 5 μL ATP ( 100 mM or 2-fold serial dilutions thereof , Sigma ) , 5 μL methionine ( 100 mM or 2-fold serial dilutions thereof , Sigma ) , and 5 μL purified enzyme ( 3 μM or 1 . 5 μM , total volume 100 μL ) . Absorbance at 340 nm was measured continuously on a Synergy H1 plate reader ( BioTek ) at 37°C . Rate was determined from the linear fit to the data collected over 10 min . Steady-state kinetic constants ( Km and kcat ) were determined by fitting substrate versus velocity data to the Michaelis-Menten equation in GraphPad Prism . RNA was purified as previously reported [26] . Briefly , cells ( ≥5 × 107 ) were washed 3x with 10 mL of PBS ( 10mM Na2HPO4 , 1 . 9 mM KH2PO4 , 137 mM NaCl , 2 . 7 mM KCl , pH 7 . 4 ) , resuspended in 100 μL PBS , and 1 mL of Trizol ( Life Technologies/Invitrogen ) was added . Samples were incubated at RT for 5 min . For long-term storage , samples were flash flash-frozen in liquid N2 . Samples were extracted with chloroform and purified using RNeasy RNA Purification Kit ( Qiagen ) per manufacturer's protocol . Total RNA was quantified by measuring OD at 260/280 nm . Primers for qPCR are listed in S4 Table . cDNA was prepared as previously reported [26] . Briefly , 2 μg RNA was treated with DNaseI ( Invitrogen ) and quenched with EDTA . cDNA was synthesized with random hexamers ( Invitrogen ) using Moloney Murine Leukemia Virus Reverse Transcriptase ( M-MLV RT ) ( Invitrogen ) . cDNA levels were quantified using iTaq Universal SYBR Green Supermix ( Bio-Rad ) on a CFX 96 ( Bio-Rad ) or QuantStudio 7 Flex ( Applied Biosystems ) with a standard curve on each run for each primer . Relative mRNA levels were determined using the Pfaffle method [69] and Telomerase Reverse Transcriptase ( TERT ) was used as the reference gene [70] . For the TbAdoMetDC null+Hs line , α-Tubulin was used as the reference gene . Cells ( 107−108 ) were pelleted by centrifugation ( 2 , 000 × g , 5 min ) , washed 2x with PBS ( 137 mM NaCl , 2 . 7 mM KCl , 10 mM Na2HPO4 , 1 . 8 mM KH2PO4 , pH 7 . 4 ) , resuspended in 30–50 μL lysis buffer ( 50 mM HEPES , 100mM NaCl , pH 8 . 0 , freshly supplemented with 5mM β-mercaptoethanol , 2mM PMSF , 1 μg/mL leupeptin , 2 μg/mL antipain , 10 μg/mL benzamidine , 1 μg/mL pepstain , and 1 μg/mL chymostatin ) and lysed by 3 freeze/thaw cycles using liquid nitrogen . Cell debris was pelleted by centrifugation at 4°C , supernatant collected and protein concentration quantitated using Bio-Rad Protein Assay reagent . Samples ( 30 μg per lane ) were separated by SDS-PAGE on a 12% gel and transferred to PVDF using the iBlot transfer system ( Invitrogen ) , program P3 . Membranes were blocked using 5% Blotting-Grade Blocker ( Bio-rad ) in TBST ( 50 mM Tris , 150 mM NaCl , pH 7 . 4 , 0 . 1% ( v/v ) Tween-20 ) followed by incubation with primary antibody overnight at 4°C . After washing 3x with PBS ( 10 mL , 10 min ) , membranes were incubated with secondary antibody , either Protein A conjugated to HRP ( 1:1000 , AbCam ) or goat α-rabbit conjugated to HRP ( Sigma ) for 30 min at RT . Membranes were washed 3x with PBS , Supersignal West Pico substrate ( ThermoFisher ) was added and signal was imaged on a LAS 4000 imager ( GE Healthcare ) . Quantification of western blots was performed using ImageQuant TL 8 . 1 ( GE Healthcare ) . α-TbAdoMetDC , α-prozyme , and α-HsAdoMetDC antibodies have been previously described [26 , 34 , 71] . Antibody dilutions used were as follows: α-TbAdoMetDC ( rabbit , polyclonal , 1:2000 ) , α-prozyme ( rabbit , polyclonal , 1:2000 ) , α-BiP ( 1:50000 , BiP ) , α-HsAdoMetDC ( rabbit , polyclonal , 1:2000 , a gift from David Feith ) , α-TbAdoMetSyn ( rabbit , polyclonal , 1: 2000 ) . Cells ( 2 × 107 per sample ) were harvested in extraction buffer ( 80% MeOH , 0 . 1% formic acid for AdoMet and dcAdoMet targeted analysis and 80% methanol ( MeOH ) only for broad metabolite analysis ) and subjected to 5 freeze/thaw cycles . Cell debris was removed by centrifugation ( >17 , 000 × g , 10 min , 4°C ) , and the supernatant dried by vacuum centrifuge . For broad metabolite analysis , LC-MS/MS was performed as previously described [72] to provide analysis of 112 metabolites , excluding dcAdoMet . For dcAdoMet and AdoMet targeted analysis , samples were suspended in 150 μL solvent identical to the starting conditions of the chromatography method . Insoluble material was removed by centrifugation ( >17 , 000 × g , 10 min , 4°C ) and 10 μL of sample was injected for analysis . A Shimadzu Nexera X2 high-performance liquid chromatography ( HPLC ) coupled to a SCIEX 6500+ QTRAP was used for quantification of metabolites . Separation of metabolites was performed on a hydrophilic liquid chromatography column ( Luna HILIC , 100 x 4 . 6 mm , 3 μm , 4 Å , Phenomenex ) . The chromatography gradient consisted of two solvents: A: H2O , 0 . 2% formic acid , 5 mM ammonium acetate , B: 90% acetonitrile , 0 . 2% formic acid , 5 mM ammonium acetate . Optimal separation and detection was achieved with a flow rate of 1 . 0 mL/min and by the following gradient: 0 . 1–2 min 70% B , 2–3 min 20% B , 3–5 min 20% B , 5–5 . 1 70% B , 5 . 1–7 min 70% B . Infusion optimization was performed using standards obtained commercially ( Affymetrix ) or enzymatically synthesized as previously described to obtain optimal precursor and product ion masses for each metabolite [26] . In positive mode , multiple reaction monitoring ( MRM ) was used for detection and quantification of metabolites . The optimal linear response range of both dcAdoMet and AdoMet was determined using the authentic standards . At least two of the most abundant product ions were monitored and the calculated peak areas were normalized to uridine monophosphate as a spiked internal standard and the amount of total protein in the extracted pellet determined by bicinchoninic acid ( BCA ) assay . The relative abundance of each metabolite was determined by normalization of dcAdoMet and AdoMet signals to the untreated control . The following pairs of precursor/product ions were monitored: AdoMet ( 399/250 , 399/136 ) and dcAdoMet ( 355/250 , 355/298 , 355/136 ) . To measure methionine levels in FBS ( Gemini Bio-Products , Lot A29F ) , extracts were prepared with 200 μL 100% MeOH per 100μL serum and vortexed vigorously to precipitate protein . Insoluble material was removed by centrifugation ( >17 , 000 × g , 10 min , 4°C ) and the supernatant dried by vacuum centrifuge . LC-MS/MS analysis of methionine from FBS was performed as previously described [73] . The following pairs of precursor/product ions were monitored: Methionine ( 150/104 , 150/133 ) . Cells were cultured at 1 × 106 cells/mL in leucine-free HMI-19 medium with 10% dialyzed FBS and supplemented with 10 μM iso-propyl-d7 ( ( ( CD3 ) 2CDCH2CH ( NH2 ) COOH ) , 2H7-leucine ) ( CDN Isotopes ) , determined as the minimum L-leucine required to maintain prozyme upregulation by Genz-644131 ( S1B Fig ) . Cells ( 107−108 ) were harvested and processed for western blot analysis as above . Samples ( 50 μg per lane ) were separated by SDS-PAGE on a 4–20% gradient precast gel ( Bio-Rad ) . The gel was then stained with GelCode Blue Stain Reagent ( ThermoFisher ) and a 10 mm slice of the lane centered around 37 kDa was analyzed for unlabeled and d7-labeled prozyme by selected reaction monitoring ( SRM ) . The tryptic peptide sequences chosen for analysis were SAFPTGHPYLAGPVDR ( residues 157–172 ) and LEGFTVVHR ( residues 297–305 ) . These peptides were chosen because they contain only one leucine each , eliminating any complication from peptides that might potentially have a mix of heavy and light leucine . Additionally , we avoided peptides that were prone to missed cleavages ( consecutive R or K , for example ) , peptides that contained methionine ( potential oxidation ) , and peptides that contained cysteine ( potential for incomplete carbamidomethylation ) . After preliminary studies we settled on LEGFTVVHR ( residues 297–305 ) for quantification due to its lower limit of detection . Stable heavy-isotope-labeled peptides were synthesized as standards by 21st Century Biochemicals with purities of >97% as determined by HPLC . All peptides were synthesized with a C-terminal [13C6 , 15N4] arginine , and were used without further purification . Protein gel pieces were reduced and alkylated with DTT ( 20 mM ) and iodoacetamide ( 27 . 5 mM ) . A sufficient volume of 0 . 05 μg/μL solution of trypsin ( Pierce ) in 50 mM triethylammonium bicarbonate ( TEAB ) was added to completely cover the gel . The gel was allowed to sit on ice for 30 min and then 50 μL of 50 mM TEAB was added and the proteins were digested overnight . Peptides were then extracted from the gel and dried . Samples were reconstituted , spiked with 100 fmol of each heavy-isotope labeled peptide , and solid-phase extraction was performed with an Oasis HLB μelution plate ( Waters ) . Samples were dried and reconstituted in 10 μL of 2% ( v/v ) acetonitrile ( ACN ) and 0 . 1% trifluoroacetic acid in water for SRM analysis . The top seven transitions for each heavy-labeled peptide were determined by monitoring peak areas for all singly and doubly charged b and y ions below m/z = 1 , 250 and for all doubly and triply charged peptide ions below m/z = 1 , 000 , for a mix of the heavy-labeled peptide standards . These data were analyzed using Skyline v4 . 1 ( http://skyline . maccosslab . org ) [74] , and collision energies were optimized by a subsequent sample injection . Transitions that had interference from impurities or noise peaks were not included when performing peptide quantifications . Spiked samples were separated on a Dionex Acclaim PepMap100 reverse-phase C18 column ( 75 μm × 15 cm ) using an Ultimate 3000 RSLCnano HPLC system . The HPLC was controlled using Chromeleon Xpress ( version 6 . 8 SR10 ) and Dionex Chromatography MS Link v . 2 . 12 . Separation of peptides was carried out at 200 nL/min using a gradient from 0%–25% B for 15 min , 25%–35% B for 5 min , and 35%–80% B for 5 min , where mobile phase A was 2% ACN , 0 . 1% formic acid in water and mobile phase B was 80% ACN , 10% trifluoroethanol , 10% H2O , and 0 . 1% formic acid . Mass spectrometric analysis was performed on an AB Sciex 6500 QTRAP mass spectrometer in positive-ion low-mass mode , using a NanoSpray III source with a New Objective precut 360 μ PicoTip emitter ( FS360-20-10-N20-10 . 5CT ) . The source settings were as follows: curtain gas = 30 , ion spray voltage = 2 , 450 , ion source gas 1 = 6 . Analyst Software v . 1 . 6 was used to run the mass spectrometer . SRM data were analyzed using Skyline v4 . 1 . | Trypanosoma brucei is a single-celled eukaryotic pathogen and the causative agent of human African trypanosomiasis ( HAT ) . Polyamines are organic polycations that are essential for growth in T . brucei to facilitate protein translation and to maintain redox homeostasis . The pathway is the target of eflornithine , a current frontline therapy for treatment of HAT . Polyamine biosynthetic enzymes are regulated at multiple levels in mammals ( e . g . transcription , translation and protein turnover ) , but in contrast , T . brucei lacks these mechanisms . Instead in T . brucei a central enzyme in polyamine metabolism called AdoMetDC must form a complex with a sister protein ( termed a pseudoenzyme ) to be active . Herein , we show that cellular levels of this sister protein we call prozyme are in turn feedback regulated by both AdoMetDC and by its reaction product in response to cell treatments that reduce pathway output . This regulatory paradigm highlights how pseudoenzymes can evolve to play an important role in metabolic pathway regulation and in organismal fitness . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"rna",
"interference",
"protein",
"metabolism",
"chemical",
"compounds",
"aliphatic",
"amino",
"acids",
"messenger",
"rna",
"enzymology",
"parasitic",
"protozoans",
"organic",
"compounds",
"trypanosoma",
"brucei",
"leucine",
"protozoans",
"amino",
"acids",
"epigenetics",
... | 2018 | A dual regulatory circuit consisting of S-adenosylmethionine decarboxylase protein and its reaction product controls expression of the paralogous activator prozyme in Trypanosoma brucei |
PCR-Restriction Fragment Length Polymorphism ( RFLP ) analyses targeting multiple nuclear genes were established for the simple and practical identification of Leishmania species without using expensive equipment . This method was applied to 92 clinical samples collected at 33 sites in 14 provinces of Ecuador , which have been identified at the species level by the kinetoplast cytochrome b ( cyt b ) gene sequence analysis , and the results obtained by the two analyses were compared . Although most results corresponded between the two analyses , PCR-RFLP analyses revealed distribution of hybrid strains between Leishmania ( Viannia ) guyanensis and L . ( V . ) braziliensis and between L . ( V . ) guyanensis and L . ( V . ) panamensis , of which the latter was firstly identified in Ecuador . Moreover , unexpected parasite strains having the kinetoplast cyt b gene of L . ( V . ) braziliensis and nuclear genes of L . ( V . ) guyanensis , L . ( V . ) panamensis , or a hybrid between L . ( V . ) guyanensis and L . ( V . ) panamensis were identified . This is the first report of the distribution of a protozoan parasite having mismatches between kinetoplast and nuclear genes , known as mito-nuclear discordance . The result demonstrated that genetically complex Leishmania strains are present in Ecuador . Since genetic exchanges such as hybrid formation were suggested to cause higher pathogenicity in Leishmania and may be transmitted by more species of sand flies , further country-wide epidemiological studies on clinical symptoms , as well as transmissible vectors , will be necessary .
Leishmaniasis , caused by protozoan parasites of the genus Leishmania , is a neglected tropical disease widely distributed worldwide , especially in tropical and subtropical areas , affecting at least 12 million people in 96 countries [1] . Approximately 20 Leishmania species belonging to the subgenera Leishmania ( Leishmania ) , Leishmania ( Viannia ) and Leishmania ( Mundinia ) are pathogenic to humans [1 , 2] . Since infected parasite species is known to be the major determinant of clinical outcomes in leishmaniasis [1] , identification of the causative parasite is important for appropriate treatment and prognosis . Leishmania species have been classified conventionally by multilocus enzyme electrophoresis ( MLEE ) [3 , 4] . Genetic analysis of kinetoplast and nuclear targets , such as cytochrome b ( cyt b ) , cysteine protease ( cpb ) , heat shock protein 70 ( hsp70 ) genes and the internal transcribed spacer ( ITS ) regions of ribosomal RNA , has commonly been used for species identification due to its sensitivity , simplicity and reliability [5–13] . In addition , a simple PCR-Restriction Fragment Length Polymorphism ( RFLP ) , which does not require costly equipment , was developed for species identification , and the ITS region and hsp70 gene are widely applied to epidemiological studies [11 , 14–19] . In Ecuador , leishmaniasis is endemic in Pacific coast , Andean highland , and Amazonian areas , and eight species , Leishmania ( Leishmania ) mexicana , L . ( L . ) amazonensis , L . ( L . ) major-like , L . ( Viannia ) guyanensis , L . ( V . ) panamensis , L . ( V . ) braziliensis , L . ( V . ) naiffi , and L . ( V . ) lainsoni , have been recorded as causative agents of cutaneous leishmaniasis ( CL ) and mucocutaneous leishmaniasis ( MCL ) [8 , 20 , 21] . Of these , distribution of L . ( L . ) amazonensis and L . ( L . ) major-like have been reported to be localized , and infections by them have not been reported recently [8 , 21] . Infection by L . ( V . ) guyanensis together with its closely-related species , L . ( V . ) panamensis , has been identified from CL patients in Pacific coast areas by MLEE [21–24]; however , our recent cyt b gene analysis revealed a wide range distribution of L . ( V . ) guyanensis , without detecting any L . ( V . ) panamensis in these areas [8] . These results suggest that endemic species may change , or the reported results may be caused by the discordance between the MLEE analysis and kinetoplast cyt b gene analysis employed for species identification . Recently , a countrywide epidemiological study was carried out based on the cyt b sequence analysis and it identified L . ( V . ) guyanensis and L . ( V . ) braziliensis widely in Pacific coast and Amazonian areas and L . ( L . ) mexicana in Andean high lands as current major causative species in Ecuador [8] . Additionally , L . ( V . ) naiffi and L . ( V . ) lainsoni were recently recorded in Amazonian areas [8 , 20 , 25] . In this study , a simple and practical method for the identification of Leishmania species in Ecuador was established on the basis of PCR-RFLP analyses targeting mannose phosphate isomerase ( mpi ) and 6-phosphogluconate dehydrogenase ( 6pgd ) genes , and the result was compared with that obtained by the cyt b gene sequence analysis . This study demonstrated the presence of genetically complex Leishmania strains in Ecuador , and strongly suggested the importance of applying multiple target approaches to enhance the reliability of species identification and to characterize more detailed genetic properties of the parasite .
Frozen stocks of 24 parasite strains of five Leishmania species [L . ( V . ) guyanensis , L . ( V . ) panamensis , L . ( V . ) braziliensis , L . ( L . ) major-like , L . ( L . ) mexicana] that were isolated from CL patients in Ecuador and identified at the species level by MLEE [22–24] ( Table 1 ) were spotted on an FTA Classic Card ( Whatman , Newton Center , MA ) and subjected to sequence analysis . Three strains of L . ( V . ) naiffi identified by cyt b gene analysis [25 , 26] were also utilized ( Table 1 ) . Most of the clinical samples employed in this study were collected from patients suspected of CL in the previous study [8 , 20] , and each 3 samples newly obtained from Provinces of Manabi and Santo Domingo de los Tsachilas , all of which were identified as L . ( V . ) guyanensis by the cyt b gene analysis , were included in this study . Leishmania parasites were identified on the basis of cyt b sequence analysis [8 , 20] . The samples were collected at 33 sites in 14 provinces of Ecuador ( S1 Fig ) . Residual tissue materials were spotted onto an FTA Classic Card , after taking scraped margin samples of active lesions for routine diagnosis . Two-mm-diameter disks of FTA card were punched out from each filter paper , washed three times with an FTA Purification Reagent ( Whatman ) , and subjected to PCR amplification . PCR primers for amplification of cyt b , hsp70 , mannose phosphate isomerase ( mpi ) and 6-phosphogluconate dehydrogenase ( 6pgd ) gene fragments were designed based on the sequence regions conserved among species ( Table 2 ) . PCR amplification with a pair of outer primers was performed with 30 cycles of denaturation ( 95°C , 1 min ) , annealing ( 55°C , 1 min ) and polymerization ( 72°C , 2 min ) using Ampdirect Plus reagent ( Shimadzu Biotech , Tsukuba , Japan ) . Each 0 . 5-μl portion of the PCR product was reamplified with inner primers under the same condition described above . The products were cloned into the pGEM-T Easy Vector System ( Promega , Madison , WI ) and sequences were determined on both strands by the dideoxy chain termination method using a BigDye Terminator v3 . 1 Cycle Sequencing Kit ( Applied Biosystems , Foster City , CA ) . Primers for amplification of a partial sequence of the kinetoplast cytochrome oxidase subunit II-NADH dehydrogenase subunit I region ( COII-ND1 ) were also designed based on the sequences conserved among species ( Table 2 ) . The COII-ND1 sequences were determined on both strands by direct sequencing with inner primers , L . COII-2S and L . COII-2R . Restriction enzyme mapping was performed in silico by using BioEdit Sequence Alignment Editor to obtain species-specific RFLP patterns . Clinical samples spotted on FTA cards , in which parasites were identified by cyt b gene analysis in a previous study , were subjected to PCR-RFLP analysis . PCR amplifications targeting mpi and 6pgd were performed as described above using a high fidelity DNA polymerase , KOD plus ( Toyobo , Osaka , Japan ) . The PCR products were digested by restriction enzymes HaeIII , HapI , and BstXI for the mpi gene and Bsp1286I and HinfI for the 6pgd gene , and resulting restriction fragment patterns were analyzed by 2% agarose gel electrophoresis . GeneRuler 100 bp Plus DNA Ladder ( Thermo Fisher Scientific , Waltham , MA ) was used as a DNA size marker . The gel was stained with GelRed Nucleic Acid Gel Stain ( Biotium , Hayward , CA ) , and DNA fragments were visualized with UV transilluminator . Differentiation between L . ( V . ) guyanensis and L . ( V . ) panamensis was performed by restriction enzyme-digestion of the hsp70 gene fragment [27] . Briefly , the hsp70 gene fragment was amplified by a nested PCR using sets of outer primers ( L . HSP-Ty1S and L . HSP-OR ) and inner primers ( L . HSP-Ty2S and L . HSP-IR2 ) ( Table 2 ) . The amplicons were digested with a restriction enzyme , BccI , and resulting fragment patterns were analyzed by 3% agarose gel electrophoresis . Clinical samples were collected by local physicians and well-trained laboratory technicians of health centers of the Ministry of Health , Ecuador . For routine parasitological diagnosis , scratching smear samples of skin lesions were taken from suspected leishmaniasis patients at health centers . In this study , only residual tissue materials were collected after the routine procedure to minimize the burden on patients . Signed consent was obtained from the adult subjects and from the children’s parents or guardians , prior to the diagnostic procedures at each health center of the Ministry , providing information on the process of diagnosis and Leishmania species analysis , following the guidelines of the Ethics Committee of the Ministry . The subjects studied were volunteers in routine diagnosis/screening and treatment programs promoted by the Ministry . All routine laboratory examinations were carried out free of charge , and treatment with specific drug , meglumine antimoniate ( Glucantime ) was also offered free of charge at each health center . The study was approved by the ethics committee of the Graduate School of Veterinary Medicine , Hokkaido University ( approval number: vet26-4 ) and Jichi Medical University ( approval number: 17–080 ) [8] .
Leishmania cyt b , hsp70 , mpi and 6pgd partial gene sequences were amplified from 27 strains of 6 species isolated in Ecuador . Sequences of these fragments showed high degrees of homology ( 88–100% , 82–100% , 83–100% and 94–100% in cyt b , mpi , 6pgd and hsp70 genes , respectively ) with corresponding leishmanial genes registered in GenBank . The restriction enzyme mapping was performed in silico to see if species-specific enzyme sites could be found in cyt b , mpi , 6pgd and hsp70 gene fragments obtained in this study . Species-specific RFLP patterns could not be obtained for the cyt b gene because of intraspecies genetic variations through the sequences . On the hsp70 gene , restriction enzymes to differentiate Leishmania species were found; however , RFLP patterns including several smaller fragments ( < 300 bp ) were similar among species . Therefore , it seems difficult to identify the species based on RFLP patterns of hsp70 using agarose gel electrophoresis in some cases because of the resolution . On the other hand , restriction enzyme sites that can differentiate Leishmania species in Ecuador were identified in mpi and 6pgd genes , except for two very closely-related species , L . ( V . ) guyanensis and L . ( V . ) panamensis . Different RFLP patterns were obtained in L . ( V . ) guyanensis/L . ( V . ) panamensis , L . ( V . ) braziliensis/L . ( V . ) naiffi , L . ( L . ) major-like and L . ( L . ) mexicana for digested mpi gene fragments with a restriction enzyme HaeIII ( Fig 1A ) . Although an RFLP polymorphism was observed in one ( strain PT27 ) of seven L . ( L . ) mexicana strains , it did not affect species identification ( Table 3 ) . L . ( V . ) braziliensis and L . ( V . ) naiffi , showing the same RFLP patterns as HaeIII digestion , were differentiated by HpaI digestion ( Table 3 , Fig 1B ) . Although L . ( V . ) lainsoni , a recently reported species in the Ecuadorian Amazon [20] , showed the same RFLP patterns as L . ( V . ) guyanensis/L . ( V . ) panamensis when digested with HaeIII and HpaI , BstXI-digestion successfully differentiated it from L . ( V . ) guyanensis/L . ( V . ) panamensis , as reported in Peruvian strains ( S2 Fig ) [28] . Digestion of the 6pgd gene with Bsp1286I resulted in distinct gene fragment patterns of L . ( V . ) guyanensis/L . ( V . ) panamensis , L . ( V . ) braziliensis , L . ( V . ) naiffi , L . ( L . ) major-like and L . ( L . ) mexicana; however , the patterns between L . ( V . ) guyanensis/L . ( V . ) panamensis and L . ( V . ) naiffi were similar and difficult to discriminate because of only about a 50 bp difference in a fragment of approximately 1 kbp ( Fig 2A ) . The two species were successfully differentiated by digesting with HinfI ( Fig 2B ) . Although L . ( V . ) guyanensis and L . ( V . ) panamensis were not discriminated by PCR-RFLP of mpi and 6pgd genes , PCR-RFLP of the hsp70 gene with a restriction enzyme , BccI , successfully differentiated the two species as reported previously ( Fig 3 ) [27] . PCR-RFLP analyses of mpi gene with restriction enzymes , HaeIII and HpaI , and 6pgd gene with Bsp1286I and HinfI were applied to 92 clinical samples collected at 33 sites in 14 provinces of Ecuador . PCR-RFLP analysis of the hsp70 gene with a restriction enzyme , BccI , was used for differentiation between L . ( V . ) guyanensis and L . ( V . ) panamensis . The results obtained by PCR-RFLP analyses were compared with those obtained by the cyt b gene sequence analysis . The results of the species identification obtained by the two nuclear genes always agreed with each other . The identification by PCR-RFLP analyses completely matched with that obtained by the cyt b gene sequence analysis in all of L . ( V . ) naiffi ( 2 samples ) and L . ( L . ) mexicana ( 3 samples ) ( Table 4 ) . Of the 73 samples identified as L . ( V . ) guyanensis by cyt b gene analysis , 72 samples were identified as L . ( V . ) guyanensis by PCR-RFLP analyses , whereas one sample from a Pacific coast area showed a hybrid pattern of L . ( V . ) guyanensis and L . ( V . ) panamensis based on the PCR-RFLP of the hsp70 gene ( Figs 3 and 4 ) . The sequence of the hsp70 gene fragment was analyzed by direct sequencing , and a single nucleotide polymorphism was confirmed , showing “C” in L . ( V . ) guyanensis but “T” in L . ( V . ) panamensis , whereas a sample having a hybrid RFLP pattern had both “C” and “T” peaks at the corresponding position ( S3 Fig ) , indicating the presence of a hybrid strain of L . ( V . ) guyanensis and L . ( V . ) panamensis in Ecuador . On the other hand , of the 14 samples identified as L . ( V . ) braziliensis by cyt b gene analysis , only 6 samples were identified as L . ( V . ) braziliensis by RFLP analyses ( Table 4 ) . In the other 8 samples identified as L . ( V . ) braziliensis by the cyt b gene analysis , three samples showed hybrid patterns in PCR-RFLP analyses of both the mpi and 6pgd genes ( Fig 5A and 5B ) . The sequences of mpi and 6pgd gene fragments were analyzed by direct sequencing , and a single nucleotide polymorphism was confirmed , showing “C” in L . ( V . ) guyanensis but “T” in L . ( V . ) braziliensis of the mpi gene , and “T” in L . ( V . ) guyanensis but “C” in L . ( V . ) braziliensis of the 6pgd gene . On the other hand , the mpi and 6pgd genes from the three samples with hybrid RFLP patterns had both “C” and “T” peaks at the corresponding position ( S4 Fig ) . From these results , the parasite species of these three samples were identified as a hybrid of L . ( V . ) braziliensis and L . ( V . ) guyanensis ( Table 4 , Fig 4 ) . In the remaining 5 samples identified as L . ( V . ) braziliensis by sequence analysis of the cyt b gene , PCR-RFLP analyses showed that one sample from a Pacific coast area was L . ( V . ) guyanensis , three samples from the northern Pacific coast and Amazonian areas were L . ( V . ) panamensis , and one sample from a northern Pacific coast area had a hybrid pattern of L . ( V . ) guyanensis and L . ( V . ) panamensis ( Table 4 , Fig 4 ) . The sequence analyses of mpi , 6pgd , and hsp70 gene fragments corresponded to PCR-RFLP analyses , indicating the presence of a mismatch between kinetoplast and nuclear genes , known as mito-nuclear discordance , in Leishmania distributing in Ecuador ( Table 4 , Fig 4 ) . To further confirm the mito-nuclear discordance , partial sequences of the COII-ND1 region were analyzed as another target of kinetoplast genes in samples showing a mismatch between kinetoplast cyt b gene and nuclear mpi , 6pgd and hsp70 genes . The sequences were compared to each two corresponding sequences obtained from L . ( V . ) braziliensis and L . ( V . ) guyanensis in this study since this region has not been well-analyzed in subgenus Viannia species . The sequences from parasites with mito-nuclear discordance showed 98 . 9–99 . 1% and 98 . 5–98 . 9% identities with those of L . ( V . ) braziliensis and L . ( V . ) guyanensis , respectively ( accession numbers: LC475135-LC475142 ) . When partial COII gene sequences in the obtained COII-ND1 region sequences were analyzed on the GenBank database , the sequences from parasites with mito-nuclear discordance showed 99 . 5% and 98 . 9% identities with those of L . ( V . ) braziliensis and L . ( V . ) guyanensis , respectively . This result strongly suggested that the kinetoplast genes of these parasites originated from L . ( V . ) braziliensis , corresponding to the result of cyt b gene analysis .
In the present study , PCR-RFLP analyses were employed for the identification of Leishmania species distributing in Ecuador in order to develop a simple and practical way for species identification independent of expensive equipment such as a genetic analyzer . As a result , mpi and 6pgd genes , for which encoding enzymes have been widely used as the gold standard of species identification , were identified as suitable targets for this purpose in the tested samples . The results obtained by the PCR-RFLP analyses of multiple nuclear targets were compared to those of cyt b gene sequence analysis [7 , 8 , 29–36] . Although most results corresponded between the two analyses , PCR-RFLP revealed distribution of hybrid and mito-nuclear discordant Leishmania strains , which could not be identified only by cyt b gene sequence analysis . The results indicated that Leishmania strains distributing in Ecuador are genetically more complex than previously thought . PCR-RFLP analysis has been employed for species identification of Leishmania species , and its utility is widely accepted [34] . The rRNA internal transcribed spacer 1 ( ITS-1 ) region and hsp70 gene are mostly used as suitable target genes , of which the former is applied mainly in the Old World [6 , 11 , 12 , 14 , 17 , 19 , 27 , 34 , 37–41] . Although the hsp70 gene is one of the most valuable genetic markers for PCR-RFLP-based species identification , intraspecific polymorphism of RFLP patterns and very similar RFLP profiles among species , which affect species identification , have been reported in some Leishmania species [42] . In this study , other nuclear genes , mpi and 6pgd genes , for which encoding enzymes have been used for MLEE , were shown to be alternative useful targets for classification by PCR-RFLP analysis . Of these , the mpi gene was reported to be the only genetic marker that can distinguish two very closely-related species , L . ( V . ) braziliensis and L . ( V . ) peruviana [7 , 43 , 44] . In addition , a recent study demonstrated that PCR-RFLP of the shorter mpi gene fragment ( approximately 500 bp ) can differentiate 4 Leishmania species [L . ( V . ) braziliensis , L . ( V . ) peruviana , L . ( V . ) guyanensis , and L . ( V . ) lainsoni] and a hybrid of L . ( V . ) braziliensis and L . ( V . ) peruviana circulating in the Department of Huanuco , Peru [28] . In the present study , PCR-RFLP analyses of longer mpi and 6pgd gene fragments ( >1000bp ) were successfully established and applied to 92 clinical samples in Ecuador . Although a polymorphic RFLP pattern , which does not affect the identification , was detected in the mpi of one L . ( L . ) mexicana strain , the variant RFLP pattern was not detected in the present clinical samples identified as L . ( L . ) mexicana . Further sample analyses from different areas and different countries will be important to confirm the utility of this analysis , although polymorphic RFLP profiles may be detectable in these genes . Since polymorphism was also reported in the hsp70 gene of several Leishmania species [42] , PCR-RFLP analyses of multiple target genes , rather than single nuclear or kinetoplast genes , will result in more accurate species identification and disclose more detailed genetic characteristics of the parasite . Several samples showing hybrid RFLP patterns were identified as hybrid strains rather than mixed infection of different Leishmania species . It is due to the following reasons: 1 ) It is little or no chance to be infected by more than one parasite in a cutaneous lesion because the lesion is typically developed at the site bitten by a sand fly transmitting specific Leishmania species , 2 ) Even if mixed infection occurs , either parasite becomes dominant in the lesion , resulting in the presence of dominant allele by the genetic analysis . However , both alleles were comparably amplified as observed in the PCR-RFLP analysis , which is indicative of a putative hybrid strain . In addition , similar results were obtained on electrograms of the direct sequencing , showing comparable fluorescence intensities of polymorphic nucleotides derived from both species . 3 ) The presence of hybrid strain has been reported in the same area as described below [45] . Isolation of putative hybrid strains as a culture is necessary for further detailed characterization of these parasites . Although multiple PCR-RFLP and cyt b sequence analyses showed corresponding results in most clinical samples , the present study revealed the distribution of several unexpected strains in Ecuador , including hybrid and mito-nuclear discordance strains . Since hybrid strains cannot be identified by the cyt b gene analysis after molecular cloning , this is another advantage of identifying parasite species by PCR-RFLP . Distribution of a hybrid strain of L . ( V . ) guyanensis/panamensis complex and L . ( V . ) braziliensis was reported in Zumba , a province of Zamora-Chinchipe in a southern part of Ecuador by using MLEE and random amplified polymorphic DNA ( RAPD ) [45] . The present study confirmed the presence of the hybrid strain in Zumba , and also in another area in the same province , Palanda . In addition , a hybrid of L . ( V . ) guyanensis and L . ( V . ) panamensis was detected in northern Pacific areas of Ecuador . This is the first report of the presence of a hybrid strain of L . ( V . ) guyanensis and L . ( V . ) panamensis in Ecuador . L . ( V . ) guyanensis and its closely related L . ( V . ) panamensis have been reported to be endemic in northern Pacific areas of Ecuador by MLEE; however , only L . ( V . ) guyanensis was identified in the same areas by cyt b gene analysis in recent studies [8 , 21 , 46] . The present study confirmed that L . ( V . ) guyanensis is dominantly present in these areas , suggesting that endemic species may change , or that there may be discordance between MLEE and genetic analysis . However , the identification of a hybrid of L . ( V . ) guyanensis and L . ( V . ) panamensis as a minor population suggests that parental L . ( V . ) panamensis may still be present in some of these areas . Another unexpected finding was identification of mito-nuclear discordant strains of Leishmania species in northern Pacific and Amazonian areas . Interestingly , mito-nuclear discordant strains were identified only in the species identified as L . ( V . ) braziliensis by cyt b gene analysis . This finding supports a recent study using cyt b gene analysis reporting increasing cases of L . ( V . ) braziliensis infection in Pacific coast areas when compared to previous studies using enzymatic MLEE analysis [8] . The hybrid strain of L . ( V . ) braziliensis and L . ( V . ) peruviana was suggested to increase disease severity when compared to parental species in an animal model [47] . Therefore , careful investigation is needed to clarify the presence of hybrid strains , including mito-nuclear discordance , and their effects on clinical courses . In addition , hybrid strains may increase the range of transmissible sand fly species if they have a potential to be transmitted by both vector species of parental parasites . Continuous vector research is important in these endemic areas , as well as parasitological and clinical studies . Further , basic parasitological research on how genetic exchange and mito-nuclear discordance occur among Leishmania species would be another interesting subject [48–51] . Mito-nuclear discordance is reported in various animals such as mammals , birds , reptiles , amphibians , fish and insects , and is inferred to result from various processes: 1 ) adaptive introgression of mitochondrial DNA , 2 ) demographic disparities , 3 ) sex-biased asymmetries , 4 ) hybrid zone movement , 5 ) an intracellular bacteria , Wolbachia infection in insects , and 6 ) human actions [52] . It provides deeper insights into the phylogenetic relationship , population structure , and evolutionary signature of these animals . Mito-nuclear discordance is also reported in helminth parasites: trematodes Schistosoma turkestanicum between populations [53] , and cestodes Taenia solium between lineages [54] , and between T . saginata and T . asiatica [55–57] . This is the first report of mito-nuclear discordance in protozoan parasites . Mito-nuclear discordance is speculated to be resulted from the similar process as hybridization of nuclear genes in protozoa . Further study is needed to disclose the mechanism of mito-nuclear discordance formation in protozoa . In addition , association of mito-nuclear discordance with the pathogenicity and vector competency of the parasites is important issues to be clarified . In this study , we established a novel PCR-RFLP-based genotyping approach to identify Leishmania species in Ecuador . Although the present PCR-RFLP analyses was shown to be practical for identification of Leishmania species in Ecuador , further study focusing on other Leishmania species and clinical samples from different countries will be needed to enhance the utility of this approach . PCR-RFLP analyses of clinical samples and subsequent comparison with kinetoplast cyt b sequence analysis revealed the distribution of genetically complex Leishmania strains having genetic characteristics of hybrid and mito-nuclear discordance . Although intraspecies genetic variation observed in the cyt b gene resulted in this gene as an unsuitable target for RFLP analysis , there is no doubt about the utility of cyt b gene sequence analysis for species identification and phylogenetic analysis since distinct interspecies genetic diversity of this gene overcomes the disadvantage of the intraspecies variation . However , the present study points to the importance of applying multiple target approaches as the combination of cyt b and the PCR-RFLP assays presented here , enhancing the reliability of species identification and characterization of genetic properties including hybrid and mito-nuclear discordance . Further studies are needed to reveal the parasitological characteristics of hybrid and mito-nuclear discordance , clinical outcomes caused by these parasites , and the range of vector species of these parasites . In addition , studies on mito-nuclear discordance in Leishmania and other protozoa may provide further insights into the mechanism of genetic exchanges of these parasites . | Leishmaniasis caused by intracellular protozoa of the genus Leishmania is a neglected tropical disease widely distributing worldwide , especially in tropical and subtropical areas . Approximately 20 species are known to be pathogenic to humans , of which eight species have been recorded as causative agents of cutaneous and mucocutaneous leishmaniases in Ecuador . Since infecting species are the major determinant of clinical outcomes , identification at the species level is important for the treatment and prognosis . The parasite species have been identified conventionally by multilocus enzyme electrophoresis ( MLEE ) and recently by genetic analysis such as sequencing and genotyping . In the present study , PCR-Restriction Fragment Length Polymorphism ( RFLP ) targeting multiple nuclear genes was employed , and the results were compared with those obtained by kinetoplast cytochrome b ( cyt b ) gene sequence analysis , which is widely applied to species identification . Although most results corresponded between the two analyses , PCR-RFLP revealed presence of unexpected genetically complex Leishmania strains having characteristics of hybrid and mito-nuclear discordance . Since hybrid strains of Leishmania were suggested to increase disease severity and may be transmitted by a wider range of sand fly species , careful epidemiological research , including clinical courses and vector research , will be needed . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"ecuador",
"geographical",
"locations",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"protozoans",
"leishmania",
"molecular",
"biology",
"techniques",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analys... | 2019 | PCR-RFLP analyses of Leishmania species causing cutaneous and mucocutaneous leishmaniasis revealed distribution of genetically complex strains with hybrid and mito-nuclear discordance in Ecuador |
The Gram-negative bacterial plant pathogen Xanthomonas campestris pv . vesicatoria employs a type III secretion ( T3S ) system to inject bacterial effector proteins into the host cell cytoplasm . One essential pathogenicity factor is HrpB2 , which is secreted by the T3S system . We show that secretion of HrpB2 is suppressed by HpaC , which was previously identified as a T3S control protein . Since HpaC promotes secretion of translocon and effector proteins but inhibits secretion of HrpB2 , HpaC presumably acts as a T3S substrate specificity switch protein . Protein–protein interaction studies revealed that HpaC interacts with HrpB2 and the C-terminal domain of HrcU , a conserved inner membrane component of the T3S system . However , no interaction was observed between HpaC and the full-length HrcU protein . Analysis of HpaC deletion derivatives revealed that the binding site for the C-terminal domain of HrcU is essential for HpaC function . This suggests that HpaC binding to the HrcU C terminus is key for the control of T3S . The C terminus of HrcU also provides a binding site for HrpB2; however , no interaction was observed with other T3S substrates including pilus , translocon and effector proteins . This is in contrast to HrcU homologs from animal pathogenic bacteria suggesting evolution of distinct mechanisms in plant and animal pathogenic bacteria for T3S substrate recognition .
Many Gram-negative bacterial pathogens of plants and animals depend on a type III secretion ( T3S ) system to successfully infect their hosts [1] . The term “T3S system” refers to both translocation-associated and flagellar T3S systems that evolved from a common ancestor [2] . Eleven components of the membrane-spanning basal body are conserved , suggesting a similar overall architecture of the secretion apparatus [1] , [3] . Main structural differences are found in the extracellular appendages associated with the basal body . The flagellar T3S apparatus is connected via an extracellular hook to the filament , the key bacterial motility organelle [4] . By contrast , the basal body of translocation-associated T3S systems is associated with an extracellular pilus ( plant pathogens ) or needle ( animal pathogens ) , which serve as conduits for secreted proteins to the host-pathogen interface [1] , [5] . Pilus and needle are proposed to be linked to the T3S translocon , a channel-like protein complex that is inserted into the eukaryotic plasma membrane and allows protein translocation into the host cell cytosol [6] , [7] . Translocation-associated T3S systems secrete two types of proteins , i . e . , extracellular components of the secretion apparatus such as needle/pilus and translocon proteins , and effectors that are translocated into the host cell [3] . Efficient secretion and/or translocation of T3S substrates depends on a signal in the N terminus , which is not conserved on the amino acid level [1] , [8] , [9] . In many cases , specific T3S chaperones bind to one or several homologous T3S substrates in the bacterial cytoplasm and promote stability and/or secretion of their respective binding partners . T3S chaperones are small , acidic and leucine-rich proteins that presumably guide secreted proteins to the secretion apparatus at the inner membrane [1] , [10] , [11] . Given the architecture of the T3S system , it is conceivable that secretion of extracellular components of the secretion apparatus precedes effector protein translocation . In translocation-associated and flagellar T3S systems from animal pathogenic bacteria , experimental evidence suggests that substrate specificity is altered by so-called T3S substrate specificity switch ( T3S4 ) proteins , e . g . , YscP from Yersinia spp . and the flagellar homolog FliK [12]–[14] . The substrate specificity switch depends on the C-terminal cytoplasmic domain of a conserved inner membrane protein of the FlhB/YscU family [12] , [13] . T3S4 proteins are not highly conserved among different pathogens and have so far only been identified in animal pathogenic bacteria [14] , [15] . It therefore remained enigmatic whether plant pathogenic bacteria employ similar mechanisms to orchestrate secretion of different T3S substrates . In our laboratory , we study T3S of the plant pathogenic bacterium Xanthomonas campestris pv . vesicatoria , the causal agent of bacterial spot disease in pepper and tomato . The T3S system of X . campestris pv . vesicatoria is essential for bacterial growth and disease symptom formation in susceptible host plants and the induction of the hypersensitive response ( HR ) in resistant plants . The HR is a rapid programmed cell death at the infection site that is triggered upon recognition of individual effector proteins , also termed avirulence ( Avr ) proteins , in plants that carry a cognate disease resistance gene [16] , [17] . In susceptible plants , effector proteins presumably modulate host cellular pathways to the pathogen's benefit and thus primarily act as virulence factors [18] , [19] . The T3S system from X . campestris pv . vesicatoria is encoded by the 23-kb chromosomal hrp ( hypersensitive response and pathogenicity ) gene cluster , which is organized in eight operons and contains 25 genes [20]–[23] . Eleven genes ( termed hrc for hrp conserved ) encode proteins that are conserved among plant and animal pathogenic bacteria [24] . Most hrp genes are essential for bacterial pathogenicity [25] . Several Hrp proteins are secreted and thus constitute extracellular components of the T3S system such as the pilus protein HrpE and the translocon protein HrpF [25]–[27] . T3S of extracellular components of the secretion apparatus and effector proteins is presumably controlled by the products of hpa ( hrp-associated ) genes that are encoded in the hrp gene cluster and contribute to pathogenicity [28]–[31] . Examples are the export control protein HpaC , which is required for the efficient secretion of translocon and some effector proteins , and the global T3S chaperone HpaB , which promotes secretion and translocation of multiple effector proteins [30] , [31] . In this study , we analyzed HrpB2 , which is an essential pathogenicity factor of X . campestris pv . vesicatoria . HrpB2 is a 13 . 7-kDa protein that is encoded by the second gene of the hrpB operon and is secreted by the T3S system [25] , [32] . Homologous proteins are present in Xanthomonas spp . , Burkholderia spp . and Ralstonia solanacearum , suggesting that HrpB2 also plays an important role in other pathogens . In X . campestris pv . vesicatoria , HrpB2 is essential for pilus formation and T3S and is therefore presumably one of the first proteins that travel the T3S apparatus [25] , [26] . However , the mode of HrpB2 action is unknown because HrpB2 does not share significant sequence or structural similarity with a protein of known function . Here , we provide experimental evidence that secretion of HrpB2 is required for bacterial pathogenicity . Secretion of HrpB2 is significantly enhanced in the absence of the export control protein HpaC . Protein-protein interaction studies showed that HrpB2 binds to HpaC and to the C-terminal domain of the conserved inner membrane protein HrcU , which also interacts with HpaC . Our data suggest that the interaction between HpaC and the C-terminal domain of HrcU promotes a switch in substrate specificity of the T3S system from HrpB2 secretion to secretion of translocon and effector proteins .
Previously , we identified HrpB2 as a T3S substrate of X . campestris pv . vesicatoria strain 85-10 [25] . Infection studies with hrpB2 deletion mutant strains revealed that HrpB2 is crucial for disease symptoms in susceptible and the HR induction in resistant pepper plants [25] . Similar results were obtained with strains 85* and 85*ΔhrpB2 , which carry hrpG* , a mutated version of the key regulatory gene hrpG in the bacterial chromosome ( Fig 1A ) . hrpG* leads to constitutive expression of the T3S system and is key for the analysis of in vitro T3S [33] . It is noteworthy that in planta growth of hrpG* strains is like wild-type [34] . The hrpB2 mutant phenotype could be complemented by ectopic expression of hrpB2 , suggesting that loss of pathogenicity was specifically due to the deletion of hrpB2 and did not result from a polar effect of the mutation on expression of other genes in the hrpB operon ( Fig . 1A ) . The fact that secretion of the effector protein AvrBs3 is abolished in hrpB2 deletion mutants suggested that HrpB2 is involved in T3S [25] . To investigate the contribution of HrpB2 to secretion of additional T3S substrates , strains 85* and 85*ΔhrpB2 were incubated in secretion medium , and total cell extracts and culture supernatants were analyzed by immunoblotting using specific polyclonal antibodies . We tested secretion of the putative translocon proteins HrpF and XopA , and the pilus protein HrpE . For technical reasons , HrpE was studied as a fusion protein consisting of the N-terminal 50 amino acids of HrpE and the reporter protein AvrBs3Δ2 , which is a derivative of AvrBs3 . AvrBs3Δ2 lacks the first 152 amino acids and thus the secretion and translocation signal [35] . It was previously demonstrated that the N-terminal 50 amino acids of HrpE restore secretion but not translocation of AvrBs3Δ2 , indicating that they contain a functional T3S signal [36] . Fig . 1B shows that HrpF , XopA and HrpE1–50-AvrBs3Δ2 were present in the culture supernatant of the wild-type strain but were not detectable in the supernatant of the hrpB2 deletion mutant , suggesting that HrpB2 is essential for secretion of these proteins . Since HrpB2 is secreted and is also required for T3S , it is presumably one of the first substrates that travel the T3S apparatus [25] . Notably , the amount of HrpB2 present in the culture supernatant of strain 85* is at the detection limit of the HrpB2-specific antibody [25] . Similar results were observed for a C-terminally c-Myc epitope-tagged version of HrpB2 , suggesting that HrpB2 is only weakly secreted by the T3S system ( Fig . 2A ) . To investigate whether HrpB2 secretion is regulated by the known export control proteins HpaB and HpaC , we performed in vitro T3S assays with strains 85* , the hpaB deletion mutant 85*ΔhpaB and the hpaC deletion mutant 85*ΔhpaC . We did not detect any influence of the global T3S chaperone HpaB on secretion of HrpB2 . Interestingly , however , significantly increased amounts of HrpB2 were secreted by strain 85*ΔhpaC ( Fig . 2 ) . This was not due to a general increase of T3S in strain 85*ΔhpaC since secretion of the translocon protein HrpF was reduced when compared to the wild-type strain 85* ( Fig . 2A and C ) . This is in agreement with the previous finding that HpaC is required for the efficient secretion of translocon and some effector proteins [31] . Oversecretion of HrpB2 in strain 85*ΔhpaC was specifically due to deletion of hpaC since the secretion phenotype was complemented by ectopic expression of hpaC-c-myc ( Fig . 2B ) . We did not detect HrpB2 in the culture supernatant of the T3S double mutant 85*ΔhpaCΔhrpE , which additionally lacks the Hrp pilus gene hrpE ( Fig . 2C ) . We therefore conclude that increased HrpB2 secretion in strain 85*ΔhpaC was mediated by the translocation-associated T3S system . Next , we investigated whether secretion of HrpB2 is crucial for protein function . For this , we analyzed N-terminal HrpB2 deletion derivatives . Surprisingly , deletion of the N-terminal 10 amino acids of HrpB2 did not abolish its secretion in wild-type and hpaC deletion mutant strains ( data not shown ) . By contrast , secretion of a HrpB2 deletion derivative lacking amino acids 10 to 25 was severely reduced in strain 85*ΔhpaC , suggesting that amino acids 10 to 25 harbour at least part of the secretion signal ( Fig . 2D ) . Notably , HrpB2Δ10–25 did not complement the hrpB2 mutant phenotype with respect to disease symptom formation in susceptible and HR induction in resistant pepper plants ( Fig . 1A ) . This was not due to the presence of the C-terminal c-Myc epitope since complementation studies were performed with untagged HrpB2 and derivatives . Immunoblot analysis of bacterial total cell extracts revealed that HrpB2Δ10–25 was stably synthesized in strain 85*ΔhrpB2 ( Fig . 2D ) . We therefore conclude that amino acids 10 to 25 are crucial for efficient secretion of HrpB2 and that HrpB2 secretion is presumably required for protein function . To investigate whether oversecretion of HrpB2 in the hpaC deletion mutant was due to increased hrpB2 transcript levels , we performed reverse transcriptase ( RT ) -PCR analysis of strains 85* and 85*ΔhpaC grown under secretion-permissive conditions . Fig . 3A shows that hrpB2 transcript levels were comparable in both strains , suggesting that deletion of hpaC did not affect the transcriptional regulation of hrpB2 . We therefore studied whether there is an interaction between HrpB2 and HpaC proteins using glutathione S-transferase ( GST ) pull-down assays . For this , GST and a GST-HpaC fusion protein were synthesized in Escherichia coli , immobilized on glutathione sepharose matrix and incubated with an E . coli lysate containing HrpB2-c-Myc . Bound proteins were eluted from the matrix and analyzed by immunoblotting using c-Myc epitope- and GST-specific antibodies . HrpB2-c-Myc specifically eluted with GST-HpaC but not with GST alone , indicating that HrpB2 interacts with HpaC ( Fig . 3B ) . Similar results were obtained with an N-terminal HrpB2 deletion derivative that lacks the first 26 amino acids and thus at least part of the T3S signal ( Fig . 3B; see above ) . The interaction between HpaC and HrpB2 is reminiscent of our previous finding that HpaC binds to different T3S substrates including translocon and effector proteins [31] . We did not observe an interaction between HrpB2 and the global T3S chaperone HpaB ( Fig . 3C ) , which is in line with the fact that HpaB does not control HrpB2 secretion ( see above ) . In animal pathogenic bacteria T3S substrate recognition is mediated by members of the conserved FlhB/YscU family of inner membrane proteins [37]–[39] . YscU , FlhB and their homologs contain four predicted transmembrane domains and a C-terminal cytoplasmic protein region that is cleaved between the asparagine and proline residues of the conserved NPTH motif [39]–[43] . To investigate a possible cleavage of the YscU/FlhB homolog HrcU from X . campestris pv . vesicatoria , we synthesized a C-terminally c-Myc epitope-tagged HrcU derivative in both E . coli and X . campestris pv . vesicatoria and analyzed protein extracts by immunoblotting using a c-Myc-specific antibody . We detected two proteins of approximately 50 kDa and 20 kDa in E . coli and X . campestris pv . vesicatoria extracts irrespective of the growth medium ( Fig . 4A ) . Both proteins presumably correspond to full-length HrcU ( 39 kDa+5 kDa epitope tag ) and the predicted C-terminal cleavage product ( 10 kDa+5 kDa epitope tag ) . The HrcU proteins migrate slower than predicted , which was previously also reported for other T3S system-associated proteins from X . campestris pv . vesicatoria [27] , [29] . Because yeast two-hybrid-based interaction studies of proteins from Xanthomonas axonopodis pv . citri suggested an interaction between HrpB2 and the C-terminal domain of HrcU [44] , we performed GST pull-down assays with HrpB2 and HrcU from X . campestris pv . vesicatoria . For this , we generated expression constructs encoding GST-HrcU , GST-HrcU-c-Myc and GST-HrcU255–357 , the latter of corresponds to the C-terminal cytoplasmic domain of HrcU . To test for proteolytic cleavage , GST-HrcU and GST-HrcU-c-Myc were analyzed by immunoblotting of E . coli protein extracts , using GST- and c-Myc-specific antibodies . Both proteins and several degradation products were visualized by a GST-specific antibody ( Fig . 4B ) . Furthermore , GST-HrcU-c-Myc and a smaller protein of approximately 20 kDa were also detected by a c-Myc specific antibody . The smaller protein presumably corresponds to the C-terminal cleavage product of HrcU ( see Fig . 4A ) , indicating that GST-HrcU fusions are proteolytically cleaved ( Fig . 4B ) . For protein-protein interaction studies , GST-HrcU and GST-HrcU255–357 ( HrcU C-terminal domain ) , immobilized on glutathione sepharose , were incubated with HrpB2-c-Myc . HrpB2-c-Myc eluted with GST-HrcU and GST-HrcU255–357 , but not with GST alone , suggesting that HrpB2 interacts with the C-terminal domain of HrcU ( Fig . 5A and B ) . HrcU homologs from animal pathogenic bacteria are involved in the T3S substrate specificity switch [37] , [39] . We therefore tested a possible interaction between HrcU and HpaC , which presumably acts as a T3S4 protein ( see also below ) . When GST-HrcU was immobilized on glutathione sepharose and incubated with HpaC-c-Myc , we did not detect HpaC-c-Myc in the eluate ( Fig . 5C ) . By contrast , HpaC-c-Myc coeluted with GST-HrcU255–357 , suggesting that it interacts with the C-terminal domain of HrcU but not with the full-length protein ( Fig . 5D ) . Since GST-HrcU is proteolytically cleaved ( see Fig . 4B ) , we assume that the protein is correctly folded . Our data therefore suggest that the interaction between HpaC and HrcU depends on a certain conformation of the HrcU C terminus that is altered in the context of the full-length protein . Next , we investigated whether the C-terminal domain of HrcU also interacts with other secreted proteins , e . g . , the putative translocon protein XopA , the pilus protein HrpE and the effector protein XopF1 . For this , GST , GST-HpaC , GST-XopA and GST-HrpE were immobilized on glutathione sepharose and incubated with HrcU-c-Myc . Fig . 6A shows that the C-terminal cleavage product of HrcU ( see above ) was detected in the eluate of GST-HpaC but not of GST-XopA or GST-HrpE . This suggests that the C-terminal domain of HrcU interacts with HpaC but not with HrpE and XopA . We did not detect full-length HrcU-c-Myc in the eluate of GST-HpaC ( Fig . 6A ) , which confirms our previous observation that HpaC specifically interacts with the C-terminal domain of HrcU but not with the full-length protein ( see Fig . 5C and D ) . To investigate a possible interaction between HrcU and the effector protein XopF1 , we expressed XopF1 as a C-terminally c-Myc epitope-tagged derivative because a GST-XopF1 fusion protein was unstable in E . coli . XopF1-c-Myc was incubated with GST-HrcU , GST-HrcU255–357 and GST-HpaB , which was used as a positive control for the interaction assay . GST-HpaB was previously shown to interact with XopF1 [31] . As expected , XopF1-c-Myc was detected in the eluate of GST-HpaB but did not coelute with GST-HrcU and GST-HrcU255–357 ( Fig . 6B ) . Taken together , our results suggest that the C-terminal domain of HrcU does not interact with the T3S substrates XopA , HrpE and XopF1 . This is in contrast to the C-terminal region of the flagellar HrcU homolog FlhB , which interacts with several secreted proteins and is therefore presumably involved in substrate recognition [38] . The finding that HpaC is involved in control of T3S substrate specificity and interacts with the C-terminal domain of HrcU suggests that it acts similarly to T3S4 proteins that were identified in translocation-associated and flagellar T3S systems from animal pathogenic bacteria . Despite limited sequence conservation , known T3S4 proteins harbour a structurally conserved T3S4 domain in the C terminus , which is responsible for the substrate specificity switch [15] , [45] . PSI-BLAST searches and hydrophobic cluster analysis showed that the T3S4 domain is not only present in proteins from animal pathogenic bacteria but also shares weak sequence similarity with the C terminus of HpaP from Ralstonia solanacearum [15] . HpaP is 27% sequence-identical to HpaC . A pairwise sequence alignment of HpaP and HpaC revealed that most conserved amino acids in the predicted T3S4 domain of HpaP are also present in HpaC or are substituted by amino acids with similar chemical properties ( Fig . S1 ) . To investigate whether the predicted T3S4 domain of HpaC participates in the interaction with the C terminus of HrcU , we performed GST pull-down assays with C-terminal HpaC deletion derivatives , which are shown in Fig . 7A . HpaC1–182-c-Myc , which is deleted in the C-terminal 30 amino acids and thus lacks part of the predicted T3S4 domain , coeluted with GST-HrcU255–357 , but not with GST alone ( Fig . 7B ) . However , when compared to the full-length HpaC protein , which has a strong affinity for HrcU255–357 , the interaction between HpaC1–182-c-Myc and GST-HrcU255–357 was significantly reduced ( Fig . 7B ) . By contrast , binding of HpaC1–182-c-Myc to other known HpaC interaction partners such as HpaB , XopF1 , XopA , HrcV and also the HpaC self-interaction was not affected ( Fig . 7B and C ) [31] . Next , we analyzed a HpaC deletion derivative , HpaC1–118-c-Myc , which lacks the C-terminal 94 amino acids and thus the complete T3S4 domain . The fact that HpaC1–118-c-Myc was not detectable in the eluate of GST-HrcU255–357 suggests that the predicted T3S4 domain of HpaC is important for the interaction with the C terminus of HrcU ( Fig . 7D ) . To address whether the predicted T3S4 domain of HpaC is also important for protein function , we expressed HpaC and deletion derivatives in the hpaC deletion mutant . Fig . 8A shows that both HpaC1–118-c-Myc and HpaC1–182-c-Myc failed to complement the hpaC mutant phenotype with respect to ( i ) disease symptom formation and the HR induction in the plant , and ( ii ) oversecretion of HrpB2 in vitro ( Fig . 8A and B ) . Furthermore , HpaC1–118-c-Myc and HpaC1–182-c-Myc did not restore the deficiency in HrpF secretion in strain 85*ΔhpaC ( Fig . 8B ) . We therefore speculate that the T3S4 domain of HpaC and thus the interaction with the C-terminal domain of HrcU is essential for the HpaC-dependent substrate specificity switch .
In this study , we analyzed the pathogenicity factors HrpB2 and HpaC from X . campestris pv . vesicatoria . We discovered that HrpB2 is not only crucial for secretion of effectors but also of extracellular components of the secretion apparatus , i . e . , the putative translocon proteins XopA and HrpF and the pilus protein HrpE . Since HrpB2 is itself secreted by the T3S system , it is presumably one of the first substrates that travels the secretion apparatus [25] . The analysis of N-terminal HrpB2 deletion derivatives revealed that the secretion signal of HrpB2 is located between amino acids 10 to 25 and is crucial for protein function . It is therefore possible that HrpB2 is an extracellular component of the secretion apparatus that promotes pilus assembly . However , HrpB2 is probably not a major pilus subunit since only low amounts of HrpB2 are secreted by the T3S system . Notably , the pilus protein HrpE is required for HrpB2 secretion and vice versa , suggesting that HrpB2 is not part of an extracellular needle-like structure below the pilus . An analogous finding was recently reported for the symbiotic bacterium Rhizobium strain NGR234 . Pilus assembly and T3S in strain NGR234 depends on the secreted protein NopB that presumably associates with NopA , which is the major pilus subunit [46] , [47] . The second important finding is that secretion of HrpB2 from X . campestris pv . vesicatoria is suppressed by the export control protein HpaC , which promotes secretion of translocon and effector proteins [31] . Proteins that differentially regulate secretion of different T3S substrates were described for flagellar or translocation-associated T3S systems and include , e . g . , the flagellar chaperone FliS and T3S4 proteins from animal pathogenic bacteria [12]–[14] , [48] , [49] . We speculate that HpaC acts similarly to T3S4 proteins and alters the specificity of the secretion apparatus from early ( HrpB2 ) to later ( translocon and effector proteins ) T3S substrates . We believe that this substrate specificity switch takes place at the protein level because HpaC interacts with HrpB2 and with the C-terminal domain of HrcU , which belongs to the FlhB/YscU family of inner membrane proteins [3] , [12] , [13] . This is in agreement with our previous finding that HpaC binds to different T3S substrates including translocon and effector proteins and also interacts with conserved inner membrane components of the T3S system such as HrcV [31] . It was therefore proposed that HpaC acts as a linker between secreted proteins and the secretion apparatus [31] . However , HpaC is dispensable for secretion of HrpB2 . Targeting of HrpB2 to the secretion apparatus is presumably mediated by the C-terminal domain of HrcU , which interacts with both HrpB2 and HpaC . The latter interaction presumably depends on a certain conformation of the HrcU C terminus since we did not detect binding of HpaC to full-length HrcU . The analysis of HpaC deletion derivatives revealed that the HrcU-binding site is located in the C terminus of HpaC , which contains the putative T3S4 domain [15] and is required for protein function . This observation suggests that the interaction between HpaC and the C-terminal domain of HrcU is required for HpaC-mediated suppression of HrpB2 secretion . Our data are reminiscent of the finding that the T3S4 protein FliK from Salmonella spp . interacts with the C-terminal domain of the HrcU homolog FlhB [37] , [38] . It was proposed that binding of FliK induces a conformational change in the C–terminal cytoplasmic domain of FlhB and thus alters the substrate specificity of the flagellar T3S system from secretion of hook components to filament proteins [12] , [50] . Since the C-terminal domain of FlhB interacts with several secreted proteins , it presumably serves as a docking point for T3S substrates [37] , [38] . This clearly differs from the FlhB homolog HrcU from X . campestris pv . vesicatoria since the C-terminal domain of HrcU does not interact with translocon and effector proteins that were tested in this study . It is conceivable that T3S substrate binding in X . campestris pv . vesicatoria is mediated by other conserved inner membrane components of the T3S system such as HrcV or the putative ATPase HrcN [3] . The precise mechanism underlying the HpaC/HrcU-mediated substrate specificity switch in X . campestris pv . vesicatoria remains to be determined . We speculate that after activation of the T3S system , binding of HpaC to the C-terminal domain of HrcU inhibits the interaction between this domain and HrpB2 and thus blocks secretion of HrpB2 . Preliminary GST pull-down assays revealed that HrpB2-c-Myc coelutes with GST-HrcU255–357 irrespective of the presence of HpaC , suggesting that all three proteins can form a complex . It still remains to be investigated whether both proteins simultaneously bind to the C-terminal domain of HrcU or whether they are both present in the eluate because they interact with each other . Taken together , our data suggest that plant and animal pathogenic bacteria share similar mechanisms to switch the substrate specificity of the T3S system but that they differ in the components that recognize T3S substrates . Another important difference between plant and animal pathogenic bacteria concerns the length control of extracellular structures associated with the membrane-spanning secretion apparatus . In translocation-associated and flagellar T3S systems from animal pathogenic bacteria , the substrate specificity switch is coupled to length control of needle and hook structures . In the flagellar T3S system from Salmonella spp . , for instance , FliK activates secretion of filament proteins after hook formation . Deletion of fliK leads to elongated hook structures , suggesting that FliK is required for hook length control [51]–[53] . Similarly to FliK , the T3S4 protein YscP from Yersinia spp . determines needle length in the translocation-associated T3S system [12] . Since YscP is itself secreted it was proposed that the N terminus of YscP anchors to the tip of the growing needle while the C terminus of the protein remains attached to the secretion apparatus and activates the substrate specificity switch [54] . According to this model , T3S4 proteins act as molecular rulers that are coupled to a substrate specificity switch [12] , [15] . The molecular ruler model was challenged by the finding that the T3S4 protein InvJ from Salmonella typhimurium is required for formation of the inner rod of the T3S apparatus . It was suggested that formation of the inner rod triggers a conformational change in the secretion apparatus that leads to the substrate specificity switch [55] . This model is supported by the recent finding that the T3S4 protein YscP from Yersinia controls secretion of the predicted inner rod protein YscI [56] . Wood et al . identified YscI point mutants that allow effector secretion in the absence of a detectable needle structure , suggesting that the needle is not required for the substrate specificity switch . The future challenge is to investigate the molecular mechanisms underlying the HpaC-mediated T3S substrate specificity switch in X . campestris pv . vesicatoria . Since HpaC is not secreted by the T3S system , it presumably does not act as a molecular ruler protein [31] . Electron microscopy studies have suggested that pilus length is not controlled by HpaC [26] . Furthermore , it should be emphasized that secretion of the Hrp pilus subunit HrpE is not affected in hpaC mutants [31] . In contrast to the relatively short ( approximately 50 nm ) T3S needle from animal pathogenic bacteria , the Hrp pilus from plant pathogens can reach a length of up to 2 µm that cannot be bridged by a single proteinaceous molecular ruler . We therefore speculate that HpaC acts as a T3S4 protein that is not involved in length control of extracellular structures of the T3S system . This hypothesis is supported by the fact that secretion-deficient derivatives of the T3S4 proteins YscP and FliK are still active , indicating that length control and substrate specificity switch functions can be uncoupled [54] , [57] .
Bacterial strains and plasmids used in this study are listed in Table 1 . E . coli cells were cultivated at 37°C in lysogeny broth ( LB ) . X . campestris pv . vesicatoria strains were grown at 30°C in NYG medium [58] or in minimal medium A [59] supplemented with sucrose ( 10 mM ) and casamino acids ( 0 . 3% ) . Plasmids were introduced into E . coli by electroporation and into X . campestris pv . vesicatoria by conjugation , using pRK2013 as a helper plasmid in triparental matings [60] . For the generation of strain 85*ΔhpaCΔhrpE , pOK-hrpEΔ9-93 , which is a derivative of the suicide plasmid pOK1 ( see Table 1 ) , was introduced into the genome of X . campestris pv . vesicatoria strain 85*ΔhpaC by conjugation . Double cross-overs resulted in deletion mutants that were selected as described [28] . Antibiotics were added to the media at the following final concentrations: ampicillin , 100 µg/ml; kanamycin , 25 µg/ml; rifampicin , 100 µg/ml; spectinomycin , 100 µg/ml; tetracycline , 10 µg/ml . The near-isogenic pepper cultivars Early Cal Wonder ( ECW ) , ECW-10R and ECW-30R [61] were grown and inoculated with X . campestris pv . vesicatoria as described previously [22] . Bacteria were hand-infiltrated into the intercellular spaces of leaves at concentrations of 2×108 cfu/ml in 1 mM MgCl2 if not stated otherwise . The appearance of disease symptoms and the HR were scored over a period of three to five days after inoculation . For better visualization of the HR , leaves were bleached in 70% ethanol . For RT-PCR analysis , bacteria were grown in secretion medium . RNA extraction and cDNA synthesis were performed as described [62] and hrpB2 transcripts were amplified by PCR . To exclude that RNA preparations contained genomic DNA , total RNA was used as a template in a control PCR using hrpB2-specific primers . The lack of detectable hrpB2 amounts suggested that the RNA preparations were DNA-free ( data not shown ) . Sequences of primers used in this study are available upon request . For the generation of hrpB2 expression constructs , hrpB2 and N-terminal deletion derivatives were amplified by PCR from X . campestris pv . vesicatoria strain 85-10 and cloned into the EcoRI and HindIII sites of pDSK602 . To create c-Myc epitope-tagged derivatives of HrpB2 , hrpB2 and truncated gene fragments were subcloned into the EcoRI/SacI sites of pC3003 , in frame with a triple-c-myc epitope-encoding sequence , and the resulting inserts were introduced into the EcoRI/HindIII sites of pDSK602 . For the generation of expression constructs encoding HrpB2Δ10–25 and HrpB2Δ10–25-c-Myc , full-length hrpB2 cloned into pUC119 or pC3003 was used as template for a PCR . PCR products were religated and the respective inserts were cloned into pDSK602 . For the generation of GST fusion proteins , full-length hrcU and a fragment encoding amino acids 255 to 357 , respectively , were amplified by PCR and cloned into the EcoRI/XhoI sites of pGEX-2TKM , respectively , downstream and in frame with the GST-encoding sequence . To construct a C-terminally c-Myc epitope-tagged HrcU derivative , hrcU was amplified by PCR , inserted into pENTR/D-TOPO and recombined into pDGW4M using Gateway technology ( Invitrogen , Carlsbad , Calif . ) . pDGW4M is a Gateway-compatible derivative of pDSK602 containing attR sites , chloramphenicol resistance and ccdB genes and the 4× c-Myc-encoding sequence of vector pGWB16 inserted into the EcoRI/HindIII sites . To generate a GST-HrcU-c-Myc expression construct , hrcU was amplified by PCR , subcloned by SacI and partial EcoRI digest in pC3003 , which contains a triple c-myc-encoding sequence and introduced into the EcoRI/SacI sites of pGEX-6P-1 , in frame with a gst-encoding sequence . In vitro secretion assays were performed as described [27] . Total cell extracts and culture supernatants were analyzed by SDS-PAGE and immunoblotting . We used polyclonal antibodies specific for HrpF [27] , XopA [29] , AvrBs3 [63] and HrpB2 [25] , respectively , and monoclonal anti-c-Myc and anti-GST antibodies ( Amersham Pharmacia Biotech , Freiburg , Germany ) . Horseradish peroxidase-labelled anti-rabbit , anti-mouse and anti-goat antibodies ( Amersham Pharmacia Biotech ) were used as secondary antibodies . Antibody reactions were visualized by enhanced chemiluminescence ( Amersham Pharmacia Biotech ) . To ensure that no bacterial lysis had occurred , blots were routinely reacted with an antibody specific for the intracellular protein HrcN ( data not shown ) [25] . GST pull-down assays were performed as described previously [31] . Briefly , GST and GST fusion proteins were expressed in E . coli and bacterial cells from 50 ml cultures were broken with a French press . GST and GST fusions were immobilized on glutathione sepharose and incubated with a c-Myc epitope-tagged derivative of the putative interaction partner . Bound proteins were eluted with 10 mM reduced glutathione . 5 µl total protein lysates and 20 µl eluted proteins were analyzed by SDS-PAGE and immunoblotting . For the generation of GST-HrpE , hrpE was amplified by PCR and cloned into the EcoRI/XhoI sites of pGEX-2TKM . The same blot was always incubated with an anti-c-Myc and an anti-GST antibody , respectively . | The Gram-negative plant pathogenic bacterium Xanthomonas campestris pv . vesicatoria is the causal agent of bacterial spot disease in pepper and tomato . Pathogenicity of X . campestris pv . vesicatoria depends on a type III protein secretion ( T3S ) system that injects bacterial effector proteins directly into the host cell cytosol . The T3S system is a highly complex nanomachine that spans both bacterial membranes and is associated with an extracellular pilus and a translocon that inserts into the host cell membrane . Given the architecture of the secretion apparatus , it is conceivable that pilus formation precedes effector protein secretion . The pilus presumably consists of two components , i . e . , the major pilus subunit HrpE and HrpB2 , which is required for pilus assembly . Secretion of HrpB2 is suppressed by HpaC that switches substrate specificity of the T3S system from secretion of HrpB2 to secretion of translocon and effector proteins . The substrate specificity switch depends on the cytoplasmic domain of HrcU , which is a conserved inner membrane protein of the T3S apparatus that interacts with HrpB2 and HpaC . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"microbiology/plant-biotic",
"interactions"
] | 2008 | HpaC Controls Substrate Specificity of the Xanthomonas Type III Secretion System |
Cyclic nucleotides ( cAMP and cGMP ) regulate multiple intracellular processes and are thus of a great general interest for molecular and structural biologists . To study the allosteric mechanism of different cyclic nucleotide binding ( CNB ) domains , we compared cAMP-bound and cAMP-free structures ( PKA , Epac , and two ionic channels ) using a new bioinformatics method: local spatial pattern alignment . Our analysis highlights four major conserved structural motifs: 1 ) the phosphate binding cassette ( PBC ) , which binds the cAMP ribose-phosphate , 2 ) the “hinge , ” a flexible helix , which contacts the PBC , 3 ) the β2 , 3 loop , which provides precise positioning of an invariant arginine from the PBC , and 4 ) a conserved structural element consisting of an N-terminal helix , an eight residue loop and the A-helix ( N3A-motif ) . The PBC and the hinge were included in the previously reported allosteric model , whereas the definition of the β2 , 3 loop and the N3A-motif as conserved elements is novel . The N3A-motif is found in all cis-regulated CNB domains , and we present a model for an allosteric mechanism in these domains . Catabolite gene activator protein ( CAP ) represents a trans-regulated CNB domain family: it does not contain the N3A-motif , and its long range allosteric interactions are substantially different from the cis-regulated CNB domains .
Cyclic adenosine monophosphate ( cAMP ) is an important second messenger , which regulates a large variety of cellular processes , including metabolism , cell shape transformation , gene transcription , photoreception and chemosensation [1]–[5] . All cAMP-binding proteins in both pro- and eukaryotes share a small module – the cyclic nucleotide binding domain ( CNB domain ) , which is typically fused to another domain . The CNB domain contains a contiguous β-subdomain and a non-contiguous α-subdomain ( Figure 1 ) . The former is a relatively rigid eight-stranded β-sandwich , which accommodates the cyclic nucleotide molecule . The flexible helical α-subdomain can accept substantially different configurations , which translates the allosteric signal [6] . Recent structure studies of cAMP-dependent protein kinase ( PKA ) demonstrated , that CNB domains toggle between two stable conformations: bound to cAMP ( so called B-form [7] ) , or to catalytic subunit of PKA ( H-form ) [8] , [9] . The intermediate , non bound form ( apo-form ) is characterized by high backbone flexibility [10]–[12] and is apparently represented by a dynamic ensemble of multiple configurations . The β-subdomain contains a highly conserved element , the Phosphate Binding Cassette ( PBC ) , which is 14-residues long and contains a short flexible α-helix ( B'-helix ) . Ribose-phosphate of cAMP , protected by the β-sandwich from the outside solution , forms six strong hydrogen bonds to the PBC . Due to these interactions , the B'-helix moves towards the cAMP molecule and adopts a compact conformation ( Figure 1 ) . Such movement causes a substantial rearrangement of the α-subdomain both in its N- and C-terminal parts . The latter contains the so called “hinge” [13] , which consists of two consecutive α-helices ( B and C ) . These helices are remarkably flexible and due to a strong connection to the B'-helix perform a swing-like motion: moving towards cAMP in the B-form ( Figure 1A ) . The N-terminal part of the α-subdomain , which typically consists of two ( short ) helices and was called the “N-terminal helical bundle” [14] , moves in an opposite way; in the B-form it moves away from the PBC , facilitating the hinge closure . In the H-form the N-terminal helices move towards the cAMP and make a contact to the B'-helix , filling the void space that results from the hinge opening ( Figure 1B ) . The other part of the cAMP molecule , the adenine ring , acts as a hydrophobic moiety , which stacks against a “capping residue” in all known CNB domain structures [15] ( Figure 1A ) . Mutation studies have established the importance of this contact for stabilization of the B-form and cooperative cAMP-induced activation of the PKA holoenzyme [8] . A recent review summarizes this information into a general model for the CNB domain allosteric mechanism for PKA , Rap guanine nucleotide-exchange factor ( Epac ) and hyperpolarization-activated cyclic-nucleotide-modulated channel ( HCN ) [14] . Although this model is in a good correspondence with much experimental data , two important issues remained unclear . First , as the authors mentioned , the N-terminal helical bundle is replaced by a single helix in the catabolite gene activator protein ( CAP ) . This raises a question about the role of the helical bundle and its functional and structural conservation . Is it a universal part of the CNB domain or it is a part of protein-protein interface between the CNB domain and the host protein ? The second problem is related to the loop located between β2 and β3 strands . A series of publications demonstrated , that it is an important element of allosteric mechanism in the PKA RIα∶A-domain [16]–[18] , but it was not included in the model described by Rehmann et al [14] and was not considered to be a universal element . To elucidate the cAMP induced allosteric mechanism that is conserved in different CNB domains , we used a recently developed method for protein structure comparison: Local Spatial Patterns ( LSP ) alignment that is capable of detecting similar patterns made up by amino acid residues in space . It is fast and does not require preliminary sequence or structural alignment of the compared proteins . Earlier we used it for comparison of protein surfaces of several CNB domains in the B-form and detected a conserved set of hydrophobic residues protecting the cAMP ribose-phosphate [15] . Here we considered both water accessible and buried residues of both B- and H-forms of four different CNB domains: PKA , HCN , Epac and bacterial cyclic nucleotide modulated potassium channel ( MloK1 ) [19] . The recently reported structures of two PKA holoenzymes [8] , [20] have allowed us for the first time to analyze both conformational states of multiple CNB domains . Our analysis has shown that there are four elements conserved in all known CNB domains , with the exception of the CAP: the PBC , the hinge , the β2 , 3-loop and the “N3A-motif” . The latter consists of the A-helix , a preceding eight residue loop and a short N-terminal helix . The loop contains a set of 310-turns , and is termed “the 310-loop” . Based on these results , we propose a general model for the allosteric mechanism in CNB domains , which we called cis-regulated domains . In CAP the N3A-motif is reduced to a single A-helix . The difference between CAP and other CNB domains is discussed .
The LSP alignment is a new method to compare protein molecules . It is based on a graph-theoretical representation of protein structure , and the result of this alignment is a pair of isomorphic graphs . Vertices of the graphs correspond to the residues which form similar spatial patterns in both proteins . Each vertex/residue is connected to the rest of the graph by several edges . They indicate the residue neighbors whose positions and orientation in space are conserved with respect to this residue . As we have shown earlier [15] , [21] , functionally important residues of protein kinases have numerous connections on the similarity graphs . In the previous works we considered only surface exposed residues . Here we analyze all residues . This allows us to recognize conserved motifs that are buried in the protein core . We define a term “involvement score” ( IS ) of a particular residue , which is equal to the number of edges for the corresponding vertex on the graph provided by the LSP alignment procedure . It reflects the extent of participation of this residue in formation of invariant spatial patterns and corresponds to AA and AI scores used in the previous work [21] , where we compared active and inactive protein kinases . Earlier we used the LSP alignment for comparison of different proteins having similar functions . In this work we present an alternative way of using the LSP alignment program . Our purpose was to quantify cAMP-induced structural rearrangements in different CNB domains . This was made by aligning two different conformations of the same protein . As the IS reflects only local structural similarities ( in our case within 10 Å range between Cα-atoms ) any large scale rearrangements in the protein do not change the score significantly . Therefore , residues which form rigid structures inside the protein and maintain their relative positions will have a high level of IS . In contrast , those residues located in points of protein flexibility will have low IS , reflecting the loss of similarity between the two protein structures . One can speculate that the residues with the lowest IS can play an important role in the allosteric mechanism , as such elements like “hinges” or “switches” have to accept two distinctive “on” and “off” conformations . Figure 2 presents the results of LSP alignment of PKA ( RIα ) H- and B-forms . As expected , residues from the rigid β-sandwich had the highest IS values , reflecting the rigidity of the β subdomain . In contrast , four regions showed a significant decrease in IS: 1 ) B'-helix of the PBC; 2 ) the hinge region; 3 ) the β2 , 3-loop and 4 ) N-terminal part of the α-subdomain . These results are in good correspondence with the NMR studies of the RIα A-domain [16] . Our analyses show that both A- and B-domains have similar IS profiles , although the drop in the β2 , 3-loop in B-domain was less prominent . Similar results were obtained for the A-domain of PKA ( RIIβ ) and the potassium channel MloK1 ( Figure S1 ) . The decrease of IS in the β2 , 3-loop in the potassium channel CNB domain was not as striking as in A-domains of PKA ( both RIα and RIIβ ) , but similar to the B-domain of RIα . This is in a good agreement with the earlier observation that B-domains of PKA are more similar to the rest of CNB domains , than the A-domains [15] . The N-terminal helical structure which earlier was called an “N-terminal helical bundle” has not been considered as a conserved element [14] . It is not a part of the current CNB domain nomenclature for two reasons: first – it is not present in CAP , and second – in the different CNB domains it forms slightly different secondary structures . For example in the B-form of RIα , RIIβ , MloK1 and Epac2 it contains a short 310-helix . In the H-form of RIα∶A it contains a set of 3-turns , which do not form the classical 310-helix and is considered to be a loop , while the B-domain retains its 310-helix configuration . In H-form of PKA RIIα∶A and RIIβ∶A this element qualifies as an α-helix , but in the B-form of RIIβ∶A and RIIβ∶B it is a 310-helix . However , a close look at the middle part of the bundle shows that the geometry of its backbone in different CNB domains ( both H- and B-forms ) , is rather conserved ( Figure 3A ) . Moreover , this element , which we will define as a “310-loop” , has a distinctive pattern of phi/psi angles: Figure 3B shows a sequence alignment of different N-terminal helical bundles . One can see that A-helix of the presented CNB domains is preceded by an eight residue long loop . Its first conserved feature is that both ends of the loop contain residues with negative chirality ( a characteristic of β-strands ) : one at the N-terminus and two at the C-terminus . There is also a large conserved hydrophobic residue ( phenylalanine or leucine ) in the middle of the loop ( F136 Figure 3C ) , which plays a central role in the hydrophobic cluster formed by the A-helix and the preceding α-helix , which did not have an established name . As in the PKA-RIα A-domain , it was called “αXn-helix” [16] or “X∶N-helix” [6] , here we call it “N-helix” , and the combination of the N-helix , the 310-loop and A-helix structure – the N3A-motif . The characteristic feature of this motif is the presence of multiple X–X pairs in its sequence ( where X represents a hydrophobic residue or a residue with large hydrophobic segment such as arginine or asparagine ) ( Figure 3B ) . Such residues are closely positioned on one side of the helix and provide a secure connection between the N3A-motif elements ( Figure 3C ) , a feature similar to the tetratricopeptide repeat [22] or the leucine-zipper [23] motifs . Analysis of the recently discovered holoenzyme structures of PKA shows that the 310-loop residues positioned between the X–X pairs are usually involved in important interactions . For example , in RIa∶A-domain V134 , L135 and H138 stack against a large hydrophobic cluster formed by the C-subunit and the PBC ( Figure 3D ) . In the RIa∶B-domain S252 forms five hydrogen bonds to the PBC , the C-terminal helical structure of the domain and the β-sandwich . The suggested conservation of the N3A-motif in different CNB domains raises a question about the definition of A and B domains in PKA-R . Until now , the beginning of the B domain was associated with the first residue in it's a-helix ( e . g . W260 in RIα or V280 in RIIβ ) . Here we suggest a new boundary between the A and B domains , which will reflect the conservation of the N3A-motif . It is known that the RIα- ( 94-244 ) construct retains its functionality and is capable of both binding to cAMP and regulating PKA [24] . In addition , the C-helix of cAMP-bound RIα has a kink between Y244 and E245 . An identical kink exists in the cAMP-bound RIIβ ( between Y265 and E266 ) . It seems logical to suggest that this kink indicates the border between the A and B domains , therefore defining E245 as the beginning of N3A-motif for the B-domain of RIα . Such a definition supports the observation made earlier by Huang and Taylor that RIα “residues 245–260 at the end of cAMP binding domain A are structurally more a part of domain B than domain A” [24] . After the new definition of CNB domains , we used the LSP alignment to detect residues involved in formation of conserved spatial patterns as we did previously for PKA-C [21] . The A-domain of the RIα holoenzyme was taken as a reference structure . It was compared to five cAMP-bound CNB domains: RIα∶B , RIIβ∶A , RIIβ∶B , HCN and MloK1 . To detect the regions , which respond to the cAMP presence , we also compared our reference structure to five cAMP-free CNB domains: six domains of PKA-R taken from the corresponding holoenzyme complexes: RIα∶B , RIIα∶A , RIIα∶B; and two apo-structures: Epac and MloK1 . Involvement scores were accumulated and presented in Figure 4 . The highest involvement scores were detected in the PBC and neighboring β6 and β7 strands . This area also had the largest cAMP-induced changes of IS , reflecting the leading role of the PBC in the allosteric mechanism . Reduced scores in the middle of PBC agree with the sequence variability profiles obtained earlier for PKA-R [25] . The second major region with highly scored residues was in the strands β2 , β3 and the loop between them . The result was rather unexpected as , until recently , this area was not considered as an important part of CNB domains . The scores , in general , did not depend on the presence of cAMP , which indicates overall conservation of the loop geometry . This suggests , that the β2 , 3-loop is an important element not only in PKA-R , as it was pointed by Das et . al [16] , [26] , but in all CNB domains . Figure 5 shows a comparison between β2 , 3-loops of different CNB domains . As was detected by the DSSP program [27] , all loops have the same pattern of their main chain chirality , indicating a high level of their geometry conservation . The most distinctive common feature for all of them is a 3-turn between residues #2 ( D164 ) and #5 ( D167 ) ( Figure 6A ) . It contains an invariant glycine residue #4 ( G166 ) which makes a conserved hydrogen bond to the PBC-arginine carbonyl . The reason for strict conservation of the glycine is evident as its dihedral angles are ruled out for any other type of residue ( φ = 89 . 4° and ψ = −26 . 8° for RIα∶A ) . Another conserved hydrogen bond is formed between the PBC-arginine amide and carbonyl of the residue #5 . The third polar contact , which can be found in all CNB structures , is formed between the PBC-arginine guanidinium group and the carbonyl of residue #9 ( N171 ) . The hydrogen bond between the arginine and the side chain of residue #8 ( D170 ) , which was found to be important for RIα∶A , is not conserved: it can be seen only in the A domains of RIα and RIIβ . This residue , however , often binds to the side chain of residue #9 and provides communication between the β2 , 3-loop and the “hinge” as its carbonyl is always bound to the amide group of the first residue in the B-helix . Besides polar interactions with the PBC-arginine , residue #1 ( I163 ) makes a conserved hydrophobic contact to the arginine side chain . This residue is a member of a conserved hydrophobic core formed by the highly scored residues: V162 , I163 , Y173 , F198 and V213 ( Figure 6B ) . The important detail is that another member of that cluster: Y173 ( one of the three highest scores obtained ) – makes a conserved hydrogen bond to the residue #9 ( N171 ) , thus closing the circle around the PBC arginine . Glycine is the predominant residue in position #7 , except for the B domains of RIIα and RIIβ , where it is substituted by alanine . Any increase of the side chain would lead to a steric clash with highly conserved alanine residues from the PBC ( A202 ) , which in its turn also has a hydrophobic contact to the PBC-arginine in the cyclic nucleotide bound configurations . The hinge , which is a well known element of the allosteric mechanism , demonstrated medium levels of IS and a strong dependence on the presence of cAMP . The preceding β8-strand received almost the same level of scores . It contains a set of conserved hydrophobic residues , which face the N3A-motif and provide a secure connection of this element to the β-sandwich . The NMR-study of RIα showed that two residues from the β8-strand ( W222 and I224 ) had a substantial chemical shift , in response to cAMP binding and are a part of the allosteric mechanism . Our results support this conclusion and demonstrate conservation of the hydrophobic interface through different CNB domains . Another highly scored residue positioned between B-helix and β8-strand ( M151 ) is also a part of this interface . As we showed earlier the N3A-motifs of different CNB domains have very similar geometry , and conserved sequence motifs . However , this region , except for the C-terminus of A-helix , demonstrated low levels of IS . This indicates that the residues , conserved in the sequence , do not form a rigid spatial motif . This conclusion is supported by the fact that the N-helix and 310-loop in RIα∶A-domain have an elevated level of hydrogen-deuterium exchange [10] , [16] . Apparently , the N3A-motif is a rather flexible element , which can adopt slightly different conformations accommodating the PBC and the hinge movements . In all comparisons the loop between β4 and β5-strands received zero level of IS ( Figure 4 ) . It is consistent with the fact that this part of the CNB domain is the least conserved in terms of sequence and structure [28] . Our data show that in many CNB domains the N-terminus of the β4–5-loop involved in the conserved anchoring of the β2–3-loop via its #3 residue ( Q165 ) ( Figures 5 and 6A ) . The β4–5-loop spatially comes close to the C-terminus of the CNB domain and in the RIα∶B-domain may provide a docking site for another protein . We deliberately excluded the CAP CNB domain from the current analysis as it does not contain the classic N3A-motif . There are several distinctive features that distinguish the CAP CNB domain from those discussed above . The major difference is that in functionally active CAP the two identical CNB domains form a homodimer with the interface being formed mainly by their C-helices . Regulatory subunits of PKA also contain two CNB domains , but their mutual interaction is rather limited: e . g . A-helix of B-domain contacts cAMP and the hinge from the A-domain . In contrast , in CAP the interaction between the two monomers is the most important allosteric contact between the PBC and the hinge . Figure 7 shows that the major binding partner for each PBC is not the hinge from its own CNB domain but the hinge from the opposite monomer . Such interactions separate the CAP CNB domain from the other CNB domains studied in this work , which we propose to call cis-regulated CNB domains while defining CAP CNB domain as a trans-regulated CNB domain .
The LSP-alignment of H- and B-forms of different CNB domains revealed four conserved structural motifs: the PBC , the Hinge , the N3A-motif and the β2 , 3-loop . These elements were found in all studied cis-regulated CNB domains . The N3A motif is not present in CAP , which represents a trans-regulated CNB domain family . We propose a generalized allosteric mechanism for cis-regulated domains as follows: a ) The PBC is a primary element , which binds sugar-phosphate moiety of cAMP . b ) The β2 , 3-loop regulates the cAMP binding to the PBC via the conserved PBC-arginine . c ) Both the PBC and the β2 , 3-loop communicate with the Hinge , which transfers the allosteric signal further to the N3A motif . d ) The N3A-motif is the most malleable element of a CNB domain as it provides communication to the host protein .
The following structures were used in the current work: PKA∶RIα B-form [32] ( PDBID – 1RGS ) ; PKA∶RIα H-form [8]; PKA∶RIIα H-form [20]; PKA∶RIIβ B-form [13] ( 1CX4 ) ; PKA∶RIIβ H-form ( Brown et al . , unpublished results ) ; HCN B-form [33] ( 1Q43 ) ; MloK1 B-form [19] ( 1VP6 ) ; MloK1 H-form [19] ( 1U12 ) ; Epac2 H-form [34] ( 2BYV ) . LSP-alignment was made by previously reported algorithm for surface matching [21] . All residues ( both water accessible and buried inside protein ) were included in the analysis . For that reason , the water accessibility cut-off was taken equal to zero . Residues were represented by their Cα–Cβ vectors . The maximum distance between Cα atoms was 12 Å . Tolerance for Cα–Cα distance was 0 . 4 Å . Tolerance for Cα–Cβ distance was 0 . 75 Å . Tolerance for the dihedral angle between the vectors was 30° . Residues with the BLOSUM62 score greater than or equal to 1 were considered to be similar . Calculations were made on a personal computer ( Pentium 4; 1 . 8 GHz; 1 Gb RAM ) under Linux OS . Molecular graphics were prepared using PyMOL ( DeLano Scientific , San Carlos , CA ) . | Cyclic nucleotides are small regulatory molecules which transmit signal from receptors positioned on a living cell membrane into the cell interior and regulate multiple biological processes ranging from bacteria to humans . Such regulation occurs through binding of the cyclic nucleotides to the corresponding proteins . All such proteins contain a relatively small domain responsible for the cyclic nucleotide binding . The most important task is to understand how cyclic nucleotide binding ( CNB ) domains translate the signal into a biological response . In this work , we studied changes in different CNB domains induced by the cyclic nucleotides using a new method for comparison of protein structures: local spatial patterns alignment . This novel method compares protein molecules and detects conserved spatial patterns comprised of similar amino acid residues . Moreover , it ranks the detected residues with respect to their functional or structural importance . Our results show that there are at least two different families of CNB domains . The first family has four structural elements which perform the signal translation . Two of these elements were known previously and two are novel . The second family can translate the signal using only three elements , but they have to work in pairs to provide interaction between the functional elements . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"computational",
"biology/macromolecular",
"structure",
"analysis",
"cell",
"biology/cell",
"signaling",
"molecular",
"biology/bioinformatics",
"biochemistry/cell",
"signaling",
"and",
"trafficking",
"structures",
"biochemistry/bioinformatics",
"biochemistry/biomacromolecule-ligand",
... | 2008 | A Generalized Allosteric Mechanism for cis-Regulated Cyclic Nucleotide Binding Domains |
A significant number of environmental microorganisms can cause serious , even fatal , acute and chronic infections in humans . The severity and outcome of each type of infection depends on the expression of specific bacterial phenotypes controlled by complex regulatory networks that sense and respond to the host environment . Although bacterial signals that contribute to a successful acute infection have been identified in a number of pathogens , the signals that mediate the onset and establishment of chronic infections have yet to be discovered . We identified a volatile , low molecular weight molecule , 2-amino acetophenone ( 2-AA ) , produced by the opportunistic human pathogen Pseudomonas aeruginosa that reduces bacterial virulence in vivo in flies and in an acute mouse infection model . 2-AA modulates the activity of the virulence regulator MvfR ( multiple virulence factor regulator ) via a negative feedback loop and it promotes the emergence of P . aeruginosa phenotypes that likely promote chronic lung infections , including accumulation of lasR mutants , long-term survival at stationary phase , and persistence in a Drosophila infection model . We report for the first time the existence of a quorum sensing ( QS ) regulated volatile molecule that induces bistability phenotype by stochastically silencing acute virulence functions in P . aeruginosa . We propose that 2-AA mediates changes in a subpopulation of cells that facilitate the exploitation of dynamic host environments and promote gene expression changes that favor chronic infections .
Bacteria excrete small molecules that act as specific chemical signals to positively regulate specialized processes [1] , including the production of virulence factors important for pathogenic infection , host colonization , and interspecies microbial interactions [2] . Using interconnected multi-layered regulatory networks , such as quorum sensing ( QS ) networks , bacteria sense and respond to external and internal bacterial cell signals as well as environmental cues , thereby adapting to exploit target hosts . Adaptation and coordination of gene expression is particularly important for pathogenic microorganisms that need to colonize changing host environments since their ability to sense and respond to host environmental cues is crucial for their survival . Pseudomonas aeruginosa is an opportunistic human pathogen that causes chronic and acute infections , and is a major agent of morbidity and mortality in cystic fibrosis ( CF ) patients . Establishment of chronic P . aeruginosa respiratory or wound infections requires a complex adaptive process that mediates essential physiological changes allowing bacterial cells to survive and persist in a dynamic host environment . Although insights into chronic infection pathways have been reported [3]–[7] , the specific bacterial signals that may promote the transition and/or adaptation of known pathogens from an acute to a chronic infection remain unknown . Many P . aeruginosa virulence factors associated with acute infections are controlled by QS [8] . This pathogen has an extensively studied complex QS-communication network that facilitates cross-talk between organisms and impacts many P . aeruginosa group-related behaviors including virulence [8] . There are at least three known QS systems in P . aeruginosa: two are dependent on the acyl-homoserine-lactone ( AHL ) QS transcription factors LasR and RhlR [9] , and a third is dependent on the 4-hydroxy-2-alkylquinolines ( HAQs ) LysR-type transcription factor MvfR ( multiple virulence factor regulator ) [10] . MvfR is critical for P . aeruginosa acute infection through its control of genes involved in the production of secreted virulence factors and in iron assimilation [10]–[13] . This regulator controls HAQ signaling and its own activity by positively regulating the expression of genes in the pqsABCDE [14] and phnAB [10] operons . These operons encode enzymes that catalyze the biosynthesis of at least 59 distinct low molecular weight compounds , most of which are structurally related to HAQs . While two of the most abundant HAQs , [4-hydroxy-2-heptylquinoline ( HHQ ) and 3 , 4-dihydroxy-2-heptylquinoline ( Pseudomonas Quinolone Signal-PQS ) ] [14]–[16] , function in vivo as ligands that bind and activate MvfR [14] , [17] , the biological functions of other PqsABCD/PhnAB biosynthetic products remain elusive . In this study , we show that one of these abundant MvfR-regulated non-HAQ low molecular weight molecules , 2-aminoacetophenone ( 2-AA ) , that is used to diagnose P . aeruginosa infections in humans [18] , reduces acute virulence by negatively fine-tuning the transcription and synthesis of the MvfR ligand HHQ , and promotes changes that are critical for pathogen adaptation and important for chronic infection .
To determine the functions of the abundant MvfR-regulated small molecules , we first compared the liquid chromatography/mass spectrometry ( LC/MS ) total ion chromatograms of culture-free supernatants from highly pathogenic wild-type P . aeruginosa ( PA14 ) cells versus those of isogenic mvfR mutant cells . As shown in Figure 1A , the mvfR versus PA14 supernatant lacked HHQ and PQS , as well as three other abundant low molecular weight compounds: the HAQ molecule 4-hydroxy-2 heptylquinoline N-oxide ( HQNO ) [19] , 2 , 4-dihydroxyquinoline ( DHQ ) [20] , and the non-HAQ molecule 2-AA . 2-AA is a relatively simple , non-HAQ volatile molecule responsible for the grape-like odor of P . aeruginosa cultures as well as burn wounds infected with P . aeruginosa [18] , [21] . Along with DHQ and HQNO , 2-AA is also produced and excreted by the HAQs-producing bacterium Burkholderia thailandensis [23] ( Figure S1 ) . Because HHQ and PQS both induce pqsABCDE expression [14] , we investigated whether these three additional abundant molecules also induce expression of these genes . Maximum levels of 2-AA in the cell supernatant vary from micromolar to millimolar range depending on the growth medium used [18] , [22] . We used Luria-Bertani ( LB ) broth media in experiments examining the effects of exogenous 2-AA supplementation on pqs operon gene transcription . We chose to use LB broth because it supports lower levels of 2-AA production ( 37 . 5 µM = 5 µg/ml Figure 1B ) than other media [18] . Using a pqsA-green fluorescence protein ( GFP ) ( ASV ) -transcriptional reporter fusion in a pqsA::H double mutant background that does not produce any of these molecules ( Figure S2 ) , we exogenously added to LB media the above molecules . HQNO only modestly induced pqsABCDE expression , while DHQ and 2-AA did not induce pqsABCDE expression ( Figure 1C ) . This finding indicates that unlike HHQ and PQS , 2-AA and DHQ molecules may have biological roles other than activating pqs operon transcription . The kinetics and dose-dependency effects of exogenously added 2-AA and DHQ on pqsA regulation were examined in greater detail using WT cells carrying the pqsA-GFP ( ASV ) transcriptional reporter fusion . Figure 2A shows that 2-AA , but not DHQ ( Figure S3A ) , greatly reduced pqsA expression , in a dose-dependent manner , with 2-AA achieving the strongest inhibition at 200 µg/ml ( 1 . 5 mM ) . Bacterial growth was unaffected by either 2-AA or DHQ ( Figure S3B and C ) . In light of these results , we focused our subsequent experimental efforts on 2-AA . We visualized pqsA-GFP expression under a fluorescent microscope in the presence of a range of 2-AA concentrations . Surprisingly , but in accordance with the established dose-dependent effects of 2-AA , only a subpopulation of the cells showed a shut-down of pqsA-GFP activity in the presence of 50 µg/ml ( 0 . 375 mM ) 2-AA ( Figure 2B and C ) ; a statistically significant number of cells had turned off the pqsA expression though some still fluoresced strongly . No fluorescing cells were seen at 200 µg/ml 2-AA ( Figure 2B and C ) . Thus 2-AA appears to promote phenotypic heterogeneity in a genetically “homogenous” population and silence pqs operon expression in a fraction of the cells , suggesting that MvfR active and inactive cells co-exist to achieve a bistable phenotype [24] . The size of the 2-AA responsive population increased with higher 2-AA concentrations . The need for higher concentrations of 2-AA to produce significant detectable transcriptional changes may be due to the compound affecting only a subpopulation of cells . 2-AA inhibition of MvfR is likely mediated via negative feedback regulation given that MvfR controls 2-AA synthesis ( Figure 1A ) . To corroborate our pqsA expression data , we measured levels of HHQ in 2-AA treated cells and found that HHQ production was also inhibited by 2-AA in a dose-dependent manner ( Figure 2D ) . Consistent with the view that 2-AA down-regulates the MvfR regulon , we observed that 2-AA decreased production of pyocyanin ( Figure 2E ) and pyoverdine ( Figure 2F ) , virulence factors whose synthesis depends on MvfR . A time course study of 2-AA production ( Figure 1B ) showed that 2-AA levels did not significantly decrease as occurs for the MvfR activators HHQ and PQS ( Figure 1B ) , indicating that the production kinetics and stability of 2-AA are distinct from those of the MvfR activators . Together , this convergence of data strongly suggests that 2-AA has the novel biological activity of silencing the MvfR regulon . To elucidate the mechanism by which 2-AA inhibits the MvfR regulon , we administered 2-AA together with HHQ to PA14 isogenic pqsA::pqsH double mutant cells , which do not produce any HAQs [14] or 2-AA ( Figure S2 ) but have functional MvfR . The expression of the pqsA reporter in pqsA::pqsH cells requires activation of MvfR . While exogenous addition of HHQ to these cells induced pqsA reporter expression via activation of MvfR , co-addition of 2-AA , at 100 µg/ml ( 0 . 75 mM ) and above , attenuated this expression in a dose-dependent manner , suggesting that 2-AA may negatively impact the MvfR-regulated operon pqsABCDE at the transcriptional level ( Figure 3A ) via MvfR . Unlike other anthranilic acid analogs , which act on PqsA activity and inhibit the synthesis of MvfR ligands [25] , the apparent 2-AA inhibition described here can be regarded as independent of PqsA since the pqsA::pqsH mutant was used in these experiments . Moreover , 2-AA did not perturb MvfR protein levels or MvfR ligand stability ( data not shown ) . To assess whether down-regulation of MvfR activity by 2-AA is due to reduced ligand levels , we engineered mvfR mutant cells to synthesize HHQ independently of MvfR by constitutively expressing pqsABCD . These cells produced HHQ in the absence of 2-AA whereas HHQ production decreased in the presence of 2-AA in a dose-dependent manner ( Figure 3B ) , indicating that inhibition of the MvfR regulon can also occur independently of MvfR . This post-transcriptional inhibition could be mediated through interference with ligand biosynthesis . The pqsABBCDE operon gene pqsA is required for the synthesis of 2-AA and HHQ ( Figure S2 ) . In our effort to understand 2-AA biosynthesis , we created various mutants in the PQS operon . Surprisingly , the pqsB::C mutant produced 2-AA in the absence of HHQ ( Figure S2 ) , demonstrating that neither pqsB and pqsC is required for 2-AA synthesis . Addition of PQS to pqsB::C mutant cultures induced pqs operon transcription , resulting in WT levels of 2-AA ( Figure S2 ) . However , as expected the addition of PQS did not result in production of HHQ ( data not shown ) . We used pqsB::C mutant cells with an additional , more sensitive reporter system to further assess the effects of endogenous 2-AA produced by the pqsB::C mutant on transcription of the mvfR regulon and its biological relevance in the absence of HHQ , as well as to quantify 2-AA effects on the pqs operon . We fused the pqsA promoter to the Bacillus subtilis sacB gene that codes for the levansucrase product , which is toxic when cells are grown in the presence of sucrose . The sacB gene has previously been incorporated into allelic exchange vectors as a means of counter-selection [23] . As shown in Figures 3C and S4 , PA14 cells with the pqsA-sacB fusion gene incorporated stably into their chromosome did not grow in the presence of sucrose , while the corresponding isogenic pqsB::C mutant cells did grow significantly , corroborating the findings that 2-AA suppresses pqsA promoter activation and indicating that this effect is stronger in the absence of HHQ . Bacterial cell proliferation occurred even following exogenous addition of HHQ , which promotes further induction of the pqs operon ( Figure 3C and S4 ) . Importantly , the endogenous level of 2-AA in the pqsB::C mutant cells in presence of the inducer , had a suppression efficacy comparable to 100 µg/ml of exogenously added 2-AA Figure 3C . These results are physiologically important and consistent with the putative role of 2-AA in down-regulation of MvfR regulon . These results also suggest there may be limited uptake of exogenously added 2-AA thereby requiring addition of higher concentrations of 2-AA to generate a physiological response by exogenously added 2-AA . The last gene in the pqs operon , pqsE , is essential for pyocyanin production [16] , [26]–[27] , but dispensable for HAQ synthesis [12] , [15] . Importantly , 2-AA did not reduce pyocyanin in an mvfR mutant that constitutively expressed pqsE , even though it efficiently inhibited pyocyanin production when MvfR was constitutively expressed ( Figure 3D ) . Hence we can deduce that 2-AA negative regulation of the MvfR regulon is a result of the down-regulation of pqsABCDE operon expression and interference with MvfR activity upstream of PqsE , via inhibition of HHQ biosynthesis . We have shown previously that P . aeruginosa pathogenesis can be studied in Drosophila melanogaster [28]–[30] . Drosophila shares striking similarities with mammals in terms of its overall physiology and innate immunity signal transduction pathway components [28] , [31] . P . aeruginosa strain PA14 is highly virulent in flies , causing high mortality; meanwhile , mvfR mutants exhibit reduced virulence , causing reduced mortality [30] . Therefore , since 2-AA negatively regulates MvfR , we first tested whether it reduces P . aeruginosa virulence in Drosophila . As shown in the Figure 4A , flies co-injected with P . aeruginosa and 2-AA succumb to infection significantly later than flies injected with only P . aeruginosa . To confirm the in vivo efficacy of 2-AA in attenuating the virulence of PA14 cells and validate the above findings in a mammalian model of infection , we used the well-studied acute mouse burn and infection model [32] , in which we have shown previously that mvfR mutant cells cause attenuated virulence [10] , [25] . Indeed , as shown in Figure 4B , 2-AA also attenuated the virulence of PA14 cells in mice . Mice inoculated with PA14 and injected once with 2-AA survived significantly longer than control mice inoculated with PA14 alone . Importantly , this effect could not have been due to reduced PA14 proliferation since 2-AA did not affect the bacterial colony forming units ( CFUs ) in the muscle tissue underlying the burn injury and infection site ( Figure 4C ) . Importantly , 2-AA greatly hampered the systemic spread of bacteria in the blood of the infected mice ( Figure 4D ) , which is a significant problem in humans with P . aeruginosa infections . Silencing of MvfR regulon activity , and thus of acute virulence factor gene expression , is likely responsible in large part for the reduced systemic dissemination of PA14 cells and attenuated PA14 virulence observed . One may wonder why a pathogen such as P . aeruginosa would produce a molecule that decreases its own virulence . However , given that successful adaptation of an organism depends on its ability to regulate gene expression in response to its changing environment and thereby maximize its long-term survival , the existence of such a mechanism should not be surprising . There is ongoing debate regarding the role of QS in chronic infection due to inactivation of LasR in P . aeruginosa isolates from CF sputum , while a lack of QS has been proposed to facilitate adaptation during a chronic infection [33] . Since 2-AA down-regulates the expression of QS-related acute virulence functions , we questioned if the same molecule could promote other functions that support adaptation during chronic infections . We therefore examined a variety of phenotypes associated with chronic infection by P . aeruginosa . A significant fraction of P . aeruginosa cells isolated from chronically infected CF patients accumulate multiple mutations in genes affecting acute virulence functions , including mucA , the regulator of alginate production , the QS regulator LasR , type III secretion system , multidrug efflux pumps , genes involved in motility , and DNA repair genes such as mutS [2] , [34]–[38] . Mutations in the lasR gene are particularly notable , not only because they accumulate in bacteria colonizing the lungs of chronically infected CF patients , but also because they have been shown to promote long-term growth and survival of CF isolates in vitro [33] . Therefore , we first examined the frequency of occurrence of lasR mutations in populations of P . aeruginosa grown in the presence of increasing concentrations of 2-AA . We found that PA14 cells exposed to increasing 2-AA concentrations for 10 d accumulated significantly more lasR mutations than untreated cells ( Figure 5A ) . Sequence analysis showed that these lasR mutant lines harbor simple deletions or single nucleotide non-synonymous mutations in the lasR gene that produce inactive LasR protein ( Table 1 ) . One of the lasR mutant lines also carried mutations in the intergenic region of the fleQ flagellar gene regulator , which is important for motility , and five clones carried mutations in the intergenic region of mexT , a regulator of multidrug efflux . Mutations in these genes lead to loss of flagellar motility and antibiotic resistance , respectively [39]–[40] . However , no mutations in the DNA repair gene mutS , in type III secretion system genes ( exsA , pscQ and popD ) , or in the virulence factor vfR or rpoN , which are also associated with chronic infections [2] , were identified in any of the lasR mutant lines sequenced . Additionally , 2-AA does not appear to be mutagenic , as it did not stimulate mutagenesis in a standard Ames test ( data not shown ) . We did not identify any mutation in lasR within a day of incubation when silencing of mvfR was observed , suggesting that it's silencing is not due to the loss of lasR . Nevertheless , the ability of 2-AA to promote lasR mutations was particularly pronounced after 10 d of incubation in isogenic mutS cells ( Figure 5A ) , which are defective in DNA repair function . In contrast to the 2-AA dependent non-linear accumulation of lasR mutants observed in the WT background , a more linear accumulation of lasR mutants in the mutS mutant background is observed ( Figure 5A ) . These data suggest that once a DNA repair mechanism is compromised , as in mutS mutants commonly found in P . aeruginosa cultures from CF sputum [37]–[38] , 2-AA effect is more prominent . Although the mechanisms and advantages of lasR mutations are not clearly understood , it has been proposed that lasR mutation may provide a growth advantage [33] . We further tested the effects of 2-AA on long-term survival of P . aeruginosa . When grown under the condition of limited iron and aeration , PA14 cells reach stationary phase within 10–12 h and undergo significant lysis after 40 h ( Figure 5B ) . However , addition of 2-AA inhibited entry into the lytic phase and the treated cells remained in stationary phase , promoting bacterial cell long-term survival ( Figure 5B ) . To investigate whether 2-AA can also promote long-term survival in vivo , we developed a non-vertebrate , whole animal persistence infection assay in Drosophila melanogaster . Using the fly-P . aeruginosa feeding assay [28] , flies were fed for 2 d with P . aeruginosa WT ( PA14 ) , 2-AA non-producer PA14 isogenic mutant strain pqsA , and the 2-AA producer strain pqsB::C . Flies were then transferred onto bacteria-free food , and the CFUs were quantified at 7 d post-feeding . Bacterial load decreased over time in the pqsA-infected flies , while flies infected with the 2-AA producing strains PA14 and pqsB::C sustained ∼3 log greater CFUs ( Figure 5C ) , indicating that 2-AA helps bacteria to survive and persist , and thus promotes better fitness in a dynamic host environment . Due to the longer incubation duration , we cannot rule out the possibility that the persistence seen here may be due to lasR mutation induced by 2-AA . Together with the above in vitro studies , these in vivo results corroborate the view that 2-AA plays a key role in switching cells to a phase that promotes concomitant adaptations that enable P . aeruginosa to persist in a chronic infection . Production of the 2-AA has been reported in plants , invertebrate animals , and vertebrate animals , including humans [41] . Moreover , several plant and human eubacterial pathogens , including Pseudomonas , Bordetella , Burkholderia , Ralstonia , Streptomyces , and Mycobacteria , as well as the Archeae Sulfolobus , encode putative PqsA and MvfR homologues , and in some cases , pqsBCD-related loci [42] . As such , these species might also produce 2-AA and use this molecule to down-regulate their respective virulence functions . Indeed , Burkholderia thailandensis , highly related and often used as a non-pathogenic surrogate for the level 3 pathogen B . pseudomallei , was confirmed to produce 2-AA ( Figure S1 ) . We assessed whether this 2-AA could inhibit HAQ biosynthesis in B . thailandensis , and found that exogenous 2-AA inhibited production of the HHQ and HNQ methylated analogs HMHQ and HMNQ ( Figure S5A and B ) [23] , suggesting that 2-AA may perform analogous functions in B . thailandensis as observed in P . aeruginosa . 2-AA producing species may have a selective advantage in mixed microbial communities by adversely influencing the expression or activities of fitness traits of their neighbors . Here we provide evidence that 2-AA , a small volatile molecule produced by P . aeruginosa ( and B . thailandensis ) , is a QS regulated molecule and an important modulator of acute and chronic virulence functions . Although much work has focused on how QS cell-cell signaling leads to the activation of a plethora of virulence factors , little is known about the silencing of these factors . Degradation of AHL by an AHL acylase [43] and interference of LasR and RhlR binding to their promoter by the transcription factors QscR [44] and RsaL [45] have been shown to down-regulate AHL mediated QS . We recently demonstrated the existence of an interplay between QS systems , their components , and environmental factors that negatively impact HAQs [27] . However , no signaling molecule has been identified to date that negatively regulates acute virulence QS related functions or promotes bacterial adaptation and long-term persistence . Furthermore , no volatile compounds have been reported to be involved in QS regulation and virulence . Our results demonstrate that the low molecular weight molecule 2-AA silences the MvfR regulon via a negative feedback mechanism , most likely via repressed synthesis of MvfR ligands transcriptionally and post-transcriptionally , which restricts HAQ-mediated QS signaling and thus acute virulence functions , thereby enabling chronic infection . We propose the idea that there may be QS regulated aerial communication among bacteria , as 2-AA is a QS-regulated molecule and a volatile signal [46] . Such aerial communication may occur and be critical in a microenvironment such as the one found in biofilms and in infected wounds or CF lungs . It is important to note that suboptimal inhibitory concentration of 2-AA affects only a subpopulation of cells , suggesting that MvfR active and inactive cells co-exist causing a bistable phenotype . Similar bi/multistability phenotypes have been described as naturally occurring for several well known systems regulated by feedback mechanism , such as regulation of the Lac operon in E . coli , mucoidy in P . aeruginosa , lysogeny of the lambda phage , and transformation competency and sporulation in B . subtilis [24] , [47] . In our model ( Figure 5D ) , we propose that 2-AA inhibits MvfR via feedback regulation and promotes longer cell survival in vitro and in vivo in flies , presumably by interfering with cell lysis and inducing mutations in lasR . Additionally , the accumulation of lasR mutations that allow adaptation to low oxygen conditions [48] suggest that this mechanism contributes to overall pathogen fitness in chronic infections . Whether , lasR mutations arise in WT cells , MvfR silenced cells , and/or metabolically altered cells , remains to be determined . In corroboration , the Drosophila persistence assay showed that 2-AA promotes bacterial long-term survival and as such fitness . Assessment of the potential role of 2-AA in the establishment and persistence of chronic infections has been hindered by a lack of clinically relevant chronic infection models . Nevertheless , the Drosophila persistence assay , using 2-AA producing and non-producing isogenic strains , permits us to at least determine the impact of this small molecule in a dynamic host environment that shares many immune functions with the human . Our identification of a P . aeruginosa QS regulated volatile molecule that limits virulence and invasive functions associated with acute infections while promoting phenotypic and genetic changes associated with chronic infection elucidates novel avenues for combating bacterial adaptation and survival in the chronic infection environment . These results also suggest that interfering with the MvfR pathway could prevent both acute and chronic infections and that 2-AA synthesis provides a potential target for the development of new anti-virulence drugs in the combatance of P . aeruginosa , and possibly of other pathogenic bacteria such as Burkholderia . This study uncovers insights that paradigmatically pave the way for the search for 2-AA-like volatile small molecules that promote pathogen adaptation and establishment of chronic infections caused by foreboding human pathogens .
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 Ethics of Animal Experiments of the Massachusetts General Hospital ( Permit Number: 2006N000093/2 ) . All Procedures were performed under sodium pentobarbital anesthesia , and all efforts were made to minimize suffering . A P . aeruginosa strain known as RifR human clinical isolate UCBPP-PA14 ( PA14 ) was used in the present experiments [11] . All of the PA14 mutants described in this paper are isogenic to UCBPP-PA14 . The bacteria were grown at 37°C on LB broth or on plates of LB agar containing appropriate antibiotics unless otherwise indicated . The overnight PA14 cultures were grown in LB and diluted the following day in fresh media with or without 2-AA . Bacterial growth kinetics was determined by taking OD600 nm measurements . Quantification of 2-AA and HAQs in bacterial culture supernatants was performed as described previously [14] , [49] . The HAQs were separated on a C18 reverse-phase column connected to a triple quadrupole mass spectrometer , using a water/acetonitrile gradient [49] . Positive electrospray in MRM mode with 2×10−3 mTorr argon and 30 V as the collision gas and energy was employed to quantify 2-AA and HAQs , using the following ion transitions: 2-AA 136>91 , HHQ 244>159 , HHQ-D4 248>163 , PQS 260>175 , and PQS-D4 264>179 . To quantify HAQs , Pseudomonas PA14 cells were grown in LB supplemented with different concentrations of 2-AA ( Sigma , USA ) , with untreated LB being used as a negative control . The culture supernatant was collected at different stages of growth mixed with an equal volume of methanol , and the compounds were analyzed by LC/MS . All samples were analyzed in triplicate . Bacterial cells were diluted 1/100 from overnight cultures and grown in six replicates in 20 µl in a 96-well plate ( Corning , Inc . , Corning , NY ) . Bacterial growth was measured as a function of optical density at OD600 nm using Tecan F200 automated plate reader ( Infinite F200 , Tecan Group Ltd , Männedorf , Switzerland ) . For the pqsA-sucB assay , the bacteria were diluted from overnight cultures to OD600 nm 0 . 1 in NaCl-free LB with tetracycline ( 50 µg/ml ) and grown ( Sigma , USA ) , with or without 10% sucrose ( 200 µl ) . Growth was measured in triplicate every 30 min for up to 25 h , using a Sunrise plate reader ( Tecan Group Ltd , Männedorf , Switzerland ) . For lysis experiments , the LB medium was treated with Chelex 100 resin 100–200 mesh , sodium glutamate ( 50 g/l ) ( BioRad , Hercules , CA ) for 1 h followed by filtration with 0 . 2 µm filters ( Corning , NY ) . Bacterial cells were diluted 1/100 in presence of various concentrations of 2AA . Triplicates ( 200 µl volumes ) were inoculated in 96 wells plates . The plates were incubated and growth was recorded every 30 min for 3 d using a Sunrise plate reader ( Tecan Group Ltd , Männedorf , Switzerland ) . PA14 or pqsA::pqsH cells carrying pAC37 plasmids with pqsA promoter fused to short-lived GFP {pqsA-GFP ( ASV ) } [50] were grown overnight in LB supplemented with gentamycin ( 50 µg/ml ) . The cultures were diluted in the morning with the test compound at desired concentrations and aliquoted ( 200 µl ) into 6 replicates in a 96-well assay plate ( Corning , Inc . Corning , NY ) . Green fluorescent protein ( GFP ) fluorescence ( excitation at 485 nm , emission at 535 nm ) and OD600 nm for growth was measured every 30 min using a Tecan F200 automated plate reader ( Infinite F200 , Tecan Group Ltd , Männedorf , Switzerland ) . Values presented were above the background fluorescence from the empty strain . For measurement of ß-galactosidase activity , pqsA::pqsH cells harboring pGX5 , which carries the pqsA-lacZ reporter gene [51] , were diluted to OD600 nm = 0 . 05; and OD600 nm and ß-galactosidase activity ( expressed as Miller Units ) were measured at the indicated optical densities . Assays were performed in triplicate . All of the experiments were repeated at least three times . The mutant was constructed as described by Lesic et al [52] such that most of the pqsB and pqsC genes were replaced with a Kanamycin resistance marker . The deletion was made from nt 151 in pqsB to nt 456 in pqsC ( within the coding regions ) , leaving 150 bp of the 5′ end of the pqsB coding region and 149 bp of the 3′ end of the pqsC coding region . The mutants were selected on LB plates containing Kanamycin ( 200 µg/ml ) . PA14 cells harboring the plasmid pAC37 , which contained the pqsA-GFP ( ASV ) reporter fusion , were grown to mid-logarithmic phase in the presence of various concentrations of 2-AA . The cells were washed in phosphate buffered saline ( PBS ) and their membranes were stained with FM-64 ( Invitrogen ) according to the manufacturer's instructions . Five-microliter aliquots of bacterial cells were spotted onto slides and covered with Poly-L-lysine coated coverslips ( Sigma Aldrich , US ) . The bacteria were visualized using an Eclipse E800 ( Nikon ) microscope with FM-64-stained membranes visualized as red and pqsA-GFP visualized as green . The pictures were processed using Spot V4 . 0 . 9 ( Diagnostic Instruments ) software . To constitutively express pqsABCD , a genomic fragment containing pqsABCD was amplified by PCR using pqsABamHI 5′ CATGGATCCAACGTTCTGTCATGTCCACG3′ and PqsDPst1 5′CGACTGCAGTCAACATGGCCGGTTCAC3′ primers from an H44 cosmid [53] as template . The BamH1 and Pst1 digested PCR product was ligated to a pDN18 vector digested with BamH1 and Pst1 . The construct was introduced into E . coli Top10 ( Invitrogen ) cells and P . aeruginosa mvfR mutant cells by electroporation . To construct pqsA-sucB reporter fusion , the pqsA promoter was amplified using the primers 5′GACTAGTCGAGCAAGGGTTGTAACGGTTTTTG3′ and 5′GAAGATCTGACAGAACGTTCCCTCTTCAGCGA3′ and the PA14 chromosome was used as template . The sacB gene was amplified using the primer pairs 5′GAAGATCTATGAACATCAAAAAGTTTGCA3′ and 5′AAACTGCAGGTTGATAAGAAATAAAAGAAAATGCC3′ from pKOBEG-sacB plasmid [54] . PqsA promoter ( PpqsA ) was then digested with SpeI/BglII and the fragment containing sacB was digested with BglII/PstI . The two fragments were ligated to the CTX ( TetR ) plasmid digested with SpeI/BglII . The ligated vector was eletroporated into E . coli SM10 lambda pir and used to integrate the CTX-PpqsA-sacB to PA14 chromosome . Rif and TetR plates were used to select for the PA14 CTX-PpqsA-sacB clones and the presence of CTX-PpqsA-sacB in the PA14 chromosome was further confirmed by PCR . Overnight PA14 cultures were diluted to OD600 nm = 0 . 05 in 5 ml LB or LB +200 µg/ml 2-AA . The bacteria were grown in triplicate at OD 3 . 0 . Pyocyanin was extracted with chloroform from 5 ml cell culture supernatant and then extracted with an equal volume of HCl ( 0 . 2 N ) ; optical density was measured at OD520 nm . The amount of pyocyanin was quantified by multiplying the OD520 nm value by 17 . 072 to obtain values in µg/ml [55] . PA14 bacterial cells ( 200 µl ) were grown in 96-well plates in D-TSB medium with a range of 2-AA concentrations . The plates were incubated and pyoverdine production and growth were measured every 30 min by a Tecan F200 automated plate reader ( Infinite F200 , Tecan Group Ltd , Männedorf , Switzerland ) . Pyoverdine levels were measured using excitation at 400 nm and emission at 460 nm . The values were normalized to cell growth ( OD600 nm ) . Pyoverdine concentrations were calculated using a calibration curve of fluorescence for a range of pyoverdine concentrations ( Sigma Aldrich , USA ) . Fly feeding on a P . aeruginosa-containing solution was performed as described previously [28] . Briefly , 5–7-day-old female Oregon-R flies ( N = 26 per group ) were fed for 1 d with a mixture of 0 . 05 ml of LB bacterial culture at OD600 nm = 1 . 8 , 1 ml of 20% sucrose , and 4 ml of water . Thus , the feeding mix contained a final concentration of 1% LB , ∼3×107 bacteria/ml , and 4% sucrose . A sterile cotton ball was placed at the bottom of each fly vial and was impregnated with 5 ml of the feeding mix . The flies in each treatment group were sub-divided into three fly vials ( 13 flies per vial ) , sealed with a clean cotton ball , and incubated at 25°C . A day later , the flies were transferred to 50 ml plastic screw-cap tubes ( 10 flies per tube ) . The tubes were perforated with a heated 0 . 9-mm needle to enable aeration , while the caps were perforated with a heated 1 . 2-mm needle and covered with a 2 . 3 cm Whatman disc ( Fisher scientific , USA ) . A 0 . 2-ml volume of 4% sucrose solution was dropped onto the Whatman disc and covered with parafilm to provide food for the flies with minimal contamination . New fly tubes were prepared daily and the flies were incubated at 25°C . CFU counts per fly were measured at the indicated time points by dipping 6–9 flies per time point in 95% ethanol for 3 s and letting them dry out on paper tissues ( Kimwipes ) . Then each fly was ground up using a 1 . 5-ml tube pestle in 0 . 1 ml of LB , and 1:10 dilutions were plated on LB plates . Statistical analysis of the CFU data was performed using the Student's t-test . PA14 and mutS mutant were grown overnight in LB . The saturated culture was diluted 1:10 in triplicate in 4% sucrose with varying concentrations of 2-AA , or without 2-AA . The diluted cultures were incubated statically at 25°C for 10 d . The cells were then diluted and plated to allow the numbers of total colonies to be counted the following day . After 2 d of plating , the lasR cells were identified by their excessive pyocyanin production . The percentage of lasR mutants in each sample was calculated and statistical significance was determined by Student's t-test , assuming equal variance . The lasR mutant colonies were further confirmed by sequencing of the complete lasR gene . In addition , the complete sequences of mexA , mexE , mexS , mexT , mexZ , wspF , fleQ , exsA , accC , vfR , popD , mutS , nfxB , rpoN , rhlR , pqsA , pqsb , pqsE , pqsH , and mvfR were also determined in these lasR mutant colonies . | P . aeruginosa causes acute as well as chronic infections in humans . In this paper we report the identification of a P . aeruginosa small molecule , 2-AA , that modulates this pathogen's virulence to promote chronic infections . We show that the synthesis of 2-AA , responsible for the grape-like odor of P . aeruginosa cultures and of wound infections , is controlled by the multiple virulence factor regulator ( MvfR ) important for virulence in acute infections . 2-AA reduces the production of MvfR-regulated acute virulence factors , and attenuates acute virulence by negatively fine-tuning the MvfR regulon activity . Moreover , we show that 2-AA adapts P . aeruginosa for chronic infections by promoting mutations in a key acute virulence gene ( lasR ) and by prolonging bacterial survival . The findings presented here reveal the function of a new MvfR-regulated molecule , and highlight MvfR's importance as a highly promising target for the development of inhibitors that can simultaneously halt acute and chronic infections caused by P . aeruginosa , and possibly by other pathogenic bacteria . This study uncovers insights that paradigmatically pave the way for the search of 2-AA-like small volatile molecules that promote pathogen adaptation and establishment of chronic infections caused by foreboding human pathogens . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"gram",
"negative",
"microbial",
"pathogens",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"bacterial",
"pathogens",
"pathogenesis"
] | 2011 | A Quorum Sensing Regulated Small Volatile Molecule Reduces Acute Virulence and Promotes Chronic Infection Phenotypes |
Fibroblastic reticular cells ( FRCs ) form the cellular scaffold of lymph nodes ( LNs ) and establish distinct microenvironmental niches to provide key molecules that drive innate and adaptive immune responses and control immune regulatory processes . Here , we have used a graph theory-based systems biology approach to determine topological properties and robustness of the LN FRC network in mice . We found that the FRC network exhibits an imprinted small-world topology that is fully regenerated within 4 wk after complete FRC ablation . Moreover , in silico perturbation analysis and in vivo validation revealed that LNs can tolerate a loss of approximately 50% of their FRCs without substantial impairment of immune cell recruitment , intranodal T cell migration , and dendritic cell-mediated activation of antiviral CD8+ T cells . Overall , our study reveals the high topological robustness of the FRC network and the critical role of the network integrity for the activation of adaptive immune responses .
Efficient interactions between the immune system and microbial antigens are initiated and maintained in secondary lymphoid organs ( SLOs ) that are strategically positioned at routes of pathogen invasion . Lymph nodes ( LNs ) , for example , are found at convergence points of larger lymph vessels , which drain extracellular fluids from peripheral tissues [1] . The interaction of naïve T cells with antigen-presenting dendritic cells ( DCs ) in LNs needs to be well coordinated because T cells with a particular specificity are rare [2 , 3] . Optimal communication between immune cells relies to a large extent on the fibroblastic reticular cell ( FRC ) network that provides specialized microenvironments for cellular interactions . For example , FRCs regulate T cell migration and survival in the T cell zone by producing homeostatic chemokines and cytokines [4–6] . Moreover , FRCs located in and around B cell follicles coordinate B cell trafficking and activity [7–9] . Importantly , while the role of FRCs in the regulation of immune responsiveness has been studied extensively ( reviewed in [10 , 11] ) , the underlying principles of the FRC network topology and its contribution to general LN functionality have remained unexplored . In order to determine the topological properties of networks , the theoretical framework of the graph theory can be utilized [12 , 13] . The theory of complex networks has been applied in the study of real-world networks , including the internet [14 , 15] , scientific collaboration [16] , power grid systems [17] , and the worldwide air transportation network [18] . Moreover , graph theory has been instrumental for the analysis of various biological systems , such as metabolic networks [19 , 20] , protein–protein interactions [21] , and neuronal cell connectivity [22 , 23] . Different classes of networks can be defined based on the nature of their topology . Random networks are described by the Erdos-Renyi model [24] in which objects ( nodes ) form random connections ( edges ) between each other with the same probability . Hence , most nodes will have approximately the same number of connections , centered on the network average with a Poisson degree distribution . In contrast , scale-free networks [25 , 26] are characterized by a power-law degree distribution with most nodes possessing few connections and very few nodes showing large numbers of connections . These few highly connected nodes are called hubs , and they maintain the whole network structure . Networks with less-centralized structures are called small-world networks [27] , where any two nodes can be reached with only a few steps in the network . A key feature of complex networks is their robustness to perturbation , which denotes the ability of a network to remain operational when nodes are functionally impaired or destroyed [14] . Such topological robustness is determined by the organizational principles of the network and has an impact on overall network functionality [13] . Interestingly , most real-world networks exhibit small-world topology , a property that is thought to provide networks with high resilience to external perturbation [28] . In contrast to engineered systems , understanding biological robustness is a difficult challenge due to the multilayered complexity of the system in which functionally relevant measures of robustness need to be established [29] . The FRC network can be almost completely destroyed during viral infection [4] or substantially altered during chronic infection with parasites [30] leading to severe immune deficiency . It is therefore important to assess the topological robustness of the FRC network and to define those parameters that determine network resilience . To address these questions , we have utilized the Ccl19idtr mouse model [7] , which enables diphtheria toxin ( DT ) -mediated ablation of FRCs expressing C-C motif chemokine 19 ( CCL19 ) . Graph theory-based analysis showed that the LN FRC network forms a lattice-like small-world network that exhibits complex network topology , substantial connectivity , and high capacity for clustering . Moreover , in silico analysis and thorough in vivo validation revealed substantial topological robustness of the FRC network .
The different microenvironments of the LN , e . g . , T or B cell zones or the subcapsular region , are built by distinct FRC subpopulations [10] . Importantly , all LN FRCs can be specifically targeted in vivo using the Ccl19-cre mouse model ( Fig 1A ) , whilst transgene expression is absent in hematopoietic cells [5] including CD11c+MHCIIhigh DCs ( not shown ) . For the structural network analysis , we have focused on the classical podoplanin ( PDPN ) -expressing T cell zone FRCs that orchestrate the interaction of DCs and T cells [31] and provide important survival factors for T cells , such as interleukin 7 ( IL-7 ) [6 , 32] . Three-dimensional reconstruction of high-resolution confocal microscopy Z-stacks covering a volume of 304 x 304 x 32 μm was applied to analyze the FRC network structure ( Fig 1B ) by defining nodes as the enhanced yellow fluorescent protein ( EYFP ) -positive FRC centers of mass and edges as physical connections between adjacent FRCs ( Fig 1C and S1 Video ) . Small-world networks exhibit the intrinsic property that most nodes can be reached from every other node by a small number of steps , even though most nodes are not direct neighbors . This enables small-world networks with fast and efficient information transfer , which is characterized by small shortest path lengths ( node-to-node distances ) . These networks also exhibit high capacity for clustering ( i . e . , connectivity between neighboring nodes ) , which is strikingly different from random networks in which all nodes have the same probability of containing an edge . Thus , a specific network can be classified as a small-world network by comparing network-level statistics to equivalent random and lattice networks . Moreover , small-worldness can be described by the σ and ω parameters ( see S1 Table ) , which classify a network as small world if σ > 1 and ω ≈ 0 ( range −0 . 5 to +0 . 5 ) [33–35] . Accordingly , random networks will show σ ≈ 1 and positive 0 < ω < 1 , while lattice networks will have σ > 1 like small-world networks but negative −1 < ω < 0 . As shown in Fig 1D ( left panel ) , a representative FRC network sample contained 176 nodes ( N ) and 685 edges ( E ) with σ and ω values of +6 . 7 and −0 . 27 , respectively . Note that network connectivity is color coded with highly connected nodes ( E ≥ 12 ) depicted in red . The equivalent random network with the same number of nodes and edges as the FRC network has both σ = 1 and ω = 0 . 93 positive ( Fig 1D , middle ) . A regular ring lattice network with the same number of nodes and eight edges per node connecting to nearest neighbors fulfills the condition for small-worldness with σ = 3 . 53 , while the negative ω = −0 . 76 identifies the lattice structure , as expected ( Fig 1D , right ) . This initial network analysis with σ = 6 . 128 ± 0 . 659 and ω = −0 . 308 ± 0 . 069 ( n = 6 mice ) indicates that FRCs of the T cell zone form a small-world network with lattice-like properties . Since the functional properties of a network are determined by its structure [36] , we assessed first whether the FRC network structure is hardwired and will be reestablished after removal of all nodes . To this end , we used specific FRC ablation in mice that express both the diphtheria toxin receptor ( DTR ) and EYFP under the control of the Ccl19 promoter ( Ccl19eyfp/idtr ) [7] . To achieve complete ablation of FRCs at the start of the experiment ( i . e . , day 0 ) , 8 ng DT per g body weight were injected intraperitoneally on days −5 and −3 ( S1A Fig ) . PDPN+EYFP+ FRCs in T cell zones were removed , while PDPN expression in and around high endothelial venules was partially maintained ( S1B and S1C Fig ) . The FRC network was partially restored on day 14 ( Fig 2A , S1B and S1C Fig ) , with approximately 32% of the EYFP volume restored ( Fig 2B ) . Importantly , the FRC network had been rebuilt on day 28 to an extent that was indistinguishable from controls ( Fig 2A and 2B , S1B and S1C Fig ) . However , basic single-cell parameters , namely FRC surface area ( Fig 2C ) and volume ( Fig 2D ) , had not yet reached the levels of controls , while other morphological parameters such as cell sphericity had returned to normal values ( Fig 2E ) . Moreover , the FRC network had reached the original cell distribution with identical intercellular distances ( Fig 2F ) and number of connected protrusions per cell ( Fig 2G ) , suggesting that the FRC network structure can be restored from scratch within approximately 4 wk . Indeed , topological network analysis ( Fig 3A ) confirmed that essential network parameters such as the number of nodes ( Fig 3B ) and edges ( Fig 3C ) had almost completely returned to the levels of controls . Further network properties such as the number of edges per FRC ( Fig 3D ) and the local clustering coefficient ( Fig 3E ) had been restored as well . Likewise , small-worldness , as determined by the σ ( Fig 3F ) and ω factors ( Fig 3G ) , was maintained after 28 d , indicating that the FRC network small-world structure is an imprinted trait of the LN infrastructure . The next set of experiments was performed to determine the structural stability of the FRC network under conditions of partial removal of nodes . Graded doses of DT were applied , and FRC morphology and topology were assessed . As shown in Fig 4A and S2 Fig , application of 0 . 5 ng/g DT resulted in moderate FRC ablation , while doses above 2 ng/g resulted in substantial damage to the FRC network . Global morphological analysis confirmed the drastic effect of DT doses >2 ng/g on the EYFP+ cell population ( Fig 4B ) . FRC numbers decreased by 37% , 67% , 70% , 91% , and 100% in mice treated with 0 . 5 , 1 , 2 , 4 , and 8 ng/g DT , respectively ( Fig 4C ) . Single-cell analysis revealed a steady increase in FRC volume with higher DT doses ( Fig 4D ) . Moreover , other cellular parameters such as compactness and surface area also increased with decreasing FRC density ( Fig 4E ) , while FRC sphericity was decreasing ( Fig 4F ) . It is most likely that these morphological changes are a consequence of FRC relaxation by which the cells compensate for the loss of neighboring cells or the need to cover more space [7 , 10 , 37] . Interestingly , minimal distances between neighboring FRCs substantially increased when cell loss was higher than 70% ( Fig 4G ) . Moreover , connectivity between FRCs was almost completely lost at DT doses >2 ng/g ( Fig 4H ) , suggesting that the FRC network had been substantially disintegrated . Topological network analysis confirmed that a distinct threshold for FRC network integrity exists , as the network structure was destroyed when more than 70% of the cells were ablated ( Fig 5A ) . Interestingly , the number of nodes ( Fig 5B ) and edges ( Fig 5C ) dropped substantially when only 37% of the FRCs were ablated . However , other network parameters such as the number of edges per FRC ( Fig 5D ) and the local clustering coefficient ( Fig 5E ) were not profoundly altered at the DT dose of 0 . 5 ng/g . Likewise , small-worldness as determined by the σ factor was not significantly affected when the FRC network was mildly perturbed by the low-dose DT injection , while >50% FRC loss ( i . e . , DT doses of 1 ng/g and 2 ng/g ) resulted in a substantial change in this network parameter ( Fig 5F ) . It appears that the ω factor is not sensitive to strong alterations in the FRC network introduced by partial node removal ( Fig 5G ) , suggesting that the FRC network remains preferentially latticed . Nevertheless , the topological analysis based on increasing FRC ablation indicates that the essential FRC network features remain stable when <40% of the cells are removed , while an ablation of >70% of FRCs results in complete network failure . Network failure occurs when nodes lose their function in a random fashion or as a consequence of targeted destruction of particular nodes . Importantly , both network topology and the nature of node loss determine the robustness of the network [14] . Here , we reasoned that rapid loss of LN FRCs , e . g . , during a viral infection [4] , occurs in a random fashion . Likewise , we considered DT-mediated removal of FRCs in the Ccl19idtr model as arbitrary . Therefore , we first performed an in silico perturbation analysis by sequentially removing nodes from the FRC network model in a randomized manner ( S2 Video ) . Network fragmentation kinetics were followed during removal of nodes and their associated connections , in order to evaluate the topological properties of the residual network at each step ( Fig 6A ) . As nodes are removed , network fragments are generated ( blue ) that are disconnected from the largest cluster of nodes ( green ) ( Fig 6A ) . For each 3-D-reconstructed FRC network , 1 , 000 simulations of randomized node removal were performed ( Fig 6B and 6C ) . These datasets permitted estimation of the network integrity threshold across all fractions of nodes removed , corresponding to the maximal value of average shortest path length of the largest cluster ( Fig 6B ) . The analysis revealed that the network started to lose the characteristic path length when approximately 50% of the nodes were removed ( Fig 6B ) . Note that 50% node removal corresponds to FRC ablation with DT doses between 0 . 5 and 1 ng/g DT , which lead to reduction of FRC numbers by 37% and 67% , respectively ( Fig 4C ) . In addition , we determined the fragmentation curve as the relative size of the largest cluster compared to the size of the starting network and fraction of nodes removed ( Fig 6C ) . In this type of analysis , a network will have higher robustness the closer the curve is to the minimal damage line ( Fig 6C , dashed line ) . The perturbation analysis demonstrated that the FRC network exhibits high robustness to random node removal , indicated by a robustness value R of 0 . 439 ( Fig 6C ) . Note that the estimated network robustness ranges between maximal vulnerability ( R = 0 ) and maximal robustness ( R = 0 . 5 ) . The network robustness for all phosphate-buffered saline ( PBS ) -treated controls was estimated 0 . 437 ± 0 . 005 , n = 6 mice ( Fig 6D ) . Importantly , the topological model predicts that network robustness is not significantly reduced when 37% of the FRCs are ablated , while ablation of >50% of FRCs , i . e . , at doses of 1 and 2 ng/g DT , will lead to a significant reduction of network robustness ( Fig 6D ) . Collectively , the in silico model predicts that the FRC network displays significant topological robustness against random node removal and is able to tolerate up to half of the network being destroyed . LNs control distribution of immune cells in the body by attracting lymphocytes and myeloid cells via afferent lymph and blood . In addition , cellular content in the LN is influenced by cell release into efferent lymph [38] . To assess how LN FRCs impinge on the immune cell content of LNs , we determined the numbers of CD45+ hematopoietic cells ( Fig 7A ) , CD8+ T cells ( Fig 7B ) , and CD11c+ DCs ( Fig 7C ) following graded FRC ablation . Interestingly , ablation efficacy <40% ( i . e . , at 0 . 5 ng/g DT ) did not lead to significantly reduced cell numbers , while FRC ablation above 70% precipitated profound changes in LN cellularity ( Fig 7A–7C , S3A and S3B Fig ) . Plotting FRC density against hematopoietic cell numbers under conditions of graded FRC depletion revealed a clear dependence of immune cell aggregation in LNs on FRC network integrity ( Fig 7D ) . Next , we determined whether and to what extent intranodal T cell migration depends on the presence of FRCs . To this end , TCR transgenic CD8+ T cells [39] were adoptively transferred into DT-treated Ccl19idtr mice , and T cell behavior was assessed by two-photon microscopy ( S3 and S4 Videos ) . Cell tracking analysis revealed comparable T cell speeds and arrest coefficients with DT doses of ≤1 ng/g . In contrast , a significant decrease in T cell speeds was observed at DT doses of ≥2 ng/g , with a concomitant increase in cell arrest ( Fig 7E and 7F ) . Accordingly , T cell tracks exhibited decreased motility coefficients ( Fig 7G ) , a measure of scanning efficacy , and decreased meandering index ( S3C Fig ) , a measure of movement straightness , when >70% of FRCs were ablated . Overall , analysis of these data indicated that substantial changes in intranodal T cell migration occurred when >70% of FRCs were lost , i . e . , at DT doses of ≥2 ng/g . To assess how FRCs affect DC-mediated activation of antiviral CD8+ T cells , we resorted to a viral vector system that facilitates exclusive in vivo targeting of DCs [40 , 41] . Propagation-deficient coronavirus particles were injected subcutaneously into FRC-depleted mice , and the activation of antiviral CD8+ T cells was assessed in draining LNs . As shown in S3D and S3E Fig , T cell receptor transgenic Spiky cells were closely associated with the FRC network . Strikingly , T cell expansion was highly dependent on the presence of an intact FRC network because an ablation of >50% of FRCs resulted in an almost complete failure to expand the antiviral T cell population ( Fig 7H ) . Labeling of the CD8+ T cells with an intracellular dye revealed that proliferation of the cells was substantially affected at DT doses of ≥1 ng/g ( Fig 7I ) , suggesting that activation of naïve CD8+ T cells by DCs can be maintained as long as approximately 50% of the FRC network remains intact . The high correlation between topology and biological function as shown in Fig 7D ( r2 = 0 . 9448 , p = 0 . 00117 ) prompted us to further assess overall correlation between FRC morphology , topology , and function . As shown in the heat map in S4 Fig , most parameters are highly correlated with increasing doses of DT ( Pearson r > 0 . 8 ) , indicating that they are decreasing with declining FRC numbers . Four parameters showed high anticorrelation , namely arrest coefficient , cell surface area , volume , and cell-to-cell distances , due to their increase with decreasing numbers of FRCs . Only the omega factor did not significantly correlate with any other parameter as it is not sensitive to DT treatment ( Fig 5G ) . Overall , this analysis demonstrates the intricate connection between LN functionality and FRC topology .
Phenotypical characteristics of biological systems arise from complex interactions between cells that are orchestrated in a highly organized spatial and timely manner . Hence , it is a major challenge to understand the structure and dynamics of cellular networks and infer the function of particular tissues and organs . Results of the present study show that the physical scaffold of LNs formed by FRCs is critical for the maintenance of LN functionality . It is conceivable that the structure of the FRC network optimizes area/volume scanning by T cells by improving accessibility to distant regions [42] . Moreover , recent findings suggest that FRCs regulate the motility of DCs through PDPN- ( C-type lectin-like receptor 2 ) CLEC2 interaction [43] . Our results are in line with a previous study that demonstrated profound effects of complete FRC network ablation on T and B cell activation [7] . However , the complete destruction or ablation of components does not reveal the extensive complexity of a system and the role of specific components in its robust performance . In particular , global systems parameters such as topological organization and robustness need to be considered in order to design strategies for system modulation or regeneration . The theory of complex networks , i . e . , graph theory , offers a novel conceptual framework for biological systems and can be used as a powerful tool to dissect the quantifiable patterns of interaction between cells and the structure–function relationship of biological systems [13 , 44] . The present graph theory-based analysis revealed that LN FRCs form a small-world network with lattice-like properties . These properties were fully restored following complete removal of all FRCs , indicating that the basic FRC network topology with substantial connectivity and high capacity for clustering is an imprinted structural trait . It is possible that FRC network regeneration is guided by collagen fibers that are produced by FRCs [45] and remain visible after FRC ablation [7] . Assessing the interdependence between FRCs and the collagen fiber network will reveal further basic principles of LN organization and functionality . Our topological analysis was restricted to representative samples of T cell zone FRCs , mainly because a substantial part of the LN FRCs can be found in multilayered sheaths around blood and lymph vessels [10] . Even high-resolution confocal microscopy did not provide the means to separate single perivascular FRCs and to assess their morphology and topology . However , the methodology applied in our study is suitable to assess FRC network topology in B cell zones where distinct subsets such as C-X-C motif chemokine 12 ( CXCL12 ) -producing FRCs [9] control B cell migration . We envision that utilization of extended-volume imaging systems [46] or selective plane illumination microscopy [47] at high resolution will provide means to achieve an extended topological analysis of the LN FRC network . Nevertheless , the topological model based on samples of the T cell zone FRC network , as applied here , predicted with high accuracy the functional consequences of FRC loss , indicating that the sample area was adequately large to infer the behavior of the whole network . The morphological and topological parameters generated here will help to further advance development of mathematical LN models and could stimulate further research in modeling cell migration and fluid transport phenomena in other SLOs . Several modeling approaches have been described that focus on the description of processes that occur in LN subcompartments such as DC-driven T cell migration [48] or differentiation of Th cell subsets [49] . However , in order to obtain a more holistic view on LN functionality , the complexity of multilayered processes needs to be captured in mathematical models , an endeavor that requires representation of the whole LN [50] . Simple models have addressed this challenge by symbolizing the basic structural elements ( either in 2-D or 3-D ) as a regular orthogonal lattice [51 , 52] . Such agent-based models can describe the behavior of a variety of different cell types under the provision of distinct rules for their interaction . In addition , hybrid approaches have utilized ellipsoid , 3-D lattice models combined with agent-based modeling of immune cell interaction that facilitated simulation of antigen encounter under inflammatory conditions [53] . Clearly , steadily increasing computer power combined with novel imaging techniques provides a wealth of information describing immune cellular location [54] and principles of structural organization such as the LN vasculature [46 , 55] . Hence , it will be possible to generate extended mathematical models that describe multiple , interdependent immune reactions in LNs based on realistic geometry . Our study demonstrates that graph theory-based analysis of LN structures such as the FRC network not only provides important information on basic organization principles but also facilitates accurate prediction for the outcome of immune reactions . This suggests that the R index of the FRC network can be considered as a biologically relevant and consistent measure of robustness with global functional implications in the immune system . The suitability of this approach for in-depth analysis of critical biological processes has been shown in studies on neuronal networks [44] . Interestingly , neurons form—as FRCs—physically connected small-world networks that determine the function of the whole organ [56] . It is possible that these physical , non-random networks might have developed under evolutionary pressure to establish their structure and achieve optimal functionality . Overall , we anticipate that implementation of graph theory-based approaches in the investigation of those cellular elements that determine LN structure and functionality will fill the gaps in the understanding of critical immune processes . Moreover , generating improved mathematical models that permit prediction of complex system behavior will promote the development of rationally designed immune therapies and impinge on therapeutic intervention in diseases with involvement of immune system components .
C57BL/6N ( B6 ) mice were purchased from Charles River Laboratories . BAC-transgenic C57BL/6N-Tg ( Ccl19-Cre ) 489Biat ( Ccl19-Cre ) [5] crossed to iDTR mice [57] and TCR transgenic mouse strain C57BL/6N-Tg ( Tcra , Tcrb ) 577Biat ( Spiky ) [39] have been previously described . DT was applied at days −5 and −3 via IP injection at the indicated doses following established protocols [7] . All mice were maintained in individually ventilated cages and were used at the age of 6 to 9 wk . Experiments were performed in accordance with federal and cantonal guidelines ( Tierschutzgesetz ) under the permission numbers SG13/05 and SG13/04 following review and approval by the Veterinary Office of the Canton of St . Gallen and under permission BE48/11 granted by the Veterinary Office of the Canton of Bern . For flow cytometric analysis of LN cellularity , inguinal LNs from individual mice were pooled and digested on 37°C in RPMI containing 2% FCS , 20 mM Hepes ( all from Lonza ) , 1 mg/ml Collagenase Type P ( Sigma-Aldrich ) , and 25 μg/ml DNaseI ( AppliChem ) for 20 min . After enzymatic digestion , cell suspensions were washed with PBS and stained using the following antibodies: CD3-PE ( BD Bioscience ) , CD8-PeCy7 , CD4-FITC , CD45-APC-H7 , MHCII-PE , CD11c-PeCy7 , B220-APC ( BioLegend ) . In flow cytometric analyses , 7-amino-actinomycin D ( 7AAD; Calbiochem ) was used to discriminate dead cells . Samples were analyzed by flow cytometry using a FACSCanto flow cytometer ( BD Biosciences ) ; data were analyzed using FlowJo software ( Tree Star ) . Single-cell suspensions from spleens were prepared by mechanical disruption of the organ and subjected to hypotonic red blood cell lysis . For in vivo proliferation , splenocytes were labeled using CFSE or intracellular dye Alexa-670 ( Molecular Probes ) according to the manufacturer’s protocol , and 2 x 107 cells ( corresponding to 2 x 106 CD8+ TCR transgenic T cells ) were transferred intravenously ( IV ) into FRC-depleted recipient mice . Twelve hours post adoptive transfer , the mice were subcutaneously injected with 3 x 106 of non-replicating coronaviral particles in both flanks . A second injection of non-replicating coronaviral particles was performed 12 h following the first one . After 72 h from the first viral particle injection , inguinal LNs from individual mice were collected and analyzed using FACS . LNs were fixed overnight at 4°C in freshly prepared 4% paraformaldehyde ( Merck-Millipore ) under agitation and subsequently washed in PBS for one additional day . Fixed tissues were embedded in 4% low melting agarose ( Invitrogen ) in PBS and sectioned with a vibratome ( VT-1200; Leica ) . Forty μm sections were blocked in PBS containing 10% FCS , 1 mg/ml anti-FcRγ ( BD ) , and 0 . 1% Triton X-100 ( Sigma-Aldrich ) . Sections were incubated overnight at 4°C with the following antibodies: anti-gp38/PDPN , anti-B220 ( Biolegend ) , and anti-YFP ( Clontech ) . Unconjugated antibodies were detected using Alexa-fluor labelled secondary antibodies ( Jackson Immunotools ) . To visualize nuclei , sections were stained with 4′ , 6-diamidin-2-phenylindol ( DAPI ) ( Sigma-Aldrich ) . Microscopic analysis was performed using a confocal microscope ( LSM-710; Carl Zeiss ) , and the datasets were processed with ZEN 2010 software ( Carl Zeiss ) . Three-dimensional reconstructions of the T cell zone FRC network were performed using Imaris ( Bitplane ) . One to two T cell zones were acquired per LN per mouse by confocal laser scanning microscopy . The total surface area and the volume of the EYFP+ FRC network were calculated using the Surfaces module by reconstructing the FRC network in 3-D with an automatic threshold for fluorescent intensities and surface area detail of 0 . 3 μm . In order to remove background noise and cell fragments , a volume filter <10 μm3 was used . Single-cell morphometric analysis was used to calculate morphological parameters for FRCs . Single FRCs were isolated as separate 3-D Surface objects by using the cutting tool in the middle of the connected protrusions . DAPI staining was utilized by masking it to the EYFP channel in order to identify cell nuclei belonging specifically to FRCs and determine number of FRCs per imaged T cell zone . FRCs that were cut at the dataset borders encompassing more than half of the central body and diving and apoptotic cells , as well as perivascular FRCs , were excluded from further analysis . Centers of mass for each FRC were calculated using the Surfaces module , and minimal distances between single FRCs were determined using the Spots module and "Spots to Spots Closest Distance" XTension in Imaris . Sphericity was calculated in Imaris , indicating how spherical a 3-D object is . The compactness measure was calculated as ( area3/volume2 ) , which is minimized by a sphere . Connected protrusions per FRC were determined by utilizing the EYFP and PDPN channels , counted before the first branching point and connected to another FRC . Detailed information about morphological parameters is available in S1 Table . The topological model of the FRC network structure was created as an undirected , unweighted graph with no isolates in Imaris by defining nodes as the EYFP+ FRC centers of mass and edges as physical connections between neighboring FRCs . PDPN was utilized in order to visualize cell-to-cell connections more accurately . Topological analysis of the network was performed using the igraph package in R and RStudio . Small-world organization of the network was determined according to σ and ω parameters as described in [33–35] . For the calculation of the small-world parameters , the values of shortest path length and clustering coefficient were averaged over 100 realizations of an equivalent Erdos-Renyi random network for each FRC network dataset per mouse . Detailed information about topological parameters is available in S1 Table . Network perturbation analysis was performed using the igraph package and procedures as described in [14] , in order to assess error tolerance by sequentially removing increasing number of nodes randomly from the network . Network robustness was estimated using the R parameter [58] , and the threshold point was determined at the maximal value of the network average shortest path length as fractions of nodes are removed . Because of randomized node removal , the perturbation analysis was performed for 1 , 000 simulation runs for each FRC network dataset per mouse . 3 x 106 CellTracker Orange/CMTMR- or CellTracker Blue/CMAC-labelled P14 TCR transgenic T cells [59] were IV transferred into sex-matched Ccl19idtr mice , which had received two injections of indicated dose of DT or PBS 3 and 5 d prior to T cell injection . Three to twenty-four hours after T cell injection , the right popliteal LNs of the recipient mice were surgically exposed as described previously [60] . Prior to image acquisition , 10 μg of AlexaFluor 633-labeled Meca79 antibody was injected IV in order to label HEVs . Each image sequence lasted for 30 min . Acquired 3-D time-lapse images were tracked using Imaris software with Spot and ImarisTrack function . Average single cell speeds were calculated from 3-D coordinates of tracked cells using Matlab [61] . T cells attached to HEVs were not included in the analysis . Arrest coefficient , motility coefficient , and meandering index were calculated as summarized in S1 Table . One-way ANOVA or a Kruskal-Wallis test was used for all multiple group comparisons . Post-tests are indicated in figure legends . Numerical data and statistical analyses of all figures are available in S1 Data . Differences with a p-value < 0 . 05 were considered statistically significant . GraphPad Prism 5 was used for all statistical analyses . | Fibroblastic reticular cells ( FRCs ) in lymph nodes are organized in a highly connected cellular network that not only acts as a scaffold for lymphocyte migration but also provides key factors for induction and maintenance of immune responses . By utilizing high-resolution microscopy coupled with computational approaches to complex network analysis , we determined the topological properties and robustness of the FRC network . The underlying structure of the FRC network has been identified as a small-world network analogous to many other biological networks . Moreover , we demonstrate that this distinct structural organization is an imprinted trait of the FRC network , which is capable of fully regenerating after complete FRC ablation . In silico perturbation analysis of the FRC network confirmed that lymph nodes are able to tolerate FRC loss of approximately 50% . In vivo experiments corroborated these findings by demonstrating substantial impairment of immune cell recruitment , migration , and dendritic cell-mediated activation of antiviral CD8+ T cells , after critical loss of FRCs . In conclusion , the present study reveals the extraordinary topological robustness of the FRC network , crucial for establishing effective immunity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"blood",
"cells",
"flow",
"cytometry",
"infographics",
"cell",
"motility",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immunology",
"developmental",
"biology",
"routes",
"of",
"administration",
"luminescent",
"proteins",
"yellow",
"fluorescent",
"protei... | 2016 | Topological Small-World Organization of the Fibroblastic Reticular Cell Network Determines Lymph Node Functionality |
Currently , the efficacy of syphilis treatment is measured with anti-lipid antibody tests . These can take months to indicate cure and , as a result , syphilis treatment trials require long periods of follow-up . The causative organism , Treponema pallidum ( T . pallidum ) , is detectable in the infectious lesions of early syphilis using DNA amplification . Bacteraemia can likewise be identified , typically in more active disease . We hypothesise that bacterial clearance from blood and ulcers will predict early the standard serology-measured treatment response and have developed a qPCR assay that could monitor this clearance directly in patients with infectious syphilis . Patients with early syphilis were given an intramuscular dose of benzathine penicillin . To investigate the appropriate sampling timeframe samples of blood and ulcer exudate were collected intensively for T . pallidum DNA ( tpp047 gene ) and RNA ( 16S rRNA ) quantification . Sampling ended when two consecutive PCRs were negative . Four males were recruited . The mean peak level of T . pallidum DNA was 1626 copies/ml whole blood and the mean clearance half-life was 5 . 7 hours ( std . dev . 0 . 53 ) . The mean peak of 16S rRNA was 8879 copies/ml whole blood with a clearance half-life of 3 . 9 hours ( std . dev . 0 . 84 ) . From an ulcer , pre-treatment , 67 , 400 T . pallidum DNA copies and 7 . 08x107 16S rRNA copies were detected per absorbance strip and the clearance half-lives were 3 . 2 and 4 . 1 hours , respectively . Overall , T . pallidum nucleic acids were not detected in any sample collected more than 56 hours ( range 20–56 ) after treatment . All patients achieved serologic cure . In patients with active early syphilis , measuring T . pallidum levels in blood and ulcer exudate may be a useful measure of treatment success in therapeutic trials . These laboratory findings need confirmation on a larger scale and in patients receiving different therapies .
Syphilis is a multi-stage disease caused by the spirochete Treponema pallidum subspecies pallidum ( T . pallidum ) which is transmitted both sexually and from mother-to-child [1] . From the late 1990s , the United Kingdom recorded large ( ten-fold or more ) increases in diagnosed cases during epidemics characterised by male-to-male transmission and high rates of HIV-1 co-infection ( currently 64% in the UK ) [1 , 2] . Globally , the disease has remained prevalent with an estimated incidence of 10 . 6 million cases each year [3] . This resurgence , and the need to reduce mother-to-child transmission , has renewed interest in improved testing and treatment strategies . The aim of the current study was to develop a novel way of safely investigating syphilis treatments by directly measuring bacterial clearance following treatment . The diagnosis of syphilis relies on serological tests; direct visualisation by dark ground microscopy ( DGM ) or PCR detection in infectious lesions [4] . Following direct contact with an infected lesion , T . pallidum invades at the point of entry to produce the chancre of primary syphilis and disseminates widely through the blood-stream and lymphatics , leading to the multi-system secondary stage [5] . The infection is subject to significant but incomplete immune clearance during the early stages and , if untreated , passes into an asymptomatic latent phase [6] . Monitoring treatment can be achieved clinically , by observing the resolution of symptoms and signs , and serologically by measuring a two-dilution ( four-fold ) fall in the titre of anti-cardiolipin tests , namely the venereal disease research laboratory ( VDRL ) or rapid plasma reagin ( RPR ) assays [7] . The requisite fall in RPR titre may not be observed for several months as seen in a recent treatment study where a fifth of the ( HIV-1 uninfected ) patients were termed ‘serofast’ at six months [8] . This creates uncertainty about the adequacy of treatment , especially as only a small proportion ( 13% ) of these patients responded ( serologically ) to re-treatment with benzathine penicillin [9] . In recent years , PCR has increasingly replaced culture for the identification of pathogens . Quantitative PCR ( qPCR ) is commonly used to quantify viruses , such as HIV-1 and Hepatitis C in order to monitor treatment response . The quantification of bacteraemia by PCR not widely reported , but has been described for Streptococcus pneumoniae , Neisseria meningitidis , methicillin-resistant Staphylococcus aureus ( MRSA ) , and Acinetobacter baumannii [10–13] . It was suggested for these infections that bacterial load is associated with disease severity and that the rate of clearance may predict clinical outcome . A recent meta-analysis of 46 studies found the sensitivity of T . pallidum PCR to be highest in primary genital or anal chancres ( 78 . 4% , 95% CI 68 . 2%–86 . 0% ) . In blood , the highest sensitivities found were for congenital ( 83 . 0% , CI 55 . 0%–95 . 2% ) and secondary ( 52 . 2% , CI 37 . 3%–66 . 7% ) disease [14] . For syphilis too , then , bacteraemia may be associated with disease activity . Quantitative measurement of T . pallidum in whole blood established a range of 195 to 1954 polA gene copies/ml in one cross-sectional study of secondary disease , and a mean of 516 tpp047 copies/ml whole blood in another [15 , 16] . Until now , serial measurements following treatment have not been made , thus a potential correlation between bacterial clearance and serologic cure was unknown . Using qPCR , we measured T . pallidum bacterial load in blood and ulcer exudate samples from four patients with early syphilis both before and up to 150 hours after treatment with single-dose parenteral penicillin . We compared the rate of bacterial clearance with clinical outcome and standard serological follow-up .
Patients with a microbiologically confirmed diagnosis of primary or secondary syphilis were invited to participate . Primary disease was defined as a genital , or peri-anal chancre in which T . pallidum was identified by DGM . Secondary disease was diagnosed in patients with consistent symptoms and signs in the context of a positive enzyme immuno-assay ( EIA ) result and a RPR titre of greater than one in eight . Following informed consent , subjects were admitted to an in-patient facility and invited to donate 4 ml whole blood , of which 3 ml were collected into a Tempus RNA preservation tube ( Life Technologies , UK ) , and 1 ml into EDTA . Each patient then received 2 . 4 megaunits of benzathine penicillin by intramuscular injection . Subsequently , blood samples were drawn , as above , two and six hours post-penicillin administration , then four-hourly for 48 hours ( 11 samples ) ; six-hourly for 24 hours ( four samples ) and , finally , 12-hourly for up to four days ( maximum eight samples ) . Samples were placed on dry ice immediately following collection and were processed in batches of up to eight . For the first patient , samples were collected at all time-points . Subsequently , the duration of sample collection was informed by T . pallidum clearance and ceased when two consecutive samples were negative by qPCR . Ulcer exudate was collected with a filter paper Snostrip . First , the ulcer was abraded with sterile gauze and saline . Next , exudate from the ulcer margin was absorbed up to the notch of the strip , the tip removed with a sterile blade and placed into 1 . 3ml of RNAlater solution ( Ambion , UK ) . Following discharge , patients were invited to attend for RPR testing at one , three and six months post-treatment . Serological cure was defined as a two-dilution ( four-fold ) reduction in RPR titre . Ethical approval for the study was obtained from the NRES Committee South East Coast ( Brighton and Sussex ) ( ref . 11/LO/0358 ) . Informed , written consent was obtained from all study participants . Within 24 hours of collection , T . pallidum DNA was extracted from 500μl aliquots of whole blood . Aliquots were added to 500μl of lysis buffer ( 20mM TrisHcl , pH 8 . 0; 0 . 2M EDTA; 1%SDS ) and incubated with 100μl of Proteinase K ( Qiagen ) for three hours at 65°C . Following extraction with a proprietary extraction method ( QiAmp , Qiagen , Crawley , UK ) , DNA was purified by ethanol precipitation and resuspended in 60μl of elution buffer ( Qiagen ) . RNA was extracted from blood in Tempus tubes according to the manufacturer’s protocol and resuspended in 100μl of the supplied elution solution . Positive and negative extraction controls were included with each batch . DNA and RNA from Snostrip samples were extracted in a single procedure with an ‘all-prep mini kit’ ( Qiagen ) . Snostrips in RNAlater were centrifuged at 14 , 000g and 4°C for 30 minutes and the upper 1 . 1ml of RNA later discarded . The remaining 200μl were added to 600μl or buffer RLT ( Qiagen ) which contained 0 . 143M beta-mercaptoethanol . Following homogenization with a Qiashredder column ( Qiagen ) the resultant DNA/RNA mixture was separated and extracted according to the manufacturer’s protocol . RNA and DNA were eluted into 60μl and 100μl of RNAse-free water , respectively . T . pallidum DNA quantification was achieved with qPCR using the CFX real-time system ( Biorad ) to detect amplification of the tpp047 gene , which encodes a conserved 47 kDa outer membrane protein . Each reaction contained 12 . 5μl 2x Quantitect PCR mix ( Qiagen ) ; 0 . 4μM of the primers TP1 ( 5-CGAGGAATACAAGATTACGAACG-3 ) and TP2 ( 5-ACGTGCAGAAAAACTATCCTCAG ) and 0 . 2μM of the hydrolysis probe TP_P ( FAM-CGGCCTCGCTCAGAGATGAGC-TAMRA ) . Primer specificity had been assessed previously [15] . Cycling conditions consisted of 15 minutes at 95°C then 44 cycles of 95°C for 15 seconds and 80 seconds at 60°C . Quantification was achieved using an ‘in-run’ plasmid standard of a 1:10 dilution series from 105 to ten copies per reaction of the tpp047 target sequence . RNA quantification was performed in a single-step RT-qPCR . Each 25μl reaction contained 1μl enzyme mix ( reverse transcriptase and taq polymerase ) ( Qiagen ) ; 0·4mM DNTP mix; 5μl reaction buffer; 0 . 6μM of the primers 16S_F ( CTCTTTTGGACGTAGGTCTTTGAG ) and 16S_R ( TCACCCTCTCAGGTCGGATA ) , and 0 . 2μM of the hydrolysis probe 16S_P3 ( FAM-CGGCCTCGCTCAGAGATGAGC-TAMRA ) . Cycling conditions began with 50°C for 30 minutes , followed by 95°C for 15 minutes then 44 cycles of 95°C for 15 seconds and 60°C for 90 seconds . Absolute RNA quantification was achieved with a 1:10 dilution series of T . pallidum 16S rRNA produced using T7 polymerase on the recombinant plasmid containing the target sequence . For both DNA and RNA quantification , samples were analysed in triplicate and two no-template ( water ) controls were included in each experiment . Results were analysed using SPSS v19 . Statistical analysis is limited to descriptive statistics and a two-tailed student’s t-test .
The T . pallidum qPCR was found to have an analytical sensitivity of at least ten tpp047 copies/reaction . The mean inter and intra-assay coefficients of variation ( for the detection of 104 tpp047 copies/μl ) were 2 . 35% ( std . dev . 0·63 ) and 0·66% ( std . dev . 0·17 ) , respectively . The quantification standard curve was 103% efficient with an average of 3 . 17 cycles between each 1:10 dilution . The RT-qPCR could detect a minimum of 23 16S rRNA copies/reaction with an inter-assay variation coefficient ( for the detection of 2 . 3x104 copies ) of 3 . 05% ( std . dev . 0 . 87 ) and intra-assay variation of 0 . 34% ( std . dev 0 . 09 ) . The 16S rRNA absolute quantification standard curve was 87 . 7% efficient with an average of 3 . 66 cycles between each 1:10 dilution . To compare the kinetics of T . pallidum DNA and rRNA clearance , two parameters were calculated . The first , time to clearance , was the time elapsed between administration of treatment and the point half way between the first of the consecutively negative time-point and the last detectable time-point . The second , clearance half-life , was based on the exponential decay equation: N ( t ) =N0 ( 12 ) t/t12 N ( t ) is the quantity remaining after a time t; N0 is the initial quantity , and t1/2 is the half-life of the decaying quantity . By measuring the coefficient of variation ( R2 ) for an exponential regression line , the rate of nucleic acid clearance in both ulcer and blood samples was found to be exponential , thus half-life was an appropriate measure of clearance . Three of four patients recruited were found to have detectable T . pallidum bacteraemia pre-treatment and it was possible to document changes in the bacterial load over time . The fourth patient recruited was bacteraemic at time points four , six , eight and nine ( between six and 30 hours post-treatment ) , but initially negative and was excluded from analyses . Baseline characteristics for all patients are presented in Table 1 . Samples from patients STS1 , STS2 , and STS3 showed similar patterns of T . pallidum nucleic acid clearance from blood following treatment ( Table 2 and Fig . 1 ) . The mean peak tpp047 level , which occurred two to ten hours post-treatment , was 1626 copies/ml whole blood ( std . dev . 652 ) . Peak 16S rRNA levels ( 8879 copies/ml whole blood , std . dev . 11109 ) were detected at similar time-points . While the peak of tpp047 DNA was similar for all three patients , that of 16S rRNA for patient STS3 was ten-fold higher than for STS1 and STS2 . Following peak bacteraemia , both RNA and DNA levels fell quickly , with neither detectable in any patient after 56 hours . The mean time to tpp047 DNA clearance was 34 hours ( std . dev . 8 ) and 29 hours ( std . dev . 23 . 86 ) for 16S rRNA . The 16S rRNA time to clearance measured for patient STS3 was 56 hours , as a result of low-level ( 35 copies/ml ) 16S rRNA detection at 56 hours following two negative samples at 42 and 46 hours . This 16S rRNA amplification at 56 hours was observed in two of three technical replicates during two separate experiments . Furthermore , both the negative Tempus™ extraction control and the two no-template PCR negative controls were negative . The mean half-life of T . pallidum tpp047 DNA clearance from blood was determined to be 5 . 68 hours ( std . dev 0 . 53 ) and was significantly longer than the 3 . 89 hours ( std . dev . 0 . 84 ) calculated for 16S rRNA ( p = 0 . 035 , two-tailed t-test ) . In order to exclude the possibility of a late recrudescence in bacterial load , blood from patient STS1 was sampled for both DNA and RNA quantification until 150 hours post-treatment . Neither was detectable following initial clearance at 32 ( RNA ) and 40 ( DNA ) hours . Patient STS3 presented with a DGM positive genital ulcer and samples were collected concurrently with each blood sample . Table 2 and Fig . 2 demonstrate an initial ulcer tpp047 load of 34 , 000 DNA copies/strip , which peaked ( 67 , 400 DNA copies/strip ) at 14 hours and remained high until 26 hours , thereafter decreasing rapidly . At 50 hours following treatment , tpp047 DNA was no longer detectable . The decay of tpp047 DNA fitted an exponential regression line ( R2 = 0 . 878 ) and the clearance half-life was 3 . 2 hours . The quantity of 16S rRNA followed a similar pattern , although this peaked earlier , at two hours post-treatment , and higher , at a level 1000 times greater than for DNA ( 7 . 08x107 copies/strip ) . After 40 hours , the level fell exponentially ( R2 = 0 . 839 ) to become undetectable by 56 hours . Clearance half-life was 4 . 1 hours . The clearance of both DNA and 16S rRNA was measured both to improve T . pallidum detection ( the 16S ribosomal RNA target was predicted to be present at 5000–10000 copies per organism ) and to compare clearance , predicted to be shorter for RNA which is inherently less stable [17] . The half-life of T . pallidum DNA clearance was found to be significantly longer than 16S rRNA ( 5 . 7 hours versus 3 . 9 hours , p = 0·035 ) . Patients STS1 , STS2 and STS3 all developed an inflammatory Jarisch-Herxheimer reaction ( JHR ) eight to 12 hours after penicillin administration . This included a worsening of rash for patients STS1 and STS2 over a one-to-two day period . At one month , clinical signs and symptoms had resolved in all three patients and serological results were consistent with cure ( Table 3 ) . This serological response was maintained at three and six months .
We present the development of a qPCR assay for quantifying T . pallidum in clinical samples and some early observations from a small , uncontrolled clinical study of patients with early syphilis treated with Benzathine penicillin , which highlights the potential clinical application of the assay . Confirming adequate treatment for syphilis with serological tests can be slow ( up to 12 months ) and problematic due to the non-specific nature of serum RPR testing [7 , 8] . Using serial sampling and novel qPCR assays , we have demonstrated clearance of T . pallidum nucleic acids from blood and ulcer exudate within 56 hours of treatment with benzathine penicillin in four patients . We have also described the kinetics of this response and found the half-life for blood clearance of T . pallidum to be 5 . 7 hours for DNA and 3 . 9 hours for RNA . For an ulcer , bacterial DNA and RNA clearance half-lives were 3 . 2 and 4 . 1 hours , respectively . All patients were cured by their treatment , showing both a clinical and serologic response . Syphilis disease activity varies through the course of the infection giving rise to early ( symptomatic ) and late ( asymptomatic ) clinical stages . The secondary stage , characterised by systemic disease ( rash , hepatitis , neurologic involvement ) is arguably the most active and is the stage during which patients are most likely to be bacteraemic [14 , 18] . Blood bacterial load has been measured by qPCR in a number of infections and correlated with disease activity . In patients with meningococcal disease , the level of bacteraemia at admission was significantly higher in individuals with severe disease and in those who died [11] and in MRSA bacteraemia where mecA DNA levels were significantly higher in non-survivors compared with survivors [12] . When the bacterial load of Acinetobacter baumannii in critical care patients was followed longitudinally , a slower rate of clearance was associated with increased mortality . Moreover , the use of appropriate antibiotics resulted in quicker bacterial clearance [13] . The sensitivity of T . pallidum PCR in blood is too low for it to be a reliable diagnostic test , especially in primary and latent disease . However , our data suggest that in selected patients with active early disease it may prove a useful marker of disease activity . Moreover , in the context of a clinical study of early syphilis treatment , qPCR may also conceivably be used to monitor bacterial clearance and quickly identify potential treatment failure . The primary chancre of syphilis develops at the site of initial infection following bacterial division and invasion and , as such , has a high bacterial load . PCR detection of T . pallidum in these lesions is now established as a diagnostic tool for the disease . The chancre is also the most likely source of bacteria for onward transmission of the disease , thus the period of infectiousness following treatment is of interest . In the current study DNA and RNA from an ulcer were undetectable 50 and 56 hours following treatment , respectively . If proven in a larger clinical study , this may be of use for counseling patients regarding abstinence following syphilis treatment . To our knowledge , there is no previous description of T . pallidum qPCR detection in animal or human samples to measure treatment response following therapy . The intensity of monitoring in this study enabled the observed rapid bacterial clearance to be measured accurately and serologic follow-up of patients was robust enabling a potential association between bacterial clearance and serologic cure to be identified . At present , our demonstration of bacterial clearance is limited to the study of four patients and requires confirmation on a much larger scale to become clinically useable . Moreover , whilst the clearance of bacterial DNA was measured , it is unknown whether the DNA detected was from living or dead organisms . As T . pallidum is non-culturable , Rabbit Infectivity Testing ( RIT ) , where test animals are injected with clinical material in order to propagate and identify T . pallidum , could have been used to assess persistence following treatment [19 , 20] . This approach would , however , have required the sacrifice of a large number of animals . A final limitation is that all four patients in the current study received benzathine penicillin treatment . Whilst this was deliberate , in order to define the precise moment of treatment , we acknowledge that bacterial clearance time following treatment with oral antibiotics may be different . In recent clinical trials of syphilis treatments , enrolled patients waited for a minimum of six months following the administration of the trial drug before cure ( defined as a four-fold reduction in RPR titre ) was diagnosed [21 , 22] . The investigation of new syphilis treatments is much needed , but to leave patients potentially untreated for six months while waiting for a serologic response is far from ideal . We hypothesise that for patients with early stage syphilis , which is PCR-positive at baseline , a clearance half-life of less than six hours or absence of bacterial nucleic acids at three days indicates bacterial clearance and is potentially a new measure of treatment success . In order to prove this , a larger observational study of syphilis treatment response is required . This proposed study would require both HIV-1 uninfected and uninfected patients with untreated primary , secondary and early latent syphilis . We propose that five sample time-points based on these preliminary data ( pre-treatment , 12 , 24 , 56 and 72 hours post-treatment ) should be used to compare bacterial clearance following treatment with both parenteral penicillin and oral antibiotics , such as doxycycline . Bacterial clearance would be correlated with serologic and clinical cure . | Syphilis is an infection that is spread both sexually and from mother-to-child . Worldwide , it affects an estimated 11 million people each year . Treatment is available , but relies heavily on penicillin and may not be as effective where the infection involves the brain or nervous system . Clinical trials are needed to assess new treatments , but proving that people are cured with current tests can take months . We developed a new test to measure the success of treatment for the early stage of syphilis . We then used it in four patients and found that the syphilis bacteria were cleared from the blood and ulcer samples collected by 56 hours . All four patients were followed up with normal tests and found to have been cured . Our new test is the first to show the speed of bacterial clearance after treatment and we plan to use it in clinical trials of new treatments . Our data may also mean that patients with syphilis are unable to pass on the infection three days after treatment , but this also needs proving on a larger scale . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Rapid Treponema pallidum Clearance from Blood and Ulcer Samples following Single Dose Benzathine Penicillin Treatment of Early Syphilis |
Colicins are protein antibiotics synthesised by Escherichia coli strains to target and kill related bacteria . To prevent host suicide , colicins are inactivated by binding to immunity proteins . Despite their high avidity ( Kd≈fM , lifetime ≈4 days ) , immunity protein release is a pre-requisite of colicin intoxication , which occurs on a timescale of minutes . Here , by measuring the dynamic force spectrum of the dissociation of the DNase domain of colicin E9 ( E9 ) and immunity protein 9 ( Im9 ) complex using an atomic force microscope we show that application of low forces ( <20 pN ) increases the rate of complex dissociation 106-fold , to a timescale ( lifetime ≈10 ms ) compatible with intoxication . We term this catastrophic force-triggered increase in off-rate a trip bond . Using mutational analysis , we elucidate the mechanism of this switch in affinity . We show that the N-terminal region of E9 , which has sparse contacts with the hydrophobic core , is linked to an allosteric activator region in E9 ( residues 21–30 ) whose remodelling triggers immunity protein release . Diversion of the force transduction pathway by the introduction of appropriately positioned disulfide bridges yields a force resistant complex with a lifetime identical to that measured by ensemble techniques . A trip switch within E9 is ideal for its function as it allows bipartite complex affinity , whereby the stable colicin:immunity protein complex required for host protection can be readily converted to a kinetically unstable complex whose dissociation is necessary for cellular invasion and competitor death . More generally , the observation of two force phenotypes for the E9:Im9 complex demonstrates that force can re-sculpt the underlying energy landscape , providing new opportunities to modulate biological reactions in vivo; this rationalises the commonly observed discrepancy between off-rates measured by dynamic force spectroscopy and ensemble methods .
Protein-protein interactions are integral to diverse cellular processes such as catalysis , transport , and signalling . For complexes of low affinity , changes in the relative concentrations of one ( or more ) binding partners , or alterations in the environment , are sufficient to trigger complex dissociation , allowing spatial and temporal control of the processes in question . More stable complexes require the input of chemical energy such as that provided by AAA+ proteins for their dissociation [1] . For high affinity complexes without direct access to an energy source , it remains unclear how dissociation can be induced on a biologically relevant timescale . This problem is exemplified by the rapid dissociation ( lifetime≈minutes ) of highly avid bacterial colicin:immunity protein complexes ( Kd≈10−14 M; lifetime≈days ) upon binding to the outer membrane of target bacteria [2] . Colicins are protein antibiotics synthesised by E . coli strains to target and kill related bacteria during environmental stress [3] . The E-type colicins of group A , which include E2 , E7 , E8 , and E9 , exert their toxicity via nuclease activity . These multi-domain proteins ( Figure 1A ) contain a receptor domain ( R ) required for initial binding ( to BtuB ) , a translocation domain ( T ) used to bring about translocation via interaction with OmpF and a cytoplasmic DNase domain ( C , residues 480–582 of full length colicin , termed E9 herein ) that results in death of the competing cell subsequent to translocation of this domain to the victim's cytoplasm ( Figure 1B ) . To prevent host suicide , colicins are expressed alongside specific immunity proteins ( Im2 , Im7 , Im8 , and Im9 ) [2] , which inactivate colicin enzymatic activity by binding to an exo-site adjacent to the active site ( Figure 2A and 2B ) [4] . The binding interface is typical for protein-protein complexes , covering a surface area of 1 , 575 Å2 for E9:Im9 , which spans residues 72 to 98 of the nuclease domain [5] . The highly homologous cognate colicin:immunity protein pairs have high affinities ( Kd≈10−14 M ) [6] , while non-cognate pairs bind less tightly ( e . g . , for E9:Im2 , Kd≈10−7 M ) [7] . The binding affinity of colicin:immunity protein complexes is determined by two binding “hotspots” on the immunity protein that interact with a distinct binding epitope on E9 . Firstly for all colicin:immunity protein complexes , stabilising interactions are formed with residues in helix III of the immunity protein . This helix is identical in sequence in all E-type immunity proteins ( apart from one residue in Im7 [6] ) . Residues within helix III of the immunity protein contact Phe86 and residues in the surrounding hydrophobic pocket of E9 . This pocket comprises Tyr 83 , Val 98 , and the alkyl chains of Lys 89 and Lys 97 of the colicin DNase domain ( Figure 2B , highlighted in lilac ) . Colicin:immunity protein affinity is modulated by stabilising ( cognate complexes ) or destabilising ( non-cognate complexes ) interactions between specificity-determining residues of helix II in the immunity protein ( which differs significantly in sequence in different proteins; Figure 1C ) and the binding interface of the colicin DNase domain ( Figure 2 ) [8] , [9] . As the on-rate for cognate and non-cognate colicin:immunity protein complexes is diffusion-limited ( kon≈10−8 M−1s−1 ) [10] , [11] , the observed differences in affinity , which span almost ten orders of magnitude ( Kd = 10−14 ( E9:Im9 ) to 10−4 M ( E9:Im7 ) ) [7] , [8] , [11] are manifested in off-rates that differ by seven orders of magnitude ( koff = 10−6 to 101 s−1 for the cognate and non-cognate complexes , respectively ) . The large differences ( 107 ) in off-rates of different colicin:immunity protein complexes render this system an excellent model for investigating the molecular determinants of molecular recognition and , in particular , for exploring how highly avid complexes can be remodelled in vivo in the absence of an external energy source to allow rapid dissociation when required for biological activity . While the precise mechanism of E-type colicin:immunity protein dissociation is unclear , it is known that colicin invasion uses molecular mimicry to subvert a series of protein-protein interactions that result in linkage of the colicin ( bound to the outer membrane ) to the TolQRA complex of the energised inner membrane ( a translocon; Figure 1B ) . As TolQRA function and colicin intoxication both require a proton gradient across the inner membrane [12] , it has been postulated that the energy of the proton motive force ( PMF ) may be harnessed to drive colicin:immunity protein dissociation , a necessary prerequisite of translocation of the nuclease to the cytoplasm . “Inside-out” energy transduction mechanisms are exemplified by the Ton system , which is highly homologous to TolQRA ( both require a functioning PMF to carry out their function [13] ) . In the case of the Ton system , PMF-driven remodelling of the plug domain of the outer membrane protein BtuB allows siderophore import . Remodelling is also thought to play a role in E9 colicin intoxication as cross-linking residues 20 and 66 of the nuclease domain prevents insertion into planar lipid bilayers and protects against cellular toxicity [14] . In accord with a requirement for structural remodelling in the mechanism of colicin invasion , immunity protein release that is usually triggered by formation of the translocon in the presence of the PMF , is prevented by cross-linking of the N- and C- termini of the R-domain ( Figure 1B ) [15] . Here we use single molecule force methods ( using atomic force microscopy ) to investigate the requirement for structural remodelling in the dissociation of single E9:immunity protein complexes under defined rates of loading and pulling geometry . The effects of structural re-arrangements in proteins can be investigated by many approaches , but most apply a “peturbant” globally . Single molecule force methods ( which use mechanical extension as a peturbant ) are ideally suited for such an investigation as force is applied locally to the complex at positions determined by the sites of linker attachment . Using this approach , we show here that a low level of force ( <20 pN ) commensurate with that applied by protein molecular motors [16]–[18] increases the dissociation rate of the E9:Im9 complex in vitro by a remarkable 106-fold . Using mutagenesis and disulfide cross-linking , we also elucidate the force transduction path through E9 , which catalyses complex dissociation , and show that this involves conformational remodelling of E9 triggered by mechanical deformation of its terminal region . The data show that mechanical force can be exploited to enable rapid dissociation of the high affinity colicin:immunity protein interaction by application of force at the N-terminus of E9 .
We used atomic force microscopy to measure the dynamic force spectrum of the unbinding of single complexes of the nuclease C-domain of colicin E9 bound to its cognate immunity protein ( Im9 ) , together with non-cognate complexes of E9:Im2 and several variants of Im9 containing point mutations in the binding site [7] , [8] . The experimental design is depicted in Figures 1D and S1 . Briefly , single cysteine residues were introduced into E9 ( the wild-type protein lacks cysteine residues ) and pseudo–wild-type variants of Im9 and Im2 in which the single naturally occurring cysteine was first mutated to alanine ( C23A ) , to enable immobilisation of each protein specifically to the substrate or cantilever . Sites chosen for mutation to allow immobilisation ( S3C , S30C , or S108C in E9 and T38C or S81C in the immunity protein ) were solvent accessible and distal to the E9:immunity protein binding interface ( Figure 2A and 2B ) . The immunity protein and E9 were next attached to the atomic force microscope ( AFM ) tip and substrate , respectively , using hetero-bifunctional polyethylene glycol ( PEG ) linkers of variable length ( Materials and Methods; Figure S1 ) . Gel filtration was used to compare the ensemble off-rates of the wild-type complex and one containing mutated E9 derivatised with PEG linker ( kioff = 1 . 8×10−6 s−1 and 5 . 8×10−6 s−1 for wild-type [8] and derivatised [E9 S3C:Im9 ( S81C ) ] complexes [Figure 3A] , respectively ) . These data , together with a nuclease assay ( Materials and Methods; Figure 3B ) showed that neither sequence changes nor PEG derivatisation significantly affected the properties of E9 alone or in complex with Im9 . Complexes were repeatedly formed and dissociated by approach-retract cycles of the functionalised AFM tip towards and away from the surface at a defined velocity ( Materials and Methods ) . Unbinding resulted in a single force peak characteristic of a single molecule unbinding event ( Figures 4A , bottom , lower plot , and Figure S2 ) for 99 . 5% of all force-extension profiles that showed any evidence of interaction between the tip and substrate ( typically 10% of all approach-retract cycles ) . All force-extension data are presented and analysed after accounting for the deflection of the AFM tip ( i . e . , the distance between apex of the AFM tip and the substrate surface ) . Initially , force-extension profiles that displayed a detectable unbinding event greater than 5 nm from the surface ( to avoid non-specific tip-surface interactions , see Text S2 ) were analysed without further filtering . Figure 4B shows a scattergram contour plot and individual frequency histograms for the unbinding force and contour length at rupture for every unbinding event in each force-extension profile of a single dataset . These data show that unbinding events occurred over a narrow range of extensions ( mode = 10 nm ) suggesting that unbinding occurs by a single pathway ( Figure 4A , top ) . Interestingly , the measured contour length is significantly shorter than that expected ( Figure 4A bottom , black solid line ) based on the sum of the lengths of each linker ( 6 . 62 nm each ) and the through space distance between the extension points on E9 and Im9 . This distance is 4 . 72 nm when extending the complex from residue 3 on E9 and residue 81 on Im9 . This complex is denoted 3:81 ( similar nomenclature is used throughout ) . To predict the expected contour length more accurately , it is necessary to account for the ability of Im9 to be immobilised anywhere between the apex and the base of the AFM tip and the distribution of end-to-end lengths within the ensemble of the polymeric linkers ( Figure S3 ) . Taking these effects into consideration ( using a Monte Carlo simulation; Figure S4 and Text S1 ) yielded a contour length significantly shorter than observed ( 7 . 57 nm , Figure 4A bottom , red line ) . Importantly , these calculations suggest that the complex undergoes deformation or elongation prior to dissociation ( see Discussion ) . In order to quantify the unbinding forces and loading rates at rupture , force-extension profiles were subsequently analysed using an automated analysis script whereby events ( <13% of total force-extension profiles; ) were filtered from non-specific interactions on the basis of their force-extension profiles . To be binned for further analysis ( Text S2 ) a force-extension profile was required to ( i ) have a rupture force larger than the thermal noise of the experiment ( 18 pN; Figure S5 and Text S3 ) ; ( ii ) have a distance to the rupture event from the hard tip-surface contact that was between 5 and 32 . 5 nm or 5 and 40 nm for protein complexes immobilised using ( PEG ) 4 and ( PEG ) 12 , respectively . The lower limit avoids the analysis of any non-specific tip-sample interactions and the upper limit is significantly greater than the expected rupture distance so that all events are analysed; and ( iii ) display only a single unbinding event . The ability of the experimental setup and data analysis method to recognise only specific E9:Im9 unbinding events was verified using two controls . Firstly , the addition of excess immunity protein to the solution between the AFM tip and substrate resulted in a decrease of the frequency of unbinding events from more than 12% to less than 1% ( Figure 4B ) . Secondly , the addition of EDTA was found to decrease the event frequency 3-fold . EDTA sequesters divalent metal cations from E9 , destabilising the protein substantially ( Tm = 36°C and 68°C for apo- and zinc-bound E9 [19] ) leading to a loss of binding to Im9 . Addition of excess Zn2+ restored E9 stability and , consequently , event frequency ( unpublished data ) . The force and loading rate at unbinding of E9:Im9 were measured for each event from force-extension data ( Text S2 ) . Force-frequency distributions ( Figure S6 ) were subsequently calculated for each dataset ( typically 100–200 events; Table S1 ) , allowing the extraction of the most probable unbinding force ( Figure S6; Text S4 ) and the loading rate at rupture . The dynamic force spectrum of each complex was then revealed by quantifying how the force at rupture varies as a function of the force loading rate between 700 and 180 , 000 pNs−1 ( Text S5 ) . The dynamic force spectrum of E9:Im9 dissociation was initially measured by immobilising E9 close to the N-terminus ( residue 3 ) ( Figure 2 ) as this region is immediately adjacent to the R-domain and contiguous with the T-domain , which is translocated during colicin intoxication in vivo ( Figure 1B ) . No force is likely to be applied directly to the immunity protein in vivo . A pulling location was thus selected for Im9 ( residue 81 ) ( Figure 2 ) that is solvent exposed and distal to the binding interface . Sample force-frequency histograms that span the range of loading rates used ( 700–180 , 000 pNs−1 ) and the resultant dynamic force spectrum for this complex ( 3:81 ) are shown in Figures S6 and Figure 5A , respectively . Two force regimes are evident . At low loading rates ( <5 , 400 pNs−1 ) , dissociation occurs at low forces with a shallow dependence of unbinding force on the loading rate . This allows rapid dissociation of an avid complex at biologically accessible loading rates [16] , [17] , [20] , or by application of biologically accessible forces ( see [16]–[18] and references therein ) . For example , at a force of 20 pN , the lifetime of E9:Im9 is approximately 12 ms , in contrast to 4 . 1 d in the absence of force . At higher loading rates ( >5 , 400 pNs−1 ) the complex is highly force resistant and the unbinding force is strongly dependent on the loading rate . The simplest explanation for these observations is that unbinding occurs by a three-state mechanism: at low forces unbinding rates are limited by a barrier in the energy landscape distal to the bound ground state of the complex ( a large xu; Figure 5B , xuo ) . At higher forces , tilting of the energy landscape results in a previously hidden inner barrier ( a small xu ) becoming rate limiting ( Figure 5B , xui ) . Such a mechanism is consistent with the dual-recognition ( un ) binding pathway for E9:Im9 determined using ensemble fluorescence experiments ( Figure 5B , top ) [8] . In this mechanism , the affinity of the initial encounter complex is determined by interactions between residues of the E9 binding interface ( Figure 2 , highlighted in lilac ) and helix III of the immunity protein ( residues S50 , D51 , I53 , and Y55 ) [5] . Rigid body rotations of the initial encounter complex then allow the formation of stabilising ( cognate ) or less stabilising ( non-cognate ) interactions between E9 and specific residues in helix II of the immunity protein [7] . Accordingly , the outer barrier measured by DFS that is rate determining at low rates of forced unbinding is expected to report on the free energy difference between the native state and the barrier for dissociation of E9 from helix III of Im9 , while the inner barrier that is rate determining at high rates of forced unbinding is expected to report on the energy gap between the native state and the barrier for dissociation of E9 from helix II ( Figure 5B ) . To confirm the apparent similarity between the force- and thermally activated unbinding mechanisms of E9:Im9 , each linear region of the dynamic force spectrum was fitted to the Bell-Evans equation [21] ( Equation 1 , where f* is the most probable unbinding force , rf is the force loading rate at rupture , T is the temperature , and kB is Boltzmann's constant ) . This allows the dissociation rate constants in the absence of force ( k0Foff ) and the “distance” along the free energy landscape from the bound state to the barrier that is rate limiting for dissociation ( xu ) to be obtained ( Table S1 ) . ( 1 ) Obtaining k0Foff values by this method assumes that the outer barrier observed in the dynamic force spectrum remains rate-limiting at lower loading rates that are inaccessible to this technique ( see [22] , [23] for example ) . These parameters were found to be k0Foff = 50±17 s−1 , xu = 0 . 9±0 . 2 Å , and k0Foff = 4 . 9±1 . 3 s−1 , xu = 5 . 8±0 . 4 Å for the inner and outer barriers of E9:Im9 dissociation , respectively . If forced unbinding ( at the loading rates applied in these experiments ) occurs over the same energy landscape as for the thermally induced pathway , the extrapolated k0Foff for the outermost barrier ( the rate determining step at low force ) , determined by DFS should be identical to that measured by ensemble methods ( kioff ) [22] , [24]–[28] . Remarkably , forced unbinding of E9:Im9 results in a k0Foff that differs from kioff by a striking six orders of magnitude ( ≈100 and 10−6 s−1 , respectively ) . To examine whether the rapid rate of force-induced dissociation of E9:Im9 is observed for other E9:immunity protein complexes , the dynamic force spectrum of a non-cognate complex ( E9:Im2 ( D33A ) , kioff = 0 . 054 s−1 ) [7] was next examined ( Figure 6A ) . This variant was selected since it has a higher affinity for E9 compared with wild-type Im2 ( Kd = 1×10−9 M and 1 . 5×10−7 M , respectively ) [7] . Again two force regimes were observed in the dynamic force spectrum of this complex , each of which has a similar xu value to that observed for each barrier of the cognate E9:Im9 complex . This indicates that force-induced unbinding of the cognate and non-cognate complexes occurs by a similar three-state mechanism . At low loading rates that probe the rate determining outer barrier , the unbinding forces ( and thus k0Foff ) were closely similar for the cognate and non-cognate complexes ( Figure 6A; Table S1 ) . Under force , E9:Im2 ( D33A ) thus behaves identically to E9:Im9 ( Figure 7 , solid dark grey and orange bars , respectively ) , despite kioff values that differ by four orders of magnitude . Similar to the behaviour of E9:Im9 , k0Foff for E9:Im2 ( D33A ) is also faster than its known kioff ( 7 . 6 s−1 versus 0 . 054 s−1 ) [7] , indicating that the underlying energy landscape for immunity protein dissociation from E9 is highly sensitive to the effects of force , regardless of the nature of the bound immunity protein . By contrast with the dynamic force spectrum of colicin:immunity proteins at low loading rates ( governed by the outer barrier ) , the inner barrier for dissociation of E9:Im2 ( D33A ) is reduced significantly compared with that of the cognate complex and only becomes visible at loading rates >22 , 000 pNs−1 ( Figure 6A ) . In these kinetic unbinding experiments each force regime is assumed to probe the free energy difference between the bound state and each barrier to unbinding . As the energy difference between the bound and free states is reduced for E9:Im2 ( D33A ) relative to E9:Im9 ( Kd is reduced 105-fold [7] ) , a reduction in unbinding force would be expected for both the inner and outer barriers . Instead , application of force at residue 3 of E9 appears to decouple the dual recognition sites of helices II and III of the immunity protein with E9 . The inner barrier measures the “strength” of the specificity residues in helix II of the immunity protein and E9 ( which are stabilising for Im9 and less stabilising for Im2 ( D33A ) ; Figure 1C ) , while the outer barrier height is determined by the stability of interactions between E9 and immunity protein helix III ( identical in sequence across all DNase E-colicins except a Thr substitution for Ser at position 51 of Im7 ) [6] . As discussed above , the inner and outer barriers appear to be due to the dissociation of immunity protein helices II and III , respectively , from the binding surface of E9 . To confirm this assignment , DFS was used to measure E9 unbinding from Im9 variants containing single point mutations that destabilise either the cognate specificity interactions of the inner barrier ( V34A in helix II ) , or interactions that define the outer barrier ( D51A in helix III ) [8] ( Figure 6B and 6C ) . As predicted , unbinding forces for E9:Im9 ( V34A ) were identical to those for wild-type E9:Im9 at loading rates <5 , 400 pNs−1 , but were reduced by more than 35 pN compared with E9:Im9 at loading rates higher than this ( xu remained constant for both barriers ) . By contrast , at loading rates <3 , 000 pNs−1 , the unbinding forces for E9:Im9 ( D51A ) were reduced to a level below the thermal noise limit of the instrument . However , at higher loading rates , E9:Im9 ( D51A ) behaves similarly to the wild-type E9:Im9 complex . These data are consistent with the proposed dual-site recognition process for colicin:immunity protein ( un ) binding [8] with force effectively uncoupling the unbinding of helices II and III of Im9 from the E9 binding surface . The ability to assign each regime of the dynamic force spectrum to the unbinding of the two recognition sites of the complex previously identified by ensemble methods [8] renders the presence of an additional hidden barrier unlikely . If present , a hidden barrier would require an xu value of greater than 3 . 7 nm to obtain a k0Foff value commensurate with kioff . This is larger than the outermost barrier previously observed for the dissociation of biotin from avidin [22] . In contrast to the characteristically flat recognition surface of E9:Im9 that is typical of protein-protein interactions in general , biotin resides in a deep pocket within avidin . We thus consider the presence of an additional barrier unlikely . Overall , therefore , the results indicate that the application of force distal to the E9:Im9 interface enables rapid dissociation of this tight binding complex such that the dissociation rate is enhanced by greater than a million-fold to a timescale commensurate with the kinetics of cell killing by colicins ( within minutes ) [2] , [15] , [29] . Force induced conformational changes are known to trigger catalysis [30] or expose “cryptic” binding sites [31] in some proteins . These remodelling events are usually very sensitive to the points of force application as proteins are known to display anisotropic force responses . Thus , when extended in different directions proteins can appear to be mechanically strong or weak [32]–[35] . To investigate whether this effect is the origin of the force-induced lability of E9:immunity protein complexes , the effect of altering the pulling location on the dynamic force spectrum of the E9:Im9 complex was examined . Accordingly , different residues on E9 and Im9 ( positions 3 , 30 , and 108 on E9 and positions 38 and 81 on Im9 ) were mutated individually to Cys to enable immobilisation to the surface at different points ( Figure 2 ) . These experiments showed that k0Foff for the outer ( rate-limiting ) barrier remains 105- to 107-fold higher than kioff regardless of the pulling location employed ( Figures 7 and 8; Table S1 ) . Nonetheless a small , but significant , dependence of the unbinding force ( Figure 8A ) and k0Foff ( Figure 7 ) on the immobilisation site on E9 was observed , with the highest k0Foff values occurring when E9 was pulled from an N-terminal location ( residue 3 , k0Foff = 4 . 9 s−1 ) and lower values occurring when E9 was immobilised at position 108 or 30 ( k0Foff = 1 . 5 and 0 . 4 s−1 , respectively; Figure 7 and Table S1 ) . By contrast , k0Foff was insensitive to the pulling location on Im9 ( Figure 8B ) . The anisotropy in k0Foff in relation to the E9 pulling location , together with the large disparity between k0Foff and kioff values and the increase in chain length upon dissociation being greater than expected based on linker length ( Figure 4A ) , suggest that remodelling or partial unfolding of E9 takes place under force . This then yields a dissociation pathway with a smaller activation free energy than is accessible in the absence of remodelling . The ability to alter the unbinding kinetics by force-induced substrate remodelling has recently been postulated [36] , [37] . The results presented here show a striking example of this phenomenon , with rate enhancements of a million-fold caused by application of only 20 pN force , at most . The data described above demonstrate that the level of acceleration in the dissociation rate of E9 from Im9 is sensitive to the precise location of force application and that the rate enhancement is greatest when force is applied close to the N-terminus of E9 . Examination of the structure of E9 shows that the N-terminal 30 residues ( highlighted in red , Figure 2B ) do not contact the immunity protein binding interface directly . Leu 23 and Ala 26 of the N-terminal region of E9 do , however , form part of hydrophobic core of E9 formed around Trp 58 that also encompasses residues of the binding interface ( Val 79 , Pro 85 , and Tyr 99 ) ( Figure 2B ) . The N-terminal region of E9 may thus relay the force trigger to an allosteric site of affinity modulation . To understand the signal transduction pathway in more detail , and to identify the location of the allosteric site that translates the mechanical stimulus to an increase in dissociation rate , a series of mutant E9 domains were produced containing disulfide bonds in different locations of the protein structure ( Figure 2A and 2B ) . Disulfide bond formation in all of these variants was shown to be spontaneous and to proceed to completion using Fourier transform ion cyclotron mass spectrometry ( Figure S7 ) . The dynamic force spectrum of E9:Im9 complexes extended from the N-terminal region of E9 ( 3:81 ) engineered to contain a disulfide bond that links the N-terminal region of the polypeptide chain to the remainder of the folded globular region of E9 ( linking residues 13–117 or 20–66; Figure 2 ) are shown in Figures 9A and 9B , respectively . Remarkably , both of these cross-linked E9 variants display a simple monotonic dynamic force spectrum over the entire accessible loading rate range with significantly increased unbinding forces ( ΔF≈100 pN relative to wild-type complexes ) . The value for xu , however , is similar to that observed for the outer barrier in the dynamic force spectrum of the uncross-linked , wild-type E9:Im9 complex . Fitting these data to the Bell-Evans model yields k0Foff values for these complexes that are increased by ≈106-fold , resulting in values for k0Foff that are similar to those measured using ensemble techniques ( k0Foff = 4 . 6×10−6 s−1 and 1 . 4×10−6 s−1 for 313–117:81 and 320–66:81 , respectively; Figure 7 ) . Values of kioff measured using gel filtration experiments under identical conditions to those employed for the AFM experiments are 3 . 0×10−6 and 5 . 8×10−6 s−1 for E920–66:Im9 and pseudo wild-type E9:Im9 derivatised with methyl- ( PEG ) 12-maleimide , respectively ( Materials and Methods; Figure S8; Table S1 ) . Addition of 4 mM DTT reversed this mechanical strengthening , leading to unbinding forces identical to those of wild-type E9:Im9 ( Figure S9 ) . To localise the region of E9 involved in force remodelling more precisely , the dynamic force spectrum of a third E9:Im9 ( 3:81 ) complex that contains a disulfide cross-link distal to the N-terminal region of E9 ( 31–122; Figure 2B ) was analysed . In this case no force enhancement was observed . Instead , a dynamic force spectrum with a single force regime was obtained , identical to that of the outer barrier of the wild-type uncross-linked complex ( Figure 9C ) . These data localise the allosteric trigger to residues 21–30 or to residues 118–121 in E9 . As E9 is extended from the N-terminus in these experiments we consider the latter site to be unlikely as the site of the trigger . The data presented here reveal that insertion of a disulfide bond is able to modulate how force is propagated through the nuclease domain of colicin E9 preventing remodelling of E9 and thus facile complex dissociation . Extension of the complex in a geometry that propagates the force via a different path would thus be expected to render the cross-link between residues 20 and 66 less effective . In accord with this hypothesis , k0Foff values for E920–66:Im9 complexes were found to be dependent on the position at which force is applied to both E9 and Im9 ( k0Foff = 4 . 0×10−5 s−1 , 1 . 4×10−6 s−1 , and 5 . 3×10−4 s−1 for 320–66:38 , 320–66:81 , and 10820–66:81 , respectively; Figures 7 and 9D; Table S1 ) . The sensitivity of k0Foff to the pulling location on E9 and to the presence of cross-links in the N-terminal region of this protein demonstrates that force can act as an allosteric trigger for E9:Im9 complex dissociation . We have identified residues 21–30 as the most probable location of this trigger , a region that both links to the N-terminus and contacts residues involved in the binding interface . To be an effective transducer of mechanical signals , the N-terminus of E9 ( residues 1–20 ) would be expected to be mechanically labile . Analysis of the sequences of colicin E2 , E7 , E8 , and E9 reveal that the N-terminal region of all four nuclease domains is highly conserved and has a high content of small aliphatic amino-acids ( RNKPGKATGKGKPVGD; Figure 10A ) . This region of the protein thus docks against the remainder of the globular domain with little side-chain interdigitation , commensurate with the requirements of a trigger activated at low forces [38] . The sequence thus appears ideally suited to transmitting mechanical signals to the binding interface at low force . The close equivalence of k0Foff and kioff of E9:Im9 upon cross-linking the N-terminal region of E9 to the remainder of this globular protein suggests that unbinding under ambient conditions and that induced by force occur by the same pathway . In this case , co-operativity between binding hot-spots on the immunity protein helices II and III is restored and mutation of residues in either helix should yield changes in the rate-limiting outer barrier that correlate with the change in affinity for that complex . Analysis of 320–66:81 E9:immunity protein complexes that vary in their binding affinity from the tightly bound Im9 ( Kd = 1 . 6×10−14 M ) , through Im2 ( D33A ) ( Kd = 1 . 0×10−9 M ) , to the weakly bound Im2 ( Kd = 1 . 5×10−7 M ) each yield k0Foff values close to the previously measured ensemble kioff values ( Figures 7 and S10 ) . Together , these data provide further evidence that cross-linking switches the force-induced unbinding pathway ( which involves remodelling of the E9 subunit within the complex ) to a cooperative event that closely matches the thermally induced unbinding mechanism . Only a single monotonic force regime is observed in the dynamic force spectrum of all complexes that contain a disulfide bridge , irrespective of their mechanical phenotype or Kd value that varies over seven orders of magnitude . This finding may reflect the absence of the second inner barrier ( due to the co-operativity between each binding hotspot ) , or changes in the relative height of each barrier that results in an altered route of force propagation that moves the inner barrier to a loading rate beyond the dynamic range available to AFM . Irrespective of these changes to the energy landscape , the observation of k0Foff values that concur with previously determined kioff values reveals that the rate-limiting step for the forced and thermally activated pathways is similar when the structural pliability of E9 is minimised by bolstering the E9 structure with disulfide cross-links . This observation has important implications for interpreting dynamic force experiments on proteins with mechanically labile structures and helps to explain the differences in off-rates frequently observed obtained by ensemble and DFS methods [39] , [40] . The million-fold increase in koff measured for E9:Im9 represents a striking example of this phenomenon .
The evolution of protein sequences has generated a rich repertoire of finely tuned protein-protein interactions whose binding affinities span ≈13 orders of magnitude [41] . Some complexes ( barnase-barstar , for example ) have evolved to bind tightly and to have a long lifetime ( ≈1 . 5 d at pH 8 . 0 ) [42] . Other , equally avid , complexes ( for example SNARE complexes ) need to dissociate more frequently for biological function [43] . Whilst altering protein sequence can modulate the binding affinity and the on- or off-rates of protein complexes , in some cases by many orders of magnitude [8] , [44] , [45] , force-induced substrate remodelling offers further opportunities to tune the energy landscape of complex formation and dissociation . For example , interactions can become stronger ( catch bonds ) [46] , or weaker ( slip bonds ) [27] , [28] upon the application of force , and hidden epitopes required for binding can be exposed by forced unfolding ( cryptic motifs ) [31] . In the case of cell-cell adhesion mediated by protein-protein interactions , combinations of these have also been identified [47] . Here we have shown a striking example of how force-induced substrate remodelling can modulate complex stability ( Figure 10B and 10C ) . At low loading rates and forces ( <20 pN [48] , [49] ) , highly avid cognate E9:immunity protein complexes dissociate in tens of milliseconds . Such lifetimes are ≈106-times shorter than the thermally induced off-rate for the same complex ( 4 . 1 d ) revealing a remarkable sensitivity of lifetime to force . These force-induced lifetimes are commensurate with the timescale for colicin intoxication of bacteria . By altering the points of immobilisation and introducing disulfide cross-links at different locations , we identify the N-terminal region of E9 as a force transducer and suggest residues 21–30 as the location of the allosteric effector of force-triggered dissociation . The N-terminal region of E9 is highly conserved with a high content of Ala and Gly residues that render this region of the protein conformationally pliable . Such a sequence provides the ideal circuitry to relay a conformational trigger to the allosteric switch that lies close to the protein complex interface . This conformational rearrangement results in contour lengths at dissociation of all the complexes that are greater than expected ( Figure S11; Text S1 ) . This could reflect a degree of local unfolding in one or both proteins involved in the complex , or could result from deformation/elongation of the intact complex under force application prior to its dissociation . In this study we have used an AFM to apply a stimulus to trigger remodelling of the E9:Im9 interface . In vivo , this triggering force may be driven by conformational re-arrangements caused by changes in the environment or by changes in other domains of the colicin that are transmitted to the DNase domain . Indeed , introduction of a disulfide cross-link across the N- and C-terminal regions of the R-domain of colicin , which is N-terminal to the nuclease domain ( Figure 1B ) prevents immunity protein release upon translocon formation [15] . In addition to local conformational changes upon formation of the translocon complex , induced conformational changes may drive immunity protein dissociation by ( i ) differential rates or extent of diffusion of the inner and outer membranes ( or protein domains within these ) [20] that are linked by the docked colicin:BtuB:OmpF:TolB translocon , or ( ii ) an energised motor-like function of TolQRA domains on the inner membrane . However , other stimuli may also result in the remodelling of the allosteric trigger . For example , facile dissociation of a colicin:immunity protein complex has also been reported for E3:Im3 . In this case , binding of the complex to a strong anion exchange resin was suggested to induce conformational changes in the immunity protein that resulted in colicin release [50] . The responsiveness of E9 to its environment has been further demonstrated by the observation that insertion into a negatively charged membrane ( required for colicin intoxication ) is prevented by introduction of one the disulfide linkages ( 20–66 ) that we show here to prevent E9 remodelling . In addition to its biological implications for colicin intoxication , our study provides direct experimental evidence that force can induce changes in the energy landscape measured by dynamic force spectroscopy using the AFM and , for E9:Im9 , provides a mechanism by which this occurs . The data show that , in addition to tilting of a “zero force” landscape as predicted and quantified by Bell , force can re-sculpt the underlying energy landscape . For E9:Im9 dissociation , we show that these changes allow facile dissociation of an avid complex at low forces . For example at 25 pN the dissociation rates of wild-type and cross-linked E9:Im9 complexes are 163 s−1 and 1 . 8×10−4 s−1 . The ability to re-sculpt the energy landscape by force provides biology the opportunity to break apart highly avid complexes in the absence of a direct source of energy . In support of this , discrepancies between kioff and k0Foff are observed for many complexes [39] , [40] but , in contrast to the 106-fold difference in off-rates observed for E9:Im9 , these differences are typically relatively small ( 102 at most ) . This suggests that colicin sequence and structure have evolved to enable triggered unbinding that is required for their biological function . In rare cases , such as the dissociation of an antigen from a kinetically and mechanically stable single-chain antibody ( an immunoglobulin-like domain ) excellent agreement between off rates is observed between ensemble and dynamic force spectroscopy methods [28] . This study , together with the identity of k0Foff and kioff for the dissociation of Im9 from disulfide bridged E9 variants ( residues 13 and 117 or 20 and 66 ) adds further support that conformational remodelling can drive dissociation in vivo . The remodelling force could be generated in many ways , such as by energy-dependent remodelling enzymes ( AAA+ proteins , for example ) , by the binding of new ligands leading to changes in the dynamics or conformation of the complex , or by changing the chemical environment . The force-induced switching between populations of protein complexes with distinct mechanical properties has been observed previously for the nuclear transport complex Ran:importin β [51] and subunits of von Willebrand factor involved in blood clotting [52] . While the mechanism underlying the force switch varies in these two cases , the application of force results in a switch to a more force resistant slip “bond” or , for von Willebrand factor , to a flex “bond . ” ( Note: these are not single bonds but a series of non-covalent interactions ) . For E9:Im9 the situation is reversed in that force induces a transition from a high resistance scenario ( low koff ) to a low force resistance slip bond ( high koff ) . This is akin to a trip wire ( a “trip bond” ) , whereby small forces trigger the remodelling of an interface that is very stable in the absence of force . The identification of a trip bond thus adds to the repertoire of behaviour of biomolecules under force that has emerged over the last decade [30] , [46] , [47] , [52] , [53] and provides a mechanism to explain the discrepancy in off-rates often observed between ensemble and DFS measurements . For colicin function , the force response of a trip bond meets the seemingly mutually exclusive requirements to provide long term protection to the host , yet permit the facile dissociation of immunity protein that is required for cell invasion of its competitors .
Triple cysteine variants of E9 were designed using “Disulfide by design” software [54] . All proteins were created and purified as described previously [14] . Silicon substrates were first cleaned by sonication in chloroform for 30 min and silicon nitride AFM cantilevers were cleaned by rinsing with chloroform for a minimum of 5 min . The substrates and AFM probes were then exposed to UV radiation ( 254 nm ) for 30 min . Following this , surfaces to be functionalized were held under vacuum in the presence of 80 µl ( 3-aminopropyl ) triethoxysilane ( APTES ) and 20 µl of N , N-diisopropylethylamine ( DIPEA ) for a period of 2 h . After this time the APTES and DIPEA were removed and the treated surfaces were left to cure under a nitrogen atmosphere for 24 h . These aminosilinated surfaces were then reacted with a heterofunctional PEG linker ( NHS- ( PEG ) n-maleimide [n = 4 or 12 , Thermo Scientific] ) by adding 15 µl of 250 mM PEG linker in DMSO to 1 ml of chloroform in which the surfaces were incubated for 1 h . After functionalization with the PEG linker , the surfaces were washed using chloroform , dried under nitrogen and held under PBS until required . To avoid hydrolysis of the maleimide groups , functionalized surfaces were used within 1 h of their preparation . When required , functionalised surfaces and AFM probes were incubated with protein ( at a concentration of 1 mgml−1 in PBS which , is in excess with respect to maleimide groups on the surfaces ) for 30 min and then washed with PBS . All AFM measurements were conducted on an Asylum MFP-3D microscope using Si3N4 cantilevers with nominal spring constants of either 30 or 100 pNnm−1 ( Bruker MLCT ) . For each cantilever used , the spring constant was determined using the thermal method [55] , [56] via inbuilt fitting software . Retraction velocities of 200–8 , 000 nms−1 were employed for dynamic force spectroscopy analysis . For velocities between 200 and 5 , 000 nms−1 a PEG linker composed of 12 monomers was used , and for retraction velocities of 8 , 000 nms−1 a shorter PEG linker ( four monomers ) was used in order to increase the loading rate that could be applied . All experiments were conducted under PBS at 25°C . For retraction velocities less than 5 , 000 nms−1 , 30 pNnm−1 nominal spring constant cantilevers were employed . At retraction velocities greater than 5 , 000 nms−1 , 100 pNnm−1 nominal spring constant cantilevers ( which have a smaller cross section ) were used in order to reduce the hydrodynamic drag experienced by the cantilever , which becomes significant for the 30 pNnm−1 cantilevers at retraction velocities above 5 , 000 nms−1 . A minimum of three separate functionalized AFM tips and surfaces were used for the collection of each dynamic force spectrum measured . Size exclusion chromatography ( SEC ) was used to quantify the release of E9 DNase into solution over time from an E9 DNase:Im9 complex incubated in the presence of excess full length colicin E9 . This procedure was used to measure kioff for both E9 ( S3C ) :Im9 ( S81C ) derivatised with a PEG linker and E920–66 domains ( see text ) in complex with Im9 , under conditions identical to those employed for the DFS experiments . E9 and Im9 were derivatised with methyl- ( PEG ) 12-maleimide ( ( MM ( PEG ) 12 ) , Thermo Scientific ) by incubation of the protein with a 20-fold molar excess of MM ( PEG ) 12 overnight at room temperature in 25 mM Tris . HCl buffer , 1 mM MgCl2 ( pH 7 . 5 ) . Following this , derivatised protein was separated from un-labelled protein and excess MM ( PEG ) 12 by size exclusion chromatography . E9 DNase domains were first incubated with Im9 at a molar ratio of 1∶2 in order to form the E9 DNase:Im9 complex . The E9 DNase:Im9 complex was then purified from the excess Im9 . A 25-µM solution of the E9 DNase:Im9 complex and 125-µM of full length colicin E9 in PBS buffer ( pH 7 . 3 ) , 0 . 01% ( w/v ) azide and a protease inhibitor cocktail ( set III , EDTA free , Calbiochem ) was then incubated for different lengths of time . Samples were removed at various times between 0 and 144 h and analysed via SEC . The intensity of the elution peak that corresponded to the free E9 DNase domain ( competed from the complex by the addition of excess of full length colicin E9 ) was quantified as a function of time to calculate an apparent dissociation rate constant ( kioff ) . An example dataset is shown in Figure S8 . E9 nuclease activity was assessed by monitoring the conversion of supercoiled DNA into other forms upon addition of E9 DNase domain ( E9 ( S108C ) or E9 ( S108C ) derivatised with MM ( PEG ) 12 as described above . DNA that was predominantly in the supercoiled conformation was isolated using a Hi-speed midi prep kit ( Qiagen ) at 4°C as described [57] . Nuclease activity was measured by addition of 30 nM E9 ( final concentration ) to purified DNA at a concentration of 50 µg/ml in 25 mM Tris . HCl buffer containing 1 mM MgCl2 ( pH 7 . 5 ) in the presence or absence of 50 nM Im9 . The reaction was arrested by adding 10 µl of this solution to 5 µl of solution containing 20 mM EDTA and agarose gel electrophoresis loading dye . Multiple time points were taken and the presence of supercoiled , linear , and open circular DNA was assessed by visualisation using agarose gel electrophoresis . | Many proteins interact with other proteins as part of their function . One method of modulating the activity of protein complexes is to break them apart . Some complexes , however , are extremely kinetically stable and it is unclear how these can dissociate on a biologically relevant timescale . In this study we address this question using protein complexes between colicin E9 ( a bacterial toxin ) and its immunity protein Im9 . These highly avid complexes ( with a lifetime of days ) must be broken apart for colicin to be activated . By using single-molecule force methods we show that pulling on one end of colicin E9 drastically destabilises the complex so that it dissociates a million-fold faster than its intrinsic rate . We then show that preventing this destabilisation ( by the insertion of cross-links that pin the N-terminus of E9 in place ) yields a kinetically stable complex . It has previously been postulated that force can destabilise a protein complex by partially unfolding one or more binding partners . Our work provides new experimental evidence that shows this is the case and provides a mechanism for this phenomenon , which we term a trip bond . For the E9:Im9 complex , trip bond behaviour allows a stable complex to be rapidly dissociated by application of a surprisingly small force . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biomacromolecule-ligand",
"interactions",
"physics",
"biochemistry",
"protein",
"interactions",
"proteins",
"biology",
"microbiology",
"biophysics",
"bacterial",
"pathogens"
] | 2013 | A Force-Activated Trip Switch Triggers Rapid Dissociation of a Colicin from Its Immunity Protein |
Recently , genome sequencing of many isolates of genetically monomorphic bacterial human pathogens has given new insights into pathogen microevolution and phylogeography . Here , we report a genome-based micro-evolutionary study of a bacterial plant pathogen , Pseudomonas syringae pv . tomato . Only 267 mutations were identified between five sequenced isolates in 3 , 543 , 009 nt of analyzed genome sequence , which suggests a recent evolutionary origin of this pathogen . Further analysis with genome-derived markers of 89 world-wide isolates showed that several genotypes exist in North America and in Europe indicating frequent pathogen movement between these world regions . Genome-derived markers and molecular analyses of key pathogen loci important for virulence and motility both suggest ongoing adaptation to the tomato host . A mutational hotspot was found in the type III-secreted effector gene hopM1 . These mutations abolish the cell death triggering activity of the full-length protein indicating strong selection for loss of function of this effector , which was previously considered a virulence factor . Two non-synonymous mutations in the flagellin-encoding gene fliC allowed identifying a new microbe associated molecular pattern ( MAMP ) in a region distinct from the known MAMP flg22 . Interestingly , the ancestral allele of this MAMP induces a stronger tomato immune response than the derived alleles . The ancestral allele has largely disappeared from today's Pto populations suggesting that flagellin-triggered immunity limits pathogen fitness even in highly virulent pathogens . An additional non-synonymous mutation was identified in flg22 in South American isolates . Therefore , MAMPs are more variable than expected differing even between otherwise almost identical isolates of the same pathogen strain .
Most taxonomic descriptions of bacterial plant pathogens and studies of their life cycle were performed at a time when it was impossible to classify bacteria precisely . Therefore , it can be difficult to determine whether plant diseases affecting crops in the field today are caused by the same pathogens described in the literature as their causal agents . Moreover , in the absence of precise classification and identification of field isolates , new pathogen variants with increased virulence may spread around the globe unobserved , presenting a potential threat to biosecurity . Furthermore , model plant pathogen strains studied for their molecular interactions with plants in laboratories may not be representative of the pathogens that cause disease in the field and genes required for pathogen success in the field may not even impact bacterial growth or virulence when evaluated under laboratory conditions , which are generally optimized for disease development . Several human diseases are caused by genetically monomorphic bacterial pathogens that evolved only after the human migration out of Africa . Genome sequencing of multiple strains belonging to each of these pathogens has elucidated their microevolution and their worldwide routes of dispersion . Examples include Yersinia pestis [1] , Bacillus anthracis [2] , and Salmonella Typhi [3] . Moreover , microevolution of clonal lineages within diverse pathogen species like Escherichia coli , Staphylococcus aureus , and Clostridium difficile have also been unraveled using single nucleotide polymorphisms identified between genomes [4] , [5] , [6] . Similar studies have yet to be undertaken for plant pathogens . Pseudomonas syringae pv . tomato ( Pto ) is the causative agent of the bacterial speck disease of tomato ( Solanum lycopersicum ) , a disease that occurs worldwide and causes severe reduction in fruit yield and quality , particularly during cold and wet springs , such as occurred in Florida and California in 2010 . Three clonal lineages of Pto have been previously described based on multilocus sequence typing ( MLST ) : T1 , JL1065 , and DC3000 [7] . Housekeeping genes of JL1065 and T1 differ in DNA sequence by only 0 . 4% while DC3000 differs from JL1065 and T1 by 0 . 9% . JL1065 and T1 were found to be the common pathogenic agents of bacterial speck in the field worldwide . Although DC3000 is a derivative of the pathotype strain of Pto and the model pathogen most commonly used to investigate the molecular basis of bacterial speck disease [8] , this lineage is only rarely found on tomato [7] . Comparing genomes of multiple isolates of the P . syringae pv . tomato ( Pto ) T1 lineage and performing a Single Nucleotide Polymorphism ( SNP ) analysis of a large collection of T1-like strains , we attempt here for the first time to unravel the microevolution and global spread of a bacterial plant pathogen .
Extending our previous MLST analysis [7] to 112 Pto isolates collected worldwide between 1942 and 2009 ( Table 1 ) we confirmed that T1 is the most common Pto lineage , followed by JL1065 and DC3000 . In fact , among all analyzed isolates only two DC3000-like strains and twenty-one JL1065-like strains were found while 89 isolates belonged to the T1 lineage . When plotting strain frequency over time ( Figure 1A ) and considering geographic origin of strains ( Table 1 and Figure 1B ) , we observed an intriguing trend: DC3000-like and JL1065-like strains were the only Pto strains isolated until 1961 when the first T1-like strain was collected in the UK . T1-strains then quickly increased in frequency becoming the most common Pto lineage . Some JL-1065 strains were still isolated in the 1980s and 1990s but all strains in our collection isolated in Europe and North America after 1999 belong to the T1 lineage . To investigate the recent evolution and virulence mechanisms of the T1 lineage , we obtained draft genome sequences using Illumina technology [9] of four T1-like strains in addition to the already sequenced genome of strain T1 [10] , which was collected in Canada in 1986 . These four newly sequenced strains are: NCPPB1108 collected in the UK in 1961 , LNPV17 . 41 collected in France in 1996 , Max4 collected in Italy in 2002 , and K40 isolated in the USA in 2005 . These strains were chosen because they represent the diversity of our strain collection in regard to time of isolation and geographic location . The genomes of NCPPB1108 , LNPV17 . 41 , and K40 were assembled and submitted to the NCBI genome database ( NZ_ADGA00000000 , ADFZ00000000 , NZ_ADFY00000000 ) , annotated , and predicted protein repertoires were compared with other P . syringae genomes . The genome of Max4 was neither submitted to NCBI nor annotated owing to significantly higher fragmentation relative to the other three genomes . A summary description of genomes can be found in Table 2 and predicted protein repertoires can be compared with additional P . syringae genomes online at genome . ppws . vt . edu . Sequencing reads were aligned against the DC3000 genome and 11 , 145 high confidence single nucleotide polymorphisms ( SNPs ) were identified between DC3000 and the five T1-like genomes using the program MAQ [11] . However , only a total of 157 SNPs were identified between any of the five T1-like strains , underscoring the close relationship among these strains ( Table S1 ) . To validate the identified SNPs we also used a second approach . This time we called SNPs between the five T1-like genomes using the T1 genome as reference for alignment , used less stringent criteria , but limited SNP identification to P . syringae core genome genes ( see details in regard to the differences between Maq settings used in the two approaches in the Materials and Methods section ) . Limiting SNP identification to the core genome allowed reliable SNP calls applying less stringent settings since genes in the core genome are present only in single copy , thus avoiding misalignment of reads typical with multigene families . 265 SNPs ( listed in Table S2 ) were identified in this way . Twenty-three of these SNPs were re-sequenced from PCR products using Sanger sequencing and all were confirmed ( data not shown ) giving us confidence in the reliability of this second approach . Since the total length of the core genome used for SNP identification in the second approach was 3 , 543 , 009 nt and the identified number of SNPs distinguishing pairs of genomes was found to be between 53 and 183 ( based on the SNPs listed in Table S2 ) , the five T1-like core genomes were determined to have pair-wise genetic distances between 0 . 000017 and 0 . 000098 . This clearly shows that Pto is a genetically monomorphic pathogen similar to , for example , Yersinia pestis or Salmonella Typhi , both of which evolved only subsequent to human migration out of Africa [12] . However , it is challenging to even estimate an approximate divergence time for the five sequenced T1-like strains since a yearly mutation rate has not yet been determined for any plant associated bacterium and data from the five genomes sequenced here are not sufficient to reliably infer a mutation rate based on the sequenced strains themselves and their time of isolation . Nonetheless , we attempted to get a rough estimate of divergence time assuming a minimum mutation rate of 3 . 4×10−9 per base pair per year as estimated for bacteria based on the E . coli and Salmonella enterica split [13] and a maximum mutation rate of 5×10−6 per bp per year , which is similar to the maximum clock rates recently inferred for a clonal methicillin resistant S . aureus ( MRSA ) lineage [14] and for Helicobacter pylori [15] and similar to a maximum clock rate assumed previously for the plant pathogen Clavibacter michiganensis subsp . sepedonicus [16] . We then used the programs IMa2 [17] , [18] and BEAST [19] to calculate divergence times for each pair of strains . The obtained results suggest divergence times of around thousand years or less using the maximum mutation rate ( Table S3 ) or around one million years using the minimum mutation rate . However , [13]considering that some of the T1-like genomes have a genetic distance from each other similar to that of the MRSA isolates analyzed by Nübel and colleagues [14] for which a divergence time of only 20 years was inferred , we believe that T1-like strains have likely evolved from their most recent common ancestor after the domestication of tomato , which must have occurred sometime before the 15th century when tomatoes were first brought from Mexico to Europe [20] . To obtain a more reliable estimate of divergence times the yearly mutation rate for plant pathogens will need to be inferred in the future based on the genomes of many more strains isolated in different years from a geographic area , where the approximate year of a single recent introduction is known , as is the case for example for P . syringae pv aesculi recently introduced into the United Kingdom [21] . A phylogenetic tree was then constructed based on the SNPs indentified by aligning sequencing reads of the five T1-like strains against the DC3000 genome ( Figure 2A ) . DC3000 was used as outgroup but only SNPs that distinguished the five T1-like strains from each other were considered ( that is , SNPs that distinguished only DC3000 from all five T1-like strains were excluded because they were not informative with respect to evolution of T1-like strains ) . Trees with identical topology were obtained using only intergenic , intragenic , synonymous , or non-synonymous SNPs ( data not shown ) , suggesting that selection did not significantly affect phylogenetic reconstruction . Typical for recently diverged bacterial genomes [22] , no homoplasies or recombination events were detected . Interestingly , strain NCPPB1108 isolated in 1961is located on the most basal branch of the tree , followed by T1 isolated in 1986 on the next branch , while the most recently isolated strains LNPV 17 . 41 ( 1996 ) , Max4 ( 2002 ) , and K40 ( 2005 ) cluster together on the most derived branch . This could suggest that in the last 50 years we have witnessed an evolution of T1-like strains whereby the strains found on tomato today may have replaced their ancestors of the recent past and may be relatively more fit . To address the question as to whether T1-like strains have evolved since 1961 , we sequenced for all 89 T1-like strains in our collection the seven informative SNP loci distinguishing strains Max4 , LNPV17 . 41 , and K40 from strains T1 , NCPPB1108 , and DC3000 ( which were identified in the alignment of the Max4 , LNPV17 . 41 , K40 , and NCPPB1108 sequencing reads against the T1 genome ) . We also sequenced for all these strains four of the SNP loci distinguishing strains T1 , Max4 , LNPV17 . 41 , and K40 from strains DC3000 and NCPPB1108 . The analyzed SNPs are highlighted in the Table S2 . Eleven different genotypes were identified among the 89 analyzed strains based on these SNP loci and SNPs in the housekeeping genes used for the original MLST analysis . Genotype sequences are listed in Table S4 and genotypes for each strain are listed in Table 1 . A maximum likelihood tree was then constructed using DC3000 and JL1065 as outgroup ( Figure 2B ) . When plotting frequency of the identified genotypes over time ( Figure 3A ) , it becomes clear that genotype frequency has changed dramatically since 1961 with different genotypes peaking at different times . Moreover , genetic distance of genotypes appears to be correlated with time since the strains identified in the 1960s and 1970s are more similar to the DC3000 outgroup than the strains isolated during the last 10 years ( Figure 3B ) . This correlation between genetic distance and time was found to be statistically significant for strains collected in Europe , the only continent where strains were consistently sampled between 1961 and 2005 . This suggests that genotypes may have evolved from each other . However , the strains from the most basal clade in the tree ( Figure 2B ) have either a 1 bp deletion or a 5 bp deletion in the gene coding for HopM1 , a type III effector known to suppress plant immunity during infection of Arabidopsis by strain DC3000 [23] , [24] , [25] . These deletions cause frameshifts leading to truncated open reading frames that are respectively 636 and 1182 bp long compared to the full length hopM1 gene in strain DC3000 , which is 2139 bp long ( Figure 4A ) . In contrast , T1-like strains on all other branches of the tree have a hopM1 allele with a nonsense mutation at bp 463 and the hopM1 allele of strain JL1065 has a 180 bp long in-frame deletion starting at position 1379 . Importantly , besides the 1 bp and 5 bp deletions and the premature stop codon all three hopM1 alleles present in the T1-like strains have 100% DNA identity to each other including the up-stream promoter region and chaperone gene shchopM1 . Therefore , three independent mutations truncated hopM1 very recently in T1-like strains and not even one T1-like strain with the ancestral full-length hopM1 allele is present in our strain collection . This suggests strong selection for loss of full-length HopM1 ( see more below ) . Interestingly , only six strain out of 89 T1-like strains have the deletions causing frameshifts leading to premature stops at codon 212 and 394 while the other 83 T1-like strains have the hopM1 allele with the early stop at codon 155 . These 83 strains thus represent the main T1-lineage that has been causing bacterial speck since 1969 , when the first member of this lineage was isolated in Switzerland . To distinguish the strains belonging to this most common T1 lineage from the other T1-like strains we call these strains from now on “T1-proper” . The world map in Figure 3C shows that several genotypes within T1-proper are present in North America and Europe , suggesting that these strains have moved with relatively high frequency between continents , possibly within seed shipments . In fact , transmission of Pto via infested tomato seed has been documented [26] . Long distance movement of Pto through the atmosphere is also a possibility since P . syringae bacteria have been isolated from rain and snow [27] . Moreover , as described above , genotypes with increasing genetic distance from the outgroup appear to have replaced one another in North America and Europe . However , members of more ancestral T1 lineages as well as JL1065-like strains have apparently persisted in developing countries in South America , Africa , and Asia ( Table 1 and Figure 3 ) . This suggests only occasional movement of Pto strains between Europe and North America on one hand and South America and Africa on the other . Moreover , the strains separated from the Pto population in North America and Europe seem to continue to adapt to tomato independently as evidenced by mutations found only in these strains ( see also results for fliC alleles from strains isolated in Colombia below ) . Is it possible that the hopM1 truncation of T1-proper strains contributed to the worldwide expansion of this lineage ? Intriguingly , the full length HopM1 protein of strain DC3000 triggers cell death in several tomato cultivars and wild tomato relatives indicating that it may function as a so-called “avirulence” gene , the product of which is recognized by a plant resistance gene leading to activation of plant defenses including programmed cell death [28] . However , given that mutating hopM1 DC3000 reduced symptom development during tomato infection and did not increase bacterial population size in planta , HopM1DC3000 has been considered a virulence factor on tomato [23] , [29] . To determine if the truncated hopM1 alleles that we identified in the T1 and JL1065 lineages lost the ability to trigger cell death in tomato , transient assays expressing all identified hopM1 alleles directly in tomato leaves using Agrobacterium-mediated expression were performed . It was found that the hopM1T1 , hopM1PT21 , hopM1NCPPB1108 , and hopM1JL1065 alleles do not trigger cell death while hopM1DC3000 triggers cell death strongly ( Figure 4B ) . However , when bacterial growth was compared under lab conditions between T1 and a T1 strain expressing hopM1DC3000 ectopically , consistent differences were not observed ( data not shown ) . We thus conclude that full-length HopM1 may be recognized by a tomato resistance gene leading to reduced bacterial growth in field conditions . Alternatively , the cell death triggered by hopM1DC3000 in the Agrobacterium-mediated expression assay may not be due to recognition but may be correlated to the known role of hopM1DC3000 in symptom formation [23] . If so , it is possible that the contribution of hopM1 to disease symptoms may actually lead to an artificial selection against full length hopM1: seedlings with severe disease symptoms infected with strains that carry the full length hopM1 allele may be less likely to be sold to farmers for planting than seedlings with mild symptoms or no symptoms at all that are infected with strains that carry a disrupted hopM1 allele . Thus , a gene like hopM1 that increases symptom severity may actually render a plant pathogen less fit in an agricultural setting . Regardless , the obvious selection for inactivation of hopM1 apparent upon analysis of multiple strains shows how characterization of pathogen populations beyond the study of a single model strain can provide new perspectives on the roles of individual virulence factors . To assess other factors potentially contributing to the success of the T1-proper strains , two additional effector genes , avrRps4 and avrPto1 , differing among the five sequenced T1 genomes were analyzed ( see Table S5 for results and Table S6 for a list of all predicted effectors in the sequenced T1-like genomes ) . Neither effector was found to be consistently present or absent in T1-proper strains compared to other T1-like strains indicating that these effectors cannot explain the recent expansion of the T1-proper lineage . Nor was there a correlation with presence or absence of the gene cluster for the biosynthesis of the phytotoxin coronatine , which is known to play an important role in the pathogenesis of strain DC3000 on Arabidopsis [30] , or avrD1 , a gene specifying the production of defense inducing syringolides [31] ( Table S5 ) . Also extending the search for differences in gene content beyond known virulence genes did not lead to plausible hypotheses in regard to what might have determined the expansion of T1-proper strains compared to all other Pto strains . Only 27 gene families , mostly coding for hypothetical proteins or bacteriophage-related proteins , are present in each of the annotated draft genome sequences of the T1-proper strains T1 , K40 , and LNPV17 . 41 but absent from the Pto strains NCPPB1108 , JL1065 and DC3000 ( as determined by using the protein repertoire comparison tool at http://genome . ppws . vt . edu/orthologsorter/ ) . However , it was striking that one of the seven informative SNPs that distinguished LNPV17 . 41 , K40 , and Max4 from T1 , NCPPB1108 , and DC3000 was in the gene fliC , resulting in a S99F mutation ( Figure 5A ) . Intriguingly , the gene fliC codes for the flagellum subunit flagellin , well known to contain microbe associated molecular patterns ( MAMPs ) that trigger an innate immune response in plants and animals [32] , [33] . The S99F mutation was found in a majority of T1-proper strains isolated from tomato after 1990 in North America and Europe ( see genotypes IPV-CT28 . 31 and LNPV17 . 41 in Figure 3 ) . Moreover , of all the mutations analyzed in the 89 Pto strains , only this particular SNP was incongruent with other SNPs: the S99F mutation is present in strains KSP53 and KS127M ( both of genotype KSP53 ) from Tanzania , although their genetic background is different from all other strains that carry this mutation . This finding suggests a recombination or parallel evolution event involving fliC ( which was not detected when sequencing the five T1-like genomes since the genomes of strains 632 and 633 were not completely sequenced ) and further supports the idea of strong directional selection on the fliC gene . Surprisingly , we even found two additional fliC mutations in T1-proper strains belonging to genotypes Colombia198 and Colombia338 isolated in different regions of Colombia in 2008 and 2009 . Both mutations are non-synonymous with one of them ( D39I ) corresponding to a highly conserved amino acid in the middle of the flg22 peptide ( Figure 5A ) , a region of the FliC protein recognized by the tomato immune receptor LeFls2 [34] . The other mutation ( A96V ) is only two codons away from the fliC mutation described above ( S99F ) . These findings suggest that even successful pathogens may be limited in their growth by the plant immune system and to be under selection pressure to further reduce induction of plant defenses . Moreover , the cluster of two mutations in a region apart from flg22 suggests a second region within flagellin besides flg22 that triggers a plant immune response . In fact , infiltrating 28 amino acid long peptides corresponding to the three alternative alleles of this region ( denoted as flgII-28 ) , we observed that the ancestral allele triggered induction of reactive oxygen species ( ROS ) indicative of a plant defense response while ROS triggered by the two derived alleles was significantly reduced and/or delayed depending on the tomato cultivar tested ( Figure 5B ) . The same trend was observed between the ancestral and derived flg22 alleles ( Figure 5B ) . Moreover , infiltration of the ancestral flgII-28 peptide into tomato leaves caused more stomatal closure than infiltration of the derived allele LNPV17 . 41 ( Figure 5C ) . Stomata are known to be important points of entry into the leaf apoplast for Pto [30] . In fact , infiltration of tomato leaves with flgII-28 peptides in advance of spraying bacteria on leaf surfaces reduced apoplastic bacterial population sizes 24 hours after inoculation ( Figure 5D ) . Although the effect of the three different alleles was not significantly different from each other , the ancestral allele consistently reduced population sizes slightly more than the two derived alleles in each of three independent experiments . Taken together , these finding suggest that the mutations in flgII-28 facilitate leaf invasion making strains that carry these mutations more competitive during this important phase of the pathogen life cycle . ROS were also induced by the ancestral flgII-28 allele in Nicotiana benthamiana but none of the flgII-28 alleles triggered ROS in Arabidopsis or bean ( data not shown ) . This indicates that flgII-28 is a MAMP , which may be specifically recognized by Solanaceae species . Whether flgII-28 is recognized by the flg22-receptor LeFL2 [34] or if it is recognized by a different receptor remains to be evaluated . The almost complete worldwide replacement of strains having the ancestral flgII-28 with strains carrying the derived allele highlights how new pathogen variants can rapidly spread around the world . Therefore , reducing movement of plant pathogens between geographic regions represents an important strategy for avoiding spread of increasingly virulent pandemic strains - even in cases when strains or variants of the same pathogen are already present in these regions . Importantly , our data also reveal that MAMPs are more variable than expected . While it was previously reported that strains belonging to the same plant pathogen species can differ in regard to the sequence of the flg22 epitope [35] , here we find that even strains belonging to the same clonal lineage can show allelic variation in flagellin . This finding also questions the recently suggested durability of immunity triggered by other MAMPs [36] . However , targeted gene engineering of the FLS2 receptor gene , and possibly other yet uncharacterized flagellin receptors , may still have potential for strengthening the plant immune response against pathogens with mutated MAMPs . We have shown how genome sequencing of multiple isolates of a crop pathogen and analysis of a large collection of isolates with genome-derived markers can yield new insights into plant pathogen evolution and molecular plant-pathogen interactions . We found that the typical bacterial speck pathogen of tomato , represented by the T1-proper lineage , is a recently evolved pathogen that rapidly spread around the world , similar to genetically monomorphic human pathogens like Yersinia pestis [1] , Bacillus anthracis [2] , or Salmonella Typhi [3] . This suggests that other bacterial plant pathogens may also have adapted to their hosts in recent history , possibly after domestication or - even more recently –after the advent of wide-spread cultivation in mono-culture of their hosts . Investigating microevolution of additional bacterial plant pathogens will make it possible to determine to what point the results obtained here for Pto are representative of bacterial plant pathogens in general . Inferring yearly mutation rates and divergence times will be essential for such studies . P . syringae pv aesculi [21] and Ralstonia solanacearum race 3 biovar 2 [37] are examples of plant pathogens that have recently spread to a new world region and for which many isolates collected in recent years from different locations are available . Therefore , these pathogen will be excellent candidates for micro-evolutionary and phylogeographic studies . Our results also highlight the value of assessing diversity in plant pathogen populations as an important complement to the study of model pathogen strains in lab conditions . This approach can lead to new hypotheses as to why some plant pathogens can cause disease and grow to high numbers on a plant species in lab conditions although they are rarely found on the same plant species in the field while other pathogens are successful both under lab conditions and in the field . Answering this question will be essential for gaining a better understanding of pathogen fitness in the field and to finding new avenues for successful control of plant diseases .
P . syringae pv . tomato strains listed in Table 1 were grown in King's Broth ( KB ) at 28°C and genomic DNA was extracted using the Gentra Puregene Yeast/Bacteria kit ( Qiagen ) following manufacturer's instructions . Fragments corresponding to the MLST loci rpoD , pgi , and gapA were PCR amplified and sequenced as previously described [7] . Genomic DNA of strains NCPPB1108 , K40 , and LNPV 17 . 41 was sequenced with Illumina technology [9] using the paired-end protocol with read length of 42nt at the University of Toronto Centre for the Analysis of Genome Evolution and Function ( CAGEF ) . Genomic DNA of strain Max4 was also sequenced with Illumina technology but using the single read protocol as previously described for T1 [10] . Genomes of strains NCPPB1108 , K40 , and LNPV 17 . 41 were assembled using Velvet 0 . 7 . 55 [38] . Insert size for paired-end reads was set to 200; expected coverage was based on the number of reads used in the assembly and the expected genome size based on strain DC3000; coverage cutoff was set to 4; minimum contig length cut off was set to 100 . A range of hash sizes was used to obtain the assembly with the highest N50 value and the lowest number of contigs for each genome . Scaffolding was turned off . Genomes were annotated using GRC [39] . SNPs between Pto strain T1 [10] and the other four T1-like strains NCPPB1108 , Max4 , K40 , and LNPV17 . 41 were identified by aligning Illumina sequence reads of T1 , Max4 , K40 , NCPPB1108 , and LNPV 17 . 41 against the DC3000 genome [8] in MAQ [11] . We only considered the 3 , 024 , 986 nucleotides in the DC3000 genome for which there was at least 20X depth of coverage by Illumina reads from each of the five Illumina datasets ( i . e . T1 , LNPV 17 . 41 , K40 , Max4 , NCPPB1108 ) and for which there was at least 95% consensus between the aligned reads . The polymorphism states of the remaining 3 , 372 , 140 nt of the DC3000 chromosome were considered to be ambiguous and we made no attempt to detect SNPs there . We considered a SNP to be present at a given site if at least 95% of the aligned reads at that site consistently called a different nucleotide from that in the reference sequence . We compared the position of each SNP against the positions of the predicted genes as specified in RefSeq:NC_004578 to determine whether it was intergenic or intragenic . For intragenic SNPs , we translated the open reading frame containing the SNP to check whether the SNP would result in a different amino acid sequence ( i . e . whether it was a non-synonymous mutation ) . The process was automated using custom Perl scripts . SNPs that were not informative to distinguish T1-like strains from each other were not considered , i . e . , all SNPs that distinguished DC3000 from the T1-like strains but that had the same nucleotide in all five T1-like strains . Only the SNP loci that distinguished T1-like strains from each other are shown in Table S1 and were used for construction of the whole genome tree shown in Figure 2A ( see below for details ) . In a second independent search for SNPs between Pto strains T1 , Max4 , K40 , NCPPB1108 , and LNPV 17 . 41 , Illumina sequence reads of the newly sequenced strains were aligned against the T1 draft genome using MAQ [11] using default parameters . The MAQ output was then parsed using a custom script eliminating all SNP calls that did not have the consensus A , C , G or T . A final list of core genome SNPs ( Table S2 ) was then assembled limiting SNPs to SNPs present in genes that were found to be present exactly one time in the P . syringae genomes T1 [10] , DC3000 [8] , B728a [40] , and 1448A [41] using OrthoMCL [42] . The total length of these genes is 3 , 543 , 009 nt . Based on silent , non-silent , intergenic , and intragenic sites , we constructed 5 bootstrapped ( 2000 replicates ) Maximum Likelihood trees for the genomes of strains T1 , Max4 , LNPN17 . 41 , K40 and NCPPB1108 using the genome of strain DC3000 as outgroup . The first four trees were based on each of the data features separately , and the remaining tree was based on the collection of all data features jointly , to which we refer to as the whole genome tree . Trees were constructed in PAUP version 4 . 0 ( http://paup . csit . fsu . edu/ ) using parameters determined by jMODELTEST [43] , [44] . Non-silent , intragenic , and the whole genome data satisfied the GTR substitution model [45]; whereas , silent and intergentic data best fit the GTR+I and SYM models [45] , respectively . A Maximum parsimony tree was built using DNAPARS of the PHYLIP 3 . 69 package ( http://evolution . gs . washington . edu/phylip . html ) . Primers were designed upstream and downstream of each of the seven SNPs that distinguished strains LNPV 17 . 41 , K40 , and Max4 from NCPPB1108 and T1 . Four primer pairs were designed for additional five SNPs ( two of them adjacent to each other ) that distinguished LNPV 17 . 41 , K40 , Max4 , and T1 from NCPPB1108 and DC3000 . The 12 SNPs are highlighted in green in Table S2 and primers are listed in Table S7 . Based on the SNPs listed in Table S4 , 11 genotypes were identified among the T1-like strains listed in Table 1 . Table S5 lists the SNP genotype for each strain . jMODELTEST [43] , [44] was used to determine the substitution model that best fit the data ( SYM ) . A maximum likelihood tree was then built in PAUP version 4 . 0 ( http://paup . csit . fsu . edu/ ) . Bootstrap analysis was performed with 5000 replicates . A Maximum parsimony tree was built using DNAPARS of the PHYLIP 3 . 69 package ( http://evolution . gs . washington . edu/phylip . html ) . Based on a 10-year sliding window , we calculated the relative frequencies of T1- , JL1065- and DC3000-like strains , for the time period 1942-2009 . Additionally , for the years 1961-2009 , T1-like strains acquired across North America and Europe according to genotypes were also analyzed based on a 10-year sliding window . Each T1-like strain was uniquely classified based on a profile of 40 SNPs . Eight genotypes of T1-like strains were observed in North America and Europe . Frequency plots were generated for these genotypes using the statistical software language R ( http://www . r-project . org/ ) . Genetic distances for all T1-like strains were calculated as compared to the DC3000 strain , under the Jukes-Cantor model . In order to investigate the relationship between these relative genetic distances and isolation year , we fit the regression model:where is the relative genetic distance , is the isolation year , and denotes independent normally distributed error . Values of which are distinguishable from zero indicate a linear temporal relationship between genetic distance ( ) and time ( ) . In order to estimate divergence times for the five sequenced T1-like stains ( Max4 , LNPV17 . 41 , K40 , T1 and NCPPB1108 ) , we used IMa2 [17] , [18] and BEAST 1 . 6 . 1 [19] . In both programs , we computed our estimates based on the nucleotides present at the concatenated SNP loci listed in Table S1 and setting the mutational clock rate ( µ ) to 1 . IMa2 [17] , [18] was run in Markov Chain Monte Carlo ( MCMC ) mode . We considered our five strains to be derived from five populations , and assumed no migration in the model . The mutation model used for this analysis is the Hasegawa-Kishino-Yano ( HKY ) model . Prior distributions were selected as uniform distributions between zero and some upper bound . Upper bounds were chosen to be far removed from the maximum likelihood estimate: 300 for , and 200 for effective population size parameters . In order to reduce auto correlations in our MCMC samples , 20 million iterations were run , with samples stored 10 , 000 iterations after a ‘burn-in’ period of 2 million generations . Multiple runs of the algorithm produced nearly identical results . In BEAST 1 . 6 . 1 [19] , prior distributions were selected as lognormal with units in% per million years . GTR was selected as substitution model . Since BEAST results are on a percent scale , results were converted to million years in order to compare to IMa2 results . To rescale program outputs to an estimated clock rate and to the length of the genome used for SNP discovery , we used:where DT is the rescaled divergence time in years; t is the estimated splitting time obtained from IMa2 or BEAST converted to years; is the mutation rate per base pair ( bp ) per year; is the length of SNPs used as input , which is 157 bp; and is the total length of the genome used for SNP discovery , which is 3 , 024 , 986 bp . Pseudomolecules were created from the draft genome sequences by concatenating contigs in the order from largest to smallest with the TIGR linker sequence “nnnnnttaattaattaannnnn” delimiting contig boundaries . Effectors were identified in the pseudomolecules using a combination of automated annotation generated by RAST ( http://rast . nmpdr . org/ ) , alignment of pseudomolecules with the DC3000 sequence visualized using the Artemis Comparison Tool , HrpL binding sites predicted as previously described [46] , and PSI-BLAST of confirmed effector sequences against the pseudomolecule sequences . Predicted effectors are listed in Table S6 . The open reading frames including the ribosome binding site but not the stop codon of hopM1 alleles were amplified by PCR from genomic DNA of Pto strains DC3000 , JL1065 , T1 , NCPPB1108 , and PT21 with the primer pairs listed in Table S7 and with nested primers to add sequences for GatewayTM ( Invitrogen ) cloning using the protocol described previously [47] . The five PCR products were then cloned into the entry vector pDNOR207 ( Invitrogen ) using the GatewayTM BP cloning kit ( Invitrogen ) . Recombined plasmids were confirmed by sequencing and cloned into the destination vector pBAV150 [47] using the GatewayTM LR cloning kit ( Invitrogen ) . hopM1-containing pBAV150 were mated from Escherichia coli into Agrobacterium tumefaciens C58C1 and used in transient assays of tomato leaves ( at a concentration corresponding to an optical density at 600 nm of 0 . 04 ) and in Nicotiana benthamiana leaves ( corresponding to an optical density at 600 nm of 0 . 4 ) using the same protocol as described previously for Nicotiana benthamiana [47] . Western blots were performed as described in [47] also . Peptides corresponding to alleles of flg22 and flgII-28 were ordered from EZBiolab with >70% purity ( see Figure 5 for peptide sequences ) . Peptides were resuspended in sterile water and used to measure induction of reactive oxygen species ( ROS ) in the tomato cultivar Chico III . A luminol - horseradish peroxidase assay was used to quantify ROS induction as described by Chakravarthy and colleagues [48] with small modifications: 4-mm leaf discs were punched out with a cork borer and floated adaxial side up in 200 µl ddH2O over night at room temperature in wells of a 96-well solid white plate . The ddH2O was then replaced with 100 µl of ROS testing buffer containing 1 uM of flg22 or flgII-28 peptide , 34 µg/ml of luminol ( Sigma ) , and 20 µg of horseradish peroxidase ( VI-A , Sigma ) . Luminescence was measured using a Bioteck , Synergy HT plate reader . Five leaf disks treated with the same peptide were tested in parallel . Leaf discs in testing buffer without addition of any flagellin peptide were used as a negative controls . Leaves were treated with flg22 and flgII-28 peptides as described by Melotto and co-workers [30] with slight modifications . Briefly , 4 week-old tomato plants were sprayed with water , placed in transparent plastic bags , and transferred to a 28°C incubator exposed to light to induce stomatal opening . Whole leaves were detached from plants and placed on a glass slide . The leaves were immersed in 5 uM of flagellin peptide dissolved in ddH20 , or just ddH2O for mock treatment , and then covered with a cover slip . The mounted leaves were placed at room temperature for 2 hours and then viewed at 200x magnification using an Axio Imager M1 upright microscope ( Zeiss ) . Pictures of stomata were taken using an Axiocam MRm camera ( Zeiss ) . Stomatal aperture of 20 stomata per test group per experiment were quantified using Axiovision v . 4 . 7 . 2 ( Zeiss ) . Leaves of 5-week-old tomato plants ( cv . ‘Chico III’ ) were infiltrated with flg peptides at a 1 µM concentration via a blunt end syringe while still attached to the plant . Plants were placed in a high humidity container for 24 hours . Strain NCPPB1108 was then sprayed onto leaves at a concentration corresponding to an optical density at 600 nm of 0 . 01 in 10 mM MgSO4 using a sprayer canister and placed back in the high humidity container . Bacterial invasion was assessed 24 hours after infection . 0 . 52 mm sections were punched out of the infiltrated leaves and placed in a tube with 200 µL 1% bleach with the leaf punch completely submerged . The tube was mildly vortexed for 5 seconds to remove epiphytic bacteria . The leaf punch was then removed from the 1% bleach solution , gently rinsed in ddH2O , and then placed in a separate tube containing 200 µL 10 mM MgSO4 and three 2 mm glass beads . The tube was placed in a mini bead beater ( Biospec Products , Inc . ) and shaken for 90 seconds to grind the leaf and release endophytic bacteria into the solution . Colony forming units were counted after dilution plating . HQ992994 – hopM1 operon of strain T1 HQ992995 – hopM1 operon of strain NCPPB1108 HQ992993 – hopM1 operon of strain PT21 JF268671 - hopM1 operon of strain JL1065 JF261012 – fliC allele of strain K40 JF261011 – fliC allele of strain Col198 JF261013 – fliC allele of strain Col338 | Our knowledge of the recent evolution of bacterial human pathogens has increased dramatically over the last five years . By comparison , relatively little is known about recent evolution of bacterial plant pathogens . Here , we analyze a large collection of isolates of the economically important plant pathogen Pseudomonas syringae pv . tomato with markers derived from the comparison of five genomes of this pathogen . We find that this pathogen likely evolved on a relatively recent time scale and continues to adapt to tomato by minimizing its recognition by the tomato immune system . We find that an allele of the flagellin subunit fliC that appeared in the pathogen population for the first time in the 1980s , and which is the most common allele of this gene in North America and Europe today , triggers a weaker tomato immune response than the fliC allele found in the 1960s and 1970s . These results not only impact our understanding of pathogen – plant interactions and pathogen evolution but also have important ramifications for disease prevention . Given the speed with which new pathogen strains spread and replace existing strains , limiting the movement of specific strains between geographic regions is critically important , even for pathogens known to have worldwide distribution . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"science",
"pest",
"control",
"plant",
"biology",
"genetics",
"biology",
"genomics",
"evolutionary",
"biology",
"microbiology",
"population",
"biology",
"genetics",
"and",
"genomics",
"agriculture"
] | 2011 | The Plant Pathogen Pseudomonas syringae pv. tomato Is Genetically Monomorphic and under Strong Selection to Evade Tomato Immunity |
The efficacies of many new T cell vaccines rely on generating large populations of long-lived pathogen-specific effector memory CD8 T cells . However , it is now increasingly recognized that prior infection history impacts on the host immune response . Additionally , the order in which these infections are acquired could have a major effect . Exploiting the ability to generate large sustained effector memory ( i . e . inflationary ) T cell populations from murine cytomegalovirus ( MCMV ) and human Adenovirus-subtype ( AdHu5 ) 5-beta-galactosidase ( Ad-lacZ ) vector , the impact of new infections on pre-existing memory and the capacity of the host’s memory compartment to accommodate multiple inflationary populations from unrelated pathogens was investigated in a murine model . Simultaneous and sequential infections , first with MCMV followed by Ad-lacZ , generated inflationary populations towards both viruses with similar kinetics and magnitude to mono-infected groups . However , in Ad-lacZ immune mice , subsequent acute MCMV infection led to a rapid decline of the pre-existing Ad-LacZ-specific inflating population , associated with bystander activation of Fas-dependent apoptotic pathways . However , responses were maintained long-term and boosting with Ad-lacZ led to rapid re-expansion of the inflating population . These data indicate firstly that multiple specificities of inflating memory cells can be acquired at different times and stably co-exist . Some acute infections may also deplete pre-existing memory populations , thus revealing the importance of the order of infection acquisition . Importantly , immunization with an AdHu5 vector did not alter the size of the pre-existing memory . These phenomena are relevant to the development of adenoviral vectors as novel vaccination strategies for diverse infections and cancers . ( 241 words )
Immunologic memory is critical for host defense against pathogens and tumours and underpins the design of modern vaccines . Indeed , many current T cell vaccination strategies against pathogens or tumours aim to elicit long-term CD8 T cell effector memory responses [1–9] . However , the factors that govern the long-term maintenance of T cell memory , particularly effector memory , have not been fully elucidated . It is also unclear as to whether there is an upper limit to the capacity of the memory compartment to accommodate multiple epitopes or if new epitopes are generated and expand at the expense of pre-existing T cell memory populations . There is evidence supporting both assertions; studies indicate that pre-existing memory is eroded after infection with a subset of pathogens[10–12] . By contrast others [13–15] show that the T cell compartment is able to accommodate ever-increasing numbers of specificities . What is clear and is supported by more recent data , is that the infection history of the host can influence the response to new , unrelated pathogens and vaccines . [16] Inflating memory responses represent a subset of long-lived epitope-specific CD8 T cell memory responses which were initially identified in a number of natural persistent murine and human infections[17–19] and recently in a murine model of infection with a non-replicative human adenovirus type 5 construct expressing the β-galactosidase gene [20–22] . These cells maintain a functional effector-memory phenotype , accumulate at high-frequencies in multiple compartments [23–25] and are able to provide rapid front-line protection against infection [26–28] . Pertinently , our group has shown that these inflating cells share a conserved phenotype with effector CD8 T cells raised upon immunization with a replication-deficient Chimpanzee Adenovirus subtype 3 vaccine vector encoding a Hepatitis C antigen [22] . Inflating epitopes may therefore serve as useful tools to measure effects of perturbations in the size of individual T cell epitope populations under different conditions . Two models of memory inflation have been developed in mice; the first , infection with murine cytomegalovirus ( MCMV ) , a natural mouse pathogen , gives rise to multiple populations of antigen-specific inflating T cell populations as well as conventional memory T cells and recapitulates many features of cytomegalovirus ( CMV ) infection in humans [24 , 28] . The second model uses a non-replicating Adenovirus containing the beta-galactosidase gene ( Ad-lacZ ) , and when delivered into C57BL/6 mice gives rise to an inflating population ( βgal96-103 , D8V ) and conventional CD8 T cell population ( βgal497-504 , I8V ) against beta-galactosidase [20] . By tracking these inflating populations in various co-and-sequential infection scenarios , we were able to determine the long-term fate of these populations in the face of subsequent infections with unrelated pathogens and also whether the presence of large pre-existing inflating T cell populations affect the generation of new inflating T cell epitopes . Our findings indicate that the T cell memory compartment is in principle able to accommodate multiple inflating and central memory T cell pools from unrelated pathogens . Furthermore , multiple inflation and central memory responses against epitopes against two unrelated pathogens are able to develop simultaneously . We found that AdHu5-induced inflating responses could develop without impacting the pre-existing inflation population . In contrast , acute MCMV infection results in rapid depletion of pre-existing Ad-lacZ-induced inflating memory T cells . Nonetheless , the depleted inflating population could be successfully reestablished in the blood and organs by boosting with the Ad-lacZ vector , indicating that the restriction on the number of inflationary responses does not lie at the level of ‘immunological space’ . Attrition of the existing memory T cell compartment is likely a specific by-product of infection by the pathogen and may have an important impact in shaping immunologic memory in certain settings after CMV infection and possibly related infections . Thus the order in which pathogens are encountered during the lifetime of the host can have a profound impact on long-term CD8 T cell effector memory immune responses .
To determine whether the immune system is able to accommodate multiple inflationary responses to epitopes from different pathogens , mice were infected with MCMV and Ad-lacZ simultaneously by the intravenous ( i . v . ) route . The levels of inflating and central memory epitopes for the respective viruses were measured by tetramer staining of blood lymphocytes . As shown in Fig 1 , co-infection does not appear to reduce the size of the inflating and central memory epitopes-specific responses compared to single infections . Additionally , multiple inflating memory populations may develop from a single virus . As shown in S1A and S1B Fig , C57BL/6 and BALB/c hybrid mice ( F1 ) were able to generate inflating responses specific to both parental strains . Therefore , the host appears able to support the development of multiple inflating and central memory responses against two unrelated infections at the same time . The results of the first experiments indicated that mice were able to simultaneously support multiple large , inflated responses originating from infection with multiple or single viruses . In order to determine if acquiring new infections may impact upon pre-existing memory populations , the established inflating populations of Ad-lacZ and MCMV-specific cells were employed as markers . Mice were first infected with MCMV , the infection was allowed to progress to the chronic phase and then they were injected i . v . with Ad-lacZ . Infection of MCMV and Ad-lacZ by the i . v . route typically generates large populations of effector T cells in circulation as well as in the liver and lungs [20] , thus the effect of Ad-lacZ on the development of the inflating D8V response was followed in these organs . The impact of the new inflating population on the pre-existing MCMV population was investigated by tracking the M38 response early after Ad-lacZ infection and up to 100 days afterwards . We found that subsequent infection with Ad-lacZ did not alter the size of the MCMV M38-specific population ( Fig 2A and 2B , S2A–S2C Fig ) . A similar observation was made in the pre-existing MCMV-specific conventional memory response to epitope M45 , where levels were unaltered after Ad-lacZ infection ( S2D–S2F Fig ) . The level of M38-specific T cells in circulation and in non-lymphoid organs remained similar to the size of MCMV-only mice and the population remained stable for extended periods . Therefore , the presence of the new developing population of inflated cells did not impact upon the pre-existing population of effector T cells . Similarly , despite the presence of existing populations of inflated cells , these did not limit the size of the newly developing Ad-lacZ inflating population ( Fig 2C and 2D and S2G and S2H Fig ) or the conventional memory population ( S2I–S2K Fig ) . Thus the host is able to support multiple populations of epitope-specific T cells which may develop at different periods of time without any evident cost to the size of the existing or the newly generated effector memory population . Next we reversed the order of infection , whereby mice initially infected with Ad-lacZ were then infected with MCMV ( Fig 3 ) . The impact of MCMV infection on the D8V population in the circulation and in non-lymphoid tissue was measured . In contrast to what was observed previously , the levels of D8V–specific T cells dropped markedly after infection with MCMV ( Fig 3A and 3B and S3A and S3B Fig ) . This reduction was most pronounced in the blood and liver , where the populations were reduced by approximately 50% compared to non-infected Ad-lacZ immune mice , but also occurred to a lesser extent in the lungs ( S3C Fig ) . The depletion of the existing D8V population was sustained for a long period , with the population not recovering in the blood , although there were indications of slow recovery of the numbers in the liver and lung by day 100 post-MCMV ( Fig 3B and S3C Fig ) . The pre-existing response to Ad-lacZ central memory epitope I8V was also followed and while it exhibits a trend towards reduction after acute MCMV infection , the fall was not as steep and there was large variation within the groups ( S3D–S3F Fig ) . As in the previous experiment , the development of the new inflating population ( MCMV-specific M38 epitope ) was not altered by the presence of Ad-lacZ memory cells in the blood , liver ( Fig 3C and 3D ) or lungs ( S3G and S3H Fig ) and this was also observed in the new developing central memory population ( S2I–S2K Fig ) . Where mice were first infected with MCMV , then infected with Ad-lacZ and subsequently re-infected i . v . with another dose of MCMV , no early alteration in the size of the D8V population occurred , confirming that events in primary MCMV infection were responsible for attrition ( S4A Fig ) . Attrition of pre-existing D8V tetramer+ cells was also observed after infection with lower doses of MCMV , albeit at a smaller magnitude and with delayed kinetics ( S4B Fig ) . Therefore , unlike an Ad-lacZ infection , a primary MCMV infection appears to negatively impact upon the host’s existing memory T cell compartment . In order to confirm that this effect was not specific to our MCMV-Smith strain viral stocks , we repeated the experiment with MCMV stocks derived from BACS construct and purified in a different laboratory ( a kind gift from Luka Cicin-Sain [29] ) . Similar responses were observed even at a 10-fold lower inoculum , whereby the pre-existing D8V response in Ad-lacZ immune mice dramatically decreased within 1-week post-MCMV infection ( S5 Fig ) . Persistent CMV and MCMV infection alters the proportion of naïve , central and effector memory subsets . Here , we observed that Ad-lacZ immunization also alters the proportion of effector , central and naïve subsets in blood in a similar manner . Compared to naïve animals , Ad-lacZ immunization caused a large expansion of the effector memory compartment , and also increased the central memory pool resulting in a decrease in the percentage of the naïve compartment ( Fig 4A and S6A Fig ) . In mice already immunized with Ad-lacZ , a second infection with MCMV does not cause any further alterations to the proportions of these memory compartments ( Fig 4A ) . In initial time-course experiments of Ad-lacZ-MCMV infected mice , attrition of the D8V-specific population was seen even at the earliest time point of day 7 post-MCMV infection . We therefore measured earlier time-points to determine how soon after MCMV infection the depletion of the D8V subset occurred . As shown in Fig 4B the D8V-specific effector memory population in a number of compartments is reduced by 4 days post-MCMV infection indicating that the early phase of MCMV infection is responsible for the reduction observed . We therefore examined whether acute MCMV infection produces an environment in which effector CD8 T cells are more readily killed . Splenocytes were isolated from mice immunized 45 days previously with Ad-lacZ , labeled with CFSE and then equal numbers were transferred into naïve controls or MCMV-infected mice at 24 hours post-infection . The size of the transferred CFSE+D8V tetramer+ population in the blood was followed in both groups . As shown in Fig 4C , in both groups the percentage of transferred CFSE+ CD8 T cells remained stable throughout the experiment . However , in MCMV-infected animals , there is a reduction of the transferred D8V tetramer+ cells , which occurs soon after infection in blood and other tissues ( Fig 4D and S6B Fig ) , thus indicating that acute MCMV infection may cause depletion of memory T cells . To investigate why D8V+ inflating cells may be susceptible to death in the face of acute MCMV , the expression profile between MCMV-specific M38 and Adenoviral-specific D8V+ inflating cells at the steady state , day 100 and day 50 post-infection respectively [22] ( GEO: GSE73314 ) were analysed . Apoptotic pathways play a prominent role in controlling the size of the memory CD8 T cell compartment; as such we performed GSEA ( Gene Set Enrichment Analysis ) on the two groups to determine whether there was a difference in the expression levels of genes in apoptosis pathways . As shown in Fig 5A , a subset of apoptotic genes were enriched in the D8V+ population compared to M38 inflating cells . While pro-survival genes including Bcl-2 were enriched in D8V+ cells ( Fig 5B ) , notable apoptotic mediators including FADD , TRAF2 , Bax and caspase 9 were also preferentially upregulated in the D8V+ population , at day 100 , in the steady state . FADD and TRAF2 are components of the extrinsic pro-apoptotic pathway , and at day 4 post-MCMV infection , we detected upregulation of the death receptor Fas on the surface of CD8 T cells from various compartments , most prominently on bone marrow and splenic CD8 T cells ( Fig 5C ) and also on D8V specific cells in the bone marrow and lymph nodes . Consistent with this these lymphocytes were less able to survive in unsupplemented culture media ( Fig 5D ) . To confirm the role of death pathways in mediating attrition of the D8V+ population , Ad-lacZ-immune mice were treated with blocking anti-FasL antibody at the time of MCMV infection . As shown in Fig 5E , blocking Fas-FasL interaction preserved the D8V+ population during acute MCMV infection . Innate activation can impair the host adaptive response to infection [1–9 , 30 , 31][10–12 , 32] . Mice lacking the IL-18 receptor ( IL-18R KO ) have impaired activation of the NK cell subset , with reduced levels of the activation markers NKG2ace , NKG2D , NK1 . 1 and KLRG1 ( Fig 6A ) . Making use of IL-18R knock-out mice , we repeated the co-infection experiment ( Fig 6B ) . IL-18R KO mice still experienced attrition of the effector memory population to a similar level as wildtype ( WT ) controls ( 67% vs . 64% reduction p = n . s ) suggesting that attrition occurs via an IL-18R independent pathway . Data shown above indicated that the host is able to accommodate the presence of multiple populations of inflated memory T cells arising from sequential MCMV and Ad-lacZ infections . Therefore , the attrition of the D8V-specific population after MCMV infection cannot be explained by a limit in the size of the T cell memory compartment . As the attrition of the effector memory population occurred so early , we questioned if this was a global innate effect due to the inflammatory environment induced during a viral infection . While Ad-lacZ is an infectious virus , it does not replicate in vivo and therefore may not induce a similar cytokine profile to other infectious viral pathogens . We therefore repeated the same sequential experiment but this time infected Ad-lacZ memory mice with a different pathogen , vaccinia , instead of MCMV ( Fig 7A ) . The D8V response in the blood was followed for >50 days after vaccinia infection . Here , we found that aside from a transient reduction at day 3 , the percentage of D8V T cells remained largely unchanged throughout the course of the vaccinia infection . This was also observed in the populations in the liver and the lung ( S7A Fig ) . Likewise , the size of the conventional memory response I8V was not altered by subsequent vaccinia infection ( S7B Fig ) . Critically , we did not observe early depletion of the pre-existing memory cells or long-term reduction of the pre-existing memory population . By contrast , when Ad-lacZ memory mice were infected with the intracellular pathogen Listeria-OVA , a reduction in the size of the D8V+ tetramer population was also observed , albeit at a smaller magnitude compared to MCMV infection ( Fig 7B ) . These results indicate that not all acute infections are able to impact upon pre-existing inflating memory populations in a similar manner . Nonetheless , this would imply that during the lifespan of the host , depending on the infection history , multiple episodes of attrition of the pre-existing inflating memory pools may be experienced . It has been reported that antigen-specific boosting increased the size and persistence of effector memory cells [13–15 , 33] . In light of these observations we questioned whether the MCMV-depleted Ad-lacZ-induced inflating memory population might be recovered by boosting . Therefore Ad-lacZ memory mice that were subsequently infected with MCMV were later injected i . v . with a second dose of Ad-lacZ . Boosting with homologous vector successfully recovered the depleted D8V population and the population size was maintained for long periods afterwards in the blood ( Fig 8A ) , liver and lungs ( Fig 8B ) . To determine the level of contribution of newly primed naïve cells in the expansion , we measured the levels of T cell receptor rearrangement excision circles ( TRECs ) in the D8V+ population by quantitative PCR ( qPCR ) . TRECs are stable , not duplicated during mitosis , and diluted out with each cellular division [34] . If the expansion following boosting was the result of increased proliferation of the existing memory pool , then the levels of TREC in the tetramer+ population would be lower compared to the Ad-lacZ+MCMV group . By contrast , expansion due to recruitment from the naïve pool would increase the level of TREC in the D8V+ memory pool . Comparing the levels of TREC by qPCR in D8V+ tetramer + cells 7 days after Ad-lacZ boosting with the unboosted Ad-lacZ+MCMV group indicated that TREC levels after boost was marginally higher ( 1 . 3 fold increase relative to Ad-lacZ+MCMV ) than in the unboosted Ad-lacZ+MCMV ( S1 Table ) indicating some input from de novo primed cells . However naïve cell priming could not account for the full magnitude of the increase , suggesting that the expanded population could constitute of a mix of both expanded pre-existing cells and de novo primed cells . To address this point further we analysed the phenotype of the boosted cells . Prior to boosting , the depleted D8V+ population still maintained an effector memory phenotype ( Fig 8C ) , being mostly CD44+CD62L- and downregulating CD27 and KLRG-1+ ( S8A Fig ) and lacking CD103 expression ( S8B Fig ) [16 , 35] . Analysis of the boosted D8V cells at the early timepoint Day 6 post-boost indicated that almost all ( 98 . 4% ) of the D8V+ tetramer cells were still of the effector memory phenotype , akin to the D8V+ tetramer cells in long term Ad-lacZ immune ( 96 . 8% ) and Ad-lacZ+MCMV ( 95 . 8% ) mice and very different to the make up of the D8V tetramer+ population 6 days after primary Ad-lacZ immunization ( Fig 8C ) , where CD44+ CD62L- phenotype comprised the minority population ( 16 . 8% ) . This would again imply that the expanded cells originated largely from the substantial pre-existing pool rather than large-scale de novo priming of naïve CD8 T cells . This phenotype was stable in the recovered D8V population , still evident at 50 days post-boost ( Fig 8D ) . Therefore , attrition of the pre-existing effector memory compartment by primary MCMV infection may not necessarily impact upon the protective ability of the depleted subsets , as they appear able to quickly expand and recover upon antigen restimulation .
Many new CD8 T cell immunization strategies exploit the ability to generate large numbers of long-lived antigen-specific effector memory cells [4 , 6–9 , 17–19 , 36] . This subset of CD8 T cell memory migrates into non-lymphoid tissues , such as the liver and lungs , and have been shown to be integral in controlling a number of clinically-relevant infections [3 , 20–22 , 37–39] in patients and animal models . How the host immune system copes with accommodating multiple large populations of effector memory cells at different points in time remains to be studied . For example , accumulation of CMV-specific responses over time is a well-described occurrence[23–25 , 40–42] . It has been suggested that there is finite immunological space and , over time , the large clonal expansion of CMV-specific T cells may occupy so much of it that there is little room for expansion of newly encountered antigens [26–28 , 43] . We therefore tested the capacity of the immune system to accommodate multiple inflationary epitopes from unrelated pathogens by using two different models that generate inflating memory populations in C57BL/6 mice . Owing to the large sizes of these epitope specific populations , they served as useful surrogates to measure perturbations in the memory compartment during a variety of co-infection scenarios . This system may be more representative of the dynamics in real-life sequential infections as here naïve mice , without manipulation of the precursor T cell compartment , and a natural mouse pathogen are employed . This is in contrast to approaches where either mice transgenic for a particular epitope or where large numbers of epitope-specific naïve cells are first introduced and then their development measured[22 , 44] . Our studies indicate that two inflating responses from two different infections are able to develop at the same time . In scenarios of sequential infection , newly developing inflated cells did not impact upon the pre-existing population of effector T cells . Similarly , the presence of large existing MCMV-specific populations of inflated cells did not limit the size of the new developing Ad-lacZ inflating population . Thus the host is able to support multiple populations of epitope-specific T cells which may develop at different periods of time without any cost to the size of the existing or the newly generated effector memory population . In the case of conventional/central memory cells , a pre-existing MCMV infection may result in lower numbers of memory T cells forming during the early stages of other subsequent infections but does not appear to affect the population size at the later time-points . These findings are comparable to results in human studies of CMV , where primary CMV was found to induce expansion of the memory T cell compartment in children but did not affect their responses to vaccination [24 , 28 , 45] . Likewise , CMV reactivation in transplant patients resulted in an increase in the size of the effector and effector memory CD8 T cell memory compartment but this did not impact upon the development of the regenerating naïve and central memory compartment [20 , 46] . Nevertheless , pre-existing memory may be depleted by the acquisition of certain infections . We find that primary MCMV infection is able to cause systemic depletion of the inflating effector memory as well as central memory subsets , which do not fully recover in the long term . It has already been reported that CMV and MCMV infection alters the proportion of naïve and memory cells in the host , mainly by increasing the absolute numbers of MCMV-specific memory cells [20 , 40 , 41 , 45 , 47 , 48] . This was also observed in our model MCMV infection , where the proportion of the naïve subset was very much reduced after high-dose MCMV and also after persistent Ad-lacZ immunization , which also induces large numbers of inflating memory cells . It is interesting to note that the absolute number of cells increased in the organs after MCMV infection , possibly a result of using naïve specific-pathogen-free mice . However apart from diminishing the naive subset , MCMV infection also reduced the absolute numbers of pre-existing Ad-lacZ-specific effector and central memory cells in the liver and lung[29 , 49] . Viral infections , including herpesviruses have been reported to induce lymphopenia leading non-specific attrition of the memory response [22 , 50 , 51] . Our data indicates that acute MCMV infection induces rapid loss of effector memory population and this was observed in all the compartments surveyed , except the lymph nodes . Importantly , a concomitant increase in D8V+ cells was not observed in the surveyed organs , including the lymph nodes , strongly arguing against altered trafficking of these cells . Although ex vivo analysis of D8V+ cells early after MCMV infection did not detect any large apoptotic populations in the blood or tissues , bystander attrition of memory CD8 T cells after acute viral infection has been described previously in humans [52]and mice [53] . Gene expression analysis comparing genes expressed in D8V+ cells versus naïve CD8 T cells and also MCMV-specific inflating M38 cells indicated that genes in the apoptotic pathway were more enriched in this population , and in line with this , we observed an increase in Fas expression on CD8 T cells from Ad-lacZ immune mice upon MCMV infection . Furthermore , a greater proportion of these Fas-upregulated CD8 T cells died in culture compared to their Ad-lacZ counterparts . Furthermore , involvement of this pathway was confirmed by in vivo treatment with blocking anti-FasL monoclonal antibody at the time of MCMV infection , which prevented the attrition of the pre-existing D8V+ population after MCMV infection . As MCMV has been reported to poorly infect CD8 T cells , [54] we hypothesize that this may be an indirect effect of viral infection . Responses to simultaneous MCMV and Ad-lacZ infection indicate that MCMV does not impact on actively developing memory responses . While this study has not able to pin down the full pathway leading to death of these cells , our results indicate that attrition is not dependent on IL-18 . IL-18 is an early innate cytokine which is a key component of the inflammasome pathway and is also responsible for stimulating IFN-γ production [55 , 56] . In agreement with others [32] we found that NK activation during acute MCMV infection is impaired in IL-18 receptor deficient mice , leading to increased levels of viral DNA in the salivary glands of these animals ( S9 Fig ) . Rydyznski et al [57] have reported that NK cells activated during acute viral infection inhibits the generation of long-lived virus-specific memory T cells by killing CD4 and Tfh cells shortly after infection . The increased size of the D8V-specific population after Ad-lacZ immunization may be a reflection of impairment of this process , as depleting NK cells during the acute infection increased the magnitude and quality of the memory response [30 , 57] . However , attrition of the pre-existing inflating population after MCMV infection was also observed in IL-18RKO mice . Further experiments directly targeting NK cell populations could explore whether NK cells have a non-redundant role in this process . How this depletion is maintained in the long term is still unclear . After 100 days post-MCMV , the D8V+ pool was able to recover to pre-depletion levels when injected with another dose of Ad-lacZ virus , indicating that in the long term , the reduced pool of D8V cells which survived the MCMV-mediated attrition were not impaired and still retained the capacity to recognize and proliferate in response to antigen . Also , when followed out to 100 days post-MCMV infection ( approximately 25% of a mouse’s lifespan ) , slow recovery of D8V+ CD8 T cells was observed in the liver and lung . More D8V+ cells accumulated in the lungs and this may be due to to the higher levels of beta-galactosidase expression in the lung compared to liver , as shown by Bolinger et al [20] leading to larger expansion of the D8V-specific population owing to increased antigen levels . Together , this would suggest that the limiting factor preventing recovery of this specific population is the availability of antigen presented by antigen presenting cells . Inflationary cells are associated with ongoing antigen presentation and recruitment and studies suggest that many of the inflationary cells are T cells that are accessible to the blood [23 , 54] . Maintenance of a sizeable population of effector T cells may therefore depend on contact of these cells with relevant APCs in the lymph node or other sites . During an acute MCMV infection , activated CD8 T cells and other cells in the vicinity of the activated T cells will upregulate FasL which will kill not only the D8V+ inflating cells but may impact on APCs with which they interact [58] . The data suggest that a pool of APCs are maintained at some level as the memory population is stabilized at a relatively high frequency . The difference observed between MCMV and Adenovirus-induced inflation in this study may reflect the ability of MCMV to replenish the pool of APCs , a situation which does not occur with the non-replicative vector . MCMV specific populations show remarkable resilience following in vivo depletion , likely through viral reactivation and re-encounter with relevant APCs . [59] Attrition was also observed after i . v . infection with Listeria but not after systemic vaccinia infection or Adenovector immunization , which would indicate the differential activation state induced by different infections plays a role in developing activation induced cell death . In support of this hypothesis several groups have reported that human Adenovirus serotype 5 induces lower levels of innate cytokines than other adenoviral serotypes in mice , monkeys and human cells in vitro [60–62] . The results of this study demonstrate that , in addition to the type of pathogen , the timing of pathogen acquisition may also be crucial in shaping the existing memory compartment . CMV is usually acquired very early in life and our data would suggest that in this scenario CMV might have a low impact in shaping the host’s immunologic memory . By contrast , acquisition of primary CMV later in life may have a larger impact on the immunological memory . Control of clinically relevant infections such as T . gondii [37 , 38] , Trypanosoma cruzi [39] and malaria [3] have been reported to rely on the long-term presence of large pools of antigen-specific effector CD8 T cells . Therefore , the findings of this study may explain how the timing of acquisition of certain unrelated infections may impact upon the efficacy of naturally acquired or vaccine-derived T cell immunity . Additionally , T cell vaccination strategies have been developed using recombinant Adenoviruses or CMV to generate protective inflating response in peripheral tissues against major pathogens and some have shown success to date , with translation into human trials [4–9] . The results here would suggest that immunization with adenoviral vectors allows for the generation of robust vector-specific memory while at the same time preserving the existing memory pool .
This was a controlled laboratory study to address the effects of different viral infections on pre-existing CD8 T cell memory . Groups of mice were infected with Ad-lacZ or MCMV or both in sequence . The levels of pre-existing and newly developing inflating CD8 T cell epitopes were measured during the acute ( day 7 to 14 ) and chronic phase ( day 21 to day >50 ) of the second infection either in the blood or in organs . Experiments were typically performed in duplicate or triplicate , in groups of N = 2–5 per timepoint . In these experiments we were observing large differences in the size of the inflating responses that were maintained over time , which increased the power of detection . Mouse experiments were performed according to UK Home Office regulations ( project license number PPL 30/2235 and 30/2744 ) and after review and approval by the local ethical review board at the University of Oxford . Mice were cared for in accordance with institutional guidelines . MCMV strain ( Strain Smith; ATCC: VR194 ) was used and kindly provided by Professor U . H . Koszinoswki , Department of Virology , Max von Pettenkofer Institute . MCMV was propagated and titrated on NIH 3T3 cells ( ECACC ) , stored at -80°C , and injected i . v . at a dose of 2x106 pfu . Recombinant adenovirus expressed the β-gal protein under the control of the human CMV promoter ( Ad-lacZ [63] ) . Ad-LacZ was propagated on HER-911 cells and purified with the Vivapure AdenoPack 20 ( Sartorius; Stedim Biotech ) . Virus titer was determined in a cytopathic effect assay [63] . Ad-LacZ was stored at -80°C in PBS and injected i . v . at a dose of 2x109pfus . Vaccinia virus ( VVWR , ATCC: VR1354 ) was injected i . v . at 2x106 pfu . Listeria-OVA was injected i . v . at a dose of 1x103cfu . C57BL/6 mice ( Harlan ) and IL-18R transgenic knock-out mice were obtained from Dr Kevin Maloy ( Oxford ) , kept under conventional conditions in individually ventilated cages , and fed with normal chow diet . Mice aged 6–8 weeks old were infected with Ad-lacZ or MCMV and then with the second virus or Listeria-OVA at the indicated time points . Mice were immunized with 1x109pfu Ad-lacZ , . then >55 days later mice were injected i . v or i . p with 100μg/100μl per mouse of LEAF purified anti-FasL ( clone Kay-10 , Biolegend ) or IgG2b isotype control ( clone MG2b-57 , Biolegend ) . At the time of treatment or 1 day later , mice were infected with 1x106pfu MCMV i . v . Antibody treatment was repeated 2 days post-MCMV infection . Peripheral blood lymphocytes ( PBL ) were sampled by tail vein puncture and collected in tubes containing FACS buffer ( PBS , 2% BSA and EDTA ) . Perfused livers were passed through a cell strainer ( BD ) , and lymphocytes were purified through a Percoll ( GE Healthcare ) gradient centrifugation . Lungs were minced with razor blades and incubated in PBS containing 60 U/ml DNase ( AppliChem ) and 170 U/ml collagenase II ( Life Technologies ) at 37°C for 45 min . Cell aggregates were dispersed by passing the digest through a cell strainer ( BD ) . The single cell suspension was spun down and then red blood cells were lysed with RBC buffer . The lymphocytes were stained for flow analysis . In some experiments , the cell suspensions were cultured in 10% RPMI ( Sigma ) supplemented with L-Glutamine , sodium pyruvate , non-essential amino acids and HEPES buffer ( all from Sigma ) in a 37°C , 5% C02 incubator prior to flow analysis . Single-cell suspensions were generated from the indicated organs , and cells were incubated with the indicated mAb at 4°C for 20 min . For PBL samples , erythrocytes were lysed with FACS Lysing Solution ( BD Pharmingen ) . Cells were analyzed by flow cytometry using a BD LSR II flow cytometer and FlowJo ( Treestar ) , gated on viable leukocytes using the live/dead fixable near-IR dead cell stain kit from Invitrogen . MHC class I monomers complexed with M38 ( H-2Kb ) , M45 ( H-2Db ) , βgal D8V ( H-2Kb ) I8V ( H-2Kb ) , m164 ( H-2Dd ) and pp89 ( H-2Ld ) were produced at the NIH Tetramer Core Facility ( Atlanta , Georgia , USA ) and tetramerized by addition of streptavidin-PE ( BD Bioscience ) or streptavidin-APC ( Invitrogen ) . Aliquots of 100μl of whole blood were stained using 50μl of a solution containing tetrameric class I peptide complexes at 37°C for 20 min followed by staining with mAbs . Antibodies were obtained from eBioscience , BD Bioscience , BioLegend , Abcam , R&D Systems , and Jackson ImmunoResearch Laboratories . Splenocytes were prepared from 4 mice infected with Ad-lacZ for 47 days . Spleens were passed through a cell sieve to generate single-cell suspension . Suspension was subjected to RBC lysis , spun down then the lymphocytes counted . The cells resuspended in PBS . These were labeled with 10μl CFSE at 10M , washed , counted then resuspended in DPBS . The cell suspension was passed through a 40μl cell sieve just before in vivo transfer by i . v . injection . Recipient mice were either infected with MCMV for 22 hours or naïve controls . Each mouse received 2x107 CFSE-labelled splenocytes in 200μl volume . GSEA was performed using the Broad Institute java desktop application ( http://software . broadinstitute . org/gsea/ ) [64 , 65] . Analysis used the KEGG apoptosis gene set from the Molecular Signature Database v6 . 0 [64]sourced from http://www . genome . jp/kegg/pathway . html . Microarray data used for GSEA was from sorted MCMV specific T cells at day 50 post infection and Ad-lacZ specific T cells at day 100 post infection and can be found at GEO: GSE73314 ) . Peripheral blood was collected from groups of mice ( N = 3 pooled per group ) . RBC lysis was performed and then lymphocytes stained with tetramer-PE . Positive selection using anti-PE beads ( Miltenyi ) was performed and purity confirmed by FACS analysis . Purity of CD8+ Tetramer+ cells were between 87–98% . Genomic DNA was extracted from sorted cells using the Miniprep Blood DNA kit ( Qiagen ) and quantitated . Taqman qPCR ( ThermoFisher ) measuring TREC DNA was performed using the protocol and B6-specific primer and probe sequences described by Broers et al [66] . To compensate for variations in input DNA , the constant region of the TCRA ( Cα ) gene was used as endogenous reference gene . Relative quantitation was performed using the 2-ΔΔCT equation . Statistical data analysis was conducted using GraphPad Prism6 software ( San Diego , CA , USA ) . The Student’s T-test , ANOVA and the Mann-Whitney tests were employed where indicated . A p value of <0 . 05 was considered statistically significant . | Many natural pathogens elicit large antigen-specific long-lived effector memory CD8 T cell responses that contribute to pathogen protection and control . This strategy is also employed by novel CD8 T cell vaccine vectors . In the real world , the host is likely to encounter a number of such pathogens at different times but the effect of pre-existing effector memory pools on the development of new CD8 T cell responses against unrelated infections , and vice versa , is still unclear . We address this through mouse models of sequential infection with MCMV and an Ad-Hu5 vaccine vector . We find that the host is able to accommodate and accumulate multiple new specificities of inflating effector memory populations at different times but certain infections , such as acute MCMV , appear to rapidly deplete existing memory populations via a Fas-dependent pathway . Additionally , immunization with vectors based on human adenovirus subtype 5 do not appear to detrimentally affect the host’s pre-existing effector memory pool . ( 154 words ) | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"animal",
"models",
"of",
"disease",
"immunology",
"cell",
"processes",
"microbiology",
"cytot... | 2017 | Adenoviral vaccine induction of CD8+ T cell memory inflation: Impact of co-infection and infection order |
Cryptococcus neoformans is a heterothallic fungal pathogen of humans and animals . Although the fungus grows primarily as a yeast , hyphae are produced during the sexual phase and during a process called monokaryotic fruiting , which is also believed to involve sexual reproduction , but between cells of the same mating type . Here we report a novel monokaryotic fruiting mechanism that is dependent on the cell cycle and occurs in haploid cells in the absence of sexual reproduction . Cells grown at 37°C were found to rapidly produce hyphae ( ∼4 hrs ) and at high frequency ( ∼40% of the population ) after inoculation onto hyphae-inducing agar . Microscopic examination of the 37°C seed culture revealed a mixture of normal-sized and enlarged cells . Micromanipulation of single cells demonstrated that only enlarged cells were able to produce hyphae and genetic analysis confirmed that hyphae did not arise from α-α mating or endoduplication . Cell cycle analysis revealed that cells grown at 37°C had an increased population of cells in G2 arrest , with the proportion correlated with the frequency of monokaryotic fruiting . Cell sorting experiments demonstrated that enlarged cells were only found in the G2-arrested population and only this population contained cells able to produce hyphae . Treatment of cells at low temperature with the G2 cell cycle arrest agent , nocodazole , induced hyphal growth , confirming the role of the cell cycle in this process . Taken together , these results reveal a mating-independent mechanism for monokaryotic fruiting , which is dependent on the cell cycle for induction of hyphal competency .
Cryptococcus neoformans is a basidiomycetous fungal pathogen of humans and animals that typically causes opportunistic infections in patients with cellular immune defects [1] . Infection initiates in the lungs and frequently disseminates to the brain where it manifests as a fatal meningoencephalitis if untreated . AIDS patients are at increased risk for infection , though infection rates have decreased significantly with better AIDS management [2] . However , in spite of the reduction in AIDS-related cases , cryptococcosis remains a frequent life-threatening opportunistic mycosis for these patients in underdeveloped countries , and is a recently emergent disease in the United States Pacific Northwest [3] for as yet , unexplained reasons . Naturally occurring strains of C . neoformans are heterothallic with two mating types , MATa and MATα , with both mating types being pathogenic , although most clinical isolates are MATα [4] . Although the taxonomy has been changing , four serotypes have been described ( A , B , C , D ) with serotypes A and D often referred to as C . neoformans variety grubii and variety neoformans respectively , and serotypes B and C being collectively referred to as C . neoformans variety gattii , or more recently , C . gattii [5] . For all serotypes , throughout the course of infection and under normal culture conditions , the fungus grows as an encapsulated yeast . Under appropriate in vitro conditions , however , the fungus can produce two kinds of hyphae; dikaryotic hyphae during MATa×MATα sexual reproduction and monokaryotic hyphae ( from individual MATa or MATα strains ) during monokaryotic fruiting [6] . Basidiospores can be produced from both hyphal types . Environmental factors required for sexual reproduction and monokaryotic fruiting are similar and include culture under low temperature ( 25°C ) , low moisture , and nutrient limitation [6] . Many genes , including homologs of the Saccharomyces cerevisiae pheromone response pathway , are required to produce both types of hyphae [7] , [8] , [9] , [10] , [11] . There are , however , distinct differences between the two hyphal types . Structurally , dikaryotic hyphae have fused clamp connections and a pair of nuclei ( one MATa and one MATα ) per hyphal compartment while monokaryotic hyphae have unfused clamp connections and a single nucleus per hyphal compartment . Sexual reproduction in C . neoformans has been characterized in detail and largely follows the pheromone response paradigm that has been developed from decades of S . cerevisiae research [12] . Less clear is the mechanism by which monokaryotic fruiting occurs . Recent studies have concluded that monokaryotic fruiting in C . neoformans variety neoformans can result from α-α mating [13] and may be an important part of the natural life cycle of this fungus [14] , with possible implications for human disease [15] . We recently reported that high temperature seed culture conditions could induce very robust monokaryotic fruiting in C . neoformans variety neoformans [11] and assumed these growth conditions enhanced α-α cell fusion . However , this assumption proved to be false . Instead , we found that cells arrested in the G2 stage of the cell cycle were competent to undergo monokaryotic fruiting at high frequency , in the absence of α-α cell fusion . Importantly , this mechanism proceeds through enlarged cells , a morphological phenotype that has been increasingly observed in vivo and is hypothesized to serve as a strategy for avoiding host defenses [16] , [17] , [18] . These results demonstrate that C . neoformans has evolved a number of different mechanisms for modifying cellular morphology to suit its specific environment , with some of these mechanisms contributing to the success of this fungus as a pathogen .
Previous studies of monokaryotic fruiting typically were performed by patching cells from a seed culture onto filament agar and then screening for a hyphal fringe around the periphery [6] , [11] , [19] . Our recent observation of the role of temperature in this process [11] led us to test whether or not high temperature increases the intensity of hyphae production after inoculation onto filament agar , or increases the number of cells that produce hyphae . A suspension of cells from a 24 h , 37°C seed culture was spread onto filament agar to observe monokaryotic fruiting in individual cells , which would reveal whether or not all cells from the seed culture were capable of undergoing monokaryotic fruiting . Figure 1A demonstrates that only part of the population grown at 37°C was able to undergo monokaryotic fruiting and that hyphae began to appear as early as 4 hours after plating onto filament agar ( Fig . 1B ) . This phenomenon was not a strain artifact nor was it restricted to a single serotype as all four serotypes were found to be able to produce hyphae under the above inducing conditions ( Fig . 1C–F ) . These results suggest that only a specific type of cell is capable of undergoing monokaryotic fruiting , and that this capability requires seed culture conditions that enable these cells to become competent for hyphal production . Lin et al . , have demonstrated that same-sex mating is one way in which monokaryotic fruiting can occur [13] . However , because the previous experiment showed that individual cells were still capable of monokaryotic fruiting in spite of being well separated from potential mating partners on spread plates , we hypothesized that either same-sex mating occurs during the seed culture growth period , or that another mechanism , which does not involve same-sex mating , could also lead to monokaryotic fruiting . To investigate these two possibilities , we first screened for evidence of α-α mating during monokaryotic fruiting using the α-α cell fusion assay to test different combinations of complementing MATα auxotrophs that would be predicted to yield prototrophic colonies on unsupplemented MIN agar . No fusants were observed from MATα×MATα crosses plated onto MIN agar after growth as a seed culture on YPD at 37°C for 24 h . However , assisted α-α matings showed that each of these strains was capable of undergoing α-α fusion , ruling out an α-α cell fusion defect ( Fig . 2A ) . The most likely explanation for these results was that because α-α fusion is a rare event [13] , our conditions , although not detecting an α-α fusion , were still not excluding this possible mechanism . To exclude a cryptic fusion event as an explanation for our observations , we utilized a cpk1Δ mutant to further test whether or not α-α fusion was required for temperature-induced monokaryotic fruiting . Cpk1p , the MAP kinase in the C . neoformans pheromone response pathway , is required for α-a fusion during sexual reproduction as well as α-α fusion during monokaryotic fruiting [9] , [13] . Therefore , if α-α fusion was the mechanism of monokaryotic fruiting in our system , the cpk1Δ mutant should not produce hyphae after plating onto filament agar because it could not fuse with the complementing strain . Our results showed that neither assisted nor unassisted α-α mating reactions ( WSA-2126×WSA-591×WSA-65 or WSA-2126×WSA-591 ) with the cpk1Δ mutant ( WSA-2126 ) showed evidence of α-α cell fusion ( Fig . 2B ) . However , when WSA-2126 was tested for monokaryotic fruiting ability after high temperature seed culture , this strain fruited normally ( Fig . 2B ) . These results confirmed that monokaryotic fruiting could occur independently of α-α fusion . The production of enlarged cells in C . neoformans has been reported to occur when cells are exposed to opposite mating type cells in standard mating type mixes [11] , and in confrontation assays , which are performed by streaking cells of opposite mating type in close proximity to each other [20] . Clinical studies have also found this cell type in vivo [16] , [18] , [21] , [22] . Because only certain cells produced hyphae during the quantitative monokaryotic fruiting assay , we decided to determine if there were developmental differences among cells after high temperature seed culture . Microscopic observation of wet mounts prepared from cells scraped off of filament agar after growing as a seed culture at 37°C showed that filaments always originated from enlarged cells ( Fig . 3A ) . When cells from 30°C and 37°C seed cultures were screened for enlarged cells , we only observed enlarged cells from the 37°C seed culture , although this incubation temperature produced both enlarged and smaller , normal-sized cells ( Fig . 3B and 3C ) . To confirm that the enlarged cells were responsible for the production of hyphae during monokaryotic fruiting , large and small cells from a 37°C seed culture were micromanipulated onto filament agar and then monitored for hyphae production . Analysis of hyphae production from each single cell revealed that small cells only grew into yeast cells while the large cells grew as both yeast and hyphae ( Fig . 3D ) . These results demonstrated that only a subset of the population grown at high temperature , which can be distinguished by the larger cell size , becomes competent to produce hyphae . The need for enlarged cells prior to the initiation of monokaryotic fruiting suggested that the mechanism for production of this phenotype possibly involved changes in cell ploidy since yeast ploidy has been noted to be associated with cell size [23] . In C . neoformans , monokaryotic fruiting has been shown to result in ploidy changes of yeast cells produced specifically from the hyphal filaments [13] . These observations suggested to us that monokaryotic fruiting may occur through ploidy changes , which are manifested as enlarged cells that arise from an endoduplication event , as has been previously suggested [24] . Cell cycle analysis of seed cultures grown at different temperatures ( 25°C , 30°C , 35°C , 37°C and 40°C ) all revealed only 1n and 2n DNA content peaks , with no evidence of a 4n peak ( Fig . 4A ) . This result showed that there was no ploidy change after high temperature ( 35°C , 37°C , 40°C ) seed culture growth , demonstrating that the enlarged cells , which were responsible for monokaryotic fruiting , were haploid . Additionally , we found that as seed culture temperature increased from 25°C to 40°C , the percentage of cells in G1 decreased , the percentage of cells in G2 increased , and the percentage of cells in S phase was similar until dropping almost to 0 at 37°C ( Fig . 4B ) . DAPI staining confirmed the relationship between cell size and G2 arrest as the staining patterns of large and small cells differed ( Fig . 4C ) . The smaller cells displayed a compact nuclear staining pattern while the larger cells displayed a larger , diffuse staining pattern , which has been observed in other G2-arrested fungi [25] , [26] . These results demonstrated that the effect of increasing temperature on monokaryotic fruiting involved the cell cycle , specifically G2 arrest . In spite of the FACS results showing that hyphal-competent cells were haploid , we could not rule out a change in ploidy just before , or during growth in the hyphal phase . Unfortunately ploidy determination by FACS analysis of hyphae is physically restricted by the filamentous characteristics of the cells . However , hyphal ploidy can be determined using a blastospore assay [13] . The results of this assay revealed that 76 out of 78 blastospores ( 97% ) were haploid , demonstrating that the hyphae produced during monokaryotic fruiting were haploid , which again excluded α-α mating or endoduplication as monokaryotic fruiting mechanisms in our system ( Fig . 5A ) . As a further control , the two diploid blastospores were sub cultured repeatedly for an additional two weeks and then retested by FACS , which revealed that they remained diploid , ruling out the possibility that haploid blastospores could be segregation products of diploid blastospores . While a late endoduplication event could explain the two diploid spores , this possibility is unlikely since the 78 spores were picked from 78 independent hyphae . However based on the recovery of two diploid blastospores in our assay , and the rarity of basidiospore production during monokaryotic fruiting , we hypothesized that there could be two hyphal types produced during monokaryotic fruiting , which could be distinguished by ploidy . One type could be vegetative in nature and not undergo basidiosporogenesis ( haploid ) and a second type could be generated that ultimately produced basidiospores ( diploid ) . To address these two possibilities , 37°C seed culture cells were spread onto filament agar and the resultant colonies screened for basidiospore chains . The blastospore assay was performed on blastospores recovered from hyphae with and without basidiospores . Cell cycle analysis again showed that all of the blastospores were haploid , regardless of whether or not the hypha produced basidiospores ( Fig . 5B ) , which was consistent with our previous observations that excluded α-α mating or endoduplication as the mechanisms of monokaryotic fruiting . Together , these results demonstrate that endoduplication is not required for monokaryotic fruiting as these hyphae are produced from haploid cells and remain haploid , in spite of being able to occasionally generate diploid yeast cells . To determine whether only G2-arrested cells undergo monokaryotic fruiting , we sorted G1 and G2 phase , 37°C seed culture cells according to DNA content . Microscopic observation of sorted G1 and G2 cells revealed that G2 phase cells were much larger than G1 phase cells ( Fig . 6 ) . We next sorted live 37°C seed culture cells according to cell size and performed cell cycle analysis on the two populations . The results indicated that the smaller cells had a single DNA content peak at the 1n position while the enlarged cells had a single DNA content peak at the 2n position , which lead us to conclude that the small cells were G1 phase cells and the large cells were cells in G2 arrest ( Fig . 7 ) . The two populations were then assayed for monokaryotic fruiting ability , which revealed that monokaryotic fruiting was a property only of the G2 fraction ( Fig . 7 ) . As a final confirmation that monokaryotic fruiting requires G2 arrest , cells were treated with nocodazole , a G2/M arrest agent that inhibits and disassembles microtubules [27] . Cells were grown as seed cultures at 30°C ( the non-permissive seed culture temperature ) , then assayed for monokaryotic fruiting ability . The experiment showed that cells treated with nocodazole became competent to produce hyphae on filament agar in a dose-dependent manner even though they were grown as a seed culture under conditions that did not normally lead to monokaryotic fruiting , whereas untreated cells only grew as yeasts ( Fig . 8 ) . Taken together , these results demonstrate that G2-arrested cells can serve as a starting point for cells that undergo monokaryotic fruiting .
In this study we have identified a novel mechanism in C . neoformans that leads to the production of hyphae , with or without basidiospores , by haploid cells ( monokaryotic fruiting ) . This mechanism appears to be dependent on the cell cycle and initiates from cells in G2 arrest . It occurs in the absence of α-α mating and/or endoduplication , thereby demonstrating that monokaryotic fruiting can occur asexually . Previous studies have shown that sexual reproduction can occur between cells of the same mating type , resulting in monokaryotic fruiting [13] , and that this phenomenon occurs in nature [24] . Under the specific conditions of this study , notably a 37°C seed culture temperature , we saw no evidence of an α-α cell fusion event , nor did we find evidence of endoduplication within the hyphae even though we screened hyphae that had produced basidiospores . During C . neoformans basidiosporogenesis , sexual reproduction results in meiosis in the basidium followed by successive mitotic divisions that yield the nuclei , which ultimately are inserted into spores as they form on the basidial surface [20] . Lin et al . observed that when fruiting was derived from an α-α fusant , sporogenesis was robust with spore chains that were long and phenotypically similar to α-a mating during sexual reproduction [13] . This process was found to be impaired in dmc1 mutants , which are meiotic mutants that still produce spores , but at a much lower frequency than sexually produced spores , and with truncated spore chains that sometimes occur as dyads ( two rather than four chains ) [13] . The phenotype of basidiospores produced in the dmc1 strains was strikingly similar to what we observed in this study and what was previously reported [6] . These observations may suggest that basidiosporogenesis can occur mitotically without meiosis , although we cannot exclude a duplication event in the basidium immediately followed by meiosis and sporogenesis . We did , however , test a dmc1 mutant and found that it was able to undergo monokaryotic fruiting under our conditions ( data not shown ) . Because our study was done in the same strain background as the study by Lin et al . , we reviewed the conditions of both experiments and found some differences that may explain the contrasting differences in ploidy . Our study used a high temperature seed culture condition , which results in rapid hyphae production upon filament agar plating . The seed culture conditions in Lin's study were not clear , however , their plating medium was V8 agar , which is normally used for mating C . neoformans , and their incubation period was for a period of weeks , whereas we screened at 24 hrs and observed hyphae in as little as four hours , although cells also produced hyphae on V8 agar under our conditions . Both filament agar and V8 agar have high agar contents; however , V8 agar is an undefined medium with V8 juice as the basal ingredient . Filament agar , on the other hand , is a low-glucose , defined medium with Yeast Nitrogen Base without amino acids and without ammonium sulfate as the source of vitamins and cofactors . Although both media are starvation media , they are substantially different in composition , which may be one explanation for the differences in hyphal types that we observed . The seed culture conditions , or more precisely , the cell cycle stage may be another explanation . Our initial investigation of the relationship between the cell cycle and monokaryotic fruiting focused on detecting what we presumed would be a transition to diploidy in yeast cells at some point during seed culture growth , which we believed would coincide with the appearance of enlarged cells in the seed culture and the association of this cell type with the hyphal progenitor . The high frequency of fruiting and enlarged cells in the seed culture suggested that detection of the diploidization event would be unambiguous . However , the data showed that instead of an α-α fusion or endoduplication event , which would result in a diploid cell , the actual mechanism that resulted in hyphae production was G2 arrest . The reason for the requirement of G2 arrest to induce hyphae is not clear , and while G2 arrest is required for monokaryotic fruiting , not all arrested cells produced hyphae . We suspect that the subpopulation of non-hyphal , G2-arrested cells consisted of cells that escaped G2 , and then proceeded to bud rather than differentiate into a hypha . These two outcomes resemble the decision point that a pheromone-exposed , G2-arrested Ustilago maydis cell faces with regard to which of the two developmental paths it will follow ( conjugation tube formation or budding ) [26] . How the generation of occasional diploid blastospores occurs is also not clear . Given that C . neoformans hyphae produce typical basidiomycete-like clamp connections , the incomplete fusion of these structures in monokaryotic hyphae combined with an aberrant segregation event during the budding of blastospores off of the hyphal compartment may yield the diploid cells that we observed at low frequency . Previous studies of the C . neoformans cell cycle have identified a number of stressors that cause G2 arrest , including oxygen depletion [28] , stationary growth phase [29] , and temperature [30] . Under our experimental conditions , hyphal competency occurs prior to transfer to hyphal inducing conditions ( starvation on filament agar ) and not during growth on filament agar . Therefore , stationary growth phase is not a factor nor is oxygen depletion since cultures were grown on the agar surface , and only for 24 hrs . Consequently , growth temperature seems to be responsible for inducing competency . Under our conditions , the 37°C incubation temperature differs from the original incubation temperature ( 30°C ) for monokaryotic fruiting [6] , which suggests that temperature is an important variable . In fact , while we did not see the temperature effect on all strains of C . neoformans , we were able to induce hyphae in all four serotypes . Interestingly , the reports of enlarged cells in vivo [16] , [18] , [21] , [22] reflect growth at elevated temperature in the mammalian body . Other pathways have also been shown to influence C . neoformans cell size in vitro , including cAMP , RAS , and PKA [16] , [31] , [32] , suggesting the possibility of conserved stimuli that may regulate these pathways . Presently , the STE12α signal transduction pathway seems to be the major or sole regulator of monokaryotic fruiting as this gene is required for monokaryotic fruiting regardless of inducing conditions . Fruiting still occurred normally in cpkΔ1 ( pheromone response pathway ) , and cacΔ1 ( cAMP pathway ) mutants , ruling out these pathways as regulators of temperature-induced monokaryotic fruiting . Other pathways , such as the calcineurin signal transduction pathway cannot be ruled out , but are more complicated to test since some mutants in this pathway do not grow at high temperature [33] . A key characteristic of enlarged cells in vivo appears to be polyploidy , which has been hypothesized to arise when the M phase of the cell cycle is skipped [16] , [18] . This enlarged cell phenotype appears to be a potential virulence factor as they are poorly phagocytized , if at all [18] . Perhaps increasing cell size evolved as a physical defense mechanism against predatory grazers , which in turn , protects cells from being phagocytized in vivo via the same mechanisms . It appears that the enlarged cell phenotype can be produced by multiple mechanisms: cell cycle arrest , a-α or α-α cell fusion , and endoduplication , each of which may have a different purpose . The developmental options available after cell cycle arrest may have been selected for in C . neoformans to enhance survival in its specific environmental niche while inadvertently creating an important human fungal pathogen . What remains to be determined is how high temperature generation of the large cell and monokaryotic fruiting phenotypes was incorporated into the evolution of C . neoformans . With the exception of C . neoformans , virtually all members of this genus grow poorly or not at all at mammalian ambient temperature . In contrast , all of the major human fungal pathogens grow at 37°C and virtually all of them have a hyphal phenotype . Perhaps the association of many basidiomycetes with rotting wood or decaying vegetation in general led to high temperature exposure and subsequent genetic selection during self-heating , compost-like conditions caused by microbial metabolism of organic matter . Once nutrients were consumed , hyphal extension towards additional nutrients and/or sporulation in the absence of nutrients could have completed the evolution of C . neoformans into a pathogen via development of a mechanism of infectious particle ( basidiospores ) dispersion combined with the ability to grow at elevated temperatures . Selection for the molecular linkage of pathways controlling cell cycle , nutrient sensing , and ultimately , differentiation , could have been the outcome of this lifestyle and allowed the fungus to coordinately regulate these pathways , thus enabling it to effectively exist as a saprophyte or pathogen .
YPD agar , MIN agar , V8 agar , and filament agar were prepared as described previously [6] , [7] , [11] with or without amino acids or nucleic acid supplements as required . JEC-21 is a wild type , MATα isolate that was used in the initial characterization of monokaryotic fruiting in C . neoformans [34] . WSA-79 is a serotype D clinical isolate from Maryland , WSA-522 is a serotype A clinical isolate from Thailand , WSA-533 is a serotype B environmental isolate from Australia , and WSA-2507 is a serotype C clinical isolate from Maryland . Additional strains , all of which were derived from the original JEC-21 - JEC-20 congenic pair [34] , consisted of the following genotypes: WSA-1 ( MATα lys2 ) , WSA-70 ( MATα ade2 lys2 ) , WSA-591 ( MATα ade2 ) , WSA-1226 ( MATα ura5 ) , WSA-2126 ( MATα ura5 cpk1Δ::ura5 ) , WSA-3002 ( MATα/MATα fusant from WSA-1226×WSA-70×WSA-68 ) , WSA-3019 ( MATα haploid blastospore ) , WSA-3070 ( MATα/MATα diploid blastospore recovered from JEC-21 hypha without basidiospore chains ) , WSA-3098 ( MATα haploid blastospore recovered from JEC-21 hypha with basidiospore chains ) , WSA-3112 ( MATα haploid blastospore recovered from hypha without basidiospore chains ) , WSA-3145 ( MATα ura5 cpk1Δ::ura5 CPK1::NEOr ) . MATa strains included WSA-65 ( MATa ura5 lys1 ade2 ) and WSA-68 ( MATa ura5 ade2 lys2 ) . Nocodazole ( Sigma-Aldrich , St . Louis , MO ) stock was prepared in DMSO at 1 . 5 mM and then used to prepare different dilutions in YPD broth ( YPD-0 . 075 µm nocodazole , YPD-0 . 15 µm nocodazole , YPD-0 . 30 µm nocodazole , YPD-0 . 60 µm nocodazole , YPD-1 . 2 µm nocodazole ) . JEC-21 cells were added to 1 . 5 ml nocodazole-YPD broth in 15 ml snap cap tubes ( BD Biosciences , Franklin Lakes , NJ ) , at a final concentration of 1×106 cells/ml . Tubes were shaken at 200× RPM at 30°C for 24 hrs . Five µl of overnight culture were then dropped onto filament agar and incubated at 25°C for 5 days . The quantitative assay was performed by growing cells as above , and then plating cells onto filament agar as described in the quantitative monokaryotic fruiting assay . Pictures were taken with an Olympus SZX12 stereo microscope ( Olympus , Center Valley , PA ) at 2× magnification . To perform α-α cell fusions , 1×106 cells of complementary , auxotrophic , MATα strains were mixed , cultured on YPD agar at 37°C for 24 hrs , and then transferred onto filament agar plates . The plates were incubated at 25°C for 24 hrs . The mixture was scraped from the plate , suspended in 1 . 0 ml sterile H2O , and then 200 µl of cells from this suspension were spread onto MIN agar plates , which were then incubated at 30°C for 4 days to screen for fusants . Assisted mating reactions , in which two MATα strains were induced to fuse by including a MATa helper strain , were performed according to previously described methods [19] . Plates were photographed at 0 . 5× magnification . Cells were harvested , fixed , and stained with propidium iodide ( Sigma-Aldrich , St . Louis , MO ) , and then sorted as described by Sia et al . [35] . Cell cycle analysis was performed using a BD FACSCalibur flow cytometer ( Becton Dickinson Biosciences , Sparks , MD ) . CellQuest Pro software was used for cell collection , and data analysis was performed using ModFit and FlowJo . G1 , S , and G2 phases were identified using the Dean-Jett-Fox mathematical model . G2-arrested cells were identified as described [25] , [26] , [36] , [37] , [38] . This population of cells typically shows a large cell phenotype , an increased proportion of cells with 2C DNA content when compared to controls , and a DAPI staining pattern that shows larger nuclei vs . smaller condensed nuclei of G2 phase cells ( see below ) . Cell sortings were performed on a BD FACSAria III ( Beckton Dickinson Biosciences ) cell sorter and analyzed using BD FACSDiva 6 . 1 software . Aliquots of sorted live cells were also used to perform the quantitative monokaryotic fruiting assay . All FACS analysis was performed at the Flow Cytometry Core Laboratory at The University of Texas Health Science Center at San Antonio . Cells were stained with DAPI according to the method described by Fuchs et al . [39] . Briefly , 1×107 cells were harvested from 24 h YPD agar plates , which were grown at either 30°C or 37°C , washed twice with phosphate buffer ( 0 . 1 M KH2PO4 , 1 . 25 mM EGTA , 1 . 25 mM MgCl2 at pH 6 . 9 ) , then resuspended in 400 µl fixative ( 5% paraformaldehyde in wash buffer ) followed by incubation at room temperature for 90 minutes . The cells were then washed three times in wash buffer and incubated with Lysing Enzymes ( Trichoderma harzianum , Sigma-Aldrich , 1 mg/ml in sterile distilled water ) for 20 min at 37°C . The cells were then washed once with sterile water , gently resuspended in 400 µl 0 . 3% Triton X-100 ( Sigma-Aldrich ) , and permeabilized by incubation at room temperature for 15 minutes . The suspension was then pelleted and washed three times with PBS . Cells were stained in DAPI ( Sigma-Aldrich ) ( 0 . 1 , 0 . 2 , or 0 . 5 µg/ml ) for 15 minutes , washed twice with PBS , and then resuspended in 200 µl prior to visualization . Images were captured on a Zeiss AxioImager Z1 microscope ( Carl Zeiss Microscopy , LLC , Thornwood , NY ) equipped with an AxioCam MR3_2 CCD camera , using the filters for DAPI ( 365 nm excitation , 395 nm beam splitter , 420–470 nm emission filter ) and differential interference contrast ( DIC , Nomarski contrast ) . Image analysis and adjustments were performed using Axiovision software ( Zeiss , Version 4 . 8 ) . Images were adjusted only for frame alignment ( to overlay DAPI over DIC ) , brightness , and contrast ( adjustment of +0 . 01 contrast units over default ) , and all images received the same treatment . | Fungi typically grow vegetatively as either yeast or hyphae . Many of the major human fungal pathogens can generate both morphologies and are referred to as the dimorphic fungi . Cryptococcus neoformans is a yeast-like fungus that has not been traditionally thought to be dimorphic since hyphae production typically occurs during the mating process between cells of opposite mating types . However , C . neoformans also can generate the hyphal state from haploid cells ( called monokaryotic or haploid fruiting ) in the absence of the opposite mating type . Recent studies have shown that the mechanism behind this process also involves mating , however , the mating reaction occurs between cells of the same mating type . Here we describe a unique mechanism responsible for monokaryotic fruiting that is independent of mating and does not proceed through a diploid intermediate . Instead , the key requirement for hyphal induction appears to be cell cycle arrest . Importantly , arrested cells display an enlarged cell phenotype , which has been observed in vivo in recent reports and has been hypothesized to be a novel protection strategy against host defenses . C . neoformans appears to have an extensive morphological repertoire , which likely contributes to its success as both a pathogen and a saprophyte . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"fungi",
"mycology",
"medical",
"microbiology",
"biology",
"microbiology"
] | 2013 | The Production of Monokaryotic Hyphae by Cryptococcus neoformans Can Be Induced by High Temperature Arrest of the Cell Cycle and Is Independent of Same-Sex Mating |
Highly pathogenic H5N1 influenza A viruses have spread across Asia , Europe , and Africa . More than 500 cases of H5N1 virus infection in humans , with a high lethality rate , have been reported . To understand the molecular basis for the high virulence of H5N1 viruses in mammals , we tested the virulence in ferrets of several H5N1 viruses isolated from humans and found A/Vietnam/UT3062/04 ( UT3062 ) to be the most virulent and A/Vietnam/UT3028/03 ( UT3028 ) to be avirulent in this animal model . We then generated a series of reassortant viruses between the two viruses and assessed their virulence in ferrets . All of the viruses that possessed both the UT3062 hemagglutinin ( HA ) and nonstructural protein ( NS ) genes were highly virulent . By contrast , all those possessing the UT3028 HA or NS genes were attenuated in ferrets . These results demonstrate that the HA and NS genes are responsible for the difference in virulence in ferrets between the two viruses . Amino acid differences were identified at position 134 of HA , at positions 200 and 205 of NS1 , and at positions 47 and 51 of NS2 . We found that the residue at position 134 of HA alters the receptor-binding property of the virus , as measured by viral elution from erythrocytes . Further , both of the residues at positions 200 and 205 of NS1 contributed to enhanced type I interferon ( IFN ) antagonistic activity . These findings further our understanding of the determinants of pathogenicity of H5N1 viruses in mammals .
In 1997 , the first human case of influenza caused by an H5N1 virus occurred in Hong Kong [1] , [2] . In 2003 , a new outbreak of H5N1 virus was identified in Vietnam . Since then , H5N1 viruses have spread across Asia , Europe and Africa . As of July 22 , 2010 , 501 cases of H5N1 virus infections in humans have been reported by the World Health Organization ( WHO; http://www . who . int/en/ ) , 297 of which were fatal . The mortality is , therefore , approximately 60% . H5N1 viruses have been characterized by using a variety of mammalian models [3] . In mice , enhanced HA cleavability , as well as lysine at position 627 of the polymerase subunit PB2 , plays an important role in the virulence of H5N1 viruses [4] . Viruses possessing these properties replicate systemically and cause death in mice . Ferrets are considered suitable for evaluating infection of human influenza viruses because these viruses replicate in the upper respiratory tract without adaptation in ferrets , and some strains cause severe pneumonia in these animals . Some of the H5N1 viruses isolated from humans can kill ferrets , whereas H5N1 viruses isolated from birds tend to cause mild disease in this animal model [5] , [6] . Systemic infection , high replication efficiencies , and neurovirulence are associated with the high lethality of human H5N1 viruses in ferrets . Salomon et al . [7] reported that the genes encoding the nonstructural proteins ( NS ) and polymerase complex are important for the lethality of the human H5N1 virus A/Vietnam/1203/04 in ferrets , compared with the avian H5N1 virus A/quail/Vietnam/36/04 . However , the molecular bases for the high virulence of H5N1 viruses in ferrets are not fully understood . To advance our understanding of the pathogenicity of H5N1 viruses , we compared the virulence of H5N1 influenza viruses isolated from humans in a ferret model . By generating reassortant viruses between the most virulent A/Vietnam/UT3062/04 ( UT3062 ) virus and the avirulent A/Vietnam/UT3028/03 ( UT3028 ) virus , we identified the genes responsible for high virulence in ferrets . We also performed in vitro studies to determine the molecular mechanisms by which H5N1 viruses exhibit high virulence in mammals .
To compare the virulence of H5N1 influenza viruses isolated from humans in ferrets , we intranasally inoculated 5- to 7-month-old male animals ( n = 3 ) with 107 plaque-forming units ( PFU ) of virus and observed the lethality , changes in body weight and body temperature , clinical signs , and virus shedding in the upper respiratory tract of the virus-infected animals ( Table 1 ) . The UT3062 , A/Vietnam/UT3040/04 , A/Vietnam/UT3028II/03 , A/Vietnam/UT30850/05 , A/Vietnam/UT3030/03 , A/Vietnam/UT3040II/04 , and A/Vietnam/UT3047III/04 were virulent in ferrets , causing the deaths of the virus-infected animals . These virulent viruses , with the exception of A/Vietnam/UT3028II/03 , caused mean maximum weight loss of 9 . 1%–18 . 4% and anorexia , consistent with previous studies [5] , [6] . Systemic viral infection was observed in most of the fatally infected animals ( Table S1 ) . Notably , inoculation of animals with UT3062 resulted in 100% lethality with 15 . 4±2 . 7% mean maximum weight loss ( Table 1 ) . These results demonstrate that UT3062 is the most virulent in ferrets of the viruses we tested . By contrast , six other human H5N1 viruses , A/Vietnam/UT30259/04 , A/Vietnam/UT3035/03 , A/Indonesia/UT3006/05 , UT3028 , A/Vietnam/UT30408III/05 , and A/Vietnam/UT30262III/04 did not kill any ferrets and all , except A/Vietnam/UT30259/04 , caused limited body weight loss ( 1 . 6%–6 . 6% mean maximum weight loss , median 4 . 2% ) ( Table 1 ) , indicating that H5N1 viruses isolated from humans differ in their virulence in ferrets . Among the H5N1 viruses listed in Table 1 , two viruses , A/Vietnam/UT3035/03 and A/Vietnam/UT30408III/05 , which did not kill any ferrets , were isolated from patients who recovered from their H5N1 virus infections . The rest of the viruses used were isolated from patients who ultimately died . Of note , the virulence of test viruses in ferrets generally correlated with that in mice [8] . On days 3 and 6 post-infection ( p . i . ) , we collected nasal washes from the virus-infected animals and titrated them in Madin-Darby canine kidney ( MDCK ) cells . On day 6 p . i . , the virus titers in the nasal washes of animals infected with the virulent viruses ( except for A/Vietnam/UT3028II/03 ) were generally higher than those of animals infected with viruses that were not lethal ( Table 1 ) . Sequence comparisons of the viruses used in this study revealed that the most virulent virus UT3062 was most closely related to an avirulent virus UT3028 , with 18 amino acid differences in their 9 proteins ( Table S2 ) ; there were no amino acid differences in the matrix ( M ) proteins of the two viruses . Therefore , we generated UT3062 and UT3028 by reverse genetics and confirmed their virulence by intranasally inoculating 5- to 6-month-old male ferrets with 107 PFU of the viruses . Animals infected with the UT3062 virus died on days 6–8 , resulting in 67% lethality and showed −13 . 4±3 . 1% mean maximum weight loss ( Figures 1 and 2 ) . Conversely , all animals infected with the UT3028 virus survived and showed appreciably less mean maximum weight loss ( −1 . 1±0 . 3% ) ( Figures 1 and 2 ) . These results were similar to those obtained with the respective original viruses ( Figure 2 and Table 1 ) . To determine the molecular basis for the high virulence of UT3062 , we generated reassortant viruses between the UT3062 and UT3028 viruses using reverse genetics and tested their virulence by intranasally inoculating 5- to 6-month-old male ferrets with 107 PFU of the viruses and observing them for 10 days for clinical manifestations . Since no amino acid differences were identified in M proteins of the two viruses , we used the UT3028 M gene for generating all reassortant viruses . The reassortants were named according to the origin of their UT3062 or UT3028 genes . For example , 3062 ( Pol+NP+HA+NA ) indicates a virus possessing the polymerase complex ( PB1 , PB2 , and PA ) , nucleoprotein ( NP ) , HA , and neuraminidase ( NA ) genes from UT3062 and the rest of its genes from UT3028 ( Figure 2 ) . Reassortant viruses possessing both the UT3062 HA and NS genes , 3062 ( NP+HA+NA+NS ) , 3062 ( Pol+HA+NA+NS ) , 3062 ( HA+NA+NS ) , 3062 ( HA+NS ) , and 3062 ( NP+HA+NS ) , were lethal to animals ( 33%–100% lethality ) . By contrast , those possessing either the HA or NS genes of UT3028 were not lethal to any animals . Mean maximum body weight loss in the animals infected with the former viruses ( 8 . 5%–14 . 4% , median 11 . 4% ) tended to be greater than that in the animals infected with the latter viruses ( 0 . 7%–8 . 6% , median 4 . 1% ) . Animals infected with 3062 ( HA+NS ) resulted in 83% lethality and 17 . 8%±2 . 7% mean maximum weight loss . These results show that the difference in virulence between UT3062 and UT3028 is mainly attributable to both of the HA and NS genes . Sequence comparison between UT3062 and UT3028 revealed only one amino acid difference in their HA proteins and four amino acid differences in their NS proteins , two in NS1 and two in NS2 ( Table 2 ) . Since NS1 and NS2 mRNAs are produced from the same gene segment , with the NS1 mRNA being unspliced and the NS2 mRNA being spliced , the nucleotide alterations can affect both proteins; i . e . , the amino acid differences at positions 200 and 205 of NS1 were coupled to those at positions 47 and 51 of NS2 , respectively . Therefore , it was impossible to substitute the amino acids only in the NS1 or the NS2 protein . Here , we generated two reassortant viruses with a mutation in the NS segment by reverse genetics . 3062 ( HA ) +NS1N200S possesses the UT3062 HA gene , a mutant NS segment encoding the UT3028 NS1 protein , which has an asparagine-to-serine substitution at position 200 ( and encodes NS2 with a threonine-to-alanine substitution at position 47 ) and the rest of its genes from UT3028 . 3062 ( HA ) +NS1G205R possesses the UT3062 HA gene , a mutant NS segment encoding the UT3028 NS1 protein , which has glycine-to-arginine substitution at position 205 ( and encodes NS2 with a methionine-to-isoleucine substitution at position 51 ) and the rest of its genes from UT3028 . We then tested their virulence in ferrets as described above . As shown in Figure 2 , both of the reassortant viruses , 3062 ( HA ) +NS1N200S and 3062 ( HA ) +NS1G205R , were not lethal to any animals , with 4 . 8%±1 . 4% and 2 . 3%±1 . 4 mean maximum weight loss , respectively . These results suggest that all of the amino acids in HA and NS proteins contribute to the virulence in ferrets , although it is unclear whether changes in NS1 , NS2 , or both affect virulence . In addition , 3062 ( HA+NA+NS ) ( 33% lethality and 8 . 5±3 . 4% mean maximum body weight loss ) was attenuated compared to 3062 ( NP+HA+NA+NS ) ( 67% lethality and 14 . 4±4 . 9% mean maximum body weight loss ) . Further , 3062 ( NP+HA+NS ) and 3062 ( HA+NS ) killed 100% and 83% of animals , respectively . These results suggest that the UT3062 NP gene may enhance virus virulence in ferrets . These results are consistent with previous findings that virulence of influenza virus is multigenic [7] , [9] , [10] . To understand the basis for the difference in virulence in ferrets among the viruses , we examined the in vitro and in vivo replication of the parental UT3062 , UT3028 , and the reassortant viruses . For in vitro testing , we compared their growth kinetics in mink lung epithelial ( Mv1Lu ) cells by infecting these cells with viruses at a multiplicity of infection ( MOI ) of 0 . 001 and monitoring the growth kinetics for 48 h . All of the viruses replicated to more than 108 PFU/ml at 36 or 48 h p . i . and the differences in their viral titers were less than one log PFU/ml at each time point ( Figure S1 ) , indicating that there were no substantial differences in their replicative ability in these cells . To examine viral replication in ferrets , we infected animals with 107 PFU of the parental UT3062 , UT3028 , and selected reassortant viruses ( 3062 ( HA+NS ) , 3062 ( NP+HA+NS ) , and 3062 ( HA+NA+NS ) ) . Virus titers in nasal and tracheal swabs , and organs were examined . On days 3 and 7 p . i . , three animals from each infected group were sacrificed for virus titration . As shown in Table 3 , UT3062 , 3062 ( HA+NS ) , 3062 ( NP+HA+NS ) , and 3062 ( HA+NA+NS ) were detected systemically on days 3 and 7 p . i . , whereas UT3028 was detected mainly in the upper respiratory tracts of ferrets on day 3 , but not 7 , p . i . The differences in replicative ability of these viruses in ferrets thus correlate with lethality in this animal model . When we compared the histopathology between ferrets infected with UT3062 and those infected with UT3028 , we found three major differences ( Figures 3 and 4 ) : ( 1 ) host reaction to viral exposure in the lungs on day 1 p . i . , ( 2 ) viral infection in the tracheobronchial lymph node , and ( 3 ) distribution of viral antigens and the inflammatory reaction in the lungs on day 3 and beyond p . i . Firstly , cells infiltrating the lung lesions differed between animals infected with UT3062 and those infected with UT3028 . Although substantial numbers of viral antigen-positive cells were detected in the lungs of ferrets infected UT3062 or UT3028 , the lungs of ferrets infected with UT3062 had marked infiltration of eosinophils around/in the bronchi ( Figure 3A ) . By contrast , the lung lesions of ferrets infected with UT3028 contained many neutrophils ( Figure 3B ) . Secondly , viral infection in the tracheobronchial ( pulmonary regional ) lymph node at 1 day p . i . differed between the two viruses . Although we did not detect viral antigen in the tracheobronchial lymph node of ferrets infected with UT3028 , we did find viral antigen at this site in all three ferrets infected with UT3062 ( Figure 3D and E ) . Thirdly , in animals infected with UT3028 , we did not detect viral antigen beyond 3 days p . i . , with the exception of one ferret , which was euthanized at 5 days p . i . ( Figure 4A ) . The numerous neutrophils observed on 1 day p . i . were replaced by lymphocytes , macrophages and regenerative epithelial cells during the course of infection ( data not shown ) . On the other hand , in animals infected with UT3062 , a substantial number of viral antigen-positive cells were detected in the lungs even 3 days p . i . and the areas in the lungs where the viral antigen-positive cells were detected expanded widely by 5 and 7 days p . i . in some ferrets ( Figure 4B ) . Moreover , when compared to ferret lung lesions with less viral antigen-positive cells , the lesions of ferrets with extensive viral antigen-positive cells had fewer lymphocytes and substantial pulmonary edema , hemorrhaging and fibrinous exudates ( data not shown ) . These findings indicate that there was a tendency for delay in viral clearance in UT3062-infected ferrets and consequently some animals progressed to death . The virus was , however , completely eliminated in some animals , presumably because of individual animal variability . Next , to evaluate the effect of the UT3062 HA and NS genes in vivo , we examined the pathogenicity of 3062 ( HA+NS ) virus , which possesses UT3062 HA and NS genes and its remaining genes from UT3028 . When we examined ferrets infected with 3062 ( HA+NS ) on days 3 and 7 p . i . , we found that they had pathological lesions that more closely resembled those of ferrets infected with UT3062 than those of ferrets infected with UT3028 . Namely , the ferrets had viral infection in the tracheobronchial lymph node and widely distributed viral pneumonia by 3 days p . i . ( Figure 3G to I ) . Pulmonary edema , hemorrhages and fibrinous exudates were obvious in the lung lesions rather than recruitment of lymphocytes and regenerative changes , which were characteristic of ferrets infected with UT3028 . Therefore , the UT3062 HA and NS gene products play a critical role in viral pathogenicity in this ferret model . Viruses first replicated in the lungs ( at the primary site of viral exposure ) , and infection then expanded into the tracheobronchial lymph node . Viral infection in the regional lymph node may negatively affect viral exclusion from the host , leading to continued viral replication . UT3062 , like almost all other human H5N1 viruses , has alanine at position 134 of HA ( H3 numbering ) . UT3028 , however , has threonine at this position ( Table 2 ) . These findings suggest that a single substitution at position 134 ( A134T ) of HA affects virulence in ferrets . Previously , Auewarakul et al . [11] showed that substitutions at positions 129 and 134 ( L129V/A134V ) allow virus recognition of both sialic acid liked to galactose by α2 , 3 linkage ( SAα2 , 3Gal ) and SAα2 , 6Gal , unlike the parent virus , which recognizes only SAα2 , 3Gal . Yamada et al . [12] , however , found that a single substitution at position 134 ( A134T ) did not change receptor-binding preference with the same sialylglycopolymers used by Auewarakul et al . [11] . We , therefore , performed virus elution assays using chicken and horse erythrocytes . From chicken erythrocytes , which express both SAα2 , 3Gal and SAα2 , 6Gal [13] , UT3062 and a reassortant possessing the UT3062 HA were not eluted even after 20h of incubation at 37°C . By contrast , UT3028 and a reassortant possessing the UT3028 HA were gradually released from chicken erythrocytes ( Figure 5 ) . Since the viruses possessing the UT3028 HA were eluted regardless of the origin of the NAs ( either UT3028 or UT3062 ) , this difference in elution from erythrocytes is due to the difference in the amino acid residue at position 134 of HA . When we used horse erythrocytes , all of the viruses were more rapidly eluted from these erythrocytes than from chicken erythrocytes; however , UT3028 and a reassortant possessing the UT3028 HA were eluted more efficiently from horse erythrocytes than were UT3062 and reassortants possessing the UT3062 HA ( data not shown ) . These results suggest that UT3062 HA differs from UT3028 HA in its receptor-binding property . NS1 mediates type I IFN antagonism and affects viral growth in cells . We , therefore , assessed the IFN antagonistic activity of these NS1s by using an IFN bioassay [14] , [15] , [16] , [17] , [18] , [19] , [20] . Briefly , Mv1Lu cells were infected with each virus at an MOI of 1 . 25 and the supernatants were collected at 12–24 h p . i . H5N1 viruses in the supernatants were inactivated with UV and neutralizing antibody ( A1A1 , [21] ) treatment . The supernatants were added to fresh Mv1Lu cells and cultured for 22 h , followed by infection with vesicular stomatitis virus ( VSV ) to determine VSV infectivity of the above-described supernatant-pretreated Mv1Lu cells . As a control , we used a recombinant influenza virus expressing an RNA-binding- and IFN-antagonism-defective NS1 protein within which two basic amino acids were substituted to alanines ( R38A/K41A ) on the UT3062 backbone [17] , [22] . We also generated reassortant viruses possessing the mutant NS1 of UT3028 that has either the N200S or the G205R mutation , on the UT3028 backbone and designated them 3028NS1-N200S and 3028NS1-G205R , respectively . As described above , these viruses also possessed amino acid substitution in NS2 , T47A , or M51I , respectively . At 18 h and 24 h p . i . , the supernatant from Mv1Lu cells infected with UT3028 or a virus possessing the UT3028 NS gene and the remaining genes from UT3062 ( i . e . , 3028 ( NS ) ) inhibited VSV plaque formation more efficiently than did the supernatants from cells infected with UT3062 ( statistically significant difference at P<0 . 05 , Tukey Honestly Significant Difference [HSD] test ) ( Figure 6 ) . Furthermore , the supernatant of cells infected with either 3028NS1-N200S or 3028NS1-G205R inhibited VSV plaque formation more efficiently than did that from viruses possessing the UT3062 NS gene ( i . e . , UT3062 and 3062 ( NS ) ) ( statistically significant difference , P<0 . 05 , Tukey HSD test ) , but less efficiently than that from viruses possessing the UT3028 NS gene ( i . e . , UT3028 and 3028 ( NS ) ) ( Figure 6 ) . These results indicate that both serine at position 200 and arginine at position 205 of NS1 contribute to the enhanced type I IFN antagonistic property of UT3062 NS1 , which , in turn , leads to high virulence in ferrets . To further assess the IFN antagonistic property of NS1 , we investigated the effects of the amino acids in NS1 on the expression of the firefly luciferase reporter gene under the control of an interferon-stimulated response element ( ISRE ) in 293 cells treated with IFNβ . Briefly , 293 cells were transfected with pISRE-Luc , pRL-TK , and pCAGGS NS1 or pCAGGS GFP ( negative control ) . At 24 h post-transfection , the cells were treated with recombinant human IFNβ . At 30 h post-transfection , the cells were lysed and luciferase activities were measured by using the Dual-luciferase Reporter assay system . There were , however , no significant differences in expression from the ISRE between UT3062 NS1 and UT3028 NS1 ( data not shown ) . We then investigated the effects of the amino acids in NS1 on the expression of the firefly luciferase reporter gene under the control of the IFNβ promoter in 293 cells treated with Sendai virus ( SeV ) as described previously [23] . Briefly , 293 cells were transfected with p125-Luc , pRL-TK , and pCAGGS NS1 or pCAGGS GFP ( negative control ) . At 36 h post-transfection , the cells were treated with SeV ( Cantell strain ) . At 48 h post-transfection , the cells were lysed and luciferase activities were measured by using the Dual-luciferase Reporter assay system . The results of this experiment also showed that there were no significant differences in expression from the IFNβ promoter between UT3062 NS1 and UT3028 NS1 ( data not shown ) , indicating that other mechanisms affect the IFN antagonistic property .
Here , using H5N1 viruses isolated from humans , we found that receptor-binding property and NS1 IFN antagonism play important roles in the high virulence of these viruses in ferrets . HA is a receptor-binding and fusion protein and , therefore , is required for virus entry . It is known to play a critical role in virulence [4] , [24] , [25] , [26] , [27] . In this study , we found that viruses possessing threonine at position 134 of HA were appreciably attenuated in ferrets compared to those possessing alanine . Although Yamada et al . [12] did not find differences in the receptor-binding preference between HAs with a single substitution at position 134 ( 134A or 134T ) in a direct binding assay to sialylglycopolymers , we found that this substitution affected receptor-binding property as detected by a virus elution assay ( Figure 5 ) . Since the amino acid at position 134 is located near the receptor-binding pocket but does not directly interact with sialyloligosaccharides [11] , the substitution at this residue may influence the receptor-binding property indirectly . Alanine at position 134 of HA is highly conserved in avian H5N1 viruses—only one virus is known to harbor serine at that position ( the Influenza Sequence Database ( ISD; https://flu . lanl . gov/ , registration system [28] ) ) . Similarly , most human H5N1 viruses also have alanine at this position; however , UT3028 and two other H5N1 viruses that we isolated from humans have threonine at this position ( Y . Sakai-Tagawa and Y . Kawaoka , unpublished ) . Further , eleven H5N1 viruses isolated from humans have valine and one has serine at this position ( ISD; https://flu . lanl . gov/[28] and Y . Sakai-Tagawa and Y . Kawaoka , unpublished ) . These data indicate that an amino acid substitution at position 134 of HA is more frequently observed in human H5N1 viruses than in avian viruses , suggesting that viruses possessing a substitution at position 134 of HA may be selected during replication in humans . Although NS1 is a multifunctional protein , one of its main functions is to suppress type I IFN production [29] . Recent studies revealed that NS1 plays an important role ( s ) in antiviral responses via dsRNA-dependent protein kinase R ( PKR ) and 2′5′-oligoadenylate synthetase/RNase L [30] , [31] . Here , using an IFN bioassay , we showed that both serine at position 200 and arginine at position 205 of NS1 contribute to the enhanced type I IFN antagonistic property of UT3062 that leads to high virulence in ferrets . However , we did not observe significant differences in IFNβ-stimulated expression from the ISRE or in SeV-stimulated expression from the IFNβ promoter in 293 cells between the UT3062 and UT3028 NS1 proteins . It may be that NS1 exhibits type I IFN antagonism by a mechanism other than tested in this study . Alternatively , the difference observed in the highly sensitive IFN bioassay using VSV is not detectable in other IFN assays . The amino acid residues at positions 200 and 205 of NS1 are not well conserved , although the residues at these positions in UT3062 have been observed in other human and avian H5N1 viruses ( Table S3 ) . These findings support the hypothesis that the amino acid residues determined to be important in this study are affected by the genetic background of the test viruses . Nonetheless , the HA amino acid at position 134 and the NS1 amino acids at positions 200 and 205 may now be included as virulence markers for H5N1 viruses .
Our research protocol for the use of ferrets followed the University of Tokyo's Regulations for Animal Care and Use , which was approved by the Animal Experiment Committee of the Institute of Medical Science , the University of Tokyo ( approval number: 19–29 ) . The committee acknowledged and accepted both the legal and ethical responsibility for the animals , as specified in the Fundamental Guidelines for Proper Conduct of Animal Experiment and Related Activities in Academic Research Institutions under the jurisdiction of the Ministry of Education , Culture , Sports , Science and Technology , 2006 . Madin-Darby canine kidney ( MDCK ) cells were maintained in minimal essential medium ( MEM ) with 5% newborn calf serum . Human embryonic kidney 293 and 293T cells were maintained in Dulbecco's modified Eagle's MEM ( DMEM ) with 10% fetal calf serum . Mink lung epithelial ( Mv1Lu ) cells were maintained in MEM with 10% fetal calf serum and 1% non-essential amino acids . All cells were grown at 37°C in 5% CO2 . H5N1 viruses isolated from humans in Vietnam and Indonesia were used in this study ( Table 1 ) . Virus stocks were propagated through two passages in MDCK cells for 24–48 h at 37°C . The cell supernatants were harvested , clarified by centrifugation , aliquoted , and stored at −80°C . The frozen virus stocks were thawed and titrated for virus infectivity in MDCK cells by plaque assays . Virus titers were calculated as PFU/ml . All experiments were performed under biosafety level 3+ conditions . Viral RNA was extracted directly from the supernatants of H5N1 virus-infected MDCK cell cultures with a QIAamp Viral RNA Mini Kit ( Qiagen , http://www1 . qiagen . com/ ) . Complementary DNA was generated by SuperscriptIII ( Invitrogen , http://www . invitrogen . com/ ) with the universal primers for influenza A virus genes . The resulting products were PCR-amplified using PfuUltra High-Fidelity DNA polymerase ( STRATAGENE , http://www . stratagene . com/ ) with specific primers for each virus gene and cloned into a plasmid under the control of the human RNA polymerase I ( PolI ) promoter and the mouse RNA PolI terminator ( PolI plasmids ) . We altered the NS1 gene sequence that encodes the RNA-binding site of UT3062 to create the RNA-binding defective sequence ( R38A/K41A ) as previously described [17] . All reassortant viruses and the parental UT3062 and UT3028 viruses were generated by plasmid-based reverse genetics , as described by Neumann et al . [32] . Briefly , PolI plasmids and protein expression plasmids were mixed with a transfection reagent , TransIT 293T ( Mirus Bio , http://www . mirusbio . com/ ) ; incubated at room temperature for 15 min; and then added to 293T cells . Transfected cells were incubated in OPTI-MEM I ( Invitrogen ) for 48 h . Supernatants containing infectious viruses were harvested and propagated in MDCK cells at 37°C for 48 h . The supernatants were harvested , aliquoted , and stored at −80°C . We used male ferrets , 5–7 months old ( MarshallBioResources , http://www . marshallbioresources . com/ ) in this study . All ferrets were inoculated intranasally with 107 PFU of infectious virus in 500 µl of phosphate-buffered saline ( PBS ) under anesthesia with ketamine ( 25 mg/kg ) and xylazine ( 2 mg/kg ) . Clinical signs , body weights , and body temperatures were recorded daily for 10 days post-infections ( p . i . ) . The percent changes in body weights were calculated by comparing the weights of each ferret at each time point to its initial weight on day 0 . Body temperatures were measured using a rectal thermometer . Changes in body temperature were calculated by comparing the body temperatures of each ferret at each time point to its initial body temperature on day 0 . All animals exhibiting more than 20% weight loss , hemorrhage from any body orifice , or inability to remain upright were euthanized . Surviving ferrets were euthanized under deep anesthesia at 3 weeks p . i . On days 3 and 6 p . i . , nasal washes were collected from anesthetized ferrets and titrated for virus infectivity in MDCK cells by plaque assays . Ferrets infected with the parental UT3062 and UT3028 and selected reassortant viruses ( Table 3 ) were euthanized with deep anesthesia and necropsied on days 3 and 7 p . i . Tissue samples of the brain , olfactory bulb , lungs , hilar lymph node , liver , kidney , spleen , duodenum , and descending colon were collected . A portion of each was stored at −80°C for virus titration and the rest were preserved in 10% neutral buffered-formalin for pathological examination . To prepare 10% tissue emulsions , frozen tissue samples were thawed , weighed , and homogenized in 10 volumes ( w/v ) of sterile PBS using a multi-beads shocker ( Yasui Kikai , http://www . yasuikikai . co . jp/ ) . After centrifugation of the samples at 800× g for 5 min at 4°C , the supernatants were collected . In addition , swabs from nose and trachea were collected and suspended in 1 ml of sterile PBS containing 0 . 3% BSA and penicillin ( 200 U/ml ) . After centrifugation at 800× g for 5 min at 4°C , the supernatants were collected and stored at −80°C . Virus in nasal washes , swabs and tissue samples was titrated for virus infectivity in MDCK cells by plaque assays . Virus titer was expressed as PFU/g for tissue samples and PFU/ml for nasal washes and swabs . The limitations of virus detection were 102 . 0 PFU/g for tissue samples and 101 . 0 PFU/ml for nasal washes and swabs . The formalin-fixed tissues were processed for routine paraffin embedding . The paraffin-embedded tissues were cut into 5 µm thick slices and stained using hematoxylin-and-eosin ( H&E ) . Additional sections were cut for immunohistological staining with rabbit polyclonal antibodies against an H5N1 virus ( A/Vietnam/1203/04 ) . Specific antigen-antibody reactions were visualized by means of 3 , 3′ diaminobenzidine tetrahydrochloride and the Dako EnVision system ( Dako , http://www . dako . jp/ ) . All animal experiments were approved by the Animal Research Committee of The University of Tokyo . The reassortant viruses and the parental UT3062 and UT3028 viruses were inoculated into Mv1Lu cell monolayers at an MOI of 0 . 001 PFU with MEM containing 0 . 3% bovine serum albumin , and incubated at 37°C . Cell supernatants were harvested at a given number of hours p . i . After centrifugation at 1 , 000× g for 5 min , samples were titrated for virus infectivity in Mv1Lu cells by plaque assays . The ability of viruses to be eluted from erythrocytes was assessed as previously described [33] , [34] , with some modifications . Briefly , virus stocks were diluted serially in calcium saline ( 0 . 9 mM CaCl2-154 mM NaCl in 20 mM borate buffer , pH 7 . 2 ) and 50 µl aliquots were incubated with 50 µl of 0 . 55% chicken erythrocytes at 4°C for 1 h in microtiter plates . The plates were then transferred to 37°C and monitored periodically for 20 h . Levels of IFN secreted by virus-infected Mv1Lu cells were assessed as previously described [14] , [15] , [16] , [17] , [18] , [19] , [20] , with some modifications . Briefly , Mv1Lu cells were infected with each virus at an MOI of 1 . 25 . Supernatants from infected cells were harvested 12–24 h p . i . To inactivate viral infectivity , the supernatants were treated with UV light for 20 min and then mixed with neutralizing α-VN1203HA monoclonal antibodies ( A1A1 , [21] ) . The supernatants were added to fresh Mv1Lu cells and incubated for 22 h . These pretreated Mv1Lu cells were then infected with VSV , and the VSV infectivity titers were determined . As a control , we used a recombinant influenza virus expressing an RNA-binding- and IFN antagonism-defective NS1 protein within which two basic amino acids were substituted to alanines ( R38A/K41A ) . The experiments were carried out in triplicate and were independently repeated twice . 8×104 of 293 cells were transfected with 50 ng of pISRE-Luc ( Clontech , http://www . clontech . com/ ) and 50 ng of pRL-TK ( Promega , http://www . promega . com/ ) by using TransIT293 ( Mirus , http://www . mirusbio . com/ ) . 100 ng of pCAGGS NS1 or pCAGGS GFP were co-transfected . Cells were incubated for 24 h and were treated with 100 units of recombinant human IFNβ , 1a ( PBL interferonSource , http://www . interferonsource . com/ ) . At 30 h post-transfection , cells were lysed and luciferase activities were measured by using the Dual-luciferase Reporter assay system ( Promega , http://www . promega . com/ ) according to the protocol provided by the manufacturer . Firefly luciferase values were divided by Renilla luciferase values to normalize for transfection efficiency . 2×105 of 293 cells were transfected with 250 ng of p125-Luc ( the reporter plasmid carrying the firefly luciferase gene under the control of the IFNβ promoter was kindly provided by Takashi Fujita ) and 250 ng of pRL-TK ( Promega , http://www . promega . com/ ) by using TransIT293 ( Mirus , http://www . mirusbio . com/ ) . 5–500 ng of pCAGGS NS1 or pCAGGS GFP were co-transfected . After incubation for 24 h , cells were stripped with trypsin and divided into two wells of a 12-well plate . At 36 h post-transfection , cells were treated with approximately 5×1010 focus-forming unit of SeV ( Cantell strain ) . At 48 h post-transfection , cells were lysed and luciferase activities were measured by using the Dual-luciferase Reporter assay system ( Promega , http://www . promega . com/ ) according to the protocol provided by the manufacturer . Firefly luciferase values were divided by Renilla luciferase values to normalize for transfection efficiency . | Highly pathogenic H5N1 influenza A viruses have caused more than 500 human infections with approximately 60% lethality in 15 countries and continue to pose a pandemic threat . The recent worldwide spread of pandemic H1N1 influenza A viruses raises the concern of reassortment between the H5N1 viruses and other influenza viruses . However , the molecular determinants for high virulence of the H5N1 viruses in mammals are not fully understood . We , therefore , investigated their virulence in a ferret model , which is a widely accepted animal model for assessing human influenza virus replication . We identified an amino acid in hemagglutinin and four amino acids in nonstructural proteins that are associated with high virulence of a human H5N1 virus , A/Vietnam/UT3062/04 . We also found that the amino acid in hemagglutinin changes its receptor-binding property and the amino acids in nonstructural protein 1 affect its interferon antagonistic ability . These findings provide insight into the pathogenesis of H5N1 viruses in mammals . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"virology/virulence",
"factors",
"and",
"mechanisms",
"infectious",
"diseases",
"microbiology/immunity",
"to",
"infections",
"respiratory",
"medicine/respiratory",
"infections",
"pathology",
"virology",
"virology/animal",
"models",
"of",
"infection",
"infectious",
"diseases/vir... | 2010 | The HA and NS Genes of Human H5N1 Influenza A Virus Contribute to High Virulence in Ferrets |
Plasmodium spp parasites harbor an unusual plastid organelle called the apicoplast . Due to its prokaryotic origin and essential function , the apicoplast is a key target for development of new anti-malarials . Over 500 proteins are predicted to localize to this organelle and several prokaryotic biochemical pathways have been annotated , yet the essential role of the apicoplast during human infection remains a mystery . Previous work showed that treatment with fosmidomycin , an inhibitor of non-mevalonate isoprenoid precursor biosynthesis in the apicoplast , inhibits the growth of blood-stage P . falciparum . Herein , we demonstrate that fosmidomycin inhibition can be chemically rescued by supplementation with isopentenyl pyrophosphate ( IPP ) , the pathway product . Surprisingly , IPP supplementation also completely reverses death following treatment with antibiotics that cause loss of the apicoplast . We show that antibiotic-treated parasites rescued with IPP over multiple cycles specifically lose their apicoplast genome and fail to process or localize organelle proteins , rendering them functionally apicoplast-minus . Despite the loss of this essential organelle , these apicoplast-minus auxotrophs can be grown indefinitely in asexual blood stage culture but are entirely dependent on exogenous IPP for survival . These findings indicate that isoprenoid precursor biosynthesis is the only essential function of the apicoplast during blood-stage growth . Moreover , apicoplast-minus P . falciparum strains will be a powerful tool for further investigation of apicoplast biology as well as drug and vaccine development .
The discovery of a plastid organelle , the apicoplast , in Plasmodium spp . ( responsible for 250 million cases of human malaria each year ) and other Apicomplexa parasites instantly made it a key target in the development of new therapies against these pathogens [1]–[3] . The need for new anti-malarials is particularly urgent given the documentation of developing resistance to the current last-line therapy in the deadliest species , P . falciparum [4] . Several features of this organelle make it both biologically fascinating and an attractive therapeutic target . The apicoplast is derived from secondary endosymbiosis of a plastid-bearing red algae and is therefore prokaryotic in origin , containing pathways that have no counterpart in the human host [3] , [5] , [6] . During the course of evolution , the apicoplast has lost its photosynthetic function and transferred most of its genome to the nucleus , requiring a dedicated protein targeting pathway to localize the majority of its over 500 gene products into the organelle [7] , [8] . The remaining 35 kb apicoplast genome encodes ∼50 mostly housekeeping genes , including ribosome subunits , tRNAs , RNA polymerase , and a chaperone thought to be involved in protein import [1] . Despite its minimalization , the apicoplast continues to serve essential though poorly defined metabolic function ( s ) . In Plasmodium , apicoplast function is necessary for both intraerythrocytic and intrahepatic development in the human host [9] , [10] . Whether the apicoplast is required for sexual stage development in the mosquito is currently unknown [11]–[14] . The essential role of the apicoplast has been clearly demonstrated by the effect of antibiotics on blood-stage P . falciparum . Dahl et al . showed that antibiotics that inhibit prokaryotic transcription or translation , such as doxycycline , specifically blocks expression of the apicoplast genome [9] , [15] . Parasites treated during first 48 h life cycle show no obvious defect from the loss of apicoplast-encoded gene products: Organelle morphology , genome replication , protein targeting , and segregation during cell division remain intact . Likewise , parasites progress normally through the ring , trophozoite , and schizont developmental stages , giving rise to daughter merozoites that successfully reinvade to establish infection of a new host cell . Instead , a curious “delayed death” phenotype is observed , whereby the deleterious effects of antibiotic inhibition accumulate in the progeny of treated parasites . In this 2nd life cycle following antibiotic treatment , the apicoplast genome fails to replicate . Protein import function is lost . Finally , the organelle itself fails to replicate and segregate during cell division . Because the apicoplast cannot be generated de novo and must be inherited at each cell division , the failure of organelle replication and segregation in these parasites results in loss of the apicoplast in daughter cells and parasite death . Overall , antibiotic-induced delayed death begins with specific inhibition of apicoplast transcription and translation in one life cycle and ends with irreversible apicoplast loss and death in the subsequent cycle . A similar delayed death has been observed in Toxoplasma gondii following antibiotic treatment or transgene expression that cause apicoplast loss [16] , [17] . Despite its promise as Plasmodium's “Achilles heel , ” the function of the apicoplast has eluded researchers in the nearly 20 years since its discovery . Without knowledge of specific proteins or pathways suitable as drug targets , particularly during the clinically symptomatic blood stage , efforts to develop apicoplast-directed therapies ( beyond known antibiotics ) have been stymied . An astounding 5%–10% of the nuclear genome is predicted to contain an apicoplast targeting signal , yet 70% of these apicoplast gene products are of unknown function [18]–[20] . Pathways that have been identified include those for the biosynthesis of isoprenoid precursors , fatty acids , heme , Fe-S clusters , and lipoic acid [21] . While in silico analysis has been revealing , many pathways will go undetected and the essentiality of predicted pathways throughout the parasite's complex life cycle needs to be experimentally validated . For example , inhibition by the antibiotic triclosan initially suggested that apicoplast-located type II fatty acid biosynthesis was essential in blood-stage parasites , prompting the development of fatty acid inhibitors as anti-malarials [22] . Later , genetic deletion of fatty acid biosynthetic genes definitively proved that this pathway is not required for blood stage growth and instead is critical for liver stage development [23] , [24] . Unfortunately , discovery and validation of apicoplast pathways has been hampered by the limited ability to generate knockouts of essential genes , isolate the organelle , or purify Plasmodial proteins for in vitro characterization [25]–[28] . Amongst the annotated apicoplast pathways , Plasmodium relies on the prokaryotic MEP/DOXP/non-mevalonate pathway for synthesizing isoprenoid precursors rather than the canonical mevalonate pathway used by most other eukaryotes and all mammals ( including the human host ) [29] , [30] . Both pathways produce isopentenyl pyrophosphate ( IPP ) and dimethylallyl pyrophosphate ( DMAPP ) as the final products , but the enzymes and chemical intermediates leading to synthesis of these compounds are entirely different . Fosmidomycin , an inhibitor of the MEP pathway , kills blood-stage parasites and has been tested in clinical trials as an antimalarial [31] , [32] . Inhibition by fosmidomycin suggests that isoprenoid precursor biosynthesis is essential in blood-stage infection , although the possibility of off-pathway targets as the cause of the drug effect ( as was found to be the case for triclosan ) has not been ruled out [29] , [33] . Furthermore , IPP and DMAPP are not an end onto themselves but rather building blocks used to synthesize small molecule isoprenoids with a host of functions or C15/C20 prenyl chains for the post-translational modification of proteins [34] , [35] . Once IPP and DMAPP are exported into the parasite cytoplasm , the downstream isoprenoid products in Plasmodium and their function during infection are unknown . The significance of isoprenoid precursor biosynthesis as a drug target and gateway for identifying isoprenoid products with essential functions in pathogenesis depends on a clear demonstration of its role in parasite survival . In addition to its essentiality , this pathway may represent the only direct output from the apicoplast into the cytoplasm during blood stage growth since the remaining annotated pathways function primarily for organelle maintenance , support the mitochondria , or are not essential in this stage . Given the difficulty of studying the apicoplast by traditional methods , we employed an alternative strategy using drug inhibition/chemical rescue , equivalent to genetic deletion/complementation , to establish pathway essentiality and sufficiency . Using this simple chemical genetic approach , we show that isoprenoid precursor biosynthesis is not only essential but in fact the sole essential function of the apicoplast during blood-stage growth .
To investigate the specificity of fosmidomycin for the isoprenoid precursor biosynthetic pathway , we observed the effect of supplementation with isoprenoid precursors , IPP and DMAPP , on drug inhibition of blood-stage parasites . Growth inhibition of blood-stage P . falciparum W2 by fosmidomycin occurred with an EC50 = 0 . 98 µM ( 95% confidence interval = 0 . 93–1 . 03 µM; Figure 1A ) . When drug susceptibility was performed in media supplemented with 200 µM IPP , DMAPP , or both IPP and DMAPP , only IPP ( without DMAPP ) was sufficient to completely reverse the growth inhibition in the presence of up to 100 µM fosmidomycin ( Figure 1A ) . Survival of parasites was dependent on the concentration of IPP in the media with rescue apparent at 200 µM IPP ( Figure 1B ) . DMAPP alone or in combination with IPP had no effect or was even slightly toxic ( Figure 1A ) . Addition of up to 2 mM 3-methyl-3-butenol , the alcohol analog of IPP lacking the pyrophosphate moiety , alone or in combination with 3-methyl-2-butenol , the alcohol analog of DMAPP , also did not rescue fosmidomycin inhibition ( Figure S1 ) . Finally , reversal of drug inhibition by addition of IPP was only seen with fosmidomycin and did not occur with chloroquine , a drug that does not target isoprenoid precursor biosynthesis ( Figure S2 ) . These findings establish that ( 1 ) fosmidomycin inhibition is specific for the isoprenoid precursor biosynthetic pathway , ( 2 ) isoprenoid precursor biosynthesis is essential for blood-stage P . falciparum , and ( 3 ) exogenous IPP fulfills endogenous biosynthetic function . To determine whether rescue of the isoprenoid precursor biosynthesis pathway can reverse delayed death due to antibiotics , P . falciparum W2 parasites were treated with chloramphenicol , clindamycin , or doxycycline in the presence of IPP . As described above , treatment of blood-stage parasites with these prokaryotic transcription and translation inhibitors specifically blocks apicoplast gene expression in the first 48 h cycle leading to apicoplast loss and parasite death in the second cycle [9] , [15] . As such , antibiotic treatment yielded a 48 h EC50 in growth assays due to nonspecific inhibition and a lower 96 h EC50 due to apicoplast-specific inhibition ( Table 1 ) . This shift in the EC50 values from 48 to 96 h is characteristic of the delayed death phenotype . By supplementing with 200 µM IPP , reversal of apicoplast-specific inhibition at 96 h by all drugs was observed with EC50 values reflective of only nonspecific drug effects . In fact , when addition of IPP was compared from 48–96 h or 0–96 h , IPP was only required during the second cycle consistent with a deficiency in the progeny of treated parasites ( Table 1 ) . Similar results were demonstrated with P . falciparum D10 strain , suggesting that the rescue of antibiotic inhibition by IPP is not strain-specific ( Table 1 ) . Blood-stage parasites were carried through several life cycles with simultaneous antibiotic treatment and IPP rescue to determine ( 1 ) any significant growth defects and ( 2 ) the dependence on further supplementation with IPP ( after removal of the antibiotic ) of the surviving parasites . As shown in Figure 2A , the doxycycline-treated , IPP-rescued strain showed parasitemia ≥65% of that seen in the untreated strain throughout the treatment/rescue course . ( Given the narrow concentration range between apicoplast-specific and nonspecific inhibition for doxycycline , some decreased growth due to non-specific inhibitory effects is expected at the drug concentration used . ) Moreover , the rescued strain was carried for a total of 26 days with IPP supplementation but in the absence of antibiotics with no diminishment in growth capacity . In contrast , doxycycline-treated parasites without added IPP quickly died after the 2nd cycle of treatment with undetectable parasitemia by the end of the 3rd cycle ( Figure 2A ) . Significantly , removal of IPP and doxycycline from the media at the start of the 4th cycle results in a rapid decline in parasitemia of the rescued parasites ( Figure 2A ) . With further passage in media lacking both IPP and doxycycline , the parasitemia of rescued parasites became undetectable . Again similar results could be demonstrated with P . falciparum D10 strain or treatment with chloramphenicol ( Figures S3 and S4 ) . These results show that antibiotic-treated parasites rescued by IPP supplementation have no gross growth defect but are entirely dependent on exogenous IPP for continued growth . We cannot rule out more subtle growth defects that would be difficult to assess by comparison of parallel cultures . IPP rescue of death following antibiotic treatment could be due to either ( 1 ) blocking the deleterious effects of antibiotic treatment to cause apicoplast loss or ( 2 ) compensating for the loss of the apicoplast . The irreversible dependence of the rescued parasites on IPP ( even after removal of the antibiotic ) suggests the latter—that apicoplast loss occurs but is chemically complemented by exogenous IPP . We sought to confirm whether the sequelae of apicoplast dysfunction that occurs following treatment with antibiotics ( reviewed above ) also bears out in IPP-rescued parasites [9] . A hallmark of organelle dysfunction is the loss of the apicoplast genome [9] , [16] . We used quantitative PCR for target genes on the apicoplast , mitochondria , and nuclear genomes to monitor the ratio of organelle: nuclear genomes during the course of antibiotic treatment and chemical rescue . Figure 2B demonstrates a marked decline in the apicoplast∶nuclear genome ratio after the 2nd cycle in all antibiotic-treated parasites regardless of supplementation with IPP . At the end of the 4th cycle , the ratio is reduced by at least 100-fold . In contrast , no such decline is noted in the mitochondria∶nuclear genome ratio ( Figure 2C ) . Rescued strains carried out for a total of 26 days continued to show undetectable apicoplast genome and detectable mitochondria and nuclear genomes . Thus , IPP-rescued parasites undergo a specific and irreversible loss of the apicoplast genome without concomitant loss of the nuclear or mitochondrial genomes , yet these parasites continue to be viable . A critical function of the apicoplast , required for the maintenance of its proteome , is the import of nuclear-encoded proteins into the organelle . A bipartite N-terminal sequence consisting of a signal sequence and a transit peptide is required to target proteins to the organelle [8] . Upon import into the apicoplast , the transit peptide is cleaved to produce a mature protein [8] . Protein processing is therefore a marker of successful protein import into the apicoplast . We used a transgenic D10 strain expressing GFP fused to an N-terminal apicoplast targeting sequence ( ACPL-GFP ) to assess apicoplast protein processing during the course of antibiotic treatment and IPP rescue [8] . The 33 kDa full-length GFP was cleaved to produce a predominant 30 kDa mature protein in untreated parasites ( Figure 3A ) . Parasites treated with doxycycline only began to lose protein processing function during the 2nd cycle as seen in the increased accumulation of full-length protein at 96 h but do not survive beyond this cycle ( Figure 3B ) . When doxycycline treatment was rescued with IPP , surviving parasites showed successive loss of protein processing with each treatment cycle such that only preprocessed GFP was detectable at 144 h , the start of the 4th cycle ( Figure 3C ) . A smaller , previously described degradation band also became apparent in the rescued parasites [8] . The absence of protein processing activity indicates a loss of the critical protein import function of the apicoplast in these rescued parasites . The final outcome of antibiotic treatment is a failure of apicoplast replication and segregation during cell division , resulting in loss of the organelle and death [9] . The loss of the genome and protein import function strongly suggests that parasites that survive antibiotic treatment are in fact apicoplast-minus . Localization of GFP in the D10 ACPL-GFP strain was used to visualize the apicoplast . As expected , GFP localizes to a discrete structure in the parasite in untreated cells ( Figure 4 and Video S1 ) . In contrast , in parasites that have been rescued from antibiotic death , GFP loses this discrete apicoplast localization and becomes diffuse ( Figure 4 ) . Confocal images show that numerous foci of GFP are scattered throughout the cytoplasm ( Figure 4 and Video S2 ) . The largest foci measure >200 nm , so these collections of GFP are less likely to be cytoplasmic protein aggregates but instead may represent vesicles containing protein . Combined with the absence of the apicoplast genome and protein import function , the loss of GFP localization indicates the absence of the apicoplast itself to which it is normally targeted .
Until now , the mystery of apicoplast function has been a critical gap in our understanding of malaria pathogenesis . Our findings demonstrate that the production of isoprenoid precursors is the only essential function of the apicoplast during asexual blood-stage P . falciparum growth ( Figure 5 ) . This surprising revelation has several important implications and invites a host of new questions . Because isoprenoid precursors are building blocks to synthesize cellular isoprenoid products with diverse functions , their key role now gives added urgency to the elucidation of these products and their downstream functions . At least one essential prenylated product is ubiquinone , a component of the mitochondrial electron transport chain . There are certainly other essential , as-yet-unidentified isoprenoid products since transgenic parasites which express yeast dihydroorotic acid dehydrogenase and no longer require their electron transport chain are still susceptible to fosmidomycin and antibiotics and could be rescued with IPP supplementation ( unpublished data ) [36] . Possible isoprenoid products include dolichols involved in protein N-glycosylation which have been detected in Plasmodium and prenylated proteins , such as Rab homologs required for vesicular trafficking and a recently identified tyrosine phosphatase [37]–[41] . The current findings also imply that several annotated apicoplast pathways are in fact non-essential . Amongst both identified pathways and the 70% of apicoplast gene products with unknown function , only isoprenoid precursor biosynthesis and any pathways supporting this function in blood-stage parasites ( including those required for organelle maintenance and replication ) are essential and therefore viable apicoplast drug targets [20] . Assertions that type II fatty acid and , by implication , acetyl-CoA biosynthesis were essential apicoplast functions during blood-stage growth have already been disproven [23] , [24] , [42] . A parasite-derived pathway for heme biosynthesis contains steps that occur in the apicoplast , mitochondria , and cytosol . Our results strongly imply that blood-stage parasites do not depend on de novo heme biosynthesis using this pathway but instead rely on an extrinsic de novo pathway utilizing imported host enzymes or salvage of heme from the host by an unidentified mechanism [43] , [44] . Still other pathways such as Fe-S cluster biosynthesis supply cofactors for enzymes within the organelle but are not exported outside the organelle . These pathways become “non-essential” when the need for organelle maintenance is removed . The complexity of the organelle and the simplicity of its blood-stage function pose an obvious contradiction . Approximately 5%–10% of the Plasmodium genome is predicted to encode apicoplast-targeted gene products ( although the localization and/or function of the majority of these gene products have not been validated ) [20] . In order to import these proteins into the apicoplast , parasites utilize a dedicated protein trafficking pathway [7] , [8] . In addition , the organelle undergoes complex morphological development during blood stage growth , requiring cellular machinery to faithfully replicate and segregate the organelle at every cell division [10] . Why are such huge resources consumed to maintain a single essential function ? First , while the function of the apicoplast is limited during the blood stage , the need for more extensive organelle function during other developmental stages may dictate its maintenance in intraerythrocytic parasites as the organelle cannot be generated de novo . Fatty acid biosynthesis , for example , is an essential apicoplast function in liver stage parasites [23] , [24] . Second , Plasmodium may have been evolutionarily trapped in its bondage to the apicoplast . Having acquired the plastid early in its evolution , it may have been unable to dispense of it even after adopting an increasingly parasitic lifestyle due to the transfer of even a few essential functions to the organelle . In any case , this imbalance emphasizes the value of targeting housekeeping pathways involved in organelle maintenance and replication to interfere with its function . An important consideration is whether our findings accurately reflect in vivo growth requirements of parasites during infection . Specifically , are there essential metabolites supplemented in culture which could not be acquired during in vivo growth and instead must be biosynthesized by the apicoplast ? While parasitized RBCs during infection use human plasma as a source of extracellular nutrients , our cultures were grown in RPMI medium 1640 supplemented with purified serum substitute , Albumax . We found that Albumax could be replaced with 10% human serum with no effect on the survival of apicoplast-minus parasites in the presence of IPP ( Figure S5 ) . RPMI medium contains salts , 20 amino acids , 11 vitamins , 4 other organic molecules , and glucose . The acquisition and biosynthesis of these nutrients by blood-stage Plasmodium and their essentiality for intraerythrocytic growth based on available evidence is shown in Table S1 . In general , blood-stage Plasmodium biosynthesizes very limited amounts of just 3 amino acids and is dependent on amino acids from either ( 1 ) hemoglobin degradation or ( 2 ) acquisition from patient plasma through newly established permeation pathways in the infected red cell [45]–[47] . Similarly , current knowledge of Plasmodium metabolism also suggests that the remaining organic metabolites found in RPMI medium are biosynthesized by non-apicoplast pathways or can be acquired from the host red cell or patient plasma [46] , [48]–[50] . Consequently , we believe that our findings can be extrapolated to in vivo requirements for the apicoplast to support parasite growth and development . At the very least , our results define a very minimal set of potential metabolites ( IPP and components found in RPMI 1640 medium ) that could be biosynthesized in the apicoplast . We cannot , however , rule out additional apicoplast functions ( other than those required for growth ) that would not be revealed in our blood culture system , such as functions required for immune evasion . Several aspects of the chemical rescue with isoprenoid precursors are notable . During chemical rescue , exogenous IPP could enter the parasite through permeation pathways established in the parasitized erythrocyte or other uncharacterized membrane transporters [46] , [51] . The RBC is largely metabolically inactive and unlikely to have significant ongoing isoprenoid precursor biosynthesis via the host mevalonate pathway or stores of these metabolites [52] . It is also unlikely that these high-energy pyrophosphorylated molecules would accumulate to appreciable levels in plasma ( 200 µM was required for rescue in our experiments ) . Consistent with this notion , IPP was not present in the Serum Metabolome Database ( SMDB ) , which contains 4 , 229 detectable metabolites [53] . Therefore , acquisition of isoprenoid precursors in vivo by salvage of IPP from infected blood is improbable . Once in the parasite , exogenous IPP may fulfill its function in the cytoplasm with or without uptake into the apicoplast [54] . Although both IPP and DMAPP are required to synthesize isoprenoid products , supplementation with IPP alone is sufficient to fulfill endogenonous isoprenoid precursor biosynthesis , implying the presence of an IPP isomerase in the cytoplasm that converts IPP to DMAPP . This activity may be important in establishing the optimal cellular ratio of IPP to DMAPP , as toxicity was noted with increasing DMAPP concentrations in our experiments . A putative IPP isomerase has been identified in the Plasmodium genome [55] . A recent report suggested that geranylgeraniol , the alcohol analog of a C20 prenyl chain , could rescue fosmidomycin inhibition [56] . We were unable to rescue fosmidomycin inhibition with alcohol analogs of IPP and DMAPP , indicating either poor cellular penetration of the alcohols or the absence of a kinase to convert the alcohol analogs to the pyrophosphorylated and active metabolites ( Figure S1 ) . Even with conversion of geranylgeraniol to geranylgeranyl pyrophosphate in the cell , it would seem that a C5 building block , such as IPP , would almost certainly be required to extend the supplemented C20 unit for construction of polyprenyl chains , such as that found in ubiquinone , and to construct smaller prenyl chains , such as for protein farnesylation . The reported rescue with geranylgeraniol was performed at 1 . 5 µM fosmidomycin , which is above the concentration required for 50% growth inhibition but may be below that required for adequate inhibition of the biosynthetic pathway ( since phenotypic growth inhibition can be apparent even at low levels of inhibition of the biosynthetic pathway ) [56] . Therefore , the reported results may be complicated by ongoing biosynthesis of IPP and DMAPP contributing to the precursor pool . Consistent with this , neither farnesol nor geranylgeraniol was able to rescue fosmidomycin concentrations >10 µM , and both showed dose-related parasite toxicity ( Figure S6 ) . In contrast , we were able to demonstrate IPP rescue at fosmidomycin concentrations exceeding 100 µM , well above its EC90 for growth inhibition . The consequences of apicoplast loss following antibiotic treatment and IPP rescue are no less intriguing . In the parasites that survive antibiotic treatment by chemical rescue , the organelle is irreversibly lost when it fails to segregate to daughter cells [9] . In these apicoplast-minus parasites , apicoplast gene products encoded in the nucleus may continue to be transcribed and translated . These products may properly shuttle into the secretory pathway but cannot be diverted to the organelle [8] . Based on the microscopy results , we hypothesize that proteins may be packaged into transport vesicles bound for the organelle but are unable to localize to the missing structure and therefore accumulate in the cytoplasm appearing as numerous foci . While we cannot rule out the presence of structural remnants of the apicoplast , the observed foci are unlikely to support apicoplast functions . Apicoplast-targeted proteins may require both cleavage of the long basic transit peptide and chaperones in the lumen of the apicoplast for proper folding . We observed that cleavage of the transit peptide from targeted proteins , a critical apicoplast function , does not occur in rescued parasites ( Figure 3C ) . The close physical and functional relationship between the apicoplast and the mitochondria raises the possibility that loss of the apicoplast might affect the ability of the mitochondria to replicate and divide . We were able to detect the mitochondrial genome by qPCR for the cytB3 gene and observe labeling of the mitochondria with Mitotracker by fluorescence microscopy in apicoplast-minus parasites ( Figure 2C; unpublished data ) . Despite the loss of the apicoplast , these parasites do appear to contain mitochondria . While the survival of apicoplast-minus P . falciparum invokes a slew of intriguing questions , these same parasites will be a powerful and indispensable tool for further dissection of apicoplast biology . Apicoplast-minus P . falciparum strains generated in this study can be used to assess organelle requirement during gametocytogenesis and mosquito stage development . These strains also provide novel avenues to identify isoprenoid products , generate conditional mutants of essential genes involved in apicoplast maintenance and replication , conduct metabolomic or proteomic profiling , and study protein trafficking to the organelle . With regard to drug development , our chemical rescue strategy also addresses the critical deficiency of current cell growth screening assays , namely lack of knowledge of the drug target . Candidate drug hits detected in phenotypic assays can be screened for chemical rescue of the growth inhibition . The reversal of growth inhibition by IPP supplementation specifically identifies inhibitors that target pathways involved in MEP pathway function , replication , or maintenance of the apicoplast , providing a pathway-specific drug screen to aid in discovery of new classes of anti-malarials . The ability to chemically complement the cell death phenotype will prevent false leads from off-target effects , like that seen with triclosan and its misconstrued effect on type II fatty acid biosynthesis [22] . Finally , the apicoplast-minus strains dependent on IPP for continued growth are a unique and ideal candidate for an attenuated blood-stage vaccine [57] , [58] . Unlike irradiated or drug-treated whole parasite vaccines , apicoplast-minus parasites would continue to develop in blood at most one cycle , including a single erythrocyte rupture and reinvasion , thereby stimulating a stronger immune response . However , judging by the effects of IPP withdrawal in culture , they would be unable to develop further in the absence of exogenous IPP ( Figure 2A ) . Lending support to this notion , a similar “limited survival” strategy targeting the apicoplast in liver-stage parasites has proven effective as a liver-stage vaccine candidate [59] . A significant advantage of our approach is that attenuation is achieved chemically and does not require difficult or costly genetic manipulation ( as is the case with genetically modified vaccine strains ) , thereby allowing for the possibility of incorporating circulating field strains of Plasmodium in a vaccine formulation [60] . There would also be very little risk of reversion as it would be extremely difficult to reacquire apicoplast function by mutation . In summary , we believe that the current study ushers in a new era in the investigation of the apicoplast in Plasmodium with exciting opportunities to counteract the malarial scourge on human lives .
Plasmodium falciparum W2 ( MRA-157 ) , D10 ( MRA-201 ) , and D10 ACPL-GFP ( MRA-568 ) were obtained from MR4 . Parasites were grown in human erythrocytes ( 2% hematocrit ) in RPMI 1640 media supplemented with 0 . 25% Albumax II ( GIBCO Life Technologies ) , 2 g/L sodium bicarbonate , 0 . 1 mM hypoxanthine , 25 mM HEPES ( pH 7 . 4 ) , and 50 µg/L gentamycin , at 37°C , 5% O2 , and 6% CO2 . For D10 ACPL-GFP , the media was also supplemented with 100 nM pyrimethamine ( Sigma ) . For passage of antibiotic-treated , IPP-rescued parasites , the media was supplemented with 1 . 5–2 µM doxycycline or 50 µM chloramphenicol . 48 h after initiation of antibiotic treatment , rescued strains were supplemented with 200 µM IPP ( Isoprenoids LC ) for continuous passage . For comparison of growth between different treatment conditions , cultures were carried simultaneously and handled identically with respect to media changes and addition of blood cells . Growth assays were performed in 96-well plates containing serial dilution of drugs in duplicate or triplicate . Media was supplemented with IPP or DMAPP as indicated . To determine the EC50 of fosmidomycin ( Invitrogen ) , growth was initiated with ring-stage parasites ( synchronized with 2 . 5% sorbitol treatment 48 h prior ) at 1% parasitemia ( 0 . 5%–2% hematocrit ) . Plates were incubated for 72 h . To determine the EC50 of antibiotics at 48 h , growth was initiated at 1% parasitemia and incubated for 48 h . To determine the EC50 of antibiotics at 96 h and observe the delayed death phenotype , cultures were initiated at 0 . 2% parasitemia , 75% of the media was exchanged at 48 h , and plates were incubated for an additional 48 h ( total 96 h ) . For all assays , growth was terminated by fixation with 1% formaldehyde and parasitized cells were stained with 50 nM YOYO-1 ( Invitrogen ) . Parasitemia was determined by flow cytometry . Data were analyzed by FlowJo , and EC50 curves plotted by GraphPad Prism . Parasites from 200 µL of culture were isolated by saponin lysis followed by PBS wash to remove extracellular DNA . DNA was purified using QiaAMP blood kits ( Qiagen ) . Primers were designed to target genes found on each organelle or nuclear genome: tufA ( apicoplast ) 5′-GATATTGATTCAGCTCCAGAAGAAA-3′ / 5′-ATATCCATTTGTGTGGCTCCTATAA-3′ , cytb3 ( mitochondria ) 5′-AGATACATGCACGCAACAGG-3′ / 5′-TCATTTGACCCCATGGTAAGA-3′ , and CHT1 ( nuclear ) 5′-TGTTTCCTTCAACCCCTTTT-3′ / 5′-TGTTTCCTTCAACCCCTTTT-3′ . Reactions contained template DNA , 0 . 15 µM of each primer , and 0 . 75× LightCycler 480 SYBR Green I Master mix ( Roche ) . PCR reactions were performed at 56°C primer annealing and 65°C template extension for 35 cycles on a Lightcycler 6500 ( Roche ) . Relative quantification of target genes was determined using the method of Pfaffl [61] . For each time point , the organelle∶nuclear genome ratio of the antibiotic-treated control or antibiotic-treated , IPP-rescued experiment was calculated relative to that of an untreated control collected at the same time . Ring-stage D10 ACPL-GFP parasites from 1 mL of culture were isolated by saponin lysis , washed with PBS , and resuspended in 1×NuPAGE LDS sample buffer ( Invitrogen ) . Proteins were separated by electrophoresis on 12% Bis-Tris gel ( Invitrogen ) and transferred to nitrocellulose membrane . After blocking , membranes were probed with 1∶1 , 000 polyclonal rabbit anti-GFP ( Clontech ) antibody and 1∶15 , 000 Alexa Fluor 810-conjugated anti-rabbit IgG secondary antibody ( Invitrogen ) . Fluorescent antibody-bound proteins were detected with Odyssey Imager at 800 nm ( LiCor Biosciences ) . Untreated and antibiotic treated/IPP rescued D10 ACPL-GFP parasites were incubated in 2 µg/mL Hoescht 33342 stain for 30 min at 37°C . Cells in culture media were settled onto 35 mm glass-bottom petri dishes ( MakTek ) coated with 1% polyethylenimine ( Sigma ) . Widefield epifluorescence live cell images were obtained on a Nikon Eclipse Ti-E equipped with a Coolsnap HQ2 camera ( Photometrics ) using a 100×/1 . 4 oil immersion objective . Confocal live cell images were obtained on an A1 confocal mounted on a Nikon Eclipse Ti-E using a 60×/1 . 4 oil immersion objective . Images were analyzed by NIS-Elements software ( Nikon ) . | Malaria caused by Plasmodium spp parasites is a profound human health problem that has shaped our evolutionary past and continues to influence modern day with a disease burden that disproportionately affects the world's poorest and youngest . New anti-malarials are desperately needed in the face of existing or emerging drug resistance to available therapies , while an effective vaccine remains elusive . A plastid organelle , the apicoplast , has been hailed as Plasmodium's “Achilles' heel” because it contains bacteria-derived pathways that have no counterpart in the human host and therefore may be ideal drug targets . However , more than a decade after its discovery , the essential functions of the apicoplast remain a mystery , and without a specific pathway or function to target , development of drugs against the apicoplast has been stymied . In this study , we use a simple chemical method to generate parasites that have lost their apicoplast , normally a deadly event , but which survive—“rescued” by the addition of an essential metabolite to the culture . This chemical rescue demonstrates that the apicoplast serves only a single essential function , namely isoprenoid precursor biosynthesis during blood-stage growth , validating this metabolic function as a viable drug target . Moreover , the apicoplast-minus Plasmodium strains generated in this study will be a powerful tool for identifying apicoplast-targeted drugs and as a potential vaccine strain with significant advantages over current vaccine technologies . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"biochemistry",
"infectious",
"diseases",
"chemical",
"biology",
"chemistry",
"biology",
"microbiology",
"molecular",
"cell",
"biology"
] | 2011 | Chemical Rescue of Malaria Parasites Lacking an Apicoplast Defines Organelle Function in Blood-Stage Plasmodium falciparum |
The pig parasite Ascaris suum plays and important role in veterinary medicine and represents a suitable model for A . lumbricoides , which infects over 800 million people . In pigs , continued exposure to Ascaris induces immunity at the level of the gut , protecting the host against migrating larvae . The objective of this study was to identify and characterize parasite antigens targeted by this local immune response that may be crucial for parasite invasion and establishment and to evaluate their protective and diagnostic potential . Pigs were immunized by trickle infection for 30 weeks , challenged with 2 , 000 eggs at week 32 and euthanized two weeks after challenge . At necropsy , there was a 100% reduction in worms recovered from the intestine and a 97 . 2% reduction in liver white spots in comparison with challenged non-immune control animals . Antibodies purified from the intestinal mucus or from the supernatant of cultured antibody secreting cells from mesenteric lymph nodes of immune pigs were used to probe L3 extracts to identify antibody targets . This resulted in the recognition of a 12kDa antigen ( As12 ) that is actively shed from infective Ascaris L3 . As12 was characterized as a phosphorylcholine-containing glycolipid-like antigen that is highly resistant to different enzymatic and chemical treatments . Vaccinating pigs with an As12 fraction did not induce protective immunity to challenge infection . However , serological analysis using sera or plasma from experimentally infected pigs or naturally infected humans demonstrated that the As12 ELISA was able to detect long-term exposure to Ascaris with a high diagnostic sensitivity ( 98 . 4% and 92% , respectively ) and specificity ( 95 . 5% and 90 . 0% ) in pigs and humans , respectively . These findings show the presence of a highly stage specific , glycolipid-like component ( As12 ) that is actively secreted by infectious Ascaris larvae and which acts as a major antibody target in infected humans and pigs .
Ascaris lumbricoides is the most prevalent intestinal parasitic nematode of man , infecting approximately 819 million people worldwide in developing countries [1] . Due to the high degree of morphological and genetic similarity , it is still debated as to whether A . lumbricoides from humans is a different species than A . suum from pigs [2–4] . Moreover , recent studies have shown that pig Ascaris is a zoonosis [5–8] . Even though anthelmintic treatment remains highly effective against A . lumbricoides , there is increased concern about the development of anthelmintic resistance . In addition , the high risk of reinfections after treatment calls for the development of new , long-acting solutions like vaccination . In addition , the development of more rapid and sensitive diagnostic techniques that adequately reflect the level of Ascaris exposure in a population could greatly improve our knowledge on infection dynamics and prevalence . Consequently , it would thus allow for a more precise estimate of the impact of infection and a better evaluation of a given intervention . Vaccination has proven to be the most efficient and cost-effective way of disease control [9] . Vaccination against ascariasis should in theory be feasible since pigs , repeatedly infected with A . suum , develop immunological responses at the level of the liver , lungs and intestine that stop migrating larvae from reaching adulthood . Furthermore , repeated exposure to the parasite induces an immunological response at the level of the intestine that is called the 'pre-hepatic barrier' , eventually preventing newly acquired larvae from migrating to the liver [10 , 11] . Recently , it has been shown that this immunity was associated with eosinophilia , mastocytosis and goblet cell hyperplasia in the caecum , the place where the infective stage 3 larvae ( L3 ) penetrate the intestine and start their hepatopulmonary migration [12] . However , it is still unclear which parasite products induce these immune responses or what the targets of these responses are . An increased understanding on this matter could offer important information for the development of protection by vaccination against this parasite . The need for improved methods to diagnose Ascaris infections in pigs and humans has recently been extensively discussed [13 , 14] . It was suggested that diagnostic tools detecting eggs in the stool are not useful for accurate evaluation of the level of exposure in pig farms [15] or sensitive enough for the detection of infection in humans where prevalence was low [16] . Serological tools detecting exposure to Ascaris might be more sensitive than egg based diagnostics for measuring prevalence or intensity of exposure in a human community [17] . Until now , only a handful of studies report the evaluation of antibody-based tests for ascariasis [18–23] . Recently , Vlaminck et al . , [15 , 17] showed that an ELISA detecting antibodies to Ascaris haemoglobin in plasma or serum samples appears to reflect general exposure to Ascaris on a community or herd level in humans and pigs , respectively . However , more species-specific antigens from early larval stages might increase the sensitivity and specificity of serological assays or recognize infections at an earlier stage . Hence , the main objective of this study was to use intestinal antibodies from pigs with a proven pre-hepatic barrier to identify immunogenic proteins of the infective stage larvae of A . suum and subsequently evaluate their protective and diagnostic potential .
The piglets used in this study were female and castrated male Rattlerow Seghers hybrid pigs of the local stock of the animal facility ( Ghent University ) . They were approximately 10 weeks old and weighed between 20 and 30 kg at the start of the trials . The pigs were raised indoors in a helminth free environment and had free access to a commercial feed and water . Adult female A . suum worms were collected from the intestines of naturally infected pigs from commercial farms that were being processed as part of the normal work at a local abattoir in Ghent , Belgium . Consent was acquired from abattoir management to collect the worms . Adult female A . lumbricoides worms were collected from the stool of individuals after treatment with anthelmintics in Jimma Town , Ethiopia [24] . Ascaris eggs were obtained by dissection of worm uteri and suspended in a 0 . 5% ( w/v ) potassium dichromate solution to a volume of 50 ml and placed in a culture flask at a concentration of 50 eggs/μl . The eggs were incubated at 27°C in the dark until third stage larvae were present and were subsequently used for infection trials or to extract infective stage-three larvae ( L3 ) . Fresh L3 were obtained from Ascaris eggs using the method described by Urban et al . [25] . In brief , Ascaris eggs , cultured in vitro were treated for 1 hr with commercial bleach and subsequently washed 3 times with phosphate buffered saline ( PBS ) of 37°C after which the eggs were transferred into an Erlenmeyer flask containing glass beads and a magnetic stir bar and stirred very slowly ( 60 rpm ) to induce hatching . After 15 min the suspension was poured onto a layer of cotton wool placed on top a Baermann apparatus with PBS at 37°C and left overnight . The next day , the larvae in the neck of the funnel were collected and washed 3 times in PBS . L3 extracts were prepared by grinding the collected larvae with a mortar and pestle that was placed in a bath of liquid nitrogen . The larval homogenate was transferred to a 15 ml tube and mixed with PBS and proteinase inhibitor cocktail ( 1:100 ) ( Sigma , Diegem , Belgium ) . The homogenate was then inverted at 4°C for 2 hrs followed by centrifugation for 30 min at 10 , 000g at 4°C . The supernatant ( L3 PBS ) was removed and kept on ice . The pellet was resuspended in PBS with 0 . 05% Tween-20 solution and fresh proteinase inhibitor cocktail ( 1:100 ) was added . The mixture was inverted at 4°C for 2 hrs and the supernatant ( L3 PBST ) was removed after centrifugation and stored on ice . Finally , the remaining pellet was resuspended in PBS containing Triton X-100 ( 2% ) and proteinase inhibitor cocktail ( 1:100 ) and subsequently inverted at 4°C for 2 hrs followed by centrifugation . The supernatant ( L3 Triton ) was collected and stored on ice . Subsequently , all extracts were sterilised by filtration ( 0 . 22μm ) and the filtrate concentrated at 4°C using a Centriprep centrifugal filter with YM-3 membranes ( Millipore , Overijse , Belgium ) . Protein concentration was determined by the BCA method ( Pierce , Rockford , USA ) and the extracts were stored at -80°C until use . In order to obtain L3 excretory-secretory ( E/S ) products , freshly obtained A . suum L3 were incubated at 37°C and 5% CO2 at a concentration of 5 , 000 larvae/ml in DMEM with 4 , 5 g/L Glucose , L-Glutamine and Pyruvate ( Thermo Fisher , Erembodegem , Belgium ) containing 1% Penicillin-Streptomycin ( P/S ) ( 5 , 000 u/ml Penicillin and 5 , 000 μg/ml Streptomycin , Thermo Fisher ) , 1% Kanamycin ( 10000 μg/ml , Thermo Fisher ) , 1% Amphotericin B ( 250 μg/ml , Sigma ) and 0 , 5% Gentamicin ( 10 mg/ml , Thermo Fisher ) . The culture fluid was collected daily and filtered using 0 . 2 μm membrane disc filters ( Supor 200 , Zaventem , Belgium ) , concentrated and dialysed against PBS in an Ultrafiltration Stirred Cell ( Millipore ) using a 10kDa cut-off filter membrane ( Millipore ) . The L3 E/S material was then stored in aliquots at -80°C . In order to obtain lung L3 and intestinal L4 and L5 stages , piglets were infected with approximately 100 , 000 infective A . suum eggs and sacrificed 7 , 14 and 28 days post-infection , respectively . The lung L3 and intestinal L4 were collected from minced lung tissue or intestinal content that was placed on a modified Baermann device as described by Slotved et al . [26] . L5 larvae were collected from the intestinal content by hand . Larvae were washed excessively using PBS 4°C and stored at -80°C . Adult worms were collected from the intestines of infected pigs from commercial farms that were being processed as part of the normal work at a local abattoir in Ghent , Belgium . Protein extracts from all life stages were produced as described above for the L3 stage . Finally , extracts from A . lumbricoides L3 , were obtained after sonication as previously described by Vlaminck et al . , [27] . Eight pigs were divided into 2 groups of 4 pigs . Pigs of group B were trickle infected 5 times a week with approximately 100 infective A . suum eggs in the feed for a period of 30 weeks . Pigs of group A were used as challenge controls . Two other pigs were euthanized before the start of the infection trial and used as negative controls . After the 30-week infection period , all pigs of group A and B were treated with a single dose of 5mg/kg fenbendazole ( MSD , Brussels , Belgium ) . One week after treatment , all pigs were infected with 2 , 000 infective A . suum eggs . All pigs were euthanized 14 days post challenge infection . During necropsy , the number of white spots , characteristic lesions on the liver caused by migrating A . suum larvae , was recorded and mesenteric lymph nodes from the small intestine , caecum and colon were collected and immediately processed as described below . Pieces of small intestine of approximately 1 meter were flushed three times with 50ml of PBS to rinse out all larvae in the intestinal lumen . The rinsing solution was collected and passed over a 200 μm mesh sieve . The remaining debris on top of the sieve was collected and examined for intestinal L4 larvae . After rinsing , the pieces of small intestine and the caecum and colon were cut open longitudinally and washed gently with excessive volumes of lukewarm tap water to remove any remaining intestinal content . Subsequently , mucus was collected by gently scraping the luminal side of the pieces of intestine with a microscope slide . During the experiment , blood samples for serum were collected from all pigs every 2 weeks . Single-cell suspensions of mesenteric lymph node cells ( LNC ) were prepared by mechanical disaggregation through a sterile stainless steel gauze . The LNC in ice cold PBS were centrifuged at 150 g for 10 min at 4°C , the supernatant removed and LNC resuspended in DMEM growth medium supplemented with 4 , 5 g/L Glucose , L-Glutamine and Pyruvate ( Thermo Fisher ) containing 1% P/S ( Thermo Fisher ) . The mononuclear cells ( monocytes and lymphocytes ( MNC ) ) were separated by the addition of LymphoPrep and centrifugation at 800g ( without brake ) for 30 min at 4°C . The MNC were collected from the interphase and washed twice with DMEM + 1% P/S + 2% heat-inactivated foetal bovine serum ( Moregate , Australia & New Zealand ) and the cells were subsequently suspended in DMEMcomplete . ( DMEM + 1% P/S + 1% Non-Essential Amino Acids ( Thermo Fisher ) , 1% Kanamycin ( 10 mg/ml , Thermo Fisher ) , 1% Amphotericin B ( 250 μg/ml , Sigma ) and 0 . 1% of a 0 . 35% β-mercaptoethanol solution ) . Viable MNC were counted , the concentration adjusted to 5 . 0 106 cells/ml in DMEMcomplete and cultured for 4 days in tissue culture flasks at 38°C and 5% C02 . Finally , the culture supernatant ( SN ) was collected , filtered using 0 . 2 μm membrane disc filters ( Supor 200 ) , concentrated using Centriprep centrifugal filter units with YM-3 membranes ( Millipore ) and stored in aliquots at -80°C . An equal volume of ice-cold PBS was added to the mucus collected from small intestine , caecum and colon and subsequently homogenised using an Ultrathurax mixer ( 2 min at 15 , 000 rpm ) . This mixture was then centrifuged for 15 min at 10 , 000g at 4°C , the supernatant was collected and centrifuged again as before . After this second centrifugation step , the supernatant was stored at -80°C . At a later time , antibodies were purified from this supernatant using Protein-A agarose beads ( Sigma ) following the manufacturer’s protocol . Purified mucosal antibodies were stored at -20°C until used . Protein extracts ( 5 μg ) were mixed with 5x sample buffer ( 60 mM Tris-Cl pH6 . 8 , 2% SDS , 10% glycerol , 5% β-mercaptoethanol , 0 . 01% bromophenol blue ) and put into a boiling water bath for 5 min . Afterwards , the samples were applied to 15% SDS-PAGE gels and separated by electrophoresis in Tris-Glycine buffer ( Tris 250mM , Glycine 200mM , SDS 1% w/v ) . Protein bands were visualized using SimplyBlue Safestain ( Thermo Fisher ) or SilverStain kit ( Thermo Fisher ) . Glycoproteins were stained by the use of the ProQ Emerald 300 gel stain kit ( Thermo Fisher ) . For Western blotting , SDS-PAGE gels were blot transferred to PVDF membranes ( Millipore ) or nitrocellulose membranes ( Thermo Fisher ) and blocked in PBS + 0 . 2% Tween80 ( PBSt80 ) or in PBS + 0 . 2% Tween20 ( PBSt20 ) + 5% Blotting-Grade Blocker ( BioRad ) . The blots were probed for 2 hrs ( 1ml/lane PBSt80 containing approximately 5 μg/ml antibodies purified from mucus or the concentrated culture supernatant of MNCs ) . The following conjugates and dilutions were used: goat anti-pig IgG-HRP conjugated ( Sigma ) ( 1/10 , 000 ) , goat anti-pig IgA-HRP ( Abcam , Cambridge , UK ) ( 1/5 , 000 ) . The immunoreactive antigens were visualised by chemiluminescent substrate ( 5ml 0 . 1M Tris pH 8 . 6 + 11μl 90mM p-coumaric acid ( Sigma ) + 25μl 250mM luminol ( Sigma ) + 15μl of 10% ( v/v ) H2O2 ) . L3 PBS extract was subjected to a Folch method [28] for the extraction of total lipids . In short , L3 PBS extract was mixed with chloroform/methanol ( 2/1 ) to a final volume 20 times the volume of the original extract and incubated for 20 min . The homogenate was centrifuged and the liquid phase recovered and washed with 0 . 2 volume of 0 . 9% NaCl solution . The mixture was centrifuged at 2 , 000 rpm to separate the two phases . Both upper and lower phases were collected separately and evaporated at 90°C under a fume hood . The dried lipid pellets were stored at -80°C until further use . To investigate the composition of the As12 antigen , total L3 PBS extract was incubated with different enzymes and submitted to different chemical treatments . For the enzymatic degradation , the L3 PBS extract was treated overnight at 37°C with pronase , lipase and trypsin at pH 8 . 0 and pepsin at pH 3 . 0 ( all from Sigma ) . Additionally , L3 PBS extract was treated overnight at 37°C or 60°C in 20mM periodic acid , 1M NaOH , 1M Trifluoracetic acid and 1M HCl or in PBS at 37°C , 60°C or 90°C . Purified As12 was treated in 48% aqueous HF at 4°C for two days , then lyophilized and used for Western blotting . The presence of phosphorylcholine ( PC ) on the antigen was confirmed by screening with TEPC-15 monoclonal IgA antibodies ( Sigma ) on western blot . The deglycosylation of The L3 PBS extract with PNGase F was performed under denaturing conditions according to the manufacturers’s protocol ( Promega ) . RNAse B was used for treatment control . The As12 antigen was also incubated in 50mM NaPO4 , pH 5 . 0 for 2 days at 37°C with 0 . 1 unit ß-N-acetylglucosaminidase from jack bean ( Sigma ) . After incubation , samples were boiled in 5x sample buffer , run on SDS-page , blotted onto PVDF membranes ( Millipore ) and recognition of the As12 antigen checked by incubation with mucosal antibodies from pigs of group B with 5μg/ml of antibodies/lane . For glycan analysis , aliquots of As12 were treated with PNGase-F , -A and chemical β-elimination to release N- and O-glycans , respectively . Released glycans were analysed by MALDI-TOF-MS after derivatisation with 2-aminobenzoic acid or permethylation , as described previously [29 , 30] . Approximately 2 μg As12 was hydrolysed in a glass vial with 50 μl 4 M TFA at 100°C for 4 hrs , dried under nitrogen , then monosaccharides labelled with 10 μl 2-aminobenzoic acid ( 2-AA ) labelling mix ( 48 mg/ml 2-AA , 1 M 2-picoline-borane dissolved in 30% acetic acid / DMSO ) followed by 2 hrs incubation at 65°C . For HPLC , 10 μl labelled sugars was added to 90 μl 0 . 6% sodium acetate , and 25 μl applied to a Superspher 100 RP-18 column 250 x 4 mm ( Merck ) . Buffers were: A = 0 . 1% butylamine , 0 . 5% phosphoric acid , 1% tetrahydrofuran; B = 0 . 05% butylamine , 0 . 25% phosphoric acid , 0 . 5% tetrahydrofuran , 50% acetonitrile . Run conditions were 0 . 5 ml/min , starting at 8% buffer B for 5 min , 8–25% buffer B over 25 min , 25–100% B over 2 min , maintained for 10 min . Monosaccharide standards ( 500 pmol Glc , Gal , Man , Fuc , Xyl , GlcNAc , GalNAc ) were treated with TFA , labelled , and ran as above . A total of 40 , 000 L3 were incubated for 1 hr at 37°C in 1ml RPMI ( Thermo Fisher ) with 20μg purified mucosal antibodies from pigs from group A , group B and from negative controls or with TEPC-15 antibodies ( Sigma ) at a concentration of 1/500 . After the incubation , larvae were washed three times with 1ml RPMI 37°C before being resuspended for 30 min in 1ml RPMI at 37°C containing FITC labelled anti-pig IgG ( Bethyl laboratories , Montgomery , TX , USA ) at a concentration of 1/1 , 000 or FITC labelled anti Mouse IgA at a concentration of 1/2 , 000 ( Thermo Fisher ) . All incubations with FITC-conjugated antibodies were conducted in the dark . Finally , the larvae were washed four times with 1ml of PBS 37°C to remove any remaining secondary antibodies and put on a glass slide for fluorescent microscopy analysis . All labelling experiments were performed in triplicate . To determine whether the stained L3 actively shed the As12 antigen , larvae were stained with 20μg of purified mucosal antibodies from pigs from group B as described above . After staining and washing , half of the larvae were killed by freezing them for 10 min at -80°C . The other half was kept in RPMI at 37°C . After this , the number of stained larvae in 3 aliquots of both groups was counted by fluorescence microscopy . This assessment was repeated after 1 , 2 , 3 , 4 and 24 hrs . Twelve piglets ( approximately 10 weeks old ) were divided into two groups of 6 pigs . Each pig in Group A was vaccinated with total lipid fraction purified from L3 PBS extract that was obtained from 200 , 000 L3 and dissolved in 0 . 5 ml sterile PBS + 0 . 5 ml Alhydrogel ( AlOH ) . The control pigs of group B were injected with 0 . 5 ml sterile PBS + 0 . 5 ml AlOH . Pigs were immunized three times by intramuscular injection at day 0 , 14 and 28 of the experiment . One week after the final immunization , all pigs were experimentally infected with 1 , 000 infective A . suum eggs in 5ml of tap water by oral intubation . Two weeks after infection , at day 49 of the experiment , all pigs were euthanized and L4 larvae were recovered from the small intestine as described above and counted . Blood samples were taken at the start of the trial , one week after the third immunization and at the time of necropsy to evaluate seroconversion against the As12 antigen following vaccination . Indirect ELISA determined antibody recognition of the As12 antigen by pig sera or human plasma samples . ELISA plates were coated with antigen overnight in carbonate buffer ( pH 9 . 6 ) at 4°C . Plates were coated with 1μl/ml of total lipid extract purified from 200 , 000 L3 which was dissolved in 50μl UPW . After three washes with PBSt , the plates were blocked with 100μl/well blocking buffer ( 5% milk powder ( w/v ) or 5% heat treated fetal calf serum in PBS ) for 2hrs at 4°C . Sera or plasma samples were added in duplicate at a dilution of 1/250 in PBSt for 2hrs at 4°C . The plates were washed again as before and incubated with the conjugate ( goat anti-pig IgG-HRP ( Sigma ) ( 1/10 , 000 ) , goat anti-human IgG4-HRP ( Southern biotech ) ( 1/2 , 000 ) ) in blocking buffer . Plates were incubated for 1hr at 37°C . O-phenylenediamine 0 . 1% in citrate buffer ( pH 5 . 0 ) served as substrate and after a 10 min incubation period in the dark , the development reaction was stopped by adding 50μl of 4M H2SO4 to all wells and optical density ( OD ) was measured at 490 nm . Sera from the trickle infected pigs described above were used to evaluate antibody response over time . Additional pig sera were obtained from an experimental infection trial performed by Nejsum et al . , [31] where piglets 10 weeks of age were infected twice a week in the feed for a total of 14 weeks with A . suum and Trichuris suis ( 25 and 5 embryonated eggs kg−1 day−1 , respectively ) . Serum and faecal samples were collected at the start of the trial ( W0 ) and 7 ( W7 ) and 14 ( W14 ) weeks after the first infection . Pigs were euthanized 14 weeks after the start of the experiment and the number of macroscopic worms present in the small intestine counted . The negative control was a pooled serum sample from 4 10-week old piglets without previous exposure to A . suum . The positive control was a pooled serum sample from 4 pigs after 14 weeks of daily infection with 100 A . suum eggs [12] . Reactivity of pig sera to the antigen is shown in ODr ( Optical Density ratio ) . ( ODr sample = ( OD sample−OD negative control ) / ( OD positive control−OD negative control ) ) . Plasma samples from A . lumbricoides infected humans were collected as previously described [32] from individuals living in Mainang village on Alor Island ( Province of East Nusa Tenggara , Timor , Indonesia ) , an area with high Ascaris prevalence ( >30% ) . A subset of 25 plasma samples of which all individuals had A . lumbricoides eggs in their stool was evaluated for anti-As12 IgG4 antibodies . A total of 24 plasma samples from individuals with hookworm infection used in this study were from people living in the East Sepik region of Papua New Guinea where A . lumbricoides was not present . Non-endemic control sera were from American subjects in St . Louis , MO , USA . Reactivity of human sera to the As12 antigen is shown in OD . TEPC-15 antibodies ( Sigma ) or pooled mucosal antibodies of pigs from group B of the trickle infection experiment were dissolved in PBSt at a dilution of 1/250 or 10μg/ml respectively and subsequently pre-incubated with PC-Cl salt ( Sigma ) at different concentrations ( 0 , 10 , 25 μg/ml final concentration ) for 30 min at room temperature . Following pre-incubation , the antibody mixtures were added to an As12-coated ELISA plate for 2hrs at 4°C . After washing the plates three times with PBSt , the TEPC-15 antibodies were detected by HRP labelled anti Mouse IgA at a concentration of 1/2 , 000 ( Bethyl laboratories ) and the mucosal antibodies were detected by HRP conjugated goat anti-pig IgG ( Sigma ) at a 1/5 , 000 dilution in blocking buffer for 1hr at 37°C . Finally , the plates were washed , developed and read as described above . All statistical analyses were performed using Graphpad Prism 6 . 0e for MacOSx . The Mann-Whitney U-test for pairwise comparison was used to compare the group means of the immune pigs and the challenge controls during the vaccination trial or to compare anti As12 antibody levels between different groups of human or pig plasma or serum samples . T-tests were performed to test for differences in the percentage of stained larvae or antibody reactivity of the two different experimental groups during the vaccination experiment . A paired t-test was used to detect the first time point at which infected pigs showed a significant higher ELISA reactivity to As12 antigen compared to the start of infection . Possible correlations between antibody responses and parasitological data ( EPG or worm counts ) were assessed using the Spearman's rank correlation coefficient . The use of computed Receiver Operating Characteristic ( ROC ) curves allowed for the determination of test sensitivity and specificity and to select an appropriate cut-off . Probability ( P ) values < 0 . 05 were considered to indicate significant differences . All animal experiments were conducted in accordance with the E . U . Animal Welfare Directives and VICH Guidelines for Good Clinical Practice . Ethical approval to conduct the studies was obtained from the Ethical Committee of the Faculty of Veterinary Medicine , Ghent University . The collection of adult A . lumbricoides worms was performed during a trial performed by Mekonnen et al . , in 2013 [24] . This study was approved by the ethical committee of Jimma University , Ethiopia , Ghent University and Antwerp University , Belgium . The Ethical Committee of the University of Indonesia , Jakarta approved the Alor Island study as previously described [32] . The Human Investigations Institutional Review Boards of Case Western Reserve University and the Papua New Guinea Medical Research Advisory Committee approved all protocols . The Institutional Review Board at Washington University School of Medicine in St . Louis , MO , USA approved our use of anonymized patient samples for the development of serological tests for helminth infections . Since written consent is not consistent with cultural norms on Alor Island , oral informed consent was obtained from all adults or , in the case of children , from their parents . The participant’s oral consent was noted on a survey questionnaire . The ethical board of the University of Indonesia and the institutional review boards in Germany and the USA approved the use of oral consent .
Immunity was induced in 4 pigs ( Group B ) trough trickle infection with 100 A . suum eggs 5 times a week for a period of 30 weeks . After a subsequent challenge infection with 2 , 000 infective A . suum eggs , the number of larvae and liver white spots were significantly reduced in these pigs when compared to control pigs who were not trickle infected ( Group A ) ( Table 1 ) . Challenge control pigs from Group A showed an average of 72 . 5 ± 43 . 3 liver white spots and 105 ± 81 . 3 L4’s in the intestine . In contrast , immunized pigs showed a 99% reduction in number of white spots on the liver ( 2 . 0 ± 1 . 8 ) and a 100% reduction in the number of L4 stage larvae in the intestine 2 weeks post-challenge , indicating the presence of intestinal immunity against infectious A . suum L3 in these pigs . Antibodies from the intestinal mucus and from MNC culture supernatant were used to detect immunogenic antigens in A . suum L3 protein extracts ( Fig 1 ) . Several antigens were recognized by IgG and IgA antibodies isolated from both the challenge control pigs and the trickle infected pigs . One zone of approximately 12kDa in size was recognised by IgG and IgA antibodies from the immunized pigs ( Group B ) . Although this zone was also detected by the challenge control animals ( Group A ) , recognition was generally much more intense in the immune animals ( Group B ) . This antigen , from here on out referred to as As12 , was also detected in the water-insoluble protein fractions of L3 ( S1 Fig ) . In addition , there was also increased recognition of a 21kDa antigen in the L3 PBS extract by mucosal IgA antibodies of immunized pigs . An immunoblot of the PBS extracts of the different life stages of A . suum showed that the recognition of the As12 antigen was restricted to the early L3 stage and its E/S products ( Fig 2A ) . The antigen was absent from the lung stage larvae onwards . In the L3 PBS extract from A . lumbricoides , a band of the same molecular weight as As12 was recognised ( Fig 2B ) . The As12 antigen could be visualised by a staining for carbohydrates or after immunoblotting but not by conventional protein staining methods like silver staining or coomassie staining ( Fig 3A ) . The As12 antigen could be purified from the L3PBS protein extract by performing a Folch extraction for the isolation of total lipid . The antigen appeared to be the sole recognized antigen in both the upper methanol and lower chloroform phase after Folch extraction ( Fig 3A ) . To further explore the specific properties and composition of the As12 antigen , the antigen was subjected to different chemical and enzymatic treatments . The immunoblot reaction to the As12 antigen by purified mucosal antibodies following exposure of the L3 PBS extract to a range of chemical and enzymatic treatments is shown in Fig 3B . Overnight treatment of the L3PBS extract with pronase , lipase , trypsin or pepsin did not affect antibody recognition of the As12 antigen . Similarly , heating the L3PBS extract to 60°C or 90°C overnight did not reduce the recognition of As12 , as well as treatment with 1M NaOH . Incubating the L3 PBS antigen with PNGaseF to cut off N-linked glycan structures or subjecting it to ß-N-acetylglucosaminidase treatment for the liberation of terminal ß-linked N-acetylglucosamine and N-acetylgalactosamine residues also did not diminish the recognition of the As12 antigen . The only chemical reactions that reduced or prevented recognition of As12 by porcine antibodies were treatment with 20mM periodic acid at 37°C or 60°C or 1M trifluoracetic acid and 1M HCl at 60°C . Finally , treatment of the As12 fraction with 48% HF for 48 hrs at 4°C for the removal of any PC groups on the antigen also reduced the immune recognition . The presence of a PC group on the As12 antigen was further confirmed by Western blot using specific anti PC antibodies ( TEPC-15 ) ( Fig 3C ) . Next to the As12 antigen , several other PC-bearing antigens were detected in the A . suum L3 extract . Mass spectrometric analysis of glycans released from As12 after enzymatical treatments with PNGase-F and -A or by chemical β-elimination provided no indications for the occurrence of common N- or O-glycans on As12 . Monosaccharide composition analysis indicated that the glycan portion of As12 consists mainly of GalNAc and GlcNAc in an approximate 1:2 ratio , with a minor amount of Glc ( S2 Fig ) . Freshly hatched A . suum L3 were incubated with antibodies purified from the intestinal mucus of immunized pigs ( Group B ) . On Western blot these antibodies mainly recognized the As12 antigen ( Fig 3C ) . Antibodies bound to the surface were detected with FITC labelled anti-pig total IgG . This showed complete staining of the outer surface of larvae that had shed their L2 sheath ( Fig 4A ) . Other larvae with partially shed L2 sheaths showed only incomplete staining ( Fig 4B ) . To further verify that the reactivity with mucosal antibodies was directed against the As12 antigen , L3 larvae were also incubated with monoclonal antibodies directed against PC . This labelling also resulted in staining of the L3 surface ( Fig 4C ) . Live L3 appear to shed off the antigen-antibody complexes ( Fig 4D ) . As a consequence , the number of L3 stained by mucosal antibodies diminished significantly over time when stained larvae were kept alive in culture . In contrast , when larvae were killed after staining , the number of stained L3 did not diminish over time ( Fig 4E ) . Results of the vaccination trial with the As12 antigen are shown in Fig 5A . Immunizing pigs with the total purified lipid fraction from A . suum L3 , containing the As12 antigen , did not seem to induce protection to subsequent homologue challenge infection . There was no significant difference in number of liver white spots or number of L4 recovered from the intestine 14 days after a challenge infection with 1 , 000 infective A . suum eggs . Pigs of the vaccinated group did however show a significantly stronger IgG response against the As12 antigen compared to control pigs ( Fig 5B ) . To evaluate whether the As12 antigen could serve as a serodiagnostic antigen to measure exposure to Ascaris , the serum IgG antibody responses against the As12 antigen in the immunized pigs of group A was measured over time ( Fig 6A ) . Anti-As12 antibodies were significantly elevated from 6 weeks after initial infection . Subsequently , sera from 91 experimentally infected pigs ( 31 ) were analysed using the As12 ELISA and showed that the reactivity to the As12 antigen increased significantly in pigs that were continuously infected with A . suum eggs ( Fig 6B ) . After ROC analysis , using the week 0 sera as Ascaris-negative and both week 7 and week 14 sera as Ascaris positive , the cut-off for positive individuals was placed at an ODr of 0 . 50 . Using this cutoff , the sensitivity of the ELISA was 98 . 4% ( 95% CI: 95 . 3–99 . 7 ) , the specificity was 95 . 5% ( 95% CI: 88 . 9–98 . 8% ) and the area under the curve was 0 . 99 ( ± 0 . 049 ) . No relationship was found between anti-As12 antibody levels and the faecal egg count after 7 or 14 weeks of trickle infection or the number of adult worms at necropsy . We subsequently tested whether the antibody reactivity to As12 was solely directed against the presence of PC on the antigen . For this , both specific anti-PC monoclonal antibodies ( TEPC-15 ) and pooled purified mucosal antibodies from trickle infected pigs were pre-incubated with increasing concentrations of PC-Cl salt to block the PC-binding sites . After pre-incubation , antibody preparations were used to detect As12 by ELISA . Pre-incubation of TEPC-15 antibodies with increasing concentrations of PC-Cl salt reduced binding reactivity to As12 in a concentration dependent way ( Fig 6C ) . Although pre-incubation with PC-Cl salt also affected the binding reactivity of pooled purified mucosal antibodies , the relative level of inhibition was not as pronounced as for the TEPC-15 antibodies ( 79% vs 27% and 91% vs 51% reduction of recognition at 10 and 25μg/ml PC-Cl salt respectively ) . This suggests that the recognition of As12 by Ascaris infected pigs was not solely directed against PC . Finally , the use of As12 as a serodiagnostic antigen was also validated with human samples . Twenty five human plasma samples from Indonesia from patients with proven A . lumbricoides infection ( EPG >50 ) were analysed using the As12 ELISA ( Fig 6D ) . In addition , 20 non-endemic plasma samples ( USA ) were used as negative controls and 24 plasma samples from individuals with hookworm infection were tested to evaluate possible cross reactivity . The levels of anti-As12 IgG4 in the human samples was significantly elevated in humans infected with A . lumbricoides . ROC analysis placed the optimal cut-off for the human As12 ELISA at an OD of 0 . 26 . Using this cut-off , the sensitivity of the ELISA was 92 . 0% ( 95% CI: 74 . 0–99 . 0% ) , the specificity was 90 . 0% ( 95% CI: 68 . 3–98 . 8% ) and the area under the curve was 0 . 96 ( ± 0 . 026 ) . A total of 24 out of 25 Ascaris infected individuals ( 96% ) , 2 out of 20 non-endemic plasma samples ( 10% ) and 2 out of 24 hookworm infected individuals ( 8 . 3% ) were positive for anti-As12 IgG4 antibodies when this cut-off was employed .
This study describes the identification and characterisation of As12 , a phosphorylcholine-containing glycolipid-like structure present on the surface of infective Ascaris L3 larvae . This antigen is targeted by the intestinal antibody response of pigs that have been previously exposed to infection . Both IgG and IgA isotype antibodies against As12 were produced by local antibody secreting cells and both Ig isotypes were detected in the mucus of infected pigs . Recognition of As12 seems to increase over time when pigs are continuously exposed to infection . The As12 antigen was also found in A . lumbricoides L3 and is being recognized by IgG4 from A . lumbricoides infected humans . As a result , an As12 based ELISA test could identify 98% of the Ascaris exposed pigs and 92% of Ascaris infected humans , whereas samples from non-endemic humans or hookworm infected individuals were negative except for two samples in each case . An amorphous envelope called the surface coat covers the L3 cuticle surface and it presents the greatest interface between the parasite and its host . Our results suggest that the As12 antigen is likely to be a component thereof . The surface coat or “fuzzy coat” is not derived from the cuticle but from specialized secretory glands . It is rich in carbohydrates , is dynamically responsive to changing host environments or immune attack and can be rapidly shed upon binding by antibodies and/or immune cells [33–35] . This process of shedding seems to be actively regulated in A . suum L3 since no decrease in staining was seen in dead L3 stained with immune pig antibodies whereas the percentage of live stained L3 decreased significantly over time . Similarly , antibody recognition of surface antigens on Toxocara canis L3 or Onchocerca cervicalis microfilaria was also lost in a temperature dependent manner or after incubation with antimetabolites [36 , 37] . This process of sloughing off antibody-bound antigens is likely to be one of the mechanisms used by Ascaris L3 in order to evade host immune responses . The high antigenicity of the As12 antigen , presented on the L3 cuticle surface , might engage the host immune system while the active shedding of the antigen after antibody binding prevents actual protective mechanisms to damage the mobile larvae . The antigen might also be actively secreted to influence the responses of nearby immune cells . The 35 kDa carbohydrate antigen ( CarLA ) detected on L3 of the sheep parasite Trichostrongylus colubriformis [38] shows characteristics similar to As12 as it is also very resistant to multiple enzymatic and chemical treatments , except for those that targeted carbohydrate structures . Although no glycans were detected following PNGase-F and -A digestion or chemical β-elimination , monosaccharide analysis indicated a simple composition suggesting a more polysaccharide nature with repeating units composed of GalNAc and GlcNAc . Just like As12 , CarLA is solely produced and excreted by the infective L3 stage and not by any other life stage [39] . However , unlike As12 , the CarLA antigen does not appear to possess any PC group [38] . This might explain why the As12 antigen could not be recognized by immune sheep sera or monoclonal antibodies directed against CarLA . The presence of a PC hapten is very common on nematode carbohydrates attached to a protein or lipid backbone . Previous studies have suggested that PC containing glycolipids from adult Ascaris worms have immunomodulatory properties [40–42] . It is possible that this As12 antigen has similar properties , but further research is required to confirm this . Despite the apparent antibody response against As12 in vaccinated pigs , there was no protection against subsequent challenge infection . A recent study by Masure et al . , [12] has shown strong eosinophil presence in the intestinal tissue of pigs which have developed a ‘pre-hepatic barrier’ . Possibly , the vaccination failed to induce an effective local intestinal cellular response which appears necessary to kill and thereby prevent the worms from migrating . Furthermore , systemic vaccination might not have induced the production of anti-As12 antibodies at the level of the intestine , where immune-mediated killing of the larvae is expected to occur [12] . We recognize the unfortunate fact that the presence of anti-As12 antibodies in intestinal mucus was not evaluated in this experiment . It is likely that the synergetic cooperation between both antigen specific antibodies and cellular responses at the site of infection is key in stopping the migrating larvae . The strong antibody responses that are mounted by the host against As12 may be useful for diagnostic purposes . Nearly all trickle-infected pigs showed strong development of anti-As12 antibodies , independent of whether they were harbouring adult worms in their gut or not . The onset of detectable antibodies was however first apparent after 6 weeks , which is similar to purified Ascaris haemoglobin antigen [15] . Infected humans also mounted a significant antibody response to the As12 antigen , whereas non-endemic or hookworm infected humans did not seem to recognize the antigen . This result , together with the fact that pre-incubation with PC-Cl salt of mucosal antibodies from trickle infected pigs did not completely prevent binding of As12 on ELISA suggests that recognition of As12 not only depends on the PC-group but possibly involves other parts of the antigen . More tests with sera from individuals with other well-characterized helminth infections need to be performed to further determine the species specificity of As12 as a serodiagnostic antigen . In conclusion , in this study we used locally secreted antibodies from the intestine of pigs with a pre-hepatic immunity to identify one major immunodominant antigen in the extracts and E/S material of infective Ascaris L3 . This As12 antigen is stage specific , of a glycolipid nature and is being actively secreted . Experimentally infected pigs or naturally infected humans develop a measurable antibody response against the As12 , advocating its possible use as a diagnostic antigen . The exact structure of this antigen and its biological role during infection is yet unknown and deserves further clarification . | Roundworms infect millions of humans and pigs throughout the world . The pig roundworm A . suum is a good model for A . lumbricoides infection in humans due to similar host physiology and the close genetic relationship between the worms . The aim of this study was to identify and characterize early larval antigens that are targeted by antibodies at the level of the intestine in immune pigs and to evaluate their protective and diagnostic potential . In order to do so , we generated highly immune pigs by repeatedly infecting them with A . suum for a long time ( 32 weeks ) . After necropsy , locally harvested antibodies from the gut were used to screen larval extracts . Hereby one particular antigen , named As12 , was detected . It was characterized as a molecule of glycolipid nature that is presented on , and actively secreted from , the surface of infective larvae . Pigs immunized with this antigen are not protected from subsequent challenge infection . Experimentally infected pigs or naturally infected humans do however mount a significant serological antibody response to the antigen . These findings shed light on a glycolipid-like antigen ( As12 ) that is secreted by infectious Ascaris larvae and is targeted by the immune system of infected humans and pigs . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"livestock",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"enzyme-linked",
"immunoassays",
"helminths",
"immunology",
"vertebrates",
"animals",
"mammals",
"developmental",
"biology",
"ascaris",
"ascaris",
"lumbricoides",
"antibodies",
"i... | 2016 | A Phosphorylcholine-Containing Glycolipid-like Antigen Present on the Surface of Infective Stage Larvae of Ascaris spp. Is a Major Antibody Target in Infected Pigs and Humans |
In experimental assays of angiogenesis in three-dimensional fibrin matrices , a temporary scaffold formed during wound healing , the type and composition of fibrin impacts the level of sprouting . More sprouts form on high molecular weight ( HMW ) than on low molecular weight ( LMW ) fibrin . It is unclear what mechanisms regulate the number and the positions of the vascular-like structures in cell cultures . To address this question , we propose a mechanistic simulation model of endothelial cell migration and fibrin proteolysis by the plasmin system . The model is a hybrid , cell-based and continuum , computational model based on the cellular Potts model and sets of partial-differential equations . Based on the model results , we propose that a positive feedback mechanism between uPAR , plasmin and transforming growth factor β1 ( TGFβ1 ) selects cells in the monolayer for matrix invasion . Invading cells releases TGFβ1 from the extracellular matrix through plasmin-mediated fibrin degradation . The activated TGFβ1 further stimulates fibrin degradation and keeps proteolysis active as the sprout invades the fibrin matrix . The binding capacity for TGFβ1 of LMW is reduced relative to that of HMW . This leads to reduced activation of proteolysis and , consequently , reduced cell ingrowth in LMW fibrin compared to HMW fibrin . Thus our model predicts that endothelial cells in LMW fibrin matrices compared to HMW matrices show reduced sprouting due to a lower bio-availability of TGFβ1 .
Fibrinogen occurs in three variants , whose relative abundance in fibrin affects its structure and pro-angiogenic capacity: ( 1 ) High molecular weight ( HMW , MW 340 kDa , ∼70% , Fig 1A ) ; ( 2 ) a low molecular weight form ( LMW , MW 305 kDa , ∼26% of total fibrinogen , Fig 1B ) that is formed after partial degradation of the carboxy-terminus of the fibrinogen Aα-chain; and ( 3 ) an alternative low-molecular weight form ( LMW’ ) that is formed after degradation of both Aα-chains ( MW 270 kDa , ∼4% of total fibrinogen ) [2] . HMW fibrin has a more open matrix structure , with larger openings between the fibers compared to LMW fibrin ( Fig 1A and 1B ) . LMW fibrin forms more complex networks with denser fibers . Fibrinogen composition is a key determinant of the number of ingrowth spots ( Fig 1C–1E and S2 Fig ) and the depth of sprouting [2–5] . hMVECs proliferate more and show more angiogenic ingrowth in HMW fibrin than in LMW or an unfractionated fibrin mixture [2] . The relative abundance of the three fibrinogen variants is changed in a number of pathologies . For example , the relative abundance of LMW and LMW’ has been found to be elevated in patients with vascular occlusion [6] and in patients with diabetes mellitus [7] , possibly due to enhanced vascular leakage . In cancer patients the fibrinogen levels were elevated , but no changes in the HMW:LMW ratio were found [6] . In post-operative patients [6 , 8] as well as after extensive acute myocardial infarction [8] the levels of HMW increased , followed by a delayed increase of LMW-fibrin [8] . In full-term newborns the levels of HMW have been found to be 25% lower than in adults [9] . Altogether , our in vitro evidence suggests that HMW-fibrinogen promotes angiogenesis more than LMW-fibrinogen . In vivo , increased levels of HMW are typically found in post-operative patients and after extensive myocardial infarction . It is unknown whether these changes in HMW:LMW ratios have clinical relevance , e . g . , in stimulating angiogenesis ( high HMW ) in post-operative patients or in the inhibition ( high LMW ) of angiogenesis in diabetes mellitus patients . During angiogenic ingrowth , the invading hMVEC proteolytically digest the fibrin matrix , suggesting that the low efficiency of in vitro angiogenesis in LMW fibrin is due to differential regulation of proteolysis . Cell-associated fibrinolysis is mostly performed by the trypsin-like protease plasmin [10–13] . Plasmin is the active conversion product of plasminogen , which is mainly produced by the liver and reaches fibrin scaffolds through the blood stream . Conversion of plasminogen into plasmin occurs by plasminogen activators and is highly regulated . Urokinase-type plasminogen activator ( uPA ) and tissue-type plasminogen activator ( tPA ) are secreted by ECs as inactive single-chain proteins . tPA is expressed in quiescent endothelium [14] and is primarily involved in clot dissolution [15] , whereas uPA and its cellular receptor ( uPAR ) are expressed during angiogenesis and control pericellular proteolysis [14 , 16] . ECs secrete inactive , single chain pro-uPA that binds to uPA receptors ( uPARs ) on the membrane of endothelial cells , and is subsequently converted into an active two-chained form . This active membrane-bound uPA-uPAR complex converts plasminogen into plasmin [11] . To balance fibrin degradation , ECs secrete plasminogen inhibitor type 1 ( PAI-1 ) that binds to tPA and uPA for deactivation and the PAI-1-uPA-uPAR complex is internalized [10 , 12] . Alongside plasmin , membrane-type 1 metalloproteinase ( MT1-MMP ) can perform cell-associated fibrinolysis [17] , although its role is still poorly understood: the MT1-MMP inhibitor TIMP-1 had only minor effects on sprouting in a 100% fibrin matrix , but was inhibiting when a 90% fibrin-10% collagen matrix was used [18] . Altogether , based on the available evidence we assume that hMVEC-associated fibrinolysis [2] is primarily due to the plasminogen-plasmin degradation system . To get more insight into a potential role of fibrinogen variants in regulating angiogenesis , here we ask , using mathematical modeling , what differences between HMW and LMW fibrinogen could explain the differences in angiogenic ingrowth that are observed in vitro . LMW fibrin has a reduced number of binding sites for growth factors , including latent-TGFβ1 [19] . TGFβ1 has a strong pro-angiogenic effect in hMVEC cultured on Matrigel [20] and is present in latent form in fibrin matrices . Thus , apart from the structural differences between fibrin variants discussed above , a possible difference between fibrin HMW and LMW matrices is their binding capacity of TGFβ1 . TGFβ1 upregulates PAI-1 and uPAR and is inhibited by TGFβ1 antagonist peptides . TGFβ1 also induces PAI-1 and uPAR expression in hepatic stellate cells [21] and uPA/PAI-1 levels in human tumor tissues [22] . LTBP1 ( latent transforming growth factor β binding protein 1 ) potentially binds the C-terminus of this Aα-chain: LMW fibrinogen has a reduced number of C-termini of the Aα-chain compared to HMW fibrinogen . The level of LTBP1 is dramatically reduced in LMW fibrinogen fraction I-9 , compared to commercially available fibrinogen and intact fibrinogen fraction I-2 [19] . LTBP1 sequesters latent-TGFβ1 in the plasma to fibrin , resulting in an inactive TGFβ1 reservoir within the fibrin matrix that can locally be activated and released by plasmin [23–25] . Endothelial cells also secrete TGFβ1 [25]; we here assume that this TGFβ1 fraction can be neglected relative to the high bio-availability of TGFβ1 in the matrix . Thus , the reduced number of LTBP1 binding sites in LMW fibrinogen compared to HMW fibrinogen can result in a lower bio-availability of TGFβ1 , thereby reducing angiogenesis . Based on the experimental data on cell-associated fibrinolysis and TGFβ1 that we have discussed above , we suggest that a local uPAR-plasmin-TGFβ1 positive feedback loop drives angiogenesis ( see Fig 2 ) . For simplicity , we assume that all cell-bound uPAR is active , i . e . , it is bound to uPA . Cell-bound uPAR activates plasmin ( Fig 2 , arrow 1 ) and plasmin locally degrades fibrin and releases and activates TGFβ1 from its latent binding protein ( see Fig 2 , arrow 2 ) . TGFβ1 stimulates the production/expression of uPAR in the protruding cell ( see Fig 2 , arrow 3 ) , whereas nearby cells , which experience only low TGFβ1-dependent uPAR stimulation , are silenced by self-secreted PAI-1 ( see Fig 2 , arrow 4 ) . The basic principle underlying this hypothesis is a reinforced random walk [26] , as introduced to the problem of angiogenesis previously [27 , 28]: ( 1 ) an external growth factor activates endothelial cells to enzymatically modify the ECM near the sprout , and ( 2 ) the endothelial cells move randomly , but with preference up gradient of the modified ECM . More recent models have described specifically the hMVEC-fibrin culture system in silico [29 , 30] . In both these previous models , the location of the novel capillary sprouts vascular ingrowths was specified a priori , prohibiting their use for analyzing the ‘degree’ of angiogenesis , usually measured as the number ingrowth spots in a cell culture [1] . Therefore , a detailed understanding and analysis of angiogenesis in the Koolwijk et al . [1] experimental 3D fibrin sprouting model requires two modifications of the assumptions in the previous work . Firstly , it is unpredictable which cells in the monolayer become sprout leaders ( ‘tip cells’ ) . Thus we cannot pre-assign the location of the tip cells [29 , 30] , or assign the onset of angiogenesis by punching a hole in the basal lamina [31] or by initiating its local digestion [27 , 28] . Also , previous models assumed that endothelial cells follow a gradient of VEGF . The Koolwijk et al . [1]in vitro model does not include growth factor gradients , so we have not included those in the present in silico model . This implies that both the location and the growth direction of sprouts in the present computational model emerge from local cell-cell and cell-matrix interactions . We hypothesize that such sprout initiation mechanisms may exist alongside the established role of the Dll4-Notch network in the selection of tip cells that lead the sprouts [32–35] . Altogether , to explore our hypothesis that the uPAR-plasmin-TGFβ1 positive feedback loop regulates spontaneous ingrowth , we model the plasminogen-plasmin degradation system in combination with a cell-based model of endothelial cell invasion . following previous continuum models [36–38] . We propose that a differential binding activity of TGFβ1 to HMW and LMW explains the higher ingrowth . We developed a computational model to evaluate if this sprouting mechanism can explain the reduced ingrowth in LMW compared to HMW; it is shown that it regulates the spacing of ingrowth spots and is also consistent with a number of additional observations .
To study how endothelial sprouting is reduced in LMW compared to HMW fibrin matrices , we developed a computational model that mimics the in vitro assay by Koolwijk et al . [1] and Weijers et al . [2] . Our hybrid model consists of a cell-based component to describe the endothelial cells and a continuum component to describe the plasminogen-plasmin system . The model represents a cross-section of the in vitro sprouting model ( Fig 3 ) , and is initialized with a monolayer of fifty endothelial cells on top of a fibrin matrix . Fibrin forms a physical obstruction for cells while , at the same time , offering support to the cells as they can adhere to fibrin . Using cell-based modeling , we explicitly describe cell shape , cell motility , cell-cell adhesion , and cell-fibrin adhesion . Each cell has a level of active uPAR , which homogeneously spread over the cell membrane , and each cell secretes PAI-1 into the extracellular space . PAI-1 , and the other extracellular proteins ( fibrin , latent-TGFβ1 , active TGFβ1 , plasminogen , and plasmin ) are modeled as concentration fields . The extracellular proteins interact with one another and with the membrane-bound uPAR ( Fig 2 ) . uPAR activates plasminogen , forming plasmin that degrades fibrin and locally activates latent-TGFβ1 by releasing it from the fibrin matrix . The active TGFβ1 induces the production of uPAR in nearby cells . These reactions form a local positive feedback loop that keeps the invasion of the endothelial cells going . To represent cells and their physical interactions with the fibrin matrix , the cellular Potts model ( CPM ) [40 , 41] was used . For details see Section Fibrin invasion . Briefly , the CPM represents cells on a regular lattice as patches of connected lattice sites . Cells move by copying lattice sites inward or outward , representing the extension and retraction of pseudopodia . To model the physical obstruction imposed by high concentrations of fibrin , the extension probability of a pseudopodium is reduced if it attempts to invade a lattice site with high fibrin concentration . The concentration of fibrin , f ( x → ) , is initialized at a uniform , non-dimensional concentration of f ( x → ) = 1 . 0 at all lattice sites x → . No fibrin is produced or added in the simulation , such that the concentration of fibrin will stay in the range f ( x → ) ∈ [ 0 , 1 ] . The invasion probability quickly drops for concentrations f ( x → ) > 0 . 5 , while for f ( x → ) < 0 . 3 fibrin does not hinder invasion . Fibrin is digested by the plasmin system , as illustrated in Fig 2 . The concentration of uPAR within each cell is modeled by an ordinary differential equation ( ODE ) . A concentration field for uPAR is projected on the CPM grid , with each lattice site that is occupied by a cell having the uPAR concentration of that cell . The concentration of uPAR moves along with the location of the cell after cell movement . A system of coupled partial differential equations ( PDEs , see Section Methods ) describes the reactions between fibrin , TGFβ1 , plasminogen , plasmin , PAI-1 and all fibrin-bound forms . The equations for the plasmin system were based on the continuum model by Diamond et al . [36] , which studies the penetration of uPA and tPA into a fibrin clot in the blood stream . To adopt this model to our problem , we added the uPAR-plasmin-TGFβ1 positive feedback loop , simplified the implementation of fibrinolysis , and deleted the convective terms . Time steps in our model are measured as Monte Carlo step ( MCS ) . One MCS is defined as the number of lattice site update attempts as there are sites in the lattice . It takes about 6000 MCS to simulate a 10 days long experimental assay , similar to the 3D-fibrin sprouting model of Koolwijk et al . [1]; so a MCS represents approximately 2 . 5 minutes in real time . In summary , the model is based on the following mechanistic assumptions: In the in vitro 3D-fibrin sprouting assay by Koolwijk et al . [1] , uroplasmin ( uPA ) and its receptor uPAR were localized specifically at the invading endothelial cells that lead the sprouts [46] . The selection mechanism of these ‘uPAR-rich’ cells in the monolayer is not fully understood . We therefore asked if the uPAR-plasmin-TGFβ1 positive feedback mechanism is sufficient to confine uPAR-expression to the invasive cells . We initialized the cells in our model with a uniform concentration of uPAR ( Fig 4A ) . Random cell movements change the contact-level and contact-duration with fibrin , resulting in local , random differences in the levels of plasmin activation . Fibrin is degraded at sites with a high plasmin activity ( Fig 4B ) , and TGFβ1 is released from the matrix ( Fig 4C ) . The active TGFβ1 induces the expression of uPAR in nearby cells ( Fig 4D ) . The expression of uPAR in more distant cells can also be induced by the released TGFβ1 to some degree , but such uPAR activity is counteracted by the self-secreted PAI-1 . Due to stochasticity , only a few cells in the monolayer can trigger the positive feedback loop sufficiently to overcome inhibition by PAI-1 and gain high levels of uPAR to start ingrowth ( Fig 4E ) . In absence of fibrin-bound latent-TGFβ1 , none of the cells in the monolayer , in a hundred stochastic simulations , manage to gain high levels of uPAR due to the lack of TGFβ1-induced uPAR expression . Thus , our modeling results show that a uPAR-plasmin-TGFβ1 positive feedback loop suffices to select uPAR-rich cells in a monolayer of endothelial cells to form ingrowth spots . Once uPAR-rich cells are selected spontaneously within the monolayer , the uPAR-plasmin-TGFβ1 positive feedback consolidates sprout progression in the model ( see Fig 5 ) . The cell leading the sprout , i . e . , the tip cell , has the highest concentration of uPAR ( see Fig 3E ) in agreement with experimental observations [46] . In agreement with in vitro observations , in the in silico model sprouts branch spontaneously ( see , e . g . , simulation 3 for a constant uPAR production rate of 0 . 003 Relative Units ( RU ) /MCS in Fig 5A ) . This occurs when a cell adjacent to the tip cell moves into another direction , or when a cell higher up in the sprout manages to trigger the feedback loop and starts a branch . Sprouts are not formed in every simulation: due to the stochastic fluctuations in cell shape and movement , in some cases none of the cells activate the positive feedback loop sufficiently to overcome the inhibition of PAI-1 . Similarly , ingrowth is not seen in every in vitro experiment , but is highly variable per cell donor and passage number of the cells within the same donor . In vitro , TNFα is required to induce sprouting of human endothelial cells [1] and the mean tube length increases at higher doses of TNFα . TNFα increases uPA production and the level of cell-bound uPA [1] . To test if the model correctly reproduced this in vitro observation , we mimicked the effect of TNFα by increasing uPAR expression in the endothelial cells . Fig 5A shows a set of simulation results after ten days of sprouting for a uPAR expression level of 0 . 001 , 0 . 002 , 0 . 003 , and 0 . 005 ( RU/MCS ) . In simulations with higher uPAR expression levels , the number of ingrowth spots increases . For each parameter setting , four simulation results for the same parameter settings are shown; these demonstrate the stochasticity of ingrowth frequency and sprout morphology . To quantify sprouting , we defined three measures: the angiogenesis level , the sprouting frequency and the fibrinolysis level . The angiogenesis level simultaneously reflects sprout depth and sprout count ( see Section Methods for the quantification algorithm ) . The blue curve in Fig 5B represents the mean angiogenesis level for all simulations that formed sprouts ( angiogenesis level>0 ) . The sprouting frequency is the number of simulations out of a hundred simulations that formed sprouts ( red curve in Fig 5B ) . The fibrinolysis level , defined as the mean percentage of initial fibrin lattice sites that are invaded by the endothelial cells in all 100 simulations , also increases for higher uPAR expression levels , as is expressed by the green curve in Fig 5B . In summary , an increase of the basal uPAR-bound uPA activity in all cells increases the probability that the uPAR-plasmin-TGFβ1 positive feedback loop is triggered in one of the cells in the monolayer , leading to high uPAR expression and sprout initiation . As a consequence , sprouts form more frequently and more excessively at higher uPAR expression levels . Thus , the model explains mechanically how ubiquitous stimulation of uPAR-bound uPA activity by TNFα leads to confined uPA activity and sprouting . A full quantitative validation of the model is not feasible at present , because only for a few parameters experimental estimates are available , leaving most other parameters as fitting parameters . To avoid overfitting , we have instead selected a set of ‘default’ parameter values for which the model qualitatively reproduces the fibrin culture system ( see S2 Table ) . To validate the model , we then tested if qualitative shifts in the parameter values , corresponding with published experiments , qualitatively reproduce the outcome of three published in vitro experiments of the plasminogen-plasmin degradation system . Firstly , Koolwijk et al . [1] reported that there was no angiogenic ingrowth and tubule formation in fibrin matrices that were made using plasminogen-depleted fibrinogen . In agreement with this observation , there is no ingrowth in our model for low initial level of fibrin-bound plasminogen ( Fig 6A ) . The sprouting percentage , the fibrinolysis percentage , and the angiogenesis level all increased with the initial fibrin-bound plasminogen concentration . Secondly , inhibition of uPAR-bound uPA activity by addition of uPA specific polyclonal antibodies , or prevention of the binding of uPA to uPAR by soluble uPAR or blocking antibodies inhibited capillary-like tube formation dose-dependently ( see Refs . [1 , 46] and Fig 6D ( bt + trasylol and bt + H2 ) ) . We mimicked the inhibition of uPAR activity by increasing the decay rate of uPAR . Consistent with the experimental results , Fig 6B shows that this parameter change results in a decrease of the sprouting percentage , the fibrinolysis percentage , and the angiogenesis level . Thirdly , experiments show that there is an optimum PAI-1 concentration for angiogenesis [47]: addition of PAI-1 to implants in wild-type mice enhanced angiogenesis up to 3-fold at low concentrations but inhibited angiogenesis nearly completely at high concentrations . In the 3D fibrin assay , addition of the anti-PAI-1 antibody MAI-2 shows a similar biphasic effect on angiogenesis ( Fig 6D ) : Moderate inhibition enhances tube formation , whereas strong inhibition reduces tube formation . This is due to excessive fibrinolysis , which is incompatible with normal capillary formation [48 , 49] . As for uPAR , we modeled the manipulation of PAI-1 activity by an increase of the decay rate of PAI-1 . Fig 6C shows that the fibrinolysis percentage strongly increases when the decay rate of PAI-1 is increased . High decay rate of PAI results in low PAI-1 activity , and thus in excessive fibrinolysis; no sprouts are formed , but the entire monolayer lowers simultaneously . Low decay rates of PAI-1 result in high PAI-1 activity and sprouting is completely inhibited . Only for intermediate levels of PAI-1 activity we find sprouting , indicated by the peaks in Fig 6C for the sprouting percentage and the angiogenesis level . In conclusion , the model can reproduce three essential validation experiments for the plasminogen-plasmin system . In absence of fibrin-bound latent-TGFβ1 , no sprouts are formed in all 100 simulations with a parameter set for which sprouts formed well in presence of fibrin-bound latent-TGFβ1 in Figs 5B and 6 ( constant uPAR production rate = 0 . 005 RU/MCS , initial fibrin-bound plasminogen concentration = 1 RU , PAI-1 decay rate = 0 . 01 MCS−1 , and uPAR decay rate = 0 . 0095 MCS−1 , using Relative Units , RU , and Monte Carlo Steps , MCS ) . This shows that initialization and consolidation of sprouts can be driven by activity of the proposed positive feedback loop formed by uPAR , plasmin , and TGFβ1 . Next we used our model to design new hypotheses about the mechanisms that reduce the level of angiogenic ingrowth in LMW fibrin matrices compared to HMW matrices . The level of LTBP1 is dramatically reduced in LMW fibrinogen fraction I-9 , which lacks major parts of the C-termini of the Aα-chain , compared to commercially available fibrinogen and intact fibrinogen fraction I-2 [19] . As LTBP1 sequesters latent-TGFβ1 to fibrin , this could result in a lower level of fibrin-bound latent-TGFβ1 . We hypothesize that this reduced level of fibrin-bound latent-TGFβ1 , in combination with our suggested local uPAR-plasmin-TGFβ1 positive feedback , could cause the reduced level of endothelial sprouting in LMW compared to HMW fibrin matrices . If the levels of inactive TGFβ1 in the fibrin matrix are too low , cells are not able to induce a strong enough uPAR-plasmin-TGFβ1 positive feedback loop to overcome the inhibition of PAI-1 and thus will not form sprouts . In line with this hypothesis , Fig 7A shows that the sprouting percentage , the fibrinolysis percentage , and the angiogenesis level decrease with lower initial concentrations of fibrin-bound latent-TGFβ1 in our model . In conclusion , our simulations results suggest that the angiogenic ingrowth is reduced in LMW fibrin matrices compared to HMW matrices due to a reduction in binding sites for LTBP1 . The addition of active TGFβ1 has a biphasic effect on in vitro sprouting [50] . Addition of active TGFβ1 to the assay stimulates sprouting at low doses and inhibits sprouting at high doses of TGFβ1 . To test this biphasic effect in the model , we initialized the model with a homogeneously spread concentration of active TGFβ1 . The medium containing TGFβ1 was refreshed every two days in vitro [50] . We similarly reset the TGFβ1 concentration to the initial value after every two days in the model . Fig 7B shows that TGFβ1 indeed has the reported biphasic effect on angiogenesis in the simulations . At low concentrations of added TGFβ1 ( TGFβ1 = 0 . 5 RU and TGFβ1 = 10 RU in Fig 7B ) , more sprouts are formed than without addition of TGFβ1 ( TGFβ1 = 0 in Fig 7B ) . The uPAR-bound uPA activity in all cells increases due to the overall addition of TGFβ1 , allowing some cells to overcome the inhibitory PAI-1 threshold for triggering the uPAR-plasmin-TGFβ1 positive feedback loop . This is a similar effect as was seen for the stimulation with TNFα above . The upregulation of uPAR-bound uPA activity is too strong at high doses of TGFβ1 , and consequently all cells degrade the matrix . This results in lowering of the complete endothelial cell monolayer , rather than in local sprouting ( TGFβ1 = 1000 RU in Fig 7B ) . In this case , fibrin is quickly degraded and some cells loose contact with the fibrin . Once the cells loose contact with the fibrin layer , they are no longer stimulated to migrate along with the degrading matrix and form the ‘fingers’ show in Fig 7B . In some simulations stacks of cells hovering above the monolayer were left behind . This is of course a model artifact , so we did not take those into account while quantifying the degree of spouting .
Endothelial cells in the in vitro assay of Koolwijk et al . [1] secrete PAI-1 , but it is unknown if all cells , or only the quiescent cells in the monolayer or perhaps only the invading uPAR-rich cells secrete PAI-1 . Interestingly , the uPAR-plasmin-TGFβ1 positive feedback loop resembles reaction-diffusion systems with activator-inhibitor dynamics [51–53] . Activator-inhibitor systems produce periodic patters , so such dynamics could be responsible for regular placement of tip cells in the in vitro assay . Most conditions for activator-inhibitor dynamics are met: The positive feedback loops stimulates local activation of uPAR-bound uPA , and the inhibitor PAI-1 diffuses faster than the “activator” uPAR , which is expressed intracellularly . A missing element for such activator-inhibitor dynamics is that the inhibitor ( PAI-1 ) must be produced locally , whereas we currently assumed that all cells secrete PAI-1 . We are unsure of this assumption: TGFβ1 induces production of uPAR as well as PAI-1 in MVEC cultured on Matrigel [20] , raising the possibility that uPAR-rich cells secrete most PAI-1 and that all conditions for activator-inhibitor dynamics are met . Thus future work should determine the localization of PAI-1 secretion in the 3D-fibrin sprouting assay [1] . Besides the activator-inhibitor dynamics , the closely related substrate-depletion model [52] is a well-studied theoretical model for pattern formation . In our model , plasminogen is the substrate for plasmin production . Conversion of plasminogen at sites of matrix invasion results in depletion of plasminogen in surrounding regions through diffusion . Indeed , plasminogen is a limiting factor for endothelial sprouting in the fibrin assay [1] . Plasminogen depletion has low impact in the current simulations , because we have initialized them with a high , homogeneous concentration of immobile , fibrin-bound plasminogen . However , plasminogen binds fibrin reversibly and can bind to ECs , so this mechanism might regulate the location of ingrowth spots for lower levels of fibrin-bound plasminogen . Interestingly , there is a delay in sprout initiation when the model is initialized with unbound plasminogen . It takes some time to reach high enough concentrations of fibrin-bound plasminogen , which is then converted to plasmin by uPAR for matrix degradation . A key patterning mechanism that is involved in angiogenesis is lateral inhibition by Delta-Notch signaling [32–35] . Cells that have high levels of Delta ligands on their membrane differentiate into so called ‘tip cells’ , which are the leaders of sprouts , and cells with low levels of Delta become ‘stalk cells’ [35] . Lateral inhibition occurs by interaction of Delta ligands with the Notch receptor of neighboring cells , resulting in the suppression of Delta production in those neighbors [32–35] . Lateral inhibition creates a pepper-and-salt pattern of tip and stalk cells , with tip cells surrounded by a rosette of stalk cells in monolayers in silico [54 , 55] . Thus , Delta-Notch signaling alone cannot account for the more widely spaced pattern of uPAR-rich leader cells in a monolayer as observed in vitro [46] . Possibly other regulation mechanisms , e . g . , the proposed uPAR-plasmin-TGFβ1 positive feedback loop , act alongside the Delta-Notch mechanism to distribute tip cells more sparsely . Notably , gene expression levels of Dll4 and Notch4 are significantly higher in endothelial cells cultured in LMW matrices than in HMW matrices [2] . The Dll4 and Notch4 expression differences by themselves cannot explain the reduced ingrowth in LMW fibrin matrices , as specific inhibition of Dll4-Notch was unable to induce recovery of tube formation in LMW . Inclusion of Delta-Notch signaling will likely affect sprout morphology . In simulations of our current model , cells adjacent to the tip cell are also activated by the released TGFβ1 , and they contribute to sprouting . This results in fairly wide , sometimes cyst-like sprouts . In our previous model [29] of the fibrin assay , narrow sprouts formed if only the tip cell secreted proteolytic enzymes for matrix degradation , and cyst-like sprouts formed when the stalk cells contributed to fibrin degradation as well . In this light , Delta-Notch signaling could repress proteolytic activity in cells adjacent to the tip cell , such that thinner sprouts will form . Our model explains differences in ingrowth between LMW and HMW fibrin based on the binding capacity of latent-TGFβ1 . An alternative explanation for the increased ingrowth in HMW fibrin compared to LMW fibrin could be that the ECs can invade the open matrix structure of HMW fibrin more easily . In absence of proteolysis , differences in matrix porosity can explain cell migration speed and persistence [56]; however , with small pore sizes of fibrin ( order 1 μm; see Fig 1 ) and the importance of fibrinolysis for angiogenic ingrowth , small differences in pore size are unlikely to contribute to differences in ingrowth . An alternative , or complementary explanation could lie in differences in the bulk mechanical properties of HMW and LMW fibrin . Indeed mechanical cell-cell communication [57 , 58] through strain-stiffening materials such as fibrin [59] suffices for generating vascular-like patterns [60] . In addition , individual fiber architecture , including fiber thickness and fiber density also affects cell spreading behavior on fibrin substrates independently of the bulk mechanical properties [61] , suggesting that fiber architecture differences ( Fig 1 ) could also contribute to differences in angiogenesis level on HMW and LMW fibrin matrices . The present model reproduces angiogenic sprouting by means of cell-fibrin adhesion and cell-division . A limitation is that the addition of TNFα in the in vitro model inhibits cell division [1] . The general cell invasion mechanism proposed here does not depend on cell division . Alongside the fibrinolysis-driven sprouting mechanisms proposed here , many alternative mechanisms of cell migration during angiogenic sprouting have been proposed that could act alongside or instead of cell division to replenish cells in growing sprouts . A range of models have shown that mutual attraction of endothelial cells suffices for the formation of vascular networks , e . g . , via a chemoattractant [62–70] , via mechanical forces [71 , 72] or via mechanically induced durotaxis [60] , and preferential attraction to elongated structures [73] . Our model could be extended with such sprouting and cell migration mechanisms to replace cell division . A detailed description of the plasminogen-plasmin system is included in our model , but still some simplifications were made . For instance , we did not take into account interactions with matrix metalloproteinases ( MMPs ) . Membrane-type 1 metalloproteinase ( MT1-MMP ) can perform cell-associated fibrinolysis [17] , but only plays a minor role in Koolwijk’s assay [18] . Furthermore , we neglected the low proteolytic activity of pro-uPA [11] , and only modeled active uPAR-bound uPA . Interactions between pro-uPA and plasmin could give interesting dynamics . Venkatraman et al . [37] considered a positive feedback loop in which the initial cleavage of plasminogen into plasmin is more efficient by uPA than pro-uPA , and the conversion of pro-uPA to uPA is driven by plasmin . By the use of a continuum model , they predict that uPA-plasmin activation is bistable in the presence of this positive feedback loop in combination with substrate competition for plasmin . A further limitation of the present model of the plasmin system , is that the numerical method cannot describe the advection of chemical species due to displacement of fibrin . This approximation is reasonable in the low Péclet number regime simulated here; i . e . , cell movement ( and the resulting advection of chemicals due to movement of fibrin ) is much slower than the movement of chemicals relative to the ECM due to diffusion and fibrin degradation . Because cells cannot ‘push’ fibrin , but only grow over it if fibrin is sufficiently degraded , the low Péclet number regime is ensured for fibrin and all fibrin bound growth factors . Also , cell movement is slower than the diffusive spread of the unbound growth factors , further justifying our approximation . A suitable method for modeling advective transport in the CPM due to cell movement for higher Péclet number cases has been proposed elsewhere [74] , and can be applied in future extensions of our model . It could be argued that the present two-dimensional approximation in silico does not represent Koolwijk’s three-dimensional cell culture model well , because in two-dimensional cell cultures the cellular micro-environment is usually not well represented [75] . However , note that in two-dimensional cross-section the cellular micro-environment of the endothelial cells corresponds with those in the three-dimensional cell culture . The leading cell is flanked by other endothelial cells and by the fibrin matrix ( see , e . g . , the uPAR-rich cell in Fig 4D ) , whereas the following endothelial cells are flanked by fibrin , culture fluid and cells ( see , e . g . , Fig 5A ) . Thus the two-dimensional cross-section in silico suffices as an approximation of the three-dimensional model in vitro . Nevertheless , the model will run in 3D with some adjustments , through appropriate scaling of the cell volume constraint and the adhesion parameters [76] . The positive feedback loop hypothesis , and the mechanisms involved , will both work in the same qualitative way in 3D as in 2D , since the reaction-diffusion equations have the same form . In conclusion , our model predicts that the reduced level of endothelial sprouting in LMW compared to HMW fibrin matrices can , at least in part , be explained by a reduced level of fibrin-bound latent-TGFβ1 in LMW fibrin . To validate this hypothesis experimentally , we propose to check if there is indeed a reduced level of fibrin-bound latent-TGFβ1 in the experimental setup [1 , 2] . As a second experiment , we propose to validate whether sprouting can be reduced in HMW fibrin matrices by addition of TGFβ1-antagonists . These validation experiments can bring us closer to an understanding of the mechanisms of selection of leader or ‘tip cells’ in the monolayer and sprouting in the in vitro setup .
The shape and motility of endothelial cells are modeled with the cellular Potts model ( CPM ) [40 , 41] . The model domain is a two-dimensional regular lattice Λ ⊂ Z 2 , with x → ∈ Λ the coordinates of the lattice sites . Cells and extracellular materials are projected onto the grid as patches of ( usually connected ) lattice sites , marked with the same unique identifier σ ( x → ) . Thus a generalized cell ( e . g . , a cell or ECM material ) s is defined as the set of lattice sites marked with the same identifier σ ( x → ) , C ( s ) = { x → ∈ Λ | σ ( x → ) = s } . Each identifier is further associated with a type τ ( σ ) . Here τ ( σ ) ∈ {cell , fibrin , cell , patch , border , medium}; its function is simply to define parameters and properties for categories of Potts domains , not for all domains individually . Cells move by extending or retracting pseudopodia , which include lamellipodia , filopodia and invadopodia . Pseudopodia movement is modeled by attempting to copy the state ( σ ( x → ) ) of a randomly selected lattice site x → into a lattice site x → ′ selected at random from the eight , first- and second-order neighbors . We then calculate the change , ΔH , of the Hamiltonian H = Hcontact + Hsize , which defines the force resulting from cell behaviors and properties in the model . An additional energy H0 is added to ΔH at the time of copying to represent dissipative energies ( or other copy biases ) , including those associated with physical obstruction by the fibrin matrix . The components of the Hamiltonian and H0 are described in more detail below . As in Hamiltonian systems F → ∝ ∇ → H , any copy attempt for which ΔH + H0 < 0 represents a passive force ( e . g . , due to adhesion or pressure differences ) that is sufficiently large to overcome the local , dissipative energies . These copy attempts are always accepted . In addition , cells exert active forces on their environment due to random membrane fluctuations; we assume these fluctuations are distributed according to the Boltzmann probability function , P Boltzmann ( Δ H , H 0 ) = e - Δ H + H 0 μ , ( 1 ) with μ , the rate of active random membrane fluctuations ( a . k . a . cellular temperature ) . The model includes a number of “static” cells , τ ( σ ) ∈ {cell patch , border} . Any copy attempt from or to the static states is ignored ( i . e . , updates are applied only if τ ( σ ( x → ) ) ∈ { medium , cell } ∧ τ ( σ ( x ′ → ) ) ∈ { medium , cell } ) . Copy attempts from fibrin sites ( τ ( σ ( x → ) ) = fibrin ) are also ignored; copy attempts into fibrin ( τ ( σ ( x → ′ ) ) = fibrin ) are a special case ( see Section Fibrin invasion ) . The plasminogen-plasmin system in this model is based on the continuum model by Diamond et al . [36] . We made some changes to make it suitable for our system and , most importantly , we included the uPAR-plasmin-TGFβ1 positive feedback , simplified the implementation of fibrinolysis , and removed convective terms . Fig 8 shows an overview of the binding and conversion reactions of plasminogen and latent-TGFβ1 in relation to fibrin that are included in our model . In this section we will discuss the reactions in Fig 8 to explain the PDE system that describes the plasminogen-plasmin system and the uPAR-plasmin-TGFβ1 positive feedback loop . | Therapies for a range of medical conditions , including cancer , wound healing and diabetic retinopathy can benefit from a better control over the growth of blood vessels . The chemical properties of fibrin , the material that forms scabs in wounds and can also occur in large concentrations in tumors , can regulate the degree of blood vessel growth ( angiogenesis ) . Angiogenesis can be mimicked in cell cultures . These allow us to modulate the chemical properties of fibrin and study the effect on angiogenesis . Fibrin occurs in high molecular weight ( HMW ) and in low molecular weight ( LMW ) forms . Interestingly , there is more ingrowth of angiogenic-like structures into HMW than in LMW fibrin , but the mechanisms are poorly understood . To get more insight into these , we constructed a computational model . Using the model , we propose and analyse a hypothetical mechanism for sprouting that could explain the differences in endothelial cell sprouting in LMW and HMW fibrin matrices . Our model suggests that cells digest fibrin , thus creating space for ingrowth . At the same time , digestion frees growth factors bound to fibrin , that activates further secretion of digestive enzymes by the cells . We propose that the resulting positive feedback loop spontaneously selects cells in the endothelial monolayer for ingrowth and helps the blood vessel sprout move deeper into the fibrin . This could be a complementary mechanism to lateral-inhibition by Delta-Notch for the selection of leader cells , also called ‘tip cells’ . Our model predicts that endothelial cells in LMW fibrin compared to HMW fibrin show reduced sprouting due to a lower bio-availability of TGFβ1 . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"cardiovascular",
"physiology",
"fibrinogen",
"enzymes",
"endothelial",
"cells",
"enzymology",
"fibrin",
"epithelial",
"cells",
"angiogenesis",
"developmental",
"biology",
"glycoproteins",
"plasmins",
"cellular",
"structures",
"and",... | 2018 | A local uPAR-plasmin-TGFβ1 positive feedback loop in a qualitative computational model of angiogenic sprouting explains the in vitro effect of fibrinogen variants |
Staphylococcus aureus is able to infect virtually all organ systems and is a frequently isolated etiologic agent of osteomyelitis , a common and debilitating invasive infection of bone . Treatment of osteomyelitis requires invasive surgical procedures and prolonged antibiotic therapy , yet is frequently unsuccessful due to extensive pathogen-induced bone damage that can limit antibiotic penetration and immune cell influx to the infectious focus . We previously established that S . aureus triggers profound alterations in bone remodeling in a murine model of osteomyelitis , in part through the production of osteolytic toxins . However , staphylococcal strains lacking osteolytic toxins still incite significant bone destruction , suggesting that host immune responses are also major drivers of pathologic bone remodeling during osteomyelitis . The objective of this study was to identify host immune pathways that contribute to antibacterial immunity during S . aureus osteomyelitis , and to define how these immune responses alter bone homeostasis and contribute to bone destruction . We specifically focused on the interleukin-1 receptor ( IL-1R ) and downstream adapter protein MyD88 given the prominent role of this signaling pathway in both antibacterial immunity and osteo-immunologic crosstalk . We discovered that while IL-1R signaling is necessary for local control of bacterial replication during osteomyelitis , it also contributes to bone loss during infection . Mechanistically , we demonstrate that S . aureus enhances osteoclastogenesis of myeloid precursors in vitro , and increases the abundance of osteoclasts residing on bone surfaces in vivo . This enhanced osteoclast abundance translates to trabecular bone loss , and is dependent on intact IL-1R signaling . Collectively , these data define IL-1R signaling as a critical component of the host response to S . aureus osteomyelitis , but also demonstrate that IL-1R-dependent immune responses trigger collateral bone damage through activation of osteoclast-mediated bone resorption .
Osteomyelitis , or inflammation of bone , is most commonly caused by invasive bacterial infection [1] . S . aureus is the most frequently isolated etiologic agent of both acute and chronic bacterial osteomyelitis [2 , 3] . S . aureus can colonize bone through hematogenous dissemination , contamination of bone following surgical or accidental trauma , or direct spread from a surrounding soft tissue infection [2 , 4] . Bone represents a unique niche for invading bacterial pathogens as it is constantly undergoing turnover by bone-forming osteoblasts and bone-resorbing osteoclasts . Bone also represents a distinctive immunological niche , as bone marrow houses hematopoietic stem cells that give rise to lymphocytes and myeloid cells [5] . Bone infections rarely resolve without medical intervention , and are difficult to treat due to the widespread antimicrobial resistance of S . aureus as well as induction of bone damage that effectively limits antibiotic delivery and immune cell influx [2 , 3] . Osteomyelitis elicits pathologic bone remodeling , which in addition to contributing to treatment failure , can enhance the likelihood of complications such as pathologic fractures [2 , 6–11] . In order to explore mechanisms of bone loss during osteomyelitis , we previously established a murine model of post-traumatic osteomyelitis [6] . Using this model , we identified a subset of staphylococcal toxins , the alpha-type phenol soluble modulins ( PSMs ) , that are responsible for killing primary bone cells in vitro and that enhance bone destruction in vivo [6 , 7] . However , a S . aureus strain lacking the alpha-type PSMs still incites substantial bone damage , causing approximately 70% of the bone loss that is observed in femurs infected with a wild-type ( WT ) S . aureus strain [6] . PSMα toxins , along with many other staphylococcal toxins and proteases , are regulated by the agr quorum sensing system . In rabbit and murine models of experimental osteomyelitis , inactivation of agr further reduced bone destruction [6 , 11] . However , significant cortical bone loss still occurred even with this virulence-attenuated strain in our model of post-traumatic osteomyelitis [6] . Taken together , these findings indicate that while bacterial factors directly contribute to bone damage , a substantial proportion of bone loss during osteomyelitis may be caused by host factors [12] . In order to maintain skeletal strength and structure , bone must be continuously remodeled by bone-forming osteoblasts and bone-resorbing osteoclasts [13 , 14] . In addition to a major role in bone formation , osteoblast lineage cells are the major cellular regulators of bone-resorption through the balanced production of a TNF-family cytokine known as receptor activator of NFκB ligand ( RANKL ) , and the RANKL decoy molecule osteoprotegerin ( OPG ) . RANKL production favors bone resorption by acting on osteoclasts , which are multinucleated bone-degrading cells that differentiate from the myeloid lineage . Bone remodeling occurs as a part of normal vertebrate physiology , but the kinetics of bone remodeling can be substantially altered in response to local and systemic inflammation [15] . Osteomyelitis , in particular , is associated with abundant levels of pro-inflammatory cytokines such as TNFα , IL-1β , and IL-6 [5 , 16] . These pro-inflammatory cytokines have been shown to promote skeletal cell differentiation in vitro , both directly by stimulating bone-resorbing osteoclasts and indirectly by promoting osteoblast production of RANKL to drive osteoclastogenesis [15 , 17 , 18] . IL-1 in particular was formerly referred to as “osteoclast activating factor , ” reflecting the ability of IL-1α and IL-1β to signal on osteoclast lineage cells to increase osteoclast viability and resorptive capacity [19–24] . Through these mechanisms , pro-inflammatory cytokines contribute to bone loss in vivo in non-infectious models of rheumatoid arthritis [17 , 25 , 26] , although less is known about their influence on bone loss during osteomyelitis . These observations led us to hypothesize that S . aureus osteomyelitis triggers enhanced bone loss through pro-inflammatory cytokine production and signaling . IL-1 cytokines signal downstream of the IL-1R through the adapter protein MyD88 , which also transduces signals from various Toll-like receptors ( TLRs ) after ligation by conserved microbial motifs known as pathogen-associated molecular patterns ( PAMPs ) . Thus , MyD88 is a critical component of the innate immune system , by relaying signals through the IL-1R and many TLRs . Prior research has highlighted a prominent role for MyD88 and IL-1R signaling in the activation of immune responses that are necessary to control S . aureus infection in other animal models of infection [27–35] . In part , this occurs through the ability of IL-1 to mediate neutrophil recruitment and promote proper abscess formation for containment of S . aureus [27 , 30] . Moreover , IL-1 plays a critical role in potentiating granulopoiesis , which occurs primarily in the bone marrow [36 , 37] . The expansion and recruitment of granulocytes , such as neutrophils , are regulated in part by an IL-1R-dependent mechanism by which IL-1 signals onto endothelial cells in the bone marrow to release G-CSF [38–40] . Thus , MyD88 and the IL-1R form a critical signaling cascade that is necessary to mount an effective immune response to invading pathogens . Importantly , osteoblasts and osteoclasts express innate immune receptors through which these cells sense and respond to PAMPs and inflammatory cytokines in cell culture [26] . Given the important role of IL-1 in anti-staphylococcal immunity , as well as compelling evidence demonstrating that IL-1 signaling impacts bone cells in vitro , we hypothesized that MyD88 and IL-1R signaling are required for efficient antibacterial immune responses during osteomyelitis , but paradoxically may also promote pathologic bone loss . To test this hypothesis , we used a murine model of S . aureus osteomyelitis , high resolution imaging , histologic analyses , and in vitro skeletal cell assays . We show that IL-1 is abundantly produced in bone in response to S . aureus infection , and that MyD88 and IL-1R signaling are required to limit staphylococcal burdens during osteomyelitis . Furthermore , S . aureus incites bone loss in vivo through an IL-1R-mediated increase in osteoclastogenesis . Our findings reveal that while MyD88 and IL-1R signaling are necessary for antibacterial responses in bone , they also contribute to S . aureus-stimulated osteoclastogenesis and host-mediated bone loss during osteomyelitis .
In order to determine changes in bone remodeling that occur during osteomyelitis , we compared architectural bone parameters between infected and mock infected wild-type ( WT ) C57BL/6J mice in a post-traumatic model of S . aureus bone infection [6] . We focused our analyses on two distinct anatomical sites of the infected femurs representing the two major architectural types of bone: cortical bone that comprises the mid-region ( diaphysis ) of the long bone and trabecular bone found in the distal femur ( metaphysis and epiphysis ) ( S1A–S1C Fig ) . We previously observed that mock infected WT mice display a rapid cortical bone healing response at the surgical site , in which the induced bone defect in the femoral diaphysis is replaced with new bone by 2 weeks post-surgery [6] . In contrast to this sterile cortical bone repair , mice infected with S . aureus develop osteomyelitis , are unable to restore the cortical bone defect , and experience extensive cortical bone loss surrounding the site of inoculation ( Fig 1A and 1B ) . Moreover , S . aureus infected femurs show reactive cortical bone formation surrounding the site of inoculation ( Fig 1C ) [6] . S . aureus osteomyelitis induces dramatic alterations in cortical bone surrounding the infectious focus , which was initiated in the middle of the femoral diaphysis . However , trabecular bone , located at the ends of the long bones , is the major site of homeostatic bone remodeling [41] . In order to elucidate how inflammation during osteomyelitis leads to alterations in trabecular bone architecture , we also performed micro-computed tomography ( μCT ) imaging on trabecular bone in the distal femur . To determine the amount of trabecular bone that was lost during osteomyelitis , we calculated the trabecular bone volume per total volume ( BV/TV ) , which is a standard measure of bone volume and architecture [42] . S . aureus infected femurs exhibited a dramatic loss in trabecular bone , with BV/TV markedly decreased in infected femurs compared to mock infected femurs ( Fig 1D and 1E ) . The observed decrease in BV/TV during infection is reflective of a decline in the number of bony trabeculae , which in turn increases the overall volume of space between trabeculae ( Fig 1F and 1G ) . Trabecular thickness was not significantly reduced in infected relative to mock infected femurs ( Fig 1H ) . Although skeletal histology revealed that the area of the femur encompassing the trabecular bone did not have apparent abscess formation ( S1C Fig ) , viable S . aureus cells were recoverable from the femoral epiphyses encompassing the trabecular bone ( S2 Fig ) . These data collectively reveal that S . aureus osteomyelitis induces changes in bone turnover throughout the entire infected femur , which is reflected in a significant loss of cortical and trabecular bone . One common mechanism of bone loss is mediated by an increase in the number of bone-resorbing osteoclasts residing on the bone surface . To examine if the inflammation associated with S . aureus osteomyelitis enhances numbers of osteoclasts in vivo , we collected histologic sections of infected femurs for histomorphometry , which enables quantification of the number of osteoclasts and osteoclast resorbing surface relative to intact trabecular bone . Histomorphometric analysis showed an increased number of osteoclasts per bone perimeter ( N . Oc/B . pm ) in S . aureus infected femurs relative to mock infected femurs , suggesting that enhancement of osteoclastogenesis might be one mechanism underlying trabecular bone loss during osteomyelitis ( Fig 1I ) . Taken together , these data indicate that S . aureus osteomyelitis perturbs normal bone homeostasis to induce pathologic bone remodeling in both cortical and trabecular bone . Previous studies have shown that toxin-deficient S . aureus strains retain the ability to alter bone remodeling , albeit to a lesser extent than WT S . aureus , implicating inflammation as a potential mediator of dysregulated bone remodeling during osteomyelitis [6 , 7] . To characterize the local inflammatory environment during S . aureus osteomyelitis , we conducted longitudinal , multiplexed cytokine profiling of S . aureus and mock infected femurs over the course of 14 days . Relative to mock infected femurs , S . aureus infected femurs have more abundant levels of cardinal pro-inflammatory cytokines including IL-1α , IL-1β , IL-6 , and TNFα ( Fig 2 ) . While both IL-1α and IL-1β are highly produced in infected femurs , IL-1β had a higher fold change than IL-1α throughout the timecourse when comparing S . aureus infected to mock infected femurs . Furthermore , infected femurs have increased levels of cytokines that support myeloid cell chemotaxis and expansion , including KC ( CXCL1 ) , G-CSF , M-CSF , MCP-1 ( CCL2 ) , MIP-1α ( CCL3 ) , MIP-1β ( CCL4 ) , and MIP-2 ( CXCL2 ) , compared to mock infected femurs ( Fig 2 ) . Cytokine profiling of S . aureus osteomyelitis demonstrated that inflammatory cytokines , chemokines , and growth factors are greatly increased in infected femurs by day 1 and throughout infection . A rapid and robust cytokine response to S . aureus in bone led us to focus on identifying host signaling pathways that are responsible for coordinating an innate immune response . Given the central role for the signaling adapter MyD88 in pathogen recognition and induction of innate immune responses , we first sought to determine how MyD88 signaling influences staphylococcal burdens and host morbidity and mortality during osteomyelitis . Myd88-/- mice have enhanced susceptibility to , and morbidity from , bacterial infection [27 , 43 , 44] . We therefore inoculated these mice with a range of S . aureus colony forming units ( CFUs ) , from 104−106 . Although bacterial inocula up to 106 CFUs did not cause mortality in WT mice , Myd88-/- mice were exquisitely susceptible to S . aureus osteomyelitis , with mortality observed even at inocula as low as 104 CFUs ( Fig 3A ) . For infected Myd88-/- mice that met humane endpoints prior to the experimental endpoint or succumbing to disease , bacterial CFUs were enumerated at that time in the femur , liver , and kidneys . At the time of early sacrifice , Myd88-/- mice had between 107−108 S . aureus CFUs in the femur , kidneys , and liver , which was significantly increased over CFUs recovered from WT mice ( S3 Fig ) . At the experimental endpoint ( day 14 post-infection ) , surviving Myd88-/- mice not only had significantly elevated bacterial burdens in the infected femur , but also experienced more bacterial dissemination to the kidneys and liver ( Fig 3B ) . Consequently , the inability to prevent systemic bacterial dissemination results in significantly increased mortality in Myd88-/- mice . Taken together , these data demonstrate a critical role for MyD88-dependent immune responses during S . aureus osteomyelitis . Myd88-/- mice have altered intestinal barrier function and are severely immunocompromised , and therefore may have significant microbiome differences relative to WT mice [45–47] . When considered in concert with recent studies suggesting that the microbiome may regulate bone mass [48–51] , these observations prompted us to breed Myd88-/- and Myd88+/+ littermate controls from a heterozygous colony and compare these littermate controls for susceptibility to osteomyelitis . We also examined the influence of sex as a biologic variable in these experiments . In line with results from mice bred in separate colonies , significant mortality from osteomyelitis was observed in male Myd88-/- mice , but not Myd88+/+ littermate controls ( S4A Fig ) . Of note , at day 14 post-infection , male Myd88-/- mice had no difference from Myd88+/+ littermate controls in recovered CFUs from infected femurs ( S4B Fig ) . This observation could indicate sex-dependent differences in osteomyelitis pathogenesis , or alternatively may reflect selection bias from removal of mice that succumbed to infection ( S4A Fig ) . In contrast to male mice , female Myd88-/- mice exhibited significantly higher bacterial burdens in the infected femur when compared Myd88+/+ littermate controls ( S4C Fig ) . These experiments confirm that MyD88 is critical for the control of bacterial burdens and systemic dissemination during osteomyelitis independently of any confounding variables associated with separate colony maintenance . In the absence of MyD88 signaling , mice are unable to control S . aureus infection , indicating that upstream receptors that signal through MyD88 , including S . aureus-recognizing TLRs and IL-1R , may be important for antibacterial protection . The high levels of IL-1 and IL-1-regulated cytokines present in S . aureus infected femurs led us to investigate the contribution of IL-1R signaling to anti-staphylococcal immunity in bone . We subjected WT and Il1r1-/- mice to S . aureus osteomyelitis . In contrast to the extreme systemic morbidity observed in MyD88-/- mice suffering from osteomyelitis , Il1r1-/- mice had less morbidity when compared to WT mice , in that they lost significantly less weight over the course of infection ( Fig 4A ) . To determine the role of IL-1R and the relative contributions of IL-1 isoforms ( IL-1α or IL-1β ) to control bacterial burdens in bone , WT , Il1r1-/- , Il1a-/- , and Il1b-/- mice were subjected to S . aureus osteomyelitis . Enumeration of bacterial burdens revealed that Il1r1-/- mice harbored significantly higher bacterial burdens in infected femurs than WT , Il1a-/- , and Il1b-/- mice ( Fig 4B ) . Il1a-/- and Il1b-/- mice sustained bacterial burdens that were not significantly different from WT mice ( Fig 4B ) . Unlike MyD88-/- mice , Il1r1-/- mice were protected from significant systemic dissemination to the liver or kidneys ( Fig 4C ) . To determine whether differences in WT and Il1r1-/- strains were due to background genotype or separate colony maintenance , heterozygous Il1r1+/- mice were bred to generate Il1r1+/+ and Il1r1-/- littermate controls . Infection of littermates with 106 S . aureus CFUs confirmed that Il1r1-/- mice sustained higher bacterial burdens in bone compared to Il1r1+/+ mice ( Fig 4D ) . We next investigated the kinetics of bacterial clearance between WT and Il1r1-/- mice . For this analysis we chose a lower S . aureus inoculum of 105 CFUs in an attempt to equilibrate bacterial burdens at day 14 post-infection . In both WT and Il1r1-/- mice , the initial S . aureus inoculum of 105 CFUs replicates to approximately 107 CFUs by day 1 post-infection ( Fig 4E ) . In WT mice , bacterial burdens decreased by greater than 1 log between days 3 and 5 post-infection . In contrast , bacterial burdens in Il1r1-/- mice were essentially unchanged through day 5 post-infection , and only declined between days 5 and 10 post-infection . Accordingly , WT and Il1r1-/- mice had significantly different bacterial burdens at day 5 post-infection with this lower inoculum , even though bacterial burdens were roughly equivalent at the final time point ( day 14 ) . These data reveal differences in infection kinetics between WT and Il1r1-/- mice , and suggest that Il1r1-/- mice might have a delay in bacterial control during osteomyelitis . In other S . aureus infection models , IL-1 coordinates neutrophil recruitment and is necessary for sequestration of S . aureus into mature abscesses [27 , 30] . We therefore hypothesized that the delay in bacterial clearance in Il1r1-/- mice subjected to osteomyelitis was related to differences in abscess maturation and neutrophil abundance . To visualize immune cell infiltration and abscess structure , we conducted myeloperoxidase ( MPO ) staining on histologic sections of infected femurs at day 14 post-infection . Il1r1-/- mice with osteomyelitis have differential MPO staining in comparison to WT controls , suggesting these mice have disorganized abscess structure ( Fig 5A ) . WT mice have MPO+ cells that surround and encompass the abscess , whereas Il1r1-/- mice show extensive MPO+ staining throughout the femur . To assess changes in inflammatory signatures that correspond to differences in infection kinetics , infected femur homogenates from WT and Il1r1-/- mice were analyzed using multiplexed cytokine analysis . In comparison to WT mice , Il1r1-/- mice had significantly decreased abundance of neutrophil growth factors G-CSF and GM-CSF and lower levels of the neutrophil chemokine CXCL1 at day 1 post-infection , a timepoint that precedes early bacterial control in WT mice between days 3 and 5 ( Figs 5B–5D and 4E ) . GM-CSF and CXCL1 levels then decline in WT mice by day 5 post-infection . In contrast , Il1r1-/- mice display significantly higher levels of GM-CSF and CXCL1 at day 5 post-infection when compared to WT mice , prior to the decrease in bacterial burdens that occurs between days 5 and 10 post-infection ( Figs 5C , 5D and 4E ) . Moreover , there are global changes in cytokine abundance when comparing WT and Il1r1-/- mice ( S1 Table ) . These data suggest that WT mice have an early influx and/or expansion of neutrophils , important for the control of bacterial burdens and normal abscess formation . To monitor neutrophil abundance during the course of osteomyelitis , bone marrow from infected and contralateral , uninfected WT and Il1r1-/- femurs at various time points after S . aureus infection was analyzed via flow cytometry ( S5A–S5F Fig ) . Neutrophils were identified as CD45+CD11b+Ly6G+Ly6Clo and reported as the percent of CD45+ immune cells . At day 1 post-infection , WT and Il1r1-/- mice were found to have neutrophils comprising less than 10% of CD45+ cells in the bone marrow . By day 3 post-infection , Il1r1-/- mice have significantly fewer neutrophils in the infected bone marrow compared to WT mice ( Fig 5E ) . Furthermore , differences between relative neutrophil abundance were also observed between WT and Il1r1-/- mice in the contralateral , uninfected femurs . Neutrophil abundance at day 5 post-infection is comparable between WT and Il1r1-/- genotypes , but again was significantly decreased in Il1r1-/- femurs at day 14 post-infection . Therefore , the data suggest that Il1r1-/- mice with S . aureus osteomyelitis have altered neutrophil responses , indicated by the significant decrease in relative neutrophil abundance at two timepoints post-infection . WT mice subjected to S . aureus osteomyelitis display significant cortical bone destruction and reactive bone formation at the site of infection , while sustaining alterations in osteoclast number and trabecular bone loss in the distal femur ( Fig 1A–1I ) [6] . Given the important role of IL-1R signaling in skeletal cell differentiation and function in vitro , we hypothesized that pathologic bone remodeling during S . aureus osteomyelitis is mediated , in part , by IL-1R signaling . To test this hypothesis , we first compared changes in cortical bone remodeling between WT and Il1r1-/- mice using a lower dose ( 105 CFUs ) S . aureus infection . At the site of infection , Il1r1-/- mice sustained increased cortical bone loss in comparison to WT mice ( Fig 6A and 6B ) . In areas adjacent to the cortical bone loss , Il1r1-/- mice had a dramatic increase in new bone formation . In fact , the volume of new bone formation in Il1r1-/- mice was nearly twice the volume formed in infected WT femurs ( Fig 6C ) . Importantly , these data do not completely control for differences in bacterial burdens at the site of infection , where Il1r1-/- mice harbor higher bacterial burdens at day 5 post-infection with a lower inocula ( Fig 4E ) . However , Il1r1-/- mice do equilibrate bacterial burdens to the same level at WT mice by day 14 post-infection ( Fig 6D ) . Consistent with an increase in reactive bone formation , histologic analyses revealed increased callus formation in infected Il1r1-/- femurs and also demonstrated qualitative differences in callus composition compared to infected WT femurs ( Fig 6E–6G ) . Additionally , the dramatic cortical bone alterations observed in Il1r1-/- mice were confirmed in littermate controls ( S6A and S6B Fig ) . In contrast to cortical bone remodeling changes during S . aureus infection , μCT analysis revealed no differences in cortical bone remodeling of a sterile bone defect at day 14 post-surgery , where mock infected WT and Il1r1-/- femurs had no significant differences in cortical bone loss or reactive bone formation ( Fig 6H–6J ) . Collectively , these data indicate that during S . aureus osteomyelitis , Il1r1-/- mice exhibit significantly altered cortical bone remodeling , with increased reactive bone formation , altered callus architecture , and greater cortical bone loss at the site of infection . To elucidate the cellular changes driving differences in bone remodeling between WT and Il1r1-/- mice , we next analyzed trabecular bone remodeling during osteomyelitis . Histomorphometric analysis of trabecular bone was performed in both S . aureus infected femurs and contralateral , uninfected femurs from each genotype . Histomorphometry revealed that the infected femurs from WT mice had significantly lower trabecular BV/TV than contralateral , uninfected femurs ( Fig 7A ) . In contrast , infected femurs from Il1r1-/- mice showed no significant differences in BV/TV in comparison to the contralateral , uninfected femur , suggesting that these mice were protected from infection-associated trabecular bone loss despite having significantly higher bacterial burdens in the regions encompassing the trabecular bone over time ( Fig 7A; S2 Fig ) . To determine if differences in osteoclast biology might underlie the distinct trabecular bone remodeling parameters of WT and Il1r1-/- mice , we calculated the numbers of osteoclasts present on trabecular bone surfaces in both infected and contralateral , uninfected femurs . The infected femurs in WT mice displayed greater osteoclast numbers per bone perimeter ( N . Oc/B . pm ) and osteoclast surface per bone surface ( Oc . S/BS ) compared to the contralateral , uninfected femurs , correlating with the infection-induced loss of trabecular bone volume ( Fig 7A–7C ) . In contrast , the infected femurs from Il1r1-/- mice showed no increase in N . Oc/B . pm or Oc . S/BS when compared to the contralateral , uninfected femur ( Fig 7B and 7C ) . These data suggest that S . aureus infection causes enhanced osteoclastogenesis in trabecular bone , which is dependent on intact IL-1R signaling and contributes to bone loss . Histomorphometric analysis revealed that S . aureus infection enhances osteoclastogenesis and trabecular bone loss in an IL-1R-dependent manner . However , bone volume and remodeling are also significantly impacted by osteoblast function . To determine the contribution of osteoblasts toward altered bone homeostasis and trabecular bone loss during staphylococcal osteomyelitis , we measured bone mineralization in the trabecular bone of infected WT and Il1r1-/- mice . No differences were observed in mineralizing surface , bone formation rate , or mineral apposition rate between WT and Il1r1-/- mice ( S7A–S7C Fig ) . These data indicate that IL-1R signaling does not drive differences in trabecular osteoblastic function during S . aureus infection , and that the decrease in trabecular BV/TV is not a function of decreased osteoblastic bone formation . Staphylococcal infection causes bone loss and enhanced osteoclastogenesis in trabecular bone . Accordingly , we hypothesized that secreted bacterial factors might augment osteoclast differentiation . To test this hypothesis , we measured osteoclast differentiation of RANKL-primed myeloid progenitors after stimulation with S . aureus culture supernatant or a vehicle control . To avoid induction of cell death in myeloid cells , we used a S . aureus strain lacking the alpha-type PSMs , which we previously demonstrated are both necessary and sufficient for causing cell death when staphylococcal supernatants are applied to murine bone marrow-derived macrophages ( BMMs ) [6 , 7] . Stimulation of RANKL-primed BMMs with toxin-deficient supernatant resulted in a dramatic increase in mature osteoclasts , as identified as tartrate resistant acid phosphatase positive ( TRAP+ ) multinucleated cells , relative to vehicle control in WT cells ( Fig 8A ) . To determine if this bacterial enhancement of osteoclastogenesis was dependent on IL-1R signaling , we performed similar experiments with Myd88-/- and Il1r1-/- BMMs . We discovered that Myd88-/- and Il1r1-/- cells do not undergo robust S . aureus-mediated osteoclastogenesis , despite being able to differentiate into osteoclasts via canonical RANKL stimulation ( Fig 8A , S8A–S8D Fig ) . Moreover , toxin-deficient S . aureus supernatants caused significantly less osteoclast formation in Myd88-/- and Il1r1-/- cells in comparison to WT cells ( Fig 8A ) . We next tested the role of IL-1 blockade on osteoclastogenesis using WT or Il1r1-/- cells , with or without IL-1R antagonist ( IL-1ra ) treatment . RANKL pre-commitment of WT cells with simultaneous IL-1ra treatment blunted the generation of osteoclast precursors , leading to 50% fewer S . aureus-stimulated osteoclasts ( Fig 8B ) . This decline in S . aureus-enhanced osteoclastogenesis results in differentiation to a similar level as is observed in Il1r1-/- osteoclast precursors ( Fig 8B ) . These observations indicate that MyD88 and the IL-1R are required for S . aureus-mediated enhancement of osteoclastogenesis . Therefore , although MyD88 and IL-1R are critical mediators of the anti-staphylococcal immune response , S . aureus infection also elicits osteoclast-mediated bone loss through MyD88 and IL-1R signaling pathways .
Bacterial osteomyelitis is a debilitating invasive infection of bone that is accompanied by significant damage to skeletal tissues and the surrounding vasculature . Using a model of post-traumatic S . aureus osteomyelitis , we have detailed dramatic architectural and cellular bone remodeling alterations that accompany S . aureus infection . In prior research , we determined that some of the cortical bone loss observed during infection is due to psm- and agr-dependent mechanisms [6] . This is in keeping with the findings of Gillaspy et al . , who used a rabbit model of osteomyelitis and observed significantly less bone pathology when infecting with an agr mutant [11] . However , the observation of residual bone pathology in mice infected with an agr mutant in our osteomyelitis model led us to postulate that host responses to bacterial infection may also contribute to bone loss . The focus of this work was therefore to delineate critical host responses to staphylococci in bone and to elucidate how an innate immune response might impact bone homeostasis . In this study , we focused primarily on MyD88 and IL-1R signaling cascades given their established roles in innate immune responses against S . aureus in other models of infection [27–35] , in concert with the known effects of these signaling pathways on bone cell function [19–22] . Additionally , patients with single nucleotide polymorphisms ( SNPs ) in Myd88 , Il1a , and Il1r1 genes have an increased risk of osteomyelitis and inflammatory joint disorders , further underscoring the importance of these immune pathways in skeletal homeostasis [52–57] . A robust innate immune response was observed in S . aureus infected femurs , with abundant levels of pro-inflammatory cytokines IL-1α , IL-1β , IL-6 , and TNFα detectable as soon as one day after infection , similar to what has been reported in other musculoskeletal infection models [5 , 16] . Many cytokines with increased abundance in infected versus mock infected femurs are encoded by IL-1 target genes , including IL-1α , IL-1β , IL-6 , MCP-1 ( CCL2 ) , and the murine IL-8 homologs , KC ( CXCL1 ) and MIP-2 ( CXCL2 ) [58 , 59] . In turn , IL-1 cytokines and IL-6 also promote the release of IL-17 and subsequent G-CSF production , both of which were highly abundant in S . aureus infected femurs [60] . These data suggest a role for IL-1 signaling in orchestrating downstream inflammatory responses to pathogens in bone . Growth factors and chemokines that support myeloid cell influx and expansion after infection , including M-CSF , G-CSF , MCP-1 ( CCL2 ) , MIP-1α ( CCL3 ) , MIP-1β ( CCL4 ) , KC ( CXCL1 ) and MIP-2 ( CXCL2 ) , are also highly abundant in S . aureus infected femurs relative to mock infected femurs . This observation parallels other reports demonstrating increased levels of myeloid chemokines from osteoblasts after S . aureus infection , and a role of myeloid chemokines and growth factors in supporting osteoclastogenic bone degradation by promoting expansion of osteoclast precursor cells [61 , 62] . Moreover , many of these chemokines have been shown to coordinate neutrophil responses during acute inflammation [63] . These data partially overlap with early cytokine signatures measured in a pin prosthetic implant model of S . aureus biofilm infection , with IL-1β , IL-6 , TNFα , IL-12p70 , and IL-17 detected in infected tissues . However local inflammation during biofilm infection was also characterized by increased abundance of IL-2 [64] , which was not significantly elevated in our infection model . However , the development of a biofilm has been shown to attenuate the host pro-inflammatory response in a catheter S . aureus biofilm model , as measured by decreased levels of IL-1β , TNFα , CXCL2 , and CCL2 [65] . Future studies should continue to delineate how implant-associated biofilms skew the immune response during osteomyelitis . The early increase in cardinal pro-inflammatory cytokines after S . aureus infection indicates that the post-traumatic model of osteomyelitis used in this study is most representative of acute osteomyelitis . However , the bone pathology visualized by day 14 post-infection has clear features suggestive of chronic infection , including significant reactive bone formation and the presence of sequestra [66] . Consistent with a possible shift from acute to chronic infection , we observed production of IFNγ and IL-17 at later time points post-infection , which could represent S . aureus-specific adaptive Th1/Th17 responses [64] . Delineating the cytokine milieu at later time points after infection will help to more comprehensively characterize the inflammation accompanying osteomyelitis in this model . Based on the robust early inflammatory responses to S . aureus in bone coupled with the detection of multiple IL-1 associated cytokines , we focused on the role of MyD88 and IL-1R signaling in coordinating antibacterial defenses during osteomyelitis . Furthermore , since several cardinal pro-inflammatory cytokines , including IL-1 , have direct effects on skeletal cell differentiation and function , we hypothesized that MyD88 and IL-1R-dependent signaling pathways would be necessary for control of bacterial proliferation during osteomyelitis , but that these same pathways might also contribute to pathogen-induced bone loss through actions on skeletal cells . To determine the contribution of MyD88 and IL-1R signaling towards antibacterial immune responses during S . aureus osteomyelitis , we infected mice globally deficient in MyD88 , IL-1R , IL-1α , and IL-1β and measured bacterial burdens and morbidity . We found that MyD88 and IL-1R signaling are required to control bacterial burdens in bone in mice infected with a high dose inoculum . Furthermore , even with a lower S . aureus inoculum , a timecourse experiment revealed that Il1r1-/- mice had a delay in ability to control bacterial burdens . In order to determine the relative contributions of IL-1 isoforms to IL-1R-mediated antibacterial immunity in bone , we infected Il1a-/- or Il1b-/- mice and compared bacterial burdens to those observed in WT and Il1r1-/- mice . Mice lacking either IL-1α or IL-1β sustained bacterial burdens not statistically different from bacterial burdens harbored by WT mice , but Il1a-/- and Il1b-/- mice both harbored significantly less bacterial CFUs than Il1r1-/- mice . These data suggest that the loss of both cytokines may be required to recapitulate the enhanced bacterial burdens in Il1r1-/- mice . Moreover , the extensive repertoire of innate receptors that signal through MyD88 ( e . g . TLRs ) likely promote a more effective antibacterial response to S . aureus , as MyD88 signaling is critical to prevent disseminated disease and death during osteomyelitis . Our findings that MyD88 and IL-1R mediate antibacterial protection in bone are consistent with data from previous studies demonstrating that Myd88-/- and Il1r1-/- mice have enhanced susceptibility to bacterial infection in various experimental models [27 , 43 , 44 , 67] . Several studies have also reported that IL-1R signaling contributes to the early influx of neutrophils and abscess formation to protect against S . aureus cutaneous and prosthetic joint infection [27 , 29–31] . Elevated levels of GM-CSF , G-CSF , and CXCL1 in WT mice suggest that neutrophil influx and/or expansion is also a critical early response to S . aureus in bone to prevent continued bacterial replication and spread . Interestingly , the neutrophilic cytokine response was delayed in Il1r1-/- mice in response to S . aureus , which in congruent with other Il1r1-/- mouse models in response to inflammatory stimuli [38–40] . Disorganized abscess architecture in Il1r1-/- infected femurs may indicate improper neutrophil mobilization without IL-1R signaling as an underlying mechanism for the inability to control bacterial burdens . Furthermore , we have previously shown that neutrophil depletion leads to significantly increased bacterial burdens during S . aureus osteomyelitis [7] . Here , we determined that relative neutrophil abundance was lower in Il1r1-/- mice at early and late time points in the infected femurs , suggesting that Il1r1-/- mice have altered systemic neutrophil responses that may correlate with alterations in G-CSF and GM-CSF data . Lower amounts of neutrophils in both the infected and contralateral , uninfected femurs of Il1r1-/- mice relative to WT mice support prior observations that Il1r1-/- mice have a defect in granulopoiesis [38–40] . Together these reports detail the importance of IL-1R signaling to protect against S . aureus bone infections by coordinating an effective anti-staphylococcal neutrophil response . The inability of Myd88-/- and Il1r1-/- mice to mount appropriate anti-staphylococcal immune responses and the characteristic differences in bone remodeling between WT and Il1r1-/- mice led us to confirm these phenotypes with littermate controls bred heterozygously . Contradictory reports have detailed either no difference in bone mass of Il1r1-/- mice [68] , low bone mass in Il1r1-/- mice [69] , or greater bone mass in Il1r1-/- mice [70 , 71] when compared to WT comparators . These studies used various WT comparators ( 129/J , 129/Sv , BALB/cA , C57BL/6 ) , and also varied with respect to the assessment of mouse age and gender . Additionally , Myd88-/- mice have altered intestinal barrier function and differences in the microbiome , which can lead to differences in immune function and bone mass [45 , 46 , 48–51] . Therefore , breeding of heterozygous colonies allowed us to reduce the confounding influence of mouse genotype and microbiome effects to confirm the importance of MyD88 and IL-1R signaling in antibacterial responses and bone remodeling . To investigate the mechanisms by which S . aureus alters bone homeostasis to incite bone destruction and reactive bone formation , we measured cortical and trabecular changes in bone architecture via μCT , quantified changes in skeletal cell function and activity in vivo using standard bone histomorphometry , and cultured skeletal cells in vitro to determine how S . aureus influences skeletal cell differentiation . These data reveal that S . aureus osteomyelitis induces changes in bone turnover both locally at the inoculation site , as well as in more distal areas not grossly impacted by abscess formation , leading to the significant loss of cortical and trabecular bone . Il1r1-/- mice exhibited more dramatic cortical bone changes , which may be due to differences in bone remodeling processes between WT and Il1r1-/- mice , the fact that Il1r1-/- mice harbor increased bacterial burdens over the duration of infection , or a combination of these factors . Previous studies have shown that the loss of IL-1R and MyD88 signaling enhances healing of sterile bone defects [72] , which may explain the enhanced volume of reactive callus formed on the cortical bone of Il1r1-/- mice in our findings . Although we did not observe significant differences in osteoblast-mediated bone parameters in trabecular bone , it is possible that there are significant differences in osteoblast or pre-osteoblast differentiation and function in healing cortical bone ( callus ) . Trabecular bone is the major site of homeostatic bone remodeling [41] , and bone loss here is thought to be multifactorial with potent contributions from inflammation , altered skeletal cell differentiation , and direct interaction with bacterial cells . With respect to the latter mechanisms , we detected viable S . aureus in the regions of the femur encompassing trabecular bone throughout the course of infection . S . aureus infection enhanced the number of osteoclasts as well as the actively resorbing trabecular bone surface during osteomyelitis in WT mice , thereby corroborating previous observations of enhanced osteoclast surface from S . aureus infected human bone biopsies [73] . This may reflect , in part , direct interactions with S . aureus protein A which induces osteoclastogenesis through TNFR1 and EGFR [10] . However , other bacterial factors that directly enhance osteoclastogenesis in vivo remain to be determined . In contrast , Il1r1-/- mice were protected from enhanced osteoclastogenesis and trabecular bone loss . Excitingly , although Il1r1-/- mice harbored higher bacterial burdens throughout the course of infection , they did not exhibit trabecular bone loss or increased osteoclastogenesis relative to contralateral , uninfected femurs . Taken together , these in vivo bone remodeling data indicate that S . aureus osteomyelitis enhances osteoclastogenesis and triggers trabecular bone loss in WT mice , mainly through IL-1R-dependent effects on osteoclasts . In vitro osteoclast differentiation assays further supported the observation of increased osteoclastogenesis in response to S . aureus in vivo , as staphylococcal supernatants significantly enhanced osteoclast formation from RANKL-primed WT precursor cells . These data corroborate other reports demonstrating that infection of host cells in vitro with live S . aureus enhances osteoclastogenesis and bone resorbing activity [74] , and our in vivo data now provide evidence that this enhanced osteoclastogenesis translates to bone loss during infection . Mechanistically , genetic deletion of Myd88 and Il1r1 and molecular inhibition of IL-1R signaling were found to confer resistance to S . aureus-enhanced osteoclastogenesis . Consistent with previously published reports , endogenous IL-1 has been described to promote osteoclastogenesis in vitro through synergistic signaling of the IL-1 and RANK receptors in the absence of infection [75] . Canonical osteoclast differentiation is initiated by RANK receptor signaling to activate the transcription factors NFATc1 and cFos , which in turn increase IL-1R expression . This allows IL-1 to signal on osteoclast precursors to potentiate osteoclast formation by activating osteoclast-specific genes , and IL-1 has been reported to enhance “pathologically activated osteoclasts” that favor bone loss [19–22 , 70 , 75 , 76] . In the context of infection , S . aureus and specific staphylococcal toxins have been found to stimulate the production of IL-1 cytokines [77–79] . Moreover , IL-1 cytokines have been described to promote osteoclastogenesis in vitro and lead to bone destruction in murine models of rheumatoid arthritis and autoinflammatory disorders [17 , 20 , 21 , 80 , 81] . Therefore , these data are consistent with other observations and suggest that IL-1 signals onto osteoclast precursors to enhance osteoclastogenesis and trabecular bone resorption during infection . However , residual osteoclast formation observed in Il1r1-/- cells suggests that while MyD88 is required for osteoclastogenesis in response to staphylococcal supernatant , there are both IL-1R-dependent and independent mechanisms involved . These in vitro studies support our findings that a major driver of bone loss during S . aureus osteomyelitis is coordinated by IL-1R-mediated osteoclast enhancement . Data presented in this study highlight MyD88 and IL-1R signaling as critical pathways supporting anti-staphylococcal immunity in bone , but also implicate these signaling cascades in promoting bone loss during osteomyelitis . There are a few limitations of the experimental approach outlined in this study . We used globally deficient knockout mice to elucidate how MyD88 and IL-1R signaling impact bone homeostasis and anti-staphylococcal immunity . Given that the adapter protein MyD88 is necessary to relay signals from other upstream receptors including TLRs , future studies should explore the relative contributions of other MyD88-dependent receptors in the pathogenesis of osteomyelitis . In certain S . aureus infection models , TLR2 and TLR9 have been shown to contribute to anti-staphylococcal immunity [28 , 44 , 82] . In vitro osteoclastogenesis assays imply that other MyD88-dependent receptors can sense and respond to components of S . aureus in culture to enhance osteoclastogenesis . Accordingly , osteoblast and osteoclast lineage cells are reportedly activated in vitro through TLRs and IL-1R signaling to favor bone resorption [18 , 83 , 84] . During staphylococcal osteomyelitis , it remains unclear how much pathogen-induced bone loss occurs as result of direct osteoclast stimulation versus indirect perturbations of bone homeostasis that involve osteoblasts . This could be tested using MyD88 skeletal cell lineage specific knockout mice . Furthermore , the IL-1R-expressing target cells that stimulate anti-staphylococcal immunity and the source and isoform of IL-1 that promotes bone loss remain unclear . Collectively , this study details the paradoxical roles of innate immune signaling pathways in the pathogenesis of osteomyelitis . Although MyD88 and IL-1R signaling elicit antibacterial responses during bone infection to protect against bacterial proliferation , dissemination , and systemic disease , they also contribute to host-mediated bone loss . Our findings also highlight a specific MyD88- and IL-1R-dependent mechanism of osteoclast enhancement , thereby uncovering a new mechanism for bone loss during S . aureus osteomyelitis .
All experiments involving animals were reviewed and approved by the Institutional Animal Care and Use Committee at Vanderbilt University Medical Center on the animal protocols M12059 and M1800055 . All experiments were performed according to NIH guidelines , the Animal Welfare Act , and US Federal law . The murine model of osteomyelitis required inhalational anesthesia with isoflurane ( 1–5% ) . Post-operative analgesia ( buprenorphine 0 . 5–0 . 1 mg/kg ) was provided pre-operatively and every 8–12 hours for 48 hours post-infection . Mice were euthanized by CO2 asphyxiation with secondary confirmation by cervical dislocation and observation of heart rate and breathing . C57BL/6J ( Stock #: 000664 ) , Myd88-/- ( Stock #: 009088 ) and Il1r1-/- ( Stock #: 003245 ) mice were purchased through The Jackson Laboratory . Il1a-/- and Il1b-/- mice were generated as described [85] . WT mice were bred with Myd88-/- or Il1r1-/- mice to produce Myd88+/- or Il1r1+/- mice , respectively . Heterozygous mice were bred together to create mice carrying knockout ( -/- ) , heterozygous ( +/- ) , or wild-type ( +/+ ) alleles for Myd88 or Il1r1 . Mice were bred heterozygously to reduce the confounding influence of microbiome effects associated with genotypes and maintenance of separate mouse colonies . The resulting littermates were earpunched and genotyped through Transnetyx , Inc . ( Cordova , TN ) . All infections were conducted with an erythromycin-sensitive derivative of the USA300 type S . aureus LAC clinical isolate ( AH1263 ) [86] . The toxin-deficient strain LACΔpsmα1–4 ( Δpsm ) has been previously described and was used for in vitro assays to prevent cell death [6 , 87] . Bacteria were routinely grown on tryptic soy agar ( TSA ) or shaking in tryptic soy broth ( TSB ) with or without 10 μg/mL erythromycin as detailed previously [6] . To prepare concentrated supernatants , Δpsm was grown overnight in RPMI supplemented with 1% casamino acids [7] . The murine model of osteomyelitis was performed as described previously [6 , 7] . AH1263 was sub-cultured from an overnight culture , grown for 3 hours , and then adjusted in PBS to a concentration of approximately 1x106 CFUs in 2 μl PBS , unless diluted 1:10 or 1:100 to deliver inoculum doses of 1x105 or 1x104 CFUs , respectively . Osteomyelitis was induced in 5- to 8-week old male and female mice following the introduction of a unicortical bone defect using a 21G needle , into which 2 μl of bacterial suspension or PBS ( mock infection ) was injected into the intermedullary canal . Muscle fascia and skin were closed with sutures and mice were given buprenorphine analgesic every 12 hours for 48 hours , with daily monitoring until the experimental end point . Mice were euthanized if they met human endpoints , including inability to ambulate , inability to eat or drink , loss of greater than 20% body weight , and/or hunched posture . Femurs were harvested 14 days post-infection and fixed for 48 hours in neutral buffered formalin at 4°C . Bones were scanned using a μCT50 ( Scanco Medical , Switzerland ) and analyzed with μCT Tomography V6 . 3–4 software ( Scanco USA , Inc . , Wayne , PA ) . To expand previous μCT50 analyses that assessed only the cortical bone of the femoral diaphysis [6] , here the diaphysis and distal epiphysis of each femur were visualized in the scout-view radiographs and imaged with 10 . 0 μm voxel size at 70 kV , 200 μA , and an integration time of 350 ms in a 10 . 24 mm view . Each imaging scan resulted in 1088 slices ( 10 . 88 mm ) of the femur that included the diaphysis surrounding the inoculation site , trabecular bone in the distal femur , and excluded the proximal epiphysis . Three-dimensional volumetric analyses were conducted by contouring transverse image slices in the region of interest . The diaphysis of each femur was comprised of 818 image slices . These image slices were used to quantify cortical bone destruction ( mm3 ) and reactive bone formation ( mm3 ) surrounding the cortical bone inoculation site as described previously [6] . Trabecular bone measurements were obtained in the distal femur by advancing proximally past the growth plate 30 slices . 101 slices were analyzed with an inclusive contour drawn along the endosteal surface to include trabeculae and exclude the cortical bone . Trabecular bone volume per total volume ( % ) , trabecular number ( 1/mm ) , trabecular thickness ( mm ) , and trabecular spacing ( mm ) were determined by segmentation of the image with a lower threshold of 329 mg HA/ccm , sigma 1 . 3 , and support 1 . After μCT imaging , femurs were decalcified for three days in 20% EDTA at 4°C . Decalcified bones were processed and embedded in paraffin before sectioning at 4μm thickness through the infectious nidus and bone marrow cavity using a Leica RM2255 microtome . Sectioned femurs were stained with a modified hematoxylin and eosin ( H&E ) that included orange G and phloxine for enhanced bone contrast , tartrate-resistant acid phosphatase ( TRAP ) stain with hematoxylin counterstain , or 3 , 3’-diaminobenzidine ( DAB ) immunohistochemistry to detect myeloperoxidase ( MPO ) . OsteoMeasure software ( OsteoMetrics , Inc . , Decatur , GA ) was used to manually analyze TRAP-stained histologic sections at a region of interest encompassing the trabeculae proximal to the growth plate in the distal femur . Osteoclast number , osteoclast surface , and bone perimeter were calculated and reported per ASBMR standards [42] . A Leica SCN400 Slide Scanner was used to scan stained femur sections in brightfield at 20X . Images were uploaded to and imaged with the Digital Imaging Hub ( Leica Biosystems , Buffalo Grove , IL ) and Tissue Image Analysis 2 . 0 ( Tissue IA 2 . 0 ) ( Leica Microsystems , Buffalo Grove , IL ) was used to analyze callus area of infected femurs at 20X . WT and Il1r1-/- mice were intraperitoneally injected with 20 mg/kg calcein on days 8 and 12 post-infection with 105 CFUs . Femurs were subsequently harvested , formalin fixed , and dehydrated prior to embedding in poly ( methyl methacrylate ) for sectioning , and counterstained with toluidine blue . Fluorescent labels were identified as single- or double-labeled surface . Fluorescent labels and trabecular bone were traced in the distal femur using OsteoMeasure software , and the mineralizing surface per bone surface ( MS/BS ) , mineral apposition rate ( MAR ) , and bone formation rate per bone surface ( BFR/BS ) were calculated per ASBMR standards [42] . At various time points post-infection , tissues were harvested and homogenized using a BulletBlender and NAVY lysis tubes ( Next Advance , Inc . , Averill Park , NY ) at 4°C . To enumerate bacterial CFUs in infected femurs , the whole femur or only the regions encompassing the trabecular bone ( i . e . metaphyses and epiphyses ) were homogenized in PBS . To maximize cytokine signals in femur homogenates , CelLytic Buffer MT Cell Lysis Reagent ( Sigma , Saint Louis , MO ) was substituted for PBS to specifically lyse mammalian cells . Livers and kidneys were homogenized in PBS . Femur and organ homogenates were vortexed , serially diluted in PBS , and plated on TSA for bacterial enumeration . Femur homogenates lysed in CelLytic Buffer MT and PBS showed no difference in recoverable bacterial burdens . Following homogenization , femur homogenates were centrifuged at 4000 x g for 5 minutes to remove debris and the supernatant was stored at -80°C for subsequent analysis using Milliplex MAP multiplex magnetic bead-based antibody detection kits ( EMD Millipore , Billerica , MA ) according to the manufacturer’s protocols . Cytokine quantification from bone homogenates was performed using the 32-plex Mouse Cytokine/Chemokine Magnetic Bead Panel ( MCYTMAG-70K-PX32 ) on the FLEXMAP 3D instrument . The quality controls for IL-13 failed , and these data were excluded . Cytokine levels from femurs homogenized in 500 μl volume were read as pg/mL . Femur homogenates reported as relative values were homogenized in PBS , whereas cytokines values corrected for total protein were homogenized in CelLytic Buffer MT to maximize cytokine signals . Total protein ( mg/mL ) was quantified using the Pierce BCA Protein Assay Kit ( ThermoFisher Scientific , Waltham , MA ) per manufacturer’s directions . Infected Il1r1-/- femurs were up to two times larger than WT infected femurs and four times larger than mock infected WT and Il1r1-/- femurs . Cytokine levels are therefore reported as pg cytokine/mg protein to control for femur size differences between infected Il1r1-/- and WT femurs . Following S . aureus infection ( 105 CFUs ) , femurs from WT and Il1r1-/- mice were harvested at 1 , 3 , 5 , and 14 days post-infection . Whole bone marrow ( WBM ) was flushed through a 70 μm nylon cell strainer ( Falcon , Corning , New York ) and red blood cells ( RBCs ) were lysed using the Ammonium Chloride Potassium ( ACK ) Lysing Buffer ( Lonza , Walkersville , MD ) . Bone marrow ( BM ) mononuclear cells were counted and 1 million cells were plated per well and washed in PBS supplemented with 3% FBS and 0 . 1% sodium azide ( FACS buffer ) . Cells were incubated with Anti-CD16/CD32 ( Biolegend , 1:100 , clone 93 , San Diego , CA ) to block non-specific antibody staining . Cells were then incubated with a mixture of murine-specific cell surface antibodies on ice , including Anti-Ly6G-PE ( Biolegend , 1:3200 , clone 1A8 ) , Anti-Ly6C-PE-Dazzle 594 ( Biolegend , 1:1600 , clone HK1 . 4 ) , Anti-CD68-PE-Cy7 ( Biolegend , 1:100 , clone FA-11 ) , Anti-CD11b-APC ( Tonbo 1:4800 , clone M1/70 , San Diego , CA ) , and Anti-CD45-AlexaFluor 700 ( Biolegend , 1:400 , clone 30-F11 ) . Cells were washed two times in FACS buffer , resuspended in 2% paraformaldehyde solution , and run on a 3-laser BD LSRII flow cytometer the following day . Single BM cells were identified from successive gates , including side scatter-area by forward scatter-area ( SSC-A x FSC-A ) , forward scatter area by height ( FSC-A x FSC-H ) , and side scatter area by height ( SSC-A x SSC-H ) . Next , CD45+ cells , CD11b+ cells , and Ly6G+LyClo cells ( neutrophils ) were gated sequentially . WBM was flushed from femurs of 8- to 13-week old male mice using α-MEM media . Following RBC lysis , WBM was resuspended in a 90% FBS and 10% DMSO solution and frozen in liquid nitrogen until thawed for use . BMMs were enriched by plating 8 to 13 million cells per 10 cm dish in α-MEM , 10% FBS , 1X Penicillin/Streptomycin ( P/S ) , and 100 ng/mL recombinant murine M-CSF ( R&D Systems , Minneapolis , MN , 416-ML ) for 4 days . Non-adherent cells were removed , and adherent cells were washed with PBS , scraped into fresh media , and counted prior to plating . Enriched BMMs were plated at a density of 50 , 000 cells/well in 96-well plates , and media ( α-MEM , 10% FBS , 1X P/S ) was supplemented 1:20 with CMG14-12 as an M-CSF source [88] . Osteoclastogenesis assays were performed with RANKL-primed osteoclast precursors , which were generated by plating BMMs in 35 ng/mL recombinant murine RANKL ( R&D Systems , Minneapolis , MN , 462-TR ) for 2 days . Prior to stimulation , RANKL-primed osteoclast precursors were washed twice with PBS . RANKL-primed osteoclast precursors were stimulated with either a vehicle control ( 1% casamino acid-supplemented RPMI ) or Δpsm supernatant . M-CSF was supplemented into the media containing each stimulation . To test the specific role of IL-1R inhibition on S . aureus-enhanced osteoclast differentiation , osteoclastogenesis assays in WT and Il1r1-/- cells were conducted with the addition of a vehicle control ( 0 . 1% low endotoxin BSA ) or 1 μg/mL recombinant murine IL-1ra ( Novus Biologicals , LLC , Littleton , CO , NBP2-35105 ) during the 2 days of RANKL pre-commitment or during the 4 days of Δpsm supernatant stimulation . On day 6 in culture , all RPMI- and S . aureus-stimulated osteoclastogenesis assays were fixed with a 4% formaldehyde and 0 . 05% Triton X-100 solution in PBS ( 10 minutes ) and 1:1 acetone:ethanol ( 1 minute ) , before TRAP staining with reagents from the Acid Phosphatase , Leukocyte ( TRAP ) Kit ( Sigma , Saint Louis , MO , 378A ) . In control osteoclastogenesis assays without S . aureus supernatant stimulation , cells were stimulated at the time of plating with 1:20 CMG14-12 and 35ng/mL RANKL . Fresh media , CMG14-12 , and RANKL were replenished on days 4 and 6 in culture , with cells TRAP stained on day 7 . OsteoMeasure was used to manually quantify mature osteoclasts , identified as TRAP+ multinucleated cells . Data analysis and statistical tests were conducted using Graph Pad Prism software . Unpaired t-tests were used to compare CFU burdens , measurements of bone architecture using μCT and histology , cytokine levels , neutrophil abundance , and TRAP+ multinucleated cell counts when two groups were being compared . Log-rank Mantel Cox tests compared survival curves between genotypes for each S . aureus inoculum . A one-way ANOVA with Tukey’s multiple comparison test compared CFU burdens from femurs between multiple genotypes . A two-way ANOVA was used with Fisher’s Least Significant Difference ( LSD ) test to compare the effects of genotype and infection status between histomorphometry measurements . A repeated measures two-way ANOVA with Tukey’s multiple comparisons test was used to compare TRAP+ multinucleated cell counts between genotype at each Δpsm supernatant dose . Repeated measures two-way ANOVAs with Dunnett’s multiple comparisons test were conducted on TRAP+ cell counts from each genotype , to compare Δpsm supernatant dosage effects . A three-way ANOVA with Tukey’s multiple comparisons test was used to compare cell genotype , IL-1ra pre-treatment , and IL-1ra treatment alongside Δpsm supernatant stimulation . P values of less than 0 . 05 were considered statistically significant . Details on number of data points , experimental replicates , calculated standard deviation , and statistical significance for each experiment are described in figure legends . | Osteomyelitis is a common , debilitating infection of bone that rarely resolves without prolonged antibiotics and invasive surgical procedures . This study explores the role of host inflammation during osteomyelitis . Our findings highlight innate immune responses that are critical for control of S . aureus burdens in bone , prevention of bacterial dissemination , and death . Conversely , these same immune pathways were found to contribute to disease pathogenesis by activating infection-associated bone loss . Such bone loss is associated with detrimental outcomes during osteomyelitis , including pathologic fractures and irreversible changes in bone growth plates . Alterations in bone remodeling that occur during S . aureus osteomyelitis are multifaceted , driven by bacterial toxins and inflammation-mediated changes . We found that inflammation incited by S . aureus in bone contributes to pathologic bone loss . We highlight a mechanism by which S . aureus stimulates formation of bone-resorbing osteoclasts both in vitro and in vivo . Our data illustrate connections between innate immune signaling pathways and bone homeostasis . These connections may be relevant to other autoinflammatory bone and joint conditions characterized by bone loss , such as rheumatoid arthritis . Collectively , this work outlines the fine balance between promotion of bacterial clearance and protection from collateral tissue damage . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [] | 2019 | MyD88 and IL-1R signaling drive antibacterial immunity and osteoclast-driven bone loss during Staphylococcus aureus osteomyelitis |
Human African trypanosomiasis ( HAT ) , a parasitic protozoal disease , is caused primarily by two subspecies of Trypanosoma brucei . HAT is a re-emerging disease and currently threatens millions of people in sub-Saharan Africa . Many affected people live in remote areas with limited access to health services and , therefore , rely on traditional herbal medicines for treatment . A molecular docking study has been carried out on phytochemical agents that have been previously isolated and characterized from Nigerian medicinal plants , either known to be used ethnopharmacologically to treat parasitic infections or known to have in-vitro antitrypanosomal activity . A total of 386 compounds from 19 species of medicinal plants were investigated using in-silico molecular docking with validated Trypanosoma brucei protein targets that were available from the Protein Data Bank ( PDB ) : Adenosine kinase ( TbAK ) , pteridine reductase 1 ( TbPTR1 ) , dihydrofolate reductase ( TbDHFR ) , trypanothione reductase ( TbTR ) , cathepsin B ( TbCatB ) , heat shock protein 90 ( TbHSP90 ) , sterol 14α-demethylase ( TbCYP51 ) , nucleoside hydrolase ( TbNH ) , triose phosphate isomerase ( TbTIM ) , nucleoside 2-deoxyribosyltransferase ( TbNDRT ) , UDP-galactose 4′ epimerase ( TbUDPGE ) , and ornithine decarboxylase ( TbODC ) . This study revealed that triterpenoid and steroid ligands were largely selective for sterol 14α-demethylase; anthraquinones , xanthones , and berberine alkaloids docked strongly to pteridine reductase 1 ( TbPTR1 ) ; chromenes , pyrazole and pyridine alkaloids preferred docking to triose phosphate isomerase ( TbTIM ) ; and numerous indole alkaloids showed notable docking energies with UDP-galactose 4′ epimerase ( TbUDPGE ) . Polyphenolic compounds such as flavonoid gallates or flavonoid glycosides tended to be promiscuous docking agents , giving strong docking energies with most proteins . This in-silico molecular docking study has identified potential biomolecular targets of phytochemical components of antitrypanosomal plants and has determined which phytochemical classes and structural manifolds likely target trypanosomal enzymes . The results could provide the framework for synthetic modification of bioactive phytochemicals , de novo synthesis of structural motifs , and lead to further phytochemical investigations .
Human African trypanosomiasis ( HAT ) , also known as sleeping sickness , is caused by the single-celled kinetoplastid parasites , Trypanosoma brucei , which are transmitted to humans by infected tsetse flies ( Glossina spp . ) . Two sub-species of T . brucei ( rhodesiense and gambiense ) cause the two different forms of the disease . T . b . rhodesiense is found in southern and eastern Africa while T . b . gambiense is found in the western , central and some parts of eastern Africa . T . b . gambiense now accounts for about 90% of all reported cases of sleeping sickness . A third subspecies , T . b . brucei , does not cause HAT because of its susceptibility to lysis by human apolipoprotein L1 [1] . Current chemotherapies of HAT are directed either to the early or late stages of the disease . All the clinically available HAT chemotherapeutic drugs have been noted to be ineffective , and they also have severe side-effects . The only drug candidate in clinical trials for the treatment of HAT is the nitroimidazole fexinidazole . Fexinidazole is currently in clinical study for the treatment of the late stage form of HAT [2] , [3] . It is worth noting that the number of reported cases of HAT fell in the past decade , and it has also been suggested that a possible elimination of the disease might be in sight [4] . This is a very delightful development for this “neglected” tropical disease , and it is our hope that continued research into new and effective chemotherapy against HAT remains an integral part of public health initiatives in endemic communities . Medicinal plants from Nigeria's lush rainforest , as well as her very diverse montane and savanna vegetation , continue to play a vital role in her healthcare system . For tens of millions of Nigerians , indigenous traditional medicine is the major – and sometimes the only – access to pharmacological agents [5] . There have been several published reports on the biological activity of Nigerian plants , but most of the bioactive components of those plants have not been characterized . However , the country's big and loosely-regulated traditional medicine industry continues to promote the efficacy of extracts and concoctions made from most of the plants . A number of Nigerian plants have been used traditionally in West Africa to treat protozoal infections and many of these have shown in-vitro antiprotozoal activity ( Table S1 ) . Several T . brucei protein targets have been identified and experimentally validated [6] . In addition to validated targets , several potential targets have been predicted in silico [7] . For a recent review of phytochemical agents that show activities against parasitic protozoans and protozoan biochemical targets , see [8] , [9] . Some of the potential T . brucei drug targets that we considered in this work include adenosine kinase [10] , pteridine reductase 1 [11] , dihydrofolate reductase [12] , trypanothione reductase [13] , cathepsin B [14] , heat shock protein 90 [15] , as well as sterol 14α-demethylase ( CYP51 ) [16] , nucleoside hydrolase [17] , triose phosphate isomerase [18] , nucleoside 2-deoxyribosyltransferase [19] , UDP-galactose 4′ epimerase [20] and ornithine decarboxylase [21] . In this computational study , we have evaluated the interaction of compounds that were isolated from some antitrypanosomal Nigerian medicinal plants ( Table S1 ) against potential protein drug targets in Trypanosoma brucei for which X-ray crystal structures were available from the Protein Data Bank ( PDB ) . We strove to address the questions of which phytochemical agents might be responsible for the observed antitrypanosomal activity and what are the likely targets of those phytochemicals . In doing so , we hope to identify particular classes of phytochemical agents that can be exploited for antiparasitic chemotherapy .
Protein-ligand docking studies were carried out based on the crystal structures of rhodesain ( PDB 2p7u , [22] and PDB 2p86 [23] ) , T . brucei adenosine kinase , TbAK ( PDB 2xtb and PDB 3otx [24] ) , T . brucei pteridine reductase 1 , TbPTR1 ( PDB 3jq7 [25] ) , T . brucei dihydrofolate reductase , TbDHFR ( PDB 3rg9 and PDB 3qfx [26] ) , T . brucei trypanothione reductase , TbTR ( PDB 2wow , [27] ) , T . brucei cathepsin B , TbCatB ( PDB 3hhi [28] ) , T . brucei heat shock protein 90 , TbHSP90 ( PDB 3omu [29] and PDB 3opd [30] ) , T . brucei sterol 14α-demethylase , TbCYP51 ( PDB 3gw9 [16] ) , T . brucei nucleoside hydrolase , TbNH ( PDB 3fz0 [31] ) , T . brucei triosephosphate isomerase , TbTIM ( PDB 1iih , PDB 6tim [32] , and PDB 4tim [33] ) , T . brucei nucleoside 2-deoxyribosyltransferase , TbNDRT ( PDB 2a0k , PDB 2f64 , and PDB 2f67 [19] ) , T . brucei UDP-galactose 4′-epimerase , TbUDPGE ( PDB 1gy8 [20] ) , and T . brucei ornithine decarboxylase , TbODC ( PDB 1f3t [34] , PDB 1njj [35] , and PDB 1qu4 [21] ) . All solvent molecules and the co-crystallized ligands were removed from the structures . Molecular docking calculations for all compounds with each of the proteins were undertaken using Molegro Virtual Docker v . 4 . 3 [36] , [37] , with a sphere large enough to accommodate the cavity centered on the binding sites of each protein structure in order to allow each ligand to search . If a co-crystallized inhibitor or substrate was present in the structure , then that site was chosen as the binding site . If no co-crystallized ligand was present , then suitably sized cavities were used as potential binding sites . Standard protonation states of the proteins based on neutral pH were used in the docking studies . The protein was used as a rigid model structure; no relaxation of the protein was performed . Assignments of charges on each protein were based on standard templates as part of the Molegro Virtual Docker program; no other charges were set . Each ligand structure was built using Spartan '08 for Windows [38] . The structures were geometry optimized using the MMFF force field [39] . Flexible ligand models were used in the docking and subsequent optimization scheme . As a test of docking accuracy and for docking energy comparison , co-crystallized ligands were re-docked into the protein structures . Different orientations of the ligands were searched and ranked based on their energy scores . The RMSD threshold for multiple cluster poses was set at <1 . 00 Å . The docking algorithm was set at maximum iterations of 1500 with a simplex evolution population size of 50 and a minimum of 30 runs for each ligand . Each binding site of oligomeric structures was searched with each ligand . The lowest-energy ( strongest-docking ) poses for each ligand in each protein target are summarized in Tables S2–S20 .
Phytochemical studies of Acacia nilotica [40]–[45] have shown an abundance of polyphenolic compounds ( Table S2 ) , including hydrolyzable tannins , flavonoid gallates , and flavonoid glycosides . Although these polyphenolics are notorious for being promiscuous protein complexing agents and they do show relatively strong docking to all proteins investigated in this study , some selectivity can be seen . Thus , for example , 1 , 3-digalloylglucose showed docking selectivity for TbUDPGE , 3′ , 5-digalloylcatechin was selective for TbAK , and 3′ , 7-digalloylcatechin selectively docked with TbNH and was the strongest binding ligand for that protein ( −44 . 2 kcal/mol ) . 5 , 7-Digalloylcatechin was the strongest binding ligand for TbPTR1 ( −42 . 7 kcal/mol ) and 4′ , 7-digalloylcatechin was the strongest binding ligand for TbODC ( −41 . 4 kcal/mol ) . A number of these polyphenolic ligands showed strong docking interactions with TbAK , TbPTR1 , TbCYP51 , TbNH , and TbUDPGE , and interactions with these protein targets may be responsible for the antitrypanosomal activity of A . nilotica [46] . The docking study suggests that rhodesain , TbDHFR , TbTR , TbCatB , and TbHSP90 are not targets for A . nilotica phytochemicals . Ageratum conyzoides extracts have been dominated by flavonoids and chromanes ( Table S3 ) [40] , [47]–[49] . 5 , 6-Dimethoxy-2-isopropylbenzofuran , 6 , 7-dimethoxy-2-methyl-2- ( 2-methyl-1-propanone ) -3-chromene , 6-acetyl-2 , 2-dimethylchroman , and O-methylenececalinol exhibited selectivity for TbTIM with docking energies comparable to the co-crystallized ligand , 3-phosphoglyceric acid ( −21 . 6 kcal/mol ) . The flavonoid 3′ , 4′ , 5 , 5′ , 6 , 8-hexamethoxyflavone , on the other hand , showed selective docking to TbPTR1 and TbUDPGE . Nour and co-workers [49] have examined the antitrypanosomal activities of several methylated flavonoids and a chromene from A . conyzoides . The flavonoids all have similar antitrypanosomal activities with IC50 values ranging from 3 . 0 to 6 . 7 µg/mL . The chromene , O-methylencedalinol , on the other hand , was much less active ( IC50 = 78 . 4 µg/mL ) . The docking energies for many of the protein targets was much more negative ( stronger docking ) for the flavonoids than for the chromene . Thus , for example , there is good correlation between log ( IC50 ) and docking energies of the ligands with TbPTR1 or with TbUDPGE ( R2 = 0 . 712 and 0 . 751 , respectively ) . Compounds isolated from Annona senegalensis include annonaceous acetogenins , diterpenoids , and sesquiterpenoids , and aporphine alkaloids ( Table S4 ) [40] , [50]–[53] . The acetogenins ( annogalene , annonacin , annonacin A , annosenegalin , and senegalene ) are probably responsible for the antitrypanosomal activity of the plant [54] , [55] . These compounds show a propensity for docking with TbAK , TbCYP51 , and TbUDPGE . The acetogenins are very flexible with a great deal of conformational mobility . Nevertheless , docking with these protein targets is largely hydrophobic . Key interactions of the acetogenins with TbAK include Phe337 , Gly298 , Asn295 , Asn67 , and Gly296 . Additionally , the acetogenin annogalene is one of the best binding ligands for TbDPGE ( −42 . 9 kcal/mol ) . Bridelia ferruginea has been phytochemically characterized with polyphenolic and triterpenoid constituents ( Table S5 ) [40] , [56] . The flavonoids delphinidin and ferrugin showed docking selectivity for TbPTR1 . The tannin epigallocatechin ( 7→4′ ) gallocatechin showed notably strong docking with TbCYP51 . Although they are relatively weak docking ligands , the triterpenoids friedelin and taraxerol docked selectively with TbUDPGE . Limonoids are characteristic phytochemicals of the Meliaceae , including Carapa procera ( Table S6 ) [40] , [57] , and numerous limonoids have exhibited antiprotozoal activities [58]–[62] . Six of the eleven C . procera limonoids showed notably strong docking with TbCYP51 ( docking energies<−26 kcal/mol ) . A similar trend was noted for docking of Khaya limonoids ( see below ) . Carapolides A , B , and C showed particularly strong docking with docking energies of −31 . 8 , −29 . 3 , and −28 . 5 kcal/mol , respectively; comparable to the docking energy of the co-crystallized ligand , N-[ ( 1R ) -1- ( 2 , 4-dichlorophenyl ) -2- ( 1H-imidazol-1-yl ) ethyl]-4- ( 5-phenyl-1 , 3 , 4-oxadiazol-2-yl ) benzamide [16] ( −28 . 6 kcal/mol ) , for this protein . The limonoids all dock with TbCYP51 near the heme cofactor ( Fig . 1 ) . In addition , preferential docking of individual limonoids with other protein targets include: mexicanolide with TbAK , 3β-isobutyroloxy-1-oxomeliac-8 ( 30 ) -enate with TbPTR1 , and evodulone with TbCatB . We conclude , therefore , that T . brucei sterol 14α-demethylase , TbCYP51 , is a protein target of C . procera limonoids . Enantia chlorantha is dominated by aporphine and berberine alkaloids ( Table S7 ) [40] , [63] , [64] . E . chlorantha aporphine alkaloids seem to show a propensity for docking with TbPTR1 or with TbUDPGE while the berberine alkaloids showed selectivity for TbPTR1 . Both pseudocolumbamine and pseudopalmatine docked with TbPTR1 with docking energies ( −27 . 5 kcal/mol ) comparable to the co-crystallized ligand , 6-phenylpteridine-2 , 4 , 7-triamine [25] ( −27 . 6 kcal/mol ) . These planar alkaloids dock into the active site by way of hydrophobic interactions with the NADP+ cofactor and a hydrophobic pocket formed by Phe97 , Met163 , Cys168 , Pro210 , Trp221 , and Leu209 ( Fig . 2 ) . Liriodenine and columbamine docked selectively to TbTIM with docking energies lower ( −24 . 0 and −24 . 7 kcal/mol ) than the co-crystallized ligand , 3-phosphoglyceric acid [32] ( −21 . 6 kcal/mol ) . These nearly planar alkaloids are known also to be DNA intercalators and topoisomerase inhibitors [65] . Polyphenolic compounds , flavonoids , biflavonoids , etc . , have been isolated and identified from Garcinia kola ( Table S8 ) [40] , [66] . G . kola biflavonoids docked favorably with TbAK and TbODC . The biflavonoids do not dock at the adenosine binding sites of TbAK , but rather in a pocket between the two sites bounded by residues Asn222 , Gly298 , Ala297 , Thr264 , Asp266 , Glu225 , Arg132 , and Asn195 ( see Fig . 3 ) . Likewise , biflavonoid docking with TbODC does not occur at the ornithine/putrescine binding site or the geneticin binding site , but rather in a pocket bounded by Asp243 , Asp385 , Val335 , Asp332 , Ala334 , Ala244 , and Arg277 ( Fig . 4 ) . This would suggest that if G . kola biflavonoids inhibit either TbAK or TbODC , they act as allosteric inhibitors of these proteins . The two tocotrienols garcinal and garcinoic acid , on the other hand , docked more favorably with TbUDPGE . Key interactions of the tocotrienols with the protein are hydrogen-bonding of the phenolic –OH of the ligands with Pro253 and Phe255 , hydrogen-bonding of the carbonyl group of the ligand side chains with Arg268 , hydrogen-bonding of the pyran ring oxygen atom with Arg235 , face-to-face π – π interactions of the ligand aromatic rings with Phe255 , and hydrophobic interactions of the tocotrienol ligands with Leu222 , His221 and the NAD cofactor ( Fig . 5 top ) . The prenylated benzophenone kolanone docked very strongly with TbNH ( docking energy = −37 . 1 kcal/mol ) in the nucleoside binding site ( Fig . 6 ) , a hydrophobic pocket bounded by Trp80 , Phe178 , Asn171 , Trp205 , and Val277 , with additional hydrogen-bonding with Asn171 . The phytochemical compositions of Khaya ivorensis [67] and K . senegalensis [58] , [68]–[72] , like other members of the Meliaceae , are characterized by limonoids [40] . Many of the Khaya limonoids showed markedly strong docking to TbAK as well as TbCYP51 ( see Table S9 ) . Of particular note , 3-O-acetylkhayalactone strongly docked with TbAK , TbDHFR , and TbUDPGE ( −31 . 6 , −32 . 2 , and −34 . 2 kcal/mol , respectively ) . This ligand docked in the same site in TbAK as the Garcinia biflavonoids ( above ) , but in a different position in TbUDPGE ( Fig . 5 bottom ) . Important hydrogen-bonding interactions of 3-O-acetylkhayalactone with TbUDPGE are with residues Glu214 , Ser219 , Leu102 , Thr220 , and His221 . 3-O-Acetylkhayalactone docked in the active site of TbDHFR in the same general location as the co-crystallized ligand ( Fig . 7 ) . In addition , the docking energy of 3-O-acetylkhayalactone ( −32 . 2 kcal/mol ) was lower than either of the co-crystallized ligands , 5- ( 4-chlorophenyl ) -6-ethylpyrimidine-2 , 4-diamine ( pyrimethamine ) and 6 , 6-dimethyl-1-[3- ( 2 , 4 , 5-trichlorophenoxy ) propoxy]-1 , 6-dihydro-1 , 3 , 5-triazine-2 , 4-diamine [26] ( −22 . 7 and −30 . 1 kcal/mol , respectively ) . In general , the Khaya limonoids showed weak or no docking with TbNH , TbTIM , or TbNDRT . The sterols and triterpenoids [40] from Lawsonia inermis showed preferential docking to TbCYP51 ( T . brucei sterol 14α-demethylase ) ( Table S10 ) . This is perhaps not surprising since the normal substrates for this enzyme are sterols . The laxanthones from L . inermis showed preferential docking to TbPTR1 with docking energies comparable to the co-crystallized ligand . In addition , they docked in the same positions and orientations as pseudocolumbamine and pseudopalmatine from Enantia chlorantha ( see above and Fig . 2 ) . Anthraquinones dominate the phytochemistry of Morinda lucida [40] , [73] , [74] , along with triterpenoid acids [75] ( see Table S11 ) . Anthraquinones , as a class , demonstrated significant docking affinity for TbPRT1 and TbTIM ( Fig . 8 ) . Anthraquinones are also known to be DNA intercalators and topoisomerase inhibitors [76] . The triterpenoid acids oleanolic acid and ursolic acid showed notable docking energies with TbCYP51 ( see above ) . The phenylpropanoid oruwacin docked very strongly to TbPTR1 ( docking energy = −32 . 4 kcal/mol ) and TbNH ( docking energy = −31 . 6 kcal/mol , Fig . 6 ) . The phytochemistry of Morinda morindoides [40] is dominated by flavonoid glycosides [77] and phenylpropanoid-conjugated iridoid glycosides [78] ( Table S12 ) . Of these , epoxygaertneroside and morindaoside were selectively strongly binding ligands for TbAK , and morindaoside also docked strongly to TbPTR1 . A number of gaertneroside derivatives showed docking selectivity for TbDHFR ( see Table S12 ) , while dehydroepoxymethoxygaertneroside docked very strongly with TbNH , occupying the nucleoside binding site ( Fig . 6 ) with the same hydrophobic interactions as kolanone and oruwacin ( above ) . It is unlikely that these glycosides will remain intact in vivo , and hydrolysis may be necessary for absorption and general bioavailability [79] . Of the flavonoid aglycones from M . morindoides , apigenin , chrysoeriol , kaempferol , quercetin , and ombuin selectively docked with TbPTR1 . Phytochemical investigations of Nauclea latifolia have revealed numerous indole alkaloids [40] , [80] , [81] ( Table S13 ) . The alkaloid glycosides 10-hydroxystrictosamide , cadambine , dihydrocadambine , and tetrahydrodesoxycordifoline showed notably strong docking to TbUDPGE , whereas non-glycosylated alkaloids showed preferential docking with TbPTR1and/or TbAK . 10-Hydroxyangustine , angustine , naucleamide B , and naucletine , in particular , docked more strongly with TbPTR1 than the co-crystallized ligand , 6-phenylpteridine-2 , 4 , 7-triamine [25] ( docking energy = −27 . 6 kcal/mol ) . T . brucei triosephosphate isomerase , TbTIM , is the likely protein target for the phytochemical agents of Newbouldia laevis . Both furanonaphthoquinones [82] , [83] and pyrazole alkaloids [84] , [85] from this plant showed remarkable selective affinity for this protein ( Table S14 ) . The monomeric furanonaphthoquinone ligands all occupy the same site with hydrogen bonding of the furan oxygen and C ( 9 ) carbonyl oxygen to Lys313; C ( 4 ) carbonyl oxygen with Ser513 and Val514; a van der Waals surface provided by Val533 , Gly534 , Gly535; and a hydrophobic pocket to accommodate the isopropenyl moiety provided by Ile472 , Gly512 , and Leu532 ( see Fig . 9 ) . Similarly , the pyrazole alkaloid 4′-hydroxywithasomnine has key hydrogen-bonding interactions between the pyrazole ring nitrogens and Ser513 and Val514 . The aromatic ring lies in the hydrophobic pocket made up of Ile472 and Leu532 , and there is an additional hydrogen-bonding interaction between the phenolic –OH group and His395 ( Fig . 9 ) . Withanolide triterpenoids , abundant components of Physalis angulata [40] , [86]–[88] , generally showed preferential docking to T . brucei sterol 14α-demethylase , TbCYP51 ( Table S15 ) . This is consistent with the docking of L . inermis triterpenoids and steroids , C . procera and Khaya limonoids ( see above ) . Five of the withanolides , 14-hydroxyixocarpanolide , physagulin J , physagulin L , withangulatin H , and withangulatin I , docked more strongly to TbCYP51 than the co-crystallized ligand [16] . Three withanolides , physagulin A , physagulin L′ , and withangulatin A , docked more strongly into TbODC than the co-crystallized ligand ( pyridoxal 5′-phosphate ) for that protein [34] . The pyrrolidine alkaloid , phygrine , docked with TbTIM preferentially . Picralima nitida glycosylated coumestans [89] showed strong binding to most of the protein targets , except for TbNDRT ( Table S16 ) . They are , for example , along with Acacia nilotica flavonoid gallates , the only ligands that dock to rhodesain with docking energies comparable to the co-crystallized ligand . Of the ligands examined in this work , coumestan 2 is the strongest-binding ligand for TbAK ( −44 . 1 kcal/mol ) and TbUDPGE ( −43 . 7 kcal/mol ) . The corresponding aglycones , 4 , 5 , and 6 , however , were selective for TbPTR1 as well as TbUDPGE . The phytochemistry of Prosopis africana is characterized by piperidine alkaloids ( Table S17 ) [40] , [90] . These ligands exhibited similar docking energies with all protein targets , owing presumably to the small , flexible nature of the compounds . They did , however , show slightly better affinity for TbAK . Phytochemical investigations of Rauwolfia vomitoria have revealed this plant to be replete with indole alkaloids ( Table S18 ) [40] , [91] , [92] . The structural diversity of these indole alkaloids seems to defy targeting any one particular protein . There are some notable docking results , however . 3-Epirescinnamine docked with TbPRT1 and with TbODC more strongly than the co-crystallized ligands , 6-phenylpteridine-2 , 4 , 7-triamine [25] and putrescine [34] , respectively . Isoreserpiline , raumitorine , and rauvanine also docked strongly to TbPTR1 . Ajmalimine , isoreserpiline , rauvomitine , and serpenticine had remarkable docking energies with TbNH . The trimethoxybenzoyl and trimethoxycinnamyl esters , renoxydine , rescidine , rescinnamine , reserpine , along with methyl 3 , 4-dimethoxybenzoylreserpate , were all excellent ligands for TbUDPGE , with docking energies comparable to uridine-5′-diphosphate , the co-crystallized ligand [20] . Of these , renoxydine , rescidine , and reserpine , along with neonorreserpine , docked strongly to TbCYP51 . Cinnamate esters from Securidaca longipedunculata [40] showed selective docking to TbTIM while S . longipedunculata xanthones [40] , [93] had a docking preference for TbPTR1 . Both of these protein targets have relatively small binding sites , which are more suitable for the small ligands ( Fig . 10 ) . The xanthones also docked relatively strongly with TbUDPGE . The S . longipedunculata indole alkaloids dehydroelymoclavine and alkaloid A [94] docked strongly to TbPTR1 and TbAK , respectively , as well as with TbUDPGE ( see Table S19 ) . Phytochemicals isolated from Strychnos spinosa include secoiridoids [95] , indole , pyridine , and naphthyridine alkaloids [40] , sterols and triterpenoids [96] ( Table S20 ) . Relatively small pyridine and naphthyridine alkaloids from Strychnos spinosa showed preferential docking to TbPTR1 and/or TbTIM . The indole alkaloids akagerine , 10-hydroxyakagerine , and kribine also docked preferentially to TbPTR1 , while iridoid glucosides preferred TbUDPGE . Both sterols and triterpenoids from S . spinosa selectively docked with TbCYP51 , similar to what was observed with L . inermis sterols and triterpenoids ( see above ) , but these compounds also showed notably strong docking with TbODC . Interestingly , a comparison of docking energies of triterpenoid and steroid ligands with their antitrypanosomal activities [96] shows no correlation , even comparing TbCYP51 docking or TbODC docking . Plots of log ( IC50 ) vs . docking energies gives R2 values of 0 . 043 and 0 . 007 for TbCYP51 and TbODC , respectively . It may be that inhibition of some other protein target [8] , [9] is the biochemical mechanism of activity for these compounds . In terms of natural products drug discovery , it is useful to examine whether different phytochemical classes show selectivity for particular protein targets . Simple flavonoid ligands showed docking preferences for TbPTR1 and TbUDPGE . Flavonoid gallates , on the other hand , were shown to be promiscuous docking ligands to all protein targets , but were particularly strongly docking with TbAK , TbPRT1 , TbCYP51 , and TbNH . Likewise , flavonoid glycosides tended to be promiscuous docking agents , but with preference for TbAK , TbPRT1 , and TbNH . Oligomeric flavonoids ( tannin-like polyphenolics ) showed strong docking to TbAK . The diversity of flavonoid structures has led to diverse biological activities , including antiprotozoal activity , but the modes of antiprotozoal activity have not been well elucidated [97] . As previously noted ( see above ) , triterpenoid ligands were largely selective for TbCYP51 . Withanolide triterpenoids also showed a docking preference for TbCYP51 , while limonoids preferentially docked with TbAK as well as TbCYP51 . Not surprisingly , sterols showed a propensity to dock with TbCYP51 , but also docked strongly with TbUDPGE . All of the anthraquinone ligands examined in this docking study , docked with strong binding energies to TbPRT1 . Likewise , xanthone ligands exhibited docking selectivity for TbPTR1 . Naphthoquinones , on the other hand , docked preferentially with TbTIM . Most chromene ligands also showed notable docking energies to TbTIM . The phenylpropanoids examined showed preferences for TbTIM as well as TbUDPGE , while glycoside derivatives of phenylpropanoids showed selectivity for TbDHFR . Berberine alkaloids docked preferentially to TbPTR1 while aporphine alkaloids showed some selectivity for TbPTR1 and TbUDPGE . Piperidine alkaloids were also selective for TbUDPGE . Pyrazole and pyridine alkaloids , on the other hand , preferred docking to TbTIM . A total of 93 indole alkaloids were examined in this docking study and many of them showed notable docking energies with TbUDPGE and some with TbAK and TbPTR1 . Glycoside derivatives of alkaloids also preferentially docked with TbUDPGE . Overall , the protein objects most targeted by the phytochemical ligands in this study were TbUDPGE , targeted by many alkaloids; TbPTR1 , preferred by planar-like ligands; TbCYP51 , which docked terpenoid ligands well; and TbAK , which docked many different classes of phytochemicals . Those proteins least preferred in terms of docking energies were rhodesain , TbCatB , and TbNDRT . Rhodesain and TbCatB are both cysteine proteases with relatively small binding sites . It may be that the docking energies reflect the fact that only relatively small ligands , with inherently small docking energies , can fit well into the binding sites of these two proteins . The docking energies do not , however , reflect the potential for covalent bonding to the active sites of these proteins . It is useful , therefore , to examine small electrophilic ligands for energetically favorable docking orientations that would allow for reaction of nucleophilic amino acid side chains to the electrophilic sites of the ligands . Although umbelliferone does not dock with particularly strong energies to rhodesain or TbCatB , it does dock in poses such that the nucleophilic Cys25 of rhodesain or Cys122 of TbCatB are poised to undergo conjugate addition to the pyrone ring ( Fig . 11 ) . The S atom of Cys25 is 3 . 16 Å from C ( 4 ) of docked umbelliferone in rhodesain , while in TbCatB , Cys122 is 3 . 65 Å from C ( 4 ) of umbelliferone . Coumarins have been shown to be trypanocidal agents [98] and it has been suggested that umbelliferone undergoes conjugate addition with available cysteine thiol groups [99] . Many naphthoquinones have been shown to be antitrypanosomal [100] , and are suspected to interfere with redox thiol metabolism by inhibition of TbTR [101] , [102] . There are docking poses , albeit not the lowest-energy poses , of isoplumbagin ( docking pose energy = −9 . 9 kcal/mol ) and lawsone ( docking pose energy = −8 . 6 kcal/mol ) with TbTR such that these quinone ligands are in the proximity of reduced trypanothione ( Fig . 12 ) . Similarly , both isoplumbagin and lawsone dock with the cysteine proteases rhodesain and TbCatB with the electrophilic carbons near the active-site cysteine residues ( Fig . 13 ) . N . laevis furanonaphthoquinones ( α-lapachone derivatives ) also dock with rhodesain in poses such that the nucleophilic Cys25 can undergo Michael addition to the quinone ring ( Fig . 14 ) . None of the furanonaphthoquinones docked near the trypanothione thiol groups in TbTR , however . This in-silico investigation suggests that trypanosomal phytochemicals may target different protein targets . There are several caveats to these docking results: ( a ) many of the phytochemical agents may not be bioavailable due to limited solubility , membrane permeability , hydrolysis , or other metabolic decomposition; ( b ) tannins and other polyphenolics are promiscuous protein binding agents and are likely , therefore , not selective antitrypanosomal ligands; ( c ) the docking studies do not account for synergism in bioactivity of phytochemicals; ( d ) this current study does not address the binding of ligand to human homologous isozymes , which may also be targeted; ( e ) there are likely additional phytochemicals in each of the medicinal plants that have not been isolated or identified; and ( f ) there are likely additional trypanosomal proteins or other biochemical targets that have not yet been identified . Nevertheless , this in-silico molecular docking study has provided evidence for what phytochemical classes and structural manifolds are targeting particular trypanosomal protein targets and could provide the framework for synthetic modification of bioactive phytochemicals , de novo synthesis of structural motifs , and further phytochemical investigations . | Traditional herbal medicine continues to play a key role in health , particularly in remote areas with limited access to “modern medicines” . Many plants are used in traditional Nigerian medicine to treat parasitic diseases . While many of these plants have shown notable activity against parasitic protozoa , in most cases the mode of activity is not known . That is , it is not known what biochemical entities are being targeted by the plant chemical constituents . In this work , we have carried out molecular docking studies of known phytochemicals from Nigerian medicinal plants used to treat human African trypanosomiasis ( sleeping sickness ) with known biochemical targets in the Trypanosoma brucei parasite . The goals of this study were to identify the protein targets that the medicinal plants are affecting and to discern general trends in protein target selectivity for phytochemical classes . In doing so , we have theoretically identified strongly interacting plant chemicals and their biomolecular targets . These results should lead to further research to verify the efficacy of the phytochemical agents as well as delineate possible modifications of the active compounds to increase potency or selectivity . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"Discussion"
] | [
"medicinal",
"chemistry",
"phytochemistry",
"chemistry",
"computational",
"chemistry"
] | 2012 | In-silico Investigation of Antitrypanosomal Phytochemicals from Nigerian Medicinal Plants |
During intracellular infections , autophagy significantly contributes to the elimination of pathogens , regulation of pro-inflammatory signaling , secretion of immune mediators and in coordinating the adaptive immune system . Intracellular pathogens such as S . Typhimurium have evolved mechanisms to circumvent autophagy . However , the regulatory mechanisms targeted by S . Typhimurium to modulate autophagy have not been fully resolved . Here we report that cytosolic energy loss during S . Typhimurium infection triggers transient activation of AMPK , an important checkpoint of mTOR activity and autophagy . The activation of AMPK is regulated by LKB1 in a cytosolic complex containing Sirt1 and LKB1 , where Sirt1 is required for deacetylation and subsequent activation of LKB1 . S . Typhimurium infection targets Sirt1 , LKB1 and AMPK to lysosomes for rapid degradation resulting in the disruption of the AMPK-mediated regulation of mTOR and autophagy . The degradation of cytosolic Sirt1/LKB1/AMPK complex was not observed with two mutant strains of S . Typhimurium , ΔssrB and ΔssaV , both compromising the pathogenicity island 2 ( SPI2 ) . The results highlight virulence factor-dependent degradation of host cell proteins as a previously unrecognized strategy of S . Typhimurium to evade autophagy .
Salmonella enterica serovar Typhimurium ( S . Typhimurium ) is a facultative intracellular Gram-negative pathogen , which causes gastroenteritis in humans and typhoid like disease in mice . The virulence factors of S . Typhimurium are organized in two gene clusters called Salmonella Pathogenicity Island 1 and 2 ( SPI1 and SPI2 ) , which encode two distinct , type-3 secretion systems ( T3SS ) . The effector proteins of SPI1 are critically important for invading non-phagocytic cells . SPI2 dependent effector proteins enable the pathogen to create a niche in the salmonella containing vacuole ( SCV ) for replication , which is important for intracellular survival of the pathogen [1] . Internalized pathogens are subjected to xenophagy , a special form of autophagy that targets intracellular pathogens for lysosomal degradation . Autophagy is an evolutionarily conserved process , which is essential in maintaining cellular homeostasis by eliminating damaged organelles for recycling . Hence , autophagy is vital in promoting cell survival under various stressful conditions , such as pathogen infection , nutrient and growth factor deprivation , or mitochondrial and endoplasmic reticulum stress . Autophagy occurs at basal levels in cells , but is upregulated upon stress such as pathogen invasion [2] . It also contributes to the elimination of many intracellular pathogens including Mycobacterium tuberculosis [3] . In contrast , S . Typhimurium-induced autophagy enables bacteria to obtain nutrients and replicate [4] . Various receptors such as optineurin [5] , galectin8 [6] , NDP52 [7] and ubiquitin modifiers such as FAT10 [8] have been shown to assist in targeting cytosolic S . Typhimurium into the autophagosome . Autophagy is controlled by mammalian target of rapamycin ( mTOR ) signaling pathway . mTOR senses nutrient availability and metabolic changes in the cell . Activation of mTOR results in the formation of multiprotein complexes mTORC1 and mTORC2 [9] . Inhibition of mTORC1 increases autophagy , whereas its activation results in the cessation of autophagy [10] . It has been reported that S . Typhimurium rapidly depletes intracellular amino acid pools , which results in transient inhibition of mTORC1 and activation of autophagy . It is important to note that S . Typhimurium counteracts autophagy by activating mTORC1 [11] . However , the interplay of molecular signals that control mTOR activity and promote autophagy in S . Typhimurium infected cells remains elusive . S . Typhimurium evades phagosome degradation associated with different forms of cell death including apoptosis , pyroptosis and necroptosis [12 , 13] . In macrophages , S . Typhimurium induces a type-I-Interferon-mediated , energy-depleting necroptotic cell death , which results in the loss of host’s resistance and tolerance against the pathogen [14] . Adenosine monophosphate kinase ( AMPK ) is a crucial intracellular energy sensor that is activated upon decline in ATP and increase of the AMP/ATP ratio . Activation of AMPK restores energy levels by enhancing mitochondrial biogenesis and autophagy [15] . AMPK activation is initiated upon binding of AMP to AMPK , which allows the upstream kinase , liver kinase B1 ( LKB1 ) to phosphorylate AMPK [16] . The ability of LKB1 to phosphorylate AMPK is dependent on the deacetylation of its lysine residue by Sirtuin-1 ( Sirt1 ) [17] . Sirt1 belongs to the family of lysine deacetylases and plays an important role in the activation of AMPK [18] . Sirt1 is predominantly localized in the nucleus yet translocates to the cytoplasm in response to the PI3K-AKT signaling pathway [19] . Sirt1 mainly exerts its cell autonomous functions by regulating various transcription factors such as p53 , FOXO1 , FOXO3A and NF-κB [20] in the nucleus . Sirt1 regulates cellular repair mechanisms such as mitochondrial biogenesis and autophagy [21] . It governs the formation of autophagic vacuoles by deacetylating the Atg5 , Atg7 and Atg8 ( LC3 ) complex[22] . In addition , Sirt1-dependent activation of AMPK leads to inhibition of mTOR , which also propels autophagy [18 , 23] . Notably , AMPK provides NAD+ for the activity of Sirt1 thereby establishing a positive feedback loop [24] , which is expected to result in prolonged autophagy . However , little is understood about the role of Sirt1 in pathogen-induced autophagy . In this study , we delineate how S . Typhimurium disrupts the Sirt1/LKB1/AMPK circuit acting as an mTOR checkpoint control . Specifically , we show that S . Typhimurium infection induces lysosomal degradation of Sirt1 , LKB1 , and AMPK , which unleashes mTOR and eventually results in impaired autophagy . The results of this study identify the Sirt1/LKB1/AMPK complex as a previously unrecognized target for SPI2 encoded effector proteins by which Salmonella manipulates the important checkpoint mTOR to compromise autophagic host cell defense mechanisms .
Previously we had shown that S . Typhimurium induces necrotic cell death in macrophages [14] . Because this form of cell death is correlated with energy depletion we began to investigate specific markers of metabolic energy in S . Typhimurium-infected bone marrow-derived macrophages ( BMDMs ) . Indeed , ATP as well as NAD+ levels dropped in macrophages over time upon S . Typhimurium infection ( Fig 1A and 1B and S1A Fig ) . Intracellular decline in levels of ATP and NAD+ trigger the activation of adenosine monophosphate kinase ( AMPK ) [25] . Despite sustained low levels of ATP and NAD+ in S . Typhimurium-infected macrophages , AMPK was only transiently activated at 1h and then declined to basal level at 4h as inferred from the phosphorylation of AMPK and acetyl coA carboxylase ( ACC ) , a bona fide substrate of AMPK ( Fig 1C and 1D ) . LKB1 activates AMPK [26] , therefore we asked if the biphasic AMPK activation is under the control of LKB1 . Interestingly , phosphorylated and non-phosphorylated forms of LKB1 were downregulated upon infection ( Fig 1C and 1D ) . Consistently , microscopical examinations revealed that both abundance and co-localization of LKB1 with AMPK was reduced at 4h post infection ( Fig 1E ) . Pearson’s correlation coefficient analysis confirmed decreased co-localization ( Fig 1F ) . S . Typhimurium infection also induced increased co-localization of AMPK ( Fig 1G and 1H ) and LKB1 with LysoTracker Red ( Fig 1I and 1J ) and LAMP1 ( Lysosome associated membrane protein-1 ) ( S1B–S1E Fig ) suggesting that AMPK and LKB1 were degraded in lysosomes . We confirmed the lysosomal degradation of AMPK and LKB1 ( Fig 1K and 1L ) by inhibiting lysosomal activity using concanamycin A , which also prevented the degradation of p62 a target of lysosomal degradation ( S1F Fig ) . Degradation of AMPK and LKB1 was dependent on the virulence of S . Typhimurium because the heat-killed S . Typhimurium did not alter the expression of total AMPK and LKB1 ( S1G Fig ) . In contrast , inhibiting proteasomes using MG132 did not prevent the degradation of AMPK and LKB1 ( Fig 1K and 1L ) but prevented the degradation of IκB ( S1H Fig ) . Activation of LKB1 requires deacetylation by Sirt1 [17] . Immunofluorescence analysis showed that Sirt1 co-localized with LKB1 in uninfected cells , during the early ( 1h ) and late phase of infection ( 4h ) ( Fig 2A and 2B ) . We also found that LKB1 and AMPK co-immunoprecipitated with Sirt1 , yet the abundance of the proteins were markedly reduced at 4h post infection ( Fig 2C ) . Immunoblot analysis confirmed that Sirt1 protein expression was downregulated in S . Typhimurium-infected macrophages ( Fig 2D and 2E ) . Notably , a significant change in the mRNA expression of Sirt1 was not observed ( S2A Fig ) , suggesting a post-translational mechanism by which S . Typhimurium downregulates Sirt1 . Sirt1 co-localized with S . Typhimurium containing vacuoles ( SCV ) at 1h post infection , which diminished at 4h ( Fig 2F and 2G ) . We also observed that Sirt1 and LKB1 co-localized on SCV shaped vesicles ( S2B Fig ) at 1h post infection . Immunoblot analysis of isolated S . Typhimurium-containing phagosomes revealed the presence of Sirt1 in phagosomes within 30min , which rapidly declined at later time points post infection ( S2C Fig ) . S . Typhimurium infection induced increased co-localization of Sirt1 with LysoTracker Red ( Fig 2H and 2I ) and LAMP1 ( S2D and S2E Fig ) , suggesting that degradation of Sirt1 is lysosome-mediated . We confirmed lysosomal degradation of Sirt1 by inhibiting lysosomal activity by bafilomycin A , E64D or calpeptin , all of which prevented Sirt1 degradation ( Fig 2J and 2K ) . Bafilomycin A treatment also prevented the degradation of AMPK and LKB1 ( S2F Fig ) . In contrast , degradation of Sirt1 was not prevented when treated with proteasome inhibitor MG132 ( S2G Fig ) . Heat-killed S . Typhimurium ( S2H Fig ) and LPS ( S2I Fig ) did not induce the degradation of Sirt1 . These observations indicate that S . Typhimurium induces the translocation of Sirt1 along with AMPK and LKB1 to SCVs and lysosomes followed by degradation . Importantly , Sirt1 is known to shuttle between nucleus and cytoplasm , depending on the induced stress [19] . Analysis of cytoplasmic and nuclear fractions isolated from S . Typhimurium-infected macrophages revealed that cytosolic Sirt1 presented with a slightly higher molecular weight compared to that of the nuclear fraction in the uninfected cells ( Fig 2L ) . The shift in band size is probably brought about by phosphorylation of Sirt1 by kinases , which is a prerequisite for transport out of the nucleus mediated by CRM1 [27] . Indeed , inhibition of CRM1-mediated nuclear export by leptomycin-B reduced the translocation of Sirt1 to the cytosol and its degradation ( S2J and S2K Fig ) similar to the translocation of p53 which was examined as a positive control ( S2L Fig ) . Leptomycin treatment also reduced the activation and degradation of AMPK and LKB1 ( S2M Fig ) . Taken together , our data suggest that S . Typhimurium infection stimulates the nuclear export of Sirt1 onto lysosomes for degradation . Sirt1 nucleocytoplasmic shuttling is regulated by PI3K-AKT signaling pathway [19] . S . Typhimurium infection enhanced the basal phosphorylation of AKT at S473 residue and to a minor extent at Thr308 ( Fig 3A and 3B ) , which is consistent with the idea that cytosolic translocation is mediated by AKT leading to subsequent lysosomal degradation of Sirt1 . In addition , the AKT-mTOR pathway controls lysosomal function [28] . To examine whether AKT is involved in S . Typhimurium-induced Sirt1 degradation , macrophages were treated with AKT inhibitor VIII . AKT inhibition prevented the degradation of Sirt1 ( Fig 3C and 3D ) . Consistently , AKT inhibition led to increased AMPK activity as indicated by phosphorylation of ACC ( Fig 3C and 3D ) . Confocal microscopy showed that AKT inhibitor VIII treatment significantly reduced colocalization of Sirt1 with lysosomes ( Fig 3E and 3F ) and S . Typhimurium ( Fig 3G and 3H ) . Inhibition of PI3K , an upstream activator of AKT , also prevented Sirt1 degradation ( S3A Fig ) . These observations indicate that inactivation of AKT leads to stabilization of Sirt1 resulting in sustained AMPK activation during the later phase of S . Typhimurium infection . Early activation of AKT is facilitated by SopB a virulence factor of S . Typhimurium . The question arises as to the mechanism by which S . Typhimurium activates AKT at a later time point . The pronounced phosphorylation of AKT at S473 ( Fig 3A ) suggested the involvement of mTOR . mTORC1 regulates vacuolar fission , which redistributes the luminal contents of phagosomes into the lysosome network [29] . Consistent with previous reports[4 , 11] , we observed that S . Typhimurium infection increases the activity of both mTORC1 and mTORC2 , indicated by phosphorylation of the well-established targets ribosomal S6 kinase ( S6K ) and N-myc downstream-regulated gene ( NDRG1 ) , respectively ( Fig 4A and 4B ) . Therefore , we investigated whether Sirt1 translocation on to SCVs and lysosomes is mTOR dependent . Indeed , S . Typhimurium-infected macrophages treated with Torin1 ( inhibitor of both mTORC1 and mTORC2 ) significantly decreased the co-localization of Sirt1 with S . Typhimurium ( Fig 4C and 4D ) . Inhibition of mTOR also reduced Sirt1 translocation on to lysosomes ( Fig 4E and 4F ) and attenuated its degradation ( Fig 4G ) . Moreover , S . Typhimurium-phagosomes isolated from cells treated with Torin1 showed markedly reduced Sirt1 ( S4A Fig ) . As observed with AKT inhibition , mTOR inhibition also preserved AMPK-mediated phosphorylation of ACC ( Fig 4G and 4H ) . Similarly , ectopic expression of Sirt1 showed increased activity of AMPK ( Fig 4I and 4J ) . We conclude from our findings that S . Typhimurium–induced translocation and degradation of Sirt1 in phagolysosomes is mTOR and AKT dependent , which is crucially important for the disruption of Sirt1-dependent AMPK activation . Sirt1 , AMPK and mTOR are critically involved in the regulation of autophagy , which is an important cell-autonomous defense mechanism required for pathogen clearance [30] . The biphasic activation and inactivation of Sirt1 and AMPK raised the question about the consequences for autophagy . As has been shown in HeLa cells [4 , 11] , infection of macrophages isolated from LC3-GFP expressing transgenic mice revealed that localization of LC3 on SCVs occurred only at the early time point ( 1h p . i . ) tested ( Fig 5A ) . LC3-GFP on SCVs was significantly decreased at 4h ( Fig 5A ) . Concomitantly , conversion of LC3I to II was observed at 1h post infection ( Fig 5B and 5C ) . Notably , p62 , which is a bona fide target of autophagosomal degradation declined at 1h to accumulate at 2h and 4h post infection , indicating that the autophagic flux was initially increased and subsequently impaired indicating a short and transient phase of autophagy in S . Typhimurium-infected cells ( Fig 5B and 5C ) . As degradation of AMPK and LKB1 involves lysosomes rather than the proteasome ( Fig 1K and 1L ) , we tested whether Sirt1 , AMPK and LKB1 are targeted to lysosomes via autophagy . Microscopical examinations revealed that Sirt1 ( Fig 5D and S5A Fig ) , AMPK ( Fig 5E and S5B Fig ) and LKB1 ( Fig 5F and S5C Fig ) co-localized with LC3 . Furthermore , Sirt1 , AMPK and LKB1 accumulated in autophagy deficient macrophages derived from Atg7fl/fl LysMcre+/+ mice ( Atg7-/- ) ( Fig 5G and 5H ) . These data suggest that transient induction of autophagy is sufficient to target Sirt1 , AMPK and LKB1 for lysosomal degradation . The impact of AMPK degradation on the termination of autophagy in S . Typhimurium-infected macrophages was confirmed by pharmacological activation of AMPK using AICAR . As expected AICAR highly upregulated autophagy as assessed by LC3 conversion and p62 degradation ( S5D and S5E Fig ) . An increase in the co-localization of LC3 with SCVs at 4h post-infection was also observed ( S5F and S5G Fig ) . These data suggest that S . Typhimurium suppresses autophagy upstream of AMPK . Previous reports suggested that mTOR-dependent AKT activation is dependent on virulence factors of S . Typhimurium [31] . Therefore we investigated whether the degradation of Sirt1 and subsequent inhibition of AMPK activation and autophagy could be virulence dependent . To address this we used ΔssrB and ΔssaV mutants of S . Typhimurium . SsrB is a response regulator of a two-component system that regulates the majority of the SPI2 encoded virulence factors [32] and SsaV is a component of the SPI2 type III secretion apparatus [33 , 34] . Infection of macrophages with ΔssrB ( Fig 6A and S6A Fig ) or ΔssaV ( Fig 6B and S6B Fig ) resulted in prolonged phosphorylation of ACC indicative of sustained AMPK activation . Similarly , the S . Typhimurium mutants , ΔssrB ( Fig 6C and S6C Fig ) and ΔssaV ( Fig 6D and S6D Fig ) failed to induce Sirt1 degradation and preserved the enzymatic activity of Sirt1 . Analysis of nuclear and cytoplasmic fractions of macrophages infected with ΔssrB showed reduced translocation of Sirt1 to the cytoplasm ( Fig 6E ) and subsequent targeting to lysosomes ( Fig 6F and S6E Fig ) . Notably , infection with the S . Typhimurium mutants , ΔssrB ( Fig 6G and 6H ) and ΔssaV ( S6F and S6G Fig ) resulted in increased LC3 conversion and reduced p62 expression indicating ongoing autophagy and unhampered autophagic flux , respectively . Indeed , the ΔssrB ( Fig 6I and 6J ) and ΔssaV ( S6H and S6I Fig ) mutants also co-localized with LC3 at 4h post-infection , indicating that autophagy was not impaired . Consistently , both mutants failed to activate mTOR suggesting that mTOR activation and attenuation of autophagy are SsrB and SsaV dependent ( Fig 6K and 6L and S6J and S6K Fig ) . Taken together , the results suggest that S . Typhimurium employs SsrB-dependent virulence factors of SPI2 to disrupt the Sirt1/LKB1/AMPK checkpoint of mTOR and autophagy ( Fig 7 ) .
Intracellular survival and replication within eukaryotic host cells is a hallmark of S . Typhimurium , which is sensed as a major virulence factor of Salmonella . After internalization by phagocytes , Salmonella remains in a specific membrane-bound compartment , termed Salmonella-containing vacuole ( SCV ) . By means of a type III secretion system ( T3SS ) encoded by Salmonella pathogenicity island 2 ( SPI2 ) , S . Typhimurium translocates a number of effector proteins into the cytosol that interfere with host cell defense mechanisms to avoid fusion of SCV with lysosomes and eventually bacterial killing . We here report a novel function of SPI2 which targets the AMPK-dependent activation pathway of mTOR , a prominent checkpoint of cellular homeostasis that modulates a wide array of critical cellular functions , including proliferation , metabolism , and survival . S . Typhimurium infection of macrophages resulted in early energy loss , which is immediately sensed by AMPK . Activated AMPK down-regulates mTOR , which in turn initiates a cellular stress response including autophagy . Our data reveal Sirt1 and LKB1 as essential members of a cytosolic AMPK activation complex , which are targeted by S . Typhimurium for lysosomal degradation in a SPI2 dependent manner . The physical dismantling of the AMPK activation complex allowed robust mTOR activation and subsequent cease of autophagy . Numerous studies have elucidated the significance of autophagy in the cell autonomous defense against S . Typhimurium [5 , 35 , 36] . However , the regulation of autophagy in macrophages during S . Typhimurium infection is not well understood . Initiation of autophagy depends on the activation status of mTOR , which senses the intracellular nutrient availability . It was shown recently that S . Typhimurium induces transient depletion of amino acids in HeLa cells leading to transient activation of autophagy . However , amino acids were gradually replenished resulting in activation of mTOR and inhibition of autophagy [11] . mTOR forms two functionally distinct complexes , mTORC1 and mTORC2 , the activities of both being dependent on the activation of mTOR by AKT within the complex [9] . S . Typhimurium virulence factor SopB was shown previously to activate AKT at Ser473 in an mTORC2-dependent manner at an early time point [11 , 31 , 37] . In agreement with these reports , we observed an increase in phosphorylation of AKT at Ser473 . Moreover , it has been demonstrated that activation of AKT and mTOR is regulated by focal adhesion kinase in a SPI2 dependent manner [38] . Consistently , our results with the ΔssrB and ΔssaV S . Typhimurium mutants now indicate that the sustained activation of AKT and mTOR is dependent on S . Typhimurium virulence factors encoded by SPI2 and/or the type III secretion apparatus . Increased activation of mTOR and AKT are both known to result in the inhibition of autophagy , initiated at early time periods of infection . Notably , AMPK enhances autophagy by at least two routes , that is , by mTOR-independent mechanisms , including the phosphorylation of Ulk1 at Ser317 and Ser777 [39] yet also by inhibiting mTOR through phosphorylation of TSC2 to promote the formation of an Ulk1-Atg13-FIP200 complex [40] . We here demonstrate that S . Typhimurium infection is associated with early but transient activation of AMPK secondary to rapid loss of ATP . Whereas the early drop in ATP led to an increase in the activity of AMPK , S . Typhimurium induced targeting of the AMPK-activation complex for lysosomal degradation reduced AMPK activity during the later phase of infection despite sustained low levels of ATP . A major observation of this study revealed that lysosomal targeting of AMPK and its subsequent degradation is dependent on S . Typhimurium SPI2 , as shown by the ΔssrB S . Typhimurium mutant and SPI2-type III secretion defective mutant ΔssaV [34] . Transient AMPK activation in S . Typhimurium-infected cells resulted in ineffective autophagy with no signs of autophagic flux indicated by accumulation of p62 . In contrast , pharmacological activation of AMPK using AICAR increased LC3 conversion and p62 degradation , suggesting that autophagic flux is highly dependent on sustained AMPK activation , which was counteracted by S . Typhimurium in infected macrophages . In general , S . Typhimurium survives in macrophages and establishes systemic infection by employing genes encoded on SPI2 [41 , 42 , 43] . SsrB is part of a two-component system that specifically activates multiple SPI2 localized genes , which are predominantly expressed after the SCV is acidified [32] and SsaV is a component of the type III secretion apparatus that injects the SPI2 virulence factors into the host cell [33] . Our study reveals that SPI2 encoded virulence factors dismantle an important cellular defense mechanism by targeting Sirt1/LKB1/AMPK complex for lysosomal degradation . AMPK activation is primarily regulated by the upstream kinase LKB1 [26] . We observed that LKB1 constitutively colocalized with AMPK , which is consistent with previous reports that LKB1 activates AMPK . Notably , the cytosolic localization of LKB1 depends on its previous deacetylation by Sirt1 in the nucleus . Sirt1-mediated deacetylation of nuclear LKB1 enables the export of the kinase to the cytosol , where it is phosphorylated by the protein kinase Czeta [17] . Whereas the activation of AMPK by Sirt1 has been studied in the context of mitochondrial metabolism [18] , the regulation of Sirt1 during host-pathogen interactions is not well understood . We show here that S . Typhimurium markedly down-regulates Sirt1 expression commencing within 1h post infection . Several lines of evidence indicated that S . Typhimurium induces lysosomal degradation of Sirt1 , which is consistent with previous observations that Sirt1 is cleaved by cathepsins in endothelial progenitor cells during stress induced premature senescence [1] . Whereas the translocation of Sirt1 onto SCVs results in subsequent lysosomal degradation , S . Typhimurium seems to be able to escape into the cytosol thereby avoiding lysosomal degradation . The decline in Sirt1 expression upon S . Typhimurium infection was accompanied by inhibition of AMPK . Indeed , ectopic overexpression of Sirt1 restored AMPK activity , suggesting that Sirt1 is essentially required for the activation of AMPK during S . Typhimurium infection . Apart from its role in regulating AMPK with secondary effects on autophagy , Sirt1 has been reported to directly regulate autophagy by deacetylating Atg5 and Atg7 [44] . Thus , S . Typhimurium through initiating lysosomal degradation of Sirt1 disrupts autophagic defense mechanisms at several molecular levels .
All animal procedures were in accordance with guidelines laid out by the German Animal Welfare Act and were approved by the North Rhine-Westphalian State Agency for Nature , Environment , and Consumer Protection [Landesamt für Natur , Umwelt and Verbraucherschutz ( LANUV ) Nordrhein-Westfalen; File no: 84–02 . 05 . 40 . 14 . 082 and 84–02 . 04 . 2015 . A443] and the University of Cologne . Bone marrow derived macrophages ( BMDMs ) were prepared as described [14] from C57BL/6J mice maintained and bred in the animal facility of Center for Molecular Medicine , University of Cologne . Atg7fl/fl LysMcre+/+ myeloid specific Atg7 knockout mice were a kind gift from Michael Schramm , University of Cologne . Mice were sacrificed by cervical dislocation and bone marrows from the femurs were flushed using RPMI medium . The flushed cells were centrifuged and resuspended in RPMI containing 10% FBS . Cells were seeded in tissue culture dishes and allowed to differentiate into macrophages in medium supplemented with 20% L929 cell-culture supernatant for 7 days . Non-adherent cells were removed on days 2 and 4 , and adherent macrophages were used from day 7 onwards . Macrophages were infected as described . In brief , cells were seeded into tissue culture plates and infected with S . Typhimurium ( SL1344 ) , S . Typhimurium mutants; ΔssaV or ΔssrB ( MOI , 10 ) . After 30 min , extracellular bacteria were removed and cells were incubated for 2h in medium containing 50μg/ml gentamicin and then were washed and subsequently cultured in medium containing less gentamicin ( 10μg/ml ) . At desired time points cells were collected for analysis . S . Typhimurium mutant ΔssrB generated in the lab of Brett Finlay was obtained from Subash Sad . S . Typhimurium mutant ΔssaV was obtained from the lab of Ivan Dikic . The inhibitors and activators were used 2h prior to infection until unless otherwise mentioned . bafilomycin A1 ( 100nM ) , E64d/pepstatin A ( 10μg/ml ) , calpeptin ( 10μg/ml ) , AKT inhibitor VIII ( 10μM ) , leptomycin B ( 50nM ) , Torin1 ( 10μM ) , AICAR ( 1mM ) , MG132 ( 10μM ) and wortmanin ( 1μM ) . For plasmid transfection , WT Sirt1 plasmid created in the laboratory of Toren Finkel was procured from Addgene ( cat no: 10962 ) [45] . Transfection of plasmid was done using jetPEI transfection reagent ( Polyplus-transfection ) following manufacturer’s instructions . LysoTracker deep red ( L12492 ) , Superscript III first strand synthesis system ( 18080–051 ) , ProLong Gold antifade reagents with DAPI ( P36935 ) , Goat-anti-rabbit alexafluor 488 ( A-11034 ) , 594 ( A-11072 ) , Goat-anti-mouse alexafluor 488 ( A-11017 ) , 594 ( A-11020 ) , Image-iT FX signal enhancer ( I36933 ) were obtained from Life technologies . Bafilomycin A1 ( B1793 ) , concanamycin A ( 27689 ) , MG132 ( M8692 ) , lactacystin ( L6785 ) , pepstatin A ( P5318 ) , leptomycin B ( L2913 ) , AICAR ( A9978 ) and antibody for LC3 ( L7543 ) were obtained from Sigma Aldrich . E64d ( sc-201280 ) , calpeptin ( 117591-20-5 ) and antibodies for Sirt1 ( sc-15404 ) , LAMP1 ( sc-17768 ) , Lamin B ( sc-6217 ) and Syntaxin 3 ( sc-393518 ) were purchased from Santacruz . Protease inhibitor tablets ( 88666 ) , BCA Protein Assay Kit ( 23227 ) , NEPER nuclear and cytoplasmic extraction kit ( 78833 ) , anti-LPS of Salmonella Typhimurium ( MA1-83451 ) and formaldehyde ( 28908 ) were obtained from Thermo Scientific . Antibodies for SIRT1 ( 3931 ) , phospho-NF-κB p65 ( 3033 ) , NF-κB p65 ( 4764 ) , Acetyl- NF-κB p65 ( 3045 ) , phospho-AMPK ( 2535 ) , AMPKα ( 2532 ) , phospho-acetyl-CoA Carboxylase ( 3661 ) , acetyl-CoA carboxylase ( 3662 ) , phospho AKT-T308 ( 2965 ) , phospho AKT-S473 ( 4060 ) , AKT ( 4691 ) , phospho-p70S6 kinase ( 9205 ) , p70S6 kinase ( 9202 ) , SQSTM1/p62 ( 5114 ) , phospho-4E-BP1 ( 9455 ) , 4E-BP1 ( 9452 ) , phospho-NDRG1 ( 3217 ) , phospho-mTOR ( 2974 ) , mTOR ( 2972 ) , phospho-LKB1 ( 3482 ) , LKB1 ( 3047 ) were purchased from Cell Signaling and antibody against GAPDH ( AF5718 ) was procured from R&D systems . Light Cycler 480 SYBR Green I Master ( 04707516001 ) from Roche . RNeasy mini Kit ( 74106 ) , RNase free DNase set ( 79254 ) and DNAseI ( 79254 ) from Qiagen . ATP measurements were performed at Metabolomic Discoveries , Berlin . Metabolites from S . Typhimurium-infected macrophages were extracted using an extraction buffer supplied by the company and the extract was analyzed using LC-QTOF mass spectrometer . Sample concentrations were adjusted to optimally detect ATP . ATP levels were also estimated in our laboratory using Cell Titer-Glo Luminescent Cell Viability Assay ( Promega ) following manufacturer’s instructions . The Intracellular NAD levels upon infection were measured using NAD+/NADH Assay Kit ( Abcam , San Francisco , CA ) according to manufacturer's instructions . BMDMs were grown on 12mm coverslips ( 0 . 1-0 . 2x106 cells at the time of treatment or infection ) . At desired time points , the coverslips were washed with PBS and cells were fixed with 4% ( wt/vol ) formaldehyde for 15min at room temperature . The fixed cells were washed three times with PBS and permeabilized with 0 . 3% tritonX-100 in PBS for 5 minutes at room temperature . The cells were washed with PBS followed by incubation with Image-iT FX signal followed by incubation with primary antibodies for overnight . The cells were then incubated with appropriate secondary antibodies labelled with Alexa flour 488 or 594 . The coverslips were mounted on glass slides using ProLong Gold antifade containing DAPI . Cells were imaged using an inverted Confocal microscope ( Olympus IX81 equipped with Cell^R Imaging Software; Tokyo , Japan ) using a 60x Plano Apo oil objective with 1 . 45 numerical aperture . Pearson’s correlation was calculated using Olympus fluoview fv1000 software . Phagosome preparation was done as previously described [46] . A minimum of 10x106 of BMDMs was seeded on to 10cm dishes followed by infection . After desired time points , the cells were washed with PBS and incubated with equilibration buffer ( 50 mM Pipes buffer , pH7 . 0; 50 mM KCl; 2 mM MgCl2; 5 mM EGTA; 1 mM DTT and 10 μM cytochalasin B ) on ice for 20min . After incubation , lysis buffer was added ( 50 mM Pipes buffer , pH7 . 0; 50 mM KCl; 2 mM MgCl2; 5 mM EGTA; 220 mM mannitol; 68 mM sucrose; 1 mM DTT and 10 μM cytochalasin B ) and lysed cells were scraped using a cell scrapper and collected in a tube . The macrophage lysate was passed 15 times through a 23G needle for homogenization and spun down at 400g for 5 min . The post nuclear supernatant was adjusted to 35% ( wt/vol ) by addition of 65% sucrose in HEPES/EGTA buffer . A sucrose gradient was prepared by overlaying 1ml of HEPES/EGTA buffer containing 65% sucrose , 2ml of 55% sucrose , 3ml of 32 . 5% sucrose and 3ml of 10% sucrose . The gradient was centrifuged at 28 , 500 rpm for 1h at 4°C and the phagosomal fraction at the interface between 55%-39% was harvested . The phagosomal fraction was diluted with HEPES buffer and centrifuged further at 28 , 500 rpm for 1h at 4°C and the pellet was lysed with RIPA buffer and used for western blot analysis . Cells were lysed with radio-immunoprecipitation assay ( RIPA ) buffer containing protease inhibitors . After clearing the cell lysate with protein A/G agarose beads ( Millipore ) for an hour , the beads were removed by centrifugation and the whole cell lysate ( approximately 500μg of protein ) was treated with 4 μg of antibody against Sirt1 for 18h . Protein G agarose beads were then added and incubated for an additional 1hr . The immunoprecipitated proteins along with the agarose beads were collected by centrifugation . The collected beads were washed several times with RIPA buffer . The washed samples were mixed with SDS-PAGE sample loading buffer , boiled and resolved on a 10% SDS-polyacrylamide gel and the respective proteins precipitated were identified by western blotting . Western blotting was performed on proteins extracted using RIPA buffer . BCA was done to quantify the amount of proteins in the lysates . Required samples were mixed 1:1 with 2X sample loading buffer , boiled at 95°C and resolved by SDS-PAGE . Proteins were then transferred on to a PVDF membrane blocked with 5% milk or BSA and probed with the primary antibody of interest followed by treatment with an appropriate secondary antibody conjugated to horseradish peroxidase . The blots were developed using an enhanced chemiluminescence substrate ( GE Health sciences ) and bands were identified by exposing the membrane on to an X-ray film . Densitometric analysis of immunoblots was performed using NIH ImageJ . | S . Typhimurium is a facultative intracellular pathogen which uses its type III secretion system to avoid cell-autonomous defense mechanisms such as autophagy . Here we show that S . Typhimurium induces energy depletion resulting in an early but transient activation of AMPK and autophagy . Salmonella virulence factors target Sirt1/LKB1/AMPK for lysosomal degradation , which enables sustained mTOR-activation and inhibition of autophagy . Activation of mTOR establishes a molecular feedback loop that enhances lysosomal degradation of Sirt1/LKB1/AMPK . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"phosphorylation",
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"autophagic",
"cell",
"death",
"lysosomes",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"intracellular",
"pathogens",
"pathogens",
"cell",
"processes",
... | 2017 | Salmonella Typhimurium disrupts Sirt1/AMPK checkpoint control of mTOR to impair autophagy |
Fragile sites are loci of recurrent chromosome breakage in the genome . They are found in organisms ranging from bacteria to humans and are implicated in genome instability , evolution , and cancer . In budding yeast , inactivation of Mec1 , a homolog of mammalian ATR , leads to chromosome breakage at fragile sites referred to as replication slow zones ( RSZs ) . RSZs are proposed to be homologous to mammalian common fragile sites ( CFSs ) whose stability is regulated by ATR . Perturbation during S phase , leading to elevated levels of stalled replication forks , is necessary but not sufficient for chromosome breakage at RSZs or CFSs . To address the nature of additional event ( s ) required for the break formation , we examined involvement of the currently known or implicated mechanisms of endogenous chromosome breakage , including errors in replication fork restart , premature mitotic chromosome condensation , spindle tension , anaphase , and cytokinesis . Results revealed that chromosome breakage at RSZs is independent of the RAD52 epistasis group genes and of TOP3 , SGS1 , SRS2 , MMS4 , or MUS81 , indicating that homologous recombination and other recombination-related processes associated with replication fork restart are unlikely to be involved . We also found spindle force , anaphase , or cytokinesis to be dispensable . RSZ breakage , however , required genes encoding condensin subunits ( YCG1 , YSC4 ) and topoisomerase II ( TOP2 ) . We propose that chromosome break formation at RSZs following Mec1 inactivation , a model for mammalian fragile site breakage , is mediated by internal chromosomal stress generated during mitotic chromosome condensation .
Unintended double strand breaks ( DSBs ) arise during the unchallenged life of the cell . These breaks do not arise randomly throughout the genome , but occur preferentially at loci referred to as fragile sites . Fragile sites exist in all organisms examined to date including bacteria , yeast , flies , plants , and mammals . Examples include the bacterial ter [1] , budding yeast replication slow zones ( RSZs ) [2] , and mammalian common- and rare- fragile sites [3] , [4] . Some fragile sites are loci of specialized DNA/chromosomal processes . For example , the bacterial ter function as preferred loci of replication fork termination [5] . For the majority of fragile sites however , their precise function , assuming that it exists , remains elusive . The term “fragile site” was first used to describe a heritable locus of recurrent chromosome breakage on metaphase spreads of human lymphocytes [3] . Currently , there are more than 120 fragile sites identified in the human genome [6] . Notably , not all fragile sites form breaks at the same frequency , and some are more prone to breakage than others . For example , FRA3B at 3p14 . 2 is the most fragile site in the human genome , exhibiting breaks in 50% of metaphases after a mild replication stress [6] , [7] . Reason ( s ) for the differential tendencies for breakage among mammalian fragile sites is not known . Mammalian fragile sites are classified as either rare or common , depending on their frequency within the population . Rare fragile sites are seen in <5% of the population . Most rare fragile sites are tri-nucleotide repeats , whose increased breakage is caused by expansion of the repeats [6] . Common fragile sites ( CFSs ) , on the other hand , are present in every chromosome and in all individuals . Furthermore , common fragile sites are conserved throughout mammalian evolution [8] , [9] suggesting that they might be a normal component of the chromosome [4] . Mammalian fragile sites are said to be “expressed” when they display signs of breaks or gaps on metaphase chromosome spreads . Studies have identified several conditions that may play a role in mammalian fragile site expression . These include: ( i ) the time at which a locus is replicated during normal S phase , based on the early observations that the vast majority of mammalian fragile sites replicate late [10]; ( ii ) mild inhibition of DNA replication , contributing to elevated levels of stalled replication forks and further delays in the replication of the normally late replicating fragile loci [11]; ( iii ) inactivation of checkpoint proteins such as ATR [12] or ATM [13]; ( iv ) inactivation of proteins involved in DSB repair and/or replication fork restart [14]; ( v ) premature onset of mitosis [15]; and ( vi ) anaphase and/or cytokinesis [16] , [17] . These observations led to a number of models regarding the mechanism underlying fragile site expression . In all cases , replication fork stalling is proposed to be the initiating event , with stalled forks ultimately giving rise to a DSB by a process or processes whose exact nature remains unresolved . The uncertainly is , in part , due to the fact that equally plausible hypotheses have not yet been tested in a suitable model system . The budding yeast genome , like that of other organisms , contains different types of fragile sites . These differ with respect to their structure , distribution in the genome , and genetic requirement for their stability or breakage [2] , [18]–[23] . The Replication Slow Zone ( RSZ ) is a fragile site that was identified based on its sensitivity to the loss of Mec1 function [2] . Mec1 , like its mammalian counterpart ATR , is an essential protein [24] involved in a number of fundamental processes , including genome duplication , DNA repair , recombination , meiosis , and checkpoint regulation [24]–[27] . It promotes dNTP synthesis during every G1-S transition to ensure that the cell has sufficient levels of dNTPs for genome duplication [28] , [29] . Mec1 up-regulation of dNTP synthesis is also essential during replication stress- or DNA damage- checkpoint responses [30] , [31] . Additional checkpoint functions of Mec1 include stabilization of stalled forks , coordination of repair , and preventing cell cycle progression until the damage situation is resolved [27] , [32] , [33] . The name RSZ was based on the observation that replication forks moved notably slower through these regions than through other loci during normal S phase [2] . In MEC1 cells , forks continue to progress through RSZs , eventually completing their duplication . In mec1-4 cells , replication forks progress more or less normally until they reach RSZs . At RSZs , the forks remain stalled for about 90 minutes , until the appearance of DSBs at these loci some time during G2/M [2] . Analysis of eleven RSZs identified on chromosomes III and VI suggests that they do not occur randomly along the chromosome , but occur between highly active replication origins along the entire length of the chromosome; a notable exception , however , is the centromeric region , which lacks a RSZ [2; N Hashash and R Cha , unpublished data] . RSZs and mammalian CFSs are both large genetic determinants , each comprising about 0 . 1% of the respective genome ( i . e . ∼10 kb RSZ of 1 . 5×107 bp budding yeast genome and ∼1 Mb CFS of the 3×109 bp mammalian genome ) . Some studies reported a correlation between the occurrence of some CFSs or RSZs and certain features of the genome , including high flexibility , high AT content , hairpin structure , and/or hotspots for ectopic genome integration [2] , [6] , [34] , [35] . Currently however , there are no structural or functional features that can be utilized for their a priori identification . RSZs and CFSs are both late replicating loci [2] , [10] and exhibit sensitivity to mild replication stress or deficiencies in Mec1 or ATR [2] , [12] , [19] . Largely based on these similarities , RSZs were proposed to be homologous to mammalian CFSs [2] , [19] , [35] . Here , we investigate the mechanism of RSZ breakage following Mec1 inactivation . Specifically , we tested involvement of each of the five processes implicated in mammalian fragile site expression ( above ) . Results showed that RSZ breakage following Mec1 inactivation required functions of topoisomerase II ( Top2 ) and the condensin complex; in the absence of Top2 or condensins , RSZs did not break , even though replication forks still stalled . In contrast , replication fork restart , spindle tension , anaphase , or cytokinesis were all dispensable for RSZ breakage . Based on these observations , we propose that internal chromosomal stress , generated during mitotic chromosome condensation , promotes the conversion of stalled forks at RSZs to DSBs .
Replication forks stall during unchallenged S phase either as a part of normal replication program [2] , [21] , [36] , [37] or incidentally , upon encountering a damaged template [33] or due to insufficient levels of dNTPs [19] , [30] , [32] . In either case , the resumption of DNA synthesis from stalled forks or the rescue by the firing of cryptic origins , while maintaining the integrity of stalled forks , is essential for cell's survival . Homologous recombination plays a key role in replication fork restart [38] . During this process , DSBs can be generated as an intermediate and contribute to endogenous chromosome breakage . To test whether breakage at RSZs is generated via homologous recombination or recombination-related process , we utilized a previously characterized temperature sensitive mec1 strain , mec1-4 [2] , and assessed the effects of eliminating relevant proteins; homologous recombination proteins ( Rad50 , Rad51 , Rad52 , Rad54 , Rad55 , and Mre11 ) , the Sgs1BLM-Top3 complex , the Srs2 helicase , and the Mus81-Mms4 endonuclease , a putative resolvase [38]–[42] . Thermal inactivation of Mec1-4 results in prolonged replication fork stalling at RSZs , followed by chromosome breakage at these loci . The breaks appear typically around 90–120 minutes after alpha-factor arrest/release , corresponding to the G2/M phase of the first cell cycle after the release . MEC1 and mec1-4 strains carrying a null allele of one of the genes mentioned above were arrested with a-factor at 23°C and released into fresh YPD media at 37°C , a restrictive temperature for mec1-4 [2] . Samples were collected 3 hours after the release , and the status of chromosome III ( ChrIII ) was assessed by pulse field gel electrophoresis ( PFGE ) followed by Southern hybridization using a telomere-proximal probe , CHA1 ( Figure 1A , 1B ) . As shown previously [2] , DSBs enriched for RSZs in ChrIII were observed in the mec1-4 culture ( Figure 1C , 1D ) . Elimination of various proteins with a role in the replication fork restart did not prevent break formation at RSZs ( Figure 1; data not shown ) indicating that the involvement of this process was unlikely . The spindle assembly checkpoint ( SAC ) is an evolutionarily conserved mechanism responsible for ensuring that every pair of sister-chromatids is under spindle tension prior to anaphase . The SAC monitors this process by assessing microtubule occupancy of the kinetochores and/or tension generated across chromosomes/kinetochores [43] . Like the other checkpoint systems , the SAC is a signal transduction cascade and is mediated by the Mad1 , 2 , 3 ( mitotic arrest deficient ) and Bub1 , 2 , 3 ( budding uninhibitied by benzimidazole ) proteins . In the absence of these proteins , cells proceed through mitosis irrespective of whether all sister kinetochores are under spindle tension , resulting in frequent chromosome mis-segregation and cell death . Although the SAC was originally thought to operate independently of the DNA damage checkpoint , recent evidence suggests an interplay between the two . For example , Mec1/Tel1 have been shown to inhibit anaphase by utilizing Mad/Bub proteins independently of the kinetochores in response to DNA damage [44] , which might account for the earlier observation that about 50% of mec1-4 cells undergo mitosis despite the presence of unresolved replication forks [2] . These considerations raise the possibility that RSZ breakage might occur as a result of inappropriate mitosis . If this was the case , we reasoned that inactivation of the SAC might increase RSZ breakage by allowing a greater proportion of mec1-4 cells to proceed through mitosis with unresolved replication forks at RSZs . We tested this possibility by assessing the impact of deleting MAD2 or BUB2 on RSZ breakage . MEC1 and mec1-4 strains in mad2Δ BUB2 , MAD2 bub2Δ , or MAD2 BUB2 backgrounds were arrested with alpha-factor at 23°C and released into fresh YPD media at a restrictive temperature for mec1-4 . Samples were collected 3 hours after the release , and the status of ChrIII was assessed . As expected , RSZ breakage was observed in mec1-4 control culture ( Figure 2A ) . RSZ breakage was also observed in the mec1-4 mad2Δ or mec1-4 bub2Δ cultures ( Figure 2A ) . Furthermore , the extent of breakage was comparable in the presence or absence of MAD2/BUB2 , suggesting that RSZ breakage was unlikely to be caused by compromised SAC function , leading to inappropriate mitosis . Another implication of the lack of an impact of mad2Δ or bub2Δ is that RSZ breakage might be independent of spindle tension . Assuming that inactivation of the SAC would have allowed some cells to proceed through mitosis in the absence , or with a reduced level , of spindle tension , we reasoned that mad2Δ or bub2Δ may have resulted in a reduction in RSZ breakage if spindle tension played a role . To directly address the involvement of spindle tension , we investigated the effects of microtubule depolymerising drugs such as methyl1-2-benzimidazolecarbamate ( MBC ) or nocodazole . First , we confirmed that spindle poison was effective in preventing elongation of spindles in mec1-4 cells ( Figure 2D ) . In the absence of spindle poison , the elongation in mec1-4 cells occurred reproducibly earlier than in MEC1 ( Figure 2Di ) . The reason for this remains unknown but is likely to be related to the role ( s ) of Mec1/Tel1/Rad53 in regulating spindle status in response to DNA damage or replication stress [45] , [46] . We also found that MBC blocked Clb2 degradation , a readout for mitotic exit [47] , in mec1-4 ( Figure 2E ) . The latter suggested that although mec1-4 cells were compromised in preventing the onset of mitosis in the presence of stalled forks [2] , they were competent in mediating a spindle damage-dependent SAC response . To ensure that we assessed the impact of spindle depolymerisation on RSZ breakage , rather than the impact of SAC response to the depolymerisation , we decided to examine the effects of spindle poison in the absence of the SAC . mec1-4 mad2Δ or mec1-4 bub2Δ strains were released from alpha-factor arrest as described above , except that they were released into YPD media containing either MBC or nocodazole ( Materials and Methods ) . Samples were collected 3 hours after the release and assessed for RSZ breakage . Results showed chromosome breakage in the presence of either drug ( Figure 2B; data not shown ) , demonstrating that mitotic spindles are dispensable for RSZ breakage . The dispensability of mitotic spindles prompted us to consider whether internal chromosomal stress might be involved in RSZ breakage . During mitotic prophase , the duplicated genome undergoes dramatic structural reorganization , leading to sister chromatid individualisation and chromosome compaction in preparation for segregation during anaphase [48] , [49] . To test whether the intra-chromosomal stress generated during these processes might have a role in RSZ breakage in the absence of spindle tension , we assessed the impact of inactivating relevant gene products . These included; ( i ) Scc1/Mcd1 ( hereon referred to as Scc1 ) , a component of the cohesion complex that holds sister chromatids together until their disjunction at the onset of anaphase [47] , [50] , ( ii ) Esp1 , a caspase-like cysteine protease that promotes sister chromatid separation by mediating the cleavage of Scc1 [51] , ( iii ) Ycg1 and Ysc4 , two non SMC components of the condensin complex , required for mitotic chromosome compaction [52] , [53] , and ( iv ) Top2 , a type II topoisomerase that catalyzes decatetation of DNA strands between the sister chromatids to allow their resolution and facilitate chromosome condensation [48] , [49] , [54] . A set of mec1-4 mad2Δ strains , each expressing temperature sensitive scc1 , esp1 , ycg1 , ysc4 , or top2 alleles were released from alpha-factor arrest into YPD + MBC media at 37°C , a restrictive temperature for all of the conditional alleles utilized . Samples were collected 3 hours after the release and analysed for RSZ breakage ( Figure 2C ) . As expected , RSZ breakage was observed in the mec1-4 mad2Δ control strain . The breakage was also observed in strains expressing a temperature sensitive scc1 or esp1 allele , suggesting that RSZ breakage occurred independently of the status of the cohesins . In contrast , inactivation of Top2 , Ycg1 , or Ysc4 suppressed chromosome breakage , suggesting that mitotic chromosome condensation might be involved in RSZ breakage in the absence of spindle tension . To rule out the possibility that the top2 , ycg1 , or ysc4 suppression was mediated by their impact on S phase progression , either by allowing replication forks to progress through RSZs [2] or by committing cells to inviability before forks reach a RSZ [19] , we assessed their impact on the status of S phase progression . A WT , mec1-4 mad2Δ , or mec1-4 mad2Δ strain expressing a temperature sensitive allele of either top2 or ycg1 was released from G1 arrest into YPD+MBC media at 37°C . Samples were collected at various time points after the release and subjected to fluorescent activated cell scan ( FACS ) analysis ( Materials and Methods ) . In a WT control , cells proceeded through S phase and completed bulk genome duplication by 80 minutes following alpha-factor arrest/release ( Figure 2F ) . In contrast , S phase progression in a mec1-4 mad2Δ culture , like that in a mec1-4 culture [2] was delayed ( Figure 2F ) , consistent with earlier observations that the status of MAD2 did not confer any effects on DNA replication [55] . Thermal inactivation of top2 or ycg1 did not exert a notable effect on S phase progression , suggesting that the top2 , ycg1 , or ysc4 suppression was unlikely to be due to their impact on DNA replication . Taken together , we conclude that Top2/condensin-mediated mitotic chromosome condensation triggers RSZ breakage in the absence of mitotic spindles . Results thus far showed that breakage of RSZ , like that of mammalian CFSs , can occur in the absence of spindle tension . Importantly however , breakage of RSZs or CFSs occurs during normal cell proliferation in the presence of spindle tension . Thus , it was formally possible that the Top2/condensin dependent RSZ breakage was a mechanism operating specifically in the absence of mitotic spindles . To address this , we assessed the effects of inactivating Top2 , Ycg1 , or Ysc4 in the presence of spindle tension . A set of mec1-4 MAD2 BUB2 strains expressing temperature sensitive alleles of top2 , ycg1 , or ysc4 was released from G1 arrest into a fresh YPD media in the absence of spindle poison . Samples were collected 3 hours after the release and analysed for RSZ breakage . The results showed that top2 , ycg1 , or ysc4 suppressed RSZ breakage ( Figure 3A ) . In contrast , thermal inactivation of Scc1 or Esp1 did not prevent the breakage ( Figure 3B ) . These results suggest that the genetic requirement ( and the mechanism , by extension ) of RSZ breakage in the presence or absence of spindle tension is likely to be the same . We conclude that Top2/condensin mediated mitotic chromosome condensation is required for RSZ breakage during normal cell proliferation irrespective of the status of spindle tension . As expected from the essential nature of TOP2 and condensin , inactivation of these gene products did not rescue the lethality of mec1-4 at non-permissive temperature ( Figure 3C; data not shown ) . If RSZ breakage is independent of spindle tension , then a prediction might be that it should also be independent of the events downstream of the SAC execution point , such as anaphase , mitotic exit , and cytokinesis . We tested this by monitoring the occurrence of these events in mec1-4 and MEC1 cultures as they proceeded through a synchronous cell cycle in the presence of spindle tension . Samples were collected at various time points following alpha-factor arrest/release and assayed for RSZ-expression , Scc1-cleavage , a readout for the onset of anaphase [43] , and Clb2-degradation , a readout for exit from mitosis [47] ( Figure 4 ) . In the MEC1 culture , cells completed bulk genome duplication between 60–75 minutes following alpha-factor release ( Figure 4A panels i and v ) . An Scc1 cleavage product was observed starting from 75 minutes after release ( Figure 4A panels ii and v ) . Levels of Clb2 peaked at 45 and 60 minutes following the release and decreased rapidly thereafter ( Figure 4A , panels iii and v ) . These results indicate that , in the MEC1 culture , the completion of bulk genome duplication ( 60–75′ ) , onset of anaphase ( 75–90′ ) , and exit from mitosis ( 90′ ) occurred in a temporally ordered manner . PFGE/Southern analysis of ChrIII showed that chromosome breakage in this culture remained at background levels ( Figure 4A panel iv ) . In the mec1-4 culture , the cells remained stuck in mid-S phase from about 60 minutes following the release , suggesting that anaphase did not take place ( Figure 3B panels i and v ) . Scc1 cleavage was not observed ( Figure 3B panels ii and v ) . A modest reduction in Clb2 levels was observed starting 90 min after the release , although the extent of reduction was notably less than that observed in the MEC1 culture ( Figure 3B , panels iii and v ) . In this mec1-4 culture , DSBs began to accumulate starting at t = 90–105 minutes ( Figure 3B panels iv and v ) . These results demonstrate that RSZ-expression in the presence of spindle tension occurs in the absence of Scc1 cleavage , Clb2 degradation , or , by extension , the onset of anaphase or exit from mitosis , respectively . Next , the occurrence of cytokinesis was assessed . To this end , we generated MEC1 and mec1-4 strains expressing an endogenous copy of MYO1-GFP , and monitored the appearance of binucleate cells with or without the Myo1-GFP signal ( Figure 4C panel iv versus panel v ) . MYO1 encodes a component of the actomyosin ring that localizes to the bud neck from early S phase ( e . g . Figure 4C panel ii ) . The Myo1 ring remains at the neck until cytokinesis , during which the ring constricts and Myo1 disappears from the bud neck [56] ( Figure 4C panel v ) . MEC1 MYO1-GFP and mec1-4 MYO1-GFP cells were released from alpha-factor arrest into fresh YPD media at a restrictive temperature in the absence of spindle poison . Samples were collected at various time points and assessed for the morphology of DNA ( via a DAPI stain ) and the status of Myo1-GFP ring . In both cultures , the Myo1-GFP ring appeared by 45 minutes after release ( Figure 4C panel vi ) . In MEC1 cells , the Myo1-GFP ring remained at the bud neck until 75 minutes and disappeared by 105 minutes , indicating that cytokinesis had occurred . In contrast , the Myo1-GFP ring in mec1-4 cells remained at the bud neck throughout the duration of the experiment , indicating that cytokinesis did not take place ( Figure 4C panel vi ) . In this culture , the fraction of cells that had undergone anaphase – i . e . those containing two DAPI staining bodies that were separated by MyoI-GFP ring ( e . g . Figure 4C panel iv ) remained low , in agreement with the lack of Scc1 cleavage ( Figure 4Bii ) . In the MEC1 control , the fraction of binucleate cells reached about 50% by t = 75 minutes; thereafter , the fraction decreased as the cells underwent cytokinesis . RSZ expression in the mec1-4 culture was observed at t = 120 minutes ( data not shown ) . Taken together , these results show that RSZ breakage in the presence of spindle tension occurs independently of anaphase , mitotic exit , or cytokinesis . The simplest interpretation would be that the breakage occurs before the onset of anaphase .
The aim of this study was to examine the mechanism of chromosome breakage at RSZ , a MEC1-sensitive fragile site and a model for mammalian CFSs . Specifically , we tested involvement of the following possibilities , each of which had been implicated in mammalian fragile site expression: ( i ) errors in replication fork restart , ; ( ii ) premature mitotic chromosome condensation; ( iii ) spindle tension; ( iv ) anaphase; or ( v ) cytokinesis . Evidence revealed that Top2 and condensin proteins were required for RSZ breakage; in contrast , the key proteins involved in replication fork restart , spindle tension , anaphase , or cytokinesis were dispensable . In all eukaryotes examined to date , an essential function of Top2 and condensins is chromosome compaction [48] , [49] , [53] , [54] , [57] , [58] . Although the extent of mitotic chromosome condensation in budding yeast is about two orders of magnitude less than that observed in metazoan cells ( the compaction ratio is 160 in yeast versus 10 , 000–20 , 000 in metazoans; 57 , 58 ) , inactivation of budding yeast Top2 or condensin subunits also results in a chromosome compaction defect [52] , [59] , [60] , in agreement with the notion that the mechanism is evolutionarily conserved . Taken together , we propose a model whereby two temporally and genetically distinguishable events mediate chromosome breakage at RSZs ( Figure 5 ) . In WT cells , the genome duplication is complete by the end of S phase ( Figure 5Ai ) . During mitotic prophase , the duplicated genome undergoes Top2- and condensin- dependent chromosome compaction ( Figure 5Aii ) in preparation for its disjunction during anaphase ( Figure 5Aiii ) . In the absence of Mec1 function , replication forks stall at RSZs ( Figure 5Bi ) ; despite this , the cells exit S phase and proceed through the cell cycle . During prophase , the incompletely duplicated genome of mec1-4 cells becomes subjected to Top2- and condensin- dependent chromosome compaction ( Figure 5Bii ) . We propose that internal stress generated during this process promotes the conversion of stalled forks to a DSB . The molecular mechanism underlying the catalysis of breakage is unknown , but may involve a nuclease that is yet to be identified ( below ) . Above evidence indicates that chromosome breakage is independent of spindle tension or tension-dependent events such as anaphase or cytokinesis . This breakage is also independent of the status of sister chromatid cohesins , consistent with the report that cohesion removal , while essential for sister chromatid resolution , is dispensable for mitotic chromosome compaction [61] . In the absence of Top2 or condensins ( Figure 5C ) , chromosome condensation does not take place; therefore , the incompletely duplicated genome of mec1-4 cells is not subjected to the internal stress that triggers the conversion of stalled forks to DSBs . Nevertheless , the cells die , likely due to the lack of an essential Top2 or condensin function ( s ) [52] , [62] . With regard to the dispensability of the replication fork restart process , it is important to note that the list of candidate genes examined is not exhaustive , and therefore , we cannot rigorously eliminate its involvement based on this line of evidence . Nevertheless , our results unequivocally rule out the involvement of some of the key players in replication fork restart that had previously been implicated in breakage at different types of fragile sites ( see below ) ; the RAD52 epistasis group proteins , the Sgs1BLM-Top3 complex , the Srs2 helicase , and the Mus81-Mms4 endonuclease [38]–[42] . The dispensability of spindle tension is not surprising in the light of the fact that the distribution pattern of RSZs is different from that of the spindle tension mediated breaks . Specifically , RSZs are found between active replication origins along the entire length of the chromosome except for the centromeric region ( [2]; N . Hashash and R . Cha , unpublished data ) . In contrast , spindle tension-dependent DSBs tend to occur around the centromere , the locus of greatest spindle tension [22] , [63] . Mammalian CFSs , like RSZs , are found along the chromosome arms . Furthermore , the fact that mammalian fragile sites are defined as loci of recurrent breaks or gaps on metaphase chromosome spreads , obtained from cultures treated with spindle poisons such as colchicines [3] , [4] , strongly suggest that expression of mammalian fragile sites , like that of RSZs , occurs independently of spindle tension . The amount of force exerted by a pair of microtubules at the centromere ( i . e . 20 piconewton [pN] ) is estimated to be at least an order of magnitude smaller than that required to break the chromosome ( i . e . 480 pN ) [64] , [65] . Assuming that the intra-chromosomal stress generated during mitotic chromosome compaction is less than that generated by the spindles , it is likely that the Top2/condensin-dependent RSZ breakage is mediated by an endonuclease . As a means to test whether Top2 was the responsible enzyme , we performed Top2 ChIP-on-CHIP analysis in MEC1 and mec1-4 cells , reasoning that if Top2 catalyzed the cleavage , we might observe its enrichment at RSZs . Analysis thus far has failed to show any such enrichment , suggesting that its direct involvement was unlikely ( N Hashash , R Cha , Y Katou , K Shirahege; unpublished data ) . Nevertheless , this observation alone does not eliminate the possibility , because Top2 may dissociate from the ends of the DSB after DNA cleavage , and therefore would not normally remain enriched at RSZs . Alternatively , the cleavage might be mediated by a different protein , for example , Yen1 , an evolutionarily conserved Holiday junction resolvase that is activated during M phase of the cell division cycle [66] or proteins involved in post replication repair [67] . It is also possible that the DSBs at RSZs result from cleavage of single stranded DNA associated with stalled forks [68] . A positive role for Top2 and condensin in chromosome breakage is unexpected in light of the observations that their inactivation caused , rather than prevented , DSB formation [e . g . [22] , [23] , [69] , [70] . Also surprising is the dispensability of anaphase or cytokinesis in RSZ breakage . Upon a closer examination , however , it becomes apparent that the chromosome breakage examined in each study was at different types of fragile loci in the genome , in that they differed with respect to their structure ( e . g . a hairpin or a specific protein-DNA complex ) , distribution ( e . g . chromosome arms versus the centromeres ) and/or the timing of their expression ( e . g . during S phase , before anaphase , or during cytokinesis ) [2] , [16] , [18]–[20] , [22] , [23] , [41] , [69] , [71] . These observations provide further support for the notion that both the stability and the expression of each type of fragile sites is under a specific genetic and regulatory control [19] . Among the different types of fragile sites identified and characterized to date , the RSZ appears to be the closest structural and functional homolog of mammalian fragile sites . Furthermore , among the currently proposed mechanisms of mammalian fragile site expression , the mechanism of RSZ breakage inferred in the current study seems to be most consistent with the original definition of a mammalian fragile site , that it is a heritable locus of recurrent breaks or gaps on metaphase chromosome spreads [3] . Taken together , it is tempting to speculate that the mechanism of RSZ breakage and that of CFS expression , at least for those that are sensitive to the loss of ATR or ATM functions [12] , [13] , might be conserved and that the mammalian Top2 and condensin may similarly play a role in promoting fragile site expression .
All strains were of the SK1 background except those noted . Relevant genotypes of the strains are listed in Table S1 . Unless specified otherwise , cells were grown in YPD ( 1% [w/v] yeast extract , 2% [w/v] bacto-peptone , 2% [w/v] glucose ) media . To obtain a synchronous culture for cell cycle analysis , mid-log cultures were arrested with 5 µg/ml alpha-factor for 3 hours before being released to fresh YPD media . For temperature-sensitive strains , cells were arrested with alpha-factor at 23°C before releasing to YPD media prewarmed to a restrictive temperature . To induce microtubule deploymerization , cells were grown in the presence of either 15 µg/ml nocodazole ( Sigma-Aldrich ) or 40 µg/ml carbendazim ( MBC; Sigma-Aldrich ) . Cells from 1 ml of relevant samples were fixed ( 40% [v/v] ethanol , 0 . 1 M sorbitol ) for at least 3 hours before being pelleted , resuspended in RNase solution ( 50 mM Tris-HCl pH 7 . 5 , 100 µg/ml RNaseA ) and incubated overnight at 37°C . The next day , the cells were treated with 500 µl of pepsin solution ( 50 mM HCl , 5 mg/ml pepsin ) for a minimum of 5 minutes at room temperature , before being resuspended in 1 ml SYTOX solution ( 50 mM Tris-HCl pH 7 . 5 , 1 µM SYTOX Green; Invitrogen Molecular Probes ) . The samples were incubated overnight at 4°C . The next day , they were analyzed on a Becton Dickinson FACSan using Cell Quest software ( Becton Dickinson ) . Chromosome-sized DNA in agarose plugs for PFGE was prepared as described [72] . Electrophoresis was performed at 14°C in a Bio-Rad CHEF Mapper under the following condition: a voltage gradient of 6 V/cm , switch times of 5–30 sec , a switch angle of 115° , in a 1% agarose gel in 0 . 5× TBE for 24 hours . The DNA in gels was transferred to nylon membranes and hybridized with 32P-labeled CHA1 probe , a 757 bp HindIII-BamHI fragment ( −156 to +601 of the ORF ) restricted from a pUC19 based plasmid ( pRSC38 ) . The image was visualized and signals quantified using a Storm 860 PhosphorImager and ImageJ software , respectively . 900 µl from appropriate samples was incubated with 100 µl 37% ( w/v ) formaldehyde ( Fisher Scientific ) for 10 minutes at room temperature . The cells were pelleted and washed twice with 1 ml PBS , and then resuspended in 200 µl PBS . A 10 µl sample of the cell suspension was spread onto a glass microscope slide and left to dry . Before application of the glass coverslip , 2 µl of 4 , 6-diamino-2-phenylindole ( DAPI; Sigma ) solution ( 1 . 5 µg/ml in Vectashield mounting medium [Vector Lab] ) was dotted onto the dried cells . Fluorescence microscopy was performed on a Deltavision Spectris system . Whole-cell extracts were prepared from cell suspension in 20% trichloroacetic acid by agitation with glass beads . Precipitated proteins were solubilized in SDS-PAGE sample buffer and appropriate dilutions were subjected to SDS-PAGE and Western blotting . Antibodies utilized for Western blotting were , rabbit polyclonal anti-Clb2 ( Santa Cruz Biotechnology Inc ) , mouse monoclonal anti-HA ( 12CA5; NIMR , London ) , mouse monoclonal anti-MYC ( 9E10; NIMR , London ) , and rat monoclonal anti-tubulin ( YL1/2; Abcam ) . For each antibody a 1∶1000 dilution was used for Western Blotting except for anti-tubulin , which was used at a 1∶5000 dilution . | Chromosome breakage can occur during normal cell division . When it occurs , the breaks do not arise randomly throughout the genome , but at preferred locations referred to as fragile sites . Chromosome breakage at fragile sites is an evolutionarily conserved phenomenon , implicated in evolution and speciation . In humans , fragile site instability is also implicated in mental retardation and cancer . Despite its biological and clinical relevance , the mechanism ( s ) by which breaks are introduced at mammalian fragile sites remains unresolved . Although several plausible models have been proposed , it has not been possible to ascertain their contribution , largely due to the lack of a suitable experimental system . Here , we study a yeast model system that closely recapitulates the phenomenon of chromosome breakage at mammalian fragile sites . We eliminate all but one of the currently considered models—premature compaction of the incompletely replicated genome in preparation for their segregation during cell division . We also find that the breakage required functions of three proteins involved in the genome compaction , an essential process that is evolutionarily conserved from bacteria to humans . Our findings suggest that a fundamental chromosomal process required for normal cell division can paradoxically cause genome instability and/or cell death , by triggering chromosome breakage at fragile sites . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"model",
"organisms",
"genetics",
"biology",
"microbiology",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] | 2012 | Topoisomerase II– and Condensin-Dependent Breakage of MEC1ATR-Sensitive Fragile Sites Occurs Independently of Spindle Tension, Anaphase, or Cytokinesis |
In order to cross a street without being run over , we need to be able to extract very fast hidden causes of dynamically changing multi-modal sensory stimuli , and to predict their future evolution . We show here that a generic cortical microcircuit motif , pyramidal cells with lateral excitation and inhibition , provides the basis for this difficult but all-important information processing capability . This capability emerges in the presence of noise automatically through effects of STDP on connections between pyramidal cells in Winner-Take-All circuits with lateral excitation . In fact , one can show that these motifs endow cortical microcircuits with functional properties of a hidden Markov model , a generic model for solving such tasks through probabilistic inference . Whereas in engineering applications this model is adapted to specific tasks through offline learning , we show here that a major portion of the functionality of hidden Markov models arises already from online applications of STDP , without any supervision or rewards . We demonstrate the emergent computing capabilities of the model through several computer simulations . The full power of hidden Markov model learning can be attained through reward-gated STDP . This is due to the fact that these mechanisms enable a rejection sampling approximation to theoretically optimal learning . We investigate the possible performance gain that can be achieved with this more accurate learning method for an artificial grammar task .
An ubiquitous motif of cortical microcircuits is ensembles of pyramidal cells ( in layers 2/3 and in layer 5 ) with lateral inhibition [1]–[3] . This network motif is called a winner-take-all ( WTA ) circuit , since inhibition induces competition between pyramidal neurons [4] . We investigate in this article which computational capabilities emerge in WTA circuits if one also takes into account the existence of lateral excitatory synaptic connections within such ensembles of pyramidal cells ( Fig . 1A ) . This augmented architecture will be our default notion of a WTA circuit throughout this paper . We show that this network motif endows cortical microcircuits with the capability to encode and process information in a highly dynamic environment . This dynamic environment of generic cortical mircocircuits results from quickly varying activity of neurons at the sensory periphery , caused for example by visual , auditory , and somatosensory stimuli impinging on a moving organism that actively probes the environment for salient information . Quickly changing sensory inputs are also caused by movements and communication acts of other organisms that need to be interpreted and predicted . Finally , a generic cortical microcircuit also receives massive inputs from other cortical areas . Experimental data with simultaneous recordings of many neurons suggest that these internal cortical codes are also highly dynamic , and often take the form of characteristic assembly sequences or trajectories of local network states [5]–[10] . We show in this article that WTA circuits have emergent coding and computing capabilities that are especially suited for this highly dynamic context of cortical microcircuits . We show that spike-timing-dependent plasticity ( STDP ) [11] , [12] , applied on both the lateral excitatory synapses and synapses from afferent neurons , implements in these networks the capability to represent the underlying statistical structure of such spatiotemporal input patterns . This implies the challenge to solve two different learning tasks in parallel . First it is necessary to learn to recognize the salient high-dimensional patterns from the afferent neurons , which was already investigated in [13] . The second task consists in learning the temporal structure underlying the input spike sequences . We show that augmented WTA circuits are able to detect the sequential arrangements of the learned salient patterns . Synaptic plasticity for lateral excitatory connections provides the ability to discriminate even identical input patterns according to the temporal context in which they appear . The same STDP rule , that leads to the emergence of sparse codes for individual input patterns in the absence of lateral excitatory connections [13] now leads to the emergence of context specific neural codes and even predictions for temporal sequences of such patterns . The resulting neural codes are sparse with respect to the number of neurons that are tuned for a specific salient pattern and the temporal context in which it appears . The basic principles of learning sequences of forced spike activations in general recurrent networks were studied in previous work [14] , [15] and resulted in the finding that an otherwise local learning rule ( like STDP ) has to be enhanced by a global third factor which acts as an importance weight , in order to provide a – theoretically provable – approximation to temporal sequence learning . The possible role of such importance weights for probabilistic computations in spiking neural networks with lateral inhibition was already investigated earlier in [16] . In this article we establish a rigorous theoretical framework which reveals that each spike train generated by WTA circuits can be viewed as a sample from the state space of a hidden Markov model ( HMM ) . The HMM has emerged in machine learning and engineering applications as a standard probabilistic model for detecting hidden regularities in sequential input patterns , and for learning to predict their continuation from initial segments [17]–[19] . The HMM is a generative model which relies on the assumption that the statistics of input patterns over time steps is governed by a sequence of hidden states , such that the hidden state “explains” or generates the input pattern . We show that the instantaneous state of the HMM is realized by the joint activity of all neurons of a WTA circuit , i . e . the spikes themselves and their resulting postsynaptic potentials . The stochastic dynamics of the WTA circuit implements a forward sampler that approximates exact HMM inference by propagating a single sample from the hidden state forward in time [19] , [20] . We show analytically that a suitable STDP rule in the WTA circuit – notably the same rule on both the recurrent and the feedforward synaptic connections – realizes theoretically optimal parameter acquisition in terms of an online expectation-maximization ( EM ) algorithm [21] , [22] , for a certain pair if the stochastic network dynamics describes the state sequence upon the input sequence . We further show that when the STDP rule is applied within the approximative forward sampling network dynamics of the WTA circuit , it instantiates a weak but well defined approximation of theoretically optimal HMM learning through EM . This is remarkable insofar as no additional mechanisms are needed for this approximation – it is automatically implemented through the stochastic dynamics of the WTA circuit , in combination with STDP . In this paper we focus on the analysis of this approximation scheme , its limits and its behavioral relevance . We test this model in computer simulations that duplicate a number of experimental paradigms for evaluating emergent neural codes and behavioral performance in recognizing and predicting temporal sequences . We analyze evoked and spontaneous dynamics that emerges in our model network after learning an object sequence memory task as in the experiments of [23] , [24] . We show that the pyramidal cells of a WTA circuit learn through STDP to encode the hidden states that underlie the input statistics in such tasks , which enables these cells to recognize and distinguish multiple pattern sequences and to autonomously predict their continuation from initial segments . Furthermore , we find neural assemblies emerging in neighboring interconnected WTA circuits that encode different abstract features underlying the task . The resulting neural codes resemble the highly heterogeneous codes found in the cortex [25] . Furthermore , neurons often learn to fire preferentially after specific predecessors , building up stereotypical neural trajectories within neural assemblies , that are also commonly observed in cortical activity [5]–[7] , [26] . Our generative probabilistic perspective of synaptic plasticity in WTA circuits naturally leads to the question whether the proposed learning approximation is able to solve complex problems beyond simple sequence learning . Therefore we reanalyze data on artificial grammar learning experiments from cognitive science [27] , where subjects were exposed to sequences of symbols generated by some hidden artificial grammar , and then had to judge whether subsequently presented unseen test sequences had been generated by the same grammar . We show that STDP learning in our WTA circuits is able to infer the underlying grammar model from a small number of training sequences . The simple approximation by forward sampling , however , clearly limits the learning performance . We show that the full power of HMM-learning can be attained in a WTA circuit based on the rejection sampling principle [19] , [20] . A binary factor is added to the STDP learning rule , that gates the expression of synaptic plasticity through a subsequent global modulatory signal . The improvement in accuracy of this more powerful learning method comes at the cost that every input sequence has to be repeated a number of times , until one generated state sequence is accepted . We show that a significant performance increase can be achieved already with a small number of repetitions . We demonstrate this for a simple and a more complex grammar learning task .
In this section we briefly summarize the relevant concepts for deriving our theoretical results . An exhaustive discussion on hidden Markov model theory can be found in [17]–[19] . Throughout the paper , to keep the notation uncluttered we use the common short-hand notation to denote , i . e . the probability that the random variable takes on the value . If it is not clear from the context , we will use the notation to remind the reader of the underlying random variable , that is only implicitly defined . The HMM is a generative model for input pattern sequences over time steps ( the input patterns are traditionally called observations in the context of HMMs ) . It relies on the assumption that a sequence of hidden states and a set of parameters exist , which govern the statistics of . This assumption allows to write the joint distribution of and as ( 4 ) where we suppress an explicit representation of the initial state , for the sake of brevity . The joint distribution ( 4 ) factorizes in each time step into the observation model and the state transition or prediction model [19] . This independence property is illustrated by the Bayesian network for a HMM in Fig . 1B . The HMM is a generative model and therefore we can recover the distribution over input patterns by marginalizing out the hidden state sequences . Learning in this model means to adapt the model parameters such that this marginal distribution comes as close as possible to the empirical distribution of the observable input sequences . A generic method for learning in generative models with hidden variables is the expectation-maximization ( EM ) algorithm [30] , and its application to HMMs is known as the Baum-Welch algorithm [31] . This algorithm consists of iterating two steps , the E-step and the M-step , where the model parameters are adjusted at each M-step ( for the updated posterior generated at the preceding E-step ) . A remarkable feature of the algorithm is that the fitting of the model to the data is guaranteed to improve at each M-step of this iterative process . Whereas the classical EM algorithm is restricted to offline learning ( where all training data are available right at the beginning ) , there exist also stochastic online versions of EM learning . In its stochastic online variant [21] , [22] the E-step consists of generating one sample from the posterior distribution , given one currently observed input sequence . Given these sampled values for , the subsequent M-step adapts the model parameters such that the probability increases . The adaptation is confined to acquiring the conditional probabilities that govern the observation and the prediction model . It would be also desirable to realize the inference and sampling of one such posterior sample sequence in a fully online processing , i . e . generating each state in parallel to the arrival of the corresponding input pattern . Yet this seems to be impossible as the probabilistic model according to ( 4 ) implies a statistical dependence between any and the whole future observation sequence . However , it is well known that the inference of can be approximated by a so-called forward sampling process [19] , [20] , where every single time step of the sequence is sampled online , based solely on the knowledge of the observations received so far , rather than the observation of the complete sequence . Hence sampling the sequence is approximated by propagating a single sample from the HMM state space forward in time . In this section we show that the dynamics of the network realizes a forward sampler for the HMM . We make use of the fact that equations ( 1 ) , ( 2 ) and ( 3 ) realize a Markov process , in the sense that future network dynamics is independent from the past , given the current network state ( for a suitable notion of network state ) . This property holds true for most reasonable choices of EPSP kernels . For the sake of brevity we focus in the theoretical analysis on the simple case of a single exponential decay with time constant . We seek a description of the continuous-time network dynamics in response to afferent spike trains over a time span of length that can be mapped to the state space of a corresponding HMM with discrete time steps . Although the network works in continuous time , its dynamics can be fully described taking only those points in time into account , where one of the neurons in the recurrent circuit produces a spike . This allows to directly link spike trains generated by the network to a sequence of samples from the state space of a corresponding HMM . Let the spike times produced during this time window be given by . The neuron dynamics are determined by the membrane time courses ( 2 ) . For convenience let us introduce the notation , with and by analogy , with . Due to the exponentially decaying EPSPs the synaptic activation at time is fully defined by the synaptic activation at the time of the previous spike , and the identity of the neuron that spiked in that previous time step , which we denote by a discrete variable . We thus conclude that the sequence of tuples ( with ) fulfills the Markov condition , i . e . the conditional independence and thus fully represents the continuous dynamics of the network ( see Methods ) . We call the network state . The corresponding HMM forward sampler follows a simple update scheme that samples a new state given the current observation and the previous state . This dynamic is equivalent to the WTA network model . This state representation allows us to update the network dynamics online , jumping from one spike time to the next . Using this property , we find that the dynamics of the network realizes a probability distribution over state sequences , given an afferent sequence , which can be written as ( 5 ) where is the set of network parameters . The factorization and independence properties in ( 5 ) are induced by the state representation and the circuit dynamics . We assume here that the lateral inhibition within the WTA circuit ensures that the output rate of the whole circuit is normalized , i . e . at all times . This allows to introduce the distribution over the inter-spike-time intervals independent from ( see Methods for details ) . Note , that determines the interval between spikes of all circuit neurons , realized by a homogeneous Poisson process with a constant rate . The second term in the second line of ( 5 ) determines the course of the membrane potential , i . e . it assures that follows the membrane dynamics . Since the EPSP kernels are deterministic functions this distribution has a single mass point , where ( 2 ) is satisfied . The first factor in the second line of ( 5 ) is given by the probability of each individual neuron to spike . This probability depends on the membrane potential ( 1 ) , which in turn is determined by , and the network parameters . Given that the circuit spikes at time , the firing probability of neuron can be expressed as a conditional distribution . The lateral inhibition in ( 1 ) ensures that this probability distribution is correctly normalized . Therefore , the winner neuron is drawn from a multinomial distribution at each spike time . For the given architecture the functional parts of the network can be related directly to hidden Markov model dynamics . In the Methods section we show in detail that by rewriting the membrane potential ( 1 ) can be decomposed into three functional parts ( 6 ) The lateral excitatory connections predict a prior belief about the current network activity and the feedforward synapses match this prediction against the afferent input . The inhibition implements the normalization that is required to make ( 6 ) a valid multinomial distribution . The functional parts of the membrane potential can be directly linked to the prediction and observation models of a HMM , where the network state is equivalent to the hidden state of this HMM . The WTA circuit realizes a forward-sampler for this HMM , which approximates sampling from the posterior distribution in an online fashion [20] . Its sampling is carried out step by step , i . e . it generates through each spike a new sample from the network state space , taking only the previous time step sample into account . Furthermore this forward sampling requires no additional computational organization , but is achieved by the inherent dynamics of the stochastically firing WTA circuit . Formulating the network dynamics in terms of a probabilistic model is beneficial for two reasons: First , it gives rise to a better understanding of the network dynamics by relating it to samples from the HMM state space . Second , the underlying model allows us to derive parameter estimation algorithms and to compare them with biological mechanisms for synaptic plasticity . For the HMM , this approach results in an instantiation of the EM algorithm [19] , [30] in a network of spiking neurons ( stochastic WTA circuit ) . In the Methods section we derive this algorithm for the WTA circuit and show that the M-step evaluates to weight updates that need to be applied whenever neuron emits a spike at time , according to ( 7 ) where is a positive constant that controls the learning rate . Note that the update rules for the feedforward and the recurrent connections are identical , and thus all excitatory synapses in the network are handled uniformly . These plasticity rules ( 7 ) are equivalent to the updates that previously emerged as theoretically optimal synaptic weight changes , for learning to recognize repeating high-dimensional patterns in spike trains from afferent neurons , in related studies [13] , [32] , [33] . The update rules consist of two parts: A Hebbian long-term potentiating ( LTP ) part that depends on presynaptic activity and a constant depression term . The dependence on the EPSP time courses ( 2 ) makes the first part implicitly dependent on the history of presynaptic spikes . The STDP window is shown in Fig . 1C for -shaped EPSPs . Potentiation is triggered when the postsynaptic neuron fires after the presynaptic neuron . This term is commonly found in synaptic plasticity measured in biological neurons , and for common EPSP windows it closely resembles the shape of the pre-before-post part of standard forms of STDP [11] , [12] . The dependence on the current value of the synaptic weight has a local stabilizing effect on the synapse . The depressing part of the update rule is triggered whenever the postsynaptic neuron fires independent of presynaptic activity . It contrasts LTP and assures that the synaptic weights stay globally in a bounded regime . It is shown in Fig . 4 of [13] that the simple rule ( 7 ) reproduces the standard form of STDP curves when it is applied with an intermediate pairing rate . While these M-step updates emerge as exact solutions for the underlying HMM , the WTA circuit implements an approximation of the E-step , using forward sampling from the distribution in equation ( 5 ) . In the following experiments we will first focus on this simple approximation , and analyze what computational function emerges in the network using the STDP updates ( 7 ) without any third signal related to reward or a “teacher” . In the last part of the Results section we will introduce a possible implementation of a refined approximation , and assess the advantages and disadvantages of this method . In this section we show through computer simulations that our WTA circuits learn to encode the hidden state that underlies the input statistics via the STDP rule ( 7 ) . We demonstrate this for a simple sequence memory task and analyze in detail how the hidden state underlying this task is represented in the network . The experimental paradigm reproduces the structure of object sequence memory tasks , where monkeys had to memorize a sequence of movements and reproduce it after a delay period [23] , [24] , [34] , [35] . The task consisted of three phases: An initial cue phase , a delay phase and a recall phase . Each phase is characterized by a different input sequence , where the cue sequence defines the identity of the recall sequence . We used four cue/recall pairs in this experiment . The structure of this task is illustrated in Fig . 2A . The graph represents a finite state grammar that can be used to generate symbol sequences by following a path from Start to Exit . In this first illustrative example the only stochastic decision is made at the beginning , randomly choosing one of the four cue phases with equal probabilities while the rest of the sequence is deterministic . On each arc that is passed , the symbol next to the arc is generated , e . g . AB-delay-ab is one possible symbolic sequence . Note that all symbols can appear in different temporal contexts , e . g . A appears in sequence AB-delay-ab and in BA-delay-ba . The delay symbol is completely unspecific since it appears in all four possible sequences . Therefore this task does not fulfill the Markov condition with respect to the input symbols , e . g . knowing that the current symbol is delay does not identify the next one as it might be any of a , b , c , d . Only additional knowledge about the temporal context of the symbol allows to uniquely identify the continuation of the sequence . This additional knowledge can be represented in a hidden state that encodes the required information , which renders this task a simple example of a HMM . The hidden states of this HMM have to encode the input patterns and the temporal context in which they appear in order to maintain the Markov property throughout the sequences , e . g . a distinct state encodes pattern B when it appears in sequence AB-delay-ab . The temporal structure of the hidden state can be related to the finite state grammar in Fig . 2A . The arcs of the grammar directly correspond to the hidden states , i . e . given knowledge about the currently visited arc allows us to complete the sequence . The symbols next to the arcs define the observation model , i . e . the most likely symbol throughout each state . In this simple symbolic HMM the observation model is in fact deterministic , since exactly one symbol is allowed in each state . In the neural implementation of this task , the symbolic sequences are presented to the WTA circuit encoded by afferent spike trains . Every symbol A , B , C , D , a , b , c , d , delay is represented by a rate pattern with fixed length of 50 ms , during which each afferent neuron emits spikes with a symbol-specific , fixed Poisson rate ( see Methods ) . One example input spike train encoding the symbolic sequence AB-delay-ab is shown in the top panel of Fig . 2A . The input spike times are not kept fixed but newly drawn for each pattern presentation . This input encoding adds extra variability to the task , which is not directly reflected by the simple symbolic finite state grammar . Still , the statistics underlying the input sequences follow the dynamics of a HMM of the form ( 4 ) , and therefore our WTA circuit and the spike trains that encode sequences generated by the artificial grammar share a common underlying model . The observation model of that HMM covers the uncertainty induced by the noisy rate patterns by assigning a certain likelihood to each observed input activation . The hidden state representation has to encode the context-dependent symbol identity and the temporal structure of the sequences , i . e . the duration of each individual symbol . In our continuous-time formulation the hidden state is updated at the time points . Therefore , throughout the presentation of a rate pattern of 50 ms length , several state updates are encountered during which the hidden state has to be maintained . In principle this can be done by allowing each hidden state to persist over multiple update steps by assigning non-zero probabilities to . However , this approach is well known to result in a poor representation of time as it induces an exponential distribution over the state durations , which is inappropriate in most physical systems and obviously also for the case of deterministic pattern lengths , considered here [17] , [19] . The accuracy of the model can be increased at the cost of a larger state space by introducing intermediate states , e . g . by representing pattern B in sequence AB-delay-ab by an assembly of states that form an ordered state sequence throughout the pattern presentation . Each of these assemblies encodes a specific input pattern , the temporal context and its sequential structure throughout the pattern , and with sufficiently large assemblies the temporal resolution of the model achieves reasonable accuracy . We found that this coding strategy emerges unsupervised in our WTA circuits through the STDP rule ( 7 ) . To show this , we trained a WTA circuit with afferent cells and circuit neurons by randomly presenting input spike sequences until convergence . In this experiment , the patterns were presented as a continuous stream of input spikes , without intermediate pauses or resetting the network activity at the beginning of the sequences . Training started from random initial weights , and therefore the observation and prediction model had to be learned from the presented spike sequences . Prior to learning the neural activity was unspecific to the patterns and their temporal context ( see Fig . 2B ) . Fig . 2C shows the evoked activities for all four sequences after training . The output of the network is represented by the perievent time histogram ( PETH ) averaged over 100 trial runs and a single spike train that is plotted on top . To simplify the interpretation of the network output we sorted the neurons according to their preferred firing times ( see Methods ) . Each sequence is encoded by a different assembly of neurons . This reflects the structure of the hidden state that underlies the task . Since the input is presented as continuous spike train , the network has also learned intermediate states that represent a gradual blending between patterns . About 25 neurons were used to encode the information required to represent the hidden state of each sequence . This coding scheme installs different representations of the patterns depending on the temporal context they appeared in , e . g . the pattern delay within the sequence AB-delay-ab was represented by another assembly of neurons than the one in the sequence BA-delay-ba . Small assemblies of about five neurons became tuned for each pattern and temporal context . This sparse representation emerged through learning and is not merely a consequence of the inherent sparseness of the WTA dynamics . Prior to learning all WTA neurons are broadly tuned and show firing patterns that are unordered and nonspecific ( see Fig . 2B ) . After learning their afferent synapses are tuned for specific input patterns , whereas the temporal contexts in which they appear are encoded in the excitatory lateral synapses . The latter can be seen by inspecting the synaptic weights shown in Fig . 2D . They reflect the sparse code and also the sequential order in which the neurons are activated . They also learned to encode the stochastic transitions at the beginning of the cue phase , where randomly one of the four sequences is selected . These stochastic switches are reflected in increased strength of synapses that connect neurons activated at the end and the beginning of the sequences . The behavior of the circuit is further examined in Fig . 3 . The average network activity over 100 trial runs of the neurons that became most active during sequence AB-delay-ab are shown in Fig . 3A . In addition the spike trains for 20 trials are shown for three example neurons . The same sorting was applied as in Fig . 2 . Using the hidden state encoded by the network it should be possible to predict the recall patterns after seeing the cue , if it correctly learned the input statistics . We demonstrate this by presenting incomplete inputs to the network . After presentation of the delay pattern the input was turned off and the network was allowed to run freely . The delay pattern was played three times longer than in the training phase . During this time the network was required to store its current state ( the identity of the cue sequence ) . After this delay time the input was turned off – no spikes were generated by the afferent neurons during this phase , the network was purely driven by the lateral connections . Since the delay time was much longer than the EPSP windows the network had to keep track of the sequence identity in its activity pattern throughout this time to solve the task . Fig . 3B shows the output behavior of the network for sequence AB-delay-free ( where free denotes a time window with no external input ) . After the initial sequence AB was presented , a small assembly of neurons became active that represents the delay pattern that was associated with that specific sequence . After the delay pattern was turned off , the network completed the hidden state sequence using its memorized activity , which can be seen by comparing the evoked and spontaneous spike trains in Fig . 3A and B , respectively . In order to quantify the ability of the network to reproduce the structure of the hidden state , we evaluated the similarity between the spontaneous and evoked network activity using the rank order correlation coefficient , which is a similarity measure normalized between and , where means that the order is perfectly preserved . This measure has been previously proposed to detect stereotypical temporal order in neural firing patterns [6] . Fig . 3C shows the histograms over the correlation coefficients for all four sequences . The histograms were created by calculating the rank order correlation between the spontaneous sequences and the PETH of the evoked sequences . It can be seen that the temporal order of the evoked sequence was reliably reproduced during the free run . To that end , for each of the input sequences , a stable representation has been trained into the network , that is encoded in the lateral synapses . This structure emerged completely unsupervised using the local STDP rule , solely from the intrinsic dynamics of the network . The first experiment demonstrated that through STDP , single neurons of a WTA circuit get tuned for distinct input patterns and the temporal context in which they appear . The neural code that emerged is reminiscent of some features found in cortical activity of monkeys solving similar tasks , namely the emergence of context cells that respond specifically to certain symbols when they appear in a specific temporal context [34] , [36] , [37] . However , the overall competition of a single WTA circuit hinders the building of codes for more abstract features , which are also found in the cortex in the very same experiments where neurons in the same cortical area encode different functional aspects of stimuli and actions . They seem to integrate information on different levels of abstraction which results in a diverse and rich neural code , where close-by neurons are often tuned to different task-related features [25] . We show that our model reproduces this mixed selectivity of cortical neurons if multiple interconnected WTAs are trained on a common input . The strong competition is restricted to neurons within every single WTA , whereas there is no competition between neurons of different circuits and lateral connections allow full information exchange between the circuits . Therefore , the model is extended by splitting the network into smaller WTA groups , each of which receives input from a distinct inhibitory feedback loop that implements competition between members of that group . In addition all neurons receive lateral excitatory input from the whole network . Every WTA group still follows the dynamics of a forward sampler for a HMM . Each of these WTA circuits adapts its synaptic weights through STDP to best represent the observed input spike sequences . In addition , the lateral connections between WTA groups introduce a coupling between the network states of individual groups . The dynamics of the whole network of WTA circuits can be understood as a forward sampler for a coupled HMM [38] , where every WTA group encodes one multinomial variable of a compound state such that from one time step to the next all single state variables have influence on each other [20] , [38] . In the first experiment we have seen that the WTA circuit learned to use about of the available neurons to encode each of the four sequences . We have also seen that the network used small assemblies of neurons to represent each of the patterns in favor of a finer temporal resolution . This implies that WTA circuits of different size can learn to decode the input sequence on different levels of detail , where small circuits only learn the most salient features of the input sequences . To show this we trained a network with WTA groups of random size between 10 and 50 units , giving a total network size of , on the simple object sequence memory task ( Fig . 2A ) . The neural code that emerges in this network after training is shown in Fig . 4 . The output rates of the circuit neurons were measured during the presentation of pattern a appearing in the sequence AB-delay-ab , BA-delay-ba , shown in Fig . 4A , B respectively . Three classes of neurons can be distinguished: 10 neurons were tuned to pattern a in the context AB-delay-ab only ( shown in red ) , 12 neurons were tuned to pattern a exclusievly in the context BA-delay-ba ( shown in blue ) and 5 additional neurons encode pattern a independent of its context ( green ) , i . e . they get activated by the pattern a in both sequences AB-delay-ab and BA-delay-ba . The remaining neurons were not significantly tuned for pattern a ( average firing rate during pattern a was less than , not shown in the plot ) . To pinpoint the computational function that emerged in the network we compared the spontaneous activity of individual neurons from different WTA circuits . Spike trains for one context-specific and one non-specific neuron are compared in Fig . 4C and D , respectively . Both panels show spike raster plots over 20 trial runs and averaged neuron activities ( PETH ) for sequences AB-delay-free and BA-delay-free . The neuron in Fig . 4C belongs to a small WTA group with a total size of 15 neurons and shows context unspecific behavior , whereas the neuron in Fig . 4D which belongs to a larger WTA group ( 42 neurons ) is context specific ( see Fig . 4A , B ) . This behavior is also reproduced during the free run , when the neurons are only driven by their lateral synapses . The neuron in Fig . 4D remains silent during BA-delay-free and thus shows the properties of context cells observed in the cortex , whereas the neuron in Fig . 4C is active during both sequences . Still , during spontaneous replay that neuron correctly reproduces the temporal structure of the input sequences . In sequences starting with AB the neural activity peaks at after the onset of the free run – the time pattern a was presented in the evoked phase . If the sequence starts with BA this behavior is modulated and the activity is delayed by roughly 50 ms , to the time point a would appear in the recall phase . The required information to control this modulation was not available within the small WTA group the neuron belongs to , but provided by neighboring context-specific neurons from other groups . To see this we trained a linear classifier on the evoked activity during the delay phase of AB-delay-ab and BA-delay-ba ( see Methods for details ) . If the neurons reliably encode the sequence identity a separating plane should divide the -dimensional space of network activities between the sequences . Training the classifier only on the 15-dimensional state space of the group the neuron in Fig . 4C belongs to , did not reveal such a plane ( the classification performance was ) . Therefore , this small WTA circuit did not encode the required memory item to distinguish between the two sequences after the delay phase . However , the whole network of all WTA groups reliably encoded this information and the classifier trained on the -dimensional state space could distinguish between the delay phases of AB-delay-ab and BA-delay-ba with accuracy . To illustrate the different emergent representations , we compared linear projections of the state of the small WTA group with 15 neurons and the state of the whole network in Fig . 4E , F , respectively . The plots show the network activity during the delay phase for all four sequences . Each line corresponds to a trajectory of the evoked network activity , where the line colors indicate the sequence identity . The state trajectories were projected onto the first two dimensions of the dynamic principal component analysis ( jPCA ) , that was recently introduced as an alternative to normal PCA that is applicable to data with rotational dynamics [39] . Empirically , we found this analysis method superior to normal PCA in finding linear projections that separate the network states for different input sequences . One explanation for this lies in the dynamical properties of WTA circuits . Due to the global normalization which induces a constant network rate , the dynamics of the network are roughly energy-preserving . Since this implies that the corresponding linear dynamical system is largely non-expanding/contracting , a method that identifies purely rotational dynamics such as the jPCA was found to be beneficial here . Fig . 4E shows the first two jPCA components of the neural activities during the delay phase for the WTA circuit with 15 neurons , which the neuron in Fig . 4C belongs to . This circuit was not able to distinguish between all four input sequences , since it activated the same neurons to encode them . This is also reflected in the jPCA projections shown in Fig . 4E , which show a large overlap for sequences AB-delay-ab and BA-delay-ba . On the other hand , the network state comprising all neurons reliably encoded the sequence identities ( see Fig . 4F ) . The delay state for each sequence spans an area in the 2-D projection and therefore the network found a state space that allows a linear separation between the sequences . Such a representation is important since the neuron model employs a linear combination of the network state in the membrane dynamics ( 1 ) and therefore provides the information required by the neurons in Fig . 4C , D to modulate their spontaneous behavior . Information about transient stimuli is often kept available over long time spans in trajectories of neural activity in the mammalian cortex [5] , [7] , [26] , [40] and in songbirds [41]–[43] . In the previous experiment we saw that our model is in principle capable to develop such trajectories in neural assemblies ( see Fig . 3B ) , which emerged to encode salient input patterns and the temporal structure throughout them . However , in that experiment the input sequences comprised a rich temporal structure , since each pattern was only shown for a time bin which might have facilitated the development of these activity patterns . In this section we study whether a similar behavior also emerges when the input signal is stationary over long time spans . In analogy to the previous experiment we generated two input sequences A-delay and B-delay . The patterns A , B were played for and the pattern delay for 500 ms . As in all other experiments , the patterns were rate patterns , i . e . each input neuron fired with a constant Poisson rate during the pattern and spike times were not kept fixed throughout trials . One example input spike train is shown in Fig . 5A . Although the input was stationary for during the delay pattern , we could still observe the emergence of neural trajectories in the network after training . Again , we used a network composed of multiple interconnected WTA circuits to learn these patterns . We employed a network of WTA groups of random size in the range from 10 to 100 neurons . The total network had a size of circuit neurons and we used afferent cells . Fig . 5B shows the sorted average output activity after training . For each of the two sequences a distinct assembly of neurons emerged and the neurons composing these assemblies fired in a distinct sequential order . Fig . 5C shows the rank order correlations between the evoked and spontaneous activities . The trajectories of neural firing were reliably reproduced during spontaneous activity , but only about 100 neurons were used for each of the two assemblies , leaving the remaining 500 neurons ( almost ) perfectly silent . The emergence of these trajectories can be further enhanced using a homeostatic intrinsic plasticity mechanism which enforces that on average all network neurons participate equally in the representation of the hidden state . This can be achieved by a mechanism that regulates the excitability of each neuron , such that the overall output rate of neuron ( measured over a long time window ) converges to a given target rate . ( see [44] and the Methods section ) . Augmenting the dynamics of the network with this intrinsic plasticity rule prevents neurons from becoming inactive if their synaptic weights decrease and by that assures that each neuron joins one of the assemblies . This can be seen in Fig . 5C , D which shows the output activity after training with STDP augmented with the homeostatic mechanism . The neurons formed a fixed ordered sequence and thus showed a clear preference for a certain point in time within the pattern . Even though the delay pattern had no salient temporal structure ( the rates of all afferent neurons were constant throughout the pattern ) these trajectories were formed by imprinting the sequential order of the neural activity into the lateral excitatory connections . As in the first experiment each neuron has learned to fire after a distinct group of preferred predecessors , resulting in neural trajectories through the network . Therefore , the time that has elapsed since the delay pattern started could be inferred from the neural population activity . In addition the identity of the initial pattern was also memorized , since about half of the population became active for each of the two sequences . The finite state grammar used in the previous experiments ( Fig . 2A ) did not utilize the full expressive power of HMMs since it only allowed stochastic switches at the beginning of each sequence . In this section we consider the problem of learning more general finite state grammars in WTA circuits , a problem that has also been extensively studied in cognitive science in artificial grammar learning ( AGL ) experiments [45] . Fig . 6A shows the artificial grammar that was used in [27] to train subjects using different stimulus modalities ( visual , auditory and tactile ) . There it was shown that humans can acquire the basic statistics of such grammars extremely fast . On this particular task humans showed a performance of 62% to 75% percent ( depending on the stimulus modality that was used ) after only a few dozens of stimulus presentations [27] . We show that our network model can extract the basic structure of this grammar . This internal representation can be subsequently used to classify unseen sequences as grammatical or not . Through STDP the network adapts the parameters such that they reflect the statistics underlying the training sequences , and the emergent HMM can then be used to evaluate the sequence likelihood . The ability of the network to distinguish between grammatical and ungrammatical sequences was assessed by applying a threshold on the sequence log-likelihood , an approximation of which was computed over a single sample from ( 5 ) ( see Methods ) . The threshold was assigned to the mean of the log-likelihood values computed for all test sequences . Likelihoods that laid above that threshold were reported as grammatical . In this experiment we used a sparse input coding , where only a small subset of afferent neurons is activated for each of the symbols . This representation could be realized by another WTA circuit used as input for the network to decode more complex input patterns . We trained a single WTA circuit with neurons on this sparse input . Using this model , we were able to achieve high learning speeds . In each training iteration one of the 12 training data sets from [27] ( using only the first sequence of each match/mismatch pair ) was chosen at random and presented to the network . For testing we used the 20 test sequences from [27] to evaluate the learning performance . Training was interrupted after every sequence presentation to assess the classification performance . The resulting learning curve is shown in Fig . 6B . The classification rate of that was reported in the behavioral experiment was exceeded after only 80 iterations . By training the network beyond this point performances up to were reached . Note that none of the training sequences appeared in the test set . Therefore the network has not just learned a fixed set of sequences , but extracted relevant statistical features that allowed it to generalize to new data . So far in all experiments the simple forward sampling approximation was used for learning the model parameters . Although this learning paradigm has shown to be surprisingly powerful , it is limited and will not be sufficient if the network is required to learn more complex tasks or acquire probabilistic models with a high level of detail . In this section we derive the refined approximation toward evaluating the HMM E-Step in a recurrent WTA circuit based on rejection sampling . Exactly solving the E-step requires to evaluate the posterior probability of , given by ( 8 ) where is the HMM joint distribution , given by equation ( 4 ) . A stochastic EM update is realized by drawing a state sequence from the posterior for which the M-step parameter updates are performed . However , directly sampling from ( 8 ) is not possible for a spiking neural network , since it requires the integration of information over the whole state sequence and thus , looking into the future . This can be seen by noting that the integral in ( 8 ) runs over the state space of the whole sequence . To that end , the network is not able to sample from this distribution directly . Nevertheless , it is possible to indirectly evaluate ( 8 ) using samples generated from ( 5 ) , which can be expressed by ( 9 ) where denotes the expected value over , which in this context is called a proposal distribution since it is used to propose samples , which are then used to indirectly evaluate the target distribution . The scalar is the importance weight between the target and the proposal distribution , which is used to scale the influence of the sample [19] , [20] , [46] . The expectation in the denominator of ( 9 ) is again not easy to evaluate , since it requires us to integrate over multiple sequences . The most pragmatic solution to this problem is to approximate this term using a single sample from the proposal distribution . Under this approximation the importance weight in ( 9 ) cancels out and we arrive at the trivial approximation , i . e . each sample from the proposal distribution is accepted as a valid sample from the posterior . This is the forward sampling approximation that was used so far throughout all experiments . In order to improve this approximation we use the stochasticity of the network , which assures that different state sequences are proposed if the same input sequence is presented several times . Rejection sampling utilizes this stochasticity and preferentially selects sequences with high likelihood throughout the whole input . The required information to do this selection is a global quantity that must be tracked over the whole sequence . The probability to accept a state sequence is directly proportional to the importance weight , which computes to ( 10 ) Note that ( 10 ) can be easily computed forward in time , since in each time step , it only needs to be updated using the instantaneous input likelihood . Further note that this is a measure for surprise or prediction error – the probability of observing the current input given the previous state . The information to decide whether to accept is the accumulated prediction error over the whole sequence . This approach also naturally extends to the case of multiple interconnected WTAs . There , the contributions to the importance weight of every single circuit have to be multiplied in every time step and therefore , a possible rejection is in that case effective for the whole network of all WTAs at once . Since the importance weights need to be accumulated over the whole sequence of spike events of length , the weight update rules ( 7 ) can not be applied instantaneously . In the neural implementation we achieved this using a synaptic eligibility trace as proposed in [47] . Instead of updating the weights directly they are tagged and consolidation of the tags is delayed until the whole sequence is read . The probability to accept these tags is proportional to the importance weights , i . e . ( 11 ) where is a constant that scales the acceptance rate . If a sequence is accepted , the synaptic tags are consolidated . If the circuit decides not to accept , the synaptic weight changes for the whole sequence have to be discarded . This result is analytically similar to [15] , where the importance weights ( 10 ) were introduced by weighting the eligibility traces with a deterministic scalar factor ( importance sampling ) . Here , in the rejection sampling framework a stochastic variant of this method is used . The advantage of the rejection sampling method is that it is not necessary to explicitly compute the normalization in ( 8 ) . The normalization can be approximated by replaying in every training iteration the input sequence multiple times until it gets accepted once , instead of using a constant number of replays as with importance sampling . In practice however it is necessary to adapt the parameter throughout learning in order to get a reasonable number of replays . We used a simple linear tracking mechanism for throughout the experiments ( see Methods ) . A performance comparison of these different sampling approximations is provided at the end of the Results section . We assume that the circuit interacts with a mechanism that allows the replay of the afferent stimulus multiple times . By enforcing that each input is accepted once , we guarantee that the network learns the statistics of all input sequences with equal accuracy . This view allows us to make an interesting theoretical prediction: when an input is not well represented by the network it is more likely to be rejected and therefore , the number of rejected and resampled sequences represents a notion of novelty . Literally speaking , the network pays more attention to novel inputs , by resampling them multiple times ( see Methods for details ) . In the following experiments we investigate the possible performance gain that can be achieved if the network has access to this rejection sampling mechanism . We have previously seen that the grammar from Fig . 1 in [27] can be learned almost perfectly using pure forward sampling . However , this data set had a very simple structure . To distinguish between grammatical and ungrammatical sequences only required the analysis of the local statistics of the input . E . g . it is easy to see that the sequence DEAC is not grammatical since it contains the bigram DE , which never appears in the training data . Each of the ungrammatical sequences contains at least one illegal bigram and thus can be classified based on a simple model of symbol transitions . This simple structure was already recovered with the online learning scheme and therefore using rejection sampling on that task did not result in a significant performance increase ( see Fig . 6 ) . To demonstrate the advantage of rejection sampling , we created a grammar that required integration of information over a longer time span , shown in Fig . 7A . Although this grammar only allows to create four sequences AABC , BBAC , ABAD and BABD , the underlying structure is more complex than in the previous tasks . The identity of the last symbol can only be inferred if the identity and context of the first symbol is integrated and memorized over the whole sequence . To that end , the rejection sampling algorithm that allows the network to propagate information over the whole sequence , should bring a definite benefit over forward sampling for this task . The quantity that is needed to update the importance weights ( 10 ) and also to estimate the sequence likelihood for classifying grammatical against ungrammatical inputs , is given by the instantaneous input likelihood ( see Methods ) . As pointed out earlier , this quantity is a measure for surprise , i . e . the probability of observing the current input pattern given the network state . The ability of the network to exploit this prediction error to classify sequences is illustrated in Fig . 7 . The input-output behavior of a network after training with rejection sampling is shown for the grammatical sequence BBAC and the ungrammatical sequence BBAD , in Fig . 7C , D respectively . The bottom plots show traces of the instantaneous input log-likelihood . Throughout the grammatical sequence in Fig . 7C the trace stays near baseline , which indicates that the network is capable of predicting the sequence . Within the patterns , the trace only shows small deviations due to input noise . Switches between the input patterns e . g . at the border from pattern A to C cause modest levels of surprise , due to the sudden change of the network state . However , the illegal transition to pattern D in Fig . 7D causes a strong negative peak . At this point the network is not capable of predicting the final pattern . Thus the input is assigned to a low overall sequence likelihood and will therefore be classified as ungrammatical . In the rejection sampling algorithm this quantity is also used throughout training , to learn preferably from sequences that are best capable of predicting the input sequences . To quantify the advantage of this method over online learning we compared the performance on the AGL task . As in the previous experiment , the ability of the network to distinguish between grammatical and ungrammatical sequences was evaluated by applying a threshold on the sequence likelihood . The threshold was assigned to the mean of the log-likelihood values computed for all tested sequences . The network parameters were tuned such that the number of rejected samples in each iteration , averaged over the whole training session was equal to the desired number of samples ( see Methods ) . The classification errors are compared in Fig . 7B for learning with forward and rejection sampling . The parameter that scales the number of rejected samples was tracked to give an average number of 10 rejected samples per iteration . Despite this relatively small number of times the sequences is resampled , it can be seen that the performance on this task significantly increased with rejection sampling . Online learning achieved a classification rate of . With rejection sampling the network achieved classification rate . Hence we confirmed , that having access to the rejection sampling mechanism allows the network to learn the input statistics with higher levels of accuracy . Furthermore , for the example given here , this was achieved with a relatively small average number of resampled state sequences . In order to give a quantitative notion of how the sampling approximations affect the learning performance , we applied the methods to solve a generic HMM learning task . To allow a direct comparison with standard machine learning algorithms for HMMs , we used a time-discrete version of our model in this section . Therefore , we set the inter-spike-intervals to a fixed constant value and used rectangular EPSP kernels of the same length . With this modification our model is equivalent to a discrete input , discrete state HMM , commonly considered in the machine learning literature [19] . We created random HMMs and used them to generate a training and a test data set . Using this data we compared the training performance of different approximation algorithms . The accuracy of the rejection sampling algorithm crucially depends on how the parameter in equation ( 11 ) is selected . If it is set to a very large constant value , every sample gets accepted and we arrive at the simple forward sampling approximation . We compared this forward sampling algorithm with the simple tracking algorithm that was used in the previous experiment and with the optimal mechanism , which computes over a batch of sampled sequences ( see Methods ) . In addition we compared these methods with the importance sampling algorithm considered in [15] , where the scalar values of the importance weights were directly used to weight synaptic tags . All sampling methods were compared for an average number of 10 and 100 resampled sequences . Furthermore we applied standard EM learning for HMMs ( the Baum-Welch algorithm ) as reference method [19] , [31] . The results of the eight different training algorithms are compared in Fig . 8 . The figure shows the log-likelihood on the test data averaged over the 50 learning trials . As can be seen , pure forward sampling shows poor performance on this task compared with Baum-Welch learning , but with increasing number of samples the approximation approaches the performance of the exact EM updates . Interestingly we found that importance sampling and rejection sampling show almost the same performance . We believe that the reason for this lies in the high variance of the importance weights . The weights of consecutive samples can differ several orders of magnitude . After normalization , effectively only the sample with the highest importance weight has non-zero influence on the weight updates . Therefore the two algorithms are numerically almost identical for the task considered here . Using the tracking mechanism for resulted in decreased performance compared to the exact algorithm . Still , a significant performance gain can be observed with increased average number of samples .
The close relation between HMMs and recurrent neural networks was previously discovered and employed for deriving models for Bayesian computation in the cortex . These studies targeted the implementation of Bayesian filtering [56] , [57] , capturing the forward message of the belief propagation algorithm in a rate-based neural code , or using a two-state HMM to capture the dynamics of single neurons [58] , [59] . In the present study we directly analyzed spikes produced by WTA circuits in terms of samples from the state space of a HMM . For the HMM this results in an arguably weaker form of inference than belief propagation , but led in a straightforward manner to an analysis of learning in the network . The emergence of predictive population codes in recurrent networks through synaptic plasticity and their importance for sequence learning was previously suggested and experimentally verified [60] , [61] . In [62] it was shown that spiking neurons can learn the parameters of a 2-state HMM using synaptic plasticity , thereby implementing an online EM algorithm [63] , [64] . In [14] learning of temporal models was implemented through a variational approximation , and revealed STDP-like learning rules . In [15] it was shown that a network of neurons can learn to encode and reproduce a sequence of fixed spike times . The learning rules were derived using an importance sampling algorithm that yielded synaptic updates similar to the third-factor STDP rule presented here . The crucial difference between [15] and our approach is the usage of WTA circuits as building blocks for the recurrent network instead of individual neurons . Due to the possibility to use multiple WTAs our model has the freedom to factorize the multinomial HMM state space into smaller coupled variables , whereas [15] always fully factorizes the state space down to single binary variables . However , under the assumption of linear neurons the state-transition probabilities in all these models are always represented by only recurrent synapses . Thus the expressive power of all these models ( with the same number of neurons ) should be more or less identical . The optimal factorization of the state space may strongly depend on the task . Our experiments suggest that the restriction on the number of possible activity patterns due to the usage of WTAs seems minor compared to the crucial advantage of their intrinsic stabilizing effects of the network's activity . To the best of our knowledge this stabilization is the reason why the pure forward sampling learning approach performed so well in our experiments . The theoretical framework that we have introduced in this article provides a new and more principled understanding for the role of STDP in a generic cortical microcircuit motif ( ensembles of pyramidal cells with lateral excitation and inhibition ) : Even in the absence of global signals related to reward , STDP installs in these microcircuit motifs an approximation to a HMM through forward sampling . The underlying theoretical analysis provides a new understanding of the role of spikes in such WTA circuits as samples from a ( potentially very large ) set of hidden states that enable generic cortical microcircuits to detect generic neural codes for afferent spike patterns that can reflect their temporal context and support predictions of future stimuli . A remarkable feature of our model is that it postulates that noise in neural responses plays a very important role for the emergence of such “intelligent” temporal processing: We have shown that it provides in WTA circuits the basis for enabling probabilistic inference and learning through sampling , i . e . through an “embodiment” of probability distributions through neural activity . Thus stochasticity of neural responses provides an interesting alternative to models for probabilistic inference in biological neural systems through belief propagation ( see [65] for a review ) , i . e . through an emulation of an inherently deterministic calculation . The rejection sampling algorithm that was proposed here as a method for emulating the full power of HMM learning requires in addition a mechanism that allows to replay input patterns multiple times . Such replay of complex spatiotemporal patterns is well documented in the hippocampus and was proposed as a mechanism for memory consolidation in the cortex [66] . This view is also supported by findings that showed that coordinated reactivation of temporal patterns can be observed in the cortex [8] , [9] , [67] , [68] . In our framework , samples generated by the WTA circuit must be replayed several times until the network produces a spike train that provides a sequence of hidden states that gives satisfactory explanations and predictions for all segments of the sequence . The number of times a sequence is replayed is proportional to the prediction error accumulated over the sequence , which is a measure for the sample quality . Thus , sequences that are novel and to that end not well represented in the network should be replayed more often and thus , they get more attention in the learning process . This view is supported by experimental data that revealed that transient novel experiences are replayed more prominently than familiar stimuli [69]–[71] . Altogether our results show that hidden Markov models provide a promising theoretical framework for understanding the emergence of all-important capabilities of the brain to understand and predict hidden states of complex time-varying sensory stimuli .
In this section we provide additional details to the derivations of the network model and its stochastic dynamics . For the sake of simplicity , throughout the theoretical analysis we use a simple EPSP kernel of the form ( 12 ) Thus , a kernel with a single exponential decay with time constant . Here , determines the Heaviside step function which is for and zero else . The derivation provided here can be extended to more complex EPSP shapes , if two prerequisites are fulfilled . First , a suitable Markov state must be found that describes the dynamics of the EPSP kernel , i . e . a state must exist for which we can write . In fact , this property holds true for any deterministic function , although the required Markov state can be very complex . Second , the statistics of the EPSPs induced by the kernel must be readily described by an exponential family distribution . For this latter requirement the same considerations as for the afferent synapses apply , which have been addressed in [13] , [32] , [33] . The simplest case for which these conditions are fulfilled is the one considered in the last experiment where rectangular EPSPs and constant inter-spike intervals of the same length were used . In that case the network state collapses to , which follows a multinomial distribution as considered in [32] . All simulations were done in Matlab ( Mathworks ) , directly implementing the derived equations without discrete time approximations . The population output rate was tuned to give an average output rate of 5–20 Hz per neuron . Prior to learning all weights were set to small equally distributed random values . The weight updates were incorporated using a constant learning rate . Other than in the theoretical analysis where synaptic delays were neglected for the sake of simplicity , we used synaptic delays of 5 ms for the lateral excitatory synapses in the numerical experiments . We also used a more realistic double exponential EPSP kernel of the form [28] ( 38 ) where and are the time constants of the falling and rising edges of the EPSP kernel , respectively . The above theoretical analysis applies equally to this kernel , but would be slightly more complex since each of the two exponential decay terms comprises a piece of memory which has to be reflected in the network state . The diagonal of the weight matrix was set to zero and these weights were excluded from learning . Instead , a refractory mechanism was used with a kernel given by [28] ( 39 ) where is the maximum amplitude of the refractory kernel , is the refractory time constant and is the time elapsed since the last output spike . Equation ( 39 ) was subtracted from the membrane potential ( 1 ) . | It has recently been shown that STDP installs in ensembles of pyramidal cells with lateral inhibition networks for Bayesian inference that are theoretically optimal for the case of stationary spike input patterns . We show here that if the experimentally found lateral excitatory connections between pyramidal cells are taken into account , theoretically optimal probabilistic models for the prediction of time-varying spike input patterns emerge through STDP . Furthermore a rigorous theoretical framework is established that explains the emergence of computational properties of this important motif of cortical microcircuits through learning . We show that the application of an idealized form of STDP approximates in this network motif a generic process for adapting a computational model to data: expectation-maximization . The versatility of computations carried out by these ensembles of pyramidal cells and the speed of the emergence of their computational properties through STDP is demonstrated through a variety of computer simulations . We show the ability of these networks to learn multiple input sequences through STDP and to reproduce the statistics of these inputs after learning . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"circuit",
"models",
"computer",
"and",
"information",
"sciences",
"neural",
"homeostasis",
"neural",
"networks",
"computational",
"neuroscience",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"neuroscience"
] | 2014 | STDP Installs in Winner-Take-All Circuits an Online Approximation to Hidden Markov Model Learning |
Lipid droplets ( LDs ) are important cellular organelles that govern the storage and turnover of lipids . Little is known about how the size of LDs is controlled , although LDs of diverse sizes have been observed in different tissues and under different ( patho ) physiological conditions . Recent studies have indicated that the size of LDs may influence adipogenesis , the rate of lipolysis and the oxidation of fatty acids . Here , a genome-wide screen identifies ten yeast mutants producing “supersized” LDs that are up to 50 times the volume of those in wild-type cells . The mutated genes include: FLD1 , which encodes a homologue of mammalian seipin; five genes ( CDS1 , INO2 , INO4 , CHO2 , and OPI3 ) that are known to regulate phospholipid metabolism; two genes ( CKB1 and CKB2 ) encoding subunits of the casein kinase 2; and two genes ( MRPS35 and RTC2 ) of unknown function . Biochemical and genetic analyses reveal that a common feature of these mutants is an increase in the level of cellular phosphatidic acid ( PA ) . Results from in vivo and in vitro analyses indicate that PA may facilitate the coalescence of contacting LDs , resulting in the formation of “supersized” LDs . In summary , our results provide important insights into how the size of LDs is determined and identify novel gene products that regulate phospholipid metabolism .
Lipid droplets ( LDs ) are dynamic organelles that govern the storage and turnover of lipids [1] . They also play important roles in membrane and lipid trafficking , protein storage , protein degradation and the replication of hepatitis C and dengue viruses [1] , [2] , [3] , [4] , [5] . All LDs comprise a core of storage neutral lipids , i . e . triacylglycerols ( TAG ) and sterol esters ( SE ) , which are wrapped by a monolayer of phospholipids containing embedded proteins . LDs are believed to originate from the endoplasmic reticulum ( ER ) , although the exact mechanism underlying their biogenesis remains to be determined [6] . LDs of various sizes have been observed in different tissues or within the same cell type under different ( patho ) physiological conditions [7] , [8] . A giant ( up to 200 µm in diameter ) , unilocular LD often occupies the entire cytoplasm of white adipocytes , specializing in energy storage . In contrast , many much smaller LDs ( usually less than 10 µm in diameter ) are found in brown adipocytes . Small LDs are also found in normal liver cells; however , the size of liver LDs increases dramatically in hepatic steatosis , e . g . in the ob/ob mice [9] . The physiological significance of LD size has not been well recognized and little is known about the molecular mechanisms that influence LD size . Recent studies have begun to shed light on the control and physiological relevance of LD size . Deletion of FSP27 ( fat-specific protein of 27 kDa ) resulted in many smaller LDs in white adipocytes , enhanced lipolysis and protection from diet-induced obesity and insulin resistance [8] , [10] . A genome-wide RNA interference ( RNAi ) screen in Drosophila S2 cells identified enzymes of phospholipid biosynthesis as determinants of LD size and number [11] . Interestingly , screens of the viable yeast deletion library found extensive clustering of LDs and formation of “Supersized” LDs ( SLDs ) that are up to 50 times the normal volume in cells deleted for FLD1 [12] , which encodes a functional homologue of a human lipodystrophic protein: seipin [13] , [14] . Congenital generalized lipodystrophy ( CGL ) is characterized by a nearly complete absence of adipose tissue and a range of metabolic changes such as extreme insulin resistance [15] . Genome-wide linkage analysis identified two loci for CGL: CGL type 1 ( CGL1 ) is caused by mutations in the 1-acylglycerol-3-phosphate-O-acyl transferase 2 ( AGPAT2 ) gene [16] and CGL2 by mutations in BSCL2 which encodes seipin [17] . AGPAT2 catalyzes the formation of phosphatidic acid ( PA ) but knocking down AGPAT2 led to elevated levels of several phospholipid species including PA [18] , [19] . In mice , mutations in the lipin-1 gene which encodes a phosphatidate phosphatase are responsible for severe lipodystrophy [20] , [21] . Therefore , both AGPAT2 and lipin-1 appear to control adipogenesis through modulation of the synthesis of phospholipids and triacylglycerol precursors , especially PA . In contrast , although Fld1p ( yeast seipin ) has been implicated in lipid metabolism , little information is available on its molecular function [12] , [22] . Previous genome-wide studies of yeast LDs covered only non-essential yeast genes and used only one culture condition [12] , [13] . Here , a revised screen of yeast deletion mutants for the formation of “supersized” lipid droplets ( SLDs ) identifies known and novel proteins in phospholipid metabolism . These mutants including fld1Δ share one common feature: an increase in the level of PA . Reducing the amount of PA invariably led to a significant reduction in SLD formation in all mutants . Finally , a critical role of PA in LD coalescence is confirmed when LD formation is reconstituted in vitro .
We previously identified yeast gene deletions that led to a reduced number of cytoplasmic LDs , and we noticed that one of these mutants , fld1Δ , developed very large LDs ( Figure 1A and [12] ) . Whereas the diameter of LDs in wild type ( WT ) cells typically ranges between 0 . 3 to 0 . 4 µm , and rarely exceeds 0 . 5 µm [23] , fld1Δ cells often synthesize LDs with a diameter larger than 1 . 0 µm [12] . We arbitrarily define LDs with a diameter greater than 1 . 0 µm as “supersized” LDs ( SLDs ) , whose volume is over 30 times the average of wild type LDs . About 20% of fld1Δ cells cultured in rich ( YPD ) medium contained SLDs , and the percentage increased to ∼70% when cells were grown in minimal ( synthetic complete/SC ) medium ( Figure 1A and Table 1 ) . Given the effects various nutrients may have on the dynamics of LDs , we reasoned that growing cells on defined , minimal media ( SC ) would reduce the impact of nutrients and uncover additional genes . We have therefore screened the entire collection of viable yeast deletion mutants ( ∼4800 ) grown on minimal ( SC ) media for SLDs . In addition , our previous screen focused only on the viable yeast deletion mutants , representing ∼80% of the genome . In order to identify essential genes impacting LD size , we now included the collection of mutants where all essential genes are controlled by the TetO7-promoter , which can be switched off efficiently [24] . Besides fld1Δ , we identified nine additional mutants ( sld2-10 ) that produced supersized LDs ( SLDs ) ( Figure 1 and Table 1 ) . Except for two previously uncharacterized genes ( RTC2/SLD6&MRPS35/SLD7 ) , the majority of the SLD genes appear to function directly or indirectly in the metabolism of phospholipids , especially phosphatidylcholine ( PC ) . PC is synthesized in yeast via two pathways: the Kennedy pathway and the phosphatidylethanolamine N-methyltransferase ( PEMT ) pathway ( see Figure S1 ) . In the PEMT pathway , phosphatidylethanolamine ( PE ) is methylated to PC in three steps by two methyltransferases , Cho2p and Opi3p . Besides cho2Δ and opi3Δ mutants , ino2Δ and ino4Δ mutants are also defective in PC synthesis via PE methylation since Ino2p and Ino4p are transcription factors that positively regulate the PEMT pathway [25] . When cki1Δ , pct1Δ , cpt1Δ , cho2Δ , opi3Δ , ino2Δ , and ino4Δ cells were cultured in rich ( YPD ) medium , none of these mutants accumulated SLDs . In contrast , when grown in SC medium ( no choline and hence little Kennedy pathway activity ) , approximately 60% of cho2Δ cells , 90% of opi3Δ cells , 97% of ino2Δ and ino4Δ cells produced SLDs , whereas cki1Δ , pct1Δ , and cpt1Δ cells did not ( Figure 1B and 1E ) . From these results , PC synthesis does appear to be critical in regulating the size of LDs , in agreement with a recent study examining LD dynamics in Drosophila S2 cells [11] . Interestingly , of the 825 essential genes examined , SLDs were observed only upon knocking-down CDS1 ( encoding CDP-diacylglycerol synthase ) ( Figure . 1D and 1E ) . Therefore , the synthesis of not only PC , but also other phospholipids could be important for LD growth . The SLDs observed in fld1Δ , cho2Δ , opi3Δ , ino2Δ , ino4Δ , and TetO7-CDS1 ( thereafter referred to as cds1 when repressed by doxycycline ) strains were further confirmed by electron microscopy ( Figure 1E ) . The levels of TAG and SE of all mutants were also examined , and the level of TAG was significantly increased in all mutants ( Figure 1F ) . As compared to many small LDs in WT cells , the formation of SLDs limits the surface area that is accessible to lipases . Therefore , the mobilization of TAG may be impaired in sld mutants . TAG breakdown in fld1Δ , ino4Δ , and cds1 strains was monitored in the presence of 10 mg/L cerulenin that prevents their de novo synthesis as described [26] . Our results show that TAG mobilization in the mutants is significantly slower than that of WT ( Figure S2 ) . Our finding that YPD media invariably decreased the percentage of cells displaying SLDs in all sld mutants suggested that certain components present in rich YPD media , but absent or low in SC media , suppressed SLD formation . Considering that Cds1p , Cho2p , Opi3p , Ino2p , and Ino4p are either enzymes or transcription factors involved in phospholipid biosynthesis , we speculated that these components might be precursors of phospholipids . To examine this possibility , we cultured WT and mutant strains in SC media supplemented with 1 mM choline , 1 mM ethanolamine , or 75 µM inositol . Interestingly , inositol treatment reduced the SLD formation in all mutants ( Figure 2 ) . In contrast , ethanolamine addition had an opposite effect; it enhanced SLD formation in most of the mutants . As expected , choline addition completely blocked the formation of supersized LDs in cho2Δ , opi3Δ , ino2Δ and ino4Δ strains ( Figure 2A and 2B ) , since exogenously added choline restored PC synthesis in these mutants through the Kennedy pathway . Surprisingly it also had similar effect in rtc2Δ and mrps35Δ strains , suggesting that these two genes may also function in PC metabolism . Choline addition also partially inhibited SLD formation in cds1 , ckb1Δ , and ckb2Δ cells , but had little effect in fld1Δ cells . One notable common feature among cho2Δ , opi3Δ , ino2Δ , ino4Δ and cds1 mutants is the accumulation of PA ( Figure 3A and [25] , [27] ) . PA is a cone-shaped lipid that alters the curvature of the membranes , and has been shown to promote both SNARE-dependent and -independent membrane fusion events [28] , [29] . A previous study implicated PA in the assembly of lipid droplets from newly synthesized TAGs in a cell-free system [30] . To examine whether PA is a key player in the formation of SLDs , we first analyzed the cellular level of PA in the SLD mutants by LC-MS . Indeed a significant elevation of PA was seen in all mutants except fld1Δ cells , where the level of PA is only moderately elevated ( Figure 3A and 3B ) . Inositol treatment reduces the cellular PA pool through increased synthesis of phosphatidylinositol ( PI ) and also through the activation of a Mg2+-dependent PA phosphatase [25] , [31] . Consistent with the implication of PA in SLD formation , inositol treatment resulted in a significant reduction of SLD formation in all mutants including fld1Δ ( Figure 2 ) . In addition , when two PA phosphatases ( PAH1 and DPP1 ) were overexpressed under a GAL1 promoter [20] , [32] , both greatly reduced SLD formation in ino2Δ and ino4Δ cells , and also in fld1Δ cells ( Figure 3C ) . Overexpression of PAH1 and DPP1 did not change the level of PC , PE , PS and PI , but significantly reduced the cellular level of PA in fld1Δ cells ( Figure S3 ) . These results imply that the increased amount of PA may account for the formation of SLDs , and that the level of PA in subcellular organelles such as the endoplasmic reticulum where LDs originate may have changed in fld1Δ cells , despite an insignificant increase in overall PA in this mutant ( see below ) . SLDs in yeast were originally identified in fld1Δ cells; but the molecular function of Fld1p/seipin remains elusive [22] . To gain more insights into the function of Fld1p , mRNA microarray analysis was performed in WT and fld1Δ cells . Of ∼5800 transcripts examined , INO1 and OPI3 were the only transcripts whose levels were significantly upregulated in fld1Δ cells ( Figure 4A ) . Quantitative real-time PCR confirmed a ∼5-fold increase in the INO1 mRNA level in fld1Δ cells ( Figure 4B ) . INO1 gene expression is derepressed when intracellular PA concentration rises [25] . We therefore examined the level of PA on the ER where the Opi1p-Scs2p regulatory complex of INO1 expression exists [33] . Indeed , a significant increase of PA was observed in microsomes isolated from fld1Δ cells ( Figure 4C ) . These results show that Fld1p , as other SLD mutants , can regulate PA metabolism , and also suggest that both the level and location of PA are relevant to droplet formation . A recent study in Drosophila also revealed a possible role for dSeipin in PA metabolism [34] . Another strategy to increase PA is through inactivation of Pah1p , the PA phosphatase and ortholog of mammalian lipin proteins [20] , [21] . Deletion of PAH1 leads to a dramatic increase in the level of its substrate , PA , but causes a dramatic reduction in the amount of TAG [20] . Although the number of LDs was significantly reduced in pah1Δ cells , LD size was comparable to that in WT cells . Remarkably , SLDs were detected consistently in ∼3% of pah1Δ cells , though its TAG synthesis was decreased by over 50% ( Figure 5A–5C ) . In contrast , no SLDs were ever observed in dga1Δ lro1Δ cells , which have little diacylglycerol ( DAG ) acyltransferase activity [35] , [36] . Interestingly , when pah1Δ cells were supplemented with oleate and DAG which bypasses the lack of PA phosphatase activity , the number of cells producing SLD increased to ∼30% , whereas oleate alone had no effect ( Figure 5D ) . These results further indicate that PA plays an important role in SLD formation , and this role is more pronounced when the biosynthesis of TAG is not severely compromised . The result that ethanolamine addition enhanced SLD formation in nearly all mutants ( Figure 2 ) suggested that an elevated phosphatidylethanolamine ( PE ) concentration could have a role in SLD formation , given that PE is also a cone-shaped phospholipid that can increase membrane curvature , thereby promoting LD monolayer coalescence [11] . Consistent with this notion , mutants known to accumulate PE also displayed a higher percentage of cells forming SLDs , particularly opi3Δ , ino2Δ , and ino4Δ ( Table 1 ) . As shown in Figure 6A , our lipidomic analysis further revealed that lipid droplets isolated from cho2Δ , ino2Δ , and ino4Δ also had a higher PE to PL ( total membrane phospholipids ) ratio than those of WT cells . In addition , ethanolamine treatment significantly increased the proportion of PE on lipid droplets of fld1Δ and cds1 cells . Even in LDs of ino4Δ cells , ethanolamine addition still moderately increased the PE to PL ratio , though its PE level was already much higher than WT . However , elevated PE alone was not able to induce SLD formation since inositol addition completely abolished the biogenesis of supersized LDs despite that a higher PE to PL ratio persisted in ino4Δ cells ( Figure 6A ) . Moreover , SLDs were abundant in cds1 where the PE/PL ratio was lower than that of the WT ( Figure 6A and Figure S4 ) . A decreased phospholipid ( PL ) to TAG ratio could also induce SLD formation , since coalescence may be induced to decrease the surface-to-volume ratio of droplets when phospholipids are limiting [11] . This model appears to be true for the mutants grown in SC media . However , when grown in YPD media , SLDs disappeared in cho2Δ , ino2Δ and ino4Δ strains but the decreased PL to TAG ratio persisted ( Figure 6B ) . In addition , inositol supplementation did not increase the phospholipid to TAG ratio in cds1 or ino4Δ mutant ( Figure 6B ) . Choline addition completely inhibited SLD formation in rtc2Δ and mrps35Δ , in a manner similar to cho2Δ , opi3Δ , ino2Δ , and ino4Δ , strains known to be defective in the methylation of PE into PC ( Figure 2 ) . This phenotype suggests that deletion of RTC2 or MRPS35 might affect PC synthesis through the PEMT pathway . As expected , lipidomic analysis revealed that rtc2Δ and mrps35Δ strains displayed a 2 . 5-fold increase of PE to PC ratio , indicating these two gene products are involved in PC synthesis through the PEMT pathway ( Figure 7 ) . rtc2Δ and mrps35Δ cells also synthesized ∼60% more phosphatidylinositol ( PI ) than WT , possibly resulting from the accumulation of CDP-DAG due to a blocked PEMT pathway ( Figure 7 and Figure S4A ) . We also examined the phospholipid profiles of ckb1Δ and ckb2Δ strains and found that both synthesized less PC and PE than WT without causing significant changes in the PE to PC ratio ( Figure 7 and Figure S4A ) . We next investigated how changes in phospholipids in the sld mutants may lead to the formation of SLDs . One possibility could be enhanced fusion activities as previously suggested [12] . Indeed , fusion of Nile red-stained LDs could be observed in cho2Δ , opi3Δ , ino2Δ , ino4Δ , and cds1 strains ( Figure 8 , Videos S1 and S2 ) , and also in rtc2Δ , mrps35Δ , ckb1Δ and ckb2Δ strains ( not shown ) . The fusion frequency of LDs in each mutant was similar ( ∼10 out of 200 adjacent pairs of LDs ) to that of fld1Δ [12] . Furthermore , LDs isolated from representative strains demonstrated fusion activities in vitro ( Videos S3 and S4 ) . It should be noted that no fusion events in wild type yeast cells were ever observed with the methods employed . To obtain direct evidence that PA induces coalescence of small LDs to form supersized droplets , artificial LDs were made and their stability tested . After generating the artificial droplets by sonication , we removed liposomes formed at the same time by density gradient centrifugation . This fractionation also concentrated the artificial droplets . From this starting point , we followed the stability of LDs by light scattering , which directly measures the number of LDs . When we increased the concentration of PA in artificial LDs , their number decreased significantly during incubation ( Figure 9 ) . In the presence of PE , a smaller fraction of PA ( ∼3% molar ratio , PC:PE:PA 3∶1∶0 . 13 ) achieved a similar effect compared to the coalescence observed in the presence of PC covered LDs ( ∼5% PA ( PC:PA 20∶1 ) . These results show that PA reduces the stability of LDs and mediates coalescence of LDs , and that this property of PA may be modulated by the phospholipid composition of LDs .
Lipid droplets are dynamic organelles whose number and size undergo constant changes in response to internal and external cues [1] , [3] . The physiological relevance of the size of the LDs is not well understood , and far less is known about how the size of LDs is determined at the molecular level . In this study , we identify key proteins that govern the size of LDs in yeast by modulating phospholipid metabolism . We also identify proteins previously unknown to regulate phospholipid metabolism . Most importantly , we provide in vivo and in vitro evidence that phosphatidic acid can influence the size of the LDs . SLDs provide an efficient form of fat storage in terms of surface to volume ratio . Do cells automatically generate SLDs upon lipid loading to economize on the synthesis of phospholipids that form the surface of lipid droplets ? While large lipid droplets are typical in white adipocytes , most other cell types ( brown adipocytes , hepatocytes , myocytes ) store lipids in numerous small LDs . In WT yeast cells , a dramatic increase in the number but not size of LDs is often observed in growth conditions favoring neutral lipid synthesis and storage , such as starvation . Maintaining small LDs may be physiologically important: upon starvation , yeast cells convert phospholipid intermediates and sterols to neutral lipids which in turn can be hydrolyzed to release fatty acids and sterols for immediate membrane synthesis and cell growth when glucose becomes available . It has also been recently reported that lipolysis is important for efficient cell-cycle progression in yeast [26] , and lipolysis occurs more efficiently for small LDs ( Figure S2 and [8] , [11] ) . Therefore , it appears that SLDs are only formed in highly specialized , non-dividing cells ( e . g . fully differentiated white adipocytes ) or under pathological conditions such as severe hepatic steatosis . Genetic factors that regulate the size of LDs were identified in two separate screens in yeast and in Drosophila S2 cells [11] , [12] . Decreased PC synthesis , and consequently an increased PE to PL ratio ( or a decrease in PC/TAG ) , has been associated with SLD formation ( Figure 1 ) [11] . LDs are phase-separated organelles in the cytoplasm . Thus , unlike for other organelles , the steady state and lowest energy state is probably to have only one droplet ( this would minimize interfacial surface energy ) . Some phospholipids ( specifically PC ) shield droplets from coalescence . Fusogenic lipids such as PA and PE could overcome this effect . In agreement with this notion , we found that PE and PA both have an effect on SLD formation . First , treatment of ethanolamine further increased PE to PL ratio and enhanced SLD formation in most mutants ( Figure 2 and Figure 6 ) ; in addition , strains with increased levels of both PE and PA also had a higher percentage of cells producing supersized LDs ( Table 1 ) . Decreased levels of PLs also could lead to SLD formation , since phospholipids levels may not be sufficient under these conditions to prevent the hydrophobic TAG phases to fuse . If the amount of PLs on LDs is not sufficient , fusion would occur until the surface-to-volume ratio of LDs is reflecting the ratio of phospholipids to TAG . At this point the monolayer would shield the LD from any further fusion . Identification of the cds1 mutant through the screening of the knock-down collection of essential yeast genes turned our attention to PA , whose critical role in SLD formation was confirmed by the strong “size-reduction” effect of inositol supplementation , an efficient and reliable way to reduce PA in yeast . The essential role of PA in SLD formation was further confirmed when all mutants that develop SLDs were found to accumulate PA ( Figure 3 ) , including the seipin-deficient ( fld1Δ ) ( Figure 4 ) and lipin-deficient ( pah1Δ ) ( Figure 5A and 5B ) cells . PA is a central intermediate in the synthesis of major glycerolphospholipids and TAG , as well as an important signaling lipid [25] , [37] , [38] . Different pools of PA and distinct PA subclasses may account for the diversity of PA function . For instance , the yeast Opi1p ( ER localized transcription repressor ) senses only a PA pool on the ER but not the plasma membrane PA pool regulated by the yeast phospholipase D Spo14p [33] . Deleting or overexpressing SPO14 did not have any impact on the formation of SLDs ( data not shown ) , suggesting that an intracellular ( ER/LD ) PA pool is responsible for SLD formation . This appears to be also the case for the fld1Δ mutant , in which the level of microsomal PA , but not overall PA , is significantly increased . The level of PA on the ER could very well reflect the amount of PA on the LDs , given that LDs are believed to originate from the ER . In summary , we find here that increased PA levels may overcome the effect of phospholipid shielding , and that the location of PA also matters . Besides establishing a strong link between PA and the size of LDs , our screen also reveals that Rtc2p and Mrps35P can regulate the PEMT pathway of PC synthesis . Both Rtc2p and Mrps35p were found to associate with mitochondria by proteomic studies [39] , [40] . We are currently undertaking experiments to understand the role of these two proteins in phospholipid metabolism . Exactly how ckb1Δ and ckb2Δ mutants cause a significant increase in cellular PA and thereby the formation of SLDs is not clear , although the key transcription factor that regulates phospholipid synthesis in the yeast , Opi1p , can be phosphorylated and regulated by casein kinase 2 [41] . Finally , the mechanistic link between changes in the level of PA/PE and the formation of SLDs was investigated . LDs are covered by a monolayer of phospholipids , whose composition may have a profound effect on the dynamics of the LDs . Both PA and PE are cone-shaped , fusogenic lipids that can alter the curvature of the membranes . PA , in particular , has been shown to promote both SNARE dependent and independent membrane fusion events [28] , [29] . It is possible that a higher level of PA on the monolayer of the LDs would promote spontaneous fusion of contacting LDs . Indeed , in vivo microscopic observation found increased incidents of LD fusion in mutants with increased level of PA ( Figure 8 and Videos S1 , S2 , S3 , S4 ) . Remarkably , when artificial LDs are made with different ratios of PA , PE and PC , it is clear that even a small amount of PA could significantly increase the size of the LDs ( Figure 9 ) . In summary , studies described herein identify novel protein and lipid regulators of the size of the LDs , an important lipid-storage organelle . Knowing how LD size is determined may provide invaluable insights into how human cells/tissues handle abnormal influx of lipids in today's obesogenic environment .
S . cerevisiae wild type strain BY4741 ( MATa; his3Δ1; leu2Δ0; met15Δ0; ura3Δ0 ) and its derived non-essential gene-deletion strains were either obtained from EUROSCARF or generated in this study . The latter included pah1Δ ( PAH1::HIS3MX6 ) , and dga1Δlro1Δ ( DGA1::kanMX4 , LRO1::hphNT1 ) . The Tet-promoter strains used for expression of essential genes under the regulatable TetO7 promoter were obtained from Open biosystems . Yeast extract , peptone , dextrose , and yeast nitrogen base were purchased from BD . Nile red , choline , ethanolamine , inositol , doxycycline , cerulenin , and Ficoll 400 were from Sigma . 1 , 2-dioctanoyl-sn-glycerol 3-phosphate ( PA 8∶0/8∶0 ) , 1 , 2-dioleoyl-sn-glycero 3-phosphate ( PA 18∶1/18∶1 ) , oleoyl-L-α-lysophosphatidic acid , 1 , 2-dioctanoyl-sn-glycerol ( DAG 8∶0/8∶0 ) , phosphatidylethanolamine , phosphatidylcholine , and triolein were purchased from Avanti Polar Lipids . To screen for yeast mutants that generate supersized LDs , cells were cultured in synthetic complete media ( 0 . 67% yeast nitrogen base , 2% dextrose , and amino acids ) in 96-well plates at 30°C till stationary phase . For TetO7-regulated strains , 15 µg/ml doxycycline was added to repress specific genes . Cells were stained with 20 µg/ml Nile red for LDs and observed by fluorescence microscopy . Yeast strains found to contain SLDs were recultured in synthetic complete media with aeration in 10 ml culture tubes to confirm the phenotype . These strains were also grown in YPD media ( 1% yeast extract , 2% peptone , and 2% dextrose ) to examine the morphology of LDs . For phospholipid precursor treatment , synthetic complete medium was supplemented with 1 mM choline , 1 mM ethanolamine , or 75 µM inositol . Fluorescent imaging was performed under a Leica CTR5500 microscope ( Wetzlar , Germany ) with an EL6000 fluorescent lamp . Images were taken with a DFC300 FX digital camera and a Leica LAS AF software . Yeast cells were viewed under a×100/1 . 30 oil immersion objective lens . A 450−490-nm bandpass excitation filter , a 510 dichromatic mirror , and a 515-nm longpass emission filter ( Leica filter cube I3 ) were chosen to observe Nile red-stained LDs . For statistical presentation of the percentage of cells containing supersized LDs , 200 cells were observed and percentage was calculated . The experiments were done in triplicates and the result was shown as mean ± SD . Mammalian LDs were stained with Bodipy 493/503 ( Invitrogen ) and observed with a 470/40-nm bandpass excitation filter , a 500-nm dichromatic mirror , and a 525/50-nm bandpass emission filter ( Leica filter cube GFP ) . To observe and record LD fusion , 3 µl of mid-log phase cells ( OD600∼1 . 5 ) or purified LDs were stained with Nile red , spotted on a slide and covered with a coverslip . Under the microscope , cells in which two or several LDs lay close together were targeted . Images were collected at 0 . 5 second intervals . Cells were grown in rich medium until stationary phase , harvested , fixed with 2 . 5% glutaraldehyde and postfixed with 2% ( w/v ) osmium tetroxide . The samples were subsequently dehydrated in a series of graded ethanol and embedded in Spurr's Resin . 80 nm ultrathin sections were stained with uranyl acetate and lead citrate and examined under a JEM-1230 Joel electron microscope . Total RNA was extracted using the RNeazy Plus kit ( QIAGEN ) . cDNA was generated from total RNA using a SuperScript VILO cDNA Synthesis Kit ( Invitrogen ) . PCR reaction was performed using Rotor-Gene RG-3000A ( Qiagen ) . Threshold cycle value for each gene was acquired at the log phase and gene expression was normalized to reference genes as indicated . For Affymetrix Array processing and analysis , samples were prepared according to the Affymetrix GeneChip® Yeast Genome 2 . 0 Array protocol . Differences between WT and fld1Δ strains were compared using one-way ANOVA and adjusted for false discovery rate at 0 . 05 level . Array data is deposited on Gene Expression Omnibus ( http://www . ncbi . nlm . nih . gov/geo/ ) . Microsomes were isolated as described [42] . Briefly , WT and fld1Δ cells were cultured in SC media till log phase ( OD600∼1 . 0 ) and harvested . Cells as 0 . 1 g ( wet weight ) /ml in 0 . 1 M Tris SO4 were sequentially incubated in pH9 . 4/10 mM DTT for 10 min , and in 1 . 2 M sorbitol/20 mM Tris Cl , pH7 . 5/1x SC medium/zymolase 100 T ( 15 mg/ml ) for 30 min . Spheroplasts were lysed in HEPES lysis buffer ( 10 mM HEPES/KOH , pH6 . 8/50 mM potassium acetate/100 mM sorbitol/2 mM EDTA ) with the aid of a Dounce homogenizer . After removal of cell dedris , lysates were centrifuged at 30 000 g for 10 min at 4°C . P30 000 g membrane pellets were resuspended in HEPES lysis buffer , loaded onto 1 . 2 M/1 . 5 M sucrose ( prepared in HEPES lysis buffer ) gradients , and centrifuged at 100 000 g for 1 h at 4°C . ER membranes were collected at the 1 . 2 M/1 . 5 M sucrose interface . Lipid droplets were isolated as described previously [12] . Lipids were extracted from zymolase-digested lyophilized yeast cells , isolated ER microsomes or lipid droplets . Briefly 900 µl ice-clod chlorofom:methanol ( 1∶2 ) was added to samples . Mixtures were vigorously vortexed for 1 min , and incubated for 2 h in vacuum container with rotary shaking at 4°C . Then 400 µl ice-cold water and 300 µl chloroform were added , vortexed and incubated on ice for 1 min . After centrifugation at 12000 rpm for 3 min at 4°C , the lower organic phase was collected . Subsequently , 50 µl 1 M HCl and 500 µl chloroform were added to the remainder , vortexed , and incubated on ice for 3 min . The lower organic phase was also collected after centrifugation at 12000 rpm for 3 min at 4°C , and combined with the first extract . The extracted lipids were blown dry with nitrogen gas , and resuspended in solvent for mass spectrometry analysis . Lipidomic analysis , and quantitative measurement of neutral lipids via thin layer chromatography ( TLC ) were performed as described [12] , [43] . TLC plates were developed in chloroform/methanol/water ( 65∶25∶4 ) to separate phospholipid species . To prepare lipid emulsions , lipids were mixed in chloroform/methanol ( 2∶1 ) , dried under a stream of N2 , resuspended in buffer ( 150 mM NaCl , 50 mM Tris/HCl , pH 7 . 5 , 1 mM EDTA , ) , and sonicated . For emulsions , the molar ratio of TAG to total phospholipids was 2∶2 . 5 . Contaminating vesicles were removed , and LDs were concentrated by ultracentrifugation at 100 , 000 g for 15 min . For light scattering , lipid concentration was 25 mM phospholipids and 20 mM TAG before centrifugation . All data are presented as mean ± SD . Statistical comparison between the two groups was performed using Student's t-test . Microarray data were analyzed using one-way ANOVA and adjusted for false discovery rate at 0 . 05 level . | Lipid droplets ( LD ) are primary lipid storage structures that also function in membrane and lipid trafficking , protein turnover , and the reproduction of deadly viruses . Increased LD accumulation in liver , skeletal muscle , and adipose tissue is a hallmark of the metabolic syndrome . Enlarged LDs are often found in these tissues under disease conditions . However , little is known about how the size of LDs is controlled in eukaryotic cells . In this study , we use genetic and biochemical methods to identify important gene products that regulate the size of the LDs . Notably , a common feature among these mutants with “supersized” LDs is an increased level of phosphatidic acid ( PA ) . We also show that a small amount of PA can increase the size of artificial LDs in vitro . Overall , our study identifies important lipids and proteins in determining LD size . These results provide valuable insights into how human cells/tissues handle abnormal influx of lipids in today's obesogenic environment . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"lipids",
"model",
"organisms",
"genetic",
"screens",
"genetics",
"biology",
"genetics",
"and",
"genomics"
] | 2011 | A Role for Phosphatidic Acid in the Formation of “Supersized” Lipid Droplets |
Genome structure variation has profound impacts on phenotype in organisms ranging from microbes to humans , yet little is known about how natural selection acts on genome arrangement . Pathogenic bacteria such as Yersinia pestis , which causes bubonic and pneumonic plague , often exhibit a high degree of genomic rearrangement . The recent availability of several Yersinia genomes offers an unprecedented opportunity to study the evolution of genome structure and arrangement . We introduce a set of statistical methods to study patterns of rearrangement in circular chromosomes and apply them to the Yersinia . We constructed a multiple alignment of eight Yersinia genomes using Mauve software to identify 78 conserved segments that are internally free from genome rearrangement . Based on the alignment , we applied Bayesian statistical methods to infer the phylogenetic inversion history of Yersinia . The sampling of genome arrangement reconstructions contains seven parsimonious tree topologies , each having different histories of 79 inversions . Topologies with a greater number of inversions also exist , but were sampled less frequently . The inversion phylogenies agree with results suggested by SNP patterns . We then analyzed reconstructed inversion histories to identify patterns of rearrangement . We confirm an over-representation of “symmetric inversions”—inversions with endpoints that are equally distant from the origin of chromosomal replication . Ancestral genome arrangements demonstrate moderate preference for replichore balance in Yersinia . We found that all inversions are shorter than expected under a neutral model , whereas inversions acting within a single replichore are much shorter than expected . We also found evidence for a canonical configuration of the origin and terminus of replication . Finally , breakpoint reuse analysis reveals that inversions with endpoints proximal to the origin of DNA replication are nearly three times more frequent . Our findings represent the first characterization of genome arrangement evolution in a bacterial population evolving outside laboratory conditions . Insight into the process of genomic rearrangement may further the understanding of pathogen population dynamics and selection on the architecture of circular bacterial chromosomes .
Genome arrangement has profound effects on organismal phenotype . Genome arrangement likely impacts gene expression [1] , [2] , [3] , and can result in total loss of gene function when a rearrangement breakpoint occurs inside a reading frame . Moreover , rearrangements are known to affect linkage and introduce genetic incompatibility in eukaryotes [4] . Similar recombination-stifling effects have been proposed in prokaryotes [5] , [6] , whose capacity for genetic exchange among divergent taxa has only recently been appreciated [7] . In naturally competent microbes which undergo frequent homologous recombination , genome arrangements themselves may be better indicators of vertical inheritance than other molecular characters . Our ability to measure gene order and chromosome structure has undergone several revolutions , beginning with careful study of linkage maps [8] , later moving towards direct observation by microscope , FISH , Radiation Hybrid , paired-end genome sequencing , and Optical Mapping techniques [9] , [10] , [11] , [12] . The continued improvement in measurement technology has offered revelations regarding the pattern and extent of genome rearrangement in organisms ranging from bacteria [13] to mammals [14] . In circular bacterial chromosomes , DNA replication divides the circular chromosome into two domains called replichores . Replication begins when DNA polymerase holoenzymes anneal to the origin of replication ( ori ) . Two holoenzymes then simultaneously copy the circular chromosome in opposite directions , and initially the DNA polymerase holoenzymes are co-localized in the cell in a so-called “replication factory” [15] . Each holoenzyme copies about half the chromosome , and they eventually meet each other in the Ter macrodomain . The Ter macrodomain spans a large portion of the chromosome opposite the origin of replication and contains several ter sites which bind proteins that halt procession of DNA polymerase [16] . In cases where homologous recombination has taken place during replication , the XerCD molecular machinery resolves the chromosome dimer at the dif site [17] , [18] . Moreover , the predominant site of replication termination appears to be at or near the dif site [19] . We refer to each half of the chromosome , delineated by ori and dif , as a replichore . Hereafter we will use the word “terminus” or phrase “terminus of replication” to refer to the approximate location of the dif site . Genome sequencing has revealed that rearrangements do not occur with uniformly distributed endpoints on circular prokaryotic chromosomes . Instead , a striking pattern of inversions with endpoints biased by the origin and terminus of replication has commonly been observed [20] , [21] , [22] , [23] . Several explanations for the observed pattern have been devised , all of which focus on the nature of DNA replication in circular chromosomes . An inter-replichore inversion refers to a chromosomal inversion with one endpoint in each replichore . Such inversions swap the relative orientations of the origin and terminus . If the inversion endpoints are equally distant from the origin , then replichore sizes remain unchanged—a so-called “symmetric inversion” . Previous genome analyses indicate that inversions typically occur with breakpoints in oppositely oriented repetitive elements [24] , [25] , [26] . When DNA damage occurs , the homology-dependent recombination-repair machinery recruits another copy of the repetitive element as a repair template . Inversions , deletions , and duplications occur when the resulting Holliday junction is incorrectly resolved . Whereas recombination among inverted repeats leads to inversions , recombination among direct repeats leads to deletion . When the recombination among direct repeats occurs during replication , the segment becomes deleted from one chromosome and duplicated in the other . Bacterial DNA replication appears to induce a multitude of mutational biases and selective forces with respect to their chromosome architecture [27] . Chromosomes are thought to remain small due to a general deletion bias [28] . Strand-specific oligomers such as χ sites [29] assist with DNA repair , while KOPS/AIMS [30] , [31] have roles in DNA replication and chromosome segregation . Such sequence signals would be disrupted by inversions within a single replichore , but not by inter-replichore inversions . Moreover , a large survey of Salmonella genomes in culture has provided evidence that genomes with equal-sized replichores ( balanced replichores ) may be under positive selection [32] . It is currently unknown whether symmetric inter-replichore inversions are frequently observed simply because they occur more frequently than other rearrangements ( a recombination bias ) , or whether other patterns of rearrangement commonly occur but are strongly selected against [26] . The observed frequency of rearrangement relative to neutral substitution is highly variable in different organisms . The frequency of observed rearrangement in modern genomes correlates with the presence of repeats induced by mobile genetic elements [26] , [33] . Interestingly , mobile genetic elements ( IS elements/transposons ) are also associated with the generation of pseudogenes , genome reduction , and adaptive evolution of niche change [34] . Large-scale inversion and deletion are both driven by homologous recombination among repeat elements . Taken together , these associations suggest that methods to predict episodes of ancient genome rearrangement may be able to uncover historical genome reduction and transitions in ecological niche . Studies of Yersinia have revealed extensive genomic rearrangement relative to conspecific isolates , and IS elements have been implicated in the rearrangement process . The recent availability of several finished Yersinia genome sequences offers the possibility to investigate patterns and biases associated with genomic rearrangement . Yersinia pestis played a role as the causative agent of the three major plague epidemics which together resulted in millions of deaths over the past two millenia [35] . Previous molecular studies have indicated that Yersinia pestis is a recently emerged clone of Y . pseudotuberculosis , with an estimated divergence less than 20 , 000 years ago [36] , although some ambiguity in the branching order of Y . pestis isolates remains [37] . Given its pathogenic lifestyle , Y . pestis population dynamics are different from those of non-pathogens and the effect of population dynamics on genome arrangement warrants consideration . Upon infection of a human host , Y . pestis likely undergoes expansive population growth . Transmission to a new host is usually mediated by a flea vector which can viably harbor only a small number of Yersinia cells compared to an infected human . As such , modern Y . pestis may have undergone several cycles of unconstrained population growth followed by extreme transmission bottlenecks . The unconstrained growth phase could permit generation of cell lines with genomic rearrangement , which are subsequently fixed by the transmission bottlenecks . Such population dynamics would serve to increase the observed rate of rearrangement . Previous experimental work has characterized patterns of genome arrangement in isolates of E . coli and Salmonella whose genomes were artificially perturbed in the laboratory [38] . Our study represents the first attempt to quantify selection and recombination bias acting on genome arrangement in a naturally evolving population .
We apply a Bayesian MCMC sampler to investigate selection and recombination bias acting on genome rearrangements in sequenced Yersinia isolates . At the time of this study , nine finished Yersinia genomes were publicly available , listed in Table 1 , and several more had been sequenced to draft quality . As the Yersinia pestis are very recently diverged , only a small number of nucleotide substitutions have been observed in fully sequenced genomes [39] , and efforts to reconstruct the Yersinia phylogeny have consequently been forced to integrate presence/abscence patterns of IS elements and VNTR sequences [37] . Pairwise comparisons of Yersinia genomes have revealed a large number of genomic rearrangements [25] , [40] which may be suitable phylogenetic characters . As large-scale genome rearrangement is thought to be a low-homoplasy molecular character [41] impervious to lateral exchange by homologous recombination , even a small number of rearrangements may suffice to resolve phylogenetic tree topology . In order to compute a rearrangement history , we require genomes to be encoded as a signed permutation matrix indicating order and orientation of homologous segments in each genome . We used the Mauve multiple genome alignment software to identify and align 84 Locally Collinear Blocks ( LCBs ) shared among the 9 Yersinia genomes . Differential gene content among Yersinia lineages precludes a nine-way alignment that completely covers each genome . On average 81 . 5% of each genome is contained within LCBs , and the remaining lineage-specific regions reside in breakpoint regions . The breakpoint regions cannot be unambiguously assigned to either neighboring LCB , and the uncertainty about their placement in ancestral genome arrangements causes corresponding uncertainty in ancestral replichore sizes . While Y . pestis and Y . pseudotuberculosis share a majority of their gene content , Y . enterocolitica has substantial differential content relative to the other eight taxa [42] . To mitigate inference problems related to differential gene content ( see Methods ) , we removed Y . enterocolitica from our analysis and computed an alignment on the remaining 8 taxa using a procedure described in Methods . The alignment of eight Y . pestis and Y . pseudotuberculosis strains , shown in Figure 1 , consists of 78 LCBs ( 79 before considering genome circularity ) that cover an average of 93 . 3% of each genome . The distribution of LCB lengths ( Figure 2 ) appears to be geometric , consistent with expectation under the Nadeau-Taylor random breakage model [14] . For the purpose of inferring ancestral replichore sizes , we divide each of the 78 breakpoint regions in half and assign each half to a neighboring LCB . The origin and terminus of replication in each genome were assigned on the basis of a consensus prediction and homology ( see Methods ) . We used a modified version of the BADGER 1 . 01b software to sample the posterior probability distribution of phylogenetic trees , mutation rate , and genome arrangement histories using inversions as mutation operations . The model treats all inversion events to be equally likely a priori , with no explicit preference for rearrangements that maintain or improve replichore balance . The prior distribution on branch lengths creates a strong preference for histories with fewer inversions . Like other Bayesian MCMC samplers for phylogenetics , the method used here creates an initial phylogenetic tree with mutation events mapped onto the branches , then repeatedly proposes modifications to the current tree topology , mutation history , and branch lengths . Any proposed modifications are accepted with probability dictated by the Metropolis-Hastings ratio [43] , [44] . The initial proposed reconstruction of inversion history typically has very low likelihood and proposed modifications are generally accepted until the likelihood reaches a local maxima . The initial period of sampling is commonly referred to as burn-in . Samples taken during burn-in are discarded since the Markov-chain has not yet converged to the true posterior distribution . As applied to the 78 Yersinia LCBs , we ran chains with 1 , 510 , 000 modification proposal steps , discarded the first 10 , 000 steps of each chain as burn-in and then subsampled every 50 steps ( details in Methods ) . The resulting posterior sampling consists of 30 , 000 complete genome arrangement histories . Each sampled history contains a tree topology with inversion events mapped onto the branches . In total , the sampled histories contain 30 , 000 tree topology estimates and 2 , 520 , 185 genome arrangements , of which 2 , 280 , 185 are inferred ancestral arrangements and 240 , 000 are modern genome arrangements . Visualization of the posterior distribution of trees using SplitsTree v4 [45] reveals a small amount of topological ambiguity as a splits network ( Figure 3 ) . Contributing to topological ambiguity are seven different tree topologies with parsimonious inversion histories of 79 events . All seven parsimonious topologies differ in their grouping of Y . pestis isolates . Nonetheless , the Y . pestis are found to be monophyletic , with subgroupings that are consistent with previously published genome analyses [39] . Application of a maximum parsimony algorithm to reconstruct inversion phylogeny recovers one of the seven parsimonious topologies identified by BADGER , also with 79 inversions [46] , [47] . Internal branches of the Y . pestis clade are very short relative to external branches , a phenomenon which could have numerous explanations including exponential population growth , population subdivision , an ancestral selective sweep , or recently accelerated mutation rates possibly associated with pathogen population dynamics or relaxed selection in culture . Of note , SNP phylogenies also exhibit short internal branches [39] . To quickly scan for patterns in the genome rearrangement history of Yersinia , we developed a 3D video system to visualize the series of rearrangement events . The posterior sampling of inversion history contains 30 , 000 samples . We selected the one history with maximum a posteriori probability and rendered the series of rearrangement events on each branch of the phylogeny using custom Java software . The chromosome is rendered as a torus with positions of the replication origin and terminus marked . The replichores present in an ancestral node of the tree are colored distinctively , left replichore in blue , right replichore in green . The intensity of the colors changes on a gradient from origin to terminus , such that segments near the origin in the ancestor are dark blue or green , while segments near the terminus are light . Supplementary Videos S1 , S2 , S3 , S4 , S5 , S6 , S7 , and S8 show the inversion history along each external branch of the maximum a posteriori tree estimate . Several striking patterns of rearrangement can be seen in the videos , especially those representing longer branches such as the branch leading to Y . pestis 91001 ( Video S3 ) . First , the terminus remains positioned mostly opposite the origin throughout the rearrangement history . Second , light-colored segments which were near the terminus in the ancestral genome arrangement tend to remain near the terminus . Third , when large inversions happen within a single replichore , they appear to be quickly followed by a second inversion that reverts the first . We now describe statistics to quantify the significance of these observations , along with other aspects of genome arrangement evolution that are not as easily recognizable through visualization . When the terminus of replication lies opposite the origin on the circular chromosome , replichore sizes are equal and the genome is said to be balanced . If we assume the origin is at positions 0 and 1 on the circular chromosome and the terminus dif site lies at some position b where 0<b<1 , we can quantify the degree of imbalance as the deviation from perfect balance: . Thus , a perfectly balanced genome with b = 0 . 5 will have 0 imbalance , and imbalance increases to 1 as the terminus dif site position b approaches 0 or 1 . Of the 2 , 520 , 185 sampled ancestral arrangements , 77 . 9% of the arrangements have a replichore within 20% of perfect balance , and 88 . 5% are within 30% of perfect balance . The full distribution of balance for ancestral arrangements can be gleaned from the historic terminus position plot in Figure 4A . To prove that the ancestral positioning of the terminus can not be explained by a series of inversions with arbitrary endpoints , we performed 30 , 000 simulations of replichore balance evolution in a genome that undergoes inversions with uniformly chosen endpoints . Comparison with the null model suggests it can not explain the observed data ( KS test , median p-value<10−1 ) . Even when the simulated terminus dif site position is restricted to the range observed in modern genomes , the null model cannot explain the observed genomic balance ( KS test , median p-value≈0 . 0001 ) . Not all modern genomes are balanced genomes . Y . pestis Pestoides F is conspicuously imbalanced , with a terminus position of 0 . 77 ( 54% imbalance ) . As such , we might ask whether the imbalance observed in ancestral genome arrangements is confined to the Y . pestis Pestoides F lineage . Figure 4B shows the imbalance observed on each external branch of the phylogeny , with internal branches pooled . Clearly all lineages undergo imbalance , although the Pestoides F isolate has a greater fraction of imbalanced genomes in its history . Surprisingly , the Y . pseudotuberculosis exhibit a high degree of imbalance as well . As they are sister taxa to Pestoides F , the imbalance could be attributed to imbalance at the common ancestor . In fact , the common ancestor is frequently predicted to have an imbalanced genome , and reconstructions with a balanced common ancestor require intermediate states of imbalance on branches leading to the modern Y . psuedotuberculosis genomes . Alternative explanations for the unusual terminus position in Y . pestis Pestoides F could be entertained , one such explanation being assembly error . As the assembly has been validated using a 40 kb Fosmid library , we do not believe this to be the case ( P . Chain , personal comm . ) . Another alternative is that the primary replication terminus has shifted to a different location in the Y . pestis Pestoides F lineage . Visual inspection of the rearrangement pattern for Y . pestis Pestoides F in Figure 1 reveals several instances of local overlapping inversions characteristic of symmetric inversion about the terminus ( seen as a “fan” pattern of crossing lines ) . If Pestoides F has indeed switched to a new primary terminus site it would introduce some error in our calculation of the historic replichore balance distribution . However , since only about 10% of inversions occur on the branch leading to Y . pestis Pestoides F , the error would be negligible . The error would serve to overdisperse the estimated balance distribution and result in weaker apparent bias towards replichore balance . Substantial ambiguity exists in the phylogenetic tree topology reconstructed from the Yersinia genome arrangements . BADGER found seven parsimonious topologies , and in total 48 unique topologies were sampled with inversion counts ranging from 79 to 87 . Parsimony has enjoyed a long history as a guiding philosophy in evolutionary inference , so it is of interest to know whether parsimonious reconstructions agree with our expectation of replichore balance in genome arrangements . The mean estimate of imbalance turns out to be slightly smaller for parsimonious histories and the variance is much lower , as shown in Table 2 . The difference in balance between parsimonious and other reconstructions is significant ( KS test , p<2e-16 ) but the difference is small ( KS D = 0 . 016 ) . If we believe that strong selection for balanced genomes exists and inversions not affecting balance are neutral , then parsimonious reconstructions appear slightly more favorable . Previous studies have suggested that DNA replication introduces a recombination bias that favors inversions with endpoints that are equally distant from the origin of replication [22] , [20] , so-called symmetric inversions . Given our inferred inversion histories , we can formally test for an excess of symmetric inversions . To do so , we introduce the following notation . Let V be the ordered set of inversions mapped onto tree branches in a sampled reconstruction of the inversion history , and let vi represent the ith inversion . Then we define a symmetry statistic for inter-replichore inversions as: ( 1 ) where OL ( vi ) is the distance between the origin and the left-end of the ith inter-replichore inversion , while OL ( vi ) is the distance between the origin and the inversion's right-end . Thus , the equation expresses the distance between inversion endpoints and the origin in each replichore , and computes the squared-difference of distances . Equation 1 assigns a perfectly symmetric inversion a value of zero , while asymmetric inversions take on large values . Incidentally , the symmetry statistic is agnostic to the choice of which replichore is the left or right . We would like to know whether the observed inversions are more symmetric than expected by chance . To do so , we use permutation to generate a distribution of symmetry statistics that represent the null hypothesis of lack of symmetry . We compute the symmetry statistic on arbitrary pairs of left and right inversion endpoints from inter-replichore inversions , according to the following equation: ( 2 ) More formally , we compute a null distribution by sampling x and y uniformly without replacement from the set of possible inter-replichore inversions . OL ( vx ) represents the distance from the origin to the left-side of inversion x , and OR ( vy ) is the distance from the origin to the right-side of inversion y . If the inversion endpoints on the two replichores were independent from each other , then we would not see a significant deviation from the null distribution . Deviation towards larger values would imply fewer symmetric inversions than expected , whereas deviation towards smaller values implies more symmetric inversions than expected . Comparison of symmetry statistics generated by Equations 1 and 2 demonstrates that within-replichore inversions are more likely to be symmetric than expected by chance ( KS test , median p = 0 . 0001 , mean D = 0 . 47 ) . The observed symmetry statistic distribution and the corresponding null distribution are shown in Figure 5 . Our inference method does not estimate event times but only relative event ordering , thus we are unable to directly infer the actual amount of time ancestral genomes have spent in a balanced state . However , if we define a state of imbalance as a percentage deviation from perfect balance , say a 20% deviation , then we can quantify the number of imbalance episodes that the organisms have undergone . The posterior estimate of the number of imbalance episodes the eight Yersinia have undergone is 3 . 26 ( σ = 1 . 82 ) , not counting episodes which span a bifurcation event in the tree . The posterior distribution is shown at left in Figure 6 . Similarly , we can define the duration of an imbalance episode as the number of mutation events ( inversions ) experienced before the chromosome returns to a balanced state . The length of imbalance episodes observed in our posterior sampling is shown at right in Figure 6 . If imbalance is strongly selected against , we might expect episodes of imbalance to be very short and more frequent than expected by chance given the total number of imbalanced arrangements . To determine whether the number and duration of imbalance episodes was unusual , we designed a permutation test in which the balance states along branches of reconstructed trees were randomly permuted ( see the Methods section for details ) . The permutation gives a null model of an organism which freely transitions to and from balance , spending the same total amount of time in each state as the Yersinia genomes . Surprisingly , we find the exact opposite of our initial expectation . There are fewer imbalance episodes than expected under the null model , and episodes of imbalance are longer than expected given the null model . The pattern is robust to choice of a particular balance threshold , as other thresholds up to 40% give similar results . Explanations might be that imbalance is only mildly detrimental , or that transmission bottlenecks periodically fix suboptimal genome arrangements in lineages of Y . pestis , despite their fitness disadvantage . Once imbalanced , several inversions typically occur before balance is restored . Given that the Y . pestis chromosome is littered with repetitive DNA , the observation is consistent with the notion that picking an arbitrary pair of oppositely oriented repeats is unlikely to yield an inversion that restores balance . Under such a hypothesis , the number of inversions occurring before restoration of balance should rise with the frequency of oppositely oriented repetitive DNA . Assuming that no selection or recombination bias acts on inversion length , the distribution of inversion lengths could be modeled as the distance between two uniformly chosen points on a circle with circumference 1 . However , 46 . 3% of sampled inversions act within a single replichore and we might expect such inversions to be short relative to inter-replichore inversions . Although they do not affect balance , inversions within a replichore act to reverse the polarity of x sites [29] , KOPS/AIMS [30] , [31] , and they also change leading/lagging strand A/T and G/C biases [48] , relative gene density [27] , and gene expression levels . As shown in Figure 7 , the observed length distribution for within-replichore inversions does indeed indicate that they are shorter than inter-replichore inversions . However , we expect inter-replichore inversions to be longer than within-replichore by definition , because inter-replichore inversions must have one endpoint in each replichore . To determine whether within-replichore inversions are significantly shorter than inter-replichore inversions , we develop a null model of inversion length that accounts for replichores . Replichore sizes change as the position of the terminus dif site changes over the course of evolution , thus the expected length of within-replichore and inter-replichore inversions changes . We assume that inversion endpoints are uniformly distributed and that no inversion acts on more than half the chromosome , otherwise a shorter complementary inversion operates on the other side of the circular chromosome . We can then define the expected length of a within-replichore inversion as: ( 3 ) ( 4 ) where 0<b<1 is the position of the terminus dif site relative to the origin of replication . We define a similar measure of expected length for inter-replichore inversions: ( 5 ) We provide a detailed derivation of these equations in the Methods section , and the values given by each equation for 0<b<1 are shown at left in Figure 8 . Knowing the expected length for each inversion , we compute the ratio of observed length to expected length for each inversion in the posterior sampling . The distribution of ratios for within- and inter-replichore inversions is given at right in Figure 8 . Both classes of inversion are shorter than would be expected under the null model . Comparison among within- and inter-replichore inversions reveals that within-replichore inversions are much more so than inter-replichore inversions ( KS test , median p = 0 . 002 , mean D = 0 . 41 ) . Previous study of Salmonella isolates has demonstrated that inversion of the origin relative to the terminus does not have a noticeable fitness impact , so long as balance is maintained [32] . Despite that , eight of the nine Yersinia genomes have the origin and terminus in identical relative orientation , which we term the canonical OriDif configuration ( see Table 1 ) . The configuration can be readily observed in Figure 1 by noticing that blocks containing the dif site ( purple ) are shifted upwards in every genome except Y . pseudotuberculosis IP31758 , as are blocks containing the origin ( extreme left and right in Figure 1 ) . If the canonical OriDif offers no selective advantage over the non-canonical configuration , then observation of the canonical OriDif can be modeled with a binomial distribution . Under the binomial , the probability of observing eight of nine genomes with the canonical OriDif is 0 . 018 , suggesting that a preference for the canonical OriDif configuration must exist . The genomes of Y . pestis Angola and Y . pseudotuberculosis YPIII were finished while this manuscript was under review and they too exhibit the canonical OriDif configuration , bringing the tally to 10/11 and p<0 . 01 . Of note , studies of mutation patterns in diverse bacteria suggest that replication terminates near the dif site itself , despite the presence of many additional ter sites [19] . Although it is tempting to generalize the canonical OriDif idea to other bacterial genomes , a cursory examination of related heavily rearranged Shigella genomes did not reveal a preference for a canonical OriDif configuration . That modern isolates favor the canonical OriDif configuration suggests that ancestral Yersinia would favor it as well , and probably also spend a noticeably greater amount of time in such a configuration . Most genome rearrangements in Yersinia ( 53 . 7% ) are inter-replichore inversions which swap canonical and non-canonical OriDif configurations . As such , the number of arrangements with the canonical OriDif is not substantially different from those which have the non-canonical arrangement . Given that modern genomes tend towards balance and a canonical OriDif , we might expect an association between balance and OriDif because an inversion that disrupts balance must also change the relative orientation of the origin and terminus . The left panel of Figure 9 shows overall balance of arrangements as a function of OriDif configuration . A significant association between balance and canonical OriDif can be seen ( KS test , median p = 0 . 0015 , mean D = 0 . 4 ) . Interestingly , when arrangements at internal nodes of the phylogeny are compared to branch arrangements , the association between canonical OriDif and balance appears to be more pronounced ( Figure 9 right ) . However , a comparison of balance at internal node arrangements with canonical OriDif versus branch arrangements with canonical OriDif fails to demonstrate a significant difference ( KS test , median p = 0 . 67 , mean D = 0 . 33 ) . Failure to find a significant difference may be due to lack of inferential power , since each inversion history sample has only six internal node arrangements from which to estimate the balance distribution . Additional finished Yersinia genome sequences would provide greater statistical power . The most-parsimonious inversion histories inferred by BADGER contain 79 inversion events , yet only 78 gene-order breakpoints exist in the Yersinia genomes . Clearly , some breakpoints must be used repeatedly . Previous breakpoint re-use studies [49] , [50] have typically relied on inferring the mere existence of reuse rather than identifying rearrangement hotspots . To do so , we must shift focus from breakpoints to inversion endpoints . Every inversion event acts to reverse one or more consecutive LCBs . The left side of the left-most and right side of the right-most reversed LCBs constitute the inversion endpoints . As such , we can count the number of times a given LCB boundary is used in an inversion history . By definition , every LCB boundary must be the endpoint of at least one inversion , however some LCB boundaries may be used more than once . Figure 10 shows the posterior estimate of usage for individual LCB boundaries , mapped according to their occurrence in the Yersinia pestis KIM genome . A striking pattern emerges in which inversion endpoints lie proximal to the origin of replication much more frequently than to the terminus . While inversions with endpoints near the terminus of replication do occur , they are comparatively rare . Experimental studies of genome rearrangement in E . coli and Salmonella have pointed towards the existence of chromosomal domains near the terminus that can not tolerate inversion endpoints [38] , termed the “impermissible zones” , or “non-divisible zones” . Yersinia appear to have a similar constraint , visible as the region immediately surrounding dif having 0 or 1 inversion endpoints . An alternative and very plausible explanation is the presence of AIMS proximal to the terminus of replication [31] . AIMS are polarized motifs that direct chromosomal segregation during cell division , and the density of such motifs increases with proximity to the terminus dif site . Reversal of a large AIMS-rich segment could severely disrupt chromosome segregation . In other Enterobacteriacae , frequent chromosomal inversion has been attributed to the presence of rRNA operons proximal to the origin [51]; as they are conserved in sequence , these operons provide a large substrate for homologous recombination . To investigate whether ribosomal RNA operons were involved in the large number of observed rearrangements we assessed the presence of rRNA operons in modern isolates . In Figure 10 , inversion endpoints which have an annotated ribosomal RNA gene within 1500 bp of the endpoint have been highlighted red and marked with R . Although the most commonly used inversion endpoint does border a ribosomal operon , the majority of heavily used endpoints do not . Instead , all but one of the remaining inversion endpoints have an annotated transposase or IS element ORF within 1500 bp . Thus the difference in observed inversion rate among ribosomal operons and transposable elements is not appreciable . If inversions with endpoints near the terminus are forbidden , then the relative terminus position has limited range with respect to the origin . Thus , we might revisit the question of whether the observed replichore balance distribution can be explained by a neutral model of inversion . As with the unconstrained model , simulations of replichore balance evolution which restrict the relative terminus position to the range of [0 . 25 , 0 . 75] fail to explain the observed distribution of replichore balance ( KS test , median p-value = 0 . 0001 ) . The Bayesian posterior distribution of the terminus position ( Fig . 4A ) shows that replichore balance has been largely maintained during the evolution of Yersinia genomes . To demonstrate that the observed pattern does not result from inversion followed by an immediate reversion with approximately the same endpoints , we introduce the following statistics . As above , let V be the ordered set of inversions for all edges in the tree and let vi refer to the ith inversion . We refer to the left endpoint of inversion vi as L ( vi ) and the right endpoint as R ( vi ) . Note that genome coordinates range from 0 to 1 , so that 0≤L ( vi ) ≤R ( vi ) ≤1 . We compute the following statistic for consecutive pairs of inversions vi and vi+1: ( 6 ) The value in Equation 6 is smallest when consecutive inversions have identical endpoints , in which case the second inversion effectively reverts the first inversion . However , since our Bayesian model of genome rearrangement favors histories with fewer overall inversions , it will only very rarely sample histories that contain consecutive inversions that perfectly cancel each other out . It will , however , sample consecutive inversions with nearby endpoints in an unbiased manner . Such a pattern of inversion could be caused by an unknown mutational or selective force that favors immediate reversion of inversions , such as a rebalancing inversion . Figure 11 compares the observed distribution for Equation 6 to a permuted distribution generated by pairing L ( vi ) −L ( vi+1 ) values with R ( vj ) −R ( ji+1 ) for i , j sampled uniformly without replacement . The observed distribution appears to be very similar to the permuted distribution . The difference is not significant ( KS test , median p = 0 . 86 , mean D = 0 . 1 ) , indicating that consecutive inversions with nearly equal endpoints are not observed more frequently than would be expected by chance alone .
We have identified several inversion patterns which deviate substantially from null expectation that all inversions are equally likely . Do our observations result from selection against some inversions , or is there a recombination bias which causes some inversions to occur more frequently than others ? Our statistics can not directly quantify the relative contributions of these two evolutionary forces . We might argue that balanced replichores result from weak-to-moderate positive selection . Our observation that episodes of imbalance are less-common than expected and last longer than expected could indicate that in general , imbalance is selected against , but when it occurs it is only mildly deleterious because balance is usually not immediately restored . Occasional relaxed selection on balance could be a function of pathogen population dynamics . On the other hand , a similar pattern could be induced by a recombination bias which usually preferred inversions with endpoints equidistant from the origin . Imbalance would be occasionally introduced by an inversion with endpoints of unequal distance from the origin , and because rebalancing requires a second inversion with endpoints of unequal distance from the origin , it may take many inversions to restore balance . Our observation that Yersinia has a canonical OriDif configuration seems most easily explained by natural selection . A recombination bias introducing such a pattern would have to cause inter-replichore inversions to occur almost exclusively in pairs , and to our knowledge , no plausible molecular mechanism has been described which could achieve such a feat . Incidentally , if the canonical OriDif results from selection it implies that some symmetric inversions may be mildly deleterious in Yersinia . Our observation that inversions with endpoints near the terminus are much less frequent than inversions with endpoints near the origin could be explained by selection against such inversions . If Yersinia is under reduced selection for growth rate , it may be more tolerant of inversions near the origin . Closely related organisms such as E . coli are known to have several ter binding sites throughout the half of the chromosome surrounding the terminus dif site . The ter sites are polarized motifs , such that they halt replisome procession only in one direction [16] . As such , a within-replichore inversion involving a ter site may result in a lethal disruption of DNA replication . A similar deleterious effect could be envisioned when inverting AIMS-rich segments . We might also entertain recombination bias as an explanation for the excess of inversions with endpoints near the origin . Fast-growing bacteria are known to have multiple replication forks [52] . If the regions near the origin of replication exist in higher copy number they may be more prone to rearrangement , but higher copy number would also result in higher effective population size ( Ne ) which might be expected to counteract the effect of a higher mutation rate . In any case , Figure 10 exhibits a precipitous shift from high inversion rate to low rate moving away from the origin . Although a plausible mechanism exists for selection against within-replichore inversions proximal to the terminus , the reasoning does not apply to inter-replichore inversions , which account for over half of all inversions . Given that the rate of inversion is about three times higher near the origin , it seems likely that additional unknown forces of recombination bias or selection play a role in reducing the inversion rate near the terminus . Accurate genome arrangement phylogenies have the potential to provide a reference phylogenetic tree topology against which hypotheses of recombination , gene conversion , and lateral gene transfer can be tested . Chaisson et al [53] demonstrated that carefully filtered mammalian microinversion markers could be used as binary characters that form a perfect phylogeny , and a similar approach could be envisioned for microbes . Although Chaisson et al claim that rearrangements are low-homoplasy characters based on the ability of their ( carefully filtered ) data to pass the four-gamete test , three confounding factors stymie such simple approaches to rearrangement phylogeny when studying complete genome arrangements . First , rearrangement mutations frequently overlap each other , creating inter-dependence and thus precluding a clear representation of mutations as binary characters . Second , population-level variability in genome arrangement has been reported in both microbes [54] and mammals [55] , implying that lineage-sorting effects may yield genome arrangement trees that do not match the species tree . Finally , programmatic rearrangement [56] , [57] not only introduces population-level variability , but can repeatedly invert the same chromosomal segment , potentially resulting in frequent homoplasy . It should be emphasized that PCR-based assays have identified mixtures of genome arrangements in laboratory cultures of Y . pestis [54] . If genome rearrangements such as symmetric inversions are nearly-neutral mutations , we would expect their frequency in the population to approximately follow a Wright-Fisher model . Thus , populations with a high rearrangement rate are likely to have more than one genome arrangement present . To our knowledge , no evidence of programmatic rearrangement mutations in Y . pestis has been reported that would be likely to cause frequent reversion and homoplasy in large-scale rearrangement mutations . Such effects have been observed as part of phase variation in other microbes [56] . Whilst rich stochastic models of nucleotide sequence evolution have been developed , comparatively little effort has gone into development of stochastic models of genome arrangement evolution . Inversions are known to affect a variety of genomes , including mitochondria [58] , plastids [59] , [60] and bacteria . However , mutational processes such as transposition or segmental duplication and loss [61] can also result in genomic rearrangement , and can have an especially profound effect on eukaryotic and mitochondrial gene order . Future efforts to model genome arrangement evolution should undoubtedly address duplication/loss . Although bacteria are usually unichromosomal , they also have plasmids and other short circular chromosomes that might play an important role in rearranging the genetic material . Therefore a Bayesian MCMC method for multichromosomal genome arrangement phylogeny would also be desirable . Pairwise models of multi-chromosomal rearrangement via circular intermediates have recently been derived , although not in a Bayesian context [62] , [63] , [64] . The rearrangement patterns inferred by our study should prove valuable as a guide for phylogenetic inference when the inversion history signal has become saturated . The Yersinia genomes studied here appear to lie precisely on the verge of saturation , as seven parsimonious topologies were discovered . Just as codon models and gamma-distributed rate heterogeneity have aided phylogenetic inference on nucleotides , models of rearrangement which explicitly acknowledge that not all genome arrangements are equally likely may be useful to disambiguate phylogenetic signal in saturated inversion histories . Pairwise study of eukaryotic genome arrangement has demonstrated preference for particular types of rearrangement events [65] , and methods similar to ours could conceivably be extended to identify selection on arrangement from phylogenies of multi-chromosomal eukaryotic genomes . A non-phylogenetic , pairwise model of rearrangement by inversion has previously been used to investigate the preference for historic replichore balance in bacteria [66] . Using randomly simulated genome arrangements as a baseline , the authors conclude that historical replichore balance has been significantly maintained in a variety of bacteria , but not all . Our Bayesian method improves on their model by allowing us to gauge more rigorously the degree of statistical confidence and uncertainty in reconstructions of inversion history . Moreover , our method avoids a systematic bias when exploring possible inversion histories . The distribution sampled by the Ajana et al method is not uniform over equally parsimonious inversion scenarios , but is skewed to favor particular mutation events . The difference between their sampling distribution and the uniform distribution can grow exponentially in some cases ( [67] , section 5 . 2 ) .
We used the Progressive Mauve algorithm [68] to compute an alignment of the nine genomes listed in Table 1 . Analysis of the resulting alignment indicated that Y . enterocolitica 8081 contains substantial gene content differences with respect to the other Yersinia genomes , with only 81 . 5% of an average Yersinia genome conserved among all nine taxa . Current Bayesian models of genome arrangement do not model gain and loss of genetic material , thus we removed Y . enterocolitica 8081 from further analysis . An alignment of the eight Y . pestis and Y . pseudotuberculosis genomes was constructed using the default mauveAligner parameters . The resulting LCBs were inspected using the Mauve alignment viewer and the minimum LCB weight was adjusted to a value which eliminates LCBs consisting of only repetitive elements ( LCB Weight 600 ) . We then computed a full alignment with minimum LCB weight 600 , and processed the resulting XMFA alignment file into a permutation matrix in BADGER format ( Dataset S1 ) . We apply the Bayesian model of genome rearrangement by inversion implemented in the BADGER software [69] . BADGER models genomic inversions as a continuous-time Markov process occurring along branches of an unrooted phylogenetic tree which relates organisms . All inversion events are modeled to be equally likely a priori . This enables us to calculate the likelihood of a genome rearrangement history mapped onto a tree given the tree and mutation rates , see e . g . [70] . Branch lengths are measured as the number of mutations on a branch , with lengths modeled using an exponential distribution . The mean value of the exponential distribution is given a hyper-prior which creates a strong preference for shorter overall branch lengths and thus assigns higher posterior probabilities to parsimonious inversion histories . BADGER samples from the joint posterior distribution of tree topologies , inversion histories , and mutation rates using Metropolis-coupled Markov-chain Monte Carlo , also known as MCMC with Parallel Tempering [71] . Accurate inference using MCMC methods requires Markov-chain convergence and adequate mixing . In general , MCMC samplers for genome rearrangement appear to mix very slowly because the likelihood surface can be rugged , and good proposal mechanisms for transitioning between peaks may not exist . The use of heated parallel chains ( Metropolis coupling ) can alleviate the problem to some extent [72] . The Parallel Tempering method first considers the Bayesian posterior distribution as a Boltzmann distribution at unit temperature . The probability of a particular state X in a Boltzmann distribution is defined as ( 7 ) where ΔG ( X ) is the free energy , e is the natural number and T is the temperature . Since we are talking about hypothetical energies and temperatures , we omit the Boltzmann- or gas-constant ( k or R ) in the formula . Setting T = 1 leads to defining the free energy of a state as ( 8 ) After defining the free energy for each state , the Parallel Tempering runs several chains with different temperatures , the unheated chain has temperature 1 , the heated chains have higher temperatures . All chains converge to their own prescribed Boltzmann distribution , but sometimes they swap states . The swapping is governed by the Metropolis rule ( [43]; hence the name , Metropolis-coupled MCMC ) , which guarantees that swapping does not change the convergence to the prescribed distributions . The probability surface is flat at high temperatures , which provides fast mixing in the state space , while the swappings between the unheated and heated chains allow the possibility that the unheated chain can jump from one local minimum into another one . In our application to the Yersinia LCBs , we used a Metropolis-coupling scheme with temperatures ranging from 1 to 1 . 18 to ensure adequate mixing . A comparison of runs with 3 , 5 , 19 , and 49 heated chains revealed that only runs with 19 or 49 heated chains discovered all seven parsimonious topologies within 500 , 000 MCMC steps . Monitoring the log-likelihood plot and comparison among the runs suggests that the chains have converged and mixed sufficiently to support the inferences described in the present work . To make inference on ancestral genome arrangements , we modified the BADGER C++ code to record inversion histories at each subsample point . Additional software was implemented to summarize the resulting posterior samples of genome arrangement . All software is available from http://bioinformatics . org . au/barphlye . Despite exclusion of Y . enterocolitica from the genome rearrangement phylogeny , it remains a potentially useful outgroup for rooting the tree using a molecular character such as nucleotide substitutions . Debate rages over the proper method to infer phylogenies using large multi-gene or whole-genome datasets . Recombination , lateral exchange , lineage sorting , and other natural processes can result in a phylogenetic signal that varies widely from gene to gene . One attempt to acknowledge and mitigate the impact of such effects is the recently proposed concordance factor approach , which provides a method to infer the fraction of a genome supporting a given hypothesis of vertical inheritance [73] . We apply Bayesian tree concordance statistics to estimate support for alternative rootings of the phylogenetic network shown in Figure 3 . An analysis of 30 randomly selected genes gives an a posteriori concordance factor of 19 . 4 ( out of 30 , 90% confidence interval [10] , [28] ) supporting a root on the branch leading to Y . pseudotuberculosis IP31758 . An alternative rooting on the branch leading to Y . pseudotuberculosis IP32768 garners a concordance factor of only 7 . 5 , with a 90% confidence interval of [0 , 17] . The concordance factor analysis suggests that recombination and lineage sorting in Yersinia has caused inconsistent phylogenetic signal throughout the genome , but that a greater fraction of sampled genes support a rooting on Y . pseudotuberculosis IP31758 . Such frequent large-scale homologous recombination has recently been reported as a common feature in other Enterobacteriacae [74] , [75] . Interestingly , the concordance tree splits weakly support placement of Y . pestis Pestoides F as a sister taxa to Y . pestis KIM , whereas the inversion phylogeny places the Pestoides F lineage as ancestral to the remaining Y . pestis with high confidence . Although we discarded Y . enterocolitica due to presence of differential gene content , the eight remaining genomes contain some lineage-specific content as well . Differences in gene content imply that observed LCB lengths are different in each modern genome . Moreover , breakpoint regions may contain lineage specific content . To perform inference on ancestral replichore balance with a model that lacks gene gain and loss , it was necessary to assign a length to each LCB and to account for the portion of each chromosome in breakpoint regions . We took a reference-genome approach based on Y . pestis KIM , which represents a median in terms of genome size among the eight Yersinia genomes studied . We assigned half of each breakpoint region to its neighboring LCB in Y . pestis KIM , and took the resulting LCB lengths as representative of all genomes . An average of 6 . 7% of each modern genome lies in breakpoint regions , and genome size deviates from Y . pestis KIM by +/− 3% . Thus , our use of a reference genome introduces some error into estimates of ancestral replichore sizes . In the worst case , the error could be as large as 10% , but the average error is small enough that it does not affect the main conclusions described here . The origin and terminus of replication in Y . pestis KIM was previously identified as occurring at approximately 1 bp and 2 . 324 Mbp , respectively [25] . Here , the terminus refers to a point on the chromosome where strand-specific oligomer skew shifts abruptly to the opposite strand . Others have reported that the change in oligomer skew typically occurs near the terminus dif site [19] , and so we use the site of strand bias change as a proxy for the true dif site . The ori and dif sites were assigned in other genomes on the basis of homology to Y . pestis KIM . The predicted dif site lies in the middle of a large 140 Kbp segment conserved among all Yersinia genomes at >95% sequence identity ( see Figure 1 ) . Similarly , the predicted origin lies in the middle of a 53 Kbp segment conserved among all Yersinia at >95% sequence identity . Comparison of our origin and terminus predictions to those made by an automated prediction system [76] reveals that our predictions agree with those made by the automated system within 1 kbp in nearly all cases . Discrepancy occurs in the terminus prediction for Y . pestis 91001 . The discrepancy seemingly results from numerous recent rearrangements having disrupted the signal of strand-specific oligomer skew and in turn confusing the automated system . We report analysis on 30 , 000 samples from the posterior distribution of inversion histories . We assume that Yersinia has one true evolutionary history , and that at most one of the inferred histories represents the true history . As such , when comparing the distributions of quantities of interest , we do so on a per-sample basis using the Kolmogorov-Smirnov test . We take the median p-value over the 30 , 000 tests to be an estimator of the p-value which would be obtained had the test been applied to the one true history . We report mean D values as average estimates of the difference between target distributions . We use random permutation to generate a null distribution of the number and duration of episodes of imbalance . A tree sample with inversions mapped onto its branches has one genome arrangement for each leaf ( 8 in total ) , one arrangement for each internal node ( 6 in total ) , and some number of intermediate arrangements along each branch of the tree . For each sample in the posterior distribution of trees and inversion histories , we assign imbalance values the intermediate genome arrangements in the sample . For each branch of a given tree sample , we generate a permuted distribution by randomly shuffling the imbalance values of intermediate genome arrangements on that branch . We then count the number of transitions to and from imbalance along the original branch and along the branch with permuted values . Thus , the randomly permuted data have the same total number of balanced and imbalanced states with the same balance values , but any clusters of imbalanced states will be uniformly random . Our permutation approach disregards the actual inversion events , but generates random permutations with the same overall balance values . It is not possible to construct a random permutation of imbalance values by shuffling the inversion events themselves , since overlapping inversion events have strong ordering constraints and violation of these constraints would often change the imbalance values . Moreover , a strategy which samples inversion events uniformly at random would not yield a set of balance values consistent with the set we desire to permute . Assume the endpoints of an inversion are in positions x and y , with x , y∈[0 , 1] . The inversion length can be expressed as the function min{|x−y| , 1−|x−y|} , since the inversion occurs on a circular chromosme of length 1 and for any inversion longer than 0 . 5 , a complementary inversion with shorter length exists . If we assume that the inversion endpoints are uniformly distributed , then the expected length is the integral average of the function min{|x−y| , 1−|x−y|} over the appropriate area A: ( 9 ) where |A| denotes the size of the area . In the case of within-replichore inversions , area A is the union of the two squares as delineated by the dashed line of Fig . 12 , in case of inter-replichore inversions , A is the union of the two rectangles . For simplicity we suppress the full details of integration , and the resulting equations for within- and inter-replichore inversions are given in Equations 4 and 5 , respectively . | Whole-genome sequencing has revealed that organisms exhibit extreme variability in chromosome structure . One common type of chromosome structure variation is genome arrangement variation: changes in the ordering of genes on the chromosome . Not only do we find differences in genome arrangement across species , but in some organisms , members of the same species have radically different genome arrangements . We studied the evolution of genome arrangement in pathogenic bacteria from the genus Yersinia . The Yersinia exhibit substantial variation in genome arrangement both within and across species . We reconstructed the history of genome rearrangement by inversion in a group of eight Yersinia , and we statistically quantified the forces shaping their genome arrangement evolution . In particular , we discovered an excess of rearrangement activity near the origin of chromosomal replication and found evidence for a preferred configuration for the relative orientations of the origin and terminus of replication . We also found real inversions to be significantly shorter than expected . Finally , we discovered that no single reconstruction of inversion history is parsimonious with respect to the total number of inversion mutations , but on average , reconstructed genome arrangements favor “balanced” genomes—where the replication origin is positioned opposite the terminus on the circular chromosome . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"genetics",
"and",
"genomics/microbial",
"evolution",
"and",
"genomics",
"evolutionary",
"biology/evolutionary",
"and",
"comparative",
"genetics",
"genetics",
"and",
"genomics/chromosome",
"biology",
"evolutionary",
"biology/genomics",
"computational",
"biology/evolutionary",
"... | 2008 | Dynamics of Genome Rearrangement in Bacterial Populations |
Clonorchiasis is among the most neglected tropical diseases . It is caused by ingesting raw or undercooked fish or shrimp containing the larval of Clonorchis sinensis and mainly endemic in Southeast Asia including China , Korea and Vietnam . The global estimations for population at risk and infected are 601 million and 35 million , respectively . However , it is still not listed among the Global Burden of Disease ( GBD ) and no disability weight is available for it . Disability weight reflects the average degree of loss of life value due to certain chronic disease condition and ranges between 0 ( complete health ) and 1 ( death ) . It is crucial parameter for calculating the morbidity part of any disease burden in terms of disability-adjusted life years ( DALYs ) . According to the probability and disability weight of single sequelae caused by C . sinensis infection , the overall disability weight could be captured through Monte Carlo simulation . The probability of single sequelae was gained from one community investigation , while the corresponding disability weight was searched from the literatures in evidence-based approach . The overall disability weights of the male and female were 0 . 101 and 0 . 050 , respectively . The overall disability weights of the age group of 5–14 , 15–29 , 30–44 , 45–59 and 60+ were 0 . 022 , 0 . 052 , 0 . 072 , 0 . 094 and 0 . 118 , respectively . There was some evidence showing that the disability weight and geometric mean of eggs per gram of feces ( GMEPG ) fitted a logarithmic equation . The overall disability weights of C . sinensis infection are differential in different sex and age groups . The disability weight captured here may be referred for estimating the disease burden of C . sinensis infection .
Clonorchis sinensis is one of the important medical trematodiases and can cause clonorchiasis . Humans are infected through ingesting raw or undercooked fish or shrimp containing the larval ( metacercariae ) [1] . The adult worms live mainly in the intrahepatic bile ducts and the gallbladder , occasionally in the extrahepatic and pancreatic ducts . Most of the infected are asymptomatic , but some can develop into various symptoms and signs/complications [1] , [2] . That is associated with infection intensity , duration , number of re-infection and susceptibility of the host [2]–[4] . The symptoms include fever , abdominal discomfort , weakness , anorexia , nausea , diarrhea , acute pain in the right upper quadrant and so on . Chronic infection can lead to mild changes such as inflammation , light change of liver parenchyma , thickening and expansion of ducts , ambiguity and thickening of gallbladder walls and severe sequelaes including gallstone , cholecystitis , cholangitis , hepatomegaly , cyst of liver , hypertrophy of gallbladder , polyp of gallbladder , etc [1]–[5] . The most severe clinical sequelae is cholangiocarcinoma ( CCA ) . Owing to several studies demonstrating the relation between C . sinensis infection and CCA [6]–[9] , it was reassessed as “carcinogenic to humans” ( Group 1 ) by the International Agency for Research on Cancer ( IARC ) in 2009 [10] . Nearly 1 000 CCA cases related to C . sinensis infection occur annually in People's Republic of China ( P . R . China ) , which may still be underestimated [11] , [12] . Although C . sinensis is mainly endemic in Southeast Asia , including P . R . China , Korea and Vietnam [1] , [2] , [13] , clonorchiasis may occur in other parts of the world where there live immigrants from endemic areas [14]–[16] . The global estimations for population at risk and infected are 601 million and 35 million , respectively [1] , [17] , [18] . A population of 12 . 5 million was estimated to be infected in P . R . China mainland in 2003 , especially in southeast and northeast areas [19] . Owing to long term treatment and control efforts , other parasitic diseases such as schistosomiasis and soil transmitted helminthiases declined significantly in P . R . China [20] , [21] . However , the prevalence of C . sinensis increased by 75% from 1990 to 2003 , based on national sampling surveys [19] . Clonorchiasis is among the most neglected tropical diseases [22] . It has not been listed among the Global Burden of Disease ( GBD ) [18] , [23] . No completely clear information yet on how severe clonorchiasis is makes it difficult to compare its impact with the morbidity of other parasitic diseases . To understand the clear epidemiological picture , WHO launched the initiative to assess the global burden of foodborne diseases including clonorchiasis in 2006 [24] . Nevertheless , up to date , it is still lack of evidence-based estimation on burden of clonorchiasis , especial the population-based data . As hoc , disability weight is still not available . Disability weight reflects the average degree of loss of life value due to certain chronic disease condition and ranges between 0 ( complete health ) and 1 ( death ) [23] . It is crucial parameter for calculating the morbidity part of any disease burden in terms of disability-adjusted life years ( DALYs ) that is a standard unit for health measurement in GBD . Therefore , one community-based study in P . R . China was carried out to capture the disability weight of C . sinensis infection through model simulation .
DALYs as a standard unit for health measurement is consisted of premature mortality ( years of life lost , or YLLs ) and disability ( years of life living with a disability , weighted by the severity of the disability , or YLDs ) [25] . YLLs caused by C . sinensis infection is mainly attributable to CCA , which can be estimated separately . CCA can also contribute to YLDs , which can also be calculated separately with the disability weight referring to other cancers . Therefore , this study attempted to capture the overall disability weight of non-fatal sequelaes caused by C . sinensis infection ( Figure 1 ) . Based on the probability and corresponding disability weight of single sequelae , the overall disability weight of non-fatal sequelaes can be calculated as followings [26]:Where Psequelae k represents the probability of the kth sequelae , Dsequelae k represents the disability weight of the kth sequelae , and n is the total number of sequelaes included . Additive effect is adopted when co-sequelaes occurs in the same person , which is implicitly indicated in the formula and the current practice of GBD . A village called Shibo in Shunde district , Guangdong province , P . R . China was selected as study area . The economic development of the village was relatively high and the per capita annual net income reached 11 800 RMB ( about 1 815 dollars ) in 2010 . The selection is based on following reasons . First , Guangdong province , located in the southeast of P . R . China , has the highest prevalence of C . sinensis [19] . Second , Shunde district ranks among the top infection in Guangdong due to special diet habits of local people [27] . Third , the Shibo village has not yet received administration of mass drug . Villagers were motivated by local administrators to participate in this study . Firstly , one stool sample was collected from each voluntary participant . Triple Kato-Katz thick smears were prepared for each sample , and then microscopically examined by skilled technicians to qualify and quantify helminthes' eggs [28] . Eggs per gram of feces ( EPG ) was calculated by multiplying the egg number of every smear by 24 and then computing the average of three smears . Secondly , those infected with C . sinensis but without other helminthes were further investigated through questionnaire and ultrasound examination by trained investigators and clinical doctors , respectively . The questionnaire included 3 parts , i . e . demographical characteristics , recent health status ( past one month ) and past medical history . The content of health status included symptoms such as diarrhea , pain in the right upper quadrant , weakness , headache , abdominal distension , anorexia and nausea , which were used to assess sequelaes attributable to C . sinensis infection . The past medical history included hepatitis , diabetes , hypertension , hypotension , gastrosis and gynecological disease , which were used to exclude important confounders . The ultrasound examination involved liver and gallbladder . Any evidence of light change of liver parenchyma , thickening and expansion of ducts , ambiguity and thickening of gallbladder walls , gallstone , cholecystitis , cholangitis , cholecystectomy , hepatomegaly , cyst of liver , hypertrophy of gallbladder and polyp of gallbladder was recorded . After the investigation , the examination results were fed back to the participants . Those infected with C . sinensis and ( or ) soil transmitted helminthes ( STHs ) were treated with albendazole , free of charge , according to Guangdong provincial guideline for parasitic diseases control . Those with severe symptoms and ( or ) signs/complications were advised to search for further examination and treatment in hospital . The disability weights of diarrhea , low back pain/abdominal pain , gallbladder and bile duct disease , and mild/moderate hepatomeglay were available in GBD , Australia Burden of Disease ( ABD ) or other literatures [26] , [29]–[31] . The sequelaes without corresponding disability weight , namely cyst of liver and hypertrophy of gallbladder , were assigned the same disability weight of similar sequelae , i . e mild/moderate hepatomeglay . Other symptoms such as weakness , headache , abdominal distension , anorexia and nausea were not included for analysis because of their mildness and non-specificity . Similarly , other signs such as light change of liver parenchyma , thickening and expansion of ducts , ambiguity and thickening of gallbladder walls were also excluded . In total , two symptoms and eight signs/complications were included for estimating the overall disability weight caused by C . sinensis infection ( Table 1 ) . To estimate the uncertainty of the results , Monte Carlo simulation was applied by WinBUGS software ( http://www . mrc-bsu . cam . ac . uk/bugs/winbugs/contents . shtml ) . Eight subgroups were analyzed , i . e . total population , the male and female , the age group of 5–14 , 15–29 , 30–44 , 45–59 and 60+ . Following 2 000 pre-iterations , 20 000 iterations were run , when the models were all convergent . The frequency and disability weight of each sequelae were fed to the models and uniform distribution was applied . Due to the lack of priors , they were set as follows . Firstly , it is probable that all infected will fall with the two symptoms , i . e . diarrhea and pain in the right upper quadrant . Consequently , their priors were set from 0 to 1 . Secondly , unlike the symptoms , the development of signs/complications was chronic and accumulated progress [32] , so their priors should be narrowed . Here , they were set based on the data from the community investigation . To attain conservative results , prior of each sign/complication was set from 0 to the value which was the maximum probability of the eight signs/complications in each group . For example , in the total population , the probability of hypertrophy of gallbladder had the maximum value of 0 . 104 in the eight signs/complications , so priors of all eight signs/complications were set from 0 to 0 . 104 . The same means was adopted for other groups except the age group of 5–14 . In the age group of 5–14 , the investigated probabilities of all eight signs/complications were 0 , due to lower infection intensity and chronic progress in development of morbidity [32] . It was reasonably assumed the probabilities of above signs/complications were also approaching to 0 in this group even huger population involved . Therefore , their priors were more narrowed , set from 0 to 0 . 0001 . The predicted probabilities , disability weights and 95% confidence interval ( CI ) were outputted . The WinBUGS codes used for Monte Carlo simulations were listed in File S1 . Data were double-entered and cross-checked by EpiDate3 . 1 software ( http://www . epidata . dk/ ) . Analysis was run in SPSS for Windows ( version11 . 0; SPSS Institute , Inc . , Chicago , IL ) . The EPG was logarithmically transformed into lg ( EPG ) , which approached to normality . Student's t test was used for comparing infection intensity in different sex . Analysis of variance ( ANOVA ) was employed for comparing it in different age groups , and then Dunnett test was adopted for comparison between the group of 5–14 and others . Pearson test was used to assess the association between various categories . Statistical significance was given at a p-value of 0 . 05 . To attain geometric mean of EPG ( GMEPG ) , the average of lg ( EPG ) was calculated and inversely logarithmically transformed . To detect important sequelaes in the models , attributable proportion was analyzed , which denoted the proportion of disability weight of single sequelae taking in the overall one . Those contributing over 10% were considered as important ones . The relationship between disability weights and GMEPG was explored through mathematical function . Ethical clearance had been granted by the Ethics Committee of the National Institute of Parasitic Diseases , Chinese Center for Disease Control and Prevention in Shanghai , P . R . China ( Ref No: 20100525-1 ) . The objectives , procedures and potential risks were orally explained and informed to all participants . And a written consent form was also obtained from each participant with signature of him or his proxy .
Out of 6 882 villagers , 1 385 participated in fecal examination . 519 persons were infected with C . sinensis and ( or ) STHs , and none with Schistosoma japonicum . Among 505 mono-infected with C . sinensis , 293 participated in the questionnaire investigation and ultrasound examination . Because hepatitis may result in some similar sequelaes as C . sinensis infection , 28 persons with hepatitis were excluded from further analysis . In addition , 6 persons without ultrasound examination result were also excluded . Therefore , 259 persons were included for final analysis , of which 167 were negative for any of the ten sequelaes . The number of persons with 1 , 2 , 3 , and 4 sequelaes was 65 , 23 , 3 and 1 , respectively ( Figure 2 ) . There was no difference for sex composition in different age groups ( p>0 . 05 ) . The GMEPG of the male and female were 435 and 281 , respectively ( p<0 . 05 ) . The GMEPG increased with the age ( p<0 . 05 ) and it was significantly higher in age group of 30–44 , 45–59 and 60+ compared to that in age group of 5–14 ( p<0 . 05 ) ( Table 2 ) . The frequencies and probabilities of ten sequelaes included for analysis through community investigation were listed in Table 3 , Table 4 and Table 5 . In addition , the frequencies of weakness , headache , abdominal distension , anorexia and nausea were 20 , 43 , 23 , 15 and 5 , respectively , while that of light change of liver parenchyma , thickening and expansion of ducts , ambiguity and thickening of gallbladder walls were 53 , 63 and 64 , respectively . The frequencies of diabetes , hypertension , hypotension , gastrosis and gynecological disease were 4 , 22 , 7 , 39 and 10 , respectively . The predicted probabilities , disability weights and 95% CI were listed in Table 3 , Table 4 and Table 5 . The overall disability weight was 0 . 075 ( 95% CI: 0 . 060–0 . 091 ) ( Table 3 ) . The overall disability weights of the male and female were 0 . 101 ( 95% CI: 0 . 079–0 . 126 ) and 0 . 050 ( 95% CI: 0 . 035–0 . 067 ) , respectively ( Table 3 ) . The overall disability weights of the age group of 5–14 , 15–29 , 30–44 , 45–59 and 60+ were 0 . 022 ( 95% CI: 0 . 008–0 . 043 ) , 0 . 052 ( 95% CI: 0 . 033–0 . 071 ) , 0 . 072 ( 95% CI: 0 . 049–0 . 097 ) , 0 . 094 ( 95% CI: 0 . 069–0 . 122 ) and 0 . 118 ( 95% CI: 0 . 079–0 . 165 ) , respectively ( Table 4 and Table 5 ) . Due to the difference of probability and disability weight of each sequelae , their attributable proportion was also differential ( Table 3 , Table 4 and Table 5 ) . In the total population , gallstone and cholangitis took the most proportion , in sum of 56% . Gallstone and cholangitis also took the most proportion in the male , while gallstone , cholecystitis and polyp of gallbladder all took over 10% in the female . In the age group of 5–14 , clinical symptoms , i . e . diarrhea and pain in the right upper quadrant , were only attributable proportion . In the age group of 15–29 , diarrhea , gallstone , cholecystitis , cholangitis , cholecystectomy and polyp of gallbladder all took over 10% but less than 20% . In the age group of 30–44 , cholangitis , gallstone and polyp of gallbladder ranked the top three proportion . In the age group of 45–59 , gallstone and cholangitis took 36% and 22% , respectively . In the age group of 60+ , gallstone , cholangitis and hypertrophy of gallbladder took 40% , 11% and 11% , respectively . To demonstrate the relationship between disability weight and infection intensity quantitatively , mathematical function was constructed . Following the increase of GMEPG in different age groups , disability weight raised proportionally . They fitted a logarithmic equation as follows: y = 0 . 0362ln ( x ) −0 . 1269 , where y stood for disability weight and x represented GMEPG . The correlation coefficient ( R2 ) arrived at 0 . 9757 ( p = 0 . 002 ) ( Figure 3 ) .
Liver flukes infection including C . sinensis has not been included in the GBD [18] , [23] in spite of their heavy morbidity including CCA [10] . Therefore , it is the first time to present the evidence-based disability weight of liver flukes infection captured from population-based survey and model simulation . The overall disability weight of non-fatal sequelaes caused by C . sinensis infection is 0 . 075 . The overall disability weights of the male and female are 0 . 101 and 0 . 050 , respectively . And they are 0 . 022 , 0 . 052 , 0 . 072 , 0 . 094 and 0 . 118 in the age group of 5–14 , 15–29 , 30–44 , 45–59 and 60+ , respectively . Obviously , the disability weight of the male is higher than that of the female , and the more age is , the more disability weight is . That is coherent with the difference of the GMEPG in different groups . The difference of disability weight essentially demonstrates the variance of infection intensity , in other words , the worm burden . Compared to the female , the male prefer more to eating raw fish and especially they have more chance to doing this at gathering [19] , [33] . Therefore , higher exposure leads to higher infection intensity . The adult worms of C . sinensis can survive in human body for many years [1] . The development of major sequelaes , especially the important signs/complications , is chronic progress [32] . Furthermore , following the age's raise , the re-infection will increase . Hence , more morbidity occurs in the older . Gallstone is one of the most characteristic and frequent pathological features in clonorchiasis [5] , [34] . Therefore , it takes the major attributable proportion in overall disability weight in most groups . What is an exception is in the age group of 5–14 , where no clinical signs/complications take attributable proportion . That is due to their chronic clinical evolution reflecting the accumulation of worms in the body through subsequent rounds of infection [32] . The lost follow-up is quite high and not equal in different age groups with lower in the older . Those aged 45+ only took 49% in all population infected with C . sinensis in fecal examination , while it took 58% in the 259 persons included for final analysis . Therefore , the overall disability weight of 0 . 075 should not be extrapolated directly to the source population , while the age-specific disability weight will be more reasonable . Similarly , when extrapolated to other population with different age composition , age-specific disability weight should also be applied . The relationship between disability weight and infection intensity has also been explored , which fits a logarithmic equation . However , due to the lack of more points , the conclusion is not invulnerable . However , this trend deserves further study . If it occurs indeed , a mathematical model will likely become gold standard for future burden evaluation . Various methods can be adopted for calculating disability weight . In the GBD , the person trade-off technique through expert panel was introduced , which may underestimate the disability weights of neglected tropical diseases [35] . Patients-based determination of quality of life is also valuable [36] , which has been adopted in chronic schistosomiasis japonica successfully [37] . However , there is also controversy on it [38] . In a recent application for infection with STHs , no expected result was captured [39] . Thus , the operation is strict and needs high quality control . Some researchers have captured the disability weight of schistosomiasis japonica by decision-model according to the probabilities and disability weights of different morbidities [26] and the result conformed to that of patients-based quality of life assessment [37] . The new recaptured disability weight of schistosomiasis challenged existing one in GBD [40] . In this study , the similar approach was adopted . There exists one important concern how to deal with the co-sequelaes . It is reasonable to argue that different outcomes including additive effect , synergistic effect or antagonistic effect may occur due to the difference of co-sequelaes . Especially , there exists the risk that the overall disability may be over 1 when additive effect is adopted . Consequently , to avoid this problem , the antagonistic approach was adopted in ABD [31] , [41] . However , here we still adopted the additive approach . Firstly , there exists no perfect and standard way to solve this problem . Secondly , the additive approach is adopted by current practice in GBD . Thirdly , what is the most important is co-sequelaes was not serious in this study shown in Figure 2 . Multiparasitism is a pervasive problem , which makes the attribution of commonly encountered symptoms and signs of morbidity to particular parasites impossible [42]–[44] . For example , schistosoma japonicum is endemic in part of P . R . China and causes some similar sequelaes as C . sinensis infection , such as diarrhea , abdominal pain and hepatomegaly [26] . In addition , STHs infection can introduce similar problem . Therefore , during the design of this study , in order to avoid the confounders , the field without schistosomiasis was selected , which can be seen from the result of fecal examination . Furthermore , those infected with STHs were also excluded . Although we haven't excluded all other parasitic infections such as protozoa that can also lead to diarrhea and cyst of liver , it should not be serious problem due to the high economic development of the field with per capita annual net income over 10 000 RMB . Hepatitis is another important confounder , so persons with hepatitis were also excluded from final analysis . However , someone with other diseases such as diabetes , hypertension , hypotension , gastrosis and gynecological disease were still included . It is not reasonable and necessary to exclude them . Firstly , excluding them will diminish the sample size and lead to bias . Secondly , what is most important is that these diseases are not related to the ten sequelaes included for analysis . We have already excluded sequelaes such as weakness , headache , abdominal distension , anorexia and nausea , which could not be attributed to C . sinensis infection specifically . There exist several limitations in this study . Firstly , not all sequelaes attributable to C . sinensis infection were included for analysis , which will lead to underestimation of the overall disability weight . Some symptoms and signs were excluded due to their mildness and non-specificity . Because the ultrasound examination only focused on liver and gallbladder , other sequelaes caused by C . sinensis infection such as pancreatitis [45]–[48] would be omitted . Furthermore , to exclude the potential confounders , all persons with hepatitis were excluded . However , although hepatitis is not frequently caused by C . sinensis infection , the potential one or more cases still exist [18] . Secondly , the disability weight of single sequelae referred to different sources . Only diarrhea is listed among the disability weights of GBD , so we have to refer to other sources such as ABD and other literatures . However , the disability weights in ABD and other literatures are not only monotonic , but also can't be discriminated for different age groups . Consequently , we have to assign the same disability weight of 0 . 349 to gallbladder and bile duct disease including gallstone , cholecystitis , cholangitis , cholecystectomy and polyp of gallbladder , and 0 . 060 to hepatomegaly , cyst of liver and hypertrophy of gallbladder for all age groups . However , we assume this is the best evidence-based . Thirdly , to elaborate which sequelaes are indeed attributable to clonorchiasis and not due to other factors , control should be introduced . But it is too difficult to search a completely matched control in field . To overcome this weakness , several potential important confounders including schistosomiasis , STHs and hepatitis were excluded in design and analysis . Although the confounding of another important factor-alcoholism has been partially avoided through exclusion of hepatitis , it may still affect the accuracy of the study , which should be solved in further study . However , the overall disability weight captured here is still not overestimated , due to incomplete inclusion of sequelaes mentioned above and conservative priors set in model simulation . Fourthly , due to the limit of sample size , a result presented as sex- and age-specific category hasn't been offered , which should also be explored in further study . In the first WHO report on neglected tropical diseases , it is said the absence of conclusive information on the geographical distribution and burden of foodborne trematode infection including C . sinensis means their public–health impact may have been underestimated for decades [32] . Therefore , the disability weight captured here is expected to promote the further studies and benefit the final estimation of disease burden , which will promote health awareness and implementation of intervention . | Clonorchiasis is caused by eating incompletely cooked fishery product which carries the larval of Clonorchis sinensis . Millions of people are estimated to suffer in Southeast Asia . However , it is still among the most neglected tropical diseases due to the lack of clear evaluation , of which no disease burden available is one important reason . Our study is the first attempt to estimate the disability of C . sinensis infection , which reflects the average loss of life value due to some conditions and is crucial for calculating disease burden in terms of disability-adjusted life years ( DALYs ) . After obtaining the probability and disability of single sequelae caused by C . sinensis infection through community investigation and literatures reviewing respectively , the overall disability was captured through model simulation . It was showed the overall disability of the male was higher than that of the female , positive correlation occurred between disability and infection intensity , and gallstone took the major attributable proportion . Thus , C . sinensis infection can cause apparent disability . The disability captured here may promote the further studies and benefit the final estimation of disease burden , which will promote health awareness and implementation of intervention . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"neglected",
"tropical",
"diseases",
"clonorchiasis",
"parasitic",
"diseases",
"disease",
"informatics"
] | 2011 | Disability Weight of Clonorchis sinensis Infection: Captured from Community Study and Model Simulation |
Fascioliasis , caused by the liver fluke Fasciola hepatica , is a neglected tropical disease infecting over 1 million individuals annually with 17 million people at risk of infection . Like other helminths , F . hepatica employs mechanisms of immune suppression in order to evade its host immune system . In this study the N-glycosylation of F . hepatica’s tegumental coat ( FhTeg ) and its carbohydrate-dependent interactions with bone marrow derived dendritic cells ( BMDCs ) were investigated . Mass spectrometric analysis demonstrated that FhTeg N-glycans comprised mainly of oligomannose and to a lesser extent truncated and complex type glycans , including a phosphorylated subset . The interaction of FhTeg with the mannose receptor ( MR ) was investigated . Binding of FhTeg to MR-transfected CHO cells and BMDCs was blocked when pre-incubated with mannan . We further elucidated the role played by MR in the immunomodulatory mechanism of FhTeg and demonstrated that while FhTeg’s binding was significantly reduced in BMDCs generated from MR knockout mice , the absence of MR did not alter FhTeg’s ability to induce SOCS3 or suppress cytokine secretion from LPS activated BMDCs . A panel of negatively charged monosaccharides ( i . e . GlcNAc-4P , Man-6P and GalNAc-4S ) were used in an attempt to inhibit the immunoregulatory properties of phosphorylated oligosaccharides . Notably , GalNAc-4S , a known inhibitor of the Cys-domain of MR , efficiently suppressed FhTeg binding to BMDCs and inhibited the expression of suppressor of cytokine signalling ( SOCS ) 3 , a negative regulator the TLR and STAT3 pathway . We conclude that F . hepatica contains high levels of mannose residues and phosphorylated glycoproteins that are crucial in modulating its host’s immune system , however the role played by MR appears to be limited to the initial binding event suggesting that other C-type lectin receptors are involved in the immunomodulatory mechanism of FhTeg .
Infection with parasitic worms ( helminths ) modulates the host immune system by biasing T helper ( Th ) cells towards a Th-2/Treg immune response [1 , 2] while simultaneously impairing pro-inflammatory Th1/Th17 immunity . This polarisation is due to the interaction of helminth derived molecules with pattern recognition receptors on innate immune cells such as dendritic cells , macrophages and mast cells that drive the polarisation of T-cells [3] . Many helminth proteins and lipids are glycosylated , and the initial interaction of these molecules with innate immune cells is mainly via C-type lectin receptors ( CLRs ) [4 , 5 ) ] . Helminth-derived glycoconjugates , i . e . N- and O-glycoproteins and glycolipids , often contain a mixture of glycan motifs similar or identical to those present in the host , and structurally distinct pathogen-related motifs [6] . With respect to structural analysis of parasitic helminth glycans , Schistosoma mansoni has been most extensively studied [7 , 8] . Although glycan profiles are extremely complex and specific for each life-stage or preparation ( e . g . excreted vs . somatic antigens ) , common features of schistosome glycans are the presence of a high proportion of Lewis X elements , N , N’-diacetyllactosamine motifs ( LacdiNAc; LDN ) α1–3 substituted with single Fuc or unique Fucα1-2Fuc difucosyl units , and N-glycan core modifications with xylose and fucose [7 , 8 , 9 , 10 , ] . While in most cases both glycans and proteins have a role to play [11 , 12 , 13 , 14 , ] , some isolated glycan structures have been proven to directly activate host cells when presented as multivalent arrays on carrier proteins [15] . Several reports have pointed to a role for CLRs in mediating immune regulatory processes driven by helminth-derived glycoconjugates [11 , 16 , 17] . Schistosoma mansoni soluble egg antigens signal through several CLRs , dendritic cell-specific intercellular adhesion molecule-3-grabbing non-integrin ( DC-SIGN ) , Macrophage galactose lectin ( MGL ) and Mannose receptor ( MR ) , inhibiting dendritic cells maturation and influencing the development of Th2 immune response . Similarly , Trichuris suis antigens signal through MR and DC-SIGN which are involved in inhibiting a pro-inflammatory dendritic cell phenotype [17] . Recent studies demonstrated the direct interaction of MR with Fasciola hepatica and specific S . mansoni derived antigens [11 , 18] . Fasciola hepatica is a parasitic flatworm that infects humans and livestock worldwide . The economic burden of F . hepatica infection to the agricultural industry is estimated at $3 billion per year [19] while an estimated 1 million people are infected worldwide [20 , 21] . F . hepatica shares with other helminths [22 , 23 , 24] the ability to polarise Th2 immune responses within hours post infection while simultaneously impairing the ability of innate immune cells to promote Th1/Th17 immune responses [25 , 26] . We are interested in the tegumental antigens ( FhTeg ) that are released continuously by F . hepatica during infection and exposed to host immune responses . FhTeg exhibits potent Th1 immune suppressive properties in vivo by suppressing serum levels of the Th1 mediators IFNγ and IL-12p70 in the mouse model of septic shock [27] . FhTeg-activated dendritic cells and mast cells are hypo-responsive to TLR activation thereby suppressing the production of inflammatory cytokines and co-stimulatory molecules important in driving adaptive immune responses [27 , 28] . FhTeg mechanism of action is independent of TLRs and has been linked to the suppression of NF-κB and MAPK pathway [29] and enhanced expression levels of suppressor of cytokine signalling ( SOCS ) 3 , a negative regulator of the TLR and STAT3 pathway [30] . More recently it was shown that SOCS3 expression was enhanced in the liver of infected mice [31] . FhTeg is a biological matrix rich in glycoconjugates that remains continually exposed to the host immune system and is unique to each Fasciola species [32] . While proteomic analyses confirm the high abundance of glycoproteins in the tegument preparation [33 , 34] , detailed glyco-analytical data are available only for the glycolipid fractions of Fasciola spp . [35 , 36 , 37] . With a view to further understanding F . hepatica’s mechanism of immune suppression , in this study we investigated the N-glycosylation profile of FhTeg and explored the role played by oligomannose and negatively charged glycans in immunomodulatory mechanism of FhTeg .
FhTeg was prepared as previously reported [27] . In brief , F . hepatica adult worms following collection from sheep at a local abattoir were washed in sterile phosphate-buffered saline ( PBS ) and incubated in 1% Nonidet P-40 ( NP-40 [Sigma Aldrich] in PBS ) for 30 min . Supernatant was collected and NP-40 removed using BIO-RAD detergent-removing biobeads ( BIO-RAD ) , and the remaining supernatant was centrifuged at 14 , 000 × g for 30 min at 4°C prior to being filtered/concentrated using compressed air , and then stored at −20°C . All protein concentrations were determined using a bicinchoninic acid ( BCA ) protein assay kit ( Pierce , Fischer Scientific , Dublin , Ireland ) and all antigen tested for endotoxin using a using the Pyrogene endotoxin detection system ( Cambrex ) . FhTeg gave endotoxin levels similar to background levels and were less than the lower limit of detection in this assay ( <0 . 01 EU/ml ) . N-glycans were isolated by sequentially digesting FhTeg-derived glycopeptides with PNGase F and PNGase A ( Sigma-Aldrich , Ireland ) , as previously described [38] . The purified N-glycans were subjected to labelling with 2-aminobenzoic acid ( 2-AA ) using a previously described protocol [39] . 2-AA-labeled N-glycans were analyzed by matrix-assisted laser-desorption ionization ( MALDI ) time-of-flight ( TOF ) mass spectrometry ( MS ) with an Ultraflex II MALDI-TOF-MS instrument ( Bruker Daltonics; Bremen , Germany ) operating in the negative-ion reflector mode using DHB ( Bruker Daltonics ) as a matrix . Aliquots of FhTeg 2-AA-labeled-N-glycans were incubated with Jack bean α-mannosidase , neuraminidase from Vibrio cholera , β-N-acetyl glucosaminidase from Canavalia ensiformis ( all from Sigma-Aldrich , Wicklow , Ireland ) , or Jack bean β- ( 1–4 , 6 ) galactosidase ( Prozyme-Glyko , Hayward , CA , USA ) and analyzed by MALDI-TOF-MS . MALDI-TOF-MS spectra were annotated in terms of monosaccharide composition ( FxHyNz ) applying the Glyco-Peakfinder tool [40] , followed by manual interpretation in-line with the exoglycosidase treatment results , and LIFT fragmentation analysis of selected ion species , using Bruker Daltonics FlexAnalysis software ( Bruker Daltonics ) . All glycan signals were detected as [M-H]- . Fourier transform ion cyclotron resonance mass spectrometry ( FT-ICR-MS ) was performed on a Bruker 12T solariX XR high-resolution MALDI-FT-ICR-MS instrument equipped with a ParaCell ( Bruker Daltonics , Bremen , Germany ) . The system was controlled by ftmsControl software and equipped with a Smartbeam-II laser system operating at 200 Hz . The glycan sample was spotted onto a MALDI ground steel target plate ( Bruker Daltonics ) in 1 μL water and mixed on plate with 1μL super-DHB ( Sigma-Aldrich ) ( 2 . 5 mg/mL solution in acetonitrile/water , 1:1 containing 1mM NaOH ) . Biotinylated Concanavalin A lectin ( ConA ) used in this study was purchased from Vector Laboratories ( Peterborough , UK ) . It was selected as the plant lectin with highest affinity for mannose-rich carbohydrate epitopes . IRDye 800 Streptavidin was purchased from Li-COR Biosciences ( Lincoln , MA , USA ) . Precast 4–20% sodium dodecyl sulphate ( SDS ) -polyacrylamide gels ( Pierce ) were run under standard conditions ( see supplementing material ) . The gels were transferred onto nitrocellulose membranes by iBlot Dry blotting system ( Invitrogen , Carlsbad , CA , USA ) . After standard western blotting procedure , the membranes were scanned using Odyssey Infrared Imaging System ( Li-COR Biosciences ) . Data analysis was performed with Odyssey V 3 . 0 software ( Li-COR Biosciences ) . The following protocol was adapted from a previously described method [41] . Briefly , adult liver flukes were flat-fixed in 4% paraformaldehyde and incubated with fluorescein-labelled ConA ( Vector Laboratories ) . Parasites were mounted on glass microscope slides with Vectashield anti-fading solution ( Vector Laboratories ) . Specimens were viewed using a Leica DM IL LED microscope using 10x , 20x , and 40x HI PLAN I objectives ( Leica Microsystems , Wetzlar , Germany ) equipped with epifluoresce source and a filter system for FITC fluorescence . Images were processed with Adobe Photoshop CS4 software ( Adobe System Inc . , San Jose , USA ) . BALB/C mice , 6–8weeks old ( Charles River , Carrentrilla , Ireland ) , were kept under specific pathogen-free conditions at the Bioresource Unit , Faculty of Health and Science , Dublin City University ( DCU ) , Ireland . All mice were maintained according by European Directive 2010/63/EU . Ethical permission for the use of animals and experimental protocols were approved by the Health Products Regulatory Authority , Ireland ( licence number B100/2833 ) and Dublin City University Ethics committee ( reference number: DCUREC/2010/033 ) . We adhered to all agreed protocols . Bone marrow-derived immature DCs were isolated from the femurs and tibia of BALB/C mice according to Lutz et al . protocol [42] , yielding >95% pure BMDC population ( on the basis of CD11c expression ) . The same protocol was also performed on C57/BL mice or from MR knockout mice on a C57/BL background ( a gift from Professor Padraic Fallon , Trinity College Dublin ) . CHO cell line stably expressing murine MR [43] and untransfected controls were maintained in RPMI 1640 Glutamax ( Gibco , Life Technologies , Bleiswijk , Netherlands ) containing 10% foetal calf serum ( FCS; Bodinco B . V . , Alkmaar , Netherlands ) . All media were supplemented with penicillin ( Astellas Pharma B . V . ) and streptomycin ( Sigma-Aldrich ) , and transfected cell lines were continuously kept under selection of 0 . 6 mg ml−1 geneticin ( Gibco ) . Cellular adhesion assay was performed as previously reported [43] . Briefly , FhTeg and BSA were fluorescently labelled with PF-647 or PF-488 using the Promofluor labelling kits according to the manufacturer’s recommendations ( Promokine , Heidelberg , Germany ) . Where indicated cells were pre-incubated with EGTA ( 10mM , Sigma-Aldrich ) , anti-MR ( 1 μg ml-1 , clone: 15–2 , Abcam , Cambridge , UK ) , mannan ( 0 . 1–1 mg ml−1 , Sigma-Aldrich ) , GalNAc-4S ( 1-25mM , Sigma-Aldrich ) , for 45 min at 37°C prior to addition of fluorescently labelled FhTeg in the stated concentrations at four degrees . As control for non-specific binding , cells were incubated with fluorescently labelled BSA at four degrees . After extensive washes , binding was analysed by flow cytometry ( BD FACSAria or FACSCanto , BD Biosciences ) , using FacsDiva ( BD Biosciences ) and FlowJo Software ( TreeStar , Ashland , OR , USA ) . For microscopy studies , after the final incubation , cells were washed with PBS and fixed in 4% paraformaldehyde . After extensive rinsing , cells were resuspended in Vectashield anti-fading solution with DAPI ( Vector Laboratories ) , mounted on slides and viewed using a Leica DM IL LED microscope as described above . BMDCs were stimulated with or without GalNAc-4S ( 1mM ) , Mannan ( 50 μg ml-1 ) or Man-6P ( 50 μg/ml ) , 30 min prior to stimulation with FhTeg ( 10 μg ml-1 ) . Cells were harvested after 2 . 5 h and washed before RNA extraction with high pure RNA isolation kit ( Roche Diagnostics , Burgess Hill , UK ) according to the manufacturer’s instructions . cDNA was synthesized using a reverse transcriptase kit ( Roche ) according to the manufacturer’s protocol . The quantitative PCR ( qPCR ) transcription analysis was carried out on a real-time thermal cycler LightCycler96 ( Roche ) , using Real Time Ready primers ( Roche ) specific for SOCS3 and three internal reference genes: β-actin ( NM_007393 . 3 ) , GAPDH ( NM_008084 . 2 ) , and Gusb ( β-glucuronidase , NM_010368 . 1 ) . The Pfaffl method was used to calculate fold changes compared to the three reference genes . BMDCs were stimulated with FhTeg ( 10 μg/ml ) for 2 . 5 h prior to stimulation with LPS ( Escherichia coli R515; 100 ng ml-1; Enzo Life sciences , Farmingdale , USA ) , for 18 h . Alternatively , cells were preincubated with GalNAc-4S ( 10 mM ) , Mannan ( 500 μg ml-1 ) or Man-6P ( 50 μg/ml ) for 30 min prior to stimulation with F . hepatica antigens . Control cells were treated with medium , FhTeg antigens ( also in combination with carbohydrate inhibitors ) or LPS alone . Supernatants from cultured DCs were tested for the production of IL-12p70 , and TNF ( BD OptEIA ELISA sets; BD Biosciences ) . All data were analysed for normality prior to statistical testing by Prism 6 . 0 ( GraphPad Software Inc , La Jolla , CA , USA ) software . Where multiple group comparisons were made , data were analysed using one-way ANOVA . For comparisons between two groups , the Student’s t test was used . In all tests , p < 0 . 05 was deemed significant .
This study is the first to present the MS analysis of FhTeg derived N-glycans . The spectrum of PNGase-F released glycans contains major signals indicative of oligomannose type structures ( Man4-9GlcNAc2 , m/z 1192 . 6 , 1354 . 7 , 1516 . 8 , 1678 . 8 , 1840 . 9 , 2003 . 0 [M-H]- , respectively ) ( Fig 1 ) , trimannosyl glycans with and without α1–6 core fucose ( Fuc ) ( m/z 1176 . 6 , 1030 . 5 , respectively ) , and complex type glycans with single or truncated antenna ( m/z 1233 . 6 , 1395 . 7 , 1436 . 7 ) and their core α1-6-fucosylated variants ( m/z 1379 . 7 , 1541 . 8 , 1582 . 8 ) . In addition , relatively abundant signals were observed at m/z 1313 . 6 and 1459 . 7 , which indicate the occurrence of a sulphate/phosphate modified form of the truncated monoantennary glycans . These assignments , based on monosaccharide composition , were fully supported by exoglycosidase digestions ( S1A–S1C Fig ) , and LIFT MS/MS fragmentation analysis of selected ions m/z 1313 . 8 , 1459 . 9 , m/z 1541 . 9 , 1582 . 9 ( Fig 2 ) . In each glycan the number of antenna was confirmed by removing accessible mannoses with α-mannosidase treatment ( S1A Fig ) . Similarly , the presence of the terminal , non-substituted Gal and/or GlcNAc residues was confirmed by β-hexosaminidase and β-galactosidase treatments ( S1B and S1C Fig ) , and these incubations also confirmed location of Fuc ( if present ) exclusively on the core GlcNAc . Notably , the sulphated/phosphorylated ions m/z 1313 . 8 and 1459 . 9 were retained after β-hexosaminidase digestion ( S1B Fig ) . This suggests that terminal GlcNAc residues are substituted with the anionic group , with core fucosylation on peak m/z 1459 . 9 . The MS/MS fragmentation however indicates that the S/P group can be present on both the terminal GlcNAc as well as the terminal Man residue ( Fig 2 ) . The latter observation is confirmed by the α-mannosidase treatment , which does not remove the substituted mannose ( m/z 1459 . 6 in S1A Fig ) . The spectrum of PNGase A released glycans didn’t reveal any additional peaks if compared to PNGase F digestion ( S2 Fig ) , thus suggesting the absence of core α1-3-fucosylated glycans . Minor peaks were also annotated and depicted in supplementing material , S1 Table . Notably , minor signals matching with N-acetylneuraminic acid ( NeuAc ) or N-glycolylneuraminic acid ( NeuGc ) containing N-glycans were detected ( S1 Table ) . In particular , two signals at m/z 2052 . 05 and 2068 . 05 correspond to sialylated diantennary glycans bearing one NeuAc and one NeuGc , respectively; while signals at m/z 2359 . 2 and 2375 . 2 correspond to disialylalted glycans . The presence of these capping sugars was validated by neuraminidase digestion for the signals at m/z 2052 . 050 and 2068 . 0 ( S1D Fig ) . The presence of sialylated glycoproteins in F . hepatica has been reported in scattered studies using different experimental techniques [37] . However , since other helminths lack the biosynthetic machinery required for sialylation , the sialic acid-containing glycans in the liver fluke might be derived from host glycoproteins ( such as sialylated antibodies ) , as observed for other parasites [38] . Since high mannose glycans were abundantly present in the N-glycan profile of FhTeg , we incubated flat fixed adult flukes with fluorescein-conjugated ConA , a plant lectin with high affinity for oligomannose type glycans that bind to the carbohydrate-recognition domains on MR ( CRD ) in order to confirm the presence of oligomannose N-glycans on the surface of F . hepatica coat . ConA bound to the external morphological features of the adult fluke such as the oral and ventral suckers , tegumental spines and the underlaying tegumental coat ( Fig 3A–3D ) . The abundant expression of glycan motifs on the tegumental spines enabled the visualisation of spinelets , features usually observed only by scanning electron microscopy [44] . In the lectin blot ( Fig 3E ) , ConA recognised a wide array of FhTeg glycoproteins ( >15 bands ) with electrophoretic mobility ranging between 260 and 17 KDa , with the most intense protein bands at approximately 18 and 40 KDa . To study the potential interaction between FhTeg and MR , the binding of FhTeg by MR-transfected CHO cells was investigated in presence and absence of EGTA and two carbohydrate MR-binding inhibitors , mannan and GalNAc-4S that interact with the CRD ( carbohydrate-recognition domains ) and CR-MRdomain ( N-terminal cysteine-rich domain ) on MR , respectively [45] . As measured by flow cytometry , FhTeg significantly bound to CHO cells expressing MR compared to untransfected s CHO cells ( S2A Fig ) . This binding was reversed in the presence of EGTA and a combination of mannan and GalNAc-4S ( Fig 4A ) , suggesting the potential involvement of both carbohydrate binding domains in FhTeg-MR interaction . Given the apparent abundance of mannose ( for CRDs ) and the potential presence of sulphated GlcNAc ( for CR-MR ) , on the surface of adult flukes , we also examined the interaction of the FhTeg with BMDCs . Both flow cytometry and microscopy analyses ( Fig 4B and 4C ) revealed that FhTeg strongly adheres to BMDCs compared to fluorescently labelled BSA which was used as a negative control ( Fig 4D ) . Flow cytometric analysis revealed that this interaction is mediated by calcium as the binding was significantly reduced upon pre-incubation with EGTA ( p≤ 0 . 001 ) . The involvement of MR was assessed by pre-incubation of cells with anti-MR blocking antibody or the carbohydrate inhibitors ( mannan and GalNAc-4S ) prior to stimulation with FhTeg . Pre-incubation of BMDCs with all inhibitors significantly reversed the binding ( Fig 4B ) , implying the involvement of CLRs . This indicates that MR could contribute to FhTeg binding to BMDCs , however it is acknowledged that a variety of other CLRs ( e . g . Dectin-1 , DC-SIGN , MGL etc ) may also be involved . In previous studies we demonstrated that FhTeg up-regulates SOCS3 in DCs and mast cells in vitro [29] . SOCS3 [29] is an intracellular protein regulating the duration or intensity of cytokine-induced signal via a negative feedback inhibition mechanism [46] . SOCS proteins are induced via JAK/STAT signaling and , among other signals , stimulation of TLRs [47] . Here we determine if the induction of SOCS3 RNA can be inhibited by the addition of mannan to DCs in culture prior to stimulation with FhTeg ( 10 μg ml-1 ) . Notably , incubation with mannan did not alter SOCS3 expression ( Fig 5A ) . Another property of FhTeg is its ability to suppress the LPS-induced production of pro-inflammatory cytokines in BMDCs [27] . Pre-incubation of BMDCs with mannan prior to LPS and FhTeg stimulation reversed IL12p70 secretion ( Fig 5B ) . To clarify the role of MR in FhTeg-BMDCs interactions , we obtained BMDCs from MR knockout mice . The absence of MR resulted in a statistically significant decrease in the binding of FhTeg to the surface of BMDCs ( Fig 5C and 5D ) . However in the absence of MR , FhTeg retained the ability to induce the expression of SOCS3 ( Fig 5E ) , and to suppress LPS induced IL12p70 expression in BMDCS ( Fig 5F ) . In order to evaluate if the phosphorylated oligosaccharides contribute to the immunological properties of FhTeg , a panel of monosaccharide inhibitors , Man-6P , GlcNAc-4P and GalNAc-4S where investigated . Despite the absence of sulphated glycans on FhTeg , GalNAc-4S , a known inhibitor of the Cys-MR domain of the mannose receptor , was included in the panel as anionic control . Preincubation of BMDCs with Man-6P failed to suppress FhTeg immunological properties ( Fig 7A and 7B ) . Notably , GlcNAc-4P interfered with culture media pH and expression level of housekeeping genes in CHO-MMR and BMDCs , respectively , resulting in unattendable results ( S3 Fig ) . Interestingly , incubation with 1mM GalNAc-4S for 30 min prior to FhTeg stimulation failed to inhibit FhTeg ability to suppress the LPS-induced production of pro-inflammatory cytokines in BMDCs ( Fig 7C ) but reversed SOCS3 transcription to basal levels ( p = 0 . 019 ) , ( Fig 7D ) . However the absence of sulphated glycans on FhTeg and the limited role played by MR in the immunomodulatory activity of the tegumental antigen suggest the involvement of a different CLR targeted by negatively charged glycoconjugates and inhibited by GalNAc-4S .
Host-pathogen interactions , immune cell development and function are mediated by glycans , glycolipidsand glycoproteins [48 , 49] and in more recent years glycomics approaches have been used to facilitate the unravelling of these processes . In this study N-glycans were isolated and analysed by mass spectrometry for the first time from the tegumental antigen ( FhTeg ) of adult liver flukes to reveal that FhTeg-derived N-glycans are composed predominantly of oligomannose type and truncated structures . Mannosylated glycans are widely distributed on the tegument of adult flukes with remarkable abundance especially on the spines and suckers and are present on a wide panel of glycoprotein components of the FhTeg preparation . As these represent the fluke’s morphological features in most direct contact with the host , a role in cell adhesion and/or signalling for these glycoproteins is hypothesized . This study support previous findings by Allister et al . ( 2011 ) who using a panel of lectins examined the glycoprofile of F . hepatica gastrodermis that also exhibited an abundance of mannosylated glycoproteins [50] . Glycosylated antigens from F . hepatica excetory-secretory products and from other helminths ( S . mansoni , T . suis , T . crassiseps , T . spiralis ) have been shown to take part in immunomodulatory processes of parasitic infections via the mannose receptor ( MR ) [11 , 18 , 51 , 52 , 53] . MR is a type-I membrane protein with a cytoplasmic domain involved in antigen processing and receptor internalisation and three different types of binding domains at its extracellular region . In particular , MR features multiple C-type lectin-like carbohydrate-recognition domains ( CRDs ) responsible for Ca2+-dependent binding to terminal mannose , fucose or N-acetylglucosamine [54] , a fibronectin type II ( FNII ) domain involved in collagen binding [55] and an N-terminal cysteine-rich ( CR-MR ) domain that mediates Ca2+-independent binding to sulphated sugars such as SO4-3-Gal or SO4-3/4-GalNAc [56] . In the case of helminth antigens , binding to MR has been always attributed to the interaction of antigen-derived glycans with the multiple CRDs that mediate MR binding to saccharides with terminal mannose , fucose or N-acetylglucosamine residues . Here we show that blocking MR carbohydrate binding domains with specific glycan inhibitors ( i . e . mannan for CRDs and GalNAc-4S for CR-MR ) reversed the binding of FhTeg to BMDCs and CHO+MMR cells and a reduced binding was also observed in BMDCS isolated from MR knockout mice . However , it appears that MR plays a role in the binding of FhTeg to DCs but not in the signalling pathways , as FhTeg immune modulatory properties weren’t inhibited by preincubation with mannan in vitro and were retained in the absence of MR . These evidences suggest that other CLRs may play key roles in FhTeg immune modulation . Recently , the interaction of F . hepatica excretory-secretory products with macrophages demonstrated that MR and Dectin 1 blocking antibodies could inhibit the induction of M2 macrophages [18] . However , it has also been shown that TLR2 plays a role in the FhES inhibition of macrophages during Mycobacterium bovis activation [57] . S . mansoni soluble egg antigens have been shown to signal through several CLRs; DC-SIGN , MGL and MR to inhibit DC maturation and induce a Th2 immune response while the nematode T . suis antigens are involved in CLR signalling by inducing a DC phenotype which inhibits bacterial TLR activation and the activation of an inflammatory immune response [17] . Dectin 2 has also been implicated in the induction of IL-1β production from DCs following S . mansoni SEA stimulation [58] . The most extensively studied immunomodulatory glycans are conjugated and non-conjugated variants of the human milk sugar LNFPIII and its functional trisaccharide motif LeX that is expressed among many other glycans on various schistosome antigens [59] . In different forms LeX and/or LNFPIII induce M2-like macrophage [60] , promote Th2 immune responses [61] and attenuate a range of inflammatory disorders including psoriasis [62] and transplant rejection [63] . While LeX binds to SIGNRI , this receptor does not have a role in its uptake by macrophages and it mechanism of action is TLR4 dependent [64] . LeX is not found in F . hepatica but in general helminth antigens are composed of a complex mix of glycans , proteins and lipids , and it is reasonable to assume that multiple TLR and CTRs are involved in its immune properties with a number of redundancy pathways . The unexpectedly limited role played by the mannose receptor in FhTeg mechanism of immunomodulation prompted a further mass spectrometric investigation to elucidate the nature of other terminal sugars . The most notable aspect of the MS spectrum of FhTeg-derived N-glycans was the discovery of structures modified with anionic groups on terminal residues . Negatively charged glycans on the glycocalyx of adult liver flukes have been previously observed by electron microscopy when staining acid carbohydrates with cationic dyes such as colloidal iron , ferric chloride or ruthenium red [31] . In his pioneering study on F . hepatica tegument , Threadgold could not investigate the nature of the anionic groups due to the lack of adequate experimental tools . FT-ICR-MS was conducted on FhTeg-derived N-glycans and unequivocally ascertained the presence of phosphate and not sulphate groups on terminal Man or GlcNAc monosaccharides . Interestingly , an acidic glycolipid fraction containing the highly antigenic phosphodiester , GlcNAc ( α1-HPO3-6 ) Gal ( 1–1 ) ceramide , has been previously reported in F . hepatica [35] . In this study , MS/MS fragmentations suggest that , differently from the glycolipid , in FhTeg the phosphate group is bound terminally as monoester to a single monosaccharide ( Man or GlcNAc ) . Notably , phosphorylation of N-glycans is predominantly found on high-mannose type glycans in the form of Man-6P monoester or as phosphodiester capped by α-GlcNAc ( i . e . GlcNAc-P-Man ) . These phosphorylated glycan epitopes are responsible for the targeted trafficking of hydrolases to lysosomes through recognition by P-type lectins , CI-MPR and CD-MPR [65] . The presence of parasitic lysosomal hydrolases in FhTeg antigen preparation cannot be excluded , but it has to be noted that Man-6-phosphorylation has also been reported on non-lysosomal proteins [66 , 67 , 68 , 69] . Interestingly , Man-6-phosphorylation is observed on complex-type N-glycans of envelope glycoproteins of varicella zoster virus involved in viral entry [70 , 71 , 72] . Scattered reports have described the presence of phosphorylation on monosaccharides other than mannose , however no clear understanding of its role has emerged thus far [73 , 74 , 75] . In particular , in the case of the synapse-specific clathrin assembly protein AP180 , the presence of two sites of O-linked glycosyl phosphorylation ( GlcNAc-6P ) contributes to an increased protein net negative charge and hydrophilicity , suggesting a potential inhibitory effect in synaptic vesicle endocytosis processes [75] . In this study , the nature and the potential biological role played by these negatively charged glycoconjugates was further exploredsince oligosaccharide phosphorylation hasn’t been observed in any other helminth for which glycomics investigations have been carried out thus far , . Notably , inhibition assays revealed that some key immune modulatory properties exhibited by FhTeg are independent from Man-6P , thus excluding a potential role for Mannose 6-Phosphate Receptor in Fasciola infection mechanism . FhTeg-induced expression of SOCS3 was reversed upon pre-incubation of DCs with GalNAc-4S , however this did not reverse the immune modulatory properties of FhTeg suggesting that SOCS3 despite been induced during infection may not be important to its immune modulatory properties . However further studies are required to confirm this while other negative regulators of cytokines signaling such as SOCS2 have yet to be examined . GalNAc-4S was originally included in this study merely as a control to explore the anion specificity of the inhibition mechanism and the effect upon FhTeg activated DCs was not expected as sulphated anionic glycans were not identified in FhTeg and the role of MR was excluded in these studies . A possible explanation is that many CLRs can interact with more than one type of glycan structure and the addition of GalNAc-4S to our assays may give insight into the type of receptor but not the type of sugar important in this process . However , to date evidence supports the binding of GalNAc-4S to the CD of MR suggesting that this glycan may bind to an alternative receptor . While our data points towards a possible new pathway linking SOCS3 expression and CLR stimulation we have yet to identify the exact receptor . This is supported by work by Rolls et al . ( 2006 ) that show increased SOCS3 expression and decreased secretion of IFN-γ and TNF in T cells stimulated with chondroitin sulfate proteoglycan ( CSPG-DS ) in vitro [76] . No cellular receptor has been identified yet for this antigen . SOCS3 was also found to be induced in immature DCs as result of DC-SIGN stimulation by HIV-1 envelope glycoprotein gp120 thorough a complex pathway involving other factors such as Ras , Raf , and NF-κB-induced IL-10 secretion [77] . This supports our observation that SOCS3 induction through CLRs stimulation by glycoconjugates could be a driving mechanistic step in this process [78 , 79] but the exact receptor needs to be identified . In summary , while this study suggests a role for CLRs in FhTegs interaction with BMDCs , MR is not directly involved in the immune modulatory properties measured in the parameter of this study . While we have not identified the receptor involved in the induction of SOCS3 or in the suppression of pro-inflammatory cytokine , we believe that given the spectrum of N-glycans identified in FhTeg a variety of pathways may be activated simultaneously in a glycan-dependant manner by the heterogeneous mixture of glycoproteins that compose FhTeg . Further studies and a simplification of the tegumental antigen are required to fully dissect the effect of individual components . The discovery of phosphorylated N-glycans in FhTeg flukes highlight a potential a novel role for these molecules during helminth infection . | Fascioliasis , caused by the liver fluke Fasciola hepatica , is a neglected tropical disease infecting over 1 million individuals annually with 17 million people at risk of infection . These worms infect the liver and can survive for many years in its animal or human host because they supress the host’s immune system that is important in clearing worm infection . Worms are similar to humans in that they are made of proteins , fats and sugars , and while there are many studies on worm proteins , few studies have examined the sugars . We are interested in the sugars because we believe that they help the parasite survive for many years within its host . To examine this , we have used a technique called mass spectrometric analysis to characterise the sugars present in F . hepatica . We also have developed systems in the laboratory to test if these sugars can suppress the host’s immune system . We conclude that F . hepatica sugars are crucial in suppressing its host’s immune system; however , the exact way the sugars can do this requires further studies . These studies are important for the development of worm vaccines or therapies . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"cell",
"binding",
"phosphorylation",
"cell",
"physiology",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"chemical",
"compounds",
"pathology",
"and",
"laboratory",
"medicine",
"helminths",
"rna",
"extraction",
"carbohydrates",
"organic",
"compounds",
"animals... | 2016 | Fasciola hepatica Surface Coat Glycoproteins Contain Mannosylated and Phosphorylated N-glycans and Exhibit Immune Modulatory Properties Independent of the Mannose Receptor |
Neurosteroids are endogenous modulators of neuronal excitability and nervous system development and are being developed as anesthetic agents and treatments for psychiatric diseases . While gamma amino-butyric acid Type A ( GABAA ) receptors are the primary molecular targets of neurosteroid action , the structural details of neurosteroid binding to these proteins remain ill defined . We synthesized neurosteroid analogue photolabeling reagents in which the photolabeling groups were placed at three positions around the neurosteroid ring structure , enabling identification of binding sites and mapping of neurosteroid orientation within these sites . Using middle-down mass spectrometry ( MS ) , we identified three clusters of photolabeled residues representing three distinct neurosteroid binding sites in the human α1β3 GABAA receptor . Novel intrasubunit binding sites were identified within the transmembrane helical bundles of both the α1 ( labeled residues α1-N408 , Y415 ) and β3 ( labeled residue β3-Y442 ) subunits , adjacent to the extracellular domains ( ECDs ) . An intersubunit site ( labeled residues β3-L294 and G308 ) in the interface between the β3 ( + ) and α1 ( − ) subunits of the GABAA receptor pentamer was also identified . Computational docking studies of neurosteroid to the three sites predicted critical residues contributing to neurosteroid interaction with the GABAA receptors . Electrophysiological studies of receptors with mutations based on these predictions ( α1-V227W , N408A/Y411F , and Q242L ) indicate that both the α1 intrasubunit and β3-α1 intersubunit sites are critical for neurosteroid action .
Neurosteroids are cholesterol metabolites produced by neurons [1] and glia [2] in the central nervous system ( CNS ) that are thought to play important roles in both nervous system development and behavioral modulation [3] . Neurosteroid analogues are also being developed as sedative hypnotics [4] , antidepressants [5] , and anticonvulsants [6] . Gamma amino-butyric acid Type A ( GABAA ) receptors , the principal ionotropic inhibitory neurotransmitter receptors in the CNS , have been identified as the primary functional target of neurosteroids . The major endogenous neurosteroids—allopregnanolone and tetrahydroxy-desoxycorticosterone ( THDOC ) —are positive allosteric modulators ( PAMs ) of GABAA receptors , potentiating the effects of GABA at nanomolar concentrations and directly activating currents at micromolar concentrations . GABAA receptors are members of the pentameric ligand-gated ion channel ( pLGIC ) superfamily and are typically composed of two α subunits , two β subunits , and one γ or δ subunit [7] . There are 19 homologous GABAA receptor subunits ( including six α , three β and two γ isoforms ) , with each subunit composed of a large extracellular domain ( ECD ) , a transmembrane domain ( TMD ) formed by four membrane-spanning helices ( TMD1–4 ) , a long intracellular loop between TMD3 and TMD4 , and a short extracellular C-terminus . These distinctive structural domains form binding sites for a number of ligands: GABA and benzodiazepines bind to the ECD , picrotoxin to the channel pore [8] , and general anesthetics—such as propofol [9 , 10] , etomidate [11] , barbiturates [12] , and neurosteroids—to the TMDs [13–18] . Substantial evidence indicates that neurosteroids produce their effects on GABAA receptors by binding to sites within the TMDs [13–15 , 19 , 20] . Whereas the TMDs of β-subunits are critically important to the actions of propofol and etomidate [11 , 21–26] , the α-subunit TMDs appear to be essential for neurosteroid action . Mutagenesis studies in α1β2γ2 GABAA receptors identified several residues in the α1 subunit , notably Q241 in TMD1 , as critical to neurosteroid potentiation of GABA-elicited currents [14 , 27] . More recent crystallographic studies have shown that , in homo-pentameric chimeric receptors in which the TMDs are derived from either α1 [16] or α5 subunits [17] , the neurosteroids THDOC and pregnanolone bind in a cleft between the α-subunits , with the C3-hydroxyl substituent of the steroids interacting directly with α1Q241 . Neurosteroids are PAMs of these chimeric receptors , and α1Q241L and α1Q241W mutations eliminate this modulation . These studies posit a single critical binding site for neurosteroids that is conserved across the six α-subunit isoforms [14 , 27] . A significant body of evidence also suggests that neurosteroid modulation of GABAA receptors may be mediated by multiple sites . Site-directed mutagenesis identified multiple residues that affect neurosteroid action on GABAA receptors , suggestive of two neurosteroid binding sites , with one site mediating potentiation of GABA responses and the other mediating direct activation [14 , 27] . Single channel electrophysiological studies as well as studies examining neurosteroid modulation of [35S]t-butylbicyclophosphorothionate ( TBPS ) binding , have also identified multiple distinct effects of neurosteroids with various structural analogues producing some or all of these effects , consistent with multiple neurosteroid binding sites [28–30] . Finally , neurosteroid photolabeling studies in the bacterial pLGIC , Gloeobacter ligand-gated ion channel ( GLIC ) , demonstrate two neurosteroid binding sites per monomer [31] , one analogous to the canonical intersubunit site and one located in an intrasubunit pocket previously shown to bind propofol [32 , 33] and the inhalational anesthetics [33 , 34] . Both of these sites contribute to neurosteroid modulation of GLIC currents , suggesting the possibility of analogous sites in GABAA receptors . We have developed a suite of neurosteroid analogue photolabeling reagents with photolabeling groups positioned around the neurosteroid ring structure to identify all of the neurosteroid binding sites on GABAA receptors and determine the orientation of neurosteroid binding within each site . Photolabeling was performed in membranes from a mammalian cell line that stably expresses α1His-FLAGβ3 receptors ( rather than in detergent-solubilized receptors ) to optimize the likelihood that the receptors were in native conformations and environment . Finally , we deployed a middle-down mass spectrometry ( MS ) approach , coupled with a stable-heavy isotope encoded click chemistry tag for neurosteroid-peptide adduct identification to circumvent challenges associated with MS identification ( predominantly neutral loss ) and quantification of neurosteroid-peptide adducts [35] . Using these approaches , we have identified three clusters of neurosteroid-photolabeled residues on the human α1β3 GABAA receptor . Computational docking studies , guided by the photolabeling data , were used to describe three binding sites and the orientation of the neurosteroids within each site . The docking studies were also used to predict critical residues to test the contribution of each of these sites to neurosteroid modulation of GABAA currents . Site-directed mutagenesis of these sites and electrophysiological studies indicate that at least two of three structurally distinct sites contribute to allosteric modulation of GABA currents .
Allopregnanolone ( 3α-hydroxy-5α-pregnan-20-one ) is a potent , endogenous PAM of GABAA receptors ( Fig 1A ) . We synthesized three photolabeling analogues of allopregnanolone in which photolabeling moieties were placed at various positions around the steroid backbone . KK123 has a 6-diazirine photolabeling group on the C5-C6-C7 edge of the sterol , which is a likely binding interface with α-helices [36] and minimally perturbs neurosteroid structure [37] . KK123 is , however , an aliphatic diazirine and , as such , may preferentially label nucleophilic amino acids [38] . The two other reagents , KK202 and KK200 , incorporate a trifluoromethylphenyl-diazirine ( TPD ) group at either the 3- or 17-carbon . These were designed to sample the space in the plane of the steroid off either the A-ring ( KK202 ) or the D-ring ( KK200 ) . Following UV irradiation , TPD groups generate a carbene which can insert into any bond [39 , 40] . Thus , while the TPD groups are bulky and removed several angstroms from the neurosteroid pharmacophore , they should form an adduct precisely at their binding site in the GABAA receptor . Where feasible ( KK123 , KK202 ) , an alkyne was incorporated in the photolabeling reagents to allow attachment of a fluorophore , purification tag , or an MS reporter tag ( FLI-tag ) via click chemistry [35] . A useful photoaffinity labeling reagent must bind to the same site on a protein as the ligand it mimics and should produce the same effects on protein functions . To determine whether our photoaffinity labeling reagents mimic allopregnanolone as modulators of GABAA receptor function , we assessed modulation of α1β3 GABAA receptors currents in Xenopus laevis oocytes , and enhancement of [3H]muscimol binding in human embryonic kidney ( HEK ) cell membranes expressing α1β3 GABAA receptors . KK123 enhanced GABA-elicited ( 0 . 3 μM ) currents 4 . 2 ± 3 . 3-fold at 1 μM ( n = 5 cells ) and 8 . 2 ± 6 . 7-fold at 10 μM ( n = 7 ) . KK123 ( 10 μM ) also directly activated α1β3 GABAA receptors , eliciting 6 . 3% ± 3 . 8% ( n = 5 ) of the maximum current elicited by a saturating concentration of GABA . KK123 potentiation of GABA-elicited currents and direct activation were absent in α1Q242Lβ3 GABAA receptors , indicating that KK123 closely mimics the actions of allopregnanolone ( the human α1Q242L mutation is equivalent to rat α1Q241L and is known to selectively prevent neurosteroid action ( Fig 1B and Table 1 ) [14 , 27] ) . KK200 and KK202 also potentiated GABA-elicited currents at 1 and 10 μM and directly activated the channels at 10 μM ( Table 1 ) . Positive allosteric modulation by KK202 was somewhat surprising , given that an ether-linked TPD group replaces the 3α-OH group thought to be critical for neurosteroid action [41 , 42] . While the effects of KK200 were abolished in α1Q242Lβ3 receptors , the potentiation by KK202 was reduced by 50% in α1Q242Lβ3 receptors , suggesting that KK202 may have actions at both the canonical neurosteroid site and other binding sites . Because photolabeling experiments were performed in membranes prepared from cells expressing α1β3 GABAA receptors , we also examined the ability of the photolabeling reagents to enhance [3H]muscimol binding in these membranes . A stable HEK-293 cell line was established with tetracycline-inducible expression of human α1His-FLAG β3 GABAA receptors ( See Materials and methods ) ; receptor density in these membranes was 20–30 pmol [3H]muscimol binding/mg membrane protein . Consistent with previous determinations [43] , the average stoichiometry of the receptors was estimated at two α1 subunits and three β3 subunits using MS label-free quantitation [44] ( spectral count ) . Allopregnanolone enhanced [3H]muscimol binding to these recombinant receptors 4-fold with a half maximal effective concentration ( EC50 ) of 3 . 9 ± 5 . 6 μM ( S1 Fig ) . KK123 , KK200 , and KK202 all enhanced [3H]muscimol binding with EC50 values similar to or lower than allopregnanolone ( S1 Fig ) . Collectively , the electrophysiology and radioligand binding data indicate that KK123 , KK200 , and KK202 are functional mimetics of allopregnanolone . To determine whether KK123—which contains an aliphatic diazirine—photolabels GABAA receptors , we utilized the butynyloxy ( alkyne ) moiety on KK123 to attach a biotin purification tag for selective enrichment of photolabeled GABAA receptor subunits . HEK-293 cell membranes containing α1β3 GABAA receptors were photolabeled with 15 μM KK123 , solubilized in SDS , and coupled via Cu2+-catalyzed cycloaddition to MQ112 ( S2A Fig ) , a trifunctional linker containing an azide group for cycloaddition , biotin for biotin-streptavidin affinity purification , and a cleavable azobenzene group for elution of photolabeled proteins . The photolabeled-MQ112-tagged receptors were bound to streptavidin beads and eluted by cleavage of the linker with sodium dithionite . The purified , photolabeled GABAA receptor subunits were assayed by western blot using anti-α1 and anti-β3 . A band at 52 kDa was observed with both α1 and β3 subunit antibodies in the KK123 photolabeling group ( S2B Fig ) , indicating that both α1 and β3 subunits are photolabeled by KK123 . In control samples photolabeled with ZCM42—an allopregnanolone photolabeling analogue containing a diazirine at the 6-carbon but no alkyne ( S2C Fig ) —neither α1 nor β3 subunits were purified . These data indicated that KK123 can photolabel both α1 and β3 subunits and is thus an appropriate reagent to use for site identification . A 35 kDa band was intermittently observed in replicate anti-α1 western blots ( S2B Fig ) ; this is likely to be a proteolytic fragment of the α1-subunit that retains the antibody-recognition epitope but was not further analyzed . Identification of sterol adducts in hydrophobic peptides has been impeded by multiple challenges , including peptide insolubility during sample digestion , ineffective chromatographic separation of hydrophobic TMD peptides , and neutral loss of sterol adducts from small hydrophobic peptides during ionization and fragmentation . To circumvent these problems , we employed middle-down MS to analyze GABAA receptor TMD peptides and their sterol adducts . This approach identifies each TMD as a single , large peptide and attenuates neutral loss of adduct , facilitating identification of the sites of neurosteroid incorporation . In our studies , α1His-FLAGβ3 GABAA receptors were photolabeled in native HEK cell membranes . The photolabeled proteins were then solubilized in n-dodecyl-β-D-maltoside ( DDM ) -containing lysis buffer . The pentameric GABAA receptors were purified using anti-FLAG agarose beads , and eluted receptors were digested with trypsin in the presence of the MS-compatible detergent DDM . These conditions generated peptides containing each of the GABAA receptor TMDs in their entirety . The peptides were separated using PLRP-S nano-liquid chromatography and analyzed on a Thermo ELITE orbitrap mass spectrometer . This workflow ( S3 Fig ) minimized protein/peptide aggregation , simplified MS1-level identification of TMD-sterol adducts , and optimized fragmentation of TMD peptides and their adducts . All eight of the TMD peptides were reliably sequenced with 100% residue-level coverage . In addition , the covalent addition of neurosteroid to the TMD peptides increased the hydrophobicity of TMD peptides and shifted their chromatographic elution to later retention times ( S3 Fig ) . The delayed retention time was used as a critical criterion for identification of photolabeled peptides . Two photolabeled peptides were found in the mass spectra of tryptic digests of α1β3 GABAA receptors photolabeled with KK123 ( Fig 2A and S4 Fig ) . A KK123 adduct of the α1-TM4 peptide , 398IAFPLLFGIFNLVYWATYKK123LNREPQLK423 ( m/z = 875 . 503 , z = 4 ) , was identified ( add weight of KK123 = 316 . 27 ) . Site-defining ions in the fragmentation spectra identified the site of KK123 insertion as Y415 , at the C-terminus of α1-TM4 ( underlined in the sequence; see Fig 2A ) . In a separate series of experiments , α1β3 receptors were photolabeled with KK123 , which was then coupled to FLI-tag using click chemistry . FLI-tag , an azide-containing tag , adds both charge and a heavy/light stable isotope pair to a photolabeled peptide , enhancing identification by creating doublets in the MS1 spectra [35] . MS1 level search for pairs of ions differing by 10 . 07 mass units found two peptide ion features ( m/z = 1 , 073 . 246 and m/z = 1 , 076 . 580 , z = 3 ) that had identical chromatographic retention times ( Fig 2B ) . Fragmentation spectra revealed both of these peptides as β3-TM4 peptide ( 426IVFPFTFSLFNLVYWLYKK123YVN445 ) with a KK123-FLI-tag adduct ( adduct mass = 672 . 432 and mass = 682 . 441 ) on Y442 ( Fig 2C ) . In the fragmentation spectrum , ions containing KK123 plus light FLI-tag ( Fig 2C , black ) were different by 10 . 07 mass units from the corresponding fragment ions from KK123 plus heavy FLI-tag ( Fig 2C , red ) , confirming that KK123 photolabels Y442 of the β3 subunit . β3-Y442 is located on the C-terminus of β3-TM4 in a homologous position to α1-Y415 , the KK123 photolabeling site in α1-TM4 ( Fig 2D , upper right panel ) . Thus , KK123 labeling data identified two discrete sites , one in α1 and the other in β3 . We employed additional photolabeling reagents containing TPD groups arrayed around the sterol backbone to confirm whether the KK123-labeled residues represent neurosteroid binding sites and to determine the orientation of the neurosteroids in these sites . KK200 , which has a TPD photolabeling group attached at C17 on the steroid backbone , has been previously used to map neurosteroid binding sites on GLIC [31] . Analysis of α1β3 receptors photolabeled with 15 μM KK200 detected two photolabeled TMD peptides: an α1-TM4 peptide , 398IAFPLLFGIFNKK200LVYWATYLNREPQLK423 , was photolabeled with KK200 ( m/z = 898 . 002; z = 4 ) ; site-defining ions in the fragmentation spectra identified N408 as the modified residue ( Fig 3A ) . The N408 residue ( N407 in rat ) has previously been shown to be critical to neurosteroid potentiation of GABA-elicited currents [14 , 15] . A β3-TM3 peptide , 280AIDMYLMGCNEM+DTTFVFVFLALLEYAFVNYIFFGRKK200GPQR313 ( m/z = 1 , 188 . 352; z = 4; N-ethylmaleimide [NEM]; 1 , 4-dithiothreitol [DTT]; alkylation adduct ) , was also photolabeled with KK200 . Fragmentation spectra narrowed the possible sites of adduction to G308 or R309 , both at the junction of TM3 with the M3–M4 intracellular loop ( Fig 3B ) . Analysis of GABAA receptors photolabeled with KK202 ( Fig 3C and 3D ) , identified two photolabeled peptides eluting two minutes apart . Both peptides were identified as the β3-TM3 peptide , 278VKAIDMYLMGCNEMFVFVFLALLEYAFVNYIFFGRGPQR313 ( m/z = 811 . 453 , z = 6 ) . Fragmentation spectra of the earlier eluting peptide localized labeling to a three-residue sequence , 278VKA280 , at the N-terminus of β3-TM3 ( Fig 3C ) . The fragmentation spectrum of the later eluting peptide , identified L294 as the site of adduction ( Fig 3D ) . ( The different retention time of the two photolabeled peptides is likely due to differences in peptide conformation and surface hydrophobicity resulting from incorporation of the photolabeling reagent into different residues . ) An important test of whether the photolabeled sites constitute specific allopregnanolone binding sites is the ability of excess allopregnanolone to competitively prevent photolabeling . Photolabeling studies for site identification were performed using 15 μM photolabeling reagent and achieved levels of labeling efficiency varying from 0 . 06% to 3 . 0% ( S1 Table ) . Because allopregnanolone has limited aqueous solubility ( about 30 μM ) and a large competitor excess is needed to demonstrate competition ( particularly with an irreversibly bound ligand ) , we were limited to studying competition at the photolabeled residues that could be detected following photolabeling at a concentration of 3 μM . Accordingly , we measured the photolabeling efficiency obtained following photolabeling of α1β3 GABAA receptors with 3 μM KK123 , KK200 , or KK202 in the presence or absence of 30 μM allopregnanolone . KK123 photolabeled both α1-Y415 ( 0 . 77% efficiency ) and β3 -Y442 ( 0 . 37% efficiency ) . For both of these residues , photolabeling was reduced by >90% in the presence of excess allopregnanolone ( Fig 4A ) . KK200 photolabeled β3 -G308/R309 ( 0 . 19% efficiency ) , and labeling was reduced by 98% in the presence of allopregnanolone . KK202 labeled both β3-L294 ( 0 . 29% efficiency ) and β3-278VKA280 ( 0 . 21% efficiency ) in TM3; labeling of both of these sites was undetectable in the presence of 30 μM allopregnanolone . Studies were also performed to determine whether the orthosteric agonist GABA ( 1 mM ) enhanced photolabeling by 3 uM KK123 or KK200 . Labeling efficiency was not significantly enhanced in the presence of GABA . This suggests that there is a small difference in neurosteroid affinity for closed versus open/desensitized states , which is consistent with the fact that neurosteroids have very low efficacy as direct activators of GABAA receptors [45] . Modification of ligand analogues with labeling groups at different locations has been used to determine the orientation of the ligands within their binding pockets [46] . Here , the six residues photolabeled by KK123 , KK200 , and KK202 were examined in a model of the α1β3 receptor created by threading the aligned sequence of the α1 subunit on the structure of the β3 subunit ( PDB 4COF ) [47] . The photolabeling sites grouped into the following three clusters: cluster 1 ( brown circle ) , β3-L294 ( KK202 ) and β3-G308/R309 ( KK200 ) ; cluster 2 ( red circle ) , α1-Y415 ( KK123 ) and α1-N408 ( KK200 ) ; and cluster 3 ( blue circle ) , β3-Y442 ( KK123 ) and β3-278VKA280 ( KK202 ) ( Fig 5A ) . In cluster 1 ( brown circle , Fig 5A and 5D ) , β3-L294 faces into the β ( + ) /α ( − ) intersubunit cleft , and G308/R309 is at the junction between the bottom of TM3 and the TM3–4 intracellular loop . G308/R309 is two α-helical turns below β3-F301 ( i . e . , toward the intracellular terminus of TM3 ) , a residue previously photolabeled by 6-azi pregnanolone in β3 homomeric receptors [13] . These data support neurosteroid binding in the β ( + ) /α ( − ) interface , consistent with the canonical THDOC and pregnanolone binding sites identified in crystal structures of α1 ( + ) /α1 ( − ) interfaces in chimeric proteins [16] and in substituted cysteine modification protection studies of α1β2γ2 receptors [18] . The pattern of labeling also indicates that the A-ring of the steroid is oriented upwards in the intersubunit cleft toward the center of the membrane , the D-ring is pointing toward the intracellular termini of the TMDs , and the C5-C6-C7 edge of the steroid is pointing toward the β3 ( + ) side of the cleft . Cluster 1 corresponds to a β3 ( + ) /α1 ( − ) intersubunit site . In cluster 2 ( red circle , Fig 5A and 5C ) , N408 and Y415 are both on the C-terminal end of α1-TM4 , facing toward TM1 within the same α1 subunit , consistent with an α1 intrasubunit neurosteroid binding site . N408 , the residue labeled by the C17-TPD of KK200 , is two α-helical turns closer to the center of TM4 than is Y415 , the residue labeled by the C6-diazirine of KK123 . This labeling pattern suggests that neurosteroids orient in this site with the A-ring pointing toward the ECD and the D-ring facing to the center of the TMD . Cluster 2 corresponds to an α1 intrasubunit site . In cluster 3 ( blue circle , Fig 5A and 5B ) , Y442 is located at the C-terminal end of β3-TM4 , and 278VKA280 is located on the TM2–TM3 loop near the extracellular end of β3-TM3 . The adjacency of these two photolabeling sites suggests an intrasubunit neurosteroid binding site at the extracellular end of β3 , analogous to the α1 intrasubunit site . The labeling of 278VKA280 in the extracellular loop by the C3-TPD group of KK202 suggests that neurosteroids orient in this site with the A-ring facing the ECD . Cluster 3 corresponds to a β3 intrasubunit site . A homology model of the α1β3 GABAA receptor based on the structure of a β3 homomeric GABAA receptor ( PDB 4COF ) [47] was used to examine the preferred energetic poses of neurosteroid binding to the three binding sites . The homology model was embedded in a 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine ( POPC ) bilayer and the structure refined by molecular dynamics . We then docked each of the three photoaffinity labeling reagents as well as allopregnanolone to each of the proposed binding sites , using a time course series of snapshots from the simulation trajectory to account for receptor flexibility . All of the neurosteroid photolabeling reagents docked in the three sites; the identified sites are relatively shallow with respect to the protein–lipid interface . Moreover , the neurosteroid analogues were all found to adopt multiple poses in each of the sites with minimal energy differences between the poses ( see Materials and methods ) . Photolabeling data combined with the docking scores ( binding energy ) and population of a given pose were used to guide selection of the preferred steroid orientation in each site . In the α1 intrasubunit site , the poses clustered between TM1 and TM4 . The preferred pose ( Fig 5C ) for allopregnanolone ( lowest energy cluster of poses ) shows the A-ring oriented toward the ECD with the walls of the predicted binding site lined on one side by N408 and Y415 and on the other by V227 . Docking of KK200 in this site has a similar orientation with the A-ring oriented toward the ECD and the TPD group on the D-ring proximal to N408 ( Fig 6A ) . Docking of KK-123 shows a preferred pose in which the A-ring is oriented toward the ECD and the C6-diazirine proximal to Y415 ( Fig 6B ) . These data elucidate a prior finding that mutations to N408 and Y411 eliminate potentiation by steroid analogues that lack a hydrogen bonding group on the D-ring [48] . In the β3 intrasubunit site , the poses are clustered between TM3 and TM4 . Allopregnanolone preferred a pose with the A-ring oriented toward the ECD near Y442 and the D-ring proximal to V290 ( Fig 5B ) . KK123 was found to dock at the top of the TM helices with the A-ring oriented toward the ECD , placing the 6-diazirine in proximity to Y442 ( Fig 6C ) . KK202 was found to dock in a similar orientation but lower in the TM region , with the TPD group in proximity to A280 and Y442 ( Fig 6D ) . In the intersubunit site , the preferred pose for allopregnanolone was one of the lowest energy clusters of poses with the A-ring proximal to α1-Q242 ( equivalent to rat α1-Q241 ) , the D-ring pointing toward the cytoplasmic termini of the TMDs , and the D-ring facing β3-F301 ( Fig 5D ) . Docking of KK200 showed a similar orientation although shifted slightly upwards toward the ECD , placing the A-ring near α1-Q242 , the benzene ring of the TPD near β3-F301 , and the diazirine in proximity to G308 ( Fig 6E ) . The preferred pose of KK202 was closer to TM3 of the β3 subunit with the A-ring near α1-Q242 and the D-ring near the β3-F301 placing the diazirine in proximity to β3-L294 ( Fig 6F ) . The orientation of allopregnanolone docked in our α1β3 model is nearly identical to the orientation of THDOC in the crystal structure of α1-GLIC [16] ( PDB 5OSB ) . As a confirmation of our docking , we also docked allopregnanolone to the apo-neurosteroid crystal structures of the α1-GLIC [16] and α5-β3 chimeric [17] proteins ( PDB 5OSA and PDB 5OJM , respectively ) ( S5 Fig ) . The preferred poses for allopregnanolone in the β3-α1 intersubunit site are nearly identical between the three models . The preferred poses are also very similar in the α1 intrasubunit site between our α1β3 homology model and the structures of the α-homomeric TMDs . The A-ring/D-ring orientation of allopregnanolone in the three neurosteroid sites was consistent with the orientations identified by the photolabeling data in all of the GABAA receptor structures . The calculated binding energies from the docking studies ( S4 Table ) indicate that the rank order of allopregnanolone affinity for the three sites is β3/α1 intersubunit site > α1 intrasubunit site > β3 intrasubunit site . The β3-α1 intersubunit binding site identified in our photolabeling studies has been extensively validated by site-directed mutagenesis as a functionally important site . Mutations on the α ( − ) side of the interface , including Q241 ( rat ) L/W and W246 ( rat ) L , have been shown to eliminate neurosteroid potentiation and gating of α1β2γ2 GABAA receptors [14 , 27 , 49]; mutations on the β ( + ) side of the interface , including F301A and L297A , have also been shown to partially reduce neurosteroid effect [17] . In the current study , we showed that α1Q242Lβ3 prevented the action of allopregnanolone , KK123 , and KK200 while reducing the effect of KK202 in α1β3 receptors , confirming the β–α interface as a functionally significant neurosteroid binding site and validating the relevance of our photolabeling reagents ( Fig 1B ) . Based on computational simulation and docking results , we also identified residues in the proposed α1- and β3-intrasubunit binding sites that we predicted could be involved in allopregnanolone binding or action ( S2 Table and S3 Table for all mutated subunits tested ) . N408 and Y411 in α1-TM4 line one side of the putative α1 intrasubunit site , and V227 in α1-TM1 lines the other ( Fig 5C ) . α1N407 ( rat ) A and α1Y410 ( rat ) W mutations have previously been shown to prevent neurosteroid potentiation of GABA-elicited currents in α1β2γ2 GABAA receptors [15] . Our data confirm that the double mutant α1N408A/Y411Fβ3 substantially reduces allopregnanolone potentiation of GABA-elicited currents ( Fig 4B and 4C1 , ***p < 0 . 001 versus α1β3 wild-type ) . Allopregnanolone ( 1 μM ) potentiation of GABA-elicited currents and direct activation ( 10 μM ) of α1V227Wβ3 receptors was also significantly reduced in comparison to α1β3 wild-type ( *p < 0 . 05 and **p < 0 . 01; Fig 4B , 4C1 and 4C2 ) . To test whether these mutations selectively affected neurosteroid actions , we also compared the effect of propofol in α1V227Wβ3 and α1N408A/Y411Fβ3 to its effect on wild-type α1β3 receptors . Propofol action was not different between the mutant and wild-type receptors , indicating a selective effect on neurosteroid action ( Fig 4B , 4C3 and 4C4 ) . The finding that multiple mutations lining the α1-intrasubunit binding pocket selectively reduce allopregnanolone action buttresses the evidence that the photolabeled residues identify a specific , functionally important neurosteroid binding site . Multiple mutations within the putative β3-intrasubunit binding site were also tested . However , none of the mutations significantly altered potentiation or activation by allopregnanolone ( S2 Table and S3 Table for all of the mutations that were tested ) . These data suggest that allopregnanolone occupancy of the β3 intrasubunit site does not contribute to channel gating . Direct activation of α1β2Y284Fγ2 receptors by THDOC has previously been shown to be markedly reduced in comparison to wild-type receptors [15] , although we found no significant effect of the β3-Y284 mutation in α1β3 receptors ( S2 Table and S3 Table ) . The difference in results between experiments in α1β3 and α1β2γ2 GABAA receptors suggests possible receptor subtype specificity in the functional effects of neurosteroid binding at a β-intrasubunit site .
Collectively , the photolabeling , modeling , and functional data indicate that heteropentameric α1β3 GABAA receptors contain at least seven binding sites for neurosteroids , of three different types . The use of multiple photolabeling reagents also enabled determination of the orientation of neurosteroids in each proposed class of sites . At least two of these classes are involved in producing the allosteric effect of steroids , the β3-α1 intersubunit site ( two copies per receptor ) and the α1 intrasubunit site ( two copies ) . Mutations of residues in the proposed β3 intrasubunit site ( three copies ) had no effect on modulation by allopregnanolone although residues were labeled by two photolabeling reagents and labeling was prevented by excess allopregnanolone . Accordingly , the functional significance of this proposed site is not known . Previous , site-directed mutagenesis studies using electrophysiology readout identified multiple residues , including α1-Q241 , N407 , Y410 , T236 , and β3-Y284 , that selectively contribute to the positive allosteric effects of neurosteroids [14 , 27] . Based on homology to the structure of the muscle nicotinic acetylcholine receptor [50] , it was hypothesized that there are two neurosteroid binding sites on GABAA receptors: an α1-intrasubunit site spanning Q241 and N407 and an intersubunit site between β3-Y284 and α1-T236 . Subsequent data [16–18] have clearly established the existence of a β–α intersubunit site . Our photolabeling experiments and homology modeling now show that the previously identified residues contribute to multiple distinct neurosteroid binding sites , albeit differently than originally proposed . It is noteworthy that the α1-intrasubunit site was not identified in the X-ray crystallographic structures of α1-GLIC chimeras bound with THDOC or the α5-β3 chimera bound with pregnanolone . This is likely because the proteins with steroid bound in the intrasubunit site did not form stable crystals . Mutations in either the β3-α1 intersubunit site or the α1-intrasubunit site can ablate both potentiation and direct activation by allopregnanolone , indicating that these are not distinct sites mediating potentiation and direct activation . The data also do not conform to simple energetic additivity for the two sites . The observation that mutations in either binding site can largely eliminate neurosteroid effect indicates that these two sites do not function completely independently and suggests allosteric interaction between the two sites . Development of site-selective neurosteroid analogues ( PAMs and antagonists ) should facilitate clarification of the mechanisms of allosteric interaction between these two sites . In light of the demonstration of multiple neurosteroid binding sites in α1β3 GABAA receptors , the possibility of additional isoform-specific sites must be considered . The strong sequence homology between the TMDs of the six α-subunits and three β-subunits suggests that there will not be large isoform differences in the intersubunit site [27] . In contrast , the contribution of ECD residues to the α- and β-intrasubunit sites suggests possible isoform-specific differences . The sequence homology between the γ and δ subunits and α and β subunits suggests that there may also be intrasubunit neurosteroid binding sites in these isoforms . Identification of a neurosteroid binding site on a δ-subunit would be of particular relevance because GABAA receptors containing δ-subunits are particularly sensitive to neurosteroids [51–53] . High-resolution , cryo-electron microscopy structures of α1β3γ2 GABAA receptors [54–56] have been published since initial submission of this work . The structural homology between γ2 subunits and α and β subunits suggests that there may also be intrasubunit neurosteroid binding sites in the γ2 subunit . The existence of multiple sites in which neurosteroids bind with different orientation may also offer some explanation for the difficulty in identifying neurosteroid antagonists [57] and for the differences in single-channel electrophysiological effects of various neurosteroid analogues [28 , 30] . The possibility of multiple isoform-specific sites with distinct patterns of neurosteroid affinity , binding orientation , and effect offers the exciting potential for the development of isoform-specific agonists , partial agonists , and antagonists with targeted therapeutic effects .
The human α1 and β3 subunits were subcloned into pcDNA3 for molecular manipulations and cRNA synthesis . Using QuikChange mutagenesis ( Agilent ) , a FLAG tag was first added to the α1 subunit then an 8xHis tag was added to generate the following His-FLAG tag tandem ( QPSLHHHHHHHHDYKDDDDKDEL ) , inserted between the fourth and fifth residues of the mature peptide . The α1 and β3 subunits were then transferred into the pcDNA4/TO and pcDNA5/TO vectors ( ThermoFisher Scientific , Waltham , MA ) , respectively , for tetracycline-inducible expression . For X . laevis oocytes , point mutations were generated using the QuikChange site-directed mutagenesis kit ( Agilent Technologies , Santa Clara , CA ) and the coding region fully sequenced prior to use . The cDNAs were linearized with Xba I ( NEB Labs , Ipswich , MA ) , and the cRNAs were generated using T7 mMessage mMachine ( Ambion , Austin , TX ) . The tetracycline-inducible cell line HEK T-RexTM-293 ( ThermoFisher ) was cultured under the following conditions: cells were maintained in DMEM/F-12 50/50 medium containing 10% fetal bovine serum ( tetracycline-free , Takara , Mountain View , CA ) , penicillin ( 100 units/ml ) , streptomycin ( 100 g/ml ) , and blastcidine ( 2 μg/ml ) in a humidified atmosphere containing 5% CO2 . Cells were passaged twice each week , maintaining subconfluent cultures . Stably transfected cells were cultured as above with the addition of hygromycin ( 50 μg/ml ) and Zeocin ( 20 μg/ml ) . A stable cell line was generated by transfecting HEK T-RexTM-293 cells with human α1-8x His-FLAG pcDNA4/TO and human β3 pcDNA5/TO in a 150 mm culture dish , using the Effectene transfection reagent ( Qiagen ) . Two days after transfection , selection of stably transfected cells was performed with hygromycin and zeocin until distinct colonies appeared ( usually after two weeks ) . Medium was exchanged several times each week to maintain antibiotic selection . Individual clones ( about 65 ) were selected from the dish and transferred to 24-well plates for expansion of each clone selected . When the cells grew to a sufficient number , about 50% confluency , they were split into two other plates , one for a surface ELISA against the FLAG epitope and a second for protein assay , to normalize surface expression to cell number [58] . The best eight clones were selected for expansion into 150 mm dishes , followed by [3H]muscimol binding . Once the best expressing clone was determined , the highest-expressing cells of that clone were selected through fluorescence-activated cell sorting ( FACS ) . FACS was done against the FLAG epitope , using a phycoerythrin ( PE ) -conjugated anti-FLAG antibody . Fluorescent-activated cells ( 1 ml containing about 10 million cells ) were sorted on the AriaII cell sorter ( Washington University Pathology Core ) , collecting 0 . 5% of the highest-fluorescing cells in a culture tube containing complete medium . The cells were plated in a 35 mm dish and expanded until a near confluent 150 mm dish was obtained . Cells were enriched for expression by FACS three times . A final FACS was performed to select individual cells into a 96-well plate , which resulted in only 10 colonies of cells . These colonies were expanded and assayed for [3H]muscimol binding; the highest-expressing clone was used for experiments . Stably transfected cells were plated into fifty 150 mm dishes . After reaching 50% confluency , GABA receptors were expressed by inducing cells with 1 μg/ml of doxycycline with the addition of 5 mM sodium butyrate . Cells were harvested after 48 to 72 hours after induction . HEK cells , after tetracycline induction , grown to 70%–80% confluency , were washed with 10 mM sodium phosphate/proteinase inhibitors ( Sigma-Aldrich , St . Louis , MO ) two times and harvested with cell scrapers . The cells were washed with 10 mM sodium phosphate/proteinase inhibitors and collected by centrifugation at 1 , 000 g at 4°C for 5 minutes . The cells were homogenized with a glass mortar Teflon pestle for 10 strokes on ice . The pellet containing the membrane proteins was collected after centrifugation at 34 , 000 g at 4°C for 30 minutes and resuspended in a buffer containing 10 mM potassium phosphate and 100 mM KCl . The protein concentration was determined with micro-BCA protein assay and stored at −80°C . [3H]muscimol binding assays were performed using a previously described method with minor modification [59] . Briefly , HEK cell membranes proteins ( 50 μg/ml final concentration ) were incubated with 1–2 nM [3H]muscimol ( 30 Ci/mmol; PerkinElmer Life Sciences ) , neurosteroid in different concentrations ( 1 nM-10 μM ) , binding buffer ( 10 mM potassium phosphate , 100 mM KCl [pH 7 . 5] ) , in a total volume of 1 ml . Assay tubes were incubated for 1 hour at 4°C in the dark . Nonspecific binding was determined by binding in the presence of 1 mM GABA . Membranes were collected on Whatman/GF-C glass filter paper using a Brandel cell harvester ( Gaithersburg , MD ) . To determine the Bmax of [3H]muscimol binding , 100 μg/ml of proteins were incubated with 250 nM [3H]muscimol , with specific activity reduced to 2 Ci/mmol , for 1 hour at 4°C in the dark . The membranes were collected on Whatman/GF-B glass filter papers using manifold . Radioactivity bound to the filters was measured by liquid scintillation spectrometry using Bio-Safe II ( Research Products International Corporation ) . Each data point was determined in triplicate . For all the photolabeling experiments , 10–20 mg of HEK cell membrane proteins ( about 300 pmol [3H]muscimol binding ) were thawed and resuspended in buffer containing 10 mM potassium phosphate , 100 mM KCl ( pH 7 . 5 ) at a final concentration of 1 . 25 mg/ml . For photolabeling site identification experiments , 15 μM neurosteroid photolabeling reagent was added to the membrane proteins and incubated on ice for 1 hour . For the photolabeling competition experiments , 3 μM neurosteroid photolabeling reagent in the presence of 30 μM allopregnanolone or the same volume of ethanol was added for incubation . The samples were then irradiated in a quartz cuvette for 5 minutes , by using a photoreactor emitting light at >320 nm [59] . The membrane proteins were then collected by centrifugation at 20 , 000 g for 45 minutes . All of the photolabeling experiments to identify sites of neurosteroid photolabeling were performed at least three times . The photolabeled peptides and residues described in the text were all observed in replicate experiments . The amount of 10 mg of KK123 or ZCM42 photolabeled HEK membrane proteins were solubilized in 1 ml 2% SDS/PBS and incubated at room temperature for 2 hours . The protein lysate was collected by centrifugation at 21 , 000 g for 30 minutes . FLI-tag was clicked to the KK123- or ZCM-photolabeled proteins at room temperature overnight in PBS buffer containing 2% SDS , 100 μM FLI-tag [35] , 2 . 5 mM sodium ascorbate , 250 μM Tris [ ( 1-benzyl-1H-1 , 2 , 3triazol-4-yl ) methyl]amine , and 2 . 5 mM CuSO4 . The amount of 1% Triton/PBS was added to the protein lysate to an SDS final concentration of 0 . 05% . The protein lysate was loaded onto a streptavidin agarose column . The flow through was reloaded to the column two times or till the flow through was colorless and the streptavidin column was dark orange yellow . The column was washed with 10 ml 0 . 05% Triton/PBS and eluted by 10 ml 100 mM sodium dithionite/0 . 05%Triton/PBS . The column was turned into colorless after elutions . The eluted proteins were concentrated into 100 μl with 30 kDa cutoff Centricon apparatus . The supernatant of the Centricon tube was added into SDS-sample loading buffer , loaded to a 10% SDS-PAGE , and transferred to a PVDF membrane , followed by western blot with polyclonal rabbit anti-α1 raised against a peptide mapping within a cytoplasmic domain of human GABAR α1 subunit [60] ( Santa Cruz Biotechnology ) or monoclonal anti-β3 antibody against 370–433 of mouse GABAR β3 subunit [61] ( NeuroMab ) . The photolabeled membrane proteins were resuspended in lysis buffer containing 1% DDM , 0 . 25% cholesteryl hemisuccinate ( CHS ) , 50 mM Tris ( pH 7 . 5 ) , 150 mM NaCl , 2 mM CaCl2 , 5 mM KCl , 5 mM MgCl2 , 1 mM EDTA , and 10% glycerol at a final concentration of 1 mg/ml . The membrane protein suspension was homogenized using a Teflon pestle in a motor-driven homogenizer and incubated at 4°C overnight . The protein lysate was centrifuged at 20 , 000 g for 45 minutes , and supernatant was incubated with 0 . 5 ml anti-FLAG agarose ( Sigma ) at 4°C for 2 hours . The anti-FLAG agarose was then transferred to an empty column , followed by washing with 20 ml washing buffer ( 50 mM triethylammonium bicarbonate and 0 . 05% DDM ) . The GABAA receptors were eluted with ten 1-ml 200 μg/ml FLAG peptide and 100 μg/ml 3X FLAG ( ApexBio ) in the washing buffer . The 10 ml effective elutions containing GABAA receptors ( tested by western blot with anti-α1 or anti-β3 antibody ) were concentrated by 100 kDa cutoff Centricon filters into 0 . 1 ml . The purified GABAA receptors ( 100 ul ) were reduced by 5 mM tris ( 2-carboxyethyl ) phosphine ( TCEP ) at for 30 minutes followed by alkylation with 7 . 5 mM NEM for 1 hour in the dark . The NEM was quenched by 7 . 5 mM DTT for 15 minutes . These three steps were done at room temperature . Eight μg of trypsin was added to the protein samples and incubated at 4°C for 7–10 days . The digest was terminated by adding formic acid ( FA ) in a final concentration of 1% . The samples were then analyzed by an OrbiTrap ELITE mass spectrometer ( ThermoFisher ) as in previous work [13 , 31] with some modifications . Briefly , a 20 μl aliquot was injected by an autosampler ( Eksigent ) at a flow rate of 800 nl/min onto a home-packed polymeric reverse phase PLRP-S column ( Agilent , 12 cm × 75 μm , 300 Å ) . An acetonitrile ( ACN ) 10%–90% concentration gradient was applied in the flow rate of 800 nl/min for 145 minutes to separate peptides . Solvent A was 0 . 1% FA/water , and solvent B was 0 . 1%FA/ACN . The ACN gradient was as follows: isocratic elution at 10% solvent B , 1–60 minutes; 10%–90% solvent B , 60–125 minutes; 90% solvent B , 125–135 minutes; 90%–10% solvent B , 135–140 minutes; isocratic solvent B , 140–145 minutes . For the first 60 minutes , a built-in divert valve on the mass spectrometer was used to remove the hydrophilic contaminants from the mass spectrometer . The survey MS1 scans were acquired at acquired at high resolution ( 60 , 000 resolution ) in the range of m/z = 100–2 , 000 , and the fragmentation spectra were acquired at 15 , 000 resolution . Data-dependent acquisition of the top 20 MS1 precursors with exclusion of singly charged precursors was set for MS2 scans . Fragmentation was performed using collision-induced dissociation or high-energy dissociation with normalized energy of 35% . The data were acquired and reviewed with Xcalibur 2 . 2 ( ThermoFisher ) . The MS experiments of identification of the photolabeling sites and competition of photolabeling were replicated at least three times . The LC-MS data were searched against a customized database containing the sequence of the GABAA receptor 8X His-FLAG-α1 and β3 subunit and filtered with 1% false discovery rate using PEAKS 8 . 5 ( Bioinformatics Solutions Inc . ) . Search parameters were set for a precursor mass accuracy of 30 ppm , fragmentation ion accuracy of 0 . 1 Da , up to three missed cleavage on either side of peptide with trypsin digestion . Methionine oxidation , cysteine alkylation with NEM and DTT , any amino acids with adduct of KK123 ( mass = 372 . 16 ) , KK200 ( mass = 462 . 27 ) , KK202 ( mass = 500 . 31 ) , KK123 with light FLI-tag ( mass = 672 . 4322 ) , and KK123 with heavy FLI-tag ( mass = 682 . 44 ) were included as variable modification . The GABAA receptors were expressed in oocytes from the African clawed frog ( X . laevis ) . Frogs were purchased from Xenopus 1 ( Dexter , MI ) and housed and cared for in a Washington University Animal Care Facility under the supervision of the Washington University Division of Comparative Medicine . Harvesting of oocytes was conducted under the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the National Institutes of Health . The animal protocol was approved by the Animal Studies Committee of Washington University in St . Louis ( approval No . 20170071 ) . The oocytes were injected with a total of 12 ng cRNA in 5:1 ratio ( α1:β3 ) to minimize the expression of β3 homomeric receptors . Following injection , the oocytes were incubated in ND96 with supplements ( 96 mM NaCl , 2 mM KCl , 1 . 8 mM CaCl2 , 1 mM MgCl2 , 2 . 5 mM Na pyruvate , 5 mM HEPES , and 100 U/ml + 100 μg/ml penicillin + streptomycin and 50 μg/ml gentamycin [pH 7 . 4] ) at 16°C for 1–2 days prior to conducting electrophysiological recordings . The electrophysiological recordings were conducted using standard two-electrode voltage clamp . Borosilicate capillary glass tubing ( G120F-4 , OD = 1 . 20 mm , ID = 0 . 69 mm; Warner Instruments , Hamden , CT ) were used for voltage and current electrodes . The oocytes were clamped at −60 mV . The chamber ( RC-1Z; Warner Instruments , Hamden , CT ) was perfused with ND96 at 5–8 ml min−1 . Solutions were gravity-applied from 30-ml glass syringes with glass luer slips via Teflon tubing . The current responses were amplified with an OC-725C amplifier ( Warner Instruments ) , digitized with a Digidata 1200 series digitizer ( Molecular Devices ) and were stored using pClamp ( Molecular Devices ) . The peak amplitude was determined using Clampfit ( Molecular Devices ) . The stock solution of GABA was made in ND96 bath solution at 500 mM , stored in aliquots at −20°C , and diluted as needed on the day of experiment . Stock solution of propofol ( 200 mM in DMSO ) was stored at room temperature . The steroids were dissolved in DMSO at 10 mM and stored at room temperature . The α1β3 wild-type and mutant receptors were tested ( see Table 1 and S2 and S3 Tables ) for potentiation by steroids ( 3α5α-allopregnanolone , 3α5β-pregnanolone , KK123 , KK200 , and KK-202 ) and direct activation by steroids ( allopregnanolone KK123 , KK200 , KK-202 , and pregnanolone ) . As control , several receptor isoforms were tested for potentiation by propofol . For each receptor type , we also determined constitutive open probability ( Po , const ) . To estimate Po , const , the effect of 100 μM picrotoxin ( estimated Po = 0 ) on the holding current was compared to the peak response to saturating GABA + 100 μM propofol ( estimated Po = 1 ) . Po , const was then calculated as Ipicrotoxin ÷ ( Ipicrotoxin − IGABA+propofol ) [62] . Potentiation is expressed as the potentiation response ratio , calculated as the ratio of the peak response to GABA + modulator ( steroid or propofol ) to the peak response to GABA alone . The concentration of GABA was selected to produce a response of 5%–15% of the response to saturating GABA + 100 μM propofol . Direct activation by steroids was evaluated by comparing the peak response to 10 μM neurosteroid to the peak response to saturating GABA + 100 μM propofol . Direct activation by steroids is expressed in units of open probability that includes constitutive open probability . All data are given as mean ± SD and analyzed by one-way ANOVA followed by Dunnet’s multiple comparison to the control wild-type group . A homology model of the α1β3 GABAA receptor was developed using the crystal structure of the human β3 homopentamer published in 2014 ( PDB ID: 4COF ) [47] . In this structure , the large cytoplasmic loops were replaced with the sequence SQPARAA used by Jansen and colleagues [63] The pentamer subunits were organized as A α1 , B β3 , C α1 , D β3 , E β3 . The α1 sequence was aligned to the β3 sequence using the program MUSCLE [64] . The pentameric alignment was then used as input for the program Modeller [65] , using 4COF as the template; a total of 25 models were generated . The best model as evaluated by the DOPE score [66] was then oriented into a POPC membrane , and the system was fully solvated with 40715 TIP3 water molecules and ionic strength set to 0 . 15 M KCl . A 100 ns molecular dynamics trajectory was then obtained using the CHARMM36 force field and NAMD . The resulting trajectory was then processed using the utility mdtraj [67] , to extract a snapshot of the receptor at each nanosecond of time frame . These structures were then mutually aligned by fitting the alpha carbons , providing a set of 100 mutually aligned structures used for docking studies . The docking was performed using AutoDock Vina [68] on each of the 100 snapshots in order to capture the receptor flexibility . Docking boxes were built for the β3 intrasubunit site ( cluster 3 ) , the α1 intrasubunit site ( cluster 2 ) , and the β3-α1 intersubunit site ( cluster 1 ) . The boxes were centered around the residues photolabeled by KK123 , KK200 , and KK202 and had dimensions of 25 × 25 × 25 Ångströms , large enough to easily fit the linear dimensions of all of the steroids . For docking studies of allopregnanolone , the docking boxes were placed in the same locations but had smaller dimensions of 20 × 20 × 20 Ångströms . Docking was limited to an energy range of 3 kcal from the best docking pose and was limited to a total of 20 unique poses . The docking results for a given site could result in a maximum of 2 , 000 unique poses ( 20 poses × 100 receptor structures ) ; these were then clustered geometrically using the program DIVCF [69] . The resulting clusters were then ranked by Vina score and cluster size and visually analyzed for compatibility with the photolabeling results , which is the photolabeling group oriented in the correct direction to produce the observed photo adducts . The inorganic salts used in the buffers , GABA , picrotoxin , and the steroids 3α , 5α-allopregnanolone , and 3α , 5β-pregnanolone were purchased from Sigma-Aldrich . Propofol was purchased from MP Biomedicals ( Solon ) . | Neurosteroids are cholesterol metabolites produced by neurons and glial cells that participate in central nervous system ( CNS ) development , regulate neuronal excitability , and modulate complex behaviors such as mood . Exogenously administered neurosteroid analogues are effective sedative hypnotics and are being developed as antidepressants and anticonvulsants . Gamma amino-butyric acid Type A ( GABAA ) receptors , the principal ionotropic inhibitory neurotransmitter receptors in the brain , are the primary functional target of neurosteroids . Understanding the molecular details of neurosteroid interactions with GABAA receptors is critical to understanding their mechanism of action and developing specific and effective therapeutic agents . In the current study , we developed a suite of neurosteroid analogue affinity labeling reagents , which we used to identify three distinct binding sites on GABAA receptors and to determine the orientation of neurosteroid binding in each site . Electrophysiological studies performed on receptors with mutations designed to disrupt the identified binding sites showed that two of the three sites contribute to neurosteroid modulation of GABAA currents . The distinct patterns of neurosteroid affinity , binding orientation , and effect provide the potential for the development of isoform-specific agonists , partial agonists , and antagonists with targeted therapeutic effects . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"neurochemistry",
"crystal",
"structure",
"chemical",
"compounds",
"condensed",
"matter",
"physics",
"vertebrates",
"electrophysiology",
"neuroscience",
"organic",
"compounds",
"animals",
"xenopus",
"membrane",
"proteins",
"animal",
... | 2019 | Multiple functional neurosteroid binding sites on GABAA receptors |
Myosin Ic is thought to be the principal constituent of the motor that adjusts mechanical responsiveness during adaptation to prolonged stimuli by hair cells , the sensory receptors of the inner ear . In this context myosin molecules operate neither as filaments , as occurs in muscles , nor as single or few molecules , as characterizes intracellular transport . Instead , myosin Ic molecules occur in a complex cluster in which they may exhibit cooperative properties . To better understand the motor’s remarkable function , we introduce a theoretical description of myosin Ic’s chemomechanical cycle based on experimental data from recent single-molecule studies . The cycle consists of distinct chemical states that the myosin molecule stochastically occupies . We explicitly calculate the probabilities of the occupancy of these states and show their dependence on the external force , the availability of actin , and the nucleotide concentrations as required by thermodynamic constraints . This analysis highlights that the strong binding of myosin Ic to actin is dominated by the ADP state for small external forces and by the ATP state for large forces . Our approach shows how specific parameter values of the chemomechanical cycle for myosin Ic result in behaviors distinct from those of other members of the myosin family . Integrating this single-molecule cycle into a simplified ensemble description , we predict that the average number of bound myosin heads is regulated by the external force and nucleotide concentrations . The elastic properties of such an ensemble are determined by the average number of myosin cross-bridges . Changing the binding probabilities and myosin’s stiffness under a constant force results in a mechanical relaxation which is large enough to account for fast adaptation in hair cells .
The myosin family includes at least 20 structurally and functionally distinct classes [1 , 2] . Although they all exhibit a common chemomechanical cycle , myosin molecules have remarkably diverse functions-including intracellular transport , force production in muscles , and cellular migration-as well as important roles in sensory systems [3] . To understand the emergence of these different functions , it is necessary to characterize the biophysical details of the chemomechanical cycle for each myosin class . Myosin molecules transduce chemical energy into mechanical energy through the hydrolysis of adenosine triphosphate ( ATP ) . The hydrolysis reaction and the subsequent release of inorganic phosphate ( Pi ) and adenosine diphosphate ( ADP ) induce structural changes that result in a power stroke and generate forces . The biochemical reaction rates and the response to external forces determine the specific function of each myosin [3] . On the basis of their biochemical and mechanical properties , myosins have been classified into four groups: ( i ) fast movers , ( ii ) slow but efficient force holders , ( iii ) strain sensors , and ( iv ) gates [4] . Although single-molecule experiments and structural studies have vastly advanced our understanding of force-producing molecules , we still lack a consistent description that quantitatively relates cellular functions to the molecular details . One prominent case is myosin Ic , which has been identified as a component of the adaptation motor of the inner ear [5] . Hair cells in the inner ear transduce mechanical stimuli resulting from sound waves or accelerations into electrical signals . On the upper surface of each hair cell stands a hair bundle comprising dozens to hundred of actin-filled protrusions called stereocilia . Cadherin-based tip links connect the tip of each stereocilium to the side of the longest adjacent one . When a mechanical force deflects the bundle , the resultant shearing motion raises the tension in the tip links . This tension increases the open probability of transduction channels and allows ions to diffuse into the stereocilia , depolarizing the hair cell . To retain sensitivity , a hair cell adapts to a prolonged stimulus by changing the tension in the tip links . This adaptation has a fast component lasting a millisecond or less and a slow component of a few tens of milliseconds , the molecular details of which remain uncertain . To explain slow adaptation , it has been proposed that an ensemble of myosin Ic molecules alternately step up or slide down the actin filaments inside the stereocilia to regulate the tension in the tip links . Sliding of myosin is triggered by a locally elevated Ca2+ concentration . This picture has been quantitatively supported by experimental studies on hair cells and complemented by mathematical descriptions [6–9] . Fast adaptation describes the rapid reclosure of transduction channels after abrupt stimulation of the hair bundle . This process is poorly understood and several possible explanations at a molecular level are debated [6 , 10] . One promising mechanism is the release model , in which a component of the transduction apparatus becomes more flexible and abruptly releases some of the tension in the tip links , allowing the channels to close rapidly [11 , 12] . Although myosin Ic has been implicated in both slow and fast adaptation and an ensemble of myosin Ic molecules is a good candidate for the element that releases [10] , the precise role of myosin Ic in adaptation has yet to be elucidated . The rapid response of the transduction channels to a displacement of the hair bundle suggests a direct mechanical activation through the transformation of the deflection into a force by a spring [6 , 13] . This mechanism underlies the gating-spring hypothesis that is the prevailing explanation for mechanotransduction by hair cells . The elastic property of the gating spring is the most important parameter in setting the precise relation between the deflection of a hair bundle and the open probability of the ion channels . Despite numerous studies of the molecular components of the hair bundle and their biophysical properties , we remain uncertain of the identity of the gating spring [14–18] . Every molecule that lies in series with the tip link could in principle influence the elastic properties , including the ensemble of myosin Ic molecules . These molecules bind and unbind from actin filaments and thereby change the elasticity dynamically . In order to fully explain mechanotransduction by hair cells , it is important to understand how the dynamics of single myosin Ic molecules determines the elastic properties of an ensemble and how it is regulated . Over the past few years , the biophysical properties of individual myosin Ic molecules have been characterized in optical traps , biochemical assays , and structural studies [19–24] . Like other myosin isoforms , myosin Ic displays catch-bond behavior , a prolonged attachment to an actin filament in response to increased external force [19 , 25] . The force-sensitive step in myosin Ic’s cycle is the isomerization following ATP binding , however , and not ADP release as in other slow myosins [19 , 20] . To understand how this behavior relates to the molecule’s physiological function , we introduce a consistent mathematical description of myosin Ic’s cross-bridge cycle . After the introduction of the basic framework by Huxley and Huxley , cross-bridge models have been widely used to describe the dynamics of myosin motors [2 , 26–35] . However , these models often assume irreversible transitions at fixed nucleotide concentrations that determine the input of chemical energy . In a seminal work , T . L . Hill showed how to couple a description of an enzymatic cycle to free-energy transduction in a thermodynamically consistent manner , an approach that has been applied to study muscle myosin [36–39] . We build our cross-bridge cycle for myosin Ic on these concepts and furthermore include the catch-bond behavior . Our description allows a quantitative analysis of the differences between in vitro and in vivo conditions , of Ca2+ regulation , and of cooperativity between force-producing molecules . Here we introduce a thermodynamically consistent description of myosin Ic based on single-molecule data and focus on the responses to external force , to different nucleotide concentrations , and to the availability of actin . We use this description to predict the elastic properties of an ensemble of myosin molecules and highlight the potential implication for the release model of fast adaptation .
As a functional description of myosin Ic we introduce a chemomechanical cycle consisting of five states: one state in which myosin is unbound from actin and four actin-bound states . Because we primarily focus on the force-producing states , we consider only a single , effective unbound state that combines the actin-detached ADP⋅Pi and ATP states . Each of the actin-bound states is associated with the nucleotide occupancy of the binding pocket of the myosin head ( Fig 1 ) . Myosin Ic performs its main , 5 . 8 nm power stroke upon phosphate release; a smaller power stroke of 2 nm follows ADP release . To account for the work done by these power strokes , we include a force dependence of the associated transition rates . We consider an effectively one-dimensional description in which the force acts along the coordinate of the power stroke: a positive force is oriented in a direction opposite to the power stroke . The nucleotide-binding rates depend linearly on the nucleotide concentrations and the actin-binding rates increase linearly with the actin concentration . By cycling through the five states , myosin performs work whose magnitude is bounded by the free-energy input associated with the nucleotide concentrations . We base our description on the free-energy transduction of enzymes and thus ensure thermodynamic consistency . To incorporate myosin Ic’s unique force sensitivity , we include a simple force dependence of the rate of unbinding from the filament of myosin in the ATP state . Under high forces , we expect myosin Ic to be trapped in the ATP state . Therefore we consider the ADP state ( 3 ) , the nucleotide-free state ( 4 ) , and the ATP state ( 5 ) as strongly bound . The remaining states are weakly bound or unbound ( Fig 1 ) . Our description , which captures many of the characteristics of myosin Ic , incorporates as free variables the experimentally controllable quantities external force , nucleotide concentrations , and actin concentration . This approach allows us to obtain analytic expressions for quantities that have been measured in experiments , then to use that information to determine the unknown parameter values of the model . An overview of the parameters is given in Table 1 . A mathematical description of the cross-bridge cycle and details of the estimation of parameter values are presented in the Methods section . In a single-molecule experiment using an isometric optical clamp , the lifetime of the myosin Ic-actin bond was measured for different external forces and two sets of nucleotide concentrations [20] . Because a rapid transit into and out of the weakly bound state ( 2 ) could not be resolved experimentally , this bound lifetime must be interpreted as the average time tsb that myosin Ic spends in the strongly bound states . We determined an analytic expression for the unbinding rate t sb - 1 from the strongly bound states ( Eq 53 ) as functions of force and nucleotide concentrations and fit this function simultaneously to two sets of experimental data acquired for distinct nucleotide concentrations . This unbinding rate is independent of the transition rate ω15 and of the actin concentration . Both quantities determine how often the molecule binds to the filament rather than how long it remains bound . From the average time that myosin Ic resides in the weakly bound states we estimate the binding rate ω15 for an actin concentration of 100 μM appropriate for the experiments . A detailed explanation for the fitting procedure is given in the Methods section . Fits of the unbinding rate t sb - 1 from the strongly bound states describe the experimental data well , indicating that our description is able to capture the force sensitivity of myosin Ic ( Fig 2a ) . Although none of the transition rates can account individually for the plateau around zero force , their combined effect in the cycle clearly displays such a behavior , which is characteristic of myosin Ic . The numerical values obtained in this way for the transition rates ω21 ≃ 164 s−1 and ω 51 0 ≃ 314 s - 1 suggest that in the absence of force , state ( 2 ) and state ( 5 ) are both configurations from which the myosin head rapidly detaches . The force-distribution factors ( δ ) indicate that phosphate release is only weakly dependent on force ( δ1 ≃ 0 . 12 ) and ADP release not at all ( δ1 ≃ 0 ) . The concentrations of nucleotides in cells differ from those in single-molecule experiments . We can use our description to predict the behavior of myosin molecules for different nucleotide concentrations . Although in single-molecule experiments the phosphate concentration usually remains low , the phosphate concentration in vivo is on the order of 1 mM [2] . In cells the ATP concentration is also near 1 mM and the ADP concentration is around 10 μM [2] . In the remainder of this study we refer to these numbers as the physiological nucleotide concentrations . The unbinding rate does not significantly change for higher phosphate concentrations ( Fig 2a and 2b ) . The main reason for this robust behavior is the very low rate constant for phosphate binding ( Eq 38 ) . Even for a millimolar phosphate concentration the phosphate-binding rate ω32 is very small compared to the other transition rates in the cycle . In contrast , increasing the ADP concentration decreases the overall binding rate because the molecule spends more time in the ADP state . This effect can be counteracted by an increase in the ATP concentration ( Fig 2b ) . Using the formulation given in the Methods section with the explicit solutions in Eqs 45–49 , we can determine the steady-state probability distribution for the cross-bridge cycle at different nucleotide and actin concentrations ( Fig 3 ) . For physiological nucleotide concentrations and 100 μM of actin , myosin is trapped in the ATP state ( 5 ) under forces exceeding 2 pN ( Fig 3a ) . Comparing only the strongly bound states , the molecule predominantly occupies the ADP state ( 3 ) for forces smaller than 1 . 5 pN . According to our description , myosin Ic’s cycle through the strongly bound states is limited by ADP release for forces smaller than 1 . 5 pN and by ATP release for forces larger than 1 . 5 pN . This result is consistent with experimental findings [19 , 20] . In the stereocilium of a hair cell , myosin Ic is thought to extend between the crosslinked actin filaments of the cytoskeleton and the insertional plaque to which the tip link is anchored [5 , 6 , 40] . To analyze the implications of an environment with a high concentration of actin , we determined the probability distribution for an actin concentration of 10 mM ( Fig 3b ) . Because of the increased binding probability , the unbound state ( 1 ) is depopulated . The weakly bound ADP⋅Pi state ( 2 ) dominates for forces smaller than 2 pN and the ATP state ( 5 ) for larger forces . An increased ADP concentration of 250 μM traps the myosin head in the ADP state for forces smaller than 2 pN and larger than 4 pN ( Fig 3c ) . In the intervening regime the ATP state predominates . In our stochastic description without irreversible transitions , we define myosin’s effective velocity as the average number of forward power strokes minus the average number of reverse power strokes per time . We refer to this definition as an effective velocity to emphasize that this quantity is neither the gliding velocity of an actin filament nor the ensemble velocity of several myosin Ic heads cooperating to produce a continuous movement . Every time the myosin head traverses the states ( 2 ) → ( 3 ) → ( 4 ) it performs a net power stroke of size Δx1 + Δx2 . In contrast , the reverse pathway ( 4 ) → ( 3 ) → ( 2 ) is associated with a reverse power stroke of size − ( Δx1 + Δx2 ) . The effective velocity v is accordingly given in terms of the combined local excess fluxes ΔJij ( Eq 26 ) as v ≡ Δ x 1 Δ J 23 + Δ x 2 Δ J 34 . ( 1 ) An increasing actin concentration enhances the binding of myosin and therefore decreases its cycling time , which leads to a higher effective velocity ( Fig 4 ) . The velocity saturates for an actin concentration above 1 mM . For large forces the effective velocity decreases until it becomes negative for forces larger than the stall force . According to our thermodynamic description the stall force F s = k B T Δ x 1 + Δ x 2 ln [ ATP ] K eq [ ADP ] [ P i ] ( 2 ) arises directly from Δμ = Eme , the equality of the Gibbs free energy for the hydrolysis reaction and the mechanical output . This relation reflects an implicit assumption that all of the chemical energy can be converted into mechanical energy . To account for mechanical inefficiency , the description could be extended with a loss parameter . Because we restrict our analysis to forces smaller than 6 pN , for which power strokes have been observed experimentally , we ignore the precise behavior for larger forces and consider the stall force for myosin Ic as an unknown quantity . A widely accepted definition of the duty ratio is the fraction of the total duration of an ATPase cycle that myosin spends in the strongly bound states [3 , 41–43] . Ignoring the weakly bound , actin-attached states or combining them into other states , the duty ratio is often defined as the fraction of the total cycle time during which myosin is attached to an actin filament [2 , 44–46] . Because the initiation of myosin Ic’s power stroke is limited by phosphate release , myosin Ic can bind to actin in the ADP⋅Pi state but detach without proceeding through the cycle if it detaches prior to Pi release . Such an event contributes to the attachment to the filament but not to the time that the molecule spends in the strongly bound states . The time that the molecule spends in the strongly bound states therefore differs from that spent attached to the filament . The probability Psb of occupying the strongly bound states accordingly differs from the probability Pon of being attached to actin . Our complete cycle description allows us to explicitly calculate both probabilities and to compare them . We determine Psb in terms of the fraction of the cycle that the molecule spends in the strongly bound states as P sb ≡ t sb t sb + t wb = ∑ i = 3 5 P i , ( 3 ) in which tsb is the average time spent in the strongly bound states , twb is the average time spent in the weakly bound and detached states , and Pi is the steady-state probability ( Eqs 45–49 ) . Similarly , we obtain Pon from the fraction of the total cycle time during which the myosin molecule is attached to the filament as P on ≡ t on t on + t off = ∑ i = 2 5 P i , ( 4 ) in which ton is the average time that myosin is attached to the filament , toff the average time that myosin is detached , and Pi is again the steady-state probability ( Eqs 45–49 ) . Whereas the former quantity is closely related to the duty ratio , the later quantity is important for estimation of the number of bound molecules in an ensemble . The probabilities of being attached to actin and of occupying the strongly bound states depend on the ADP concentration , on the available actin , and on the external force ( Fig 5 ) . In general , because of the catch-bond behavior an increasing force enhances the probability of attachment to actin . An elevated ADP concentration likewise traps myosin Ic in the strongly bound ADP state and increases both probabilities ( Fig 5a and 5c ) . An increased accessibility of actin enhances the binding of the myosin head , which results in a high-almost unity-probability of being bound to the filament at high actin concentrations ( Fig 5d ) . In contrast , the probability of occupying the strongly bound states saturates at a high actin concentration , for entering these states is limited by phosphate release ( Fig 5b ) . Although in vestibular hair cells myosin Ic activity is required for fast adaptation , the precise molecular details remain unknown [10] . Here we focus on two aspects that might contribute to the mechanism: the cooperative unbinding of an ensemble of myosin heads under force and a qualitative Ca2+ dependence that changes the binding probability and the elasticity of individual myosin Ic molecules [23 , 24] . In particular , we determine how these properties influence the overall elasticity of an ensemble . The myosin heads contribute to the rigidity of the adaptation motor by crosslinking the insertional plaque to the actin cytoskeleton . We think of each myosin head as a linear spring , arranged in parallel to the others , such that the overall stiffness is given by the sum of the actin-attached myosin heads multiplied by the stiffness of each myosin molecule . Because the binding and unbinding of the heads depend on the force and the nucleotide and actin concentrations , these quantities also influence the overall elastic properties of the ensemble . In general the binding process could be very complicated because of the geometry and possible steric interactions between the heads . Furthermore the helical structure of the actin filaments provides binding sites with an appropriate orientation only about every 37 nm [5] . These constraints change the number of myosin molecules that can potentially interact with actin . In our description , the total number of myosin heads is thus an effective number of molecules that can potentially bind to actin . To estimate the average number of bound myosin molecules in an ensemble , we use the attachment and detachment rates determined from our description of the chemomechanical cycle . We assume that each myosin head can bind to the filament with a binding rate kon and unbind with an unbinding rate koff . Both rates stem directly from our description , kon = ω12 + ω15 and koff from Eq 58 . Because of the stochastic binding and unbinding , the number n of bound molecules fluctuates . To describe the system as a Markov chain , we introduce a state space ( Fig 6a ) associated with the number of bound myosin heads [47] . The effective transition rates between these states are k on n ≡ ( N - n ) k on , ( 5 ) and k off n ≡ n k off . ( 6 ) Here kon depends on the actin concentration and koff on the nucleotide concentrations and on the force f per myosin molecule . We assume that an external force F applied to the ensemble is distributed equally among the attached myosin molecules , resulting in the effective force f = F/n per attached head . If one head releases from the filament then the force is redistributed among the remaining bound heads and the force per myosin molecule accordingly increases , which changes the unbinding rate koff . In general this mechanism leads to cooperative effects because the unbinding rate depends on the number n of attached myosin heads . In the case in which the myosin heads act independently , the transition rates of a single head are independent of the number of attached myosin molecules . We determine the average number of bound myosin molecules from the linear Markov chain as explained in the Methods section , n = ∑ n = 0 N n 1 + ∑ l = 0 N - 1 ∏ i = 0 l k on i k off i + 1 - 1 ∏ j = 0 n - 1 k on j k off j + 1 . ( 7 ) For the cooperative case in which koff = koff ( F/n ) , we evaluate this equation . In the independent case , in which the unbinding rate koff is independent of the number of bound myosin heads , we can simplify this expression to n = N 1 + k off / k on = N t on t on + t off = N P on . ( 8 ) Note that Pon = Pon ( f ) is a function of the force acting on a single myosin head . For the independent case , we estimate this force by f = F/N . However , in this way we underestimate the magnitude of the force per molecule because we expect that N > n . For a better approximation , we distribute the external force between the mean number of bound motors , f = F/〈n〉 , an approach that leads to an implicit equation for 〈n〉 that is not easy to solve . For physiological nucleotide concentrations and for 100 μM actin , we notice in Fig 5d that 〈Pon〉 ≈ 0 . 5 . Using this value , we estimate that in a group of 30 molecules about 〈 n ˜ 〉 ≃ 15 of them are bound on average . We then approximate the average force on a myosin molecule as f = F / 〈 n ˜ 〉 for the independent case . Note that in the independent case the force per myosin head does not depend on the number of bound heads , in contrast to the cooperative case . The mean number of bound myosin heads is influenced by the cooperative release of the molecules and the three approaches are different for intermediate forces ( Fig 6b ) . We calculate the average number of bound myosin heads as a function of force for different total numbers of myosin molecules ( Fig 7a ) . In small ensembles , the force per head is higher and therefore more heads are bound as a result of the catch-bond behavior . Increasing the concentration of available actin causes more myosin heads to attach to the filament ( Fig 7b ) . To validate our effective description , we compare our analytic results to Monte Carlo simulations as detailed in the Methods . In these simulations , each myosin head is represented as a spring that is attached to a rigid common structure . At each time step of the simulation the extensions of all springs are calculated by solving Newton’s law of force balance . In this way , we obtain for each myosin head a force that determines the transition rates of the chemomechanical cycle of that molecule . There are important differences from the analytic approach . Whereas in the simulation a myosin head proceeds stochastically through the five-state chemomechanical cycle , the heads only bind and unbind in the analytic description . As a consequence the myosin molecules step stochastically and exert fluctuating forces on each other , which in turn influences their dynamics . In our analytic model , the myosin heads are only indirectly coupled through the number of bound motors and not through an elastic interaction . The simulations show reasonable agreement with the analytic results ( Figs 6b and 7 ) . An increased coupling stiffness increases the forces between the myosin heads , which in turn result in a longer attachment because of the catch-bond behavior ( Fig 6b ) . Especially for a high actin concentration , the agreement between the simulations and the analytic description is very good . In the following , we will focus on this particular case and therefore consider only the analytic description . These results show that the average number of bound myosin heads depends on the external force , the total number of myosin molecules , the actin concentration , and-not shown here-the nucleotide concentrations . We expect that the mechanical properties of a cellular structure including myosin Ic molecules also depend on these quantities . To investigate the elastic properties of an ensemble of myosin Ic heads , we determine the force-extension relation F = n κ x , ( 9 ) in which κ is the spring constant of a single myosin head . The underlying assumption of this approach is a linear force-extension relation of the individual myosin heads , for which we take the value of κ = 500 μN/m [21] . Applying forces below 20 pN to the ensemble leads to an extension smaller than 5 nm ( Fig 8a ) . A reduced total number N of myosin molecules increases the extension because the force per myosin head is larger and stretches it farther . To test whether a mechanical release of myosin Ic molecules is related to fast adaptation , we investigate two qualitative effects of Ca2+ . First , Ca2+ could decrease the binding probabilities of the myosin head to actin [23] . Second , it could change the stiffness of myosin by initiating the dissociation of one or more calmodulin molecules from the light chains , allowing the myosin molecules to attain a more flexible conformation [24] . We next consider the mechanical release owing to Ca2+ binding , Δ x ≡ F 1 κ Ca 2 + n Ca 2 + - 1 κ n . ( 10 ) We first study the effect of a reduced binding probability on the mean number of bound myosin molecules and maintain their stiffness before and after Ca2+ binding , κ Ca 2 + = κ = 500 μ N/m . We reduce the binding probability by the factor β and determine the resulting release for N = 10 or N = 20 myosin molecules ( Fig 8b ) . A large decrease of the binding probability leads to fewer bound molecules and a larger release . The release for a group of 10 myosin molecules exceeds that for an ensemble of 20 molecules: the force on each individual myosin head is higher and stretches the molecule farther . However , the overall distance for forces smaller than 20 pN is still less than 20 nm . When we add to the 100-fold decrease of the binding probability a tenfold decrease of myosin’s elasticity and determine the resulting release for different total numbers of myosin molecules ( Fig 8c ) , the displacement is of the order of several tens of nanometers and becomes almost insensitive to force for a group of 50 myosins .
An important goal of biology is understanding how the structures and interactions of molecules result in measurable functions of cells and organisms . By combining findings on different spatial scales in a consistent manner , mathematical descriptions help us understand how physiologically relevant function is determined by the interplay of molecular components . We have constructed a quantitative description of myosin Ic’s chemomechanical cycle and studied the resulting properties at both a single-molecule and an ensemble level , which allows us to discuss important implications on the physiological function of hair cells at the whole-cell level . On the single-molecule level , it is important to understand how different members of the large myosin family display distinct biophysical properties despite a common general structure of the chemomechanical cycle . To describe myosin Ic , we constructed such a cycle and chose as control parameter the nucleotide concentrations and the external force , both of which are experimentally accessible and biologically relevant . Our simplified , one-cycle description reproduces many of the characteristic features of myosin Ic , especially the force-dependent exit from the strongly bound states . The probabilities of occupying the different states indicate that myosin Ic’s strongly bound states are dominated by the ADP state for forces below 1 . 5 pN and by the ATP state for larger forces ( Fig 3 ) . Although this behavior is in contrast to previous models in which the ADP state is the only force-sensitive state , it is nevertheless consistent with the role of myosin Ic in adaptation [5 , 21] . Increasing the ADP concentration traps the myosin heads in the ADP state , bound to actin filaments ( Figs 3 and 2b ) . This effect can be reversed by increasing the ATP concentration ( Fig 2b ) . Such a behavior accords with recordings of transduction currents in hair cells isolated from the bullfrog: changing nucleotide concentrations alters the relative occupancy of the states in the cross-bridge cycle and thus the number of bound myosin molecules , which in turn controls the tension on the mechanically sensitive ion channels . Indeed , in the presence of an ADP analog , adaptation disappears and the tension on the channels increases . Both effects can be reversed by increasing the concentration of ATP [48] . This qualitative agreement constitutes direct evidence that the model , although constructed from single-molecule measurements in vitro , captures important aspects of the behavior of living cells . Our description suggests a low effective velocity for myosin Ic . Although velocities of only tens of nanometers per second have been reported from motility experiments in vitro [10 , 49–51] , larger values have been discussed [5] . In motility assays , multiple myosin molecules work together to create motion . How the velocity measured in motility experiments is related to the effective rate of a cross-bridge cycle and to other biophysical parameters of the molecules is an open question [52–56] . However , our stochastic simulations suggest that 200 elastically coupled myosin Ic molecules , each described by the five-state chemomechanical cycle , display a motility rate of 25 nm⋅s-1 which is in good agreement with the experimental values of 16–22 nm⋅s-1 [51] . In these experiments the myosin molecules where coupled through a membrane . Greater speeds of 60 nm⋅s-1 have been reported in gliding assays , but the data were acquired at a temperature of 37°C [22] , whereas the numerical values of the biochemical rates of our model stemmed from experiments conducted at 20°C . We conclude that our description of myosin Ic constrained by single-molecule data accords with the experimental data on a larger scale . Speeds of tens of nanometers per second are too low to be consistent with rates estimated for the adaptation motor in the inner ear , which has been associated with the function of myosin Ic [5 , 10] . Depending on the species , the velocity of the adaptation motor ranges from several hundred to a few thousand nanometers per second [5 , 57 , 58] . The discrepancy between the velocities in vivo and in vitro might stem from several factors . It is still unknown to what extent these rates relate to the speed of myosin Ic molecules and to relaxations of other elastic elements . It has been suggested that the recoil of an elastic element located parallel to the myosin heads , the extent spring , contributes to the dynamics [59 , 60] . Furthermore the reaction rates of the myosin cycle could be different in vivo and in vitro . In particular , the complex composition of the cytosol and molecular modifications could lead to differences in the energy barriers between the states [61] . Another possibility is that myosin Ic , which has been shown to constitute the adaptation motor of young mice [10] , might be replaced during subsequent development by the closely related paralog myosin Ih , which has been identified as a hair-bundle protein [62] . Myosin Ih’s molecular properties have yet to be characterized and it might operate more swiftly . In hair cells the deflection by a stimulus is communicated to the transduction channel by an elastic element , the gating spring . Of uncertain origin , this elasticity displays complex behavior with implications for sensory coding [63] . Our model suggests that a cluster of myosin Ic molecules contributes to this elasticity and additionally provide the regulatory function to explain fast adaptation . If we assume that Ca2+ reduces the binding probability of myosin by a hundredfold and its stiffness by tenfold , the resultant release on the order of 40 nm accords with measurements from frog hair bundles [10] . For displacements exceeding 400 nm the extent of fast adaptation is independent of the stimulus [10] . We speculate that the insensitivity of the release to the external force for an ensemble of 50 myosin molecules is related to this experimental observation ( Fig 8c ) . Although biochemical studies have suggested that stereocilia contain around 100–200 myosin Ic molecules , the number of actively engaged molecules in an adaptation motor is probably lower [5] . Although myosin Ic’s cycle is slow , binding of Ca2+ could rapidly change the relative occupancy of specific states . Under force , most of the myosin heads are trapped in the ATP state . The binding of Ca2+ to a myosin Ic molecule triggers the release of calmodulin from the IQ domains and increases the molecule’s flexibility , as recently shown in a structural study [24 , 64] . A sudden increase of flexibility would release the myosin head from any force until all elastic elements have relaxed to a new equilibrium state . In the absence of force , our description predicts a transition rate for unbinding from the ATP state as large as ω 51 0 ≃ 314 s - 1 . This value is so great that the head would unbind immediately , probably before the forces could be redistributed among the bound myosin molecules . Although the load-free biochemical rates have been reported to be rather insensitive to Ca2+ , this mechanism might explain a possible Ca2+-induced unbinding of myosin molecules from the actin filament under force [23] . Such a fast disengagement of the myosin molecules is necessary for the adaptation motor to slide down the stereociliary actin and thus to relax the tension in the tip links in order to accomplish adaptation to an abrupt stimulus . Because the transduction channels have been localized at the lower end of the tip links , Ca2+ regulation of the adaptation motor is effective only at the next lower insertional plaque [65] . Because the tallest stereocilia lack transduction channels through which Ca2+ could enter , the forces between different rows of stereocilia are differently regulated . Inner hair cells consisting of three rows of stereocilia might therefore display less Ca2+-regulated slow adaptation [14] . How much of the hair cell’s function is impeded by the reduced regulation is an open question . Hair-bundle models based on detailed descriptions of the relevant molecular mechanisms , such as myosin Ic’s chemomechanical cycle , could provide more insight . Biochemical experiments and studies of single-molecule motility are ordinarily conducted under chemostatic conditions in which energy sources such as ATP and products such as ADP and Pi are maintained at nearly constant concentrations . In the present study , however , we have endeavored in two ways to model the behavior of myosin Ic under more lifelike conditions . First , we have imposed a thermodynamic constraint that requires the modeled reaction cycles to respect energy balance . And second , we have examined an extensive range of concentrations for the relevant nucleotides and their products . A typical stereocilium , which is about 3 μm in length and 0 . 2 μm in diameter , has a volume of only 100 aL . Even a substance found at a high concentration in the cytoplasm , such as ATP at 1 mM , can be depleted rapidly in such a small volume . When transduction channels open , for example , the plasma-membrane Ca2+ ATPase in a stereocilium confronts a flood of Ca2+ that could exhaust the available ATP in only milliseconds ! It is thus important to understand the operation of myosin Ic-based motors under realistic and potentially fluctuating conditions . A final feature of the adaptation motors that remains to be investigated is the noise associated with their activity . By pulling directly on a tip link , each motor influences the opening and closing of the transduction channel or channels at the link’s opposite end . In conjunction with thermal bombardment of the bundle as a whole and stochastic clattering of the transduction channels , the adaptation motors in a hair bundle thus contribute to the mechanical noise that interferes with the detection of faint sounds and weak accelerations [66] . It will be interesting to learn whether the activation mechanism of the myosin molecules in adaptation motors or perhaps their cooperative behavior has been optimized to mitigate this source of noise . Our study has provided new insights into biological mechanisms . The chemomechanical cycle suggests that the force-dependent unbinding rate is rather robust even under physiological nucleotide concentrations . Although the force-sensitive state is the ATP-bound state , an increased ADP concentration reduces the unbinding rate and the myosin Ic molecules are strongly bound to actin filaments . The elastic properties of an ensemble of myosin Ic molecules can be regulated by an external force and by the actin and nucleotide concentrations . Although the reaction rates of actin-bound myosin Ic are largely insensitive to Ca2+ [23] , we have shown that in an ensemble of myosin Ic molecules a possible reduction of the binding rate and elasticity could nevertheless account for fast adaptation by hair cells .
The cross-bridge cycle of a myosin molecule consists of distinct states associated with different biochemical compositions and molecular conformations . The transitions between these states involve myosin’s head binding to and unbinding from the actin filament , nucleotide binding and release , and conformational changes . We simplify the cross-bridge cycle and describe myosin’s dynamics with one state ( 1 ) in which the head is detached from actin and four states ( 2 ) − ( 5 ) , in which the head is attached ( Fig 1 ) . The four actin-bound states correspond to distinct occupancies of the nucleotide-binding pocket: in state ( 2 ) ADP and Pi are bound , whereas in state ( 3 ) only ADP is bound . State ( 4 ) is the nucleotide-free state and state ( 5 ) refers to the ATP-bound state . We represent the cross-bridge cycle as a time-continuous Markov process for which we must specify the transition rates between the states . Although all transition rates could be force- and nucleotide-dependent , it is reasonable to assume that the main effect of the nucleotide concentrations is exerted on the nucleotide-binding rates . Before introducing those transition rates , we discuss the force dependencies of the mechanical transitions . The transition rates associated with a mechanical power stroke decrease with an increase in the opposing force . Experimental data indicate that myosin Ic performs its power stroke in two steps: the lever arm is remodeled by a distance of Δx1 ≃ 5 . 8 nm upon phosphate release and then by Δx2 ≃ 2 nm upon ADP release [20] . Assuming local equilibrium , we associate the ratio of the forward and backward transition rates for the power stroke upon phosphate release with a Boltzmann factor as ω 23 ω 32 = exp ( - ( Δ G 23 + F Δ x 1 ) / k B T ) . ( 11 ) Here ΔG23 is the Gibbs free-energy difference between the states , FΔx1 is the mechanical work performed by the power stroke of distance Δx1 against the opposing load force F , and kB T is the Boltzmann constant times the temperature [67] . The equation above relies on an assumption of local equilibrium that does not indicate how the individual transition rates depend on the force . Therefore , we use the following general forms for the individual transition rates: ω 23 ≡ ω 23 0 exp ( - δ 1 F Δ x 1 / k B T ) , ( 12 ) ω 32 ≡ ω 32 0 exp ( ( 1 - δ 1 ) F Δ x 1 / k B T ) , ( 13 ) in which we introduce the force-free rate constants ω 23 0 , ω 32 0 and the force-distribution factor δ1 ∈ [0 , 1] . The restriction on the numerical values for the force-distribution factor is a consequence of the assumption that a force opposing the power stroke diminishes the corresponding transition rate [67 , 68] . For an effective description based on a projection of a high-dimensional free-energy landscape on to a single reaction coordinate , the force-distribution factor is not restricted [69] . Using the same argument as for the release of phosphate , the general forms of the forward and backward transition rates associated with the power stroke upon ADP release are ω 34 ≡ ω 34 0 exp ( - δ 2 F Δ x 2 / k B T ) , ( 14 ) ω 43 ≡ ω 43 0 exp ( ( 1 - δ 2 ) F Δ x 2 / k B T ) . ( 15 ) To account for the force-dependent behavior of myosin Ic , we must include the force sensitivity of the isomerization following ATP binding [19] . In our simplified state space this sensitivity effectively changes the unbinding rate ω51 from the ATP state . We therefore introduce a force-dependent factor g ( F ) that modifies the unbinding rate ω 51 ≡ g ( F ) ω 51 0 . ( 16 ) We require that for zero force g ( F = 0 ) = 1 and for large force g ( F ≫ 1 ) = ωoff saturates and use g ( F ) ≡ 2 ( 1 - ω off ) 1 + exp ( ξ F / k B T ) + ω off , ( 17 ) in which ξ is a characteristic length scale . In general the binding interface between the head of a molecular motor and its filament is more complicated than the idealized receptor-ligand bond considered by Bell [70] . The bond interface consists of multiple partial charges that lead to complex unbinding pathways through multiple states in the free-energy landscape [71–74] . To capture the characteristic behavior , we use the force factor of Eq 17 that has been used previously to describe the chemomechanical cycle of kinesin-1 and myosin V [75–79] . In the following we give an intuitive justification of the force factor given in Eq 17 . In our description the ATP state ( 5 ) comprises several sub-states including the binding and isomerization of ATP . The unbinding rate ω51 , which must be considered as an effective rate for proceeding through all the sub-states , therefore includes the force dependence of the isomerization step . As a first approximation , we consider that isomerization is not associated with a conformational change that would result in a displacement of an applied load . The free-energy between the state ( A ) before isomerization and the state ( B ) after isomerization accordingly does not depend on the applied force . We assume that an applied force increases the free-energy barrier between those two states without changing the difference of the energy between the states . Motivated by Kramers rate theory [80] , we use the force dependence ω A B = ω A B 0 exp ( - F ξ / k B T ) for the forward transition rate and the same force dependence for the reverse transition rate ω B A = ω B A 0 exp ( - F ξ / k B T ) . We consider the main forward pathway through these sub-states , → ( A ) ⇌ ω A B ω B A ( B ) → ω B , ( 18 ) which implies the effective transition rate ω eff = ω B 1 + ω B A 0 ω A B 0 + ω B ω A B 0 e F ξ / kB T . ( 19 ) This approach leads to a force dependence similar to that in Eq 17 . Using the same argument for the reverse pathway , we find that the transition rate ω54 has a similar force dependence . As described later in Eq 43 , the force dependence of the transition rate ω54 is imposed naturally by thermodynamic consistency . To capture the dependence on the nucleotide concentrations , we consider the nucleotide-binding steps as first-order reactions that are independent of force , leading to ω 32 0 ≡ ω ^ 32 0 [ P i ] , ( 20 ) ω 43 0 ≡ ω ^ 43 0 [ ADP ] , ( 21 ) ω 45 ≡ ω ^ 45 0 [ ATP ] . ( 22 ) Note that the units of the rate constants with a caret are M−1s−1 . Such a linear dependence of the transition rates on the reactants is motivated by macroscopic chemical-reaction laws and is widely used to describe chemomechanical cycles [2 , 67] . In a similar way , we assume a linear dependence of the actin-binding rates on the actin concentration , ω 12 ≡ ω ^ 12 [ actin ] , ( 23 ) ω 15 ≡ ω ^ 15 [ actin ] . ( 24 ) The remaining transition rate ω54 is determined by a balance condition obtained from thermodynamic consistency . We have specified above the general forms of the transition rates of our theoretical description of myosin Ic . We next introduce the dynamics and the thermodynamic constraints . We consider the stochastic dynamics of the myosin head as a continuous-time Markov process [81] . The probability Pi ( t ) of finding myosin in state ( i ) therefore evolves in time t according to the master equation d d t P i ( t ) = - ∑ j Δ J i j ( t ) , ( 25 ) with a local net flux between the states ( i ) and ( j ) given by Δ J i j ( t ) ≡ P i ( t ) ω i j - P j ( t ) ω j i . ( 26 ) A thermodynamically consistent description , which ensures that myosin does not produce more mechanical energy than the chemical energy provided by the nucleotide concentrations , implies a relation between the mechanical energy and the chemical energy . This relation provides a constraint on the transition rates of the cycle that is obtained by incorporating free-energy transduction [37] . We can express the change in the Gibbs free energy of the hydrolysis reaction for a dilute solution by Δ μ ≡ k B T ln [ ATP ] [ ADP ] [ P i ] K eq , ( 27 ) in which Keq is the equilibrium constant for the reaction , here with the numerical value Keq ≃ 4 . 9 ⋅ 105 M [2] . Note that at the equilibrium concentration of the nucleotides the change in the free energy Δμ vanishes . The mechanical energy , which is the work done by the protein against an external force F , is given by E me ≡ ( Δ x 1 + Δ x 2 ) F . ( 28 ) This mechanical energy is produced when the protein passes through a forward cross-bridge cycle , which in our description represents directed transitions through the states ( 1 ) , ( 2 ) , ( 3 ) , ( 4 ) , ( 5 ) , and finally ( 1 ) . In contrast , the backward cycle is associated with a path traversed in the opposite direction . Thermodynamically consistent coupling of the energy conversion by the protein to the hydrolysis reaction then imposes the constraint ω 12 ω 23 ω 34 ω 45 ω 51 ω 21 ω 32 ω 43 ω 54 ω 15 = exp ( ( Δ μ - E me ) / k B T ) , ( 29 ) which can further be related to the entropy production [37 , 77] . This equation has an intuitive interpretation [37]: the left side is the ratio of the average number of complete forward cycles to the average number of complete backward cycles , whereas the right side is the exponential of the difference between the chemical input energy and the mechanical output energy . At equilibrium this difference vanishes and the right side is equal to one , which requires the completion of identical numbers of forward and backward cycles . This constraint ensures that the average net cycling of the protein is thermodynamically consistent with the energy input . In our simple approach each hydrolysis reaction produces a power stroke , meaning that there are no futile cycles and therefore the chemistry is tightly coupled to the mechanics . We incorporate into our description of myosin Ic as many experimental data as possible . Some transition rates and parameter values have been reported [19 , 20]; an overview of these is given in Table 1 . The force-dependent lifetime of an actin-myosin bond has been determined with an optical trap [20] . Because of the finite time resolution of the experimental apparatus , it is reasonable to assume that the strongly bound states rather than the weakly bound ones dominate the lifetime . We accordingly interpret the reported lifetime as the time that myosin is attached to actin in the strongly bound states . As a consequence , the reported duty ratio r ≃ 0 . 11 and force-free binding time in the strongly bound states tsb ≃ 0 . 213 s provide an estimate of the time that the myosin molecule resides in the weakly bound states , t wb = t sb 1 - r r ≃ 1 . 72 s . ( 30 ) The rate constant , ω 23 0 for phosphate release has been reported for human myosin-IC [82] as ω 23 0 ≃ 1 . 5 s - 1 . ( 31 ) Because of the opaque nomenclature of myosin-I isoforms , human myosin-IC is instead myosin Ie in the nomenclature of the Human Genome Organization [83] . This value is reasonable if the rate-limiting step is phosphate release and the order of magnitude is in agreement with the considerations for myosin Ic given in [19] . We next discuss the binding probability and the free-energy difference associated with the power stroke . Because there are to our knowledge no direct measurements for myosin Ic , we use values reported for myosin II . To estimate the probability π2 that the head binds in state ( 2 ) , we refer to the cycle for rabbit skeletal muscle [2] . The reported numbers suggest a probability π 2 ≃ 0 . 998 , ( 32 ) which implies that myosin starts its cycle predominantly in state ( 2 ) . Using the definition of this binding probability π 2 = ω 12 ω 12 + ω 15 , ( 33 ) we can relate the binding rates to each other as ω 12 = π 2 ω 15 1 - π 2 . ( 34 ) Because both transition rates depend linearly on the actin concentration ( Eqs 23 and 24 ) , the actin concentration cancels and ω ^ 12 = π 2 ω ^ 15 1 - π 2 . ( 35 ) Several studies suggest that a large free-energy difference is associated with the main power stroke [2 , 28 , 32–34 , 38] . Using the value ΔG23 ≃ −15 kBT inferred from fitting a model of the myosin II cycle to experimental data acquired with frog muscle [28] , we obtain the ratio ω 23 0 ω ^ 32 0 [ P i ] = exp ( - Δ G 23 / k B T ) , ( 36 ) which provides the phosphate-binding rate constant ω ^ 32 0 = ω 23 0 [ P i ] exp ( Δ G 23 / k B T ) . ( 37 ) Assuming a phosphate concentration of [Pi] = 1 mM in frog muscle and using Eq 31 , we determine that ω ^ 32 0 ≃ 4 . 5 · 10 - 10 s - 1 μ M - 1 . ( 38 ) A complementary approach is to use the values for the rabbit muscle cycle [2]; we then obtain ω ^ 32 0 ≃ 4 . 6 · 10 - 9 s - 1 μ M - 1 , a value one order of magnitude larger . Because both rate constants are very small compared to the other transition rates of our description , they make no significant difference in our results . In a one-cycle description , the effects of changing the ATP and ADP concentrations are tightly coupled and determined by the magnitudes of the rate constants . If we assume a very fast and irreversible unbinding from the ATP state , ω51 ≫ 1 and ω15 = 0 , and an irreversible Pi release , ω32 = 0 , the average time in the strongly bound states reads t sb = ω 34 + ω 43 + ω 45 ω 34 ω 45 . ( 39 ) In the force-free case the ADP and ATP binding rates are given in Eqs 21 and 22 . Using the equilibrium binding constant KADP to estimate the rate constant for ADP binding as ω ^ 43 0 = ω 34 0 K ADP , ( 40 ) we rewrite Eq 39 as t sb = ω 34 0 ( 1 + [ ADP ] / K ADP ) + ω ^ 45 0 [ ATP ] ω 34 0 ω ^ 45 0 [ ATP ] . ( 41 ) The two rate constants and the equilibrium constant have been determined experimentally as ω 34 0 ≃ 3 . 9 s - 1 , ω ^ 45 0 ≃ 0 . 26 s - 1 μ M - 1 and KADP ≃ 0 . 22 μM [20] . Because of the low ATP-binding rate constant and the small equilibrium constant KADP for ADP release , the lifetime of the strongly bound states is very sensitive to elevated ADP concentrations . Because the experimental findings differ , we resolve this problem in our one-cycle description by using the higher value of KADP ≃ 1 . 8 μM for the equilibrium constant for ADP release [22 , 23] . Although another possibility would be to introduce a multi-cycle description , that strategy increases complexity and the number of unknown parameters . To satisfy thermodynamic consistency , we express the transition rate for ATP release in terms of all the other transition rates as given by the balance condition of Eq 29 , ω 54 = ω 12 ω 23 ω 34 ω 45 ω 51 ω 21 ω 32 ω 43 ω 15 exp ( - ( Δ μ - ( Δ x 1 + Δ x 2 ) F ) / k B T ) . ( 42 ) This transition rate is dependent on force in the same way as ω51 but independent of the nucleotide concentrations , as can be concluded by applying eqs ( 12 ) – ( 16 ) , ( 20 ) – ( 22 ) and ( 27 ) , resulting in ω 54 = ω 12 ω 23 0 ω 34 0 ω ^ 45 0 g ( F ) ω 51 0 ω 21 ω ^ 32 0 ω ^ 43 0 ω 15 K eq ≡ ω 54 0 g ( F ) . ( 43 ) We are left with seven unknown parameter values: the unbinding rates ω51 and ω21 , the binding rate ω15 , the force distribution factors δ1 and δ2 , the characteristic length ξ , and the offset rate ωoff . To estimate these values , we use analytic expressions of the average time spent in the weakly bound states and the effective unbinding rate from the strongly bound states and fit these functions to the experimental data . We derive these analytic expressions in the following section . The probability Pi of finding myosin in one of the states of the cycle is given by the solution to the steady-state master equation 0 = - ω 12 - ω 15 ω 21 0 0 ω 51 ω 12 - ω 21 - ω 23 ω 32 0 0 0 ω 23 - ω 32 - ω 34 ω 43 0 0 0 ω 34 - ω 43 - ω 45 ω 54 ω 15 0 0 ω 45 - ω 54 - ω 51 P 1 P 2 P 3 P 4 P 5 , ( 44 ) and read P 1 ≡ ( ω 23 ω 34 ω 45 ω 51 + ω 21 ( ω 34 ω 45 ω 51 + ω 32 ( ω 43 + ω 45 ) ω 51 + ω 32 ω 43 ω 54 ) / N , ( 45 ) P 2 ≡ ( ω 12 ( ω 34 ω 45 + ω 32 ( ω 43 + ω 45 ) ) ω 51 + ( ω 12 + ω 15 ) ω 32 ω 43 ω 54 ) / N , ( 46 ) P 3 ≡ ( ω 15 ( ω 21 + ω 23 ) ω 43 ω 54 + ω 12 ω 23 ( ω 45 ω 51 + ω 43 ( ω 51 + ω 54 ) ) ) / N , ( 47 ) P 4 ≡ ( ω 15 ( ω 23 ω 34 + ω 21 ( ω 32 + ω 34 ) ) ω 54 + ω 12 ω 23 ω 34 ( ω 51 + ω 54 ) ) / N , ( 48 ) P 5 ≡ ( ω 12 ω 23 ω 34 ω 45 + ω 15 ( ω 21 ω 32 ω 43 + ω 23 ω 34 ω 45 + ω 21 ( ω 32 + ω 34 ) ω 45 ) ) / N , ( 49 ) in which N is determined by the normalization ∑i Pi = 1 . To determine the effective unbinding rate from the strongly bound states , we calculate the average attachment time in the strongly bound states using a framework introduced by Hill [84 , 85] . We promote the detached state ( 1 ) and weakly bound state ( 2 ) to absorbing states by setting the transition rates ω15 , ω23 , ω12 , and ω21 to zero ( Fig 9a ) . The associated effective unbinding rate then becomes the inverse of the average time to absorption for the appropriate initial condition . The basic idea is to use an ensemble average instead of a time average . The dynamics of the correct ensemble is described by a closed diagram in which the absorbing state is eliminated by redirecting the transitions into that state to the starting states weighted with the appropriate starting probabilities [86] . For example , the transition from state ( 3 ) to state ( 2 ) is redirected with the weight 1 − π3 to state ( 5 ) and with weight π3 to state ( 3 ) . The latter transition , a self loop , cancels in a master equation and can therefore be disregarded . This procedure creates a closed diagram ( Fig 9b ) for which the steady-state probability pi of being in state ( i ) is determined from the master equation 0 = - ( 1 - π 3 ) ω 32 - ω 34 ω 43 π 3 ω 51 ω 34 - ω 43 - ω 45 ω 54 ( 1 - π 3 ) ω 32 ω 45 - ω 54 - π 3 ω 51 p 3 p 4 p 5 , ( 50 ) together with the normalization condition ∑pi = 1 . This probability distribution provides the average state occupancy before absorption . The average rate of arrivals at either of the absorbing states is therefore given by the probability current , which is identical to the effective unbinding rate from the strongly bound states t sb - 1 ≡ ω 32 p 3 + ω 51 p 5 . ( 51 ) The probability of starting in state ( 3 ) and not in state ( 5 ) is given by the relative probability current into state ( 3 ) of the complete cycle as π 3 ≡ ω 23 P 2 ω 23 P 2 + ω 15 P 1 , ( 52 ) in which P1 and P2 are given in Eqs 45 and 46 , respectively . For the sake of completeness we give the rather cumbersome expression for the effective unbinding rate from the strongly bound states , t sb − 1 = ( ( ( ω 15 ω 21 + ω 12 ω 23 ) ω 32 ω 43 + ( ω 12 ω 23 ( ω 32 + ω 34 ) + ω 15 ( ω 21 ω 32 + ( ω 21 + ω 23 ) ω 34 ) ) ω 45 ) ω 51 + ( ω 15 ω 21 + ( ω 12 + ω 15 ) ω 23 ) ω 32 ω 43 ω 54 ) / H , ( 53 ) in which H ≡ ω 15 ( ω 21 ( ω 32 ( ω 43 + ω 45 ) + ω 34 ω 45 ) + ω 23 ω 34 ω 45 + ( ω 23 ( ω 34 + ω 43 ) + ω 21 ( ω 32 + ω 34 + ω 43 ) ) ω 54 ) + ω 12 ω 23 ( ( ω 43 + ω 45 ) ω 51 + ω 43 ω 54 + ω 34 ( ω 45 + ω 51 + ω 54 ) ) . ( 54 ) The time that myosin spends in the weakly bound states can be obtained from Eq 3 as t wb = t sb 1 ∑ i = 3 5 P i - 1 , ( 55 ) in which Pi are the steady-state probabilities given in Eqs 47–49 . To determine the time during which myosin is attached to the filament , we promote state ( 1 ) to an absorbing state . The corresponding closed diagram is obtained by redirecting the transition from state ( 2 ) to state ( 1 ) with weight 1 − π2 to state ( 5 ) and the transition from state ( 5 ) to state ( 1 ) with weight π2 to state ( 2 ) . The steady-state probability si of being in state ( i ) for this closed diagram is the solution of the master equation 0 = - ω 23 - ( 1 - π 2 ) ω 21 ω 32 0 π 2 ω 51 ω 23 - ω 32 - ω 34 ω 43 0 0 ω 34 - ω 43 - ω 45 ω 54 ( 1 - π 2 ) ω 21 0 ω 45 - ω 54 - π 2 ω 51 s 2 s 3 s 4 s 5 , ( 56 ) together with the normalization condition ∑si = 1 . This probability distribution gives the probability current into state ( 1 ) , which is identical to the unbinding rate from the actin filament k off ≡ ω 21 s 2 + ω 51 s 5 . ( 57 ) Again , for completeness , we give the unwieldy expression for the unbinding rate from the filament , k off = ( ω 23 ω 34 ω 45 + ω 21 ( ω 34 ω 45 + ω 32 ( ω 43 + ω 45 ) ) ) ω 51 + ω 21 ω 32 ω 43 ω 54 A + B + C , ( 58 ) in which A ≡ π 2 ( ω 34 ω 45 + ω 32 ( ω 43 + ω 45 ) ) ω 51 + ω 32 ω 43 ω 54 , ( 59 ) B ≡ ( 1 - π 2 ) ω 21 ( ω 34 ω 45 + ( ω 34 + ω 43 ) ω 54 + ω 32 ( ω 43 + ω 45 + ω 54 ) ) , ( 60 ) C ≡ ω 23 ( π 2 ( ω 43 + ω 45 ) ω 51 + ω 43 ω 54 + ω 34 ( ω 45 + π 2 ω 51 + ω 54 ) ) , ( 61 ) π 2 ≡ ω 12 / ( ω 12 + ω 15 ) . ( 62 ) Note that π2 is independent of the time toff during which myosin is detached from the filament , and therefore independent of the actin concentration . As a consequence the unbinding rate koff is also independent of the actin concentration . To describe the ensemble of myosins as a Markov chain , we introduce a state space ( Fig 6a ) associated with the number of bound myosin heads [47] . Assuming that the heads bind and unbind independently of one another , transitions between these states can be expressed in terms of the individual binding and unbinding rates kon and koff . A transition from state ( n ) , in which n myosins are bound , to state ( n + 1 ) is associated with the binding rate k on n ≡ ( N - n ) k on . ( 63 ) The reverse transition is described by the unbinding rate k off n ≡ n k off . ( 64 ) We next incorporate an external force F into the description . As a first approximation , we assume that the myosin heads share this load equally , resulting in an effective force F/n exerted on each of the bound myosin heads and in a modified transition rate k off n ≡ n k off ( F / n ) . ( 65 ) With this specific choice of the transition rates we determine the probability Sn of being in state ( n ) from a master equation . The solution for such a finite linear Markov chain is a standard result in stochastic dynamics and can be obtained recursively or by standard methods [47 , 81] . The probability of being in state ( n ) is S n = S 0 ∏ i = 0 n - 1 k on i k off i + 1 , ( 66 ) in which S0 is determined from the normalization ∑Si = 1 as S 0 = 1 + ∑ n = 0 N - 1 ∏ i = 0 n k on i k off i + 1 - 1 . ( 67 ) From this probability distribution we obtain the average number of bound myosin heads as n = ∑ n = 0 N n S n . ( 68 ) We describe the dynamics of a myosin head with the five state chemomechanical cycle that we developed in this study . Each myosin head is coupled with a spring to a rigid common structure . The extension of this spring determines the force that is exerted on the myosin molecule and thus all transition rates of its cycle . We assume that a myosin head binds without tension to the actin filament and when it proceeds through its cycle the power-stroke transitions stretches the spring . At each time step we determine the extensions of all myosin springs from Newton’s law of force balance and adjust all transition rates accordingly . For the Monte Carlo simulations we use a Gillespie algorithm which is a standard method to simulate multi-component stochastic reactions and used for elastically coupled motor molecules [35 , 78 , 79 , 88] . After disregarding the transient behavior of our simulations , we determine the average number of myosins bound to actin from a time average . We consider all myosin molecules in state ( 2 ) to ( 5 ) bound to actin . | Myosin molecules are biological nanomachines that transduce chemical energy into mechanical work and thus produce directed motion in living cells . These molecules proceed through cyclic reactions in which they change their conformational states upon the binding and release of nucleotides while attaching to and detaching from filaments . The myosin family consists of many distinct members with diverse functions such as muscle contraction , cargo transport , cell migration , and sensory adaptation . How these functions emerge from the biophysical properties of the individual molecules is an open question . We present an approach that integrates recent findings from single-molecule experiments into a thermodynamically consistent description of myosin Ic and demonstrate how the specific parameter values of the cycle result in a distinct function . The free variables of our description are the chemical input and external force , both of which are experimentally accessible and define the cellular environment in which these proteins function . We use this description to predict the elastic properties of an ensemble of molecules and discuss the implications for myosin Ic’s function in the inner ear as a tension regulator mediating adaptation , a hallmark of biological sensory systems . In this situation myosin molecules cooperate in an intermediate regime , neither as a large ensemble as in muscle nor as a single or a few molecules as in intracellular transport . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"stiffness",
"cell",
"motility",
"mechanical",
"properties",
"actin",
"filaments",
"classical",
"mechanics",
"chemical",
"compounds",
"phosphates",
"nucleotides",
"mechanical",
"energy",
"molecular",
"motors",
"actin",
"motors",
"mathematics",
"statistics",
"(mathematics)",... | 2017 | Chemomechanical regulation of myosin Ic cross-bridges: Deducing the elastic properties of an ensemble from single-molecule mechanisms |
The dendritic spines of pyramidal neurons are the targets of most excitatory synapses in the cerebral cortex . They have a wide variety of morphologies , and their morphology appears to be critical from the functional point of view . To further characterize dendritic spine geometry , we used in this paper over 7 , 000 individually 3D reconstructed dendritic spines from human cortical pyramidal neurons to group dendritic spines using model-based clustering . This approach uncovered six separate groups of human dendritic spines . To better understand the differences between these groups , the discriminative characteristics of each group were identified as a set of rules . Model-based clustering was also useful for simulating accurate 3D virtual representations of spines that matched the morphological definitions of each cluster . This mathematical approach could provide a useful tool for theoretical predictions on the functional features of human pyramidal neurons based on the morphology of dendritic spines .
It is known that the dendritic spines ( for simplicity’s sake , spines ) of pyramidal neurons are the targets of most excitatory synapses in the cerebral cortex [1] . Numerous studies suggest that spine shape could determine their synaptic strength and learning rules and is also related to the storage and integration of excitatory synaptic inputs in pyramidal neurons [2] . Quantitative analyses have demonstrated strong correlations between spine morphological variables and synaptic structure . Specifically , the spine head volume in the neocortex is correlated with the area of the postsynaptic density ( PSD ) [3] . Both parameters are highly variable across spines . Interestingly , however , the spine head volume ( like the total spine volume ) is positively correlated with the PSD area , and there is a remarkably small variance . Moreover , PSD area is correlated with the number of presynaptic vesicles , the number of postsynaptic receptors and the readily-releasable pool of transmitters . By contrast , the length and diameter of the spine neck is proportional to the extent to which the spine is biochemically and electrically isolated from its parent dendrite [4–8] . Also , it has been shown that larger spines can generate larger synaptic currents than smaller spines [9] . Furthermore , dendritic spines are dynamic structures with volume fluctuations that appear to have important implications for cognition and memory [10–13] . Therefore , spine morphology appears to be critical from the functional point of view ( for a review , see [14] ) . There are a wide variety of spine morphologies , especially in the human cortex [15] . While many different classifications of spines have been proposed on the basis of their morphological characteristics , the most widely used was proposed by Peters and Kaiserman-Abramof [16] which groups spines into three basic categories—thin , mushroom and stubby spines—and an additional category—filopodia . However , it has also been argued that the large diversity of spine sizes reflects a continuum of morphologies rather than the existence of discrete groups [3] . Automatic clustering techniques over 2D spine representations have recently been used [17 , 18] to address this argument with the aim of avoiding the subjectivity and bias involved in manual analysis . Both studies consider that some spines cannot be clearly assigned to one of Peters and Kaiserman-Abramof’s classes because these spines are transitions between shapes . However , the geometry of spines can be more accurately determined by means of 3D reconstructions , since many morphological features are not taken into account in 2D . Ideally , 3D reconstruction using electron microscopy serial sections is the gold standard to obtain accurate estimations of the geometry of spines . However , a relatively low number of spines ( at best in the order of a few hundred ) can be reconstructed in 3D using electron microscopy in a reasonable time period , and these reconstructions can only be carried out in small segments of the dendritic arbor of the neurons . Furthermore , the quality of electron microscopy when using human brain tissue is usually suboptimal due to technical constraints . On the contrary , fluorescent labeling of neurons and the use of high power reconstruction with confocal microscopy ( or other techniques ) allow the visualization of thousands of spines with high quality along the dendritic arbor ( apical and basal dendrites ) . Thus , in this study , we used a large , quantitative database of completely 3D-reconstructed spines ( 7 , 916 ) of human cortical pyramidal neurons—using intracellular injections of Lucifer Yellow in fixed tissue—to further characterize spine geometry [15] . Here we proposed a new set of 54 features . They were selected so as to unambiguously approximate the 3D shape of spines , enabling 3D simulation of spines . A probabilistic clustering grouped the 3D reconstructed human spines according to the selected set of morphology-based features . The best number of groups for probabilistic clustering based on the Bayesian information criterion was six groups of human spines . The interpretation of the clusters in terms of their most discriminative characteristics relied on the rules generated automatically by a rule induction algorithm . Since previous studies have shown that there are selective changes in dendritic and spine parameters with aging and dendritic compartments [15 , 19–21] , we also explored the distributions of the groups according to dendritic compartment , age and distance from soma to further characterize possible variations according to these parameters . Finally , we present a stochastic method designed to simulate biologically feasible spines according to the probabilities defined by the clustering model . We introduce a procedure to shape simulated spines generating their 3D representations . To the best of our knowledge , this is the first attempt to fully characterize , model and simulate 3D spines .
We used a set of 7 , 916 3D reconstructed individual spines along the apical and basal dendrites of layer III pyramidal neurons in the cingulate cortex of two individuals ( aged 40 [C40] and 85 [C85] years ) ( Fig 1 ) . For each individual spine , a particular threshold was selected to constitute a solid surface that exactly matched the contour of each spine . In many cases , it was necessary to use several surfaces of different intensity thresholds to capture the complete morphology of a spine [15] . In such cases , spines were usually fragmented or detached from their parent dendrite ( Fig 2A and 2B ) due to the diffraction limitation of confocal microscopy . Therefore , they had to be repaired by means of a novel semi-supervised mesh processing algorithm ( see Materials and Methods ) which generated a new dataset of corrected spines . Those spines that were extremely fragmented , far removed from the dendrite or significantly deferred from their original shape were discarded . As a result , the original set of 7 , 916 spines yielded 7 , 297 ( 92 . 18% ) spines . The number and percentage of spines after repair by their dendritic compartment and age can be found as S1 Table . For the repair process , the insertion point of each spine was manually marked , approximately at the center of the created spine surface side that was in contact with the dendritic shaft . In those cases where the created spine surface did not reach the dendritic shaft , the insertion point was placed directly on the dendritic shaft where the spine emerged from the shaft , while rotating the image in 3D ( Fig 1G ) . The insertion point was useful for repairing the detached spines and computing a multiresolutional Reeb graph for feature extraction . The characterization of spines was addressed by dividing the surface of the spine into regions according to a multiresolutional Reeb graph ( Fig 2C–2F and Materials and Methods ) . Thus , regions provided local information on the spine topology while the combination of all regions gave global details of the morphology . Major morphological aspects like length , width , size or curvature were measured for each region to build a set of 36 spine features ( see Materials and methods ) . This set was complemented with 18 features , like growth direction for example . These features were included to achieve an unambiguous representation of the spine morphology . The complete set of 54 features unambiguously describes the position and orientation of all the ellipses that characterize the geometry of a spine . The software to compute the features can be found at https://github . com/ComputationalIntelligenceGroup/3DSpineMFE . To find groups of spines , we applied a model-based clustering approach which assigned spines to six clusters according to the Bayesian information criterion ( BIC ) ( Fig 3A and Materials and Methods ) . Our approach , based on probabilistic clustering , assigned a probability distribution ( p1 , … , p6 ) of belonging to each of the six clusters to each spine , where pi is the probability of belonging to cluster ( pi ∈ [0 , 1] , ∑ipi = 1 ) . Furthermore , we counted the number of spines whose maximum probability , p* = max{p1 , … , p6} , was lower than a given threshold ( Table 1 ) . We found that the membership probability of most of the spines was greater than 0 . 99 and clearly belonged to a cluster , whereas a very small number were more scattered and , consequently , their membership was not so clear . Therefore , we can conclude that with this set of features most of spines had very high membership probabilities . To gain a deeper insight into the characterization of each group unveiled by the probabilistic clustering , we identified the most representative features for each cluster . The process was based on the generation of classification rules according to the RIPPER algorithm ( see Materials and Methods ) . Each spine was attributed to its most probable cluster . Then , the RIPPER algorithm generated discriminative rules for each cluster , turning the problem into a binary supervised classification problem which pitched each cluster label against the rest . We forced the algorithm to generate a unique rule in order to improve our understanding of the differences between clusters . However , a single rule cannot be regarded as enough to characterize all the spines within a cluster because it is unable to capture all the relations between the variables defined by the model-based clustering . The result was that each cluster was characterized by only one , two or three observable features ( Fig 4 ) . The discriminative rules are available in S1 Text . An example of representative spines of the six clusters is shown in Fig 3B . The rules generated by RIPPER when it comes to classify the spines according to their cluster label , with their accuracy between parentheses , may be summarized as: The diversity of morphologies within a cluster was estimated by computing the total variance for each cluster . Fig 3D shows that Cluster 2 has the lowest total variance , denoting similarity among its spines , whereas variance in Cluster 6 has the highest total variance , suggesting more heterogeneity . To improve cluster visualization and interpretation , the distances between the membership probabilities ( p1 , … , p6 ) of the spines in a 6D space were projected to 2D according to multidimensional scaling ( see Fig 3C and Materials and Methods ) . Spines were colored in line with their probability of belonging to each cluster . Accordingly , “intermediate” spines whose membership probabilities were distributed evenly across several clusters have a mixture of colors . In this representation , we find that most of the points are clearly assigned to a cluster , as suggested by the results reported in Table 1 . Clusters 1 and 6 are outstanding examples of a clearly defined cluster , since they are quite isolated and , consequently , easy to discriminate from the other clusters . However , clusters like 3 and 4 are quite closely related . This tallies with the results reported in Table 1 , where the clusters identified as being clearly separate had a higher threshold than highly related clusters that needed a lower threshold for all their spines to be crisply assigned . To quantify the distance of the points observed by multidimensional scaling , we measured the overlap between clusters ( see Materials and Methods ) . Note that clearly defined clusters should not overlap . The results reported in S2 Table support the interpretation of multidimensional scaling . By selecting p* of each spine , the spines can be crisply assigned to a unique cluster yielding the distribution shown in Fig 5A . This bar chart represents the percentage of spines that belong to each cluster . To gain a deeper insight , we analyzed how it changes the cluster distribution of the whole population of spines ( Fig 5A ) when a dendritic compartment ( apical/basal ) , an age ( 40/85 ) or a combination of both ( Fig 5B–5D ) is selected . The study of the cluster distribution of the spines according to their dendritic compartment unveiled that the proportion of spines in Clusters 3 , 5 and 6 increase for apical dendrites and diminish for basal dendrites compared with those observed in Fig 5A , whereas the major increment for basal dendrites and decrement for apical dendrites is yielded in Cluster 1 . In order to evaluate these differences , we used χ2 hypothesis testing , that is , we tested whether the cluster distribution is independent of the dendritic compartment ( null hypothesis H0 ) . The hypothesis test returned a p-value lower than 3 . 80 × 10−34 thereby the null hypothesis H0 was rejected . The same process as applied for dendritic compartment was repeated for age . Fig 5C shows that Cluster 2 is overrepresented in C40 and Clusters 4 and 6 in C85 . On the contrary , the major decreases occur in Cluster 2 in C85 and Clusters 4 and 6 in C40 . To test if cluster distribution is independent of age , we tested the hypothesis again . Results rejected the null hypothesis ( the p-value was lower than 3 . 73 × 10−06 ) . Furthermore , we run the clustering algorithm for each subject ( C40 and C85 ) to study their distribution independently . As a result , six clusters emerged from C40 spines mostly matching those obtained for the complete population of spines and an additional one of 36 spines that only grouped spines from Clusters 5 and 6 . Clustering of C85 spines generated five clusters showing similar results to those achieved for the global population but combining spines from Cluster 2 with Cluster 4 in a unique cluster and tending to include some spines of original Cluster 6 into Cluster 5 . We then tested the cluster distribution and the combination of dendritic compartment and age for independence ( Fig 5D ) . Fig 5D shows that there is an increase of Clusters 3 and 5 for C40 apical dendrites; Clusters 3 , 5 and 6 for C85 apical dendrites; Clusters 1 and 2 for C40 basal dendrites and Clusters 1 and 4 for C85 basal dendrites with respect to the distribution observed for the whole population of spines . Additionally , from Fig 5D it can be observed that Clusters 1 and 4 are underrepresented in C40 apical dendrites; Clusters 1 and 2 in C85 apical dendrites; Clusters 5 and 6 in C40 basal dendrites and Clusters 2 , 3 and 4 in C85 basal dendrites . The null hypothesis was rejected ( p-value ≈ 4 . 11 × 10−36 ) . Hence we can reject independence between cluster distribution and dendritic compartment combined with age . In spite of the fact that the null hypothesis was rejected for all the above cases , Fig 5B–5D show that the discrepancies in the distributions are confined to only a few clusters and are not evenly spread . With the aim of pinpointing those clusters that exhibit significant differences , each one was analyzed individually . A Pearson’s χ2 test was performed cluster by cluster to check if the proportion of spines in each individual cluster was independent of the dendritic compartment , age and combination of both . The outcome of the tests is shown in Table 2 . Results confirm that only some clusters vary significantly depending on dendritic compartment , age or combination of both and indicate how strongly the hypothesis was rejected for each cluster . An example can be found for age where the null hypothesis was only rejected for Clusters 2 , 4 and 6 , showing that they are the only clusters whose distribution varies significantly with age . Furthermore , we evaluated the cluster distribution according to the distance from soma ( Fig 5E ) . The number of spines was categorized in 50 μm long sections , from 0 μm ( the beginning of the dendrite ) to 300 μm . A χ2 hypothesis test was applied in order to test the independence between cluster distribution and distance from soma . The outcome rejected the null hypothesis H0 ( p-value ≈ 8 . 00 × 10−23 ) . The number of spines assigned to each section is specified in S3 Table . Briefly , Fig 5E shows that there is a predominance of Clusters 1 and 2 at proximal distances ( 0–50 μm ) whereas Clusters 1 and 4 show a higher percentage than expected at longer distances . Model-based clustering describes the probability distributions governing each cluster . Given a cluster , a spine is simulated sampling the values for the 54 features from its probability distribution ( see Materials and Methods ) . This set of features unambiguously specifies the position and orientation of ellipses that define the skeleton of a simulated spine ( Fig 6A ) . The simulated spine is represented in 3D by surfacing the skeleton ( Fig 6B ) . However , simulated spines have an artificial appearance because the regions delimited by the ellipses are clearly distinguishable between them ( Fig 6C ) . A more accurate morphology for the simulated spine is generated by smoothing the surface ( Fig 6D ) . Examples of simulations of each cluster can be found in Fig 6E . R code , model and dataset to perform clustering and simulation of dendritic spines can be downloaded from https://github . com/sergioluengosanchez/spineSimulation . To be useful for future research , simulated spines must be geometrically equivalent to real spines . Thus , simulated and real must be indistinguishable . To test for equivalence , we state a supervised classification problem within each cluster , where the possible labels are “simulated” vs . “real” . Hence , if both groups were indistinguishable , a classifier would perform badly , having a classification accuracy of around 50% . As a result we found that both groups of spines are almost indistinguishable ( accuracy being around 60% ) , with the exception of cluster 1 ( 80% ) , where the size of simulated spines is usually somewhat larger than real spines .
This study illustrates the geometrical clustering results from over 7 , 000 complete manual 3D reconstructions of human cortical pyramidal neuron spines . Specifically , we uncovered six different classes of human spines according to a particular set of features . Additionally , we found that particular clusters were predominant in different dendritic compartments , ages and distances from soma . Furthermore , we created 3D virtual representations of spines that matched the morphological definitions of each cluster . To the best of our knowledge , this is the first time that such a large dataset of individual manually 3D reconstructed spines from identified human pyramidal neurons is used to automatically generate objective morphological clusters with a probabilistic model . Technically , serial electron microscopy is the technique of choice to obtain highly accurate measurements of the dendritic spine structure . However , it is very time-consuming and difficult , which makes it challenging to obtain large numbers of measurements . Even using high throughput 3D reconstruction of identified dendritic spines by means of automatic electron microscope techniques such as FIB/SEM technology ( combined use of focused ion beam milling [FIB] and scanning electron microscopy [SEM] ) , the number of reconstructed spines is relatively low . For example , FIB/SEM technology has permitted the full 3D reconstruction of up to 248 spines and their synaptic inputs in the adult-generated granule cells in mice [22] , which represents a major achievement in the field . Light microscopic techniques , although limited by the lower level of resolution , remain the method of choice to obtain large-scale spatial information regarding the number and distribution of dendritic spines along the dendrites ( in the order of several thousands of spines ) . Nevertheless , light microscopic studies normally estimate dendritic spine volumes from measurement of the spine head volumes , whereas spine necks are usually not included , due to the lack of software tools to reconstruct these structures accurately and because some of the spine necks have spatial dimensions of around 50–200 nm and , therefore , are not resolvable by confocal microscopy . Moreover , as discussed by Tønnesen and Nägerl [14] , image projection artifacts and limited spatial resolution mask short spine necks , leading to the false identification of stubby spines . In addition , it is difficult to distinguish the border between the head and the neck in many cases . Thus , in the present study , we used 3D reconstructed dendritic spine morphology using commercially available module software ( Imaris surface ) , which allowed us to create our own protocol to accurately represent the morphology of the spine within the limits of light microscopy ( see [15] ) . Model-based clustering methods used in this study yield six clusters based on their BIC value ( see Fig 3A ) . This criterion resulted in high cluster membership probabilities for this set of features . These included measurements of major morphological aspects like length , width or size of the spine but also other aspects such as curvature . Thus , these and previous results [3 , 15] , where only their volume and length were measured , are not comparable . Interestingly , we observed that there are particular clusters of spines that are proportionally highly represented in a particular dendritic compartment/age combination . Specifically , basal dendrites contained a higher proportion of the small Cluster 1 spines ( Fig 3B ) , whereas apical dendrites contained a higher proportion of the medium/large Clusters 3 , 5 and 6 spines . These differences would imply that their functional properties should be expected to be different in the two dendritic compartments [2] . Regarding individuals , Cluster 2 spines accounted for a higher percentage in the younger individual , whereas Clusters 4 and 6 of bigger spines had higher values than the mean percentage in the older individual . Since small spines have been reported to be preferential sites for long-term potentiation induction and large spines might represent physical traces of long-term memory [9 , 13] , the results suggest that the younger individual has a higher potential for plasticity than in the aged case . The dendritic compartment/age combination results also agreed with our previously reported study [15] that found that apical dendrites have longer spines than basal dendrites , and younger basal dendrites are significantly smaller than aged basal dendrites . For instance , small and short spines of aged basal dendrites and long spines of apical dendrites were lost . Regarding the distance from soma , there is a higher predominance of the small Clusters 1 and 2 spines than expected at proximal distances ( 0–50μm ) and the small Cluster 1 spines at distal distances . Also , distal distances showed a higher percentage of the medium-sized Cluster 4 spines than expected . Since variations in spine geometry reflect different functional properties of the spine , this particular distribution of spines might be related to the morphofunctional compartmentalization of the dendrites along the length of the dendritic pyramidal neurons . For example , it has been reported that different domains of the basal dendritic arbors of pyramidal cells have different properties with respect to afferent connectivity , plasticity and integration rules [15 , 19 , 23–26] . Thus , these results may be a reflection of a functional dendritic organization based on spine geometry . Using the technique of model-based clustering described in this study , we were able to simulate accurate spines from human pyramidal neurons . This is important for three main reasons . First , it is not necessary to store large volumes of data because all the information is summarized in the mathematical model . Second , spines are known to be dynamic structures ( see [27] for a recent review ) , and changes in spine morphology have important functional implications potentially affecting not only the storage and integration of excitatory inputs in pyramidal neurons but also mediating evoked and experience-dependent synaptic plasticity . This , in turn , has major repercussions on cognition and memory [13 , 28–32] . Thus , it is necessary to link the structural data with theoretical studies and physiological data on spines in order to interpret and make the geometrical data on spines more meaningful . Functional modeling of spines is commonly carried out according to their values of surface area , spine maximum diameter , spine neck diameter , spine length , and spine neck length . Since each cluster contains a spine population with a range of morphological features , it is necessary to model all of these morphological variations within each cluster in order to compare the possible functional differences between the clusters found in the present study . Third , one of the major goals in neuroscience is to simulate human brain neuronal circuitry based on data-driven models because ethical limitations prevent all of the necessary datasets from being acquired directly from human brains . Therefore , the implementation of this mathematical model of spines of human pyramidal cells in current models of pyramidal neurons is a potentially useful tool for translating neuronal circuitry components from experimental animals to human brain circuits . The simulation of the spines in this study represents a mathematical model that could be implemented in pyramidal cell models [33] in order to present the data in a form that can be used to reason , make predictions and suggest new hypotheses of the functional organization of the pyramidal neurons . Finally , spine heads and necks of human pyramidal cells are significantly larger in terms of their area and longer , respectively , than mouse spines [34] . Therefore , it would be interesting to compare human and non-human spines using the present model-based clustering to ascertain whether the clusters that appear are the same or different in other species , or whether there are differences between different cortical areas .
Brain samples were obtained from the Institute of Neuropathology Brain Bank , a branch of the HUB-ICO-IDIBELL Biobank and member of the Spanish Biobank network ( RETIC Biobank ) of the Institute of Health Carlos III , following the guidelines of Spanish legislation ( real Decreto 1716/2011 ) and the approval of the local ethics committee , and in accordance with recently published criteria for sample quality ( PMID: 25113170 ) . A set of 7 , 916 individually 3D reconstructed spines from layer III pyramidal neurons from the cingular cortex of two human males ( aged 40 [C40] and 85 [C85] ) were used for analyses . These cases were used as controls in a previous study unrelated to the present investigation that was dealing with Alzheimer’s disease [35] . The cause of death was traffic accident ( case C40 ) and pneumonia plus interstitial pneumonitis ( aged case , C85 ) . The tissue ( kindly supplied by Dr I . Ferrer , Instituto de Neuropatología , Servicio de Anatomía Patológica , IDIBELL-Hospital Universitario de Bellvitge , Barcelona , Spain ) was obtained at autopsy ( 2–3 h post-mortem ) . The brains were immediately immersed in cold 4% paraformaldehyde in 0 . 1 M phosphate buffer , pH 7 . 4 ( PB ) and sectioned into 1 . 5-cm-thick coronal slices . Small blocks of the cortex ( 15 × 10 × 10 mm ) were then transferred to a second solution of 4% paraformaldehyde in PB for 24 h at 4°C . After fixation , vibratome sections ( 250 μm ) from the anterior cingular gyri ( Brodmann's area 24;[36] ) were obtained with a Vibratome and labelled with 4 , 6 diamino-2-phenylindole ( DAPI; Sigma , St Louis , MO ) to identify cell bodies . Pyramidal cells were then individually injected with Lucifer Yellow ( LY; 8% in 0 . 1 M Tris buffer , pH 7 . 4 ) , in cytoarchitectonically identified layer III of the anterior cingular gyrus . LY was applied to each injected cell by continuous current until the distal tips of each cell fluoresced brightly , indicating that the dendrites were completely filled and ensuring that the fluorescence did not diminish at a distance from the soma ( for a detailed methodology of the cell injections , see [37–39] ) . Apical and basal dendrites were then scanned at high magnification by confocal microscopy and reconstructed in three dimensions using a methodology previously described in detail [15] . Sections were imaged with a Leica TCS 4D confocal scanning laser attached to a Leitz DMIRB fluorescence microscope . Fluorescent labeling profiles were imaged , using an excitation wavelength of 491 nm to visualize Alexa fluor 488 . Consecutive stacks of images ( 3 ± 0 . 6 stacks per dendrite; 52 ± 17 images ) were acquired at high magnification ( ×63 glycerol; voxel size , 0 . 075 × 0 . 075 × 0 . 28 μm3 ) to capture the full dendritic depth , length , and width of basal dendrites , each originating from a different pyramidal neuron ( 10 per case ) . The voxel size was calculated to acquire images at the highest resolution possible for the microscope which made it possible to capture the traditional diffraction limits of fluorescence microscopy ( approximately 200 nanometers ) . Regarding apical dendrites , the main apical dendrite was scanned , at a distance of 100 μm from the soma up to 200 μm ( 8 dendrites per case ) . Thus , no apical dendritic tufts were included in the analyses . As a result , a dataset containing 7 , 916 3D reconstructed individual spines along the apical and basal dendrites was built ( Fig 1 ) . We addressed the task of repairing spines by means of a semi-automatic mesh processing algorithm ( Fig 2A ) . The procedure starts by identifying fragmented spines . A spine is fragmented if there is no path between every pair of vertices on the surface of the 3D mesh , and all the vertices belong to a closed mesh . If this is the case , fragmentation is repaired by applying a closing morphological operator to each spine individually . This operator requires a binary image as input , and therefore 3D meshes are voxelized [40] . As a result of applying the closing operator to each voxel of the volumetric spine using a sphere as a structuring element , fragments are joined to form a single body . The marching cubes algorithm [41] recovers the mesh representation from the volumetric image of the repaired spine . The repair process was continued by connecting spines to dendrites by means of spine path reconstruction ( Fig 2B ) . Several points were created to attach the spine to the dendrite , using the measurement point tool in Imaris software . These are considered to be the spine insertion points . Spine reconstruction was applied to any spines whose insertion point was not on the surface of the mesh . This step in the repair process consists of filling the gap between the closest vertex of the spine to the insertion point and the insertion point according to an iterative process that grew the missing base of the spine . Specifically , each detached spine was oriented so that both points bounding the gap were aligned with the z-axis . Then , the mesh of each spine was voxelized . Each voxel slice perpendicular to the z-axis between the spine and the insertion point was filled with the result of applying a 2D Gaussian filter to the slice immediately above . The mesh representation of the completely repaired spine was recovered from the volumetric representation by the marching cubes algorithm . Finally , we smoothed the triangular mesh with a curvature flow technique [42] . As result of this process a new dataset of corrected spines is obtained . Given 3D meshes representing the surface of the spines , our goal was to extract a set of morphological features providing enough information to reconstruct an approximation of their original shapes . Our work was partially inspired by the concept of multiresolutional Reeb graph ( MRG ) [43] and its particular implementation in [44] , a technique that constructs a graph from a 3D geometric model to describe its topology ( Fig 2C–2F ) . This approach partitions a triangular mesh into regions based on the value of a function μ ( ⋅ ) . This function should preferably be the geodesic distance , i . e . , the shortest path between two points of the mesh along the surface because it is invariant to translation and rotation and is robust against mesh simplification and subdivision . We computed geodesic distance from the insertion point of the spine to each vertex of the mesh ( Fig 2C ) . The domain of μ ( ⋅ ) was divided into K = 7 equal length intervals , where ri indicates the beginning and the end of each region such that r0=[0 , 1Kα] , r1= ( 1Kα , 2Kα] , … , rK−1= ( K−1Kα , α] , where α is max μ ( ⋅ ) . This means that each of the vertices in the triangular mesh was allocated to a particular region depending on its evaluation function μ ( ⋅ ) ( Fig 2D–2E ) . At each region i , the curves defining the top and bottom bounds were assumed to be ellipses contained in the best fitting plane computed using principal component analysis . We denote Ti and Bi the top and the bottom ellipses of each region i respectively . Thus , each region provided a local description of the morphology while the combination of the information of all regions represented a global characterization of the spine . Representing a spine as a set of ellipses allows us to capture its most relevant morphological aspects while spurious details are avoided . The proposed set of 54 features must unambiguously describe the placement of the ellipses . To achieve this , at each region i a set of features was computed according to their ellipses Ti and Bi . Since the surface was required to be continuous coherence constraints were imposed on adjacent regions: ∀i , 1<i<K+1 , BiR=Ti−1R , Bir=Ti−1r . Thus , to satisfy the previous condition the following features were considered to characterize the spine ( Fig 7 ) : Height ( |hi| ) ) : This variable measures the length of the vector hi between the centroids of two consecutive ellipses . The higher the value of this variable , the longer the spine in that region . Length of major axis of ellipse ( BiR ) : Low values mean that spine is thin around BiR . Length of minor axis of ellipse ( Bir ) : It gives information about the squishiness of the spine when it is compared with BiR . If BiR and Bir have the same values the ellipse is in fact a circle while when Bir gets smaller the ellipse becomes more squished . Ratio between sections ( φij ) : It is the ratio between the area of the ellipses j and i , i . e . , φij=πBjRBjrπBiRBir . If it is higher than 1 it means that ellipse j is bigger than ellipse i . When values are between 0 and 1 it means that ellipse i is bigger than ellipse j . It can be interpreted as the widening or narrowing along the spine . We compute φ24 , φ26 and φ46 . Growing direction of the spine: The vector between ellipse centroids hi defines a direction which can be expressed in spherical coordinates , i . e . , an azimuth angle ϕi and an elevation angle θi . Ellipse direction: It is the direction of the perpendicular vector to ellipse Bi . It is obtained from BiR|BiR|×Bir|Bir| ( vectorial product ) . It is expressed in spherical coordinates as: Volume ( V ) : It is the total volume of the spine . Volume of each region ( Vi ) : It an approximation of the volume between two consecutive ellipses . It is computed from the convex hull of Ti and Bi . By generating a surface between each pair of ellipses , we get an approximation of the shape of the spine ( Fig 6B and 6C ) . Surfaces between regions can be computed by the method that we propose in the spine simulation section under Materials and Methods . Model-based clustering [45] is a probabilistic approach that assumes that data were generated by a statistical model . Its goal is to recover that model from the observed data . Finite mixture models provide a formal setting for model-based clustering . In finite mixture models , each cluster is represented by a probability distribution . The linear superposition of such distributions generates the finite mixture M . The fit of the model to the data depends on a set of parameters that are usually optimized by means of maximum-likelihood estimation . This estimation method finds the set of parameters θ that maximize the observed data likelihood , i . e . , maxθ f ( x1 , … , xN|θ ) , where x1 , … , xN are a data sample of size N . Then , we assume that the vector of features describing the spines is distributed according to a Gaussian mixture , as it can approximate any multivariate density given enough components [46] . Thus , the density is f ( x1 , … , xN|θ ) =∑c=1CπcN ( x|μc , Σc ) , where N denotes a multivariate normal distribution with prior probability πc , mean vector μc and variance-covariance matrix Σc , C is the total number of clusters and each cluster is denoted by c . Thus , the goal is to get the values for the set of parameters θ = {πc , μc , Σc}c that maximize the likelihood . This was approximated using an iterative two-step procedure called expectation-maximization algorithm [47] . To choose the most suitable number of clusters , we computed the Bayesian information criterion ( BIC ) score [48] for different values of C . BIC is a measure that adds a penalty to the log-likelihood of the model based on the number of model parameters . Therefore , it is used to select the best parameterization and number of clusters by trying to avoid the selection of overly complex models . After the clustering process , each spine has a certain probability of belonging to each cluster ( “soft” clustering ) . We used mclust , a contributed R package [49] , for model-based clustering and density estimation . In order to shed light on the features that characterize each cluster , we generated classification rules according to the RIPPER algorithm [50] . The spines were crisply assigned to a unique cluster by selecting the most probable cluster for each spine . Then , RIPPER compared each cluster against the others , generating discriminative rules . SMOTE [51] was applied as a pre-processing step before running RIPPER to avoid bias and deal with the unbalanced distribution of instances arising from data splitting ( one cluster versus the rest ) . SMOTE is a technique for adjusting the class distribution so that the set of observations of the least represented class is resampled . We also forced RIPPER to select a unique rule to improve the interpretability of each cluster , highlighting its most discriminative features . We used the RIPPER implementation included in the collection of algorithms of Weka , a software for machine learning tasks [52] . To make the clustering results graphically interpretable , we applied multidimensional scaling to represent the distance of spines to clusters according to their membership probability ( Fig 3C ) . To achieve this goal , distances between each pair of multivariate Gaussians defined by the clusters were calculated according to the Bhattacharyya distance [53] . Based on this measure , we were then able to project the above distances , originally in a 6D space , onto a 2D space using multidimensional scaling [54] . Thus , spines were placed in this space in proportion to the probability of their belonging to each cluster . Clustering aims to group similar instances and separate dissimilar instances . Therefore , method performance depends on whether the clusters overlap with each other . Non overlapping clusters are easily discovered . However , clustering algorithms have trouble separating overlapped clusters because instances cannot be clearly assigned to clusters . Hence , overlapping was understood according to [55 , 56] as the probability of misclassifying an instance from cluster i in a cluster j . Thus , the probability ωj|i of misclassifying an instance of the i-th component to the j-th component was computed as ωj|i=P[πiN ( x|μi , Σi ) <πjN ( x|μj , Σj ) |x∼N ( μi , Σi ) ] . The simulation process aimed at achieving accurate 3D representations of spines generated by the computer . This process is divided into two main phases . First , we sampled new instances from the mixture model of multivariate Gaussians . As a result of sampling , we got a dataset where each instance consisted of a vector with 54 feature values defined by a multiresolution Reeb graph . Second , we generated a 3D representation for each instance . From the set of features of a sampled spine , we built a skeleton composed of the ellipses establishing the beginning and end of regions ( Fig 6A ) . Because all the ellipses had the same number of points , each pair of consecutive ellipses was easily triangulated to obtain a closed mesh ( Fig 6B and 6C ) . Although this mesh is a 3D spine , ellipses are clearly distinguishable . We improved this result by smoothing the surface with the Loop’s subdivision algorithm [57] . Thus , we obtained a more accurate 3D representation of the spine ( Fig 6D ) . To objectively validate the realism of the simulated spines , we used a binary classifier , specifically the RIPPER algorithm . First , for each cluster , we sampled from the probability distribution of each cluster the same number of simulated spines as real spines are . Second , we combined these with real spines to generate a dataset for each cluster . Third , we applied the RIPPER algorithm with ten-fold cross-validation [58] over the datasets to discriminate between real and simulated spines . This process yields classifier accuracy , which can be regarded as the degree of realism . | Dendritic spines of pyramidal neurons are the targets of most excitatory synapses in the cerebral cortex and their morphology appears to be critical from the functional point of view . Thus , characterizing this morphology is necessary to link structural and functional spine data and thus interpret and make them more meaningful . We have used a large database of more than 7 , 000 individually 3D reconstructed dendritic spines from human cortical pyramidal neurons that is first transformed into a set of 54 quantitative features characterizing spine geometry mathematically . The resulting data set is grouped into spine clusters based on a probabilistic model with Gaussian finite mixtures . We uncover six groups of spines whose discriminative characteristics are identified with machine learning methods as a set of rules . The clustering model allows us to simulate accurate spines from human pyramidal neurons to suggest new hypotheses of the functional organization of these cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"ellipses",
"dendritic",
"structure",
"light",
"microscopy",
"geometry",
"neuroscience",
"probability",
"distribution",
"mathematics",
"ganglion",
"cells",
"microscopy",
"neuronal",
"dendrites",
"research",
"and",
"analysis",
"methods",
"animal",
"cells",
"fluorescence",
... | 2018 | 3D morphology-based clustering and simulation of human pyramidal cell dendritic spines |
Enterohemorrhagic Escherichia coli O157:H7 ( EHEC ) is an important food-borne pathogen that colonizes the colon . Transposon-insertion sequencing ( TIS ) was used to identify genes required for EHEC and E . coli K-12 growth in vitro and for EHEC growth in vivo in the infant rabbit colon . Surprisingly , many conserved loci contribute to EHEC’s but not to K-12’s growth in vitro . There was a restrictive bottleneck for EHEC colonization of the rabbit colon , which complicated identification of EHEC genes facilitating growth in vivo . Both a refined version of an existing analytic framework as well as PCA-based analysis were used to compensate for the effects of the infection bottleneck . These analyses confirmed that the EHEC LEE-encoded type III secretion apparatus is required for growth in vivo and revealed that only a few effectors are critical for in vivo fitness . Over 200 mutants not previously associated with EHEC survival/growth in vivo also appeared attenuated in vivo , and a subset of these putative in vivo fitness factors were validated . Some were found to contribute to efficient type-three secretion while others , including tatABC , oxyR , envC , acrAB , and cvpA , promote EHEC resistance to host-derived stresses . cvpA is also required for intestinal growth of several other enteric pathogens , and proved to be required for EHEC , Vibrio cholerae and Vibrio parahaemolyticus resistance to the bile salt deoxycholate , highlighting the important role of this previously uncharacterized protein in pathogen survival . Collectively , our findings provide a comprehensive framework for understanding EHEC growth in the intestine .
Enterohemorrhagic Escherichia coli ( EHEC ) is an important food-borne pathogen that causes gastrointestinal ( GI ) infections worldwide . EHEC is a non-invasive pathogen that colonizes the human colon and gives rise to sporadic infections as well as large outbreaks [1–3] . The clinical consequences of EHEC infection range from mild diarrhea to hemorrhagic colitis and include the potentially lethal hemolytic uremic syndrome ( HUS ) [4 , 5] . The prototypical EHEC O157:H7 strain , EDL933 , caused the first recognized EHEC outbreak in 1982 [6] . EDL933 has a 5 . 5 Mb chromosome and a 90 kb virulence plasmid [7 , 8] . The E . coli species pan-genome is large ( >16 , 000 genes ) , and any given isolate contains a diverse , mosaic genome with approximately 1 , 500–2 , 000 conserved “core” genes [9–14] and an additional 3 , 000–4 , 000 “accessory” genes . The pathotype E . coli O157:H7 specifically contains one or more prophages encoding Shiga toxins and the Locus of Enterocyte Effacement ( LEE ) pathogenicity island [15] . These two horizontally acquired elements are critical EHEC virulence determinants . Shiga toxins contribute to diarrhea and the development of HUS [4 , 5 , 16] . The LEE encodes a type III secretion system ( T3SS ) and several secreted effectors . EHEC’s T3SS mediates attachment of the pathogen to colonic enterocytes , effacement of the brush border microvilli , and the formation of actin-rich pedestal-like structures underneath attached bacteria ( reviewed in [17] ) . Once translocated into the host cell , T3SS effectors , which are encoded both inside and outside the LEE , target diverse signaling pathways and cellular processes [18 , 19] . A functional LEE T3SS is required for EHEC intestinal colonization in animal models as well as in humans [16 , 17 , 20–24] . In addition to the virulence factors that prompt the key symptoms of infection , EHEC also relies on bacterial factors that enable pathogen survival in and adaptation to the host environment . During colonization of the human GI tract , EHEC encounters multiple host barriers to infection , including but not limited to stomach acid , bile , and other host- and microbiota-derived compounds with antimicrobial properties ( reviewed in [25] ) . EHEC is known to detect intestinal cues derived from the host and the microbiota to activate expression of virulence genes and to modulate gene expression both temporally and spatially [26–30] . However , a comprehensive , genome-wide analysis of bacterial factors that contribute to EHEC survival within the host has not been reported . The development of transposon-insertion sequencing ( TIS , also known as TnSeq , InSeq , TraDIS , or HITS ) [31–34] facilitated high-throughput and genome-scale analyses of the genetic requirements for bacterial growth in different conditions , including in animal models of infection [35–44] . In this approach , the relative abundance of transposon-insertion mutants within transposon-insertion libraries provides insight into loci’s contributions to bacterial fitness in different environments [45 , 46] . Potential insertion sites for which corresponding insertion mutants are not recovered frequently correspond to regions of the genome that are required for bacterial growth ( often termed “essential genes” ) , although the absence of a particular insertion mutant does not always reflect a critical role for the targeted locus in maintaining bacterial growth [47 , 48] . Comparative analyses of the abundance of mutants in an initial ( input ) library and after growth in a selective environment ( e . g . , an animal host ) can be used to gauge loci’s contributions to fitness in the selective condition . Here , transposon libraries were created in EHEC EDL933 and the laboratory-adapted E . coli K-12 and used to characterize their respective in vitro growth requirements . The EHEC library was also passaged through an infant rabbit model to identify genes required for intestinal colonization . Our data indicate that during infection of the GI tract , EHEC populations undergo a severe infection bottleneck that complicates identification of genes with true in vivo fitness defects . We used two complementary analytic approaches to mitigate the noise introduced by restrictive bottlenecks to identify over 200 genes required for efficient colonization of the rabbit colon . As expected , these included the LEE-encoded T3SS and tir , a LEE-encoded effector necessary for intestinal colonization [16 , 49] . In addition , 2 non-LEE effectors and many additional new genes that encode components of the bacterium’s metabolic pathways and stress response systems were found to enable bacterial colonization of the colon . Isogenic mutants for 17 loci , including cvpA , a gene necessary for intestinal colonization by diverse enteric pathogens [45 , 50 , 51] , were constructed , validated in the infant rabbit model and tested in vitro under stress conditions that model host-derived challenges encountered within the GI tract . cvpA was found to be specifically required for resistance to the bile salt deoxycholate and therefore appears to be a previously unappreciated member of the bile-resistance repertoire of diverse enteric pathogens .
The mariner-based Himar1 transposon , which inserts specifically at the TA dinucleotide [52] was used to generate a transposon-insertion library in EDL933 . The library was characterized via high-throughput sequencing of genomic DNA flanking sites of transposon-insertion . To map the reads , we used the most recent EDL933 genome sequence [8] and annotation ( NCBI , February 2017 ) . Since this genome , unlike the initial EHEC genome [7] , has not been linked to functional information ( e . g . , the EHEC KEGG database [53] ) , we generated a correspondence table in which the new gene annotations ( RS locus tags ) are linked to the original annotations ( Z numbers ) ( S1 Table ) . This correspondence table enabled us to utilize historically valuable resources as well as the updated genomic sequence and should also benefit the EHEC research community . 137 , 805 distinct insertion mutants were identified , which corresponds to 52 . 5% of potential insertion mutants with an average of ~21 reads per genotype ( S1 Fig ) . Sensitivity analysis revealed that nearly all mutants were represented within randomly selected read pools containing ~2 million reads . Increasing sequencing depth to ~3 million reads had a negligible effect on library complexity , suggesting that a sequencing depth of ~3 million reads is sufficient to identify virtually all genotypes within this EHEC library ( S1 Fig ) . EHEC’s 6032 annotated genes were binned according to the percentage of disrupted TA sites within each gene , and the number of genes corresponding to each bin was plotted ( Fig 1A ) . As expected for a high-density Himar1 transposon-insertion library , which contains insertions at a majority of TA sites , this distribution was bimodal , with a minor peak comprised of genes disrupted in few potential insertion sites ( Fig 1A , left ) , and a major peak comprised of genes that are disrupted in most or all potential insertion sites ( Fig 1A , right ) [46] . Based on the center of the major right-side peak , we estimate that ~70% of non-essential insertion sites have been disrupted in this EHEC library , a degree of complexity that enabled high-resolution analysis of transposon-insertion frequency . Further analysis of insertion site distribution was performed using a hidden Markov model-based analysis pipeline ( EL-ARTIST , see methods and [45] ) that classifies loci with a low frequency of transposon-insertion across the entire coding sequence as ‘underrepresented’ ( often referred to as ‘essential’ genes ) or across a portion of the coding sequence as ‘regional’ ( Fig 1B ) . All other loci are classified as ‘neutral’ . Of EHEC’s 6032 genes , 895 genes were classified as underrepresented ( red ) , 407 as regional ( purple ) , and 4 , 730 as neutral ( blue ) ( Fig 1A and 1B , S2 Table ) . Neutral genes are likely dispensable for growth in LB , whereas non-neutral genes ( regional and underrepresented genes combined ) likely have important functions for growth in this media or are otherwise refractory to transposon-insertion [47 , 48] . We identified Z Numbers ( S1 Table ) and the linked Clusters of Orthologous Groups ( COG ) [54 , 55] and KEGG pathways associated with the 1302 genes classified as non-neutral ( underrepresented and regional ) ( S2 and S3 Tables ) . Each COG category was plotted against its “COG Enrichment Index” , which is calculated as the percentage of non-neutral genes in each COG category divided by the percent of the whole genome with that COG [56] . A subset of COGs , particularly translation , lipid and coenzyme metabolism , and cell wall biogenesis were associated with non-neutral genes at a frequency significantly higher than expected based on their genomic representation ( S2 Fig ) . Collectively , the COG and a similar KEGG analysis ( S3 Table ) revealed that EHEC genes with non-neutral transposon-insertion profiles are associated with pathways and processes often linked to essential genes in other organisms [57] . Non-neutral genes comprise ~22% of EHEC’s annotated genes , a proportion of the genome that is substantially larger than the 8% and 9% observed in analogous TIS-based characterizations of Vibrio parahaemolyticus and Vibrio cholerae [45 , 50] . To evaluate whether the abundance of non-neutral loci was specific to EHEC or was characteristic of additional E . coli strains , a high-density transposon-insertion library was constructed in E . coli K-12 MG1655 [58] . EL-ARTIST analysis of the K-12 library ( S1 Fig ) was implemented with the same parameters as those for the EHEC library and classified 24% of genes as underrepresented ( 786 underrepresented , 300 regional and 3397 neutral; Fig 1C , S4 Table ) . We compared gene classifications between homologous loci ( see methods and S2 Table ) and found a substantial concordance between the sets of genes with non-neutral insertion profiles: 83% ( 629/760 ) of the non-neutral EHEC genes with homologs in K-12 were likewise classified as non-neutral in K-12 ( Fig 1D ) . Thus , analyses of non-neutral loci suggest either that most ancestral loci make similar contributions to the survival and/or proliferation of EHEC and K-12 in LB or that they are similarly resistant to transposon-insertion . Previous analyses revealed that nucleoid binding proteins such as HNS , which binds to DNA with low GC content , can hinder Himar1 insertion [47] . Consistent with this observation , EHEC genes classified as non-neutral have a lower average GC content than genes classified as neutral ( S2 Fig; blue vs red distributions ) . Interestingly , the disparity in GC content between neutral and non-neutral loci is particularly marked for EHEC genes that do not have a homolog in K-12 ( divergent; Fig 1E ) . These analyses suggest that there is an association between GC content and transposon-insertion frequency in EHEC , as in other organisms , and that the prevalence of underrepresented loci among divergent loci may in part stem from the lower average GC content of these loci ( S2 Fig ) . Additional studies are necessary to determine if the association between low GC content and reduced transposon-insertion is due to the binding of HNS or other nucleoid-associated proteins , or as yet unidentified fitness-independent transposon-insertion biases . The sets of genes classified as underrepresented or regional in EHEC and K-12 transposon libraries were compared to the 300 genes classified as essential in the K-12 strain BW25113 based on their absence from a comprehensive library of single gene knockouts [59–61] . 98% of these genes ( 294/300 ) were also classified as underrepresented or regional in EDL933 ( S2 Table ) and MG1655 ( S4 Table ) . The few loci previously classified as essential but not found to be underrepresented or regional in our analysis include several small genes , whose low number of TA sites hampers confident classification . One gene in this list , kdsC , was found to have insertions across the gene in both EDL933 and MG1655 ( S2 Fig ) . kdsC knockouts have also been reported previously [62] , confirming that this locus is not required for K-12 growth despite the absence of an associated mutant within the Keio collection . Thus , underrepresented and regional loci encompass , but are not limited to , loci previously classified as essential . Several factors likely account for the frequent classification of “non-essential” loci as underrepresented or regional . First , loci can be classified as underrepresented even when viable mutants are clearly present within the insertion library ( Fig 1A ) ; insertions simply need to be consistently less abundant across a segment of the gene than insertions at other ( neutral ) sites . Loci may also be classified as underrepresented due to fitness-independent insertion biases , as discussed above [47 , 48] . Additional evidence that loci categorized as non-neutral by transposon-insertion studies are not necessarily essential for growth was provided by a recent study of essential genes in K-12 [63] . However , the more expansive non-neutral classification can provide insight into loci that enable optimal growth , in addition to those that are required . We further explored the 131 underrepresented EHEC loci ( S5 Table ) that were classified as neutral ( able to sustain insertions ) in K-12 . Most of these genes are linked to KEGG pathways for metabolism , particularly metabolism of galactose , glycerophospholipid , and biosynthesis of secondary metabolites ( S5 Table , Fig 1F ) . While this divergence could reflect the laboratory adaptation of the K-12 isolate , gene acquisition during EHEC evolution may have heightened the pathogen’s reliance on metabolic processes that are not critical for growth of K-12 . Such ancestral genes may be useful targets for antimicrobial agents , as they might antagonize EHEC growth without disruption of closely related commensal Enterobacteriaceae populations . TIS studies in additional E . coli isolates are required to determine the relative contributions of the core and accessory E . coli genome to the list of essential genes . To identify mutants deficient in their capacity to colonize the mammalian intestine , the EHEC transposon library was orogastrically inoculated into infant rabbits , an established model host for infection studies [16 , 49 , 64 , 65] . EHEC strain EDL933 causes diarrhea and similar pathology in infant rabbits as that previously described for EHEC strain 905 [16] . Transposon-insertion mutants were recovered from the colon 2 days post-infection , and the sites and abundance of transposon-insertion mutations were determined via sequencing . The relative abundance of individual transposon-insertion mutants in the library inoculum was compared to samples independently recovered from the colons of 7 animals to identify insertion mutants that were consistently less abundant in libraries recovered from the colon . Under ideal conditions , this signature is indicative of negative selection of the mutant during infection , reflecting that the disrupted locus is necessary for optimal growth within the intestine . Sequencing and sensitivity analyses of the 7 passaged libraries revealed that they contained substantially fewer unique insertion mutants than the library inoculum ( 23–38% total mutants recovered , ~30 , 000 of 120 , 000 ) ( S1 Fig ) . These data are suggestive of population constrictions that could have arisen from 2 distinct but not mutually exclusive causes: 1 ) negative selection , leading to depletion of mutants deficient at in vivo survival or intestinal colonization; and/or 2 ) infection bottlenecks , population constrictions that lead to stochastic reductions in the average number of insertions per gene independent of genotype or selective pressures . We binned genes according to the percentage of TA sites disrupted within their sequences and plotted the number of genes corresponding to each bin for both the inoculum ( Fig 2A-top ) and a representative rabbit-passaged sample ( Fig 2A-bottom ) . The passaged sample exhibited a marked leftward shift relative to the inoculum , a signature indicative of population constriction due to an infection bottleneck [46 , 66] . As infection bottlenecks can confound identification of genes with true fitness defects in vivo [45 , 66] , we analyzed the TIS data using two complementary pipelines that mitigate the effects of bottlenecks: Con-ARTIST [45] and CompTIS [67] . Con-ARTIST was developed for this purpose [45] , and we recently found that CompTIS , a PCA-based TIS analysis , is also useful for identifying genes whose inactivation leads to phenotypes that are consistent across animal replicates [67] . Both Con-ARTIST and CompTIS use iterative simulation-based normalization to compensate for experimental bottlenecks and facilitate discrimination between stochastic reductions in genotype abundance and reductions attributable to bona fide negative selection ( mutants for which there was a fitness cost in the host environment ) ; however , they use distinct methodologies to measure phenotypic consistency across animal replicates ( see methods for details ) in order to categorize genes as either conditionally depleted ( CD ) , queried ( Q ) , or insufficient data ( ID ) as compared to the inoculum library ( Fig 3 ) . Con-ARTIST utilizes relatively stringent standards to identify robust candidate genes that facilitate EHEC intestinal growth in each rabbit . Such ‘conditionally depleted’ ( CD ) genes must contain sufficient transposon-insertions for Con-ARTIST analysis ( at least 5 TA sites disrupted by transposon-insertion within the inoculum and output datasets ) and meet a standard of a 4-fold reduction in read abundance ( fold-change ) that is consistent across TA sites in a gene , where consistency is measured using a Mann Whitney U ( MWU ) statistical test ( Fig 3 ) . Queried genes contain sufficient insertions for analysis but fail to meet the fold-change or p-value threshold . In contrast , genes classified as insufficient data ( ID ) have fewer than 5 TA sites disrupted by transposon-insertion . The output of gene categorization using these thresholds is displayed for a single rabbit in Fig 2B ( additional animals in S3 Fig ) and summarized for all animals in Fig 2C . An additional criterion that genes be classified as CD in 5 or more of the 7 animals analyzed was imposed to create a consensus list of CD genes ( Fig 2C ) . In contrast to the >2000 genes classified as conditionally depleted in one or more animals , only 246 genes were classified as conditionally depleted across 5 or more animals ( S6 Table ) . CompTIS was also used to compare the seven libraries recovered from rabbit colons . CompTIS relies on PCA , a dimensional reduction approach used to describe the sources of variation in multivariate datasets , and can be used here as an alternative measure of phenotypic consistency between animals . This is particularly beneficial in this case , where the severe bottleneck limits the availability of individual transposon mutant replicates required for Con-ARTIST’s MWU p-value thresholds . In the CompTIS analysis , each gene’s fold change from the seven colon libraries was subjected to gene level PCA ( glPCA ) ( see methods and [67] ) . Genes for which fold change information is not reported in all seven animal replicates are classified as ID . glPC1 describes most of the variation in the animals ( S3 Fig ) and represents a weighted average of the fold change values for each gene across the 7 animals ( S6 Table ) . The signs and magnitudes of glPC1 were all similar ( S3 Fig ) , indicating that each rabbit contributes approximately equally to glPC1 , as expected for biological replicates . The distribution of glPC1 scores is continuous ( Fig 2D ) and describes each gene’s contribution to EHEC intestinal colonization . Most genes have a glPC1 score close to zero , suggesting that they do not contribute to colonization . However , the non-symmetrical distribution includes a marked left tail encompassing the lowest 10% of scores , which were classed as CD ( Fig 2D , S3 Fig ) . The list of 541 CD genes includes nearly all ( 85% ) of the genes classified as CD by the more conservative Con-ARTIST analysis outlined above ( Fig 3 ) . These analyses yield four groups of genes ( Fig 3 ) . Group 1 includes the 209 genes categorized as CD by both ConARTIST and CompTIS and represents the highest confidence candidate genes required for colonization . Group 2 includes 332 genes identified as CD by CompTIS but classified as Q or ID by Con-ARTIST; it includes ler ( glPC1 = -2290 ) , a critical activator of the LEE T3SS [68] , which is classified as queried by Con-ARTIST due to the relative paucity of unique insertion mutants . The lack of ler mutants is likely due to the small size of the gene ( few TA sites ) and to HNS binding occluding transposon insertion [68] . Group 3 includes the 37 genes identified as CD by ConARTIST but not by CompTIS , due to the absence of fold-change information for all seven replicate rabbits . Lastly , Group 4 includes the 5454 genes always identified as queried or insufficient data . Due to the severe bottleneck , we do not conclude that these genes are not attenuated relative to the wild type strain in vivo; it is likely that the list of CD loci called by either method is incomplete . Below , we primarily focused on Group 1 genes for further analysis , as they represent the most robust candidates for factors promoting EHEC intestinal colonization . Using our Z correspondence table ( S1 Table ) , 89% ( 186/209 ) of Group 1 genes were assigned to a COG functional category . CD genes were frequently associated with amino acid and nucleotide metabolism , signal transduction , and cell wall/envelope biogenesis , but only amino acid metabolism reached statistical significance after correction for multiple hypothesis testing ( Fig 2E ) . These genes are also associated with KEGG metabolic pathways ( particularly amino acid metabolism ) , several two-component systems , including qseC , which has previously been implicated in EHEC virulence gene regulation , and lipopolysaccharide biosynthesis [26] ( S7 Table , Fig 2F ) . 30 of the 209 CD genes are EHEC specific , whereas the remaining 179 have homologs in K-12 ( S6 Table ) , highlighting the importance of conserved metabolic pathways in the pathogen’s capacity to successfully colonize its colonic niche . Similar metabolic pathways were also found to be important for V . cholerae growth in the infant rabbit small intestine [45 , 69] , raising the possibility of targeting metabolic pathways such as those for amino acid biosynthesis with antibiotics [70–72] . To assess the accuracy of our gene classifications using Con-ARTIST and CompTIS , we examined classifications within the LEE pathogenicity island , which encodes the EHEC T3SS and plays a critical role in intestinal colonization [16 , 20–23] . The LEE is comprised of 40 genes , including genes encoding the structural components of the T3SS , some of the pathogen’s effectors , their chaperones , and Intimin ( eae ) , the adhesin that binds to the translocated Tir protein . In infant rabbits , previous studies using single deletion mutants revealed that tir , eae , and escN , the T3SS ATPase , were all required for colonization [16 , 49] . We observed a marked reduction in the abundance of insertions across nearly the entire LEE in the samples from the rabbit colons relative to the simulation-normalized input reads , indicating this locus is required for colonization ( Fig 4A ) . However , the LEE has low GC content and is regulated through HNS binding [68] which when coupled with the infection bottleneck are expected to hamper assessment of gene contributions to intestinal fitness using Con-ARTIST alone . 12/40 LEE-encoded genes were categorized as Group 1 ( CD by both ConARTIST and CompTIS ) , including 3 genes previously found to be required for colonization ( tir , eae and escN ) and 8 additional genes critical for T3SS activity , including translocon T3SS components ( espB , espD , and espA ) and structural components ( escD , escQ , escV , escI , and escC ) ( Fig 4AB , S6 Table ) [20 , 73–77] . Additionally , 15 genes were categorized as Group 2 , including many encoding chaperones critical for T3SS assembly or T3SS structural components . We also assessed the contribution of EHEC T3SS effectors on colonization . EHEC has 49 effectors: 6 genes encoded within the LEE ( espF , espG , espH , espZ , and map ) and 43 non-LEE encoded effectors ( Nle ) . Nearly all effectors ( 5/6 LEE-encoded effectors and 41/43 Nle genes ) were categorized into Group 4 ( not CD ) . Consistent with these results , previous studies have shown that the LEE-encoded effectors espG and map are dispensable for robust colonization , and that ΔespH and ΔespF only had modest colonization defects [49] . The only effectors found to be important for colonization by either Con-ARTIST or CompTIS were tir ( as expected ) and 2 Nle genes: nleA and espM1 . NleA was previously reported to be important for colonic colonization by a related enteric pathogen , Citrobacter rodentium [78] , and is thought to suppress inflammasome activity [79]; EspM1 is thought to modulate host actin cytoskeletal dynamics [80 , 81] . Interestingly , mutants in nleA and espM1 were also scored as attenuated in a TraDIS-based study of EDL933 growth in calves , where abundance of mutants in feces was analyzed [30]; however , this study used a much smaller miniTn5 library ( covering 855 genes ) and relied on a different analytic approach that used a less stringent definition of attenuation ( solely based on 2-fold reduction in transposon mutant abundance in output vs input ) , making comparison to our study difficult . Additional studies are warranted to confirm and further explore how these 2 Nle effectors play pivotal roles promoting intestinal colonization . We performed further studies of 17 conditionally depleted genes/operons that had not previously been demonstrated to promote EHEC intestinal colonization . All genes were part of Group 1 , except hupB , which was identified as CD only by CompTIS ( S6 Table ) . Mutants with in-frame deletions of either single loci ( agaR , cvpA , envC , htrA , hupB , mgtA , oxyR , prc , sspA , sufI , tolC , and RS09610 , a hypothetical gene of unknown function ) or operons with one or more genes classified as conditionally depleted ( acrAB , clpPX , envZompR , phoPQ , tatABC ) were generated . Then , each mutant strain was barcoded with unique sequence tags integrated into a neutral locus in order to enable multiplexed analysis . The barcoded mutants , along with the barcoded WT EHEC , were co-inoculated into infant rabbits to compare the colonization properties of the mutants and WT . The relative frequencies of WT and mutant EHEC within colony-forming units ( CFU ) recovered from infected animals was enumerated by deep sequencing of barcodes , and these frequencies were used to calculate competitive indices ( CI ) for each mutant ( i . e . , relative abundance of mutant/WT tags in output normalized to input ) . 14 of the 17 mutants tested had CI values significantly lower than 1 , validating the colonization defects inferred from the TIS data ( Fig 5 ) . The in vitro growth of the barcoded mutants was indistinguishable from that of the WT strain ( S4 Fig ) , suggesting that the in vivo attenuation is not explained by a generalized growth deficiency . In aggregate , these observations support our experimental and analytical approaches and provide confirmation that many of the genes classified as CD in vivo contribute to intestinal colonization . The many new genes implicated in EHEC colonization by the TIS data could contribute to the pathogen’s survival and growth in vivo by a large variety of mechanisms . Given the pivotal role of EHEC’s T3SS in intestinal colonization , as well as previous observations that factors outside the LEE can regulate T3SS gene expression and/or activity ( reviewed in [68] ) , we assessed whether T3SS function was impaired in the 11 mutants with CIs <0 . 3 ( Fig 6A ) . Translocation of EspF ( an effector protein ) fused to a TEM-1 beta-lactamase reporter into HeLa cells was used as an indicator of T3SS functionality [82] . An ΔescN mutant , which lacks the ATPase required for T3SS function , was used as a negative control . Deletions in three protease-encoded genes , clpPX , htrA , and prc , were associated with reduced EspF translocation ( Fig 6A ) . Both ClpXP and HtrA have been implicated in T3SS expression/activity in previous reports [83–86] . The ClpXP protease controls LEE gene expression indirectly by degrading LEE-regulating proteins RpoS and GrlR [87] . The periplasmic protease HtrA ( also known as DegP ) has been implicated in post-translational regulation of T3SS as part of the Cpx-envelope stress response [85 , 86] . Interestingly , prc , which also encodes a periplasmic protease [87] , also appears required for robust EspF translocation . Prc has been implicated in the maintenance of cell envelope integrity under low and high salt conditions in E . coli K-12 [88] . Consistent with this observation , in high osmolarity media a Δprc EHEC mutant exhibited cell shape defects ( S4 Fig ) . Deficiencies in the cell envelope associated with absence of Prc may impair T3SS assembly and/or function , perhaps also by triggering the Cpx-envelope stress response . Together , these observations suggest that in vivo these three proteases modulate T3SS expression/function , thereby promoting EHEC intestinal colonization . We also investigated the capacity of each of the 11 mutant strains to survive challenge with three stressors–low pH , bile , and high salt ( osmotic challenge ) –that the pathogen may encounter in the gastrointestinal tract . Relative to the WT strain , all but one ( sufI ) of the mutant strains exhibited reduced survival following one or more of these challenges ( Fig 6B and 6C ) , suggesting that exposure to these host environmental factors may contribute to the in vivo attenuation of these mutants . Many of the EHEC mutants exhibited sensitivities to external stressors that are consistent with previously described phenotypes in other organisms and experimental systems . For example , the EHEC ΔacrAB locus , which was associated with bile sensitivity in EHEC ( Fig 6B ) , is known to contribute to a multidrug efflux system that can extrude bile salts , antibiotics , and detergents [89] . Our observation that mutants lacking the oxidative stress response gene oxyR are sensitive to bile and to acid pH is also concordant with previous reports linking both stimuli to oxidative stress [90–92] . Furthermore , the heightened sensitivity to bile , acid , and elevated osmolarity of EHEC lacking the two-component regulatory system EnvZ/OmpR is consistent with previous reports that EnvZ/OmpR is a critical determinant of membrane permeability , due to its regulation of outer membrane porins OmpF and OmpC . Mutations that activate this signaling system ( in contrast to the deletions tested here ) have been found to promote E . coli viability in vivo and to enhance resistance to bile salts [93] . The EHEC ΔtatABC mutant exhibited a marked colonization defect and a modest increase in bile sensitivity . The twin-arginine translocation ( Tat ) protein secretion system , which transports folded protein substrates across the cytoplasmic membrane ( reviewed in [94 , 95] ) , has been implicated in the pathogenicity of a variety of Gram-negative pathogens , including enteric pathogens such as Salmonella enterica serovar Typhimurium [96–98] , Yersinia pseudotuberculosis [99 , 100] , Campylobacter jejuni [101] , and Vibrio cholerae [102] . Attenuation of Tat mutants can reflect the combined absence of a variety of secreted factors . For example , the virulence defect of S . enterica Typhimurium tat mutants are likely due to cell envelope defects caused by the inability to secrete the periplasmic cell division proteins AmiA , AmiC and SufI [97] . Notably , single knock-outs of any of these genes does not cause S . enterica attenuation [97] , but altogether their absence renders the cell-envelope defective and more sensitive to cell-envelope stressors , such as bile acids [98] . In EHEC , the Tat system has been implicated in Stx1 export [103] , but because Stx1 was not a hit in our screen and is not thought to modulate intestinal colonization in infant rabbits [16] , it is not likely to explain the marked colonization defect of the EHEC ΔtatABC mutant . The suite of EHEC Tat substrates has not been experimentally defined , although putative Tat substrates can be identified by a characteristic signal sequence [94 , 95] . A few substrates , including SufI , OsmY , OppA , MglB , and H7 flagellin , have been detected experimentally [103] . sufI , interestingly , was also a validated hit in our screen , and is the only CD gene that has a predicted Tat-secretion signal . However , the ΔsufI mutant did not display enhanced bile sensitivity , suggesting that attenuation of this mutant , and perhaps of the ΔtatABC mutant as well , reflects deficiencies in other processes . SufI is a periplasmic cell division protein that localizes to the divisome and may be important for maintaining divisome assembly during stress conditions [104 , 105] . E . coli tat mutants have septation defects [106] , presumably from loss of SufI at the divisome . Interestingly , envC , another validated CD gene , encodes a septal murein hydrolase [107] that is required for cell division , and the ΔenvC mutant also displayed increased bile sensitivity . Consistent with this hypothesis , in high osmolarity media , the ΔsufI , ΔenvC , and ΔtatABC mutants exhibited septation or cell shape defects ( S4 Fig ) . Collectively , these data suggest that an impaired capacity for cell division may reduce EHEC’s fitness for intraintestinal growth , and that at times this may reflect increased susceptibility to clearance by host factors such as bile . We further characterized EHEC ΔcvpA because other TIS-based studies of the requirements for colonization by diverse enteric pathogens ( Vibrio cholerae , Vibrio parahaemolyticus and Salmonella enterica serovar Typhimurium ) also classified cvpA as important for colonization , but did not explore the reasons for mutant attenuation [45 , 50 , 51] . cvpA has been linked to colicin V export in E . coli K-12 [108] as well as curli production and biofilm formation in UPEC [109] . The EHEC ΔcvpA mutant did not exhibit an obvious defect in biofilm formation or curli production ( S5 Fig ) , suggesting that cvpA may have a distinct and previously unappreciated role in pathogenicity . Initially , we confirmed that the ΔcvpA mutant exhibits an intestinal colonization defect in 1:1 competition vs the wild type strain ( ~20-fold defect , S5 Fig ) . To further characterize the sensitivity of the EHEC ΔcvpA mutant to bile ( Fig 6B ) , we exposed the mutant to the two major bile salts found in the gastrointestinal tract , cholate ( CHO ) and deoxycholate ( DOC ) ( Fig 7A and 7C ) [90 , 110] . In contrast to WT EHEC , which displayed equivalent sensitivity to the two bile salts in MIC assays ( MIC = 2 . 5% for both ) , the ΔcvpA mutant was much more sensitive to DOC than to CHO ( MIC = 0 . 08% versus 1 . 25% ) . The ΔcvpA mutant’s sensitivity to deoxycholate was present both in liquid cultures and during growth on solid media ( Fig 7A–7C ) . cvpA lies upstream of the purine biosynthesis locus purF , and some ΔcvpA mutant phenotypes have been attributed to reduced expression of purF due to polar effects [108 , 111] , but the growth of our ΔcvpA mutant was not impaired in the absence of exogenous purines ( S5 Fig ) , suggesting the cvpA deletion does not adversely modify purF expression . Moreover , the DOC sensitivity phenotype of ΔcvpA was restored by plasmid-based expression of cvpA , confirming that this phenotype is linked to the absence of cvpA ( Fig 7A ) . Bile sensitivity has been associated with defects in the bacterial envelope or with reduced efflux capacity ( reviewed in [110] ) . We assessed the growth of the ΔcvpA mutant in the presence of a variety of agents that perturb the cell envelope to assess the range of the defects associated with the absence of cvpA . The MICs of WT and ΔcvpA EHEC were compared to those of an ΔacrAB mutant , whose lack of a broad-spectrum efflux system provided a positive control for these assays . Notably , the ΔcvpA mutant did not exhibit enhanced sensitivity to any of the compounds tested other than bile salts . In marked contrast , the ΔacrAB mutant displayed increased sensitivity to all agents assayed ( Fig 7C ) . These observations suggest that the sensitivity of the cvpA mutant to DOC is not likely attributable to a general cell envelope defect in this strain . V . cholerae and V . parahaemolyticus ΔcvpA mutants also exhibited sensitivity to DOC ( S5 Fig ) , implying a similar role in bile resistance in these distantly related enteric pathogens . A variety of bioinformatic algorithms ( PSLPred , HHPred , Phobius , Phyre2 ) suggest that CvpA is an inner membrane protein with 4–5 transmembrane elements similar to small solute transporter proteins ( Fig 7D ) . Phyre2 and HHPred reveal CvpA’s partial similarity to inner membrane transporters in the Major Facilitator Superfamily of transporters ( MFS ) and the small-conductance mechanosensitive channels family ( MscC ) . The PFAM database groups CvpA ( PF02674 ) in the LysE transporter superfamily ( CL0292 ) , a set of proteins known to enable solute export . In conjunction with findings presented above , these predictions raise the possibility that CvpA is important for the export of a limited set of substrates that includes DOC . Additional studies to confirm this hypothesis and to establish how CvpA enables export are warranted , particularly because this protein is widespread amongst enteric pathogens . Here , we created a highly saturated transposon library in EHEC EDL933 to identify the genes required for in vitro and in vivo growth of this important food-borne pathogen using TIS . This approach has transformed our capacity to rapidly and comprehensively assess the contribution an organism’s genes to growth in different environments [46 , 112 , 113] . However , technical and biologic issues can confound interpretation of genome-scale transposon-insertion profiles . For example , we found that EHEC genes with low GC content or those without homologs in K-12 were less likely to contain transposon-insertions ( S2 Fig , Fig 1C ) , which may reflect processes other than reduced biological fitness . For example , the presence of nucleoid binding proteins could impair transposon-insertion into these loci . Unexpectedly , more than 100 of the genes conserved between EHEC and K-12 appear to promote the growth of the pathogen but not that of K-12 ( S5 Table ) , suggesting that strain-specific processes may have enabled divergence of the metabolic roles of ancestral E . coli genes in these backgrounds . More comprehensive comparisons between additional E . coli isolates are needed to confirm this hypothesis . In animal models of infection , bottlenecks that result in marked stochastic loss of transposon mutants can severely constrain TIS-based identification of genes required for in vivo growth . Analysis of the distributions of the EHEC transposon-insertions in vitro and in vivo ( Fig 2 ) revealed that there is a large infection bottleneck in the infant rabbit model of EHEC colonization . Both Con-ARTIST , which applies conservative parameters to define conditionally depleted genes , and a PCA-based approach , CompTIS , were used to circumvent the analytical challenges posed by the severe EHEC infection bottleneck ( Fig 3 ) . These approaches should also be of use for similar bottlenecked data that often hampers interpretation of TIS-based infection studies . Validation studies , which showed that 14 of 17 genes classified as CD were attenuated for colonization ( Fig 5 ) , suggest that these approaches are useful . Besides the LEE-encoded T3SS , more than 200 additional genes were found to contribute to EHEC survival and/or growth within the intestine , most of which have never been linked to EHEC’s colonization capacity . This set of genes , particularly those involved in metabolic processes , should be of considerable value for future studies elucidating the processes that enable the pathogen to proliferate in vivo and for design of new therapeutics .
All animal experiments were conducted in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and the Animal Welfare Act of the United States Department of Agriculture using protocols reviewed and approved by Brigham and Women’s Hospital Committee on Animals ( Institutional Animal Care and Use Committee protocol number 2016N000334 and Animal Welfare Assurance of Compliance number A4752-01 ) Strains , plasmids and primers used in this study are listed in S8 and S9 Tables . Strains were cultured in LB medium or on LB agar plates at 37°C unless otherwise specified . Antibiotics and supplements were used at the following concentrations: 20 μg/mL chloramphenicol ( Cm ) , 50 μg/mL kanamycin ( Km ) , 10 μg/mL gentamicin ( Gm ) , 50 μg/mL carbenicillin ( Cb ) , and 0 . 3 mM diaminopimelic acid ( DAP ) . A gentamicin-resistant mutant of E . coli O157:H7 EDL933 ( ΔlacI::aacC1 ) and a chloramphenicol-resistant mutant of E . coli K-12 MG1655 ( ΔlacI::cat ) were used in this study for all experiments , and all mutations were constructed in these strain backgrounds except where specified otherwise . The ΔlacI::aacC1 and ΔlacI::cat mutations were constructed by standard allelic exchange techniques [114] using a derivative of the suicide vector pCVD442 harboring a gentamicin resistance cassette amplified from strain TP997 ( Addgene strain #13055 ) [115] or a chloramphenicol resistance cassette from plasmid pKD3 ( Addgene plasmid #45604 ) [59] flanked by the 5’ and 3’ DNA regions of the lacI gene . Isogenic mutants of EDL933 ΔlacI::aacC1 were also constructed by standard allelic exchange using derivatives of suicide vector pDM4 harboring DNA regions flanking the gene ( s ) targeted for deletion . E . coli MFDλpir [116] was used as the donor strain to deliver allelic exchange vectors into recipient strains by conjugation . Sequencing was used to confirm mutations . A ΔcvpA strain was also constructed using standard allelic exchange in a streptomycin- resistant mutant ( SmR ) of V . parahaemolyticus RIMD 2210633 . A cvpA::tn mutant was used from a Vibrio cholerae C6706 arrayed transposon library [117] . HeLa cells were obtained from The Harvard Digestive Diseases Center ( HDDC ) Core Facilities at Boston Children’s Hospital and were cultured in Dulbecco’s modified Eagle’s medium ( DMEM ) supplemented with 10% fetal bovine serum ( FBS ) . Cells were grown at 37°C with 5% CO2 and routinely passaged at 70 to 80% confluence; medium was replenished every 2 to 3 days . To create transposon-insertion mutant libraries in EHEC EDL933 ΔlacI::aacC1 , conjugation was performed to transfer the transposon-containing suicide vector pSC189 [118] from a donor strain ( E . coli MFDλpir ) into the EDL933 recipient . Briefly , 100 μL of overnight cultures of donor and recipient were pelleted , washed with LB , and combined in 20 μL of LB . These conjugation mixtures were spotted onto a 0 . 45 μm HA filter ( Millipore ) on an LB agar plate and incubated at 37°C for 1 h . The filters were washed in 8 mL of LB and immediately spread across three 245x245 mm2 ( Corning ) LB-agar plates containing Gm and Kn . Plates were incubated at 37°C for 16 h and then individually scraped to collect colonies . Colonies were resuspended in LB and stored in 20% glycerol ( v/v ) at -80°C as three separate library stocks . The three libraries were pooled to perform essential genes analysis , and one library aliquot was used to as an inoculum for infant rabbit infection studies . To create TIS mutant libraries in E . coli K-12 MG1655 ΔlacI::cat , conjugation was performed as above . 200 uL of overnight culture of the donor strain ( E . coli MFDλpir carrying pSC189 ) and the recipient strain ( MG1655 ΔlacI::cat ) were pelleted , washed , combined and spotted on 0 . 45 μm HA filters at 37°C for 5 . 5 hours . Cells were collected from the filter , washed , plated on selective media ( LB Km , Cm ) , and incubated overnight at 30°C . Colonies were resuspended in LB and frozen in 20% glycerol ( v/v ) . An aliquot was thawed and gDNA isolated for analysis . Mixed gender litters of 2-day-old New Zealand White infant rabbits were co-housed with a lactating mother ( Charles River ) . To prepare the EHEC transposon-insertion library for infection of infant rabbits , 1 mL from one library aliquot was thawed and added to 20 mL of LB . After growing the culture for 3 h at 37°C with shaking , the OD600 was measured and 30 units of culture at OD600 = 1 ( about 8 mL ) were pelleted and resuspended in 10 mL PBS . Dilutions of the inoculum were plated on LB agar plates with Gm and Km for precise dose determination . An aliquot of the inoculum was saved for subsequent gDNA extraction and sequencing ( input ) . Each infant rabbit was infected orogastrically with 500 μl of the inoculum ( 1x109 CFU ) using a size 4 French catheter . Following inoculation , the infant rabbits were monitored at least 2x/day for signs of illness and euthanized 2 days postinfection . The entire intestinal tract was removed from euthanatized rabbits , and sections of the mid-colon were removed and homogenized in 1 mL of sterile PBS using a minibeadbeater-16 ( BioSpec Products , Inc . ) . 200 uL of tissue homogenate from the colon were plated on LB agar containing Gm and Km to recover viable transposon-insertion mutants . Plates were grown for 16 h at 37°C . The next day , colonies were scraped and resuspended in PBS . A 5 mL aliquot of cells was used for genomic DNA extraction and subsequent sequencing ( Rabbits 1–7 ) . Transposon-insertion libraries were characterized as described previously ( 50 , 119 ) . Briefly , for each library , gDNA was isolated using the Wizard Genomic DNA extraction kit ( Promega ) . gDNA was then fragmented to 400–600 bp by sonication ( Covaris E220 ) and end repaired ( Quick Blunting Kit , NEB ) . Transposon junctions were amplified from gDNA by PCR . PCR products were gel purified to isolate 200-500bp fragments . To estimate input and ensure equal multiplexing in downstream sequencing , purified PCR products were subjected to qPCR using primers against the Illumina P5 and P7 hybridization sequence . Equimolar DNA fragments for each library were combined and sequenced with a MiSeq . Reads for all TIS libraries have been deposited in the SRA database ( Accession Number: PRJNA548905 ) . Reads were first trimmed of transposon and adaptor sequences using CLC Genomics Workbench ( QIAGEN ) and then mapped to Escherichia coli O157:H7 strain EDL933 ( NCBI Accession Numbers: chromosome , NZ_CP008957 . 1; pO157 plasmid , NZ_CP008958 . 1 ) using Bowtie without allowing mismatches . Reads were discarded if they did not align to any TA sites , and reads that mapped to multiple TA sites were randomly distributed between the multiple sites . After mapping , sensitivity analysis was performed on each library to ensure adequate sequencing depth by sub-sampling reads and assessing how many unique transposon mutants were detected ( S2 Fig ) . Next , the data was normalized for chromosomal replication biases and differences in sequencing depth using a LOESS correction of 100 , 000-bp and 10 , 000-bp windows for the chromosome and plasmid , respectively . The number of reads at each TA site was tallied and binned by gene and the percentage of disrupted TA sites was calculated . Genes were binned by percentage of TA sites disrupted ( Fig 1A and 1C ) . For essential gene analysis , EL-ARTIST was used as in [50] . Protein-coding genes , RNA-coding genes , and pseudogenes were included in this analysis . Briefly , EL-ARTIST classifies genes into one of three categories ( underrepresented , regional , or neutral ) , based on their transposon-insertion profile . Classifications are obtained using a hidden Markov model ( HMM ) analysis following sliding window ( SW ) training ( p <0 . 05 , 10 TA sites ) . Insertion-profiles for example genes were visualized with Artemis . For identification of mutants conditionally depleted in the rabbit colon as compared to the input inoculum , Con-ARTIST was used as in [119] . First , the input library was normalized to simulate the severity of the bottleneck as observed in the libraries recovered from rabbit colons using multinomial distribution-based random sampling ( n = 100 ) . Next , a modified version of the Mann-Whitney U ( MWU ) function was applied to compare these 100 simulated control data sets to the libraries recovered from the rabbit colon . All genes were analyzed , but classification as “conditionally depleted” ( CD ) was restricted to genes that had sufficient data ( ≥5 informative TA sites ) , met our standard of attenuation ( mean log2 fold change ≤ -2 ) , met our standard of phenotypic consistency ( MWU p-value of ≤0 . 01 ) , and had a consensus classification in 5 or more of the 7 animals analyzed . Genes with ≥5 informative TA sites that fail to exceed both standards of attenuation and consistency are classified as “queried” ( Q , blue ) , whereas genes with less than 5 informative TA sites are classified as “insufficient data” ( ID ) . Gene-level PCA ( glPCA ) was performed using CompTIS , a principal component analysis-based TIS pipeline , as described in [67] . Briefly , log2 fold change values were derived by comparing read abundance in each sample to 100 control-simulated datasets as in Con-ARTIST . These fold change values were weighted to minimize noise due to variability ( for details , see [67] ) . Next , genes that did not have a fold change reported for all 7 animals were discarded . The fold change values were then z-score normalized . Weighted PCA was performed in Matlab ( Mathworks ) with the PCA algorithm ( pca ) . Genes were categorized into 4 groups based on their classifications in Con-ARTIST and CompTIS ( Fig 3 ) . The GC content of classified genes was compared using a Mann-Whitney U statistical test and a Bonferroni correction for multiple hypothesis correction when more than one comparison was made . A p-value <0 . 05 was considered significant for one comparison , p<0 . 025 for two . A Fisher’s exact two-tailed t-test was used to compare ratios of classifications between groups , where a p-value of <0 . 01 was considered significant . Barcodes were introduced into ΔlacI::aacC1and isogenic mutant strains as described previously [50 , 120] . Briefly , a 991bp fragment of cynX ( RS02015 ) that included 51bp of the intergenic region between cynX and lacA ( RS02020 ) was amplified using primers that contained a 30 bp stretch of random sequence and cloned into SacI and XbaI digested pGP704 . The resulting pSoA176 . mix was transformed into E . coli MFDλpir . Individual colonies carrying unique tag sequences were isolated and used as donors to deliver pSoA176 barcoded derivatives to EDL933 ΔlacI::aacC1and each isogenic mutant strain . Three barcodes were independently integrated into EDL933 ΔlacI::aacC1 , and three barcodes into each isogenic mutant via homologous recombination in the intergenic region between cynX and lacA , which tolerates transposon-insertion in vitro and in vivo , indicating this locus is neutral for the fitness of the bacteria . Correct insertion of barcodes was confirmed by PCR and sequencing . To prepare the culture of mixed EHEC-barcoded strains for the multi-coinfection experiment , 100 μl of overnight cultures of the barcoded strains were mixed in a flask and 1 mL of this mix was added to 20 mL LB . After growing the culture for 3 h at 37°C with shaking , the OD600 was measured and 30 units of culture at OD600 = 1 ( about 8 mL ) were pelleted and resuspended in 10 mL PBS . Dilutions of the inoculum were plated on LB agar plates with Gm and Cb for precise CFU determination . 10 infant rabbits were inoculated and monitored as described above , and colon samples collected . Tissue homogenate was plated , and CFU were collected the following day . gDNA was extracted and prepared for sequencing as in [120] . The quantification of sequence tags was done as described previously [120] . In brief , sequence tags were amplified from the inoculum culture and libraries recovered from rabbit colons . The relative in vivo fitness of each mutant was assessed by calculating the competitive index ( CI ) as follows: We compare two strains ( ΔlacI::aacC1 and isogenic mutant ) in a population with frequencies fwt and fmut , x respectively where x is one of 17 mutant strains with a deletion in gene x . For simplicity , we assume here that both expand exponentially from a time point t0 to a sampling time point ts , their relative fitness ( offspring/generation ) is proportional to the competitive index CI: ln ( fmut , x , sfmut , x , 0fwt , sfwt , 0 ) =ln ( CI ) . Here , fwt , 0 and fmut , x , 0 are the frequencies of the strains in the inoculum , measured in triplicates , and fwt , s and fmut , x , s describe the frequencies at the sampling time point in the animal host . Because the WT strain was tagged with 3 individual tags and the inoculum was measured in triplicate , we have 3x3 = 9 measurements of the ratio fwt , sfwt , 0 . The same is true for all mutant strains , such that we have 9 measurements of the ratio fmut , x , sfmut , x , 0 . In total , we therefore have 3x3x3x3 = 81 CI measurements for each mutant per animal . To determine intra-host variance in these 81 measurements , a 95% confidence interval of the CI in single animal hosts was determined by bootstrapping . For combining the CIs measured across all 10 animal hosts , we performed a random-effects meta-analysis using the metafor package [121] in the statistical software package R ( version 3 . 0 . 2 ) . The pooled rate proportions and 95% confidence intervals were calculated using the estimates and the variance of CIs in each animal determined by bootstrapping and corrected for multiple testing using the Benjamini-Hochberg procedure . Each bacterial strain was grown at 37°C overnight . The next day , cultures were diluted 1:1000 into 100 uL of LB in 96-well growth curve plates in triplicate . Plates were left shaking at 37°C for 10–24 hours . Absorbance readings at 600nm were normalized to a blank , and the average of each triplicate was taken as the optical density . T3SS functionality was assessed by translocation of the known EHEC T3SS effector protein EspF into HeLa cells as described previously [77] . Briefly , the plasmid encoding the effector protein EspF fused to TEM-1 beta-lactamase was transformed into each of the bacterial strains to be tested . Overnight cultures of each bacterial strain were diluted 1/50 in DMEM supplemented with HEPES ( 25mM ) , 10% FBS and L-glutamine ( 2mM ) and incubated statically at 37°C with 5% CO2 for two hours . This media is known to induce T3SS expression [122] . HeLa cells were seeded at a density of 2x104 cells in 96-well clear bottom black plates and infected for 30 minutes at an MOI of 100 . After 30 minutes of infection IPTG was added at a final concentration of 1mM to induce the plasmid-encoded T3SS effector . After an additional hour of incubation , monolayers were washed in HBSS solution and loaded with fluorescent substrate CCF2/AM solution ( Invitrogen ) as recommended by the manufacturer . After 90 minutes , fluorescence was quantified in a plate fluorescence reader with excitation at 410nm and emission was detected at 450nm . Translocation was expressed as the emission ratio at 450/520nm to normalize beta-lactamase activity to cell loading and the number of cells presented at each well , and then normalized to WT levels of translocation . Biofilm and curli production assays were performed as described previously [109] . For biofilm assays , bacterial cultures were grown in yeast extract-Casamino Acids ( YESCA ) medium until they reached an OD600 ~ 0 . 5 and 1/1000 dilution of this culture was used to seed 96-well PVC plates . The cultures were grown at 30°C for 48 hours and biofilm production was quantitatively measured using crystal violet staining and absorbance reading at 595nm . Relative biofilm production was normalized to the average of three WT samples . A two-tailed Mann Whitney U test was used to determine if biofilm production was significantly different . A p-value < 0 . 05 was considered to be significant . To test curli production , bacterial cultures were grown in YESCA medium until they reached an OD600 ~ 0 . 5 and then were struck to single colonies onto YESCA agar plates supplemented with Congo Red . Red colonies indicate curli production . To test if our ΔcvpA deletion had polar effects on purF , the mutant and WT were struck onto minimal media lacking exogenous purines . An adaptation of the acid shock method described in [123] was performed . Briefly , bacterial cultures were grown until mid-exponential phase ( OD600 ~ 0 . 6 ) , then diluted 20-fold in LB pH 5 . 5 and incubated for 1 hour before preparing serial dilutions and plating each culture to determine the relative percentage of survival in comparison to the wild-type EDL933 strain . The pH of the LB broth was adjusted using sterilized 1mM HCl and buffered with 10% MES . Values are expressed as percent survival normalized to WT . MIC assays were performed using an adaptation of a standard methodology with exponential-phase cultures [124] . Briefly , the different compounds to be tested ( see Figs 6B and 7C ) were prepared in serial 2-fold dilutions in 50 ul of LB in broth in a 96-well plate format . To each well was added 50 ul of a culture prepared by diluting an overnight culture 1 , 000-fold into fresh LB broth , growing it for 1 h at 37°C , and again diluting it 1 , 000-fold into fresh medium . The plates were then incubated without shaking for 24 h at 37°C . Bile salt sensitivity assays were adapted from [125] . For plate sensitivity assays , each bacterial strain was grown at 37°C until they reached mid-exponential phase of growth ( OD600nm of 0 . 5 ) and the culture was serially diluted and spot-titered onto LB agar plates supplemented with either 1% DOC or 1% CHO . Spots were air dried and plates incubated at 37°C for 24 h . For sensitivity assays done in liquid culture , each bacterial strain was grown at 37°C until it reached mid-exponential phase of growth ( OD600nm of 0 . 5 ) and then cultures were split and supplemented with either DOC , CHO or buffer ( PBS ) and bacterial growth was assessed by absorbance at 600nm . Bacterial strains were grown in either LB or LB supplemented with 0 . 3M NaCl until mid-exponential phase and analyzed by phase microscopy at 100x magnification . To test the ΔcvpA in vivo fitness defect , we competed ΔcvpA against a ΔlacZ mutant in the infant rabbits . To prepare the cultures for infection , 100 μl of overnight cultures of each strain were inoculated into 100 mL of LB Gm for . After 3 hours of growth at 37°C with shaking , the OD600 was measured and 30 units of culture at OD600 = 1 were pelleted and resuspend in 10 mL PBS . Then , 5 mL of the ΔlacZ culture was combined with 5 mL ΔcvpA to make a 1:1 mixture . Dilutions of the inoculum were plated on LB agar plates containing Gm and X-Gal ( 60 mg/mL ) for precise CFU determination and determining the input ratio of ΔlacZ ( white ) to ΔcvpA ( blue ) . 4 infant rabbits were inoculated and inoculated and monitored as described above , and colon samples collected . Tissue homogenate was plated on LB agar plates containing Gm and X-Gal ( 60 mg/mL ) and CFU were counted the following day . A competitive index was calculated by dividing the burden of ΔcvpA divided by the ΔlacZ burden , adjusted for the input ratio . cvpA was inserted into the expression vector pMMB207 [126] ( ATCC #37809 ) downstream of the IPTG-inducible promoter using isothermal assembly . The resulting plasmid , pMMB207-cvpA , as well as an empty vector control , were transformed into ΔcvpA . To rescue the ΔcvpA DOC sensitivity phenotype , we used these strains in a bile salt sensitivity assay as described above . We found that expression from this plasmid was very leaky at basal ( non-induced ) conditions and could rescue the cvpA DOC sensitivity even without addition of IPTG . To enable comprehensive functional/pathway analyses in EHEC we carried out BLAST-based comparisons between the old EHEC genome sequence and annotation system ( NCBI Accession Numbers AE005174 and AF074613 [7] ) and the new sequence and annotation system ( NZ_CP008957 . 1 and NZ_CP008958 . 1 [8] ) ( S1 Table ) . This comparison links the new annotations ( RS locus tags ) to the original ‘Z numbers’ and their associated function and pathway annotation . To make the correspondence table ( S1 Table ) between the old EHEC annotation system ( Z Numbers ) and the new system ( RS Numbers ) , local nucleotide BLAST with output format 6 was used . First , a reference nucleotide database was generated from the newest EHEC sequence and annotation ( NZ_CP008957 . 1 and NZ_CP008958 . 1 ) . The EHEC genome sequence containing Z number annotations ( AE005174 and AF074613 ) was used as the query . Of the 6032 EDL933 genes , 5508 have 85% or greater nucleotide identity to a Z Numbered locus . 4796 of these genes match only one Z Numbered locus and 712 genes match multiple Z Numbers . These cases are frequently due to repetitive genomic sequences ( such as cryptic phage genes , transposons , or insertion elements ) or situations cases where two loci have been merged into one locus . Due to gaps and ambiguity present in the original EDL933 sequence , we did not filter on alignment length in order to find the best matches . In cases where there was one matching locus , the alignment was always ≥90% of the length of the gene . For genes with multiple matches , the length of the alignment varied . There are 524 genes with no corresponding Z Number , presumably loci that are newly recognized genes , and also 71 Z Numbers not assigned to an RS Number , which are loci now recognized as intergenic regions or not present in the final assembly . To find the K-12 homolog for EHEC genes ( S2 Table ) , local BLAST was also used . A reference nucleotide and amino acid database was generated from MG1655 K-12 ( NC_000913 . 3 ) , and the newest EHEC genome sequence was used as the query . For pseudogenes and genes coding for RNA , ≥90% nucleotide identity across ≥90% of the gene length was considered a homolog . For protein coding genes , ≥90% amino acid identity across ≥90% of the amino acid sequence was considered a homolog . To find KEGG pathways and COG assignments for genes of interest , the Z correspondence table was used to look up the Z number of each gene . The Z number and corresponding functional information was searched on the EHEC KEGG database . To determine if COGs were enriched in certain groups of genes ( such as conditionally depleted genes ) , a COG enrichment index was calculated as in [56] . The COG Enrichment Index is the percentage of the genes of a certain category ( essential genes or CD genes ) assigned to a specific COG divided by the percentage of genes in that COG in the entire genome . A two-tailed Fisher’s exact test was used to determine if this ratio was independent of grouping . A Bonferroni correction was applied for multiple hypothesis testing . A p-value of <0 . 002 was considered to be significant . Sequencing saturation of TIS libraries was determined by randomly sampling 100 , 000 reads from each library and identifying the number of unique mutants in that pool . Libraries are sequenced to saturation when no new mutants are identified as additional reads are added . 2–4 million reads are sufficient to capture the depth of libraries used here . Several protein prediction programs ( PSLPred , HHPred , Phobius , Phyre2 ) [127–129] were used to analyze the CvpA amino acid sequence . Protter [130] was used to compile information from several of these searches and generate a topological diagram . PRED-TAT [131] was used to search for tat-secretion signals in the list of CD genes . | Enterohemorrhagic E . coli ( EHEC ) is an important food-borne pathogen that infects the colon . We created a dense EHEC transposon library and used transposon-insertion sequencing to identify the genes required for EHEC growth in vitro and in vivo in the infant rabbit colon . We found that there is a large infection bottleneck in the rabbit model of intestinal colonization and refined two analytic approaches to facilitate rigorous identification of new EHEC genes that promote fitness in vivo . Besides the known type III secretion system , more than 200 additional genes were found to contribute to EHEC survival and/or growth within the intestine . The requirement for some of these new in vivo fitness factors was confirmed , and their contributions to infection were investigated . This set of genes should be of considerable value for future studies elucidating the processes that enable the pathogen to proliferate in vivo and for design of new therapeutics . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"bile",
"vertebrates",
"rabbits",
"animals",
"mammals",
"animal",
"models",
"bacterial",
"diseases",
"experimental",
"organism",
"systems",
"genetic",
"elements",
"enterohaemorrhagic",
"escherichia",
"coli",
"di... | 2019 | Transposon-insertion sequencing screens unveil requirements for EHEC growth and intestinal colonization |
The B1 SOX transcription factors SOX1/2/3/19 have been implicated in various processes of early embryogenesis . However , their regulatory functions in stages from the blastula to early neurula remain largely unknown , primarily because loss-of-function studies have not been informative to date . In our present study , we systematically knocked down the B1 sox genes in zebrafish . Only the quadruple knockdown of the four B1 sox genes sox2/3/19a/19b resulted in very severe developmental abnormalities , confirming that the B1 sox genes are functionally redundant . We characterized the sox2/3/19a/19b quadruple knockdown embryos in detail by examining the changes in gene expression through in situ hybridization , RT–PCR , and microarray analyses . Importantly , these phenotypic analyses revealed that the B1 SOX proteins regulate the following distinct processes: ( 1 ) early dorsoventral patterning by controlling bmp2b/7; ( 2 ) gastrulation movements via the regulation of pcdh18a/18b and wnt11 , a non-canonical Wnt ligand gene; ( 3 ) neural differentiation by regulating the Hes-class bHLH gene her3 and the proneural-class bHLH genes neurog1 ( positively ) and ascl1a ( negatively ) , and regional transcription factor genes , e . g . , hesx1 , zic1 , and rx3; and ( 4 ) neural patterning by regulating signaling pathway genes , cyp26a1 in RA signaling , oep in Nodal signaling , shh , and mdkb . Chromatin immunoprecipitation analysis of the her3 , hesx1 , neurog1 , pcdh18a , and cyp26a1 genes further suggests a direct regulation of these genes by B1 SOX . We also found an interesting overlap between the early phenotypes of the B1 sox quadruple knockdown embryos and the maternal-zygotic spg embryos that are devoid of pou5f1 activity . These findings indicate that the B1 SOX proteins control a wide range of developmental regulators in the early embryo through partnering in part with Pou5f1 and possibly with other factors , and suggest that the B1 sox functions are central to coordinating cell fate specification with patterning and morphogenetic processes occurring in the early embryo .
The developing embryo must control gene expression to coordinate various embryonic processes such as cell fate specification , embryo patterning and morphogenesis . During the embryonic stages from the blastula to neurula , the coupling of cell lineage specification and gastrulation cell movements is particularly evident . There is also now an increased understanding of the regulatory mechanisms underlying each cell state and each morphogenetic process , but the precise mechanisms that coordinate these events have remained elusive . The group B1 SOX transcription factors are good candidates as coordinators of these embryonic processes . Indeed , they have been implicated in cell fate specification in the early embryo [1]–[7] and also patterning and morphogenetic processes [8]–[10] . B1 Sox comprises sox1a/1b/2/3/19a/19b in zebrafish and Sox1/2/3 in amniotes [11] . The sox19a/19b genes are evolutionary orthologs of mammalian Sox15 ( group G ) , although Sox15 has now been shown to have functionally diversified from the authentic B1 Sox paralogs [11] . Overall , the regulatory functions of B1 sox genes appear to be conserved as a group across vertebrate species , although the paralogs are often differentially employed in a particular process [12] . In zebrafish , sox3/19a/19b are expressed in the blastula [11] , whereas the corresponding early expression in mice is covered by Sox2 [1] . Following this stage , the B1 sox genes are thought to be important for specification of the embryonic ectoderm into the neuroectoderm lineage . During this process , their expression becomes confined to the neuroectoderm [11] . As development proceeds to the neurula stage , expression of the B1 sox genes continues in neural precursors , where they function to maintain the neural progenitor states [13]–[15] . The similarities in the characteristics of the B1 SOX proteins as transcriptional regulators [11] , [15] suggest redundant functions in tissues where they are coexpressed . In support of this notion , single Sox1 or Sox3 knockout mice display only mild abnormalities in the central nervous system ( CNS ) , presumably because of extensive coexpression of Sox1/2/3 [16]–[18] , whereas Sox2-null mouse embryos die around implantation , reflecting its exclusive expression in the ICM [1] . Consistently , a single sox2 or sox3 knockdown ( KD ) in zebrafish causes only mild developmental abnormalities [19] , [20] . Xenopus studies utilizing dominant-negative forms of SOX2 indicate a specific role of Sox2 in neuroectoderm differentiation [2] . To date , however , the overall functions of the B1 sox genes have not been systematically investigated from the blastula to early neurula stages . An important characteristic of the B1 SOX proteins is that they form a complex with co-DNA-binding partner factors to target specific sequences and this enables them to participate in the regulation of various cell states [21] . The SOX2-Oct3/4 ( Pou5f1 ) complex is a central player in regulatory networks in the ICM and ES cells [1] , [22] , [23] . Potential target genes of SOX2 and Oct3/4 in ES cells have been identified through genome-wide chromatin immunoprecipitation ( ChIP ) and microarray expression analyses [22] , [24] . The involvement of other B1 SOX-partner combinations in the regulation of specific cell states has also been reported , e . g . , B1 SOX-POUIII factors in neural precursors [15] and B1 SOX-Pax6 in lens cells [25] . However , neither B1 SOX-dependent regulatory processes nor B1 SOX target genes in the developing early embryo have been extensively investigated . In our present study , we performed single to quadruple knockdowns of sox2/3/19a/19b in zebrafish embryos and confirmed that these four genes are functionally redundant in early development . More importantly , phenotypic analyses of the sox2/3/19a/19b quadruple KD embryos uncovered developmental process-specific functions of B1 sox . In the blastula , B1 sox genes regulate the activation of the bmp2b/7 genes , which is critical for dorsoventral ( DV ) patterning . During gastrulation , B1 sox also regulate the expression of pcdh18a/18b and wnt11 , a non-canonical Wnt ligand gene , which together play a role in convergence and extension ( C&E ) movements . In neural development , the B1 sox genes are essential for the proper regulation of neural bHLH genes of both the her/Hes and proneural classes , and also for the activation of region-specific transcription factor genes such as hesx1 , zic1 and rx3 . Moreover , the activity of B1 sox is required for the neural expression of various signaling pathway genes: cyp26a1 in RA signaling , oep in Nodal signaling , shh , and also mdkb . ChIP analysis of the her3 , hesx1 , neurog1 , pcdh18a and cyp26a1 genes suggests their direct regulation by B1 SOX . These findings indicate that B1 SOX proteins play a central role in coordinating cell fate specification , embryo patterning and morphogenesis by controlling a wide variety of developmental regulators in the early embryo . We have also found an interesting overlap between the early phenotypes of the B1 sox quadruple KD embryos and the maternal-zygotic ( MZ ) spg embryos that are devoid of pou5f1 activity [26]–[29] . This highlights a broad role of the B1 SOX-Pou5f1 complex from the blastoderm to early neural stages of development .
Among the B1 sox genes of zebrafish , sox2/3/19a/19b are expressed at high levels during early development with extensive regional overlaps [11] . sox19b mRNA is maternally supplied . sox3 and sox19a are activated around the 1000-cell stage , and sox2 around the 30% epiboly ( 30%E ) stage [11] . The expression of sox3/19a/19b initially covers the entire blastoderm , but gradually disappears at the embryonic margin after 30%E ( Figure S1A ) . At the shield stage , the expression of sox2/3/19a/19b covers the future ectoderm , but then becomes confined to the presumptive neuroectoderm [11] . Expression of sox1a/1b is initiated only during late gastrulation stages ( Figure S1B ) . These expression patterns suggest that sox2/3/19a/19b are involved in early processes of zebrafish development . To investigate the function of B1 sox in early stage embryos , we knocked down sox2/3/19a/19b either individually or in combination using morpholino antisense oligonucleotides ( MO ) . Two different MOs were simultaneously used to block translation of each B1 sox gene , which ensures efficient knockdown even when using reduced amounts of MOs [20] . With this double MO strategy , an approximately 90% reduction in translation was achieved using 1 . 8 ng of a 1∶1 mixture of two MOs , as judged by their effects on luciferase reporters carrying MO-targeting 5′-UTR sequences ( sox2 [20]; sox3/19a/19b , Figure S2A ) . By western blotting , we confirmed the efficient inhibition of the synthesis of endogenous B1 SOX proteins ( Figure S2B ) . No gross abnormalities were observed in the embryo morphology when any one of sox2/3/19a/19b was knocked down ( single KD , Figure 1B ) , although the development of the CNS may be slightly perturbed and 75% of the sox2 morphants showed an up-turned tail phenotype ( Figure 1Ba ) . When any three of sox2/3/19a/19b were simultaneously knocked down ( triple KD ) , a range of morphological abnormalities was observed depending on the combination of KD targets ( Figure 1C ) . Triple KDs of sox2/19a/19b and sox2/3/19b caused only mild morphological defects ( Figure 1Cc and 1Cd ) , presumably because the remaining sox3 and sox19a genes , respectively , mostly cover the B1 sox expression domains . sox3/19a/19b morphants often showed stronger yet variable defects in their posterior structures ( Figure 1Ca and 1Cb ) , presumably reflecting the weak sox2 expression in the posterior neuroectoderm . sox2/3/19a morphants appeared normal during gastrulation , but later developed morphological abnormalities ( Figure 1Ce ) , likely because sox19b expression decreases in later stages . In contrast to the triple KDs , the quadruple knockdown of sox2/3/19a/19b ( hereafter called QKD ) using a total of 7 . 2 ng MOs resulted in very severe developmental abnormalities , suggesting essential functions of B1 sox in early embryogenesis ( Figure 1Da–1Df ) . This result also lends support to a model whereby in the triple KD embryos the remaining B1 sox gene compensates for the loss of the other three to a large extent . Taken together , our initial observations indicate that the B1 sox genes are largely functionally redundant in early zebrafish embryos . This is further corroborated by rescue experiments to be described in the next section . By reducing the amount of MOs used for QKD , hypomorphic phenotypes were produced to different extents depending on the MO levels ( Figure 1Dg and 1Dh ) . Embryos injected with a 50% concentration of MOs for QKD ( 3 . 6 ng in total ) still showed more severe abnormalities than any of the triple KD embryos . However , injection of only 25% MOs for QKD ( 1 . 8 ng in total ) resulted in phenotypes similar to the triple KD embryos in terms of severity , further demonstrating the dose-dependent knockdown effects . The specificity of the MOs we used was confirmed by injection of 5-base mismatch MOs , which caused no developmental abnormalities ( Figure 1Di ) . In addition , the observed QKD phenotypes were not altered by coinjection of p53-MO ( Figure 1Dj ) , which relieves MO-induced non-specific neural cell death [30] . Efficient elimination of B1 SOX activity under this QKD condition was also confirmed by the loss of nestin enhancer activity ( Figure S3 ) , which is regulated by B1 SOX and POU [15] . The earliest detectable morphological abnormality of the QKD embryos was a delay in epiboly , notably after the shield stage ( Figure 1D ) . At 10 hour post-fertilization ( hpf ) , when normal embryos reach the tailbud stage , the QKD embryos were still in late epiboly . The thickening of the anterior head region was less prominent in the QKD embryos ( Figure 1Dc and 1Dd ) , suggesting impairment of CNS development . Impaired CNS development was also indicated by the loss of hesx1 expression in the anterior-most neuroectoderm ( Figure 2Bd; see also Figure 3Ca ) and by the anterolateral displacement of the pax2a expression domains that mark the midbrain-hindbrain boundary ( MHB ) ( Figure 2Bd ) . An early phase of neurogenesis was also affected in the QKD embryos as indicated by the loss of proneural neurog1 expression ( Figure 2Bf ) . The QKD embryos further displayed a shortened anterior-posterior ( AP ) axis with a broadened neural plate ( marked by hoxb1b ) and broadened mesodermal structures including notochord ( marked by ntl ) ( Figure 1D and Figure 2B ) . Consistently , the gap between the prechordal plate ( marked by hgg1 ) and notochord was reduced in the QKD embryos ( Figure 2Be ) . These abnormalities commonly occur in zebrafish embryos when C&E movements are impaired during gastrulation [31] , [32] . The effects of the B1 sox KDs suggested that the B1 SOX proteins act equivalently in transcriptional regulation in the zebrafish embryo . Indeed , SOX1/2/3/19 all activate the nestin and δ-crystallin enhancers in cooperation with Brn2 and Pax6 , respectively , in cultured cells [11] . We therefore tested whether injection of a single B1 sox mRNA could rescue the QKD phenotype . Moderate amounts ( 20–30 pg ) of B1 sox mRNAs lacking the MO-target 5′ UTR sequences were individually injected with the MOs for QKD . Coinjection with any one of the sox2/3/19a/19b mRNAs dramatically rescued the QKD phenotype , as judged by the recovery of a normal morphology ( Figure 2Ca–2Cc ) . By measuring the AP axis length of the embryos at 15–16 hpf , we confirmed the recovery of axial elongation in the B1 sox mRNA-injected QKD embryos ( Figure 2D ) . In these embryos also , the expression of hesx1 , dlx3b ( neural plate border ) and neurog1 was recovered and the expression patterns of pax2a and hoxb1b were restored ( Figure 2Cd–2Cf ) . In addition , the C&E movements indicated by the expression patterns of hgg1 , ntl and dlx3b [31] were normalized in the B1 sox mRNA-injected embryos ( Figure 2Ce ) . This phenotypic rescue was efficient only to early somitogenesis stages , likely because of the gradual decrease in the exogenously supplied SOX expression [19] . Simultaneous injection of sox2/3/19a/19b mRNAs ( 5 pg each ) had essentially the same rescue effects ( data not shown ) . These observations indicate that the function of the B1 SOX proteins is interchangeable during early zebrafish development . To further explore the functions of B1 sox in early zebrafish embryogenesis , we characterized our QKD embryos by focusing on their defects in neural development and gastrulation movements . In these QKD embryos , expression domains of otx2 ( anterior neuroectoderm ) and zic2b ( entire neuroectoderm ) were expanded ventrally at 75%E , whereas expression of both foxi1 and dlx3b ( non-neuroectoderm ) was reduced ( Figure 3Aa–3Ad ) . Consistently , the neuroectodermal expression of B1 sox was also expanded in the QKD embryo ( Figure 3Ae–3Ah ) . These gene expression changes are reminiscent of the dorsalized phenotype seen in BMP-pathway mutants [33] , [34] , suggesting impairment of this pathway in the QKD embryos . At the tailbud stage in the QKD embryos , expression domains of pax2a , gbx1 ( hindbrain ) and hoxb1b ( posterior neuroectoderm ) were anterolaterally shifted , which encompass the otx2 expression domain ( Figure 2Bd and Figure 3B ) . A similar change was seen for eng2a expression in the MHB at the 3-somite stage ( Figure 3Bf ) . These observations indicate that some characteristics of the early neural plate can develop even under a severe reduction of B1 SOX activity . However , our initial analyses revealed that expression of many neural genes is abolished in the QKD embryos , including zic1 ( forebrain ) , rx3 ( eye field ) , nkx1 . 2la ( posterior neuroectoderm ) and krox20 ( rhombomere [r] 3/5 ) as well as hesx1 and neurog1 ( Figure 2B and Figure 3C ) . Injection of the MOs for QKD into embryos of the nr2f2 enhancer-trap line [35] confirmed the impairment of brain development at later stages ( Figure S4A ) . These findings together suggest that B1 sox activity is critical for neural development , although it is dispensable for the expression of some early neural genes . The normal expression levels of ntl and myod1 ( somite ) in the QKD embryos suggest that mesodermal differentiation per se can proceed ( Figure 3D ) . However , the broadened expression domains of these genes indicate defects in convergence movements , which is consistent with the widened expression of neural genes such as gbx1 and hoxb1b . Movement of the anterior prechordal plate ( marked by hgg1 ) was also impaired in the QKD embryos ( Figure 2Be ) , which is characteristic of defective C&E movements . However , hatching gland precursor cells were commonly found to aberrantly move in a dorsal direction and penetrated the ectoderm during mid-somite stages ( Figure 1De’ ) . In the severe morphants , these hatching gland cells remained as a single ball-like structure in the head ( Figure S4B ) . This phenotype is unique to the QKD embryo and may reflect a decreased adhesion of ectodermal cells , as also suggested by cell dissociation from the dorsal trunk region ( Figure 1De’ and 1Df’ ) . Interestingly , the QKD embryos show an increase in transcript levels of sox2/3/19a/19b and also sox1b at early developmental stages ( Figure 3A and Figure S1B ) . This implies a negative autoregulation of transcription among the B1 sox members , although a stabilization of these mRNAs by the MOs could not be ruled out . It is also noteworthy that even with these elevated levels of B1 sox transcripts in the QKD embryos , the MO-mediated knockdowns were effective in inhibiting their translation as revealed by western blotting ( Figure S2B ) . To further characterize the phenotype of the QKD embryos , we examined the changes in gene expression in greater detail by the combined use of in situ hybridization , RT-PCR ( summarized in Table S1 ) and microarray analysis ( Figure S5 , Table S2 and ) . Overall , these analyses indicated that a wide range of developmental processes were affected in the QKD embryos and the major phenotypes can be categorized into: ( 1 ) the early dorsoventral patterning defects; ( 2 ) defects in gastrulation movements; ( 3 ) dysregulation of early neural and neuronal regulatory genes; ( 4 ) neural patterning defects associated with the abrogated expression of signaling pathway genes; and ( 5 ) early defects resembling those observed for MZspg embryos . Further details of these phenotypes are described below . Early DV patterning of the embryo relies on a gradient of Bmp signaling , in which bmp2b/7 play a major role in zebrafish [36] , [37] . The phenotypic similarities between the QKD embryos and bmp pathway mutants described above prompted us to examine the genes in this pathway . Expression of bmp2b/7 was found to be reduced in the QKD embryos from the beginning ( Figure 4Aa–4Af ) . bmp4 expression levels were more or less normal initially , but were downregulated at late epiboly stages in the QKD embryos ( Figure 4Ag; data not shown ) . Consistently , expression of the gata2 , szl and eve1 genes , which are immediately downstream of Bmp signaling , was also reduced in the QKD embryos ( Figure 4Ah–4Aj ) . The dorsal identity of the zebrafish embryo requires activation of maternal β-catenin , which then activates expression of gsc and the Bmp antagonist genes chd and nog1 at the dorsal side . In the QKD embryos , these genes were initiated normally , although the expression of chd was slightly ventrally expanded at 30%E ( Figure 4Ak ) , which is likely secondary to the reduced expression of bmp2b/7 . At later stages in the QKD embryos , however , chd expression was rather decreased ( Figure 4Am ) , contrasting to the bmp pathway mutants [38] . To determine the relationship between the dorsalized phenotype of the QKD embryos and Bmp signaling , we injected a mixture of bmp2b/7 mRNAs ( 20 or 40 pg each ) together with the MOs for QKD . This bmp2b/7 injection rescued the expression of the Bmp downstream genes gata2 , szl and eve1 ( Figure 4B ) , indicating that signaling components acting downstream of Bmp2b/7 are not affected in the QKD embryos . Consistently , the mRNA levels of Bmp receptor genes ( acvrl1 , bmpr1aa , bmpr1ab and bmpr1ba ) and smad5 were found to be normal in the QKD embryos by microarray ( GEO accession number GSE18830 ) . These observations , together with the normal initiation of the dorsal pathway in the QKD embryos , indicate that the DV patterning defects of the QKD embryos primarily result from the reduction of bmp2b/7 expression . Components of non-canonical Wnt signaling and cell adhesion molecules are major regulators of gastrulation movements , including epiboly and C&E movements [32] , [39] . Since these movements are severely impaired in the QKD embryos , we investigated expression profiles of genes related to these processes . wnt11 and wnt5b are major Wnt ligand genes involved in C&E movements [31] , [32] . In the QKD embryos , upregulation of wnt11 that normally occurs during early gastrulation was not observed and expression of wnt5b was slightly reduced at late epiboly stages ( Figure 4C ) . We also observed decreased expression of wnt11r and wnt4a , which are important for convergence movements at later stages [40] . These data suggest that the reduction of non-canonical Wnt ligands contributes to the impairment of C&E movements . Both classical cadherins and protocadherins are involved in gastrulation movements [39] . We found that expression of pcdh18a/18b was significantly reduced in the QKD embryos ( Figure 4C ) . These genes are expressed in the epiblast at the shield stage and later in the neuroectoderm in an overlapping manner in normal embryos [41] , [42] . To examine how reduced activity of Pcdh18a/18b affects embryogenesis , we knocked down these two genes . Although only mild gastrulation defects were observed when these genes were knocked down separately , simultaneous KD caused a delay in epiboly and also C&E defects ( Figure S6; for pcdh18a single KD , see also [41] ) . Delayed epiboly has also been reported for the hypomorphic cdh1 mutants [43] , [44] , but its expression was not altered in our QKD embryos ( Figure 4C ) . These observations indicate that multiple mechanisms involved in gastrulation movements are simultaneously affected in the QKD embryos . Several neuronal genes were found to be abnormally upregulated in the QKD embryos . stmn2a is strongly expressed in CNS neurons from mid-somitogenesis stages in wild-type embryos [45] and also weakly expressed throughout the embryo at epiboly stages ( Figure 5A and Figure S7A ) . The latter early stage expression was found to be aberrantly upregulated in the QKD embryos ( Figure 5A and Figure S7A ) . The neuronal tuba1 gene was also upregulated from 75%E ( Figure 5A ) . These observations suggest that a portion of the neuronal differentiation programs is precociously initiated in the QKD embryos . Neural bHLH transcription factors are key players in the neuronal differentiation programs . Hes/her genes encode repressor-type bHLH proteins , are expressed in undifferentiated neural progenitor cells and maintain their cell state [46] . Among the zebrafish her genes , her3 , an ortholog of mammalian Hes3 , is initiated in the dorsal region of the epiblast at about 30%E , and its expression continues in bilateral inter-proneuronal domains [47] . This her3 expression is totally lost in the QKD embryos ( Figure 5A and 5Ba–5Bc ) . Proneural genes encoding activator-type bHLH proteins and participating in neurogenesis are also affected in the QKD embryos . In normal zebrafish embryos , neurog1 expression initially marks primary neurons at the end of gastrulation and then covers the proneuronal domains in a fashion complementary to her3 expression . In the QKD embryos , neurog1 expression is also lost ( Figure 2Bf and Figure 5A ) . However , not all proneural genes behave in this manner as for example ascl1a is transiently upregulated at about 75%E in a broad area of the neuroectoderm in the QKD embryos ( Figure 5A and 5B ) . Proneural genes are known to be repressed by Hes/Her [46] , but exogenous injection of her3 mRNA into QKD embryos did not repress aberrant ascl1a expression ( data not shown ) , indicating that the loss of her3 was not causal to this upregulation . The neuronal repressor REST has been implicated in suppression of Ascl1 as well as Stmn2 [48] . However , rest expression was unchanged in the QKD embryos ( Figure 5A ) , although Rest is suggested to be downstream of SOX2 in ES cells [24] , indicating that rest is not involved in the aberrant regulation of ascl1a and stmn2a . Taken together , our results indicate that the proper operation of the neuronal differentiation programs , including regulatory networks involving the neural bHLH genes , is highly dependent on the activity of B1 sox . Regional identities of the neural plate are specified through regulatory networks involving various signaling pathways and transcriptional regulators . Genes that are critical for these networks are severely affected in the QKD embryos . As described earlier , in the QKD embryos , expression of the transcription factor genes hesx1 , zic1 , and rx3 , which are required for forebrain and eye development [49]–[51] , is lost throughout early embryogenesis ( Figure 3C and Figure 5A ) . The MHB itself was established , as judged from the expression of pax2a and eng2a , but the anterolaterally-shifted expression patterns of these genes suggest an improper formation of the axes of the anterior neural plate ( Figure 2Bd and Figure 3Bf ) . The expression domain of otx2 was expanded and encircled by those of pax2a , eng2a and gbx1 ( Figure 2Bd and Figure 3B ) , which is likely due to the dorsalized phenotype caused by the decreased expression of bmp2b/7 , as the bmp2b/7 mutant embryos show similar patterns of gene expression [33] , [34] . In addition , the QKD embryos lacked the anterior neuroectoderm expression of cyp26a1 ( Figure 5C ) , which encodes an RA degrading enzyme and thereby plays a role in hindbrain patterning [52] . Hence , the reduction of cyp26a1 expression partly accounts for the hindbrain defects in the QKD embryos , e . g . , expansion of hoxb1a in r4 ( Figure 5Cc ) as observed in the cyp26a1 mutant [53] . However , B1 sox also seem to be more directly involved in gene regulation in hindbrain development as evidenced by severe downregulation of mafba ( r5/6 ) ( Figure 5Cd ) and the loss of krox20 expression in r3/5 ( Figure 3Ce ) . Nodal and Sonic hedgehog signaling are crucial for the development of ventral brain structures [54]–[56] . In normal embryos , oep , an ortholog of mouse Cripto , is strongly expressed in the anterior neural plate and is essential as a coreceptor for receiving Nodal signals [54] . Interestingly , oep expression in the ectoderm at the shield stage and the neuroectoderm at later stages is selectively lost in the QKD embryos ( Figure 5Db–5Dd ) , whereas its early zygotic and mesodermal expression was maintained ( Figure 5Da–5Dd ) . Moreover , expression of shha and shhb in the ventral floor of the brain is also lost in the QKD embryos , leaving only shhb expression in the prechordal plate ( Figure 5Df–5Di ) . These findings suggest that the loss of oep and shha/b expression leads to defective ventral brain development in the QKD embryos . It is known that shha expression in the neuroectoderm is regulated by Nodal signals from the mesoderm [57] . However , a more direct link between B1 SOX action and shh regulation is suggested , as the exogenous injection of oep mRNA into the QKD embryos did not restore the shha expression ( data not shown ) , although Nodal-encoding ndr2 is normally expressed ( Figure 5De ) . Defects in the anterior neural plate development in the QKD embryos also include the loss of mdkb expression ( Figure 5E ) . Consistent with the proposed role of mdkb in the specification of neural crest cells [58] , foxd3 expression in the neural crest was reduced in the QKD embryos ( Figure 5Ec ) . As the B1 SOX proteins primarily function as transcriptional activators [11] , [13] , [15] , [23] , [25] , the expression of direct target genes is expected to be decreased in response to the QKD . However , the upregulation of several neuronal genes such as stmn2a raised the possibility that B1 SOX might also act as repressors . To test this , we utilized dominant activator and repressor forms of SOX3 , SOX3-VP16 and SOX3-EnR ( Figure 6A ) and compared the effects of these variants under QKD conditions with those of SOX3 . As anticipated , genes that were downregulated in the QKD embryos , namely bmp2b/7 , pcdh18a/18b , her3 , hesx1 and zic1 , were efficiently recovered by the exogenous supply of either SOX3 or SOX3-VP16 but not by SOX3-EnR ( Figure 6B ) . These genes are thus likely activation targets of B1 SOX . In addition , the increased expression of stmn2a and ascl1a in the QKD embryos was suppressed in the same way . This suggests an indirect regulation of these genes by B1 SOX through the activation of repressors . However , SOX3-VP16 was less effective than SOX3 in the rescue of some genes such as pcdh18a/18b and ascl1a , suggesting that the activation process may require additional molecular interactions with the intact SOX3 protein . It is noteworthy that the morphological rescue of the QKD embryos by SOX3-VP16 was much less complete when compared to that observed for SOX3 , and that the SOX3-VP16-injected embryos showed a rather ventralized phenotype ( Figure 6C ) . In line with this observation , chd and nog1 were unexpectedly reduced in SOX3-VP16-injected embryos , but increased in SOX3-EnR-injected embryos , suggesting that the repressive action of B1 SOX may be required for the proper regulation of dorsally expressed BMP antagonist genes . These findings together indicate that the B1 SOX proteins primarily act as activators in early embryos , whereas a context-dependent repressive action of these factors is also suggested . To further explore whether the B1 SOX proteins directly regulate the potential downstream genes described above , we searched for possible B1 SOX binding sites ( containing the consensus sequence CATTGTT [21] , [59] or closely related sequences ) in the regulatory regions of these genes . We identified potential SOX-binding sites in the regulatory sequences of her3 [47] , hesx1 , cyp26a1 [60] and neurog1 [61] and also in the conserved non-coding sequences upstream of pcdh18a ( Figure 7Aa ) . To investigate the direct interaction of the B1 SOX proteins with these genomic sequences in vivo , ChIP experiments were performed using zebrafish embryos at the 70–80%E and tailbud to 2-somite stages . ChIP analysis using anti-SOX2 antibody that weakly cross-reacts with SOX3/19A/19B revealed specific binding of B1 SOX to these regulatory sequences in the zebrafish embryo ( Figure 7Ab ) . Similar results were obtained with anti-SOX3 antibody ( data not shown ) . It is thus likely that these genes are direct downstream targets of B1 SOX . To further investigate whether the activities of these regulatory sequences are dependent on B1 SOX , we created luciferase reporter vectors containing the promoter sequences for hesx1 and cyp26a1 [60] ( Figure 7Ba ) . The 0 . 8-kb promoter sequence of the zebrafish hesx1 gene used here corresponds to the chicken Hesx1 promoter that has been shown to have anterior CNS-specific regulatory activity in chicken and also zebrafish embryos [62] . The conserved non-coding sequence upstream of pcdh18a ( 412 bp ) was also cloned into the TK-luciferase reporter vector ( Figure 7Ba ) . When these reporter vectors were injected with or without the MOs for QKD into zebrafish embryos , the promoter activities of hesx1 and cyp26a1 were significantly downregulated upon B1 sox QKD ( Figure 7Bb ) . The 412-bp pcdh18a sequence showed an enhancer activity in normal embryos , whereas this activity was also reduced in the QKD embryos ( Figure 7Bb ) . These data confirm that B1 SOX proteins regulate the hesx1 , cyp26a1 and pcdh18a genes through these regulatory elements . Interestingly , POU binding sites were found abutting the SOX sites of her3 and hesx1 . These genes were found to be commonly downregulated in the QKD embryos ( Figure 3C and Figure 5A and 5B ) and also MZspg mutants ( Table S7; see also [29] ) . In the QKD embryos , pou5f1 is expressed at normal levels ( Figure S7B ) , indicating that Pou5f1 alone is insufficient to induce her3 or hesx1 . These data together suggest that B1 SOX and Pou5f1 proteins synergistically cooperate to activate her3 and hesx1 .
We found similarities in the gene expression profiles between the B1 sox QKD embryos and the MZspg embryos in a wider range of developmental stages from the blastoderm to early neural stages ( Figure 8 , Table S7 and [29] ) . This strongly suggests that their cooperation is required not only for the blastoderm stage , which may be similar to the ES cell state , but also for the early neural stage . In contrast to mouse knockouts of Sox2 or Oct3/4 , where impairment of the ICM/epiblast lineage development causes early embryonic lethality [1] , [63] , zebrafish embryos of the B1 sox QKD and MZspg mutants are viable although with severe developmental defects . This enabled us to analyze the functions of B1 sox in later developmental stages as well as the blastoderm stage . A group of key genes downstream of B1 SOX and Pou5f1 in the blastoderm was found to be bmp2b/7 , as the expression of bmp2b/7 is also severely reduced in MZspg embryos [28] . In addition , an overlap of the expression domains of B1 sox , pou5f1 and bmp2b/7 in the blastoderm strongly suggests a direct regulation of bmp2b/7 by B1 SOX and Pou5f1 . The dorsalized phenotype of the B1 sox QKD and MZspg embryos can to a large extent be ascribed to the bmp2b/7 defects , as bmp2b/7 mRNA injection into the QKD embryos rescues the expression of Bmp downstream genes ( Figure 4B ) and bmp2b mRNA injection can also rescue the MZspg embryos [28] . This co-regulation thus appears to be critical for the establishment of the early DV axis , but likely operates only during the initial activation of bmp genes , since during gastrulation the expression domains of B1 sox and bmp2b/7 segregate and eventually become complementary to each other . her3 and hesx1 , which both encode transcriptional repressors , were identified as direct targets of B1 SOX in the early phase of neural development ( Figure 5 , Figure 6 , and Figure 7 ) . The expression of these genes is also lost in the MZspg embryos ( Table S7 and [29] ) . In addition , our ChIP analysis indicated that the proximal and distal SOX-POU elements of the her3 promoter and the hesx1 promoter carrying multiple SOX and POU sites are bound by B1 SOX in vivo ( Figure 7A ) . We further verified , using a luciferase reporter assay , that the activity of hesx1 promoter is dependent upon B1 SOX . These observations together indicate that her3 and hesx1 are regulated under the cooperative action of B1 SOX and Pou5f1 . In addition , the B1 SOX and Pou5f1 complex appears to be required for expression of hypothetical repressors that inhibit neuronal differentiation , since the expression of stmn2a is aberrantly upregulated in the B1 sox QKD embryos and also in MZspg embryos ( Figure 5A , Figure S7A , and Table S7 ) . These data indicate that in the early phases of neural development B1 SOX proteins cooperate with Pou5f1 and activate the transcriptional repressor genes that inhibit the further differentiation of neural progenitor cells . In the gastrulating embryo , cell fate specification must be coupled with embryo patterning and gastrulation movements . We found that B1 SOX proteins are involved in all these processes by controlling their respective regulators in the gastrulating embryo . In the early phase of neural fate specification , the B1 SOX proteins are required for the activation of her3 and the repression of ascl1a , which by themselves are inhibitory to neuronal differentiation ( Figure 8 ) . On the other hand , B1 SOX also appear to play a role in an initial phase of neuronal differentiation by directly activating the proneural neurog1 gene . neurog1 is severely downregulated in the QKD embryos ( Figure 2 and Figure 5 ) , and in vivo binding of B1 SOX to its regulatory sequence LSE is indicated by our ChIP analysis . In addition , although B1 SOX proteins are known to counteract neurogenesis , this inhibition occurs at late steps of neurogenesis without affecting Neurog1/2 expression [13] . Taken together , these data indicate that B1 SOX are important for the successive generation of neural progenitor cells and immature neuronal cells . Another important aspect of the functions of the B1 sox genes during neural lineage differentiation is that the initiation of the transcription factor genes zic1 and rx3 depends on their activity ( Figure 3 and Figure 5 ) . rx3 may also be directly regulated by B1 SOX , since Xenopus Rx1 , a functional homolog of zebrafish rx3 , is under the direct regulation of SOX2 and Otx2 in the eye field [64] . Zic1 and Rx3 generally act as a transcriptional activator and are required for forebrain and eye lineage development [50] , [51] . Critical functions of B1 sox in embryo patterning are underscored by our present findings that the neural expression of the signaling pathway genes cyp26a1 , oep , shha/b and mdkb is dependent upon B1 sox activity ( Figure 5 ) . Signaling pathways involving these genes play a key role in cell fate decisions as well as diverse patterning processes in the developing CNS [52]–[56] , [58] . Our ChIP and promoter analyses suggest a direct regulation of cyp26a1 by SOXB1 ( Figure 7 ) . Furthermore , the expression of oep , mdkb and shha/b extensively overlaps with that of B1 sox in the neuroectoderm , also implying direct regulation by B1 SOX . Interestingly , the expression of Shh during mouse hippocampal development has recently been shown to be directly regulated by SOX2 [65] . A remarkable defect of the QKD embryos was also found to occur in gastrulation movements . Delayed epiboly and impaired C&E movements are also shared phenotypes with the MZspg embryos [26]–[28] , suggesting that these processes may also be co-regulated by B1 SOX and Pou5f1 . We speculate that a severe reduction of pcdh18a/18b in combination with a reduced expression of non-canonical wnt genes is largely responsible for the defects in epiboly and C&E movements of the QKD embryos ( Figure 4 ) . We have further shown in our present analyses that the conserved sequence block upstream of pcdh18a acts as a B1 SOX-dependent enhancer . The in vivo binding of B1 SOX to this enhancer indicated by our ChIP analysis further supports a direct regulation of pcdh18a by B1 SOX . The knockdown phenotypes of pcdh18a/18b were consistent with their important functions in gastrulation movements ( Figure S6 and [41] ) , but their molecular role in this process is still unclear . Recent studies , however , have reinforced the critical role of cell adhesion molecules in gastrulation movements [39] . Pcdh8 ( Papc ) , structurally similar to Pcdh18 , controls C&E movements of the paraxial mesoderm in cooperation with the non-canonical Wnt pathway [66] , suggesting analogous roles of Pcdh18a/18b in the ectoderm . The findings of our current study thus demonstrate that the B1 SOX proteins regulate genes that are critical for a variety of processes in early embryonic development . This suggests that these factors serve as central coordinators of gene regulatory networks in the early developing embryo by coupling cell fate specification with patterning and morphogenetic processes . In transcriptional regulation , B1 SOX proteins likely perform this coordination by partnering with a variety of factors including Pou5f1 [21] . Our microarray analysis also suggests that B1 SOX regulate additional genes and pathways that we did not investigate herein . Future studies of these genes will therefore more fully delineate their multiple functions in coordinating early embryogenesis .
MOs were obtained from Gene Tools LLC ( OR , USA ) and are listed in Table S4 . Zebrafish embryos were obtained by natural matings of wild-type TL fish and reared at 28 . 5°C in 0 . 03% Red Sea salt solution . Approximately 1 nl of solution containing various combinations of MOs , as indicated in the figures , was injected into 1-cell stage embryos . Unless otherwise noted , a 1∶1 mixture of two MOs ( 0 . 9 ng/nl each ) was used to knockdown individual B1 sox genes . To knockdown multiple B1 sox genes , the MOs were each mixed at a concentration of 0 . 9 ng/nl and injected into 1-cell stage embryos . To estimate the knockdown efficiency of B1 sox MOs , fusion mRNAs of sox3-luc , sox19a-luc and sox19b-luc were prepared by transcription of template vectors , in which the 5′-UTR sequence and a short stretch of the amino-terminal-coding sequence of the respective genes ( −131 to +32 of sox3; −151 to +32 of sox19a; −151 to +32 of sox19b ) were inserted upstream of the luciferase sequence as previously described [20] . mRNA were microinjected with the relevant MOs into embryos and luciferase activities expressed in the embryos at 10–11 hpf were measured as described [20] . Whole-mount in situ hybridizations of zebrafish embryos were performed as described previously [11] . The genes analyzed are listed in Table S1 . Total RNAs for RT-PCR and microarray analyses were prepared from 40–100 uninjected control embryos and an equivalent number of embryos injected with MOs and/or mRNAs using RiboPure kit ( Ambion , TX , USA ) . 200 ng of total RNA of each sample was reverse-transcribed using oligo-dT primer and Superscript III RT-PCR system ( Invitrogen , CA , USA ) and a 1/80 fraction of the cDNA was used for PCR templates . PCR was performed using ExTaq polymerase ( Takara , Japan ) in 25 µl ExTaq buffer containing 5% dimethyl sulfoxide , 0 . 17 mM cresol red , 10% sucrose and primers listed in Table S5 . The PCR temperature profile consisted of 5 min denaturation , 23–30 cycles of a 30 sec denaturation at 94°C , 30 sec annealing at 57–61°C and 15 sec primer extension at 72°C and lastly a 10 min extension at 72°C ( see Table S5 for cycle numbers and annealing temperatures ) . PCR products were separated in a 2% agarose gel ( 1 . 5% Methaphor agarose/0 . 5% agarose ) and stained with SYBR green I . For microarray analysis , cRNA probes were prepared using 4 µg total RNA with a one-cycle cDNA synthesis kit ( Affymetrix , CA , USA ) . Affymetrix Zebrafish Genome arrays were hybridized with 10 µg cRNA probes , and posthybridization staining and washing were performed according to the manufacturer's instructions . RNAs from two independent samples were analyzed for each embryonic stage and the data were processed using the RMA program . Fold changes of the averaged hybridization signals between control and QKD embryo samples were then determined ( Figure S5 , Table S2 and Table S3 ) . The microarray data have been deposited in the Gene Expression Omnibus ( GEO , http://www . ncbi . nlm . nih . gov/geo ) at the National Center for Biotechnology Information with the accession number GSE18830 . Western blotting with an anti-SOX2 antibody was carried out as described previously [20] . For the detection of SOX3 , SOX19A and SOX19B , an anti-SOX3 C-terminal peptide antibody [15] was used . As a loading control , the blotted PVDF membranes were stained with Coomassie Brilliant Blue R-250 . The coding sequences of the B1 sox genes and their derivatives were cloned into the pCBA3 vector [20] . The coding sequence of bmp7 was amplified by RT-PCR and cloned into pCBA3 . The cDNA clone cb670 ( Zebrafish International Resource Center [ZIRC] , OR , UAS ) was used as a template for bmp2b mRNA . mRNAs were transcribed in vitro from linearized vectors using the mMessage mMachine SP6 kit ( B1 sox and bmp7 ) or mMessage mMachine T7 Ultra kit ( bmp2b ) ( Ambion ) . For the rescue experiments , each mRNA was mixed with the MOs for QKD and injected into 1-cell stage embryos . ChIP was carried out as described previously [67] with minor modifications . Briefly , zebrafish embryos at the 70–80%E and tailbud to 2-somite stages were enzymatically dechorionated with Pronase and then fixed in 1% formaldehyde in embryo medium for 15 min at room temperature . For each immunoprecipitation experiment , approximately 200 fixed embryos were homogenized in cell lysis buffer and incubated for 15 min on ice . Nuclei were collected by centrifugation , resuspended in 200 µl of nuclei lysis buffer , incubated for 10 min on ice and then sonicated using Bioruptor ( Cosmo Bio , Japan ) to yield DNA fragments with an average size of 400–500 bases . The supernatant of the sonicated cells was diluted 10-fold with ChIP dilution buffer ( 50 mM Tris-HCl [pH 8 . 0] , 167 mM NaCl , 1 . 1% Triton X-100 , 0 . 11% sodium deoxycholate ) . 950 µl of the diluted lysate was then incubated overnight at 4°C with Protein G Dynabeads ( Invitrogen ) that had been prebound to 2 µg of anti-SOX2 antibody ( AF2018; R&D , MN , USA ) . The same volume of the lysate was precipitated with normal goat IgG as a negative control . 100 µl of the lysate was used as an input control . Beads were washed four times with RIPA buffer and once with TE buffer containing 50 mM NaCl . Bound complexes were eluted from the beads and cross-links were reversed in 200 µl of elution buffer for six hours at 65°C . Eluted DNA was then purified by treatment with RNase A , followed by proteinase K digestion , phenol∶chloroform∶isoamyl alcohol extraction and ethanol precipitation . Precipitated DNA was resuspended in 30 µl of TE buffer and 1 µl of the DNA suspension was used as a template for ChIP-PCR , which was performed using ExTaq polymerase ( Takara ) in 20 µl ExTaq buffer containing 0 . 17 mM cresol red , 10% sucrose and the primers listed in Table S6 . PCR products were separated in a 2% agarose gel ( 1 . 5% Methaphor agarose/0 . 5% agarose ) and stained with SYBR green I . The fragments of the zebrafish hesx1 promoter ( Zv8_NA6682:4150–4984 , 835 bp ) , cyp26a1 promoter ( Zv8_chr12:9333416–9335090 , 1 . 7 kbp ) and pcdh18a conserved upstream sequence ( Zv7_chr1:9017351–9017762 , 412 bp ) were amplified by PCR from the zebrafish genome using following primers ( linker sequences that incorporate restriction sites are indicated by lowercase ) : hesx1 promoter , gggagatctCGTCAAACTCTCCAAACGTGGAT and ggggtcgacCTCAAGTCCTTTAATTTAACTCCAACTG; cyp26a1 promoter , gggagatctAGTATTCCCCGTCCCATTGC and ggggtcgacGTTGAAGCGCGCAACTGATC; and pcdh18a , gggatgcatAAGGCCCGTCCCAACTGAGGG and gggagatctCTACGTCTCAATCTCCCTGACAGA . The luciferase reporter vectors were constructed using pTK200-Venusluc/ISceI , which was generated by inserting an I-SceI site downstream of the reporter poly ( A ) sequence of pTK200-Venusluc [35] . The fragments of the hesx1 promoter and cyp26a1 promoter were inserted into the upstream region of the Venusluc sequence by replacing the TK promoter sequence . The pcdh18a-TK-Venusluc vector was constructed by inserting the 412-bp pcdh18a sequence upstream of the TK promoter . To perform the luciferase assay , 17 . 5 pg of the Venusluc vectors and 2 . 5 pg of the reference Renilla luciferase vector phRG-TK ( Promega , WI , USA ) were co-injected into 1-cell stage zebrafish embryos . Half of the embryos were then subsequently injected with the MOs for B1 sox QKD to assess their effects upon the luciferase expression . Injected embryos were collected at the tail bud stage and luciferase assays were performed as described previously [20] . The NES30-TK200-nlsVenus/ISceI transgene was constructed by inserting the octamerized NES30 sequence , which is a 30-bp nestin enhancer core sequence composed of SOX and POU binding sites [15] , into pTK200-nlsVenus/ISceI . pTK200-nlsVenus/ISceI is identical to pTK200-Venusluc/ISceI except that nlsVenus [35] was used as the reporter . A transgenic line was produced by the I-SceI meganuclease method as previously described [35] . | In the developing embryo , various processes such as cell fate specification , embryo patterning , and morphogenesis take place concurrently . The embryo must control gene expression in order to coordinate these processes and thereby enable the proper organization of its structures . The B1 sox transcription factor genes , exemplified by the “stem cell gene” sox2 , are thought to play a key role in these embryonic processes from the blastoderm stage to the neural stage . However , the precise regulatory functions of these genes are largely unknown due to the lack of loss-of-function studies . In our current study , we took advantage of the zebrafish system and successfully depleted B1 sox activity from the early embryo using antisense knockdown technology . This approach enabled us to further uncover the regulatory functions of B1 sox in early embryos . We found that the activity of the B1 sox genes is required for the expression of a wide range of developmental regulators including transcription factors , signaling pathway components , and cell adhesion molecules . These findings suggest that the B1 sox functions are central to coordinating diverse embryonic processes , particularly those that occur during the development of the primordium of the central nervous system . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology/stem",
"cells",
"developmental",
"biology/morphogenesis",
"and",
"cell",
"biology",
"genetics",
"and",
"genomics/gene",
"expression",
"developmental",
"biology/pattern",
"formation",
"developmental",
"biology/cell",
"differentiation",
"neuroscience/neuro... | 2010 | B1 SOX Coordinate Cell Specification with Patterning and Morphogenesis in the Early Zebrafish Embryo |
Yellow fever ( YF ) , transmitted via bites of infected mosquitoes , is a life-threatening viral disease endemic to tropical and subtropical regions of Africa and South America . YF has largely been controlled by widespread national vaccination campaigns . Nevertheless , between December 2015 and August 2016 , YF resurged in Angola , quickly spread and became the largest YF outbreak for the last 30 years . Recently , YF resurged again in Brazil ( December 2016 ) . Thus , there is an urgent need to gain better understanding of the transmission pattern of YF . The present study provides a refined mathematical model , combined with modern likelihood-based statistical inference techniques , to assess and reconstruct important epidemiological processes underlying Angola’s YF outbreak . This includes the outbreak’s attack rate , the reproduction number ( R 0 ) , the role of the mosquito vector , the influence of climatic factors , and the unusual but noticeable appearance of two-waves in the YF outbreak . The model explores actual and hypothetical vaccination strategies , and the impacts of possible human reactive behaviors ( e . g . , response to media precautions ) . While there were 73 deaths reported over the study period , the model indicates that the vaccination campaign saved 5 . 1-fold more people from death and saved from illness 5 . 6-fold of the observed 941 cases . Delaying the availability of the vaccines further would have greatly worsened the epidemic in terms of increased cases and deaths . The analysis estimated a mean R 0 ≈ 2 . 6 - 3 . 4 and an attack rate of 0 . 09-0 . 15% ( proportion of population infected ) over the whole period from December 2015 to August 2016 . Our estimated lower and upper bounds of R 0 are in line with previous studies . Unusually , R 0 oscillated in a manner that was “delayed” with the reported deaths . High recent number of deaths were associated ( followed ) with periods of relatively low disease transmission and low R 0 , and vice-versa . The time-series of Luanda’s YF cases suggest the outbreak occurred in two waves , a feature that would have become far more prominent had there been no mass vaccination . The waves could possibly be due to protective reactive behavioral changes of the population affecting the mosquito population . The second wave could well be an outcome of the March-April rainfall patterns in the 2016 El Niño year by creating ideal conditions for the breeding of the mosquito vectors . The modelling framework is a powerful tool for studying future YF epidemic outbreaks , and provides a basis for future vaccination campaign evaluations .
Yellow fever ( YF ) is a life-threatening viral disease endemic to tropical regions of Africa and South America . The disease is transmitted in urban areas primarily via the bites of infected female Aedes aegypti mosquitoes , which is also the vector of dengue , chikungunya and Zika viruses [1–3] . Rural and intermediate YF are transmitted by sylvatic and peri-domestic Aedes species in Africa . For those infected with YF , the disease incubates in the first 3-6 days of onset , after which there is an abrupt “period of infection” of intense viremia lasting for 3-4 days ( fever , weakness , headache , nausea , muscle pain ) [4] . This is followed by a period of remission in which the symptoms reduce and settle , and most infected individuals recover at this stage . Thus , some 70-85% of YF infections are asymptomatic or have at most very mild symptoms ( i . e . , clinically inapparent ) . However , 15-25% of patients relapse and move to a “period of intoxication” characterized by abdominal pain , vomiting , jaundice ( yellow skin and eyes ) and often culminating in death . The case-fatality-ratio ( CFR ) in this latter subset is understood to be approximately 20% among the general population , and 50% among hospitalized cases [4] , although the CFR is well known to be highly variable , and dependent on the particular circumstances . Like Ebola , YF is classified as a viral hemorrhagic fever , although it is responsible for a 1000-fold more illness and death than Ebola [1] . In 2016 , YF resurged in Angola to become the largest YF outbreak on record over the last 30 years [5] . In swift response , almost all global stocks of the YF vaccine were exhausted by April 2016 . Similar to the Angolan experience , YF recently resurged in Brazil in December 2016 , where it continues to expand towards the Atlantic coast in regions not previously deemed at risk ( as of March 16 , 2017 ) [2] . Thus there is an urgent need to gain a better understanding of the transmission patterns of YF . Here we develop a mathematical model to help identify the key epidemic processes behind the Angolan outbreak in 2015-16 , and the impact and effectiveness of the vaccination campaign . The first cases of YF in Angola were seen on December 5 , 2015 but reported in the media only on January 20 , 2016 [6] . By November 2016 , the large YF epidemic of Angola and Democratic Republic of Congo , resulted in 962 confirmed infections including 393 reported deaths [2] . YF is vaccine-preventable and the vaccine can confer long-lasting immunity . The vaccine is suitable for individuals of age 9 months or older . As such , the Angolan government initiated a vaccination campaign to prevent the spread of yellow fever on the first week of February , 2016 [2 , 7] . More than 10 million doses were needed for the whole country [6] . The center of the outbreak was in Angola’s capital , Luanda province . Estimates suggest that vaccination coverage of Luanda province was 38 . 0% at the end of January 2016 , and reached 80 . 0% by mid-March 2016 , and 93 . 0% by mid-June 2016 [2 , 8–10] . Fig 1 graphs the epidemic curve of YF case numbers ( probable and confirmed; as defined in Data section ) in Luanda province as obtained from the WHO [8 , 10] . The graph peaks in February 2016 , when large-scale vaccination was introduced , and then followed by a period of rapid decline in case numbers . Despite the major vaccination effort , the epidemic proved tenacious rather than die out as predicted , and persisted for a sustained period of time forming a long “tail” in reported case numbers from April to August ( see Fig 1 ) . Also unusual is the minor peak in case numbers that occurred in May , followed soon after by an increase in deaths , despite the pressure of the vaccination and control efforts . By modelling and fitting YF time series of Luanda , our goal is to reconstruct the important epidemiological processes that help explain these different and sometimes nonintuitive features . The model allows estimation of the attack rate of the outbreak , and the basic reproduction number ( R 0 ( t ) ) , which was changing during the epidemic . Moreover , the model is able to explore the role of the mosquito vector , and the unusual waves of the YF outbreak , which we find would have become even more apparent had there been no vaccination . Some exploration of the role of climatic variables is also possible . As it is well known , the basic reproduction number ( R 0 ) is an important parameter to measure diseases’ transmissibility , and is one of the first parameters that needs to be estimated in any epidemiological study . Recall that R 0 is defined as the number of secondary cases a single typical infected individual infects over the course of its infectious period [11] . A recent study estimated R 0 to lie between 5 . 2-7 . 1 at the early stage of the 2016 YF outbreak in Angola [12] . However , R 0 was found to decrease with time as the epidemic proceeded . Kraemer et al . [13] estimated R 0 to be 4 . 8 ( 95% C . I . : 4 . 0-5 . 6 ) for Angola , although this was possibly an over-estimate given reporting rates were not stable . In summary , the literature suggests that YF is highly transmissible with direct estimates of the reproduction number being R 0 ≈ 5 , which is almost double that of pandemic influenza ( R 0 is from 1 . 5 to 3 . 6 [14–17] ) and Ebola ( R 0 is from 1 . 2 to 2 . 0 [18–21] ) . In this work , our analysis uses modern statistical inference techniques to estimate R 0 from the time-series in Fig 1 . Unlike other modelling studies , our procedure also examines how reactive protective behavior ( e . g . , insecticide , vector-control , travel restrictions possibly in response to news and media precautions ) , may lead to changes in R 0 , and allows us to explore the implications of this reaction . Any model of YF must take into account that most infected individuals are asymptomatic or mild-symptomatic ( individuals who show only fever but no jaundice ) [3 , 22–26] , making the disease difficult to detect and under-reported in the first phase . With only a slight abuse of terminology , it simplifies the modelling that follows , to classify mild symptomatic individuals as asymptomatic cases . Thus asymptomatic cases refer to all individuals who do not have severe YF ( i . e . , without clinically apparent symptoms ) . It is well understood that asymptomatic YF infections can be infectious and therefore may act as “silent sources” of YFV [23] . Asymptomatic infections , thus , have the potential to play an important role in disease transmission . It was previously understood that 6 out of 7 YF infections could be asymptomatic [26] . However , a recent meta-analysis based on 11 independent studies , suggested that the asymptomatic ratio should be 55% [27] . Given the lack of information on the proportion and infectivity of asymptomatic YF cases , we examine a number of different relevant scenarios . To the best of our knowledge , this is the first detailed modelling of YF that includes both the host and vector populations , and the asymptomatic and severe ( those exhibiting fever and jaundice ) cases in the host population . Previous models that assessed vaccination impact on YF have not included these fundamental components and pathways in a comprehensive approach . By fitting the time-series of the Angola outbreak , its evolution over time and its curtailment with vaccination , it becomes possible to statistically infer key model parameters . This in turn makes it possible to simulate alternative “what if” scenarios , and examine what might have happened under different vaccination schemes .
We study time-series of YF cases from the province of Luanda of Angola with a population of 6 , 543 , 000 in 2016 [2 , 8 , 28] . The African Health Observatory ( AHO ) published weekly YF data for Luanda province reporting 941 ( confirmed and probable ) cases and 73 deaths over the study period from December 5 , 2015 to August 18 , 2016 . Probable cases ( see [29] ) are those “with acute onset of fever , with jaundice appearing within 14 days of onset of the first symptoms and one of the followings: ( i ) presence of yellow fever IgM antibody in the absence of YF immunization within 30 days before onset of illness; or ( ii ) positive postmortem liver histopathology; or ( iii ) epidemiological link to a confirmed case or an outbreak . ” Confirmed cases are defined as those positive to serological or PCR testing . Similar to the WHO [8] and Kraemer et al . [13] , both ( weekly ) probable cases and confirmed cases are grouped together and are referred to simply as “YF cases” or equivalently “severe cases” in this study . YF vaccination coverage in Luanda province , obtained from AHO reports , increased from 38% on February 2 , 2016 when the vaccination campaign started to 93% on August 18 , 2016 ( see Fig 1 ) [8] . The vaccination coverage was determined by a linear interpolation of reported data ( see the blue dotted line in Fig 1 ) . The YF outbreak in Luanda is modeled as a Partially Observed Markov Process ( POMP ) and makes use of the Iterated Filtering and plug-and-play likelihood-based inference frameworks to fit the data [34 , 49 , 50] . These are modern state-of-the-art statistical methodologies developed for fitting complex epidemiological datasets . The Maximum Likelihood Estimate ( MLE ) for model parameters is calculated using R package “POMP” [51] . Bayesian Information Criterion ( BIC ) is employed as a criterion for model comparison , and quantifies the tradeoff between the goodness-of-fit of a model and its complexity [52] . The simulations made use of the Euler-multinomial integration method with the time-step fixed to be one day [49 , 53] . The model is first fitted to the observed YF cases and deaths , given knowledge of the true vaccination coverage . The mosquito abundance is assumed to be unknown but time-dependent , and is reconstructed . We allow the basic reproduction number of our model to be time-dependent , given that the mosquito abundance is not fixed and human behavior can impact R 0 ( t ) and change over the study period . The parameter fitting and inference process are carefully checked , thereby giving high confidence that the fits of the observed time-series are accurate for reasons that are consistent with the true underlying epidemiological processes rather than artificial model over-fitting . We conducted tests to find the best-fit model . For each asymptomatic scenario , we studied 10 different values of nm ( degrees of freedom in the m ( t ) ) , and compared them with BIC . BIC quantifies the trade off of the goodness of fitting of the model and the complexity of the model—penalizing models with more variables . A smaller BIC implies a better-fit model . For the best-fit model , the profile of maximum log likelihood was calculated as a function of the reporting ratio ( see S2 Text in SI for further details ) . The profile found is always a reasonably smooth function . The model was run 1 , 000 times with the estimated parameters , and the median of the model simulation matched the reported weekly cases . Therefore , we can be confident that the maximization of model’s log likelihood converged and the estimation is consistent . The simulated weekly reported cases Zt are modelled by Eq 1 . The corresponding weekly observed cases , Ct , as given by the WHO , are assumed to follow a Negative-Binomial ( NB ) distribution as C t ∼ NB ( n = 1 τ , p = 1 1 + τ Z t ) with mean : μ t = Z t , ( 7 ) where τ denotes an over-dispersion parameter that needs to be estimated . The weekly observed deaths , Dt , and the corresponding weekly simulated deaths , Yt , are similarly related . Finally , the overall log-likelihood function , l , is given by l ( Θ | C 1 , … , C N ; D 1 , … , D N ) = ∑ t = 1 T ln [ L t ( C ) · L t ( D ) ] , ( 8 ) where Θ denotes the parameter vector under estimation , and L t ( C ) and L t ( D ) are the probability measurement functions associated with Ct vs . Zt , and Dt vs . Yt , respectively . T denotes the total number of weeks during the study period . The confidence intervals ( C . I . ) of parameters are estimated based on parameters ranges in Table 1 , using the method of profile likelihood confidence intervals [35 , 49] . This is demonstrated in S2 Text for the severe case reporting rate ρ . Parameter estimation and statistical analysis are conducted using R ( version 3 . 3 . 3 ) . The Partial Rank Correlation Coefficients ( PRCCs ) are adopted for the model’s sensitivity analysis [34] . Firstly , 1 , 000 random samples are taken for each model parameter from uniform distributions with parameter ranges as set out in Table 1 . After that , for every random parameter sample set , the YF model was simulated to obtain the target biological quantities ( e . g . , R 0 and total number of deaths in this study ) . Finally , PRCCs were calculated between each parameter and target biological quantities .
The results for the best-fitting model under the two scenarios ( i . e . , weak and strong infectivity scenarios ) are shown in Fig 3 . The model simulation median ( of 1 , 000 simulations ) of YF cases in Luanda is plotted in red and matches well the observed patterns seen in weekly reported cases , both before and after the national vaccination campaign . The two scenarios ( for asymptomatic infectivities ) both model the data with almost the same goodness-of-fit with a ΔBIC ≈ 2 ( see S2 Text for the simulation results of strong infectivity scenario , i . e . , scenario 2 ) . That is , the observed and model time series are not significantly different for the two levels of infectivity [52] . As such the infectivity of asymptomatic cases cannot be accurately inferred from these data sets . In Table 2 , the over-dispersion τ , is notably small , which indicates the measurement model is close to a Poisson distribution ( i . e . , minor over-dispersion in measurement noise ) . This implies the reporting efforts ( i . e . , reporting ratios ) were reasonably stable over time . The analysis estimated a mean R 0 ≈ 2 . 6 - 3 . 4 and an attack rate of the whole period to be 0 . 09-0 . 15% ( % population infected by YF ) from December 2015 to August 2016 . Our estimated initial and upper bound R 0 are in line with previous studies . Asymptomatic cases were not reported , and they might be considered as a completely hidden variable . However , if the number of asymptomatic cases is very large ( e . g . , if the asymptomatic-to-symptomatic ratio is 6:1 or 7:1 ) with a weak infectivity but full immunity , this will indirectly slow down the transmission of YF in the later stages , due to herd immunity built up by these silent asymptomatic cases . If their infectivity is strong , this will increase the difficulty to control the outbreak . The model simulations of weekly deaths also fit the observed data well over the period of the main epidemic until the end of April 2016 . While the simulated median ( red line ) does not predict the two relatively small and erratic peaks at the beginning of June and end of July , nevertheless they fit reasonably within the 95% bounds . ( Note that similar peaks in death numbers appear in the delayed vaccination scenarios Fig 5b , 5d and 5f , where case numbers are higher . See next section . ) The observed YF deaths are relatively “noisy” compared to the continuously observed YF cases ( see red bars versus green dotted line in Fig 1 ) , which might be due to the lower case numbers involved or possibly spatial variation of the YF CFR . The same holds for individual model simulations . The total number of deaths was only 6% out of all reported cases ( i . e . , CFR = 6% ) , and 71% of the deaths appeared during the first wave . Although we cannot fit the final erratic mortality waves with high accuracy , our estimate of the total number of deaths is still a very good approximation . As can be seen in Table 3 , the model’s simulated cumulative death toll matches well with the observed death toll . Parameter estimates including the basic reproduction number R 0 , mean mosquito-to-human ratio 〈m ( t ) 〉 , and disease attack rate are listed in Table 2 . Our estimated mean R 0 ≈ 3 . 0 with excursions to R 0 ≈ 6 . 0 matches well other studies in the literature ( see Introduction ) . The very low attack rate is an outcome of the prompt and efficient control measures by the Angolan government [2 , 6 , 8] . The estimated mean mosquito-to-human ratio , 〈m ( t ) 〉 , is in line with previous work by Gao et al . ( see [54] ) . The estimated reporting ratio for severe cases , ρ = 70% , is reasonable , given the easily recognizable symptoms ( jaundice ) , and the control effort by the government , which managed to push the vaccination coverage to more than 90% of its population within a very short period of time . Association between the spread of YF and local climatic factors has been discussed frequently in previous studies [55 , 56] . As such , we explored the possibility that local temperature and rainfall are potential factors that consistently influence the long-term transmission dynamics . Temperature was found to have no significant effect while the effects of rainfall were significant but quantitatively mild in a first analysis . The possible reasons could be: i ) local precipitation is relatively minor during the study period but concentrated in March , and the weather is continuously hot and dry; and ii ) duration of the outbreak is short and other factors ( control measures and human reaction ) played a more prominent role . Nevertheless , in what follows , we consider the possibility that the March El Niño rainfall patterns played an important role which is difficult to untangle from our analysis . A discussion is given in S6 Text . For most years in Luanda , the rainy season is between November-May but the most accumulation of rain occurs in March-April [61] . The year 2016 was an El Niño year and it brought dramatic and unpredictable flooding events especially in the March-April period , thereby leading to conditions ideal for growth in mosquito populations . As in 1971 , YF outbreak in Luanda , local water-storage containers ( mainly the larger ones ) serving the community but also in most homes , accounted for 85% of the Ae . aegypti larval breeding sites [61 , 62] . As vividly described by Moreira [63]: “The 2016 outbreak coincided with unusually heavy rains and a severe El Niño weather pattern . We are also suffering from an economic crisis and poor sanitary conditions . All these factors created a fertile environment for an increase in the mosquito population . The outbreak reached its peak in February and has been declining since ( i . e . , population numbers , not R 0 ) . We have much more vaccine now ( in September 2016 ) than we had earlier in the epidemic . The response interventions are involving communities successfully . The dry season arrived in May ( 2016 ) and since then the mosquito population has diminished . ” Thus after the peak of the YF outbreak had passed , and the vaccination program was in progress , the local March-April El Niño rains were enhancing mosquito breeding conditions . It is surprising that simultaneously one of Luanda’s largest malaria epidemics ever was underway ( “During the first quarter of 2016 , the number of cases of malaria increased dramatically to 1 , 531 , 629 , up from 980 , 192” [64] ) . We suggest that these conditions may also be responsible for the unusual but robust second wave that is observed in the time series of R 0 ( t ) . Since we only model a single province , Luanda , and the YF transmission spread relatively rapidly throughout the province , we might assume the effects of spatial heterogeneity are likely to be minimal . However , we do not possess sufficiently detailed data to perform a careful analysis of spatial effects , and effects at the micro-scale may be important as in other diseases such as dengue [65] . Kraemer et al . [13] have discussed the importance of spatial effects for YF over all provinces of Angola , and there is a possibility that geographic waves generated from surrounding provinces could play some part in the appearance of multiple YF waves in Luanda province . Hence future work and more comprehensive data are needed to examine these possibilities . There are many possible ways to evaluate the effects of a delayed vaccination campaign when compared to the baseline scenario that was implemented in practice in Luanda . The approach followed here is to simply delay the exact same baseline scenario ( in terms of doses per week ) by a fixed time interval until the end of the observation period arrives . It is difficult to extend beyond the observation period without introducing an unacceptable rate of errors . This can be seen in the large confidence intervals for R 0 towards the end of the observation period ( see S4 Text ) . The results of 60 , 120 , and 180 days delay of the vaccination campaign for the 2016 yellow fever outbreak are presented in Fig 5 for the scenario 1 ( ψ = 0 . 1 ) . The total reported cases and total deaths are calculated for four vaccination scenarios ( including the baseline ) and outcomes are listed in Table 3 . The 180-day delay is included , because it gives an impression of what might happen when vaccination is unavailable , as mentioned . The baseline scenario ( actual vaccination or 0-delay ) results in an estimated 73 deaths associated with YF in the study period , which matches the observed number . With a 60-day delay to the vaccination roll-out , YF deaths saved were 2 . 2-fold of the observed number ( see Fig 5a and 5b ) . With a 120-day delay , the YF death saved were 4 . 5-fold of the observed number ( see Fig 5c and 5d ) . With a 180-day delay , YF deaths saved were 5 . 1-fold of the observed number ( see Fig 5e and 5f ) . The latter result is a good approximation to what might have occurred if there were no vaccination campaign in Luanda up to August 2016 . All of these results show that delaying the vaccination campaign would have greatly enhanced the epidemic in terms of infectious cases and mortality . We also investigated the “vaccination delay” situation under different scenarios ( see S2 Text ) , and found our main results of “deaths prevented” largely holds . In addition , we also considered the scenario of “what if deaths were under reported” ( i . e . , there was a constant proportion of YF deaths not reported , see S7 Text for details ) , we report our main results are also robust . A clear feature of the simulated outcomes with delayed vaccination ( red lines in Fig 5 ) , is the noticeable second wave of YF cases and deaths that appear . This feature becomes even more prominent in a situation of no vaccination ( see Fig 5e and 5f ) . Returning to Fig 1 , we also see strong signs from the observed time series of YF in Luanda , that the outbreak may have indeed occurred over two waves . Hence , even with Luanda’s large-scale vaccination campaign , the multiple-wave feature is noticeable in the observed time series , which implies considerable fluctuations in the driving force ( R 0 ) or other factors . Results of sensitivity analysis are presented in Fig 6 and indicate how model parameters impact the basic reproduction number R 0 and the death toll . R 0 is most sensitive to vector biting rate ( a ) and the vectors’ lifespan ( μ v - 1 ) , indicating the importance of the mosquitoes’ role in disease transmission . The total deaths are considerably sensitive to the proportion of severe cases ( δ ) , the case-fatality rate of severe cases ( θ ) and the initial number of susceptibles ( i . e . , the ratio Sh . 0/Nh ) . Using modern likelihood based statistical inference techniques , it was possible to fit a vector-host epidemic model successfully to the surveillance data collected for the YF outbreak in Luanda , Angola in 2016 . We were thus able to assess the success of the vaccination campaign as rolled out in Luanda . While there were in reality 73 deaths reported over the 37-week study period , the model showed that the vaccination campaign saved from death 5 . 1-fold of observed deaths and prevented from illness 5 . 8-fold of observed cases , over the study period , and no doubt many more if we were to extrapolate beyond the study period . This was determined by simulating Luanda’s YF outbreak in the absence of any vaccination . The national vaccination campaign was also found to be timely , in that delaying the availability of the vaccination any further would have greatly enhanced the epidemic in terms of number of YF cases and mortality . The change in the number of YF cases over time in Luanda suggests the possibility that the outbreak occurred in two waves over the 37-week study period . The modelling and sensitivity analysis demonstrated that this is a robust feature ( see S5 Text ) , which would have become far more prominent had the vaccination campaign been reduced in intensity . The appearance of waves implies that R 0 must oscillate to some degree in time . Reconstruction of the underlying dynamics reveals that R 0 is strongly out of phase with mortality , so that R 0 decreases when the number of deaths increase , and vice versa . Thus we hypothesize that the high death rates and number of cases influenced Luanda’s population behavioral response which in turn led to some reduction of disease transmission during the years of high mortality . Behavioral responses may typically involve using more insecticide , mosquito repellent , insecticide-treated bednets and broader vector-control programs , as outlined in S8 Text . In Luanda , it also involved cordon sanitaire , and movement restriction with the aim of reducing transmission through the wider population [66] . Such behavioral changes are able to modulate the basic reproduction number , which in turn can lead to waves in the YF case numbers . A similar phenomenon was reported for the deadly pandemic influenza ( e . g . , 1918 influenza pandemic with a fatality rate ( per infection ) 2% and an attack rate 1/3 ) [58 , 59] but never in mosquito-borne disease since either the CFR or the AR is typically low . Moreover , we showed how a simpler model that explicitly incorporated human behavior reproduces the observed data in Fig 1 ( see S5 Text ) . This may be the first example of mortality-driven basic reproduction number in a mosquito-borne disease outbreak . While this possibility appears to hold in other epidemiological contexts ( e . g . , Spanish flu [58 , 59] ) , it would be beneficial to check this further for vector-host systems . In the case of Luanda’s YF outbreak , R 0 is likely to have also been affected by the sporadic but heavy El Niño rainfall , which in turn could influence mosquito population numbers . Such a process could occur even if there is no visible long-term correlation between climate ( rainfall ) and the vector dynamics . The modelling approach described here provides a basis for future vaccination campaign evaluations . Since the YF mortality appeared to lead to oscillations in the basic reproduction number ( R 0 ) , this possibility should be considered in the development of short-term prediction tools of the spread of YF . The general approach should be of benefit in mitigating the spread and impact of YF outbreaks in the future . | An epidemic model for the transmission of yellow fever virus ( YFV ) in urban areas is formulated and implemented to study the 2016 yellow fever ( YF ) outbreak in Luanda , Angola . We explore the complex vector-host dynamics of this system taking into account mosquito abundance , vaccination and asymptomatic infections in the human population , that are generally not included in other modelling studies of YF . The model successfully fits the time series of weekly reported YF cases and deaths during the epidemic . This allows us to study the impact of the vaccination campaign in Luanda and hypothetical “delayed vaccination” scenarios . The transmission of YFV appeared to be oscillatory having a wave-like pattern in the basic reproduction number ( R 0 ( t ) ) . The oscillations are hypothesized to be due to human reaction to the reported deaths , as has been noted for other human infectious diseases , and the second wave also possibly due to El Niño rainfall patterns . We conclude that the lives saved due to the vaccination campaign before August 2016 should have been approximately 370 ( i . e . , approximately five-fold of the observed 73 deaths ) , and would have been far larger extrapolating beyond August 2016 . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"and",
"discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"atmospheric",
"science",
"angola",
"immunology",
"geographical",
"locations",
"animals",
"preventive",
"medicine",
"el",
"ni単o-southern",
"oscillation",
"infectious",
"disease",
"control",
"vaccination",
"and",
"i... | 2018 | Modelling the large-scale yellow fever outbreak in Luanda, Angola, and the impact of vaccination |
Despite the central role of alternative sigma factors in bacterial stress response and virulence their regulation remains incompletely understood . Here we investigate one of the best-studied examples of alternative sigma factors: the σB network that controls the general stress response of Bacillus subtilis to uncover widely relevant general design principles that describe the structure-function relationship of alternative sigma factor regulatory networks . We show that the relative stoichiometry of the synthesis rates of σB , its anti-sigma factor RsbW and the anti-anti-sigma factor RsbV plays a critical role in shaping the network behavior by forcing the σB network to function as an ultrasensitive negative feedback loop . We further demonstrate how this negative feedback regulation insulates alternative sigma factor activity from competition with the housekeeping sigma factor for RNA polymerase and allows multiple stress sigma factors to function simultaneously with little competitive interference .
Bacteria survive in stressful environmental conditions by inducing dramatic changes in their gene expression patterns [1 , 2] . For a variety of stresses , these global changes in gene expression are brought about by the activation of alternative σ-factors that bind the RNA polymerase core enzyme and direct it towards the appropriate stress response regulons [3] . Consequently , to ensure that these σ-factors are only active under specific environmental conditions , bacteria have evolved regulatory systems to control their production , activity and availability [3 , 4] . These regulatory networks can be highly complex but frequently share features such as anti-σ-factors , partner switching mechanisms and proteolytic activation [4] . The complexity of these networks has impeded a clear mechanistic understanding of the resulting dynamical properties . In this study , we focus on one of the best studied examples of alternative σ-factors , the general stress-response regulating σB in Bacillus subtilis [5] to understand how the structure of the σ-factor regulatory networks is related to their functional response . The σB-mediated response is triggered by diverse energy and environmental stress signals and activates expression of a broad array of genes needed for cell survival in these conditions [5] . Activity of σB is tightly regulated by a partner-switching network ( Fig 1A and 1B ) comprising σB , its antagonist anti-σ-factor RsbW , and anti-anti-σ-factor RsbV . In the absence of stress , RsbW dimer ( RsbW2 ) binds to σB and prevents its association with RNA polymerase thereby keeping the σB regulon OFF . Under these conditions most of RsbV is kept in the phosphorylated form ( RsbV~P ) by the kinase activity of RsbW2 . RsbV~P has a low affinity for RsbW2 and cannot interact with it effectively [6] . However , in the presence of stress , RsbV~P is dephosphorylated by one or both of the dedicated phosphatase complexes: RsbQP for energy stress and RsbTU for environmental stress [7–10] . Dephosphorylated RsbV attacks the σB-RsbW2 complex to induce σB release , thereby turning the σB regulon ON [11] . Notably , the genes encoding σB and its regulators lie within a σB-controlled operon [12] , thereby resulting in positive and negative feedback loops . Recently , it was shown that under energy stress σB is activated in a stochastic series of transient pulses and increasing stress resulted in higher pulse frequencies [13] . It has also been shown that increase in environmental stressor such as ethanol leads to a single σB pulse with an amplitude that is sensitive to the rate of stressor increase [14] . While it is clear that the pulsatile activation of σB is rooted in the complex architecture of its regulatory network ( Fig 1A and 1B ) its mechanism is not fully understood . Previous mathematical models of the σB network either did not produce the pulsatile response [15] or made simplifications to the network [13] that are somewhat inconsistent with experimentally observed details . As a result , it remains unclear which design features of the σB network enable its functional properties . To address these issues we develop a detailed mathematical model of the σB network and examine its dynamics to understand the mechanistic principles underlying the pulsatile response . By decoupling the post-translational and transcriptional components of the network we show that an ultrasensitive negative feedback between the two is the basis for σB pulsing . Moreover we find that the relative synthesis rates of σB and its operon partners RsbW and RsbV , plays a critical role in determining the nature of the σB response . We also use our model , together with previously published experimental data from [13 , 14] , to explain how the σB network is able to encode the rate of stress increase and the size of stochastic bursts of stress phosphatase into the amplitudes of σB pulses . We further develop this model to investigate how the network functions in the context of other σ-factors . As in many other bacteria , σB is one of the many σ-factors that complex with RNA-polymerase core that is present in limited amounts [3 , 16] . Therefore , when induced these alternative σ-factors compete with one another and the housekeeping σ-factor σA for RNA polymerase . We use our model to investigate how the design of this network enables it to function even in the presence of competition from σA which has a significantly higher affinity for RNA polymerase [17] . Lastly , we investigate how multiple alternative σ-factors compete when cells are exposed to multiple stresses simultaneously . Using our model we identify design features that are ubiquitous in stress σ-factor regulation and critical to bacterial survival under diverse types of stresses .
In a recent study , Locke et al . [13] demonstrated that a step-increase in energy stress results in pulsatile activation of σB . The study also proposed a minimal mathematical model of the network which reproduced pulsing in σB . However , this model included several assumptions inconsistent with experimentally observed details: ( i ) Phosphorylation and dephosphorylation reactions were assumed to follow Michaelis-Menten kinetics despite the fact that kinase ( RsbW ) and phosphatase concentrations are known to be comparable to substrate ( RsbV ) concentrations [18] so the approximation breaks down [19] , ( ii ) σB and RsbV are represented as a single lumped variable rather than separate species and , ( iii ) partner-switching , and the formation and dissociation of various RsbW2 complexes were not included explicitly . Though this minimal model produces pulses resembling their experimental observations , it does not depict a biochemically accurate picture of the σB network . Consequently it cannot be used to uncover the design features that enable σB pulsing . To understand the σB network response we built on our earlier study [15] to develop a detailed mathematical model that explicitly includes all known molecular interactions in the network . Note that we made one significant change to the model discussed in [15] . The model in [15] assumed that the synthesis rates for σB and its operon partners ( RsbW and RsbV ) follow the stoichiometry of their binding ratios ( i . e . RsbWT/BT = 2 and RsbWT/RsbVT = 1; where BT , RsbWT and RsbVT represent total σB , RsbW and RsbV concentrations respectively ) . However experimental measurements have shown that σB , RsbW and RsbV are produced in non-stoichiometric ratios [18] . The exact mechanism underlying these non-stoichiometric ratios is currently only incompletely understood . However , analysis of the open-reading frames in the operon showed that rsbV and rsbW may be translationally coupled due to overlapping termination and initiation codons [20] which may ensure that they are expressed in similar amounts . The same analysis also showed that the rsbW and sigB reading frames overlapped and that this overlap was preceded by a region of dyad symmetry which may form a stem-loop structure [20] . These features may interfere with sigB translation and lead to lower expression of σB than its binding partners RsbV and RsbW . To account for these features , in contrast to our earlier study , we assumed σB , RsbW and RsbV can be produced in non-stoichiometric ratios and studied how changes in relative synthesis rates of σB operon partners affect the response of the σB network to step-increases in energy stress phosphatase levels . We note that RsbX , a negative regulator of RsbTU phosphatase [21] , is not included in our model . RsbX was excluded for simplicity since it is not essential for the pulsatile response of the σB network [14] . Simulations of this detailed model showed that different combinations of RsbW:σB ( λW ) and RsbV:σB ( λV ) relative synthesis rates lead to qualitatively different dynamical responses of the σB network . For operon partner synthesis ratios similar to those estimated in [18] , ( i . e . RsbWT > 2BT and RsbWT ≈ RsbVT ) our model responded to a step-up increase of the phosphatase with a pulsatile σB response ( Fig 1C ) that resembled the experimentally observed behavior [13] . In contrast , when RsbW:σB and RsbV:σB relative synthesis rates follow the stoichiometry of their binding ratios pulsing is not observed and the σB activity monotonically increases over time ( Fig 1D ) . Pulsing also disappears when RsbW synthesis is high enough to neutralize both its binding partners ( Fig 1E ) . To understand why the pulsatile response is only observed for certain operon partner synthesis rates , we investigated our mathematical model by decoupling the network’s transcriptional and post-translational responses ( as shown in Fig 1A ) . By varying the σB operon transcription rate , while keeping the relative synthesis rates of RsbW:σB ( λW ) and RsbV:σB ( λV ) fixed , we were able to calculate the post-translational response ( Fig 2A , blue curve ) of the σB network: [σB] = Fp ( [BT] , [PT] ) . This function describes how the free σB concentration varies as a function of BT ( total concentration of σB ) and PT ( total phosphatase concentration ) . Note that although we refer only to BT for brevity , RsbWT and RsbVT are always assumed to increase in proportion to the BT for this post-translational response . This post-translational function is analogous to an in vitro assay wherein various combinations of total σB ( BT−and proportional amounts of RsbWT and RsbVT ) and total phosphatase ( PT ) are mixed together and then the resulting free σB concentration is measured . In parallel , we calculated the transcriptional response ( Fig 2A , black curve ) [BT] = FT ( [σB] ) which analogous to a transcriptional reporter construct in vivo , describes how changes in the free σB concentration affect total σB concentrations ( and RsbWT and RsbVT concentrations which are always proportional to BT ) . In this analysis framework , the steady state of the complete closed loop network can be determined by simultaneously solving the post-translational and transcriptional equations , [σB] = FP ( [BT] , [PT] ) and [BT] = FT ( [σB] ) at each phosphatase concentration PT . Graphing both functions provided the steady-state solution as their intersection point ( Fig 2A , red circle ) . This decoupling approximation allows us to quantify the sign and strength of feedback in the full model . The effective sign of the feedback in the σB network is given by the sign of the product of the sensitivities of two response functions , i . e . sign ( ( ∂FT / ∂[σB] ) · ( ∂FP / ∂[BT] ) ) . Since σ-factors function as activators of transcription , FT ( [σB] ) is a monotonically increasing function of σB ( i . e . ∂FT / ∂[σB] > 0 ) . Consequently , the sign of the feedback in the σB network is given by the sign of the sensitivity of the post-translational response to RsbBT ( i . e . ∂FP / ∂[BT] ) . In other words , if increase in the operon production leads to an increase in free σB then the feedback is positive , whereas if increase in the operon production leads to a decrease in free σB then the feedback is negative . Our results show that for the parameters chosen in Fig 1C FP is a non-monotonic function of BT ( Fig 2A , blue curve ) . At low RsbBT , free σB increases as a function of BT because RsbW is sequestered in the W2V2 complex . However at higher BT , the kinase flux dominates the phosphatase flux resulting in an increased RsbV~P and the freeing of RsbW2 from RsbV . Freed RsbW2 sequesters σB in the W2σB complex . Furthermore , in the total σB concentration range where ∂FP / ∂[BT] < 0 in Fig 2B , the post-translational response is quite steep ( Fig 2A ) , i . e . small changes in BT lead to significant decreases in free σB . This ultrasensitivity can be quantified by calculating the slope in logarithmic space , i . e . This dimensionless quantity characterizes the ratio of relative changes in σB and BT at steady state ( Fig 2B ) . The sign of LGP defines the effective sign of the feedback loop and if the magnitude of |LGP| > 1 defines an ultrasensitive response . For the σB network , in the region around the steady state LGP < −1 indicating that the σB network operates in an ultrasensitive negative feedback regime . Two types of post-translational reactions that are known to produce ultrasensitivity play a role here ( S1A and S1B Fig ) : ( 1 ) Zero-order ultrasensitivity due to competition between RsbW kinase and RsbQP/RsbTU phosphatases for RsbV and ( 2 ) molecular titration due to sequestration of σB by RsbW . Notably around the steady state , whereas both the fraction of unphosphorylated RsbV and the fraction of free σB decrease ultrasensitively as a function of increase in operon expression ( proportional to BT ) the latter is far more sensitive ( S1C Fig ) . This indicates that molecular titration between σB and its binding partners may contribute more to the ultrasensitivity of the post-translational response than the zero-order competition between RsbW and stress phosphatases . Irrespective of their relative contributions however , our results show that both mechanisms combine to ensure that near the steady state the σB network operates in an ultrasensitive negative feedback regime . Notably , negative feedback is one of the few network motifs capable of producing adaption-like pulsatile responses [22] . Moreover , ultrasensitivity of the feedback ensures homeostatic behavior—making the steady state robust to variations of parameters [22] . This explains why in Fig 1C a step-increase in the phosphatase concentration in our model leads to a σB pulse followed by return to nearly the same steady state . Plotting the trajectory of the σB pulse ( green curve , Fig 2C ) on the ( [σB] , [BT] ) plane and over the post-translational and transcriptional responses ( Fig 2C ) illustrates the mechanism driving this pulsatile response . Starting at the initial steady state ( red circle ) , an increase in phosphatase shifts the ultrasensitive post-translational response ( cyan to blue curve ) so that free σB is rapidly released from the RsbW2-σB complex whereas total σB levels remain relatively unchanged . The increase in σB operon transcription eventually causes accumulation of total σB and the anti-σ-factor RsbW . This in turn forces the σB level to decrease , following the post-translational response curve , to the new steady state ( gray circle ) which has very little free σB thereby completing the σB pulse . The same analysis can be applied for different values of relative synthesis rates , i . e . those that correspond to Fig 1D and 1E . As shown in S2 Fig these parameter values do not produce an ultrasensitive non-monotonic post-translational response . Consequently they do not lead to the emergence of overall negative feedback explaining their non-pulsing dynamics . To determine if the presence or absence of negative feedback more generally explains the different dynamical responses in Fig 1C–1E , we sampled different combinations of relative synthesis rates ( [RsbWT] / [BT] = λW and [RsbVT] / [BT] = λV ) and calculated the post-translational sensitivities . Our calculations showed that based on the sign of post-translational sensitivity ( LGP ) the relative synthesis parameter space can be divided into three regions ( Fig 2D ) . For ( λW , λV ) combinations in Region I the sensitivity is always positive . Increase in λW leads the system into an ultrasensitive negative regime ( LGP < 0 and |LGP| ≫ 1 ) in Region II . A further increase in λW or a decrease in λV transitions the system into a non-responsive ( LGP ∼ 0 ) state in Region III . Dynamic simulations for sampled ( λW , λV ) combinations confirm that pulsatile responses to step-up in phosphatase concentration are restricted to Region II where the effective feedback is negative ( S2 Fig ) . To understand the boundaries between the three regions and how the level of the phosphatase affects the network , we developed a simplified analytical model that is based on the observation that RsbW and RsbV bind strongly to each other [18] ( see S1 Text for details ) . This approximation allowed us to determine the boundaries in Fig 2D ( black and red lines ) and resulted in a clear biological interpretation of the three regions . In Region I the amount of RsbW , irrespective of phosphatase level , is insufficient to bind all of its partners and consequently some fraction of σB always remains free or unbound to RsbW . In contrast in Region II , the amount of phosphatase determines how much RsbV is in its inactive phosphorylated form RsbV~P and therefore whether the amount of RsbW is sufficient to bind all of its partners depends on the levels of RsbV~P . As a result , for this region , the ratio of kinase and phosphatase ( PT ) fluxes determines the post-translational response . Lastly , Region III is the opposite of Region I in that the amount of RsbW is more than sufficient to bind all of its partners , even when all RsbV is unphosphorylated . As a result , irrespective of phosphatase levels , very little σB is free and its level is nearly insensitive to changes in total σB . Thus negative feedback and consequently pulsing are only possible in Region II where changes in phosphatase can shift the balance between the prevalent partner complexes . The role of negative feedback in producing a pulsatile response also explains why pulsing does not occur in strains where σB operon is transcribed constitutively [13] . In this case , the σB network lacks the negative feedback necessary to produce a pulsatile response . A step-increase in phosphatase still leads to an increase in free σB due to the change in the post-translational response; however , this not followed by an increase in total σB levels ( S2C Fig ) . Consequently , an increase in phosphatase results in a monotonic increase in free σB rather than a pulse ( S2F Fig ) . The only actual measurements of λW and λV were made by Delumeau et al . [18] using a quantitative western blot assay . Interestingly they report that λW = 2 . 9 , λV = 1 . 7 in the absence of stress and λW = 2 . 4 , λV = 2 . 65 in the presence of stress . These measurements suggest that the ratios might change depending on whether cells are under stress . Although the mechanism underlying this change is unclear , our model predicts that both measured ratio pairs lie within the negative feedback regime shown in Fig 2D . Accordingly our simulations show that the network responds to step-increases in phosphatase levels with a pulsatile response for both pre-stress and post-stress ( λW , λV ) values ( S3 Fig ) . However , due to reduced ultrasensitivity of the system for these parameters , concertation of free σB following increase in the stress ( phosphatase ) does not perfectly adapt to the pre-stress value ( S3 Fig ) . In an attempt to match the near-perfect adaptation reported in Refs . [13 , 14] we’ve chosen to do further analysis with λW = 4 and λV = 4 . 5 . Notably , our simulations also showed that it is not essential for the phosphatase level to remain fixed after a stress-induced step-increase . In fact , we found that a dilution mediated decline in phosphatase level post-step-increase has little impact on the pulse amplitude ( S4 Fig ) . This observation can be explained by the relatively rapid dynamics of the post-translational response as compared to the gradual nature of dilution and suggests that pulsatile dynamics are relevant even for experimental conditions where phosphatase levels do not remain fixed in stressful conditions [14 , 23] . Further our decoupling method also sheds light on another experimental observation by Locke et al . [13]: the dependence of σB pulse amplitude on the phosphatase level . Specifically , we found that σB pulse amplitude is a threshold-linear function of the phosphatase concentration ( S5 Fig ) . Our decoupling method shows that this threshold-linear behavior arises because the σB network only operates in a negative feedback regime for phosphatase concentrations higher than a threshold . Below the phosphatase threshold , the post-translational response [σB] = FP ( [BT] , [PT] ) ∼ 0 and is insensitive to RsbBT ( S5B and S5C Fig ) . Thus , the full system lacks the negative feedback and as a result σB does not pulse . Using our analytical approximation we found that this phosphatase threshold is proportional to the basal level of RsbW kinase synthesis rate and the ratio of the kinase and phosphatase catalytic rate constants ( S5D and S5E Fig ) . Increase in the basal σB operon expression rate increases the phosphatase threshold . Further , an increase in the relative synthesis rate of RsbW ( λW = [RsbWT] / [BT] ) makes the phosphatase threshold more sensitive to the σB operon expression rate , whereas a decrease in ratio of the kinase and phosphatase catalytic rate constants makes it less sensitive ( S5D and S5E Fig ) . This shows that the phosphatase threshold represents the concentration at which the phosphatase is able to match the basal kinase flux . Altogether these results show how the ultrasensitive negative feedback plays a critical role in determining many properties of the σB network pulsatile response and how the decoupling method can facilitate the identification of essential design features that enable the existence of this negative feedback . In the preceding sections we have shown how the σB network responds to a step-increase in RsbQP or RsbTU phosphatases by producing a single pulse of activity . However , Locke et al . [13] have shown that an increase in energy stress leads to a sustained response with a series of stochastic pulses in σB activity . This study further showed that this sustained pulsing response is driven by noisy fluctuations in level of energy-stress-sensing phosphatase RsbQP . While the mean level of RsbQP is regulated transcriptionally by energy stress , its concentration in single cells can fluctuate due to the stochasticity of gene expression [8 , 13] . To determine if our model could explain this response to stochastic fluctuations in RsbQP , we modified it to include fluctuations in the concentration of this phosphatase . Based on previous theoretical [24 , 25] and experimental [26] studies we assume that fluctuating phosphatase level follows a gamma distribution which is described by two parameters—burst size ( b , average number of molecules produced per burst ) and burst frequency ( a , number of bursts per cell cycle ) . The mean phosphatase in this case is the product of burst size and burst frequency ( 〈PT〉 = ab ) . Thus , energy stress can increase mean phosphatase by changing burst size or burst frequency or both . In other words , stress conditions can increase phosphatase levels by either producing more phosphatase molecules per transcription-translation event or by making these events more frequent . While the results of [13] cannot exclude either mechanism , we can use our model to uncover which mechanisms is dominant . First , we performed stochastic simulations in which mean phosphatase concentration was varied by changing burst size . These simulations reproduced all the experimentally-observed features of the σB pulsatile response . Specifically our results show that stochastic bursts in stress phosphatase levels lead to pulses of σB activity ( Fig 3A ) . Moreover , consistent with the experimental observations of [13] , our model showed that the amplitude of σB pulses increases linearly with the stress phosphatase level ( Fig 3A and 3B ) . Finally , we found that stress-mediated increases in phosphatase concentration lead to an ultrasensitive ( effective Hill coefficient ~5 . 6 ) increase in the frequency of σB pulsing ( Fig 3C ) and an ultrasensitive ( effective Hill coefficient ~2 ) increase in the level of σB target expression ( Fig 3D ) . Next , we compared these results with stochastic simulations in which burst frequency was modulated ( Fig 3E–3H ) . These simulations also led to an increase in σB pulsing ( Fig 3E ) and a non-linear increase in the level of σB target expression as mean phosphatase level was increased with more frequent bursts ( Fig 3H ) . However , we found that σB pulse amplitude remains constant for burst frequency modulation ( Fig 3E and 3F ) unlike the ~5-fold increase for burst-size modulation ( Fig 3B ) . Moreover , the frequency of σB pulses increase linearly with phosphatase level unlike the non-linear increase observed with burst-size-increase simulations ( compare Fig 3C and 3G ) . Notably the experimental observations reported in [13] show that σB pulse amplitude does increase ( ~3-fold ) with an increase in energy stress thus suggesting that increase in phosphatase concentration at high stress is primarily the result of increase in burst size . To further reinforce the role of mean burst-size modulation in controlling the σB pulsatile response we next examined the cumulative histograms of pulse amplitudes at different phosphatase concentrations . These histograms carry different signatures for burst-size or burst-frequency encoding . The distribution of pulse amplitudes is unchanged with increase in burst frequency ( S6A Fig ) because σB pulse amplitude is determined by phosphatase burst size and not burst frequency . In contrast , if phosphatase levels are controlled by changing mean burst size then the distribution of pulse amplitudes changes accordingly . Consequently , the normalized cumulative histograms of pulse amplitudes overlap for burst-frequency encoding ( S6A Fig ) but not burst-size encoding ( S6B Fig ) . Applying this test to the data from [13] , we found that the normalized cumulative pulse amplitudes histograms do not overlap ( S6C Fig ) . These results predict that stress affects the σB network via burst-size modulation of phosphatase production which is then encoded into σB pulse amplitudes . While the molecular mechanism that introduces energy stress to the network is still not fully understood , our prediction places an important constraint on it . Our model can also be used to study the response of σB network to environmental stress . Unlike the energy stress phosphatase , the environmental stress phosphatase RsbU is regulated post-translationally by binding of RsbT [27–29] . RsbT is trapped by its negative regulators under unstressed conditions but is released upon stress . Consequently , the concentration of RsbTU complex is tightly controlled at the post-translational level and is therefore expected to be relatively insensitive to gene expression fluctuations but sensitive to the level of environmental stress . As a result , step-up increases in environmental stress agents like ethanol produce rapid increases in RsbTU and result in only a single pulse of σB activity [14] . However it has been shown that for gradual increases in stress , σB pulse amplitude depends on the rate of stress increase [14] . To explain this response , we modeled gradual stress with ramped increase in RsbTU complex concentration ( Fig 4A ) . Our simulations showed that the detailed model of σB network is indeed able to capture the effect of rate of stress increase on σB pulse amplitudes . Specifically for a fixed increase in RsbTU complex , the pulse amplitude decreases non-linearly as a function of the duration of phosphatase ramp ( Fig 4B and 4E ) . We hypothesized that this ramp rate encoding is the result of the timescale separation between the fast post-translational and the slow transcriptional responses of the σB network . During the pulsed σB activation , post-translational response is rate-limited by the phosphatase ramp . In contrast , the transcriptional response is slow and its rate is set by the degradation rate of σB operon proteins . Following a step-increase in phosphatase , the fast post-translational response ensures that σB reaches its post-translational steady state before the slow increase in RsbW sequesters σB and turns off the pulse ( Fig 4A and 4B ) . However , for a ramped increase in phosphatase the post-translational increase in σB is limited by the rate of phosphatase ramp . This allows RsbW to catch up and terminate the σB pulse earlier , thereby decreasing the pulse amplitude . To test this , we varied the degradation rate of σB operon proteins and proportionally changed the operon transcription rate to ensure that the total concentrations of σB , RsbW and RsbV are kept fixed . We found that indeed pulse amplitude decreases with increase in degradation/dilution rate ( Fig 4C and 4D ) . Our simulations showed that Kramp , the half-maximal constant for the dependence of pulse amplitude on ramp duration , was indeed sensitive to the degradation rate ( Fig 4E and 4F ) . This suggests that the timescale separation between the post-translational and transcriptional responses is the basis of ramp rate encoding into pulse amplitude . The results thus far indicate that σB network functions in the effectively negative feedback regime where increase in the operon expression decreases σB activity . Negative feedback loops have been shown to increase the robustness of the system to perturbations . We therefore decided to investigate how the σB network design affects its performance when it faces competition for RNA polymerase from other σ-factors , e . g . from the housekeeping σ-factor σA [16 , 30 , 31] . Since σA has a much higher affinity for RNA polymerase [17] , a small increase in σA can dramatically increase the amount of σB necessary to activate the transcription of the σB regulon . Thus , changes in σA can alter the input-output relationship of a stress-response σ-factor like σB ( S7A and S7B Fig ) and thereby adversely affect the survival of cells under stress . To understand how the σB network handles competition for RNA polymerase , we expanded our model to explicitly include σA , RNA polymerase ( RNApol ) and its complexes with both σ-factors . The presence of σA will affect transcriptional activity of σB but not post-translational interactions between σB operon partners ( Fig 5A , left panel ) . Therefore , post-translational response [σB] = FP ( [BT] , [PT] ) is not affected by σA . In contrast , in the transcription response , an increase in σA decreased the ‘effective affinity’ of σB for RNApol and consequently higher levels of free σB are necessary to achieve the same production rate for σB target genes . Using our model , we examined how changes in σA level affect the network response to energy stress signal , i . e . under stochastically fluctuating RsbQP phosphatase levels . Our simulations showed that phosphatase bursts lead to pulses of free σB and pulsatile transcription of σB-controlled promoters ( Fig 5B and 5C ) as the presence of σA does not affect the effective feedback sign . Notably our results also showed that the amplitudes of σB target promoter pulses are hardly affected by a ~30% increase in σA ( Fig 5C , left panel ) . This surprising insensitivity of the phosphatase-σB target dose-response to RNApol competition is the result of the ultrasensitive negative feedback between free σB and total σB . Due to the ultrasensitivity of this feedback , a small decrease in total σB levels resulting from the increase in σA causes a large increase in σB pulse amplitude ( Fig 5B left panel , Fig 5D green line ) . This increased amplitude compensates for the increased competition for RNApol and insulates the network from perturbations ( Fig 5D and 5E , green curves ) . To further illustrate the importance of the negative feedback in insulating the network , we compared the response of the wildtype network to an “in silico” mutant network wherein the σB operon is constitutive rather than σB dependent ( Fig 5A ) . Consequently this network lacks any feedback between free σB and total σB . Our simulations ( Fig 5B , right panel ) show that the free σB concentration of the no-feedback-network does not show adaptive pulsing and therefore σB concentration fluctuates along with the phosphatase levels . Increase in σA did not affect this response . This is expected since in the absence of feedback σA only affects the expression of σB targets in this network ( Fig 5A , right panel ) . Without an increase in free σB ( Fig 5D ) , the increased competition for RNApol at higher σA reduced the σB target promoter activity ( Fig 5C and 5E ) . Similarly a positive feedback network design is also incapable of increasing free σB in response to an increase in σA ( S5C , S5D and S5E Fig ) . Thus fluctuations in σA can interfere with the σB stress-response of these alternative network designs . In contrast , the wildtype σB network with its ultrasensitive negative feedback design can compensate for competition effects ( Fig 5D and 5E ) . The emergent negative feedback design of the network discussed here is not unique to σB . Transcription of many alternative σ-factors in B . subtilis as well in other bacteria is often positively auto-regulated but sigma-factor operons often include post-translational negative regulators [3 , 12 , 32–35] . For example σW , a σ-factor in B . subtilis that controls the response to alkaline shock [36] is co-transcribed with its anti-σ-factor RsiW . In the absence of stress , RsiW sequesters σW in an inactive complex . σW is activated by stress signals which trigger the cleavage and degradation of RsiW thereby releasing and activating σW target expression [37] . Although it is unknown whether the σW network functions in a negative feedback regime similar to σB or if it pulses , it is possible for this network to exhibit these design properties . If RsiW is expressed in stoichiometric excess of its binding partner σW from the σW-regulated operon which they share [38] , then similar to the σB network , σW would operate in a negative feedback regime . To determine if negative feedback control offers any advantages when multiple stress σ-factors are active , we built a new model that includes three σ-factors: σB , σW and σA ( Fig 6A ) . In this model we use σB and σW to denote two generic stress σ–factors and accordingly anti-σ-factors RsbW ( RsiW ) and other details of post-translational regulation were excluded for simplicity . Thus our results apply to any combination of alternative σ-factors with ultrasensitive negative feedback control . In this general model , regulation of free σB and σW was modeled with simplified identical versions of the negative feedback design of the σB network ( S7A Fig ) . Under this simplification , free σB and free σW are non-monotonic functions of their respective total concentrations BT and WT . These non-monotonic functions are qualitatively similar to the post-translational response function shown in Fig 2B and depend on a signaling proteins PB ( for σB ) and PW ( for σW ) . Following the previous section , this model explicitly includes σA , RNApol and its complexes with σ-factors . As a result , transcriptional activity of both σB and σW depend on σA and RNApol concentrations ( see S1 Text ) . Concentrations of RNApol and σA were chosen to ensure that amount of RNApol is insufficient to bind to all σ-factors at the same time . All other parameters of the simplified model were chosen to approximately match the full σB network model and ensure that both σB and σW operate in the negative feedback regime . Consequently for the chosen parameters this simplified model acts like our detailed model and responds to step increases in the stress signaling protein PB ( or PW ) by producing a pulse of σB ( or σW ) activity ( S7C and S7D Fig ) . To enable a comparison of the competition between σ-factors for different types of feedback we hereafter focus on only steady state response , however our conclusions are also valid for the averaged pulsatile dynamical responses that could be characteristic of the negative feedback σ-factor networks . We used this simple model to study the response when cells are simultaneously exposed to multiple stresses creating competition for RNApol . For these simulations we fixed σA levels and studied how activation signals for one alternative σ-factor affects the activity of another . As before ( S7A and S7B Fig ) , increased availability of one stress σ-factor leads to a competition for RNA polymerase and as a result reduces the activity of another stress σ-factor ( S8E and S8F Fig ) . However , when negative feedback loops are present , surprisingly , increasing the stress signal for one σ-factor did not lead to any significant change in the activity of another σ-factor . For example , increasing stress signaling protein PB while keeping PW fixed leads to an increase in free σB but also results in a small increase in free σW ( Fig 6C ) . This response can be explained by the ultrasensitive negative feedback loops controlling the two stress σ-factors . An increase in free σB by stress signaling protein PB leads to increased competition for RNApol resulting in a decrease in the production of RsbW . But since σW is regulated by a negative feedback , a decrease in total RsbW concentration actually frees up more σW thereby insulating σW target activity from the effects of RNApol ( Fig 6E ) . Similarly the dynamic response of the stress σ-factors is also insulated from competition and an increase in fixed PW levels increases the pulse amplitude of σB in response to step changes in stress signaling protein PB ( S8A–S8D Fig ) . This compensation of changes in RNA polymerase availability comes about because both σB and σW are regulated by ultrasensitive negative feedbacks in our model . As a result of this negative feedback , both σ-factor networks function as homeostatic modules . Homeostatic resistance to changes in signals is an intrinsic property of ultrasensitive negative feedback motifs . Thus the two stress σ-factors are able to function simultaneously despite the scarcity of RNApol . The mechanism minimizing competition between stress σ-factors becomes clearer when we track the changes in σ–RNApol complexes as a function of the stress signaling protein PB . As PB increases , more free-σB becomes available and binds to RNApol ( Fig 6G ) . However this RNApol must be accounted for by the RNApol lost by the other operating σW and σA factors . Comparing the contributions of each σ-factor shows that despite the fact that σA has a much higher affinity for RNApol , most of the RNApol in the σB-RNApol complex is drawn from the σA-RNApol pool rather than σW-RNApol pool ( Fig 6G ) . Thus the negative feedback design allows stress σ-factors to minimize their competition with each other at the expense of the housekeeping factor σA . The role of the negative feedback in producing this response becomes clear when we compare the response of an “in silico” mutant network with positive feedback loops between σB and BT and σW and WT ( Fig 6B ) . These positive feedback loops are expected to display no homeostatic properties and as a result , in this network activation of σB should significantly decrease σW activity . Indeed , our simulation for the positive feedback network ( Fig 6D ) demonstrates that with increase in stress signaling protein PB and the resulting increase in free σB , the free σW concentration decreases . As a result of the increased competition for RNApol and the decreased free σW , σW target promoter activity in this network decreases as a function of PB ( Fig 6F ) . Moreover comparing changes in σ–RNApol complexes as a function of stress signaling protein PB we find that most of the RNApol in the σB-RNApol complex is drawn from the σW-RNApol pool rather than σA-RNApol pool ( Fig 6H ) . Thus the negative feedback designs are essential for stress σ-factors not only to tolerate competition from σA , but also to avoid competing with each other when the cell is simultaneously exposed to multiple types of stresses . Taken together , our results show how the design of the σB network includes an implicit ultrasensitive negative feedback that plays multiple functional roles . This design enables pulsatile activation of σB in response to energy stress and rate-sensitivity to increases in environmental stress . Moreover , our model predicts that the same design feature allows the network to effectively compete with house-keeping and other alternative σ-factors for RNA polymerase core . Prompted by recent observations of the highly dynamic pulsatile response of the σB network [13 , 14] , we have developed a mathematical model that reproduces all reported features of the response including pulsatile activation in response to stress . Our model avoids making ad hoc simplifications and instead captures all the known molecular details of the network . By decoupling the post-translational and transcriptional responses in our model we were able to derive a simplified view of the network that illustrates how the pulsatile response is mechanistically based on the ultrasensitive negative feedback in the network . Using this method we identified the ratios of σB , RsbW and RsbV synthesis rates as the most critical design property , which by controlling the post-translational response determines the sign of the feedback in the network as well as all qualitative features of the network response . This highlights how ignoring non-transcriptional interactions and focusing on transcriptional regulatory interactions alone can be misleading when trying to identify or characterize network motifs . Notably , recent analyses of networks like bacterial two-component systems [39] and the sporulation phosphorelay [40] have similarly shown how the effective sign of feedback in these networks depends critically on their post-translational interactions . Interestingly previously reported measurements of the relative amounts of σB , RsbW and RsbV using a quantitative western blot assay showed that the ratios between these components may change depending upon stress [18] . Our model shows that for both pre- and post-stress values of these synthesis ratios , the σB network lies within the negative feedback regime and responds to step-increases in phosphatase levels with a pulsatile response . Thus these stress induced expression changes do not affect our conclusions about the function of the σB network . Nevertheless their mechanistic basis , whether transcriptional or post-translational , remains unclear and may add an additional layer of complexity to regulation . Together with our modeling results these observations highlight the need for more quantitative experimental methods to determine the relative synthesis rates of σB operon components . The decoupling of the post-translational and transcriptional response greatly facilitated the identification of critical design features despite the complexity of the network . This separation greatly reduces the dimensionality of the dynamical system by enabling an independent input–output analysis for the two modules . Similar methods have also been applied to deduce core functional properties in other bacterial networks comprising two-component systems and alternative σ-factors [41–43] . Interestingly our analysis revealed that the post-translational and transcriptional module structures of the σB network and the phosphorelay controlling B . subtilis sporulation are remarkably similar [40] . Despite the differences in molecular details , in both networks increase in total transcription factor levels produces a non-monotonic response in the active transcription factor . Combining this response with the transcriptional feedback produces an ultrasensitive negative feedback in both networks . The relevance of these similarities is evidenced by the fact that both networks produce dynamically similar pulsatile responses even though they are activated by entirely different stimuli . We further used our model to establish that energy stress controls σB pulses frequency by modulating the size of stochastic bursts of energy stress phosphatase rather than burst frequency . We reached this conclusion by showing that only burst size modulation can explain the experimentally observed changes in mean pulse amplitude and pulse amplitude distribution with increasing energy stress level . It should be noted that although our results provide clear evidence in favor of phosphatase burst-size modulation , direct confirmation of this mechanism necessitates the use of single-molecule techniques such as RNA-FISH [44 , 45] that can be used to estimate stochastic properties of gene expression . This result raises the question whether pulsatile σB response can achieve proportional expression of downstream genes , as was previously suggested [13 , 46] . This proportional control requires the distribution of pulse amplitudes to remain fixed even as stress levels increase . However under the burst-size encoding strategy , pulse amplitude distributions change as stress levels increase thereby negating the efficacy of a pulsed response in producing proportional expression of downstream genes . The functional significance of pulsatile response may instead lie in its ability to encode the rate of environmental stress increase . Our model showed that this rate encoding follows from the timescale separation between the fast post-translational and the slow transcriptional responses in the network . As a result cells are able to encode the rate of stress increase into σB pulses . This rate responsiveness is only possible with adaptive pulsatile responses and thus may explain the need for σB pulsing to control the general stress response . We also used our model to understand the response when placed in the larger context of other σ-factor networks and competition for RNA polymerase . Our results show how the network design is uniquely suited to insulating its response from RNA polymerase competition from the housekeeping σ-factor . Finally we demonstrated how ultrasensitive negative feedback , a ubiquitous feature of stress σ-factor regulation enables different stress σ-factors to operate simultaneously without inhibiting each other . These results are relevant not only for understanding the stress response of bacteria but also increasingly for the design of synthetic circuits . The movement towards the construction of larger genetic circuits has produced numerous recent designs that include multiple independent modules that rely on shared resources or actuators to function [47–49] . Our results highlight how competition between modules for shared resources can significantly affect the performance of these synthetic circuits . Further , inspired by the design of naturally occurring stress σ-factor network we provide new design rules that can improve the performance and robustness of the synthetic networks .
Our mathematical model of σB network is based on a previous model proposed in [15] . This ODE-based model explicitly includes all known molecular species , post-translational reactions and the transcriptional regulation of the σB operon by σB . Below we formulate the set of reactions and associated differential equations . The events shown in Fig 1A can be described by the following set of biochemical reactions: vB= v0 ( 1 + f[σB] ( K +[σB] ) ) Here ν0 is the basal synthesis rate , f is the fold change in protein synthesis due to positive autoregulation and K is the equilibrium dissociation constant for the binding of σB to the promoter DNA . The stress signals were assumed to control the concentrations of stress phosphatases RsbTU and RsbQP . For RsbQP , energy stress was assumed to regulate the transcription rate of the phosphatase and the phosphatase concentration was assumed to be subject to stochastic fluctuations resulting from gene expression noise . In contrast , RsbTU concentration is regulated by environmental stress post-translationally [27–29] , consequently RsbTU concentration was assumed to be stress-dependent but not subject to stochastic fluctuations . We assume mass-action kinetics for all the above reactions ( Eqs 1–10 ) to obtain the following set of equations that describe network dynamics: d[BT]dt=vB−kdeg[BT]d[RsbWT]dt=λWvB−kdeg[RsbWT]d[RsbVT]dt=λVvB−kdeg[RsbVT] d[σB]dt=vB+kd3[W2σB]+kb4[W2σB][V]-kb3[W2][σB]-kd4[W2V][σB]-kdeg[σB]d[W2]dt=kd[W]2+ ( kd1+kk1 ) [W2V]+kd3[W2σB]- ( kb1[V]+kb3[σB]+kdeg ) [W2]d[VP]dt=kk1[W2V]+kk2[W2V2]+kd5[VPP]− ( kb5[P]+kdeg ) [VP] d[W2V]dt=kb1[W2][V] + ( kd2+kk2 ) [W2V2]+kb4[W2σB][V] − ( kd1+kk1+kb2[V]+kd4[σB]+kdeg ) [W2V] d[W2V2]dt=kb2[W2V2][V]− ( kd2+kk2+kdeg ) [W2V2]d[VPP]dt=kb5[VP][P]− ( kd5+kp+kdeg ) [VPP] Here [σB] is the concentration of free σB; [W2] is the concentrations of dimeric RsbW; [V] and [VP] are the concentrations of unphosphorylated and phosphorylated RsbV; [W2σB] , [W2V] , [W2V2] and [VPP] are the concentrations of the corresponding protein complexes . [BT] , [RsbWT] , [RsbVT] and [PT] are the concentrations of total σB , RsbW , RsbV and phosphatase: [W2σB]+[σB]=[BT][W]+2[W2]+2[W2σB]+2[W2V]+2[W2V2]=[RsbWT][V]+[W2V]+2[W2V2]+[VP]+[VPP]=[RsbVT][P]+[VPP]=[PT] ( 11 ) All model parameters are summarized in Table 1 . To model the competition for RNA polymerase between σB and the housekeeping σ-factor σA ( Figs 5 and S5 ) , we extended the model described above and supplemented reactions ( 1–9 ) with the following reactions for σA , RNA polymerase ( RNApol ) and σ–RNApol binding: To investigate the competition between σB , σW and σA , we used a phenomenological non-monotonic function to model the post-translational regulation of stress σ-factors ( σB and σW; see S1 Text for details ) . The decoupled transcriptional and post-translational responses of the network at steady state were calculated using the bifurcation package MATCONT[50] . The post-translational response [σB] = FP ( [BT] , [PT] ) , was calculated by varying the rate of operon transcription while keeping the component synthesis rates ( λW , λV ) and the total phosphatase concentration ( PT ) fixed . Similarly , the transcriptional response [BT] = FT ( [σB] ) , was calculated by varying the free σB concentration as an independent variable to calculate the total concentrations of σB , RsbW and RsbV . The parameter values for reversible binding and phosphorylation reactions were taken from [15] or were analysis driven to obtain pulsing in σB . All the parameters used in the model are summarized in Table 1 . In the deterministic set-up ( Figs 1 , 2 , 4 and 6 , S1–S5 and S8 ) the system of differential equations was solved using standard ode15s solver in MATLAB ( The Mathworks Inc . , Natick , MA ) . For stochastic simulations in Figs 3 , 5 , S6 and S7 , the time-varying total phosphatase level [PT] ( = [P] + [VPP] ) was pre-computed using a gamma distributed Ornstein-Uhlenbeck process as in [13] . This gamma distributed Ornstein-Uhlenbeck process permits independent modulation of mean burst size ( b ) and frequency ( a ) [51] . For each phosphatase level , 50 simulations were performed each lasting 10 hours . Pulses were detected by examining local maxima and minima of the simulated trajectories , and subsequently this information was used to compute statistics for pulse amplitude and frequency . For the simulations of the effect of competition for RNA polymerase ( Figs 5 and S7 ) , the total housekeeping σ-factor concentration was varied between 5 and 15 μM . In these simulations we used λW = 4 , λV = 4 . 5 and λW = 2 , λV = 2 to simulate the wildtype ( negative feedback ) and positive feedback networks respectively . For the simulations of the no feedback network we used ( λW = 4 , λV = 4 . 5 ) and f = 0 and v0 = 8 . 64 μMhr-1 to model the σB–independent constitutive production of operon components . For the simulations of the competition between σB , σW and σA ( Figs 6 and S8 ) , the total housekeeping σ-factor concentration was kept fixed at 12 μM . The post-translational response of stress σ-factors σB ( and σW ) in this model was described using the following phenomenological function: σBfree = BT / ( 1 + ( [BT] / KB ) nb / [PB]mb ) . Here the constants nb and mb describe the non-linear dependence of free σ on total-σ . We used ( nb = 7 , mb = 5 ) and ( nb = 0 , mb = 3 ) to simulate the wildtype ( negative feedback ) and positive feedback networks respectively . KB and KW were fixed at 5μM for simulations of both networks . | Understanding the regulation of bacterial stress response holds the key to tackling the problems of emerging resistance to anti-bacteria’s and antibiotics . To this end , here we study one of the longest serving model systems of bacterial stress response: the σB pathway of Bacillus subtilis . The sigma factor σB controls the general stress response of Bacillus subtilis to a variety of stress conditions including starvation , antibiotics and harmful environmental perturbations . Recent studies have demonstrated that an increase in stress triggers pulsatile activation of σB . Using mathematical modeling we identify the core structural design feature of the network that are responsible for its pulsatile response . We further demonstrate how the same core design features are common to a variety of stress response pathways . As a result of these features , cells can respond to multiple simultaneous stresses without interference or competition between the different pathways . | [
"Abstract",
"Introduction",
"Results",
"Methods"
] | [
"cellular",
"stress",
"responses",
"engineering",
"and",
"technology",
"enzymes",
"signal",
"processing",
"cell",
"processes",
"enzymology",
"operons",
"dna-binding",
"proteins",
"phosphatases",
"polymerases",
"network",
"analysis",
"dna",
"computer",
"and",
"information"... | 2016 | Role of Autoregulation and Relative Synthesis of Operon Partners in Alternative Sigma Factor Networks |
Identification of drug-target interactions ( DTIs ) plays a key role in drug discovery . The high cost and labor-intensive nature of in vitro and in vivo experiments have highlighted the importance of in silico-based DTI prediction approaches . In several computational models , conventional protein descriptors have been shown to not be sufficiently informative to predict accurate DTIs . Thus , in this study , we propose a deep learning based DTI prediction model capturing local residue patterns of proteins participating in DTIs . When we employ a convolutional neural network ( CNN ) on raw protein sequences , we perform convolution on various lengths of amino acids subsequences to capture local residue patterns of generalized protein classes . We train our model with large-scale DTI information and demonstrate the performance of the proposed model using an independent dataset that is not seen during the training phase . As a result , our model performs better than previous protein descriptor-based models . Also , our model performs better than the recently developed deep learning models for massive prediction of DTIs . By examining pooled convolution results , we confirmed that our model can detect binding sites of proteins for DTIs . In conclusion , our prediction model for detecting local residue patterns of target proteins successfully enriches the protein features of a raw protein sequence , yielding better prediction results than previous approaches . Our code is available at https://github . com/GIST-CSBL/DeepConv-DTI .
The identification of drug-target interactions ( DTIs ) plays a key role in the early stage of drug discovery . Thus , drug developers screen for compounds that interact with specified targets with biological activities of interest . However , the identification of DTIs in large-scale chemical or biological experiments usually takes 2~3 years of experiments , with high associated costs [1] . Therefore , with the accumulation of drugs , targets , and interaction data , various computational methods have been developed for the prediction of possible DTIs to aid in drug discovery . Among computational approaches , docking methods , which simulate the binding of a small molecule and a protein using 3D structure , were initially studied . Docking methods recruit various scoring functions and mode definitions to minimize free energy for binding . Docking methods have advanced by themselves , and recently , the Docking Approach using Ray-Casting ( DARC ) model identified 21 compounds by using an elaborate binding pocket topography mapping methodology , and the results were reproduced in a biochemical assay [2] . In addition , studies have examined several similarity-based methods in which it was assumed that drugs bind to proteins similar to known targets and vice versa . One of the early methods is that of Yamanashi et al . , which utilized a kernel regression method to use the information on known drug interactions as the input to identify new DTIs , combining a chemical space and genomic spaces into a pharmacological space [3] . To overcome the requirement of the bipartite model for massive computational power , Beakley et al . developed the bipartite local model , which trains the interaction model locally but not globally . In addition to substantially reducing the computational complexity , this model exhibited higher performance than the previous model [4] . As another approach to DTI prediction models , matrix factorization methods have been recruited to predict DTIs , which approximate multiplying two latent matrices representing the compound and target protein to an interaction matrix and similarity score matrix [5 , 6] . In this work , regularized matrix factorization methods successfully learn the manifold lying under DTIs , giving the highest performance among previous DTI prediction methods . However , similarity-based methods are not commonly used at present to predict DTIs , as researchers have found that similarity-based methods work well for DTIs within specific protein classes but not for other classes [7] . In addition , some proteins do not show strong sequence similarity with proteins sharing an identical interacting compound [8] . Thus , feature-based models that predict DTI features of drugs and targets have been studied [9–11] . For feature-based DTI prediction models , a fingerprint is the most commonly used descriptor of the substructure of a drug [12] . With a drug fingerprint , a drug is transformed into a binary vector whose index value represents the existence of the substructure of the drug . For proteins , composition , transition , and distribution ( CTD ) descriptors are conventionally used as computational representations [13] . Unfortunately , feature-based models that use protein descriptors and drug fingerprints showed worse performance than previous conventional quantitative structure-activity relationship ( QSAR ) models [9] . To improve the performance of feature-based models , many approaches have been developed , such as the use of interactome networks [14 , 15] and minwise hashing [16] . Although various protein and chemical descriptors have been introduced , feature-based models do not show sufficiently good predictive performance [17] . For conventional machine learning models , features must be built to be readable by modeling from original raw forms , such as simplified molecular-input line entry system ( SMILES ) and amino acid sequences . During transformation , rich information , such as local residue patterns or relationships , is lost . In addition , it is hard to recover lost information using traditional machine learning models . In recent years , many deep learning approaches have recently been developed and recruited for omics data processing [18] as well as drug discovery [19] , and these approaches seem to be able to overcome limitations . For example , DeepDTI built by Wen et al . used the deep belief network ( DBN ) [20] , with features such as the composition of amino acids , dipeptides , and tripeptides for proteins and extended-connectivity fingerprint ( ECFP ) [21] for drugs [7] . The authors also discussed how deep-learning-based latent representations , which are nonlinear combinations of original features , can overcome the limitations of traditional descriptors by showing the performance in each layer . In another study by Peng et al . [22] , MFDR employed sparse Auto-Encoder ( SAE ) to abstract original features into a latent representation with a small dimension . With latent representation , they trained a support vector machine ( SVM ) , which performed better than previous methods , including feature- and similarity-based methods . In another study called DL-CPI by Tian et al . [23] , domain binary vectors were employed to represent the existence of domains used to describe proteins . One way to reduce the loss of feature information is to process raw sequences and SMILES as their forms . In a paper by Öztürk et al . , DeepDTA was used to represent raw sequences and SMILES as one-hot vectors or labels [24] . With a convolutional neural network ( CNN ) , the authors extracted local residue patterns to predict the binding affinity between drugs and targets . As a result , their model exhibited better performance on a kinase family bioassay dataset [25 , 26] than the previous model , kronRLS [27] and SimBoost [28] . Because their model is optimized by densely constructed kinase affinities , DeepDTA is appropriate to predict kinase affinities not to predict new DTIs with various protein classes . Furthermore , they evaluated their performances on the identical dataset , rather than on independent dataset from new sources or databases . To overcome the aforementioned problems , here , we introduce a deep learning model that predicts massive-scale DTIs using raw protein sequences not only for various target protein classes but also for diverse protein lengths . The overall pipeline of our model is depicted in Fig 1 . First , for the training model , we collected large-scale DTIs integrated from various DTI databases , such as DrugBank [29] , International Union of Basic and Clinical Pharmacology ( IUPHAR ) [30] , and Kyoto Encyclopedia of Genes and Genomes ( KEGG ) [31] . Second , in model construction , we adopted convolution filters on the entire sequence of a protein to capture local residue patterns , which are the main protein residues participating in DTIs . By pooling the maximum CNN results of sequences , we can determine how given protein sequences match local residue patterns participating in DTIs . Using these data as input variables for higher layers , our model constructs , abstracts and organizes protein features . After new protein features are generated , our model concatenates protein features with drug features , which come from fingerprints in the fully connected layer and predict the probability of DTIs via higher fully connected layers . Third , we optimized the model with DTIs from MATADOR [32] and negative interactions predicted from Liu et al . [33] . Finally , with the optimized model , we predicted DTIs from bioassays such as PubChem BioAssays [34] and KinaseSARfari [35] to estimate the performance of our model . As a result , our model exhibits better performance than previous models .
As a normal step of hyperparameter setting , we first tuned the learning rate of the weight update to 0 . 0001 . After the learning rate was fixed , we benchmarked the sizes and number of windows , hidden layers of the drug features , and the concatenating layers with the area under precision-recall ( AUPR ) on the external unseen validation dataset , which was built with MATADOR and a highly credible negative dataset . Finally , we selected the hyperparameters of the model , as shown in Table A in S1 Text , with the external unseen validation dataset , yielding an AUPR of 0 . 832 and area under the curve ( AUC ) of 0 . 852 , as shown in Fig 2 . The AUPR value of our model was less than the AUPR of the similarity descriptor; however , that does not mean that our method has lower prediction performance than the similarity method because the size of the validation is too small to evaluate the general performance . In addition , we further examined the effect of fixed maximum protein length on the prediction performance . As shown in Fig A in S1 Text , we confirmed that the prediction performance of our model is not biased to the fixed maximum protein length . Finally , the fully optimized model is visualized as a graph , shown in S1 Fig , respective to our model , the CTD descriptor , and similarity descriptors . In the same manner , we built and optimized models that use other protein descriptors with the same activation function , learning rate , and decay rate . After the hyperparameters were tuned , we compared the performance based on the independent test datasets with the different protein descriptors , the CTD descriptor ( which is usually used in the conventional chemo-genomic model ) [13] , the normalized Smith-Waterman ( SW ) score [36] , and our convolution method . The results showed that our model exhibited better performance than the other protein descriptors for all datasets , as shown in Fig 3 and Fig B in S1 Text . With the threshold selected by the equal error rate ( EER ) [37] , our model performed equally well with both the PubChem and KinaseSARfari datasets , indicating that our model has general application power . Our convolution method gave the highest accuracy score and F1 score for the PubChem dataset ( Fig 3A ) [34] and its subsets ( Fig 3B–3D ) and a slightly lower F1 score for the KinaseSARfari dataset ( Fig B in S1 Text ) [35] . The CTD descriptor gave the lowest score for any dataset and any metric , which implies that CTD is less informative and less enriched than the other descriptors . Here , we also observed that the model performance using a similarity descriptor for the KinaseSARfari dataset was similar to that of the proposed model . We can interpret this result as the similarity descriptor acts as an informative feature as a local residue pattern at the domain level , not the whole protein complex . In addition to the comparison between convolution in our model and other protein descriptors , in this section , we compared the performance of our model against recently developed deep-learning-based models . We selected three deep learning models for comparisons , SAE ( MFDR , Peng et al , 2016 ) [22] , DBN ( DeepDTI , Wen et al , 2017 ) [7] and CNN ( DeepDTA , Ozturk et al , 2018 ) . First , MFDR trains SAE in an unsupervised manner , while proteins are represented by multi-scale local descriptor feature [38] and compounds are represented by PubChem fingerprints as input and output for SAE . With trained deep representations of sparse Auto-Encoder , they performed 5-fold cross-validation by using SVM . As a result , their model gives better performances than previous bipartite local models . Because the authors do not provide the model , we implemented the MFDR model with optimized parameters the author provided in their original paper . We tested the validity of implemented MFDR and confirmed that the implemented model produces reasonably same performance compared to the results from its original work ( see Fig C S1 Text ) . Second , DeepDTI built by Wen et al . is based on DBN [20] , which is a stack of restricted Boltzmann machine ( RBM ) . DeepDTI takes amino acid , dipeptide and tripeptide compositions ( protein sequence composition descriptors , PSC ) as the protein input and ECFP with radius 1 , 2 and 3 as the compound input . We used DeepDTI with the code that the authors provided ( https://github . com/Bjoux2/DeepDTIs_DBN ) and optimized hyperparameters as the authors mentioned . Third , DeepDTA built by Ozturk et al . used stacked CNN on protein sequences and SMILES to predict affinity between target protein and compound . DeepDTA is optimized for Davis [25] and KIBA [26] dataset which contains kinases protein , their inhibitors , and dense affinity values , showing better prediction performances than previous affinity prediction models . We also used DeepDTA with the code from the original work ( https://github . com/hkmztrk/DeepDTA ) and optimized hyperparameters they provided . For the DTI prediction performance comparison , we activate the last layer with sigmoid function to predict interaction , not affinity , also we changed loss function as binary cross-entropy from mean squared error . It should be noticed that we compared the performance of all three models by training and testing with the same data set we used for a fair comparison . Results of performance comparison between our proposed model and the three related models are shown in Fig 4 , showing that performances ( accuracy , F1 ) of our model ( DeepConv-DTI ) are better than other models . MFDR which gave high AUC in 5-fold cross-validation shows decreased performances in the independent test dataset . We can speculate that SAE which learns deep representation of DTI in an unsupervised way is not appropriate for a case that datasets are composed of various protein classes . In the case of DeepDTI , DeepDTI takes physicochemical properties ( PSC ) of whole protein sequence including subsequences or domains which do not participate in the interaction with compounds , resulting in worse performance than our model which extracts local residue patterns . For DeepDTA , DeepDTA also shows worse performances than our model with having a relatively large variance . We interpret the worse performance of DeepDTA as follows . DeepDTA is optimized for a densely constructed dataset with specific protein class , while the training dataset in this comparison covers various protein classes ( kinase , protease , ion channel , nuclear receptor , GPCR , etc ) , not only kinase class . Thus , DeepDTA which is specialized for a specific protein class could not achieve better prediction performance in the generalized protein classes . In addition to the three models we compared , we also compared our model with DL-CPI [23] built by Tian et al . which used protein domain information . For proteins whose domain information is not in Pfam [39] , datasets for training , validation and test are not fully available . Therefore , we independently compared performances between DL-CPI and our model by additionally built the training , validation , and test datasets . Performance comparison results are described in Fig E in S1 Text . We confirmed that the proposed model shows better performance than DL-CPI . Because protein descriptor of DL-CPI is sparse , containing few values in large dimension , which may decrease performances . In overall , our model shows better performance than previous deep learning models in an independent test dataset from a different database , which contains distinct DTIs , dealing with DTIs with various protein classes and their interacting compounds . Because we pooled the maximum convolution results by each filter for each window , the pooled results could highlight regions of matches with local residue patterns . Although we cannot measure exactly how those values affect the DTI prediction results , the pooled maximum convolution result will affect the prediction performance by going through higher fully connected layers . Therefore , if our model is capable of capturing local residue patterns , it would give high values to important protein regions , such as actual binding sites . Examining and validating the convolution results from the intermediate layer showed that our model could capture local residue patterns that participate in DTIs . The sc-PDB database provides atom-level descriptions of proteins , ligands , and binding sites from complex structures [40] . By parsing binding site annotations , we can query binding sites between protein domains and pharmacological ligands for 7 , 179 entries of Vertebrata . From the queried binding sites and pooled maximum convolution results , we statistically test our assumption that the pooled maximum convolution results cover the important regions , including binding sites . Each window has 128 pooled convolution results , which shows bias in covering some regions . Thus , we randomly generated 128 convolution results 10 , 000 times for each sc-PDB entry and counted how many of those random results covered each amino acid in the binding sites , which resulted in the construction of normal distributions . For each normal distribution constructed by the randomly generated convolution results , considered a null hypothesis , we executed a right-tailed t-test with the number from the convolution results of our model for each window . Because we did not know which window detects the binding site , we took the most significant p-value ( minimum p-value adjusted by the Benjamini-Hochberg procedure [41] ) . The sc-PDB entry information and p-values of a window for each sc-PDB entry are summarized in the S1 File . We summarize the results of binding site detection from the most significant p-value among windows by significance level cutoff in Fig 5 . In addition , we examined sc-PDB entries with the most significant p-values for diverse window sizes . We visualized two high-score sc-PDB entries from two perspectives—the whole receptor-ligand complex and binding site-ligand perspectives—by using UCSF Chimera [42] as shown in Fig 6 . To visualize convolution results with a simplified view , first , we selected the top 5 ranked globally max-pooled results among all filters for each window because whole protein sequences are usually covered by convolution results if we select all results . Second , we rendered residues covered by convolution results by the number of covering convolution results . We visualized two sc-PDB entries , 1a7x_1 and 1ny3_1 . 1a7x_1 , representing the complex of the ion channel , protein Peptidyl-prolyl cis-trans isomerase FKBP1A ( FKB1A_HUMAN in UniProt ) , which has a short sequence length ( 108 ) , and BENZYL-CARBAMIC ACID [8-DEETHYL-ASCOMYCIN-8-YL]ETHYL ESTER ( FKA in PDB ligand ) [43] . 1ny3_1 is the complex of the kinase protein , MAP kinase-activated protein kinase 2 ( MAPK2_HUMAN in UniProt ) with sequence length 400 , and ADENOSINE-5’-DIPHOSPHATE ( ADP in PDB ligand ) [44] . Through the above evaluation , we can confirm that our proposed model is capable of capturing local residue patterns of proteins that are considered important features for DTI prediction , such as actual binding sites . From the results shown in Fig 6 , we can confirm that our model can capture the local residue patterns of proteins that participate in DTIs . Thus , to examine further characteristics of the captured protein local residue patterns , we visualized the protein features from the fully connected layer after the global max-pooling of convolution results . We visualized 1 , 527 proteins used in the training dataset categorized in various protein classes . Specifically , we visualized 257 GPCRs , 44 nuclear receptors , 304 ion channel receptors , 604 kinases , and 318 proteases . For visualization , we conducted t-distributed stochastic neighbor embedding ( t-SNE ) for dimension reduction and visualization [45] . t-SNE can map high-dimensional features to low-dimensional ones , such as 2-dimensional features , minimizing information loss during dimension reduction . Surprisingly , although our model is not intended to identify protein classes , it can roughly discriminate protein classes from the intermediate protein layer , as shown in Fig G in S1 Text .
In this work , we built a novel DTI prediction model to extract local residue patterns of whole target protein sequences with CNN . We trained the model with DTIs from various drug databases and optimized the model with an external validation dataset . As a result , the detected local features of protein sequences perform better than other protein descriptors , such as CTD and SW scores . Our model also performs better than a previous model built on DBN . In addition , by analyzing pooled convolution results and statistically and manually comparing them with annotations from sc-PDB entries , we showed that , for some proteins , our model is capable of detecting important regions , including binding sites . Therefore , our approach of capturing local residue patterns with CNN successfully enriches protein features for DTI prediction . The number of 3D structures in Protein Data Bank [46] is relatively smaller than the number of sequences , limiting 3D structure-based DTI prediction methods . For example , the number of PDB entries for Homo sapiens is 42 , 745 , while the number of protein sequences for Homo sapiens is 177 , 661 in UniProtKB . However , our method does not depend on the 3D structure of proteins because it considers only protein sequence , rather than classical protein feature descriptors such as the CTD descriptor and normalized SW score . As a result , our method can be more generally applied to predict DTIs than methods needing 3D structures . Although our model shows improved prediction performance , there is still room for improvement . First , we simply used Morgan/Circular fingerprints , which are binary and have large dimensions . Therefore , we will use more informative chemical descriptors , based on neural networks for DTI prediction , to achieve advanced performance . Second , as shown in a previous study [47] , considering 3D structure information is an effective substitution for chemical elaboration . Therefore , in the future , we will elaborate upon our model by considering 3D structure features .
To build the training dataset , we obtained known DTIs from three databases: DrugBank , KEGG , and IUPHAR . To remove duplicate DTIs among the three databases , we unified the identifiers of the compounds and the proteins . For the drugs , we standardized the identifiers of the compounds in the DrugBank and KEGG databases with the InChI descriptor . For the proteins , we unified the identifiers of the proteins as UniProtKB/Swiss-Prot accessions [48] . Among the collected DTIs , we selectively removed proteins of Prokaryota and single-cell Eukaryota , retaining only proteins of Vertebrata . Finally , 11 , 950 compounds , 3 , 675 proteins , and 32 , 568 DTIs were obtained in total . Because all collected DTIs are regarded as positive samples for training and negative DTIs are not defined in the databases above , a random negative DTI dataset is inevitably generated . To reduce bias from the random generation of negative DTIs , we built ten sets of negative DTIs exclusively from the positive dataset . The detailed statistics of the collected training dataset are shown in Table D in S1 Text . To optimize our model with the most adequate hyperparameters , we constructed an external validation dataset that had not seen DTIs in the training phase . We collected positive DTIs from the MATADOR database [32] , including ‘DIRECT’ protein annotations , and all DTIs observed in the training dataset were excluded . To build a credible negative dataset , we obtained negative DTIs via the method of Liu et al . [33] . This method selects candidate negative DTIs with low similarity to known positive DTIs . From the obtained negative dataset , we balanced the negative dataset with the positive dataset , using a negative score ( >0 . 95 ) . As a result , 370 positive DTIs and 507 negative DTIs were queried for the external validation set . The statistics of the external validation dataset are summarized in Table E in S1 Text . To evaluate our model , we built two independent test datasets from the PubChem BioAssay database [34] and ChEMBL KinaseSARfari [35]; these datasets consisted of results from experimental assays . To obtain positive DTIs from PubChem , we collected ‘Active’ DTIs from the assays with the dissociation constant ( Kd < 10μm ) [49] . Because we sought to predict whether a drug binds to a protein , among the many types of assays ( Potency , IC50 , AC50 , EC50 , Kd , Ki ) , evaluation of the dissociation constant ( Kd ) was the most appropriate assay for obtaining positive samples . For the negative samples , we took the samples annotated as ‘Inactive’ from the other assay types . Because there were too many negative samples in the PubChem BioAssay database , we first collected only negative samples whose drug or target was included in the positive samples from the PubChem BioAssay database . Second , we selected as many random negative samples as positive DTIs from PubChem BioAssay . As a result , total 36 , 456 positive and negative samples were built with 21 , 907 drugs and 698 proteins . For the performance evaluation , we created three subsets of the PubChem bioassay independent dataset for humans , which consisted of only new compounds , new proteins , and new DTIs . Detailed summaries of the PubChem dataset and its subset are shown in Table F in S1 Text . We also collected samples from KinaseSARfari . KinaseSARfari consists of assays involving a compound that binds to a kinase domain . To obtain positive samples from KinaseSARfari , we considered each assay result with a dissociation constant of ( Kd < 10μm ) as positive [49]; this value is sufficiently small to be considered positive . In contrast to the PubChem BioAssay , the number of negative samples was similar to the number of positive samples in KinaseSARfari; therefore , we did not sample the negative samples . We collected 3 , 835 positive samples and 5 , 520 negative samples with 3 , 379 compounds and 389 proteins . Detailed statistics of the KinaseSARfari dataset are shown in Table F in S1 Text . In addition , we summarize the portion of the protein class in each dataset in Fig H in S1 Text . Here , we confirmed that the training and the validation datasets were not biased toward a specific protein class . In our model , we used the raw protein sequence as the input for the protein but did not use the raw SMILES string as the input for the drug . For the drug , we used the Morgan/Circular drug fingerprint , which analyzes molecules as a graph and retrieves substructures of molecular structures from subgraphs of the whole molecular graph [21] . Specifically , we used RDKit [50] to yield a Morgan/Circular fingerprint with a radius of 2 from a raw SMILES string . Finally , each drug can be represented as a binary vector with a length of 2 , 048 , whose indices indicate the existence of specific substructures . SAE is Auto-Encoder whose distribution of latent representations is regularized with sparsity term [58] . In loss calculation , Kullback-Leibler divergence ( KLD ) loss between Bernoulli distributions each dimension in latent representation ρ^ and desired sparsity parameter ρ is added to reconstruction loss of Auto-Encoder and ridge loss for weights . Jsparse ( W , b ) =J ( W , b ) +β∑js2KL ( ρ||ρ^j ) where ρj^=1m∑i=1m[aj ( 2 ) ( x ( i ) ) ] During the training of the neural network , KLD acts as a constraint for latent representation following desired sparsity parameter . As a result , for each dimension of latent representation , only a few samples are activated , giving a more reliable representation of original input . In the previous study , MFDR used SAE to build an informative latent representation of DTI , which are composed of multi-scale local descriptors [38] and PubChem fingerprints . DBN is a generative graphical model proposed by Geoffrey Hinton [20] . DBN is actually a stack of an RBM . RBM consists of visible and hidden units , constructing a bipartite graph . In RBM , probabilistic distribution of visible units is learned in an unsupervised way , with a probabilistic distribution of visible and hidden units P ( v , h|W ) =1ZeaTv+bTh+vTWh and marginal distribution of visible units P ( v|W ) =1Z∑heaTv+bTh+vTWh to maximize the probability of visible units for V in a training set with weight matrix W argmaxW∏v∈VP ( v|W ) In DBN , during stacking of RBMs , hidden units of the previous RBM are fed as visible layers of the next RBM . In addition , RBM adopts contrastive divergence for fast training , which uses gradient descent and Gibbs sampling . In a previous study , DeepDTI , the input concatenation of drug and target protein features , PSC descriptors and ECFP with a radius of 1 , 2 and 3 , was considered a first visible layer . The authors attached logistic regression to the last hidden units to predict DTIs . To measure the prediction performance of our deep neural model based on the independent test dataset after the classification threshold was fixed , we obtained the following performance metrics: sensitivity ( Sen . ) , specificity ( Spe . ) , precision ( Pre . ) , accuracy ( Acc . ) , and the F1 measure ( F1 ) . See the formulas below: Sen . =TP/P Spe . =TN/N Pre . =TP/ ( TP+FP ) Acc . = ( TP+TN ) / ( P+N ) F1= ( Sen*Pre ) / ( Sen+Pre ) where TP is true positive , TN is true negative , FP is false positive , FN is false negative , T is positive , and N is negative . | Drugs work by interacting with target proteins to activate or inhibit a target’s biological process . Therefore , identification of DTIs is a crucial step in drug discovery . However , identifying drug candidates via biological assays is very time and cost consuming , which introduces the need for a computational prediction approach for the identification of DTIs . In this work , we constructed a novel DTI prediction model to extract local residue patterns of target protein sequences using a CNN-based deep learning approach . As a result , the detected local features of protein sequences perform better than other protein descriptors for DTI prediction and previous models for predicting PubChem independent test datasets . That is , our approach of capturing local residue patterns with CNN successfully enriches protein features from a raw sequence . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"statistics",
"enzymes",
"enzymology",
"neuroscience",
"mathematics",
"forecasting",
"artificial",
"intelligence",
"convolution",
"sequence",
"motif",
"analysis",
"pharmacology",
"extraction",
"techniques",
"res... | 2019 | DeepConv-DTI: Prediction of drug-target interactions via deep learning with convolution on protein sequences |
The rate at which HIV-1 infected individuals progress to AIDS is highly variable and impacted by T cell immunity . CD8 T cell inhibitory molecules are up-regulated in HIV-1 infection and associate with immune dysfunction . We evaluated participants ( n = 122 ) recruited to the SPARTAC randomised clinical trial to determine whether CD8 T cell exhaustion markers PD-1 , Lag-3 and Tim-3 were associated with immune activation and disease progression . Expression of PD-1 , Tim-3 , Lag-3 and CD38 on CD8 T cells from the closest pre-therapy time-point to seroconversion was measured by flow cytometry , and correlated with surrogate markers of HIV-1 disease ( HIV-1 plasma viral load ( pVL ) and CD4 T cell count ) and the trial endpoint ( time to CD4 count <350 cells/μl or initiation of antiretroviral therapy ) . To explore the functional significance of these markers , co-expression of Eomes , T-bet and CD39 was assessed . Expression of PD-1 on CD8 and CD38 CD8 T cells correlated with pVL and CD4 count at baseline , and predicted time to the trial endpoint . Lag-3 expression was associated with pVL but not CD4 count . For all exhaustion markers , expression of CD38 on CD8 T cells increased the strength of associations . In Cox models , progression to the trial endpoint was most marked for PD-1/CD38 co-expressing cells , with evidence for a stronger effect within 12 weeks from confirmed diagnosis of PHI . The effect of PD-1 and Lag-3 expression on CD8 T cells retained statistical significance in Cox proportional hazards models including antiretroviral therapy and CD4 count , but not pVL as co-variants . Expression of ‘exhaustion’ or ‘immune checkpoint’ markers in early HIV-1 infection is associated with clinical progression and is impacted by immune activation and the duration of infection . New markers to identify exhausted T cells and novel interventions to reverse exhaustion may inform the development of novel immunotherapeutic approaches .
Following infection with Human Immunodeficiency Virus Type 1 ( HIV-1 ) the rate at which an individual develops AIDS is highly variable ranging from ‘progressors’ who , if untreated , experience rapid CD4 T cell decline in months to years to ‘elite controllers’ , who spontaneously maintain undetectable plasma viraemia , often for decades . The tempo of HIV-1-associated disease progression might rest with particular characteristics of HLA class I molecules and the CD8 T cell immune responses which they dictate [1–5] . When a CD8 T cell encounters its cognate antigen , the up-regulation of T cell inhibitory molecules tightly controls the subsequent T cell activation [6–8] , and inhibits autoimmunity [9–11] . However , the persistence of antigen can overcome homeostatic controls and lead to permanent CD8 T cell dysfunction or exhaustion [12–16] . In HIV-1 infection T cell exhaustion is associated with the up-regulation of surface molecules called immune checkpoint receptors ( ICRs ) such as PD-1 , Tim-3 and Lag-3 [12 , 17–20] , which have also been associated with the size of the HIV reservoir and time to viral rebound after therapy cessation[21 , 22] . We sought to determine whether , in primary HIV-1 infection ( PHI ) , these indicators of CD8 T cell exhaustion would correlate with surrogate markers of disease ( e . g . HIV-1 plasma viral load ( pVL ) , CD4 T cell count ) and actual time to progression within a strictly defined patient population enrolled into a randomized clinical trial of early antiretroviral therapy ( ART ) . In particular , we wanted to study exhaustion in activated CD38 CD8 positive T cell populations , as CD38 expression has also been correlated with disease progression . We found significant associations between ICR expression and both pVL and disease progression , and an enhanced effect when co-expressed on activated T cells .
366 participants were enrolled into the SPARTAC trial[23] . Of 156 participants recruited at UK sites , 122 had adequate numbers of peripheral blood mononuclear cells ( PBMCs ) available for analysis at the pre-therapy ‘baseline’–the closest documented visit to the estimated date of seroconversion . Patient characteristics are detailed in Table 1 . The median ( interquartile range , IQR ) age at enrolment was 34 ( 28 , 41 ) years . Only five participants were female , and the predominant mode for HIV-1 acquisition was sex between men ( MSM ) ( 93% ) . Of the 122 , 41 were randomised to receive no immediate ART ( SOC ) , 44 12 weeks ART ( ART12 ) and 37 48 weeks ART ( ART48 ) . The median ( IQR ) time since the estimated date of seroconversion at randomisation was 77 ( 54 , 98 ) days . The median ( IQR ) times for participants in ‘early’ ( ≤12 weeks after seroconversion ) and ‘late’ ( >12 weeks after seroconversion ) were 60 ( 42 , 72 ) and 104 ( 93 , 122 ) days , respectively . MSM were significantly over-represented in the early PHI group . The median ( IQR ) baseline CD4 T cell count and HIV-1 pVL were 550 ( 435 , 675 ) cells/μl and 4 . 70 ( 4 . 0 , 5 . 3 ) log10 RNA copies/ml , respectively , without significant differences in participants enrolled ≤ or > 12 weeks into PHI . The expression of PD-1 , Tim-3 and Lag-3 on the surface of CD8 T cells and the subset of activated CD8 T cells that co-expressed CD38 was determined by flow cytometry ( Fig A in S1 Text ) and was up-regulated in PHI compared with healthy controls ( P<0 . 01 for all comparisons; Mann-Whitney ) ( Fig B in S1 Text ) . Expression of CD38 on CD8 T cells correlated with PD-1 expression ( Spearman’s rho 0 . 37; p<0 . 001 ) ( Fig 1a ) but not with Tim-3 or Lag-3 ( Fig 1b and 1c ) . Table 2 shows the percentage of CD8 T cells expressing each of the three exhaustion markers at the pre-therapy baseline time-point; 4 . 5% , 14 . 1% and 8 . 0% of all CD8 T cells and 2 . 1% , 6 . 1% and 3 . 9% of CD38 CD8 T cells ( denominator all CD8 T cells ) expressed PD-1 , Lag-3 and Tim-3 , respectively ( Table 2 ) . When stratified by time since seroconversion ( ≤12 weeks or >12 weeks ) , participants sampled earlier had ( with the exception of PD-1 and PD-1/Tim-3 co-expression for which there was a trend ) significantly higher percentages of expression of single and dual-expressed markers on CD8 CD38 T cells . The same was not observed for bulk CD8s ( Table 2 ) . Correlations between the expression of ICRs on CD8 and CD38 CD8 T cells with CD4 T cell count , pVL and time since seroconversion were assessed ( Table 3 and Fig 1d–1i ) . PD-1 and Lag-3 ( Fig 1d and 1f ) , but not Tim-3 ( Fig 1e ) , expression on CD8 T cells was associated with higher pVL ( p<0 . 001 for both ) . Expression of ICRs on CD38 CD8 T cells was significantly and more strongly associated with higher pVL compared with on the bulk CD8 positive population ( p<0 . 01 for all markers; Table 3 , Fig 1g–1i ) . Co-expression on CD8 and CD38 CD8 T cells of either PD-1/Tim-3 , PD-1/Lag-3 and Tim-3/Lag-3 was also positively associated with pVL ( Table 3 ) . PD-1 and PD-1/Lag-3 co-expression on CD8 T cells , and PD-1 , PD-1/Tim-3 and PD-1/Lag-3 expression on CD38 CD8 T cells were significantly associated with lower baseline CD4 T cell counts ( Table 3 ) . These associations persisted after adjusting for multiple comparisons apart from that between CD4 T cell count and co-expression of PD-1/Lag-3 on CD38 CD4 T cells . Next we evaluated PD-1 , Lag-3 and Tim-3 expression and progression to the SPARTAC trial primary endpoint ( either CD4 T cell count <350 cells/μl or ( re ) initiation of ART ) . PD-1 expression at baseline on CD8 and CD38 CD8 T cells predicted clinical progression ( log rank test , p = 0 . 049 and p = 0 . 014 , respectively ) ( Fig 2a and 2d ) . This effect was most evident in patients recruited within 12 weeks of infection , when compared with those recruited after 12 weeks ( Fig 2b and 2e vs 2c and 2f ) . Tim-3 and Lag-3 expression on CD8 ( Fig 3 ) or CD38 CD8 ( Fig C in S1 Text ) T cells was not associated with clinical progression when the whole cohort was analysed although , surprisingly , there was evidence for slower disease progression in participants recruited ≤12 weeks from seroconversion with increased Tim-3 expression on CD8 T cells ( log rank test , p = 0 . 01; Fig 3b ) . This could not be explained through a correlation with Tim-3 expression and PD-1 ( or Lag-3 ) ( Table C in S1 Text ) and this advantage survived ( P = 0 . 075 ) adjustment for PD-1 , Lag-3 , baseline CD4 count and therapy ( Table D in S1 Text ) . Cox proportional hazards models were used to explore further the association between baseline PD-1 expression on bulk CD8 T cells and clinical progression ( Table 4 ) . The association with PD-1 expression and clinical progression was preserved when ART alone ( HR 1 . 75; p = 0 . 045 ) or ART and baseline CD4 T cell count ( HR 1 . 76; p = 0 . 047 ) were included in the model , but not when accounting for pVL or following correction for multiple comparisons . In a model including time from seroconversion , ART and baseline CD4 T cell count there was a trend for an association between PD-1 expression and clinical progression ( HR 1 . 72 , 95% CI 0 . 98–3 . 03 ) , but this was not significant when baseline pVL was included . When the magnitude of PD-1 expression was stratified into quartiles , there was evidence of a dose effect ( HR: 1 , 1 . 60 , 1 . 71 , 2 . 16 for each quartile , respectively ) when ART and baseline CD4 T count were included in the model , but not when baseline pVL was included . We found no evidence to support a significant association with CD38 alone and time to the trial endpoint . Cox models also demonstrated an effect of bulk CD8 T cell Lag-3 expression on progression when adjusted for ART initiation and CD4 T cell count ( HR 1 . 46; p = 0 . 024 ) , but not when including pVL ( Table 4 ) . As for PD-1 expression , there were significant associations with clinical progression when including time since seroconversion , ART and baseline CD4 T cell count in the model ( HR 1 . 45; p = 0 . 03 ) or when exploring dose effect when stratified into quartiles ( HR 1 , 1 . 43 , 1 . 12 , 2 . 17 ) , but neither survived correction for baseline pVL . Tim-3 expression on bulk CD8 T cells did not predict progression . Interestingly , on restricting the Cox models to CD38 CD8 T cells , all three markers were associated with disease progression in unadjusted models or when correcting for ART and CD4 T cell count although , again , not when adjusting for baseline pVL ( Table 4 ) . Having explored the impact of single exhaustion markers on clinical outcome , we next measured whether their co-expression would predict progression ( Fig D in S1 Text and Table A in S1 Text ) . In survival analyses PD-1/Lag-3 co-expression on CD8 and CD38 CD8 T cells ( Fig Da and Dd in S1 Text; log-rank test , p = 0 . 012 and p = 0 . 027 respectively ) , and PD-1/Tim-3 co-expression on CD38 CD8 T cells ( Fig Dj in S1 Text , p = 0 . 01 ) were associated with faster time to the trial primary endpoint , although not a better predictor than PD-1 alone . When considering time since seroconversion , the effects of PD-1/Lag-3 and PD-1/Tim-3 co-expression on bulk CD8s on clinical progression seemed more pronounced in patients recruited more than 12 weeks after seroconversion , compared with those recruited within 12 weeks ( p = 0 . 030 and p = 0 . 024 , respectively ( Fig Dc and Di in S1 Text ) ) . However , when restricting the analyses to activated CD38 CD8 T cells , PD-1/Lag-3 and PD-1/Tim-3 co-expression were , in contrast , associated with progression in individuals recruited within 12 weeks of seroconversion ( p = 0 . 008 , p = 0 . 008 , respectively ( Fig De and Dk in S1 Text ) ) . There was no effect associated with Tim-3/Lag-3 co-expression . In Cox proportional hazards models , there was a strong association with high PD-1/Tim-3 expression on CD8 T cells ( above the median ) with faster progression ( HR 1 . 64 ( 95% CI 1 . 04 . 2 . 60 ) ; p = 0 . 034 ) , after adjusting for baseline CD4 T cell count , ART and pVL . The significant Kaplan-Meier analysis for PD-1/Lag-3 co-expression was not supported by Cox models when adjusting for baseline pVL . Having determined associations between PD-1 , Tim-3 and Lag-3 expression and clinical progression , we next identified the CD8 T cell memory subsets on which these markers were present ( Fig E in S1 Text and Fig 4 ) . We analysed samples from 16 participants of a second PHI cohort , HEATHER ( ‘HIV Reservoir targeting with Early Antiretroviral Therapy’ ) , with similar demographics and inclusion criteria to SPARTAC ( Table B in S1 Text ) . Similar to previous reports in PHI we found that naïve , central memory ( TCM ) , effector memory ( TEM ) and TEMRA constituted 14 . 1 , 4 . 4 , 49 . 6 and 25 . 9% ( median values ) of the CD8 T cell population , respectively ( Fig 4a ) . PD-1 and Lag-3 had similar distributions , with significantly higher expression in TEM compared with all other T cell subsets . Although Tim-3 expression was highest amongst TEM , there was relatively less expression on TCM compared with PD-1 or Lag-3 . Expression of all three markers on naïve cells was very low ( especially for PD-1 and Tim-3 ) . To determine the functional significance of PD-1 , Tim-3 and Lag-3 expression in PHI , we compared expression with the T-box transcription factor T-bet , eomesodermin ( Eomes ) and CD39 . In HIV-1 infection , CD8 T cells which are T-betdim/Eomeshi are associated with an exhausted functional phenotype with reduced polyfunctionality[24] . CD39 also identifies exhausted CD8 T cells in HIV-1 infection , and is associated with a transcriptional signature of T cell exhaustion[25] . We , therefore , explored whether these markers were co-expressed with Tim-3 , Lag-3 and PD-1 in early HIV-1 infection in the HEATHER cohort . At the nearest available pre-therapy time-point to seroconversion T-betdim/Eomeshi CD8 T cells had significantly higher levels of expression of both PD-1 and Lag-3 ( Fig 5 ) . This population also expressed significantly higher levels of CD38 and CD39 but , interestingly , lower levels of Tim-3 . We found the expression of CD39 to be bimodal with 50% of participants expressing <1% ( Fig F in S1 Text ) , likely due to polymorphisms in CD39 , as recently reported by Roederer et al [26] . There was no evidence for increased co-expression of Tim-3 , Lag-3 and PD-1 in the CD39 ‘positive’ ( i . e . >1% expression ) group compared with those expressing <1% . However , for those individuals with >1% CD39 expression , there was evidence for correlation between levels of CD39 and other exhaustion markers ( Fig F in S1 Text ) .
T cell immune exhaustion has been described as the loss of effector function as a result of repeated antigenic stimulation in persistent infections . Certain cell surface-expressed proteins—also called ‘immune checkpoint receptors’ ( ICR ) –have been associated with a continuous loss of T cell effector function [17 , 27] , and have attracted much interest as mediators of T cell exhaustion and possible targets for cancer therapies [28–30] . Here , we evaluated the association of ICRs with clinical progression in early HIV-1 infection within the SPARTAC trial . Our results show that ICR expression at the time of enrolment into the study is linked with clinical progression , but that this relationship is highly complex with different ICR markers having a variable impact depending on their co-expression with other exhaustion and activation markers , and as well as the timing of diagnosis of PHI . Expression of PD-1 , Tim-3 and Lag-3 was increased in PHI compared with HIV-uninfected controls and was predominantly found on the CD8 TEM subset . We found strong evidence to support an association between ICR expression and HIV-1 pVL , in line with previous observations [12 , 20] , although not for Tim-3 , which contrasts with reports in chronic HIV infection [18] . However , when gated on activated CD38 CD8 T cells , all ICRs correlated strongly with pVL . Expression of CD38 on CD8 T cells is a well-documented marker for HIV-1 disease progression [31] although , surprisingly , we found no evidence supporting this association in our study of primary HIV-1 infection . The increased expression of ICR markers on TEM , the association with pVL ( and also with CD4 T cell count for PD-1 ) and the increased strength of the relationship when gated on activated cells are together supportive of a close relationship between exhausted HIV-specific T cell immunity and disease progression . To test this assertion , we turned to primary endpoint data from a randomised controlled trial exploring the impact of short-course ART in PHI on clinical outcomes . In this sub-analysis of the SPARTAC RCT , there was evidence for PD-1 , Lag-3 and Tim-3 predicting disease progression as determined by the time taken to reach the trial primary endpoint ( either CD4 T cell count <350 cells/μl or ( re ) initiation of ART ) , but the strength of the association varied according to factors such as co-expression with another ICR marker , expression of CD38 , and time since seroconversion . Although not evident in the survival analyses , in Cox models only PD-1/Tim-3 co-expression significantly predicted progression when adjusting for CD4 count , ART and pVL . Despite correlating with pVL , neither Lag-3 nor Tim-3 alone were predictive of clinical progression . Lag-3 is a major histocompatibility complex ( MHC ) class II ligand that negatively interferes with the positive signal cascade derived from the TCR–MHC interaction . The role of Lag-3 in HIV-1 infection is not yet fully elucidated and data are conflicting[32][33 , 34] . Although we observed a clear correlation with pVL , sole Lag-3 expression on CD8 and CD38 CD8 T cells did not translate into an association with clinical disease progression . Tim-3 expression showed distinct associations with clinical disease progression and may even be antagonistic to PD-1 early in HIV-1 infection; Tim-3 was associated with delayed progression in participants identified within 12 weeks of seroconversion , but no effect was seen for Tim-3/PD-1 co-expression . This protective Tim-3 effect was not seen for individuals recruited later than 12 weeks after seroconversion , for whom PD-1/Tim-3 co-expression was disadvantageous . The mechanism of the immuno-regulatory effect of Tim-3 is not clear [35] , but this marker has been associated with dysfunction of HIV-1 specific CD8 T cells in chronically infected individuals [18] . Further evidence for the complex role of Tim-3 can be found in the LCMV model where it was only transiently up-regulated during acute infection [36] , whereas in chronic infection PD-1/Tim-3 T cells were more exhausted with impaired proliferative capacities than PD-1 single positive T cells [36] . Tim-3 may be crucial in the early phases of HIV-1 infection to ensure the differentiation of antigen specific effector T cells whereas later in infection continuous Tim-3 signalling may lead to T cell exhaustion [35] . Our study has several limitations . First , we evaluated the expression of ICRs on CD8 and CD8 CD38 T cells irrespective of their restriction , although we showed that the majority of ICR expression was on the TEM memory subset . Further in depth longitudinal analyses of these responses in HIV-1 specific T cell populations in conjunction with their activation status will be needed to confirm our findings . Secondly , other investigators have suggested that in acute HIV-1 infection PD-1 expression does not necessarily equate to exhaustion [37] . We could not evaluate the functionality of the immune responses to compare with exhaustion T cell status due to restricted sample availability . However , we demonstrated strong associations of these ICRs during PHI with the T-betdim/Eomeshi CD8 population and CD39 expression . This provides evidence of functional exhaustion of these ICR-expression cells as T-betdim/Eomeshi and CD39 expressing CD8 T cells have been shown to have reduced polyfunctionality , cytokine expression and exhausted transcriptional profiles [24 , 25 , 38] . Ideally , one would be able to tease out the overlap between T cell exhaustion and activation—which functionally appear to lie on a similar spectrum . However , the different pathways associated with each ICR marker , as well as their low levels of co-expression , might suggest that the phenotypes are mechanistically separate and that they are independent targets for future therapies and interventions . The SPARTAC trial consisted of different ART durations and we were not able to show any differences between trial arms , rendering the impact of different ART durations in PHI on the expression of immune exhaustion markers unanswered . Nevertheless , we did correct for the various treatment arms in the multivariable Cox model , and still found significant correlations with exhaustion marker expression and markers of disease progression . Finally , due to the lack of remaining samples from SPARTAC , we turned to a second cohort of individuals sampled during PHI ( HEATHER ) to explore ICR expression on memory CD8 T cell subsets in conjunction with T-bet , Eomes and CD39 . Different antibody clones and fluorochromes were used for these two cohorts and this will have an impact on measured levels of expression , particularly notable for PD-1 . However , in none of our analyses do we combine data from the two cohorts and as these two PHI cohorts were very similar in terms of inclusion criteria and demographics , we would expect our findings to be closely related . Of note the characteristics of the HEATHER and SPARTAC participants were comparable , both largely MSM UK infected individuals with B clade virus identified with maximum of 6 months from a previous HIV negative test all starting immediate ART at PHI diagnosis . In conclusion , our study is the first to show an association between ICR expression and clinical disease progression in a large cohort of 122 HIV-1 seroconverters . However , it is clear that the relationship between T cell immune activation , immune exhaustion , individual ICR markers and clinical outcomes is complex and further studies will be needed to clarify this . Our data suggest that interventions aimed at reversing T cell exhaustion and restoring T cell functionality may be most successful if applied shortly after HIV-1 acquisition , before an exhausted T cell phenotype is established . This may be also be pertinent for enhancing interventions which aim to reactivate the latent HIV reservoir to facilitate immune-directed killing in novel curative HIV strategies .
The SPARTAC trial was approved by the following authorities: Medicines and Healthcare products Regulatory Agency ( UK ) , Ministry of Health ( Brazil ) , Irish Medicines Board ( Ireland ) , Medicines Control Council ( South Africa ) , and the Uganda National Council for Science and Technology ( Uganda ) . It was also approved by the following ethics committees in the participating countries: Central London Research Ethics Committee ( UK ) , Hospital Universitário Clementino Fraga Filho Ethics in Research Committee ( Brazil ) , Clinical Research and Ethics Committee of Hospital Clinic in the province of Barcelona ( Spain ) , The Adelaide and Meath Hospital Research Ethics Committee ( Ireland ) , University of Witwatersrand Human Research Ethics Committee , University of Kwazulu-Natal Research Ethics Committee and University of Cape Town Research Ethics Committee ( South Africa ) , Uganda Virus Research Institute Science and ethics committee ( Uganda ) , The Prince Charles Hospital Human Research Ethics Committee and St Vincent's Hospital Human Research Ethics Committee ( Australia ) , and the National Institute for Infectious Diseases Lazzaro Spallanzani , Institute Hospital and the Medical Research Ethics Committee , and the ethical committee Of the Central Foundation of San Raffaele , MonteTabor ( Italy ) . HEATHER ( ‘HIV Reservoir targeting with Early Antiretroviral Therapy’ ) was approved by the West Midlands—South Birmingham Research Ethics Committee reference 14/WM/1104 . Ethical approvals include use of samples for the studies described . All samples were analysed anonymously . Analyses associating markers with disease progression used cryopreserved PBMCs from SPARTAC samples taken at baseline ( prior to any ART , if prescribed ) . Cell surface staining for flow cytometry was performed with: Live/Dead Pacific Blue 2ug/100ul ( Invitrogen ) , CD19 Pacific Blue 0 . 4 μg/100μl ( SJ25-C1 ) , CD3 Pacific Orange 0 . 4 μg/100μl ( UCHT1 ) , CD8 PE-Cy5 PerCP 0 . 4 μg/100μl ( SK1 ) [BD Bioscience] , PD-1 APC 0 . 5 μg/100μl ( MIH4 ) [eBioscience] , Tim-3 PE 0 . 05 μg/100μl ( 344823 ) [R&D] , Lag-3 FITC 2 μg/100μl ( 17B4 ) [LifeSpan Bioscience] and CD38 PE-Cy7 0 . 4 μg/100μl ( HIT-2 ) [Biolegend] . The isotype controls for Lag-3 was IgG2a isotype FITC 0 . 4 μg/100μl ( C45 ) [AdB serotech] , for PD-1 was IgG1 isotype APC 0 . 4 μg/100μl ( p3 ) [eBioscience] and for Tim-3 was IgG2a isotype PE 0 . 4 μg/100μl ( R35-95 ) [BD Bioscience] . Cell population gating was performed based on mean fluorescence intensity “minus one” ( FMO ) [39] and unstained controls ( Fig A in S1 Text ) . Samples were acquired on a LSR II ( BD ) with standard laser configurations and analysed using FlowJo Version 8 . 7 . 7 . Analyses examining ICR expression on T cell memory subsets used cryopreserved PBMCs from the HEATHER cohort . Cells were stained in BD Horizon Brilliant Stain Buffer ( BD ) containing all antibodies and Live/Dead Near IR at 1 in 300 dilution ( Life Technologies ) in 96 well-V bottom plates at 4°C . PBMCs were stained with the following antibodies: CD3 Brilliant Violet ( BV ) 570 0 . 16 μg/100μl ( UCHT1 ) , CCR7 Pacific Blue 1 . 8 μg/100μl ( G043H7 ) [BioLegend] , CD4 BV 605 0 . 05 μg/100μl ( RPA-T4 ) , CD8 BV 650 0 . 012 μg/100μl ( RPA-T8 ) [BD] , PD-1 PE eFluor 610 0 . 2 μg/100μl ( eBioJ105 ) , Lag-3 PE-Cy7 0 . 024 μg/100μl ( 3D5223H ) , CD45RA FITC 0 . 4 μg/100μl ( HI100 ) [eBioscience] and Tim-3 PE ( as above ) . For characterisation of T-bet/Eomes expression , PBMCs were stained as above in PBS with 5% fetal bovine serum and 1mM EDTA containing Live/Dead Near IR , anti-PD-1 , anti-Tim-3 , anti-Lag-3 , along with antibodies to CD39 BV 421 0 . 1 μg/100μl ( A1 ) and CD38 AlexaFluor 700 0 . 1 μg/100μl ( HB-7 ) [BioLegend] . Fixation and permeabilisation were performed with Foxp3 Buffer Set ( BD ) as per manufacturer’s instructions in reduced volumes to facilitate staining in 96-well plates . Staining of intracellular epitopes was performed at room temperature in PBS containing 0 . 5% BSA and 0 . 1% sodium azide with antibodies to CD3 , CD4 , CD8 ( as above ) , and T-bet FITC 2 μg/100μl ( 4B10 ) [BioLegend] and Eomes eFluor660 0 . 024 μg/100μl ( WD1928 ) [eBioscience] . All samples were acquired on a LSR II ( BD ) . Data were analysed using FlowJo Version 10 . 8 . 0r1 ( Treestar ) . Naïve T cells were defined as CD45RA+/CCR7+ , TCM as CD45RA-/CCR7+ , TEM defined as CD45RA-/CCR7- and TEMRA as CD45RA+/CCR7- . ( Fig Ea in S1 Text ) . CD8 T cells were divided into T-bethi/Eomesdim and T-betdimEomeshi populations as described by Buggert et al [24] ( Fig 5a ) . Gates for exhaustion marker positive populations were set on a partially stained anchor sample without the relevant antibody such that <0 . 1% of events were positive for the marker ( Fig Eb and Ec in S1 Text ) . Associations between T cell exhaustion markers on CD8 and CD38 CD8 positive cells and pVL or CD4 T cell count were evaluated using Spearman correlations . The fitted line , superimposed in the relevant scatterplots , was estimated through linear regression . When multiple markers were tested for correlations with other variables ( e . g . baseline CD4 T cell count or HIV-1 RNA ) , the adjusted overall critical p-values were also reported . Adjustments were performed using the method of Simes [40] targeting on an overall false discovery rate of ≤0 . 05 . Association of exhaustion markers with clinical progression ( time to trial end-point ) was assessed by Kaplan-Meier survival curves , log rank tests and Cox proportional hazards models . Covariables considered for all Cox models included ART treatment ( as a time updated binary variable ) , baseline CD4 T cell count and pVL . Effects of the time interval between the estimated date of seroconversion and a baseline exhaustion marker’s measurement ( as a binary variable i . e . >12 or ≤12 weeks ) and its first order interactions with the aforementioned covariates or the exhaustion markers under investigation were also assessed and , if significant , included in the final multivariable Cox model . Finally , we checked for any effects of the time gap between the last negative and first positive HIV tests used to estimate the seroconversion date . These analyses were performed using Stata 11 ( Stata Corp . , TX USA ) . P-values<0 . 05 were considered as statistically significant . In the analyses from the HEATHER cohort , comparison of exhaustion marker expression across three or more groups was performed using Friedman’s test ( non-parametric , paired analysis of variance ) . Where a difference was found , subsequent pairwise comparisons between groups ( Dunn’s test ) were performed with adjustment for multiple comparisons targeting on overall significance level of 0 . 05 . Expression of exhaustion markers between two T cell subsets was compared using Wilcoxon matched-pairs signed rank test . Statistical analyses were performed using GraphPad Prism 6 . 0f . | After being infected with HIV , the pace of disease progression is highly variable between individuals . Some stay well with functioning immune systems for many years , whilst others progress to AIDS quickly . Understanding the factors that underpin these differences is important and may relate to factors such as viral adaptation and immune exhaustion . Recently , there has been interest in certain molecules—called ‘exhaustion’ or ‘immune checkpoint’ markers—which reflect how well the immune system functions . Recent trials show that therapies directed against these molecules can improve anti-cancer immunity . It is known from laboratory experiments that these markers are abundant in HIV infection suggesting that the human immune response to HIV is not fully effective . The relevance of these markers in patient cohorts remains unclear . This study measures three exhaustion markers—PD-1 , Tim-3 , Lag-3 –in individuals with HIV recruited to a randomised controlled trial of therapy in early HIV infection called SPARTAC . We find a complex picture in which these markers alone , together and in combination with other markers that reflect T cell activation ( CD38 ) help predict the speed of clinical progression and immune decline , with differing effects dependent on the duration of infection . We propose that therapies directed against these markers could impact disease progression , vaccine efficacy or even newer curative strategies . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"blood",
"cells",
"t",
"helper",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"antiviral",
"therapy",
"pathogens",
"immunology",
"microbiology",
"retroviruses",
"viruses",
"i... | 2016 | Exhaustion of Activated CD8 T Cells Predicts Disease Progression in Primary HIV-1 Infection |
The NF-κB-like velvet domain protein VosA ( viability of spores ) binds to more than 1 , 500 promoter sequences in the filamentous fungus Aspergillus nidulans . VosA inhibits premature induction of the developmental activator gene brlA , which promotes asexual spore formation in response to environmental cues as light . VosA represses a novel genetic network controlled by the sclB gene . SclB function is antagonistic to VosA , because it induces the expression of early activator genes of asexual differentiation as flbC and flbD as well as brlA . The SclB controlled network promotes asexual development and spore viability , but is independent of the fungal light control . SclB interactions with the RcoA transcriptional repressor subunit suggest additional inhibitory functions on transcription . SclB links asexual spore formation to the synthesis of secondary metabolites including emericellamides , austinol as well as dehydroaustinol and activates the oxidative stress response of the fungus . The fungal VosA-SclB regulatory system of transcription includes a VosA control of the sclB promoter , common and opposite VosA and SclB control functions of fungal development and several additional regulatory genes . The relationship between VosA and SclB illustrates the presence of a convoluted surveillance apparatus of transcriptional control , which is required for accurate fungal development and the linkage to the appropriate secondary metabolism .
Velvet domain transcription factors interconnect fungal developmental programs and secondary metabolism and affect a significant part of differential gene expression during development in comparison to vegetative growth [1] . The majority of the fungal target genes of velvet domain proteins , which bind to promoters of thousands of genes by their Rel homology-like domain , is yet elusive [2 , 3] . This fungal protein family is highly conserved in ascomycetes and basidiomycetes [4 , 5] . The velvet proteins VosA ( viability of spores A ) and VelB ( velvet-like B ) can form homodimers as well as the VosA-VelB heterodimer to repress or activate gene expression [2 , 6–9] . VosA represses brlA ( bristle A ) expression encoding a master regulator for the initiation of conidia formation , which are the asexual spores of the fungus . VosA-VelB later activates within conidia the gene encoding the transcription factor VadA ( VosA/VelB-activated developmental gene ) , which downregulates brlA expression to allow the maturation of viable conidia [7] . Full suppression of conidiation during vegetative growth of the hyphae require direct binding of VosA and a second brlA-repressor , NsdD ( never in sexual development D ) to the brlA promoter [2 , 8 , 9] . Growth of fungal filaments after the germination of spores is in the first hours not responsive to external signals , because developmental regulatory genes are not expressed . De-repression of brlA accompanies the achievement of developmental competence of fungal hyphae approximately 18 to 20 h post germination [8 , 10] . This derepression is characterized by delocalization of VosA and NsdD from the brlA promoter , which allows the Flb proteins ( fluffy low brlA ) FlbB , FlbC , FlbD and FlbE to activate brlA expression [8 , 9 , 11–15] . A second layer of conidiation repression during vegetative growth is carried out by SfgA ( suppressor of fluG ) , which negatively regulates expression of the flb genes . FluG ( fluffy G ) accumulates to a certain threshold during ongoing vegetative growth , which removes the repressive effects of SfgA upon conidiation [16 , 17] . The Flb proteins activate brlA in two distinct cascades: FlbB/FlbE→FlbD→brlA and FlbC→brlA [11–15 , 18 , 19] . The fifth Flb protein , FlbA , regulates development in an indirect manner by antagonizing a G-protein mediated repression of conidiation , and thereby represses vegetative growth [20–22] . The C2H2 transcription factor BrlA activates abaA ( abacus A ) in the mid phase of conidiation [23] . AbaA activates wetA ( wet-white A ) in the late phase of conidiation , which is necessary for the synthesis of conidiospore wall components [4 , 24 , 25] . VosA is involved in time tuning of conidiation: it represses brlA until developmental competence is achieved and is activated by AbaA and WetA downstream of BrlA during late asexual growth [4 , 26] . VosA regulates conidiospore viability during ongoing spore formation in Aspergilli through activation of genes which products are important for spore maturation [4 , 6 , 27–29] . VosA and VelB are important for trehalose biogenesis [4 , 27] . Trehalose is a storage compound , which supports conidiospore viability and germination [30–32] . Velvet domain proteins couple fungal differentiation programs to specific secondary metabolisms for sexual or asexual development and a fifth of the genome is differentially expressed during development in comparison to vegetative growth [1 , 33] . Velvet domain proteins are located at the interface between development and secondary metabolism control [33–36] . A . nidulans is able to produce several secondary metabolites ( SMs ) , such as penicillins , sterigmatocystin , benzaldehydes , emericellamides , orcinol and diorcinol , diindoles , austinol and dehydroaustinol [37–43] . SM genes are often clustered in fungal genomes . Those gene clusters are controlled by cluster-specific transcription factors and master regulators , which interconnect SM biosynthesis and developmental programs in response to environmental cues , such as light [33 , 41 , 44 , 45] . A key element of this interconnection is the velvet complex , comprising the velvet proteins VeA and VelB and the methyltransferase LaeA [27 , 33 , 46–50] . Velvet proteins regulate secondary metabolite gene clusters , as well as downstream master regulators , such as the well conserved MtfA ( Master transcription factor A ) [43 , 51 , 52] . Their regulatory versatility suggests a complex hierarchy of multiple control layers of genetic networks mutually controlled by distinct transcription factors . The zinc cluster ( C6 ) protein SclB acts as activator of a genetic network , which was characterized by genome-wide transcriptional analyses and which represents a novel downstream-target for inhibition of the velvet domain protein VosA in the fungal model organism A . nidulans . SclB interconnects the formation of asexual spores and the enzymatic as well as non-enzymatic responses upon oxidative stress to a distinct secondary metabolism .
A ΔsclB strain was generated to analyze the differences in gene expression in the absence of sclB compared to A . nidulans wildtype . The complete sclB ORF in this ΔsclB strain was exchanged with a recyclable marker cassette leaving only a small six site as scar ( 100 nucleotides ) after recycling [72] . RNA of wildtype , ΔsclB and a sclB complemented ( sclB comp ) strain were extracted from submerged cultures grown for 24 h under constant agitation and sequenced to compare genome-wide transcriptional changes in the presence or absence of sclB . The reintroduction of the sclB ORF fully complemented all effects on transcription in the ΔsclB strain resulting in transcriptomes comparable to wildtype . 169 genes were significantly increased and 239 were significantly decreased in ΔsclB compared to wildtype with a threshold of at least two fold for upregulation or downregulation ( Log2 fold change ( FC ) of at least 1 ) ( S1 Table ) . Analyses employing the Aspergillus Genome Database ( AspGD ) [64] and the Fungal and Oomycete Genomics Resources Database ( FungiDB ) [73] were conducted to categorize these genes into functional groups ( Fig 2 ) . 13 genes were assigned to carbon metabolism , one to sulfur metabolism and 9 to other metabolic functions of the genes upregulated in ΔsclB compared to wildtype . Genes connected to secondary metabolism constitute the largest group ( 18 ) with an assigned function . Several genes related to the respiratory chain ( 6 ) or transmembrane transport ( 11 ) were also upregulated in ΔsclB compared to wildtype . Four genes were assigned to the response to oxidative stress and one is assigned to menadione induced stress . One gene of the group of upregulated genes in ΔsclB compared to wildtype is linked to development . The largest group among the genes downregulated in ΔsclB compared to wildtype with an assigned function is related to secondary metabolism ( 18 ) . Other large groups are constituted of genes connected to development ( 17 ) or transmembrane transport ( 15 ) . Several genes related to carbon metabolism ( 9 ) , sulfur metabolism ( 2 ) or amino acid biosynthesis ( 6 ) were found as well amongst the downregulated genes in ΔsclB compared to wildtype , as well as genes related to the response to oxidative stress ( 9 ) or to other stresses ( 6 ) . Members of eight different SM gene clusters were amongst the genes upregulated and 10 amongst the genes downregulated in ΔsclB compared to wildtype ( Table 1 and S1 Table ) . This equals approximately 25% of all predicted secondary metabolite gene clusters in A . nidulans ( Table 1 and S1 Table ) [74] . Genes encoding backbone enzymes of four of these clusters were upregulated ( AN3396 , AN3252 , AN6784 and AN1242 ) and six were downregulated ( AN6236 , AN9244 , AN8383 AN2064 , AN9226 and AN2924 ) . This equals approximately 14% of all backbone enzymes of secondary metabolite gene clusters in A . nidulans [74] . Taken together , a significant part of the transcriptome is differentially expressed when the ΔsclB strain was compared to wildtype , with even 1 . 5 times more genes with decreased than with increased transcription . Most differentially regulated genes , for which a function could be assigned , are related to secondary metabolism and genes related to development . Another large part of genes differently regulated in the absence of sclB compared to the wildtype situation are genes related to stress response , especially of the response towards oxidative stresses . The A . niger scl-2 mutant forms reduced numbers of conidiophores and structures similar to sclerotia [53] , whereas a deletion of the sclB orthologous gene in A . fumigatus ( Afu6g11110 ) did not result in any obvious phenotype when grown on minimal medium ( S2 Fig ) . Transcriptomic analyses of the ΔsclB strain compared to wildtype in A . nidulans suggested that SclB is involved in asexual development ( Fig 2 and S1 Table ) . The growth and differentiation of the ΔsclB mutant strain was examined during light and unlimited oxygen supply promoting asexual spore formation in comparison to cultivation in dark with limited oxygen supply supporting sexual development ( Fig 3A ) . A . nidulans wildtype forms high numbers of conidiophores carrying asexual spores in light and produces lower numbers of asexual spores in dark after a delay of several days [1] . The absence of sclB leads to a significantly decreased formation of conidiophores during asexual or sexual development , compared to wildtype ( Fig 3A and 3B ) . This phenotype of the A . nidulans ΔsclB strain was fully restored by re-introducing either the sclB ORF into ΔsclB ( sclB comp ) or the sclB ortholog from A . fumigatus ( Afu6g11110 ) sharing 55% similarity , indicating functional conservation ( S2 Fig ) . Quantification of conidiospore formation in light revealed that the ΔsclB strain produced less than 5% of the asexual spores produced by the wildtype after two days and reached a maximum of approximately 20% of the wildtype conidia after 10 days . A . nidulans reduces conidiophore formation during growth in the dark and favors cleistothecia formation . The ΔsclB strain produced significantly less conidiospores during growth in the dark in comparison to light suggesting that light control of development is independent of SclB . Overexpression of sclB ( sclB OE ) under control of a nitrate-inducible promoter ( PniaD ) further increases asexual spore formation in the dark , when the wildtype produced only low amounts of conidia ( Fig 3A ) . Sexual development includes nest formation and the differentiation of cleistothecia as closed fruiting bodies , which is increased in the dark and reduced in light . Cleistothecia formation is similar in the ΔsclB strain in comparison to wildtype and additional control strains suggesting that SclB control is rather targeting asexual than sexual development ( Fig 3C ) . The sclB OE strain increased the production of conidiophores significantly when grown under inhibiting and delaying conditions in the dark under limited oxygen supply , when the wildtype only produced small amounts of conidiophores and the formation of cleistothecia is favored ( Fig 3 ) . This effect in the sclB OE strain is even more pronounced when instead of point inoculated colonies leading to radial zones of different ages [75]; ( Fig 3A upper part ) , plated colonies emerging from separated germinating spores were monitored . Plated colonies form a coherent mycelium due to hyphal fusion through anastomosis tubes , and are of same age at every spot ( Fig 3A lower part , Fig 3B and 3C ) [76 , 77] . These data indicate that SclB is required for significant , efficient and accelerated conidiophore formation of A . nidulans . ChIP-on-Chip experiments showed that VosA binds the sclB promoter in vivo approximately 311 bp upstream of the sclB ORF [2] . Promoter walking electrophoretic mobility shift assays ( EMSAs ) revealed that VosA binds a 40 bp region of the sclB promoter ( marked in Fig 1 ) . EMSAs of this region and purified VosA protein verified dosage-dependent VosA binding in vitro ( Fig 4A ) . In the EMSA protein-DNA complexes run high in the gels and free DNA runs in the lower part . Possible formation of GST-VosA dimers might lead to binding of more than one DNA molecule at the same time . Two putative binding sequences were identified in this region and mutations for both of them , in which the respective putative binding sequence was deleted , showed that VosA specifically binds nine bps , spanning -337 to -329 in front of the sclB ORF ( Fig 4A ) . A vosA deletion mutant ( ΔvosA ) was constructed to analyze the impact of VosA upon sclB gene expression . Transcription levels of sclB were monitored in wildtype and ΔvosA strain with quantitative real-time PCR ( qRT-PCR ) . sclB transcription is upregulated in the absence of vosA in asexually grown colonies 24 h post induction of development ( Fig 4B ) . This indicates a repressing effect of VosA towards sclB expression during asexual development . This is in accordance with transcriptomic data showing an upregulation of sclB gene expression in conidiospores of a ΔvosA strain in comparison to wildtype published by Park and co-workers [78] . AbaA and WetA activate vosA during late asexual development . VosA together with VelB is necessary for trehalose biogenesis to support spore viability [4 , 6] . Spore viability was compared in ΔsclB and sclB OE strains on solid minimal medium . Conidiospores of the ΔsclB strain showed a rapid loss in spore viability compared to spores of wildtype , sclB comp and sclB OE strains after seven days and thereafter ( Fig 4C ) . A similar loss in spore viability was found for the ΔvosA strain , whereas conidiospores of the ΔvosAΔsclB double mutant strain showed further diminished viability after seven days and thereafter . The ΔvosA single mutant produces grey-greenish conidiospores with decreased viability [4] ( Fig 4D ) . The ΔvosAΔsclB double deletion strain supports an epistatic interaction of sclB towards vosA , because it showed the ΔsclB single mutant phenotype of reduced conidia formation with low spore viability ( Fig 4C and 4D ) . These findings place the gene encoding SclB genetically downstream of the gene for VosA . VosA binds upstream of sclB and represses sclB gene expression . VosA acts as homodimer or forms with VelB or VelC the heterodimers VosA-VelB or VosA-VelC [6 , 79] , which fulfill different functions in fungal development and interconnected secondary metabolism . Double deletions of sclB and velB or velC , respectively , were created to discriminate between SclB functions downstream of the VosA-VosA homodimer or the VosA-VelB and VosA-VelC heterodimers . veA was included into these analyses , because VeA competes with VosA for VelB and forms the VeA-VelB heterodimer . The ΔveA and ΔvelB single mutants are unable to form cleistothecia on minimal medium and are misregulated in secondary metabolism producing dark reddish pigments [6 , 33 , 52] ( Fig 4D ) . The ΔsclBΔveA and ΔsclBΔvelB double mutants both show additive phenotypes with impaired asexual and sexual development . The loss of cleistothecia formation of the ΔveA and ΔvelB single mutant is combined with increased amounts of aerial hyphae without conidia and significantly smaller greenish colony centers representing conidiophores . This indicates a SclB function for conidiophores independently of the VeA or VelB governed pathways for fruiting bodies and the corresponding secondary metabolism . The ΔvelC single mutant shows an almost wildtype-like phenotype on minimal medium combined with increased amounts of conidiophores [79] . The ΔsclBΔvelC double deletion strain shows an intermediate phenotype with a colony similar to the ΔsclB phenotype combined with an increased greenish colony center for conidiophores . Therefore , SclB functions independently of the velvet protein heterodimers VosA-VelB or VosA-VelC and is primarily a repression target of the VosA homodimer . SclB functions downstream of VosA and its absence leads to decreased conidiophore formation , whereas the sclB OE strain produces increased numbers of conidiophores during sexual development . This indicates that SclB is an activator of conidiophore formation . Strains were grown in liquid minimal medium to test whether an overexpression of sclB is sufficient to induce development under vegetative conditions . Growth in submerged cultures suppresses development in A . nidulans and results in solely vegetative growth of the wildtype ( Fig 5A ) . No conidiophores were found in wildtype , ΔsclB or sclB comp strains grown in submerged cultures . In contrast , the sclB OE strain forms conidiophores after 18 h of growth in submerged cultures ( Fig 5A ) . VosA represses gene expression of the master regulator-encoding brlA , and a ΔvosA strain forms conidiophores when grown in submerged culture conditions [4] . The expression of brlA was examined in the sclB OE mutant during vegetative growth . Strains were grown under submerged conditions what hinders asexual development in the wildtype . The wildtype only expresses basal levels of brlA under these conditions . In contrast , mRNA levels of brlA are highly upregulated in the presence of high amounts of SclB in the sclB OE strain ( Fig 5B ) . VosA represses brlA during vegetative growth and brlA gene expression was upregulated in the ΔvosA strain grown under submerged culture conditions as well ( Fig 5B ) [4 , 8] . Expression of brlA in a ΔvosA mutant in the sclB OE background was tested to examine , whether SclB is able to activate brlA gene expression . Whereas brlA expression was already upregulated about 40 times in sclB OE compared to wildtype , the ΔvosA sclB OE mutant showed even more than 400 times upregulation compared to wildtype ( Fig 5B ) . This additional upregulation indicates that SclB is able to activate brlA expression in the absence of vosA . Activation of the conidiation pathway is inhibited by the repressors VosA and NsdD during vegetative growth , which are released from the brlA promoter when the fungus becomes developmentally competent [4 , 8 , 9] . SfgA represses conidiation indirectly by regulating the genes for the Flb factors [16 , 80] . Expression levels of sfgA , nsdD and vosA were analyzed by qRT-PCR in sclB mutant strains to exclude the possibility that SclB influences the conidiation pathway by downregulating gene expression of these repressors ( S3A Fig ) . Gene expression of none of these repressor genes is altered in ΔsclB or sclB OE strains in comparison to wildtype . This demonstrates that SclB does not control the conidiation pathway through repression of its repressor genes . Taken together , the presented data indicate that SclB is an activator of the conidiation pathway through the brlA activator gene . The ΔbrlA bristle mutant phenotype of primarily stalks with diminished conidia ( Fig 5C ) is distinctly different from the ΔsclB phenotype . The ΔsclBΔbrlA double mutant resembles the ΔsclB single mutant , supporting an epistasis of sclB towards brlA ( Fig 5C ) . This underlines a function of SclB upstream of brlA in developmental programs . In addition , epistasis of sclB and abaA , a downstream factor of brlA [81] , was analyzed . ΔabaA forms brownish conidiophores with intermittent tumefactions , which are distinctly decreased in number [82] ( S3B Fig ) . The ΔsclBΔabaA mutant shows the ΔsclB single mutant phenotype but has lost the greenish colony center ( S3B Fig ) . This shows that sclB is epistatic to abaA and corroborates the finding that SclB activates the conidiation cascade upstream of its major regulator BrlA . An increased brlA expression directly leads to spore formation from vesicle-like structures [83] , whereas sclB OE activating brlA expression forms conidiophores under submerged culture conditions . Upstream activators of brlA were analyzed to examine whether SclB activates further regulatory genes of asexual development upstream of brlA . FluG is a key upstream activator of the conidiation pathway and acts as a time-dependent repressor of the conidiation-repressor SfgA [8 , 16 , 17] . The deletion of fluG leads to drastically reduced conidiation and a fluffy whitish phenotype with low amounts of conidiophores and high amounts of aerial hyphae [17] ( Fig 6A ) . The back of the colony shows a light orange color indicating an alteration in secondary metabolite production . sclB was knocked out in the ΔfluG strain to analyze epistatic interactions . The ΔfluGΔsclB double mutant strain shows an additive phenotype with large amounts of aerial hyphae , but completely failed to produce conidiophores ( Fig 6A ) . In addition , the orange color was less bright . The ΔfluG phenotype was not rescued by an overexpression of sclB ( Fig 6A ) . This indicates a function of the SclB protein downstream of FluG or the FluG-SfgA pathway . The sclB gene is presumably not a direct downstream target of FluG-mediated gene activation , as sclB OE could not rescue the loss of fluG . Transcription of fluG was increased in qRT-PCR analyses from vegetatively grown ΔsclB strain ( Fig 6B ) . This corroborates that SclB does not function as activator of fluG gene expression . SclB might have repressing effects upon fluG expression during late asexual development ( spore maturation ) , because fluG expression is upregulated in the absence of sclB during asexual growth after 24 h in comparison to wildtype ( Fig 6B ) . The sclB gene expression is decreased in the absence of fluG as well , suggesting regulatory feedback loops or cross talk between both factors and their corresponding genes ( Fig 6C ) . The Flb factors , which act downstream of FluG , activate brlA in two cascades: FlbB/FlbE→FlbD→BrlA and FlbC→BrlA [11–15 , 18 , 80] ( Fig 7 ) . Genome-wide transcriptional ana-lysis showed that flbC and flbD transcript levels are significantly lower in ΔsclB compared to wildtype during late vegetative growth when the fungus reached the state of developmental competence ( S1 Table ) . Transcription of flbB–E was analyzed in more detail through qRT-PCR measurements . flbD gene expression is distinctly lower in submerged cultures in the absence of sclB compared to wildtype ( Fig 7 ) . Moreover , flbC is downregulated in ΔsclB after 24 h of vegetative growth in submerged cultures , but upregulated in the sclB OE strain , compared to wildtype . This is in agreement with the data obtained in genome-wide transcriptomics ( S1 Table ) . Transcription of flbB and flbE is not significantly differentially regulated in the sclB mutants compared to wildtype in qRT-PCR analyses . Nevertheless , expression profiles of both , flbB and flbE in sclB mutants resemble these of flbC and flbD in their tendencies , indicating regulatory effects of SclB upon these factors as well . These analyses suggest an activating role of SclB towards the Flb cascade upstream of brlA and specifically towards flbC and flbD during late vegetative growth at the onset of conidiation . Transcription of flbB , flbC and flbD is upregulated in the absence of sclB compared to wildtype after 24 h of asexual growth . Similarly , the flbA gene for an RGS ( Regulator of G protein Signaling ) domain protein indirectly supporting conidiation [84] , is upregulated during asexual growth in the absence of sclB but not during vegetative growth . These findings indicate that SclB regulation of the conidiation cascade is part of a timely adjusted choreography of asexual development . Single and double knock out strains of the flb genes were created to further investigate the genetic relationship between sclB and the flb genes . All flb single deletions showed fluffy phenotypes [85] that are distinctly different to the ΔsclB phenotype ( Fig 7C ) . Only ΔflbC is an exception with a phenotype similar to ΔsclB , which is in agreement with the finding that SclB activates flbC gene expression . Double deletions of sclB and each of the flb genes showed phenotypes with a complete abolishment of conidiophores ( Fig 7C ) . The ΔflbCΔsclB strain resembles the phenotypes of the other ΔflbΔsclB strains , indicating that SclB functions upstream of both parts of the Flb cascade and underlines the finding that SclB activates flbC and flbD . sclB OE is not sufficient to restore the wildtype phenotype in flb knock out strains , showing that SclB acts upstream of the Flb factors ( S4 Fig ) . Taken together , these findings demonstrate that SclB activates not only brlA but also both Flb cascades through the activation of flbC and flbD , which both merge and further activate brlA . Genome-wide analysis of SclB’s influence on gene expression suggests that approximately 25% of all SM gene clusters in A . nidulans are misregulated in the absence of sclB compared to wildtype ( Table 1 and S1 Table ) . The SclB-regulated interconnection of asexual development and secondary metabolism was examined in more detail by comparing SMs from sclB mutant and wildtype strains . Extracellular SMs were extracted with ethyl acetate from wildtype and the sclB mutant strains either grown for 48 h vegetatively or three and seven days under conditions inducing asexual or sexual development in wildtype . High-performance liquid chromatography ( HPLC ) revealed that the wildtype as well as the sclB OE strain , but not the ΔsclB strain , produce austinol and dehydroaustinol after three and seven days of asexual growth in light . Both compounds were identified in samples extracted from wildtype , the sclB complemented strain and the sclB OE strain according to their masses and UV/VIS absorption maxima ( Figs 8A and S5 ) [86] . ausA , coding for a polyketide synthase producing the intermediate 3 , 5-dimethyl orsellinic acid , and ausF , required for the synthesis of both austinol and dehydroaustinol [39] are not expressed during vegetative growth in wildtype and ΔsclB , but in the sclB OE strain ( Fig 8B ) . A third SM producing gene ausH , which is necessary for austinol and dehydroaustinol production , was basally expressed in wildtype , but not in ΔsclB , whereas the sclB OE strain showed upregulation of ausH transcription ( Fig 8B ) . This is in accordance with transcriptomic data indicating that backbone enzymes of both austinol clusters are downregulated in the absence of sclB compared to wildtype ( Table 1 and S1 Table ) . This indicates that SclB activates expression of the austinol gene cluster during vegetative growth . HPLC coupled to a qToF mass spectrometer revealed that the sclB OE strain produces increased amounts of emericellamide A , C and D [87] during vegetative growth ( Figs 9A and S6 ) . The ΔsclB strain produces only traces of these compounds under tested growth conditions and no fragmentation for emericellamide A and D could be obtained from mass spectrometry ( Fig 9A and S6 ) . Expression of the four genes of the emericellamide gene cluster , easA to easD , was analyzed in vegetatively grown cultures . easA and easD are basally expressed in wildtype . Only easA , but not easB , easC or easD , was basally expressed in the ΔsclB strain . In contrast , all four genes are upregulated in sclB OE ( Fig 9B ) . Furthermore , easD was significantly downregulated in genome-wide transcriptomic analysis in the absence of sclB compared to wildtype ( S1 Tab ) . This shows that SclB acts as activator of the eas gene cluster and is necessary for emericellamide biosynthesis . Taken together , SclB activates the expression of SM clusters for emericellamides , austinol and dehydroaustinol during vegetative growth . The adaptive response to oxidative stress is required for fungal development as endogenous signal and is an important determinant for fungal fitness in corresponding environmental conditions [40 , 88] . SclB is involved in the regulation of spore viability ( Fig 4C ) and genome-wide transcriptional analyses show that several genes related to the response to oxidative stress are differentially expressed when sclB is absent ( Fig 2 and S1 Table ) . Conidiospore survival was tested during H2O2 induced oxidative stress to analyze whether SclB is involved in the regulation of the oxidative stress response as well . Conidiospores of the wildtype , the complemented and the sclB OE strain show a linear loss in spore viability over time in the presence of 100 mM H2O2 ( Fig 10A ) . In contrast , conidiospores of the ΔsclB strain show a more rapid loss in viability over time in the presence of 100 mM H2O2 . Conidiospores from wildtype , sclB comp and sclB OE strains showed survival rates of approximately 86% after 30 min of H2O2 treatment , conidiospores of the ΔsclB strain showed only 62% survival . At the same time point conidiospores of the ΔvosA and the ΔvosAΔsclB strains showed even further reduced viability of only 40% ( ΔvosA ) and 30% ( ΔvosAΔsclB ) , respectively . Similar differences were measured over the whole time period of examination . This suggests that SclB positively regulates the oxidative stress response in A . nidulans . To investigate this further , expression of genes of the oxidative stress response was tested in submerged cultures in the presence or absence of H2O2 . The glutathione and the thioredoxin system are important parts of the oxidative stress response [89–91] . The thioredoxin system is encoded by trxA ( thioredoxin ) and trxR ( thioredoxin reductase ) [90] . trxA was especially induced upon treatment with H2O2 in the sclB OE strain ( S7 Fig ) . trxR is induced in wildtype in the presence of H2O2 but not induced in the ΔsclB strain ( Fig 10B ) . It is also downregulated in the absence of sclB during unstressed growth ( S1 Table ) . The sclB OE strain stressed with H2O2 shows an increased trxR upregulation compared to wildtype ( Fig 10B ) . glrA encodes the glutathione reductase [92 , 93] , which regulation was not dependent on the presence of sclB ( S7 Fig ) . The catA gene , encoding the spore-specific catalase A , is upregulated in wildtype but not induced in ΔsclB in presence of H2O2 ( Fig 10B ) . Expression of catA in the sclB OE strain is already upregulated during unstressed growth . Several transcription factors are involved in the response to oxidative stress . napA encodes the most prominent oxidative stress regulator in A . nidulans . napA gene expression was not found to be significantly regulated under applied conditions ( S7 Fig ) . RsmA is involved in the regulation of SMs and in oxidative stress response [91 , 94] . rsmA expression is around three fold induced in wildtype when H2O2 stress is applied ( Fig 10B ) . In sclB OE the induction of rsmA expression in the presence of H2O2 is even higher ( almost six fold ) , whereas rsmA expression is not induced by H2O2 in the ΔsclB strain . sclB itself is upregulated in wildtype and in sclB OE upon addition of H2O2 in comparison to unstressed situation ( Fig 10B ) . Taken together , these data suggest that SclB is involved in the regulation of the oxidative stress response in A . nidulans and specifically acts as a positive regulator of enzyme encoding genes , such as catA and thioredoxin genes , as well as the transcription factor-encoding gene rsmA . C6 proteins are typical fungal transcription factors . In silico analyses predicted SclB to be localized in the nucleus as determined by CELLO [95] and WoLF PSORT [96] . SclB was fused N- and C-terminally to sGFP to examine subcellular localization in vivo ( S8A Fig ) . The predicted molecular mass of both versions of the SclB GFP-fusion proteins is 87 . 46 kDa . Sizes of both fusion proteins determined by western hybridization are slightly higher than bioinformatically predicted ( S8B Fig ) , indicating posttranslational modifications . Treatment of GFP-SclB crude extracts with Lambda phosphatase resulted in a band shift on a western blot , suggesting that SclB is phosphorylated during vegetative growth ( S8C Fig ) . NetPhos 3 . 1 [97] predicted 28 codons for possible phosphorylation sites ( score value between 0 and 1 , cut off >0 . 7 ) . LC-MS/MS analyses revealed three phosphorylated SclB residues S327 , T464 and S506 in samples derived from vegetatively grown cultures , supporting that SclB is phosphorylated during vegetative filamentous growth ( S9A Fig ) . However , mutation of these residues and two serines adjacent to S506 ( S504 and S505 ) to alanine to mimic constant dephosphorylation ( sclBS327A , T464A , S506A ) or aspartic acid to mimic constant phosphorylation ( sclBS327D , T464D , S506D ) did not result in any obvious phenotype ( S9B Fig ) and the function of these phosphorylation sites therefore remains elusive . Both , the N- and C-terminal GFP fusion of SclB was expressed under control of the native sclB promoter and could complement the loss of sclB , demonstrating , that the fusion proteins are functional ( S9A Fig ) . Fluorescence microscopy revealed a subcellular localization of both versions of the SclB fusion protein in nuclei of hyphae during all growth conditions tested ( vegetatively , asexually and sexually grown ) as well as in conidiospores ( Fig 11A ) and germlings ( Fig 11B ) indicating permanent nuclear localization of SclB . GFP-trap pull downs with both , the N- and C-terminally tagged SclB versions , were conducted to investigate possible interactions of SclB with other proteins . These pull downs were conducted with cultures grown vegetatively , asexually and sexually and pulled down proteins were analyzed with LC-MS/MS . The majority of identified proteins are uncharacterized ( S2 Table ) . Four importins were identified: the essential karyophorin KapF ( importin ) was identified solely in samples of vegetatively grown cultures , whereas KapJ was identified in samples from strains grown in submerged cultures , as well as in light . KapB and KapI were identified in samples grown in light or dark . Together with a predicted NES and a predicted NLS , this indicates specific control of nuclear localization for SclB . RcoA was found in samples grown in submerged cultures and in the dark , conditions inducing sexual development in the wildtype . Furthermore , it was identified in samples grown in light , but below threshold . RcoA acts as transcriptional repressor and the RcoA-SsnF co-repressor-complex , which corresponds to yeast Tup1-Ssn6 , is essential for growth in Aspergilli [98–101] . Bimolecular fluorescence complementation experiments ( Bi-FC ) were performed to verify direct interaction of SclB and RcoA in vivo . Strains were constructed for these experiments , which express fusion proteins , where one half of a split YFP ( cYFP ) was fused to SclB and the other half ( nYFP ) to RcoA [102] . Two additional strains , expressing either SclB-cYFP and free nYFP or RcoA-nYFP and free cYFP , served as controls ( S9D Fig ) . Only a signal of the joint YFP halves , indicating a physical interaction of SclB and RcoA , could be identified in nuclei of hyphae ( Fig 11C ) . This indicates that SclB can interact directly with RcoA in vivo and might execute some of its regulatory roles in developmental programs , secondary metabolism and oxidative stress response as a heterodimer .
The velvet domain protein VosA of Aspergillus nidulans binds more than a thousand fungal promoters and affects a substantial part of the transcriptome . One of these genes encodes the novel zinc cluster transcription factor SclB . VosA inhibits the expression of the sclB gene , which results in a slowdown and a decrease in asexual spore formation and a reduced production of secondary metabolites such as austinol , dehydroaustinol and emericellamides . SclB is not part of the fungal light response , which promotes the asexual program , but supports the cellular response upon H2O2 induced oxidative stress . SclB has a dual function as transcriptional activator for asexual development , but also as a repressor , presumably in combination with the repressor subunit RcoA , which we could identify as interacting partner . A genome-wide transcriptional analysis revealed that direct or indirect effects caused by the absence of the sclB gene result in more than 400 differentially expressed genes compared to wildtype ( S1 Table ) . 1 . 5 times as many of these genes are downregulated , as upregulated , in the absence of sclB . A large group of these genes are related to metabolic processes , as carbon or sulphur metabolism , or transporter activity . This most likely is a consequence of the distorted development of the ΔsclB mutant . On the other hand , several secondary metabolite and developmental genes including asexual regulatory genes as flbC or flbD , and rodA or dewA required for asexual spore formation are differentially regulated when SclB is not present in the cell . This suggests that SclB regulates asexual development and interconnected secondary metabolism in A . nidulans . SclB is localized in nuclei of germlings , conidiophores and hyphae . Four karyophorins were identified as putative interaction partners of SclB under different growth conditions and suggest a complex nuclear entry or exit control . SclB is phosphorylated at at least three residues during vegetative growth , but the function of these posttranslational modifications is yet elusive . Asexual spore formation requires the formation of the FluG protein . SclB accelerates an efficient formation of the asexual conidia in the absence of VosA by activating at least three regulatory genes downstream of FluG . Such an additional activator of conidiation had been predicted ( Fig 7B ) [11] . SclB increases flbC and flbD expression . The resulting FlbC and FlbD proteins as well as SclB activate the major asexual activator encoding gene brlA . The formation of the BrlA protein is necessary for the transition from stalk like aerial hyphae into mature conidiophores ( Fig 12 ) [83] . The molecular control mechanism by which VosA inhibits asexual differentiation is complex . VosA does not only repress the formation of the sclB gene product that acts as activator of the conidiation cascade , but also represses brlA itself during vegetative growth . De-repression only takes place , when the fungus obtains developmental competence and is triggered within a time window by the appropriate external signals for conidia formation [4 , 8] . In the further course of ongoing asexual development , the vosA gene is activated by the BrlA-downstream factors AbaA and WetA . The VosA velvet domain protein represses again the brlA and sclB genes and fulfils together with the VelB velvet domain protein its function to support spore viability [4 , 8 , 26] . SclB supports spore viability as well . One possible explanation might be that sclB gene expression is repressed by the VosA-VosA homodimer , which also represses brlA expression , whereas spore viability might be a regulatory function of the VosA-VelB heterodimer . SclB is not involved in the light control of A . nidulans , but is part of the response towards H2O2 induced oxidative stress . An internal oxidative stress signal caused by reactive oxygen species ( ROS ) serves as developmental signal in fungi and requires an appropriate fast and potent protective response [40 , 88 , 103] . ROS homeostasis therefore is crucial for the proceeding of asexual development . SclB activates elements of the fungal oxidative stress response including the thioredoxin system or catA for the spore specific catalase [89 , 90 , 104–106] . In addition , SclB activates the expression of the transcription factor RsmA during oxidative stress , which plays a similar dual role as SclB , because it is also part of the control of oxidative stress response and of secondary metabolism [91 , 94 , 107] . The SclB-mediated control for secondary metabolism includes several possible links to asexual differentiation . It is necessary for austinol , dehydroaustinol and emericellamide production and acts as activator of emericellamide , austinol and dehydroaustinol production through regulation of their gene clusters . An adduct of dehydroaustinol and diorcinol is able to overcome the conidiation defect of a ΔfluG mutant suggesting that they are involved in the FluG signal , which is crucial for the initiation of asexual development [108] . Orsellinic acid and the orsellinic acid-related diorcinol were also produced in high amounts in a ΔcsnE mutant compared to wildtype [40] . CsnE is part of the conserved COP9 signalosome ( CSN ) which controls the specificity of ubiquitin E3 cullin RING ligases for the protein degradation in the 26S proteasome [109 , 110] . CSN is required for the link between sexual development and the appropriate secondary metabolism , light control and the protection against oxidative stress [111–113] . The SclB function is involved in the alternative differentiation program . SclB connects asexual development to its specific secondary metabolism and also acts at the interphase to the response to oxidative stress . SclB interacts with RcoA in vivo . RcoA is a WD40 repeat protein , which regulates developmental programs and is required for the production of the mycotoxin sterigmatocystin as a member or the aflatoxin family [5 , 100 , 114 , 115] . A loss of rcoA in A . nidulans results in poor colony growth , impaired conidiation and the production of an orange pigment as indication of a misregulated secondary metabolism [100] . RcoA is part of the conserved SsnF-RcoA co-repressor complex corresponding to Ssn6-Tup1 in yeast , which represses numerous genes [99–101 , 116 , 117] . Target genes are repressed by several mechanisms such as through interacting with DNA-binding proteins and RNA polymerase II , through competition for promoter binding with other transcription factors , but also through histone acetylation and nucleosome positioning [118–122] . It is unclear whether there is only an RcoA-SclB heterodimer in the A . nidulans cell or whether SclB also interacts with RcoA-SsnF , because SsnF [99] could not be identified as putative SclB interaction partner . The exact molecular function of the SclB-RcoA interaction in the timely choreography of conidiation is unknown and might include as well activating as inhibiting control mechanisms during ongoing asexual development and its link to secondary metabolism and an oxidative stress response . Zinc cluster DNA-binding proteins belong to the most abundant transcription factors in the fungal kingdom [62] . SclB is present in nearly all Aspergilli and especially its C6 DNA-binding domain is highly conserved . Most C6 proteins are involved in either i ) primary or secondary metabolism or ii ) developmental programs [67] . SclB rather acts as global regulator and interconnects asexual development , secondary metabolism and the response to oxidative stress . Its C6 domain exhibits an uncommon architecture that is only found in less than 6% of all C6 proteins in A . nidulans . Other characterized A . nidulans C6 proteins with the same architecture as SclB function specifically in primary metabolic programs ( S7 Table ) [65 , 66] . Scl-2 is the SclB counterpart of A . niger . Loss of the sclB ortholog in A . niger results in reduced conidiation and impaired secondary metabolism [53] . This indicates similar regulatory effects in conidiation and secondary metabolism of A . niger Scl-2 and A . nidulans SclB . Wildtype A . niger cells form sclerotia as resting structures under very defined conditions [53 , 123] . Scl-2 also acts as a sclerotia repressor , because a corresponding scl-2 mutant strain produces sclerotia-like structures under conditions where the wildtype does not form these structures . SclB of A . nidulans is not a repressor of the formation of cleistothecia . Sclerotia have similarities with the sexual fruiting bodies of A . nidulans with the major difference that they are not linked to a sexual meiosis programme . These different control functions suggest that different fungi might have rewired the control of gene expression of this transcription factor in different developmental networks and contexts . The proposed sclB ortholog of A . fumigatus ( Afu6g11110 ) rescues the A . nidulans ΔsclB phenotype , which suggests that the molecular function of sclB therefore is conserved between A . nidulans and A . fumigatus . Some SclB functions might have changed in A . fumigatus , because it is dispensable for conidiation in this opportunistic human pathogen . Alternatively , a second redundant factor might compensate the effects of a sclB deletion , which is in agreement with other findings supporting that the conidiation cascade of A . fumigatus exhibits significant differences to its counterpart in A . nidulans . Deletion of fluG leading to diminished numbers of conidiophores in A . nidulans does not result in an obvious asexual phenotype in A . fumigatus [124 , 125] and functions of WetA , AbaA , velvet proteins or several Flb factors have changed [29 , 126] . Taken together , the VosA repression target SclB controls a novel genetic network in A . nidulans , which links conidiation to secondary metabolism and the response to oxidative stress . Further studies will broaden our understanding of the interconnection and complex mutual control of developmental programs and the production of bioactive molecules in response to environmental conditions and stresses in filamentous fungi . This is especially important , as a vast amount of bioactive natural products are still unknown and might have deleterious as well as beneficial potential to humans [38 , 127 , 128] . The SclB genetic network is a sub-network of the velvet domain network , which bridges secondary metabolism and development in fungi . In contrast , other known subnetworks of VosA , as BrlA regulating the conidiation cascade , are more specialized for a specific program . This study shows that velvet domain subnetworks include different categories as encaptic as BrlA , as well as independently acting elements as SclB . The amount of putative SclB targets and its congeneric as well as independent or even antithetic functions to VosA suggest that SclB , downstream of VosA , itself regulates a large network of downstream genes . VosA binds to more than thousand gene promoters and this network further extends through transcription factors as SclB that act themselves as master regulators .
AGB551 ( veA+ ) was used as A . nidulans wildtype . Afs35 was used as A . fumigatus wildtype . Wildtype and mutant strains ( see S3 Table ) were grown in minimal medium ( MM ) ( 1% glucose , 7 mM KCl , 2 mM MgSO4 , 70 mM NaNO3 , 11 . 2 mM KH2PO4 , 0 . 1% trace element solution pH 5 . 5 [129] ) supplemented with 0 . 1% pyridoxine-HCl , 5 mM uridine , 5 mM uracil or 4-aminobenzoic acid , when needed . Strains were grown for two days on solid MM containing 2% agar in light at 37°C and two day old spores were harvested for further experiments . For synchronized growth strains were grown in submerged cultures for 24h and subsequently shifted onto solid agar plates . Escherichia coli strains ( S4 Table ) were grown on solid lysogeny broth ( LB ) [130] medium ( 1% tryptone , 0 . 5% yeast extract , 1% NaCl ) or in liquid LB shaking on a rotary shaker at 37°C . 100 mg/ml ampicillin was added to prevent plasmid loss . For extraction of genomic DNA strains were grown over night ( o/n ) in liquid cultures . Mycelia was harvested through Miracloth filters , frozen in liquid nitrogen and ground with a table mill . Ground mycelia was mixed with 500 μl genomic DNA lysis buffer [131] and incubated 15 min at 65°C . Subsequently mycelia solution was mixed with 100 μl 8 M potassium acetate and centrifuged for 15 min at 13 , 000 rpm at room temperature ( RT ) . Supernatant was mixed with 100 μl 8 M potassium acetate and centrifuged for 15 min at 13 , 000 rpm at RT . Supernatant was mixed with 300μl isopropanol and centrifuged 10 min at 13000 rpm at RT . Pellets were washed twice with 70% ethanol and dried at 42°C before resolving in H2O at 65°C . DNA fragments for plasmid constructions were amplified with PCR from A . nidulans FGSC A4 or A . fumigatus Afs35 genomic DNA , respectively , and cloned into pBluescript SK ( + ) using the Geneart Seamless Cloning and Assembly kit , the Seamless PLUS Cloning and Assembly Kit and the Seamless Cloning and Assembly Enzyme Mix ( Invitrogen ) or via fusion PCR and subsequent cloning into pBluescript SK ( + ) with the CloneJET PCR Cloning Kit ( Thermo Scientific ) or via employment of T4 ligase ( Thermo Scientific ) according to manufacturer’s instructions . Plasmids were amplified in E . coli and extracted with the Qiaprep Spin Miniprep Kit ( Qiagen ) or the NucleoSpin Plasmid Miniprep Kit ( Macherey-Nagel ) according to manufacturer’s instructions . For the production of the plasmids pME4304 and pME4305 the pyrithiamine resistance cassette ( ptrA ) of pSK485 [72] was replaced by the nourseothricin resistance cassette ( natR ) from plasmid pNV1 [132] ( primer pair JG846/847 ) or the phleomycin resistance cassette ( phleoR ) from plasmid pME3281 [133] ( primer pair JG848/849 ) , respectively , by usage of the Seamless Cloning and Assembly Kit ( Invitrogen ) . Both cassettes additionally carried one half of the PmeI restriction site at both ends . The recyclable marker cassettes from pME4304 and pME4305 are called natRM and phleoRM , respectively , in the following . The recyclable marker cassette from pSK485 is called ptrARM in the following . For production of pME4575 , the 2 . 7 kb long 5’ and 2 . 2 kb long 3’ region of the sclB ( AN0585 ) gene were amplified with primer pairs kt208B/214 and kt211/224 , respectively , and together with the natRM cassette cloned into the EcoRV multiple cloning site of pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The deletion cassette was subsequently excised with MssI and transformed into AGB551 , resulting in the strain AGB1007 . For production of pME4578 , the 1 . 3 kb nitrate-inducible promoter ( PniaD ) , amplified with primer pair kt251/252 , the sclB open reading frame ( ORF ) itself and a small part of the 3’ region ( 1 . 8 kb ) , amplified with kt241/253 , the sclB 5’ region ( kt208b/214 ) and the natRM cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The PniaD::sclB construct was subsequently excised with MssI and transformed into AGB551 , resulting in AGB1008 . For production of pME4576 , sgfp was amplified from pME4292 with primers kt229/SR18 and , together with the sclB ORF and its 5’ flanking region ( 4 . 4 kb , primers kt208b/228 ) , the sclB 3’ region ( primers kt211/224 ) and the natRM cassette was cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . Subsequently , the sclB::sgfp construct was excised from pME4576 with MssI and transformed into AGB1007 resulting in AGB1009 . Successful transformation at the correct locus was verified by Southern hybridization . For production of pME4579 , the 5’ flanking region of sclB ( primers kt209/307 ) , sgfp ( primers SR120/121 ) , sclB ORF ( primers kt230/231 ) , the phleoRM cassette and the sclB 3’ flanking region ( primers kt211/225 ) were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . Subsequently , the sgfp::sclB construct was excised from pME4579 with MssI and transformed into AGB1007 , obtaining AGB1010 . The plasmid pME3173 was transformed into AGB1009 and AGB1010 , resulting in AGB1012 and AGB1013 , respectively , to facilitate the visualization of nuclei . pME3173 was transformed into AGB551 resulting in AGB1014 to obtain a suitable negative control for microscopy . For production of pME4577 , the sclB ORF and its 5’ UTR ( 4 . 4 kb , primers kt208b/231 ) , the sclB 3’ UTR ( primers kt211/224 ) and the phleoRM cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The sclB complementation cassette was excised from pME4577 with MssI and cloned into AGB1007 , resulting in AGB1011 . For production of pME4581 , 1 kb of the fluG 5’ flanking region ( primers kt341/342 ) , 1 kb of the 3’ flanking region ( primers kt343/364 ) and the phleoRM cassette were cloned into the EcoRV restriction site of pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The fluG deletion cassette was excised from pME4581 with MssI and integrated into AGB551 , AGB1007 and AGB1008 , resulting in AGB1016 , AGB1017 and AGB1018 , respectively . For production of pME4589 , 1 . 7 kb of the brlA 5’ region ( primers kt487/488 ) , 1 . 2 kb of the brlA 3’ region ( primers kt489/490 ) and the phleoRM cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The ΔbrlA cassette was excised from pME4589 with MssI and integrated into AGB551 and AGB1007 , resulting in AGB1031 and AGB1032 , respectively . For production of pME4591 , 1 . 2 kb of the flbB 5’ region ( primers kt515/516 ) , 1 kb of the flbB 3’ ( primers kt517/518 ) and the phleoRM cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The ΔflbB cassette was excised from the pME4591 with MssI and integrated into AGB551 , AGB1007 and AGB1008 , resulting in AGB1035 , AGB1036 and AGB1037 , respectively . For production of pME4593 , 1 . 2 kb of the flbC 5’ region ( primers kt519/520 ) , 1 kb of the flbC 3’ region ( primers kt521/522 ) and the phleoRM cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The ΔflbC cassette was excised from pME4593 with MssI and integrated into AGB551 , AGB1007 and AGB1008 , resulting in AGB1039 , AGB1040 and AGB1041 . For production of pME4595 , 1 . 1 kb of the flbD 5’ region ( primers kt523/524 ) , 1 . 2 kb of the flbD 3’ region ( primers kt525/526 ) and the phleoRM cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The ΔflbD cassette was excised from pME4595 with MssI and integrated into AGB551 , AGB1007 and AGB1008 , resulting in AGB1043 , AGB1044 and AGB1045 , respectively . For production of pME4597 , 1 . 3 kb of the flbE 5’ region ( primers kt527/528 ) , 1 . 1 kb of the respective 3’ region ( primers kt529/530 ) and the phleoRM cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The ΔflbE cassette was excised from pME4597 with MssI and integrated into AGB551 , AGB1007 and AGB1008 , resulting in AGB1047 , AGB1048 and AGB1049 . For Bi-FC plasmid construction , sclB and rcoA were amplified from cDNA instead of genomic DNA . The bidirectional nitrate-inducible promoter was excised from pME4607 in a two-step digestion with MssI and SmiI and both , the pME4607 backbone vector and the nitrate inducible promoter were utilized for all Bi-FC constructs . For production of pME4599 , the sclB ( primers kt407/415 ) and rcoA ORFs ( primers kt409/418 ) were fused to ceyfp ( primers kt416/417 ) and neyfp ( primers kt421/422 ) , respectively by fusion PCR [134] . Subsequently , sclB::ceyfp , rcoA::neyfp and the bidirectional nitrate-inducible promoter were cloned into the pME4607 backbone vector , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . pME4599 was ectopically integrated into AGB1007 resulting in AGB1051 and AGB1014 , resulting in AGB1052 . For production of pME4601 , free ceyfp ( primers kt416/SR195 ) , rcoA::neyfp and the bidirectional nitrate-inducible promoter were cloned into the pME4607 backbone vector , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . pME4601 was introduced into AGB551 and AGB1014 , resulting in AGB1054 and AGB1056 , respectively . For production of pME4600 , free neyfp ( primers kt422/SR193 ) , sclB::ceyfp and the nitrate-inducible promoter were cloned into the pME4607 backbone vector , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . pME4600 was introduced into AGB551 and AGB1014 , resulting in AGB1053 and AGB1055 , respectively . For production of pME4574 , the veA 5’ ( primers JG863/985 ) and 3’ ( primers JG865/866 ) regions and the natRM cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The ΔveA construct was excised from pME4574 with MssI and transformed into AGB551 resulting in AGB1066 . The ΔsclB cassette from pME4575 was integrated into AGB1066 , resulting in AGB1067 . For production of pME4605 , the velB 5’ ( primers SR05/06 ) and 3’ ( primers SR07/08 ) regions and the natRM cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The ΔvelB construct was excised from pME4605 with MssI and transformed into AGB551 resulting in AGB1064 . The ΔsclB cassette from pME4575 was integrated into AGB1064 , resulting in AGB1065 . For production of pME4602 , the velC 5’ ( primers kt203/145 ) and 3’ ( primers kt146/204 ) regions and the natRM cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The ΔvelC construct was excised from pME4602 with MssI and transformed into AGB551 resulting in AGB1062 . The ΔsclB cassette from pME4575 was integrated into AGB1062 , resulting in AGB1063 . For production of pME4603 , the vosA 5’ ( primers SR11/12 ) and 3’ ( primers SR13/14 ) regions and the natRM cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The ΔvosA construct was excised from pME4603 with MssI and transformed into AGB551 and AGB1007 , resulting in AGB1057 and AGB1058 , respectively . pME4578 was integrated into AGB1057 , resulting in AGB1059 . For production of pME4606 , the sclB 5’ ( primers kt215/221 ) and 3’ ( primers kt218/226 ) flanking regions and the ptrARM were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The ΔsclB cassette was excised from pME4606 with MssI and integrated into Afs35 , resulting in AfGB129 . For production of pME4580 , the sclB 5’ region and the sclB ORF , the sclB 3’region and the phleoRM marker cassette were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) : the first 1 kb part of the sclB ORF together with its 1 . 9 kb 5’ region was amplified with primers kt209/430 , introducing the first mutation in the gene product ( S327A ) . The next 431 bp of the sclB ORF were amplified with primers kt431/432 , introducing the mutation T464A in the gene product . Adjacent 135 bp were amplified with the primer pair kt433/434 and the last 172 bp of the sclB ORF were amplified with the primer pair 442/231 , introducing S504-506A in the gene product . The mutated sclB ORF and its 5’ adjacent region were fused in a series of fusion PCRs [134] from these four sequences . The complete mutated sclB ORF and its 5’ region , the sclB 3’ adjacent region ( primers kt211/225 ) and the phleoRM cassette were cloned into pBluescript SK ( + ) in a seamless cloning reaction . The sclBS327A , T464A , S504-506A cassette was excised from pME4580 with MssI and integrated into AGB1007 , resulting in AGB1015 . Similarly , the sclBS327D , T464D , S504-506D plasmid pME4610 was constructed using primers kt209/651 , kt652/653 , kt654/655 and kt657/696 . The sclBS327D , T464D , S504-506D cassette was excised from pME4610 with MssI and integrated into AGB1007 , resulting in AGB1147 . For production of pME4587 , 1 . 5 kb of the abaA 5’ region ( primers kt354/355 ) , the phleoRM cassette and 1 . 4 kb of the abaA 3’ region ( primers kt356/363 ) were cloned into pBluescript SK ( + ) , employing the Seamless Cloning and Assembly Kit ( Invitrogen ) . The ΔabaA cassette was excised from pME4587 with MssI and integrated into AGB551 and AGB1007 , resulting in AGB1028 and AGB1029 , respectively . For production of pME4609 , the sclB 5’ ( primer pair kt209/603 ) and 3’ regions were amplified from A . nidulans genomic DNA . The sclB ORF was amplified with primer pair kt254/233 from A . fumigatus genomic DNA and the three fragments were together with the natRM cassette cloned into pBluescript SK ( + ) . The construct was excised from pME4609 using MssI and transformed into AGB1007 , resulting in AGB1042 . A . nidulans was transformed by polyethylene glycol-mediated protoplast fusion as described before [135 , 136] . E . coli transformations were carried out as described in [137 , 138] . Plasmids used in this study are given in S5 Table and oligonucleotides can be found in S6 Table . Successful transformation of constructs into A . nidulans was verified by Southern hybridization [139] employing the AlkPhos Direct Labelling and Detection System according to manufacturer’s instructions ( GE Healthcare ) . Conidiospores were harvested in 0 . 96% NaCl solution containing 0 . 002% Tween 80 after 2 days and counted with a hemocytometer ( Marienfeld Superior ) . Conidiospores were diluted with 0 . 96% NaCl solution containing 0 . 002% Tween 80 , and kept at 4°C . Aliquots of 200 spores of these dilutions were plated after zero and seven days and plates were incubated for two days at 37°C in the light . This test was performed in triplicates per experimental day . For spore survival in the presence of 100 mM H2O2 , spores were diluted with 0 . 96% NaCl solution containing 0 . 002% Tween 80 in 15 ml reaction tubes and 100 mM H2O2 was added . Reaction tubes were kept in the dark at RT under constant gyration to prevent sedimentation of spores . 200 spores were plated at indicated time points and plates were incubated as mentioned above . Statistical analyses were conducted with t-tests using standard deviations of wildtype data against indicated mutant data sets . For extraction of secondary metabolites from asexually grown cultures 1*106 spores were plated and grown for 3 or 7 days in light . Spores were completely washed off and the agar was cut into small pieces . Subsequently , secondary metabolites were extracted from agar pieces with 300 ml ethyl acetate by shaking at 160 rpm at 30°C for 30 min followed by 15 min ultra-sonication at highest level . Ethyl acetate was evaporated and the crude extract was kept at -20°C . For extraction from vegetatively grown cultures , 1*107 spores were grown in submerged cultures for 48 h at 37°C on a rotary shaker and mycelia were removed with Miracloth filters . Extraction procedure was followed according to Gerke and co-workers [38] . Samples were stored at -20°C . Analytical HPLC/UV-DAD/ELSD measurements were performed using the following system: HPLC pump 420 , SA 360 autosampler , Celeno UV-DAD HPLC detector , ELSD-Sedex 85 evaporative light-scattering detector ( ERC ) ) with a Nucleodur 100–5 C18 end-capped ( ec ) column ( 250 mm x 3 mm ) and the solvent system: A = H2O + 0 . 1% ( v/v ) trifluoroacetic acid ( TFA ) , B = acetonitrile + 0 . 1% ( v/v ) TFA ( Goebel Instrumentelle Analytik GmbH ) . Secondary metabolite extracts were dissolved in 500 μl methanol and an injection volume of 20 μl was analyzed under gradient conditions ( 20% B to 100% B in 20 minutes ) with a flow rate of 0 . 5 ml/min . HPLC data was analyzed with the Geminyx III software ( Goebel Instrumentelle Analytik GmbH ) . For UHPLC-UV and UHPLC-ESI-HRMS/MS analysis crude extracts were solved in 1 ml methanol and analyzed using a Dionex Ultimate 3000 system ( Thermo Scientific ) connected to an Impact II qTof mass spectrometer ( Bruker ) . 5 μl of each sample was injected for separation on an UHPLC reversed phase column ( Acquity UPLC BEH C18 1 . 7 lmRP 2 . 1 x50 mm column ( Waters ) with an Acquity UPLC BEH C18 1 . 7 lmRP 2 . 1 x 5 mm pre-column ( Waters ) ) applying a linear acetonitril/0 . 1% formic acid in H2O/0 . 1% formic acid gradient ( from 20% to 95% acetonitril/0 . 1 formic acid in 20 min ) with a flow rate of 0 . 4 ml/min at 40°C . For internal mass calibration a 10 mM sodium formate solution was used . Data analysis and sum formula predictions were performed with Bruker Compass DataAnalysis 4 . 3 . GST tagged VosA protein was expressed and purified , as described by Ahmed and collaborators [2] . Purification was executed and monitored on an Äkta Explorer10 system ( GE Healthcare ) . Amicon Ultra Centrifugal Filter Units ( Millipore ) were used for concentration after size exclusion chromatography . EMSAs were performed as described earlier [2] . Briefly DNA probes were generated by annealing a reverse-complementary oligonucleotide pair . Protein and DNA was mixed and incubated 15 min at RT and dispersed according to molecular weight on a 6% polyacrylamide gel in 0 . 5% running buffer prior to staining with ethidium bromide . Photomicrographs were obtained with an Axiolab microscope ( Carl Zeiss Microscopy ) and a SZX12-ILLB2-200 binocular microscope ( Olympus ) . Fluorescence microscopy was performed with a Zeiss AxioObserver Z . 1 inverted confocal microscope , equipped with Plan-Neofluar 63x/0 . 75 ( air ) and Plan-Apochromat 100x/1 . 4 oil objectives ( Zeiss ) . The SlideBook 6 . 0 software ( Intelligent Imaging Innovations ) was used for picture processing . Strains were grown in 8-well borosilicate cover glass system ( Thermo Scientific ) in 400 μl MM supplemented as mentioned above , when needed , or on glass slides covered with 1 ml solid MM supplemented as mentioned above , when needed , at 37°C or 30°C . GFP-signals were normalized against wildtype background signal to subtract fungal auto fluorescence . Nuclei were visualized by ectopic integration of pgpdA::rfp::h2A into the respective strains or through staining with 0 . 1% 4’ , 6’-diamidino-2phenylindole ( DAPI ) . Conidiospore numbers were determined with a Coulter Z2 particle counter ( BECKMAN COULTER GMBH , Krefeld , Germany ) or with a Thoma cell counting chamber ( hemocytometer ) ( Marienfeld Superior ) . For quantifying cleistothecia , agar plugs of 5 mm2 were cut out from plated using the larger side of a 200 μl pipette tip and cleistothecia were individualized on a fresh agar plate and counted with help of a binocular microscope SZX12-ILLB2-200 binocular microscope ( Olympus ) . ANOVA and t-test statistical analyses were conducted using standard deviations . Mutant samples were always compared to wildtype data for two-sample comparison through t-test . For RNA isolation strains were grown vegetatively or asexually . Mycelia was harvested through sterile filters ( Miracloth ) and immediately frozen in liquid nitrogen . Frozen mycelia were ground with a table mill ( Retsch ) directly before RNA extraction . RNA was extracted with the RNeasy Plant Miniprep Kit ( Qiagen ) according to manufacturer’s instructions . cDNA was transcribed from 0 . 8 μg RNA with the QuantiTect Reverse Transcription Kit ( Qiagen ) . To measure gene expression real-Time-PCR was performed by using MESA GREEN qPCR MasterMix Plus for SYBR Assay ( Eurogentec ) in a CFX Connect Real-Time System ( BioRad ) and analysed with the CFX Manager software ( BioRad ) . Expression of the housekeeping genes gpdA ( A . nidulans and A . fumigatus ) , h2A ( A . nidulans and A . fumigatus ) and 15S rRNA ( A . nidulans ) were used for normalization . For measurement of the expression of oxidative-stress related genes , strains were grown in submerged cultures at 37°C on a rotary shaker for 24 h . Subsequently , 5 mM H2O2 was added . Control strains were left untreated . Incubation was prolonged for another 30 min shaking on the rotary shaker and mycelia were harvested as described above . Total RNA of strains grown under submerged culture conditions for 24 h at 37°C under constant agitation on a rotary shaker was isolated using the Direct-zol Miniprep Kit ( Zymo Research ) according to manufacturer’s conditions . RNA quality control was performed on a Bioanalyzer 2100 Fragment Analyzer using a Pico Chip ( RNA ) ( Agilent ) . RNA sequencing was performed at the Core Unit , the Transcriptome and Genome Analysis Laboratory , University Medical Center Göttingen . RNA integrity was assessed using the Fragment Analyzer ( Advanced Analytical ) and only samples exhibiting RNA integrity number ( RIN ) > 8 were selected for sequencing . Libraries were performed starting with 800 ng of total RNA using the TruSeq Stranded Total RNA Sample Prep Kit from Illumina ( Cat . No . RS-122-2201 ) . Library sizing ( 295–320 bp ) and quality was performed using the Fragment Analyzer ( Advanced Analytical ) . Library quantitation was performed by using Promega’s QuantiFluor dsDNA System . RNA-sequencing was performed using the Illumina HighSeq-4000 platform ( SR 50 bp; >30 Mio reads /sample ) . Demultiplexig was done using bcl2fastq2 . Raw reads were aligned using STAR version STAR_2 . 4 . 1a [140] against EnsemblFungi [141] revision 37 Aspergillus nidulans genome . Differential expression analysis was performed using edgeR [142] . Information gathered from the Aspergillus Genome Database ( AspGD ) [64] and Fungal and Oomycete Genomic Resources Database ( FungiDB ) [73] were used to categorize genes according to putative functions of their products . AspGD and FungiDB were employed for updated respective descriptions . For genetic ORFs , which were merged into a new ORF in FungiDB ( FungiDB 36; released 19 . Feb . 2018 ) , the new merged ORF was taken into consideration for all downstream analyses . Genome wide transcriptome data was submitted to EBI ArrayExpress under accession E-MTAB-6996 . Strains were grown under vegetative conditions and mycelia were harvested through sterile filter ( Miracloth ) and directly frozen in liquid nitrogen . Frozen mycelia were ground in liquid nitrogen with a table mill and approximately 200 mg was mixed with 300 μl B+ buffer ( 300 mM NaCl , 100 mM Tris pH 7 . 5 , 10% glycerol , 1 mM EDTA , 0 . 1% NP-40 ) supplemented with 1 . 5 mM DTT , complete EDTA-free protease inhibitor cocktail ( ROCHE ) , 0 . 001 mM PMSF , phosphatase inhibitor mix ( 1 mM NaF , 0 . 5 mM sodium-orthovanadate , 8 mM ß-glycerolphosphate disodium pentahydrate ) and 1 . 5 mM benzamidine , and centrifuged for 15 min at 13000 rpm at 4°C . Supernatant was transferred into fresh test tubes and protein concentration was measured with a NanoDrop ND-1000 spectrophotometer . Protein pulldowns employing GFP-trap_A beads ( Chromotek ) were conducted as described earlier [98 , 143] with some alterations . A . nidulans strains were inoculated in a concentration of 5*108 spores in 500 ml MM . Mycelia were harvested and immediately frozen in liquid nitrogen . Frozen mycelia were ground with a table mill in liquid nitrogen . Ground mycelia were mixed with B+ buffer in a ratio of 1:1 and centrifuged twice for 20 min at 4000 rpm at 4°C . Supernatant was filtered through 20 μm sterile filters ( Sartorius ) and mixed with 1:100 GFP-trap_A beads ( Chromotek ) and incubated o/n at 4°C . Equal amounts of protein were loaded on 10% SDS gels ( separation gel: 2 . 8 ml H2O , 3 . 75 ml 1 M Tris pH 8 . 8 , 100 μl 10% ( w/v ) SDS , 3 . 3 ml 30% ( v/v ) acrylamide , 10 μl TEMED , 50 μl 10% ( w/v ) APS; stacking gel: 3 . 67 ml H2O , 625 μl 1 M Tris pH 6 . 8 , 30 μl 10% ( w/v ) SDS , 650 μl 30% ( v/v ) acrylamide , 5 μl TEMED , 25 μl 10% ( w/v ) APS ) and separated at 200V . Proteins from SDS gels were blotted for 1h at 100 V ice cooled or at 35 V o/n at RT to nitrocellulose membranes ( Whatman ) . Membranes were blotted with 5% skim milk powder dissolved in TBST buffer ( 10 mM Tris-HCl pH8 . 0 , 150 mM NaCl , 0 . 05% Tween 20 ) for 1 h at RT and subsequently probed with 1:250 diluted GFP antibody ( sc-9996 , Santa Cruz Biotechnology ) . Following , membranes were washed three times in TBST and horseradish peroxidase coupled mouse antibody ( 115-035-003 , Jackson Immuno Research ) was applied as secondary antibody in a dilution of 1:2000 . Crude cell extracts were prepared as described above . B+ buffer was not supplemented with phosphatase inhibitor mix for this experiment . Crude cell extract were mixed with or without lambda phosphatase ( NEB ) according to manufacturer’s conditions and with or without phosphatase inhibitor mix in excess , and incubated for 30 min . at 30°C prior to boiling for 10 min . at 95°C together with loading dye . Subsequently , western hybridization experiments were performed as described above . Trypsin digestion of proteins was performed as published by Shevchenko and collaborators using Sequencing Grade Modified Trypsin ( Promega ) [144] . Following this procedure peptides were purified using the StageTip purification method [145 , 146] . Purified peptides were separated by reversed-phase liquid chromatography employing an RSLCnano Ultimate 3000 system ( Thermo Scientific ) followed by mass analysis with an Orbitrap Velos ProHybrid mass spectrometer ( Thermo Scientific ) as described [98 , 143 , 147 , 148] . For further details see [149] . MS/MS2 data processing for peptide analysis and protein identification was performed either with the MaxQuant 1 . 5 . 1 . 0 and Perseus 1 . 5 . 3 or the Proteome Discoverer 1 . 4 software ( Thermo Scientific ) and the Mascot and SequestHT search algorithms . Phosphosite probabilities were calculated with the phosphoRS search algorithm [150 , 151] . Three unique peptides [152] and three MS/MS counts were demanded for positive protein identification . Furthermore , only proteins identified from at least two out of three biological repetitions were considered further . Proteins also identified from the control strain ( AGB596 ) were regarded as false-positives and excluded from further consideration . | Velvet domain proteins of filamentous fungi are structurally similar to Rel-homology domains of mammalian NF-κB proteins . Velvet and NF-κB proteins control regulatory circuits of downstream transcriptional networks for cellular differentiation , survival and stress responses . Velvet proteins interconnect developmental programs with secondary metabolism in fungi . The velvet protein VosA binds to more than ten percent of the Aspergillus nidulans promoters and is important for the spatial and temporal control of asexual spore formation from conidiophores . A novel VosA-dependent genetic network has been identified and is controlled by the zinc cluster protein SclB . Although zinc cluster proteins constitute one of the most abundant classes of transcription factors in fungi , only a small amount is characterized . SclB is a repression target of VosA and both transcription factors are part of a mutual control in the timely adjusted choreography of asexual sporulation in A . nidulans . SclB acts at the interphase of asexual development and secondary metabolism and interconnects both programs with an adequate oxidative stress response . This study underlines the complexity of different hierarchical levels of the fungal velvet protein transcriptional network for developmental programs and interconnected secondary metabolism . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Material",
"and",
"methods"
] | [
"fungal",
"spores",
"secondary",
"metabolism",
"aspergillus",
"fungal",
"genetics",
"gene",
"regulation",
"oxidative",
"stress",
"regulatory",
"proteins",
"dna-binding",
"proteins",
"aspergillus",
"nidulans",
"fungi",
"experimental",
"organism",
"systems",
"transcription",
... | 2018 | Velvet domain protein VosA represses the zinc cluster transcription factor SclB regulatory network for Aspergillus nidulans asexual development, oxidative stress response and secondary metabolism |
Chronic chagasic cardiomyopathy ( CCC ) develops years after acute infection by Trypanosoma cruzi and does not improve after trypanocidal therapy , despite reduction of parasite burden . During disease , the heart undergoes oxidative stress , a potential causative factor for arrhythmias and contractile dysfunction . Here we tested whether antioxidants/ cardioprotective drugs could improve cardiac function in established Chagas heart disease . We chose a model that resembles B1-B2 stage of human CCC , treated mice with resveratrol and performed electrocardiography and echocardiography studies . Resveratrol reduced the prolonged PR and QTc intervals , increased heart rates and reversed sinus arrhythmia , atrial and atrioventricular conduction disorders; restored a normal left ventricular ejection fraction , improved stroke volume and cardiac output . Resveratrol activated the AMPK-pathway and reduced both ROS production and heart parasite burden , without interfering with vascularization or myocarditis intensity . Resveratrol was even capable of improving heart function of infected mice when treatment was started late after infection , while trypanocidal drug benznidazole failed . We attempted to mimic resveratrol’s actions using metformin ( AMPK-activator ) or tempol ( SOD-mimetic ) . Metformin and tempol mimicked the beneficial effects of resveratrol on heart function and decreased lipid peroxidation , but did not alter parasite burden . These results indicate that AMPK activation and ROS neutralization are key strategies to induce tolerance to Chagas heart disease . Despite all tissue damage observed in established Chagas heart disease , we found that a physiological dysfunction can still be reversed by treatment with resveratrol , metformin and tempol , resulting in improved heart function and representing a starting point to develop innovative therapies in CCC .
Protection against functional damage ( i . e . , disease tolerance ) represents a strategy that allows hosts to survive infection at minimum cost [1] . This strategy is successful against diseases caused by pathogens the immune system cannot eliminate and explains why individuals with similar pathogen burdens can present varying disease gravities [2] . The mechanisms underlying disease tolerance include activation of molecular pathways that are involved in tissue repair , fuel-sensing/energy production , and antioxidant defenses [1] . Activation of SIRT1 and AMPK fuel-sensing pathways contributes to cardioprotection [3] . Drugs that activate these pathways , such as resveratrol and metformin , can extend the healthspan of most organisms . Resveratrol , a polyphenol present in grapes , decreases ischemia-reperfusion injury of the heart , prevents atherosclerosis and interferes with Ca2+ handling by cardiomyocytes [4 , 5] , restoring heart contractility in cardiomyopathies and acting as an anti-arrhythmic drug [6–9] . Some of its cardioprotective effects have been attributed to its antioxidant properties , exerted by activation of the antioxidant-defense gene Nrf2 , increased expression of mitochondrial SOD2 , inhibition of NOX2 and NOX4 expression , and direct ROS scavenging [4 , 10] . The protozoan Trypanosoma cruzi infects many tissues , including the heart , and causes Chagas disease in humans . Shortly after infection , the parasite proliferates , causing acute systemic inflammation . After this initial stage , the host’s adaptive immune system controls parasite burden and patients progress to a non-symptomatic stage , but the parasite is never eliminated . Most patients remain non-symptomatic , but 30% develop heart dysfunctional disease , which affect millions of people worldwide [11 , 12] . Much as observed in humans , the hearts of rats and mice undergo severe oxidative stress in the course of experimental Chagas disease . Both NOX2 [13] and mitochondrial derived-ROS [14] contribute to produce oxidative damage in Chagas heart disease . Neutralization of oxidative stress since day 0 of infection prevents chronic contractile dysfunction [15] , but once established , no treatment has been shown to reverse heart dysfunction in Chagas disease . Several attempts to reverse the loss of heart contractility and arrhythmias in CCC targeted either parasite burden or immune response , obtaining little success [15–20] . Recently , the BENEFIT clinical trial on the effects of the trypanocidal drug benznidazole on CCC failed to prevent cardiac deterioration , despite greatly reducing parasite load [21] . Treatment to CCC remains restricted to those transposed from other cardiomyopathies , doing little to decrease morbidity . The current paradigm dictates that structural damage to heart at the chronic stage impairs its function [22] and most attempts to reverse disease aim at re-building the heart with stem cells [23] . We have previously shown that oxidative stress fuels T . cruzi acute infection [24 , 25] . Although a number of cardiomyopathies are known to be dependent on oxidative stress [26] , the case for Chagas heart disease remains obscure . Here we used the antioxidant and cardioprotective agent resveratrol to test a novel , host-targeted strategy against established Chagas heart disease: promoting disease tolerance by tuning heart physiology to a healthy pattern . As our data reveal , treatment with resveratrol greatly improves electrical and contractile activities in Chagas heart disease , while activating AMPK pathway , reducing parasite burden and oxidative stress . The general picture that emerges from this study is that despite all tissue damage , there is a relevant physiological dysfunction in established Chagas heart disease which can be reversed to improve cardiac function , opening a new perspective for Chagas disease therapy .
Murine models of Chagas disease do not usually present an indeterminate , silent stage , as happens in humans , and rather undergo a smooth transition from acute to chronic stage . We have chosen a model of infection with the Colombian strain of T . cruzi , which has been isolated from a patient in Colombia [27] and has been used in many functional heart studies in the last 25 years [16 , 28 , 29] . At 60 dpi , BALB/c infected with Colombian T . cruzi strain often present sinoatrial block , atrial abnormalities ( atrial enlargement/ interatrial block ) , second-degree atrioventricular block , sinus arrhythmia , bradycardia , prolonged QTc interval , right ventricle dilation and moderately decreased left ventricular ejection fraction ( LVEF ) , and mostly resembles the B1-B2 ( NYHA class II ) of human CCC [22 , 30] . To test whether resveratrol is effective against established Chagas disease , we infected the highly susceptible BALB/c mice with the type I Colombian strain of T . cruzi and performed individual electrocardiography ( ECG ) and echocardiography studies before starting the treatment with resveratrol at 15 mg/Kg i . p . ( at 60 days post-infection ) and after treatment cycle was complete , at 90 days post-infection ( dpi ) . Intraperitoneal route was chosen in order to increase the bioavailability of resveratrol [31] . In side studies , we tested 5 mg/Kg without success , while treatment with 10 mg/Kg presented promising results . At 90 dpi , infected mice treated with resveratrol ( RSV group ) had faster heart rate and shorter P wave duration , PR , and QT intervals when compared to infected mice treated with vehicle ( VEH group , Fig 1A , right graphs in Fig 1B , 1C , 1D and 1F ) . Non-infected controls ( NI ) are provided for comparison . Treatment with resveratrol shortened P wave duration , PR and QTc intervals , and increased individual heart rates during the interval from 60–90 dpi ( left graphs in Fig 1B , 1C , 1D and 1F ) . On the other hand , vehicle did not significantly change P wave duration ( left graphs in Fig 1C ) and did not stop the progression towards even longer PR and QTc intervals and increasing bradycardia ( left graphs in Fig 1D , 1F and 1B ) . ECG intervals did not differ between infected non-treated and infected vehicle-treated mice ( S1A Fig ) . Similar beneficial results were obtained with peroral administration of a higher dose of resveratrol ( S2 Fig , 40 mg/Kg ) . Fig 1G displays representative ECG tracings for a non-infected control and 3 infected individuals before and after treatment with vehicle ( #1–3 ) or resveratrol ( #4–6 ) . In the VEH group , incidence of sinus arrhythmia remained unchanged after treatment ( Fig 1H ) . In contrast , treatment with resveratrol led to a 35% decrease in the percentage of sinus arrhythmia-affected mice ( 16 SA-free animals out of 45 previously affected ) . Treatment with resveratrol also produced a 49% decrease in the percentage of infected mice affected by atrial and atrioventricular conduction disorders ( sinoatrial block , intra-atrial/ interatrial block , or second-degree atrioventricular block ) : 23 free animals out of 47 previously affected , while infected vehicle-treated mice had persistent abnormalities ( Fig 1H ) . At 90 dpi ( post-treatment ) , the incidence of conduction disorders was far greater among VEH than among RSV animals . Among RSV , 15/47 ( 31% ) mice presented normal ECG profiles after treatment , while among VEH 48/48 ( 100% , p<0 . 0001 ) presented sinus arrhythmia , atrial and/ or atrioventricular conduction disorders . Resveratrol instantly reversed ischemia-reperfusion arrhythmias ex vivo [6] . However , in our model , short-term treatment ( 20 h ) did not present benefits on heart electrical function ( S3A Fig ) . Treatment prorogation until 120 dpi maintained the benefits of resveratrol over vehicle on the heart electrical cycle ( S3B Fig ) . We used echocardiography to assess heart pumping efficiency . In previous studies , B-mode images were found to be more adequate to assess the geometry of chagasic hearts [32] . Before treatment ( 60 dpi ) , infected mice ( INF ) presented decreased left ventricle ejection fraction ( LVEF ) , stroke volume , and cardiac output ( Fig 2A ) , as well as dilated RV when compared with NI ( S4A Fig ) . After treatment ( 90 dpi ) , VEH presented decreased LVEF on average , while RSV presented normal values ( Fig 2A ) . Stroke volume and cardiac output were significantly improved in RSV compared to VEH animals . Heart geometry changed in response to treatment . At 90 dpi , RV area was increased in VEH compared to NI mice , but treatment with resveratrol reversed RV dilation ( S4A Fig ) . A transient decrease in LV area was found in all infected groups compared with NI mice at 90 dpi , similar to other models [33 , 34] , an effect no longer present at 120 dpi ( S4 Fig ) . VEH did not significantly differ from non-treated infected mice concerning heart geometry ( S1C Fig ) . Even when mice were treated late after infection ( 120 dpi ) with resveratrol for 40 days , they reacted positively to treatment , regaining a normal LVEF ( Fig 2B ) and presenting reduced right ventricle dilation at 170 dpi ( S4B Fig ) . On the other hand , treatment with the trypanocidal drug benznidazole for 40 days starting at 120 dpi was not capable of improving LVEF ( Fig 2B ) or reducing right ventricle dilation ( S4B Fig ) . Although treatment with resveratrol starting at 120 dpi failed to reduce P wave duration and QRS interval ( which was prolonged at 170 dpi ) , it also promoted normal heart rates , decreased PR and QTc intervals ( S5 Fig ) . Together , these results indicate that even when started late , treatment with resveratrol can still be beneficial to heart function . Altogether , these results indicate that treatment with resveratrol has a profound beneficial effect on the cardiac function of mice with established Chagas disease , being able to partially reverse both contractile and electrical dysfunctions . An inflammatory response characterized by leukocyte infiltration and tissue remodeling constitutes an important feature of CCC . Resveratrol did not significantly alter the number of invading inflammatory cells infiltrating the heart ( S6A Fig ) , heart vascularization ( S6B Fig ) , or collagen content ( measured by either 2nd harmonic or tricolor Masson , S6C and S6D Fig ) . A slight trend towards decreased number of infiltrating cells was found in RSV mice ( S6A Fig , P = 0 . 30 , 20 heart sections analyzed from each of 10–13 mice per group ) . Because resveratrol activates AMPK activation to restore cardiac function in other models [35 , 36] , we assessed AMPK phosphorylation ( Thr 172 ) in the ventricles of VEH versus RSV animals at 90 dpi . Treatment with resveratrol promoted a significant increase in phosphorylated AMPK ( normalized to the total protein levels ) compared to the VEH and NI groups ( Fig 3A ) , resulting in an increased p-AMPK/AMPK relation . SIRT1 relative expression ( Fig 3B ) and de-acetylase activity ( Fig 3C ) were not significantly different among groups . These results suggest a SIRT1-independent effect of resveratrol , and in fact treatment of infected mice with resveratrol plus EX527 , an specific SIRT1 inhibitor used as previously described [37] , was not able to reverse the significant benefits of resveratrol on ECG abnormalities ( S7 Fig ) . Together , these data support the notion that AMPK , but not SIRT1 , is activated in response to resveratrol in CCC . T . cruzi infection induces ROS production by cardiomyocytes [14] , a phenomenon that may cause cardiomyocyte death , or may act physiologically to produce electrical and pumping dysfunction . We assessed extravascular ROS production using in vivo CD105 ( endothelial marker ) /DCDFA labeling at 90 dpi . Extravascular ROS was increased in VEH group when compared to NI , while RSV had ROS levels below those found in NI mice ( Fig 3D ) . In vivo MitoSOX staining was significantly greater in VEH compared to NI mice ( Fig 3E ) . We observed a greater staining in VEH compared to RSV hearts ( P = 0 . 06 ) . RSV mice had increased heart expression of the mitochondrial enzyme SOD2 , an AMPK-controlled enzyme [38] , compared with VEH mice ( Fig 3F ) . Mitochondrial oxidative stress has been previously suggested as a possible causative factor for chagasic heart dysfunction [39 , 40] Treatment with resveratrol , in comparison with vehicle , greatly reduced lipid peroxidation in plasma , as assessed by thiobarbituric acid reactive substances ( TBARs ) ( Fig 3G ) . A sharp decrease in heart tissue parasitism was detected by quantitative PCR in RSV when compared with VEH mice at 90 dpi ( Fig 3H ) . We assessed the expression of some proteins controlled by AMPK/SIRT1 . Contrary to our expectations , there was an increase in p-ACC and GLUT4 in infected mice and resveratrol promoted a decrease towards non-infected levels ( S8 Fig ) , an effect probably due to resveratrol’s remarkable effect of reducing parasite burden . We did not find any significant differences in PGC1α expression or cardiac ATP levels . These results show that resveratrol activates the AMPK pathway and decreases parasite burden , together with reducing ROS production and lipid peroxidation . Because resveratrol activated AMPK phosphorylation and reduced ROS in our Chagas heart disease model , we tested whether activating AMPK or reducing ROS could mimic the beneficial effects of resveratrol on heart function . Metformin ( Met ) is an AMPK activator and a cardioprotective drug [41 , 42] and has indirect antioxidant activity , increasing the expression of antioxidant enzymes such as SOD2 [38] . Tempol ( Tmp ) is a SOD-mimetic drug that efficiently neutralizes ROS [43] . These drugs were administered daily by gavage . We also performed the usual i . p . treatment with RSV and respective VEH to allow a comparison of heart effects between RSV , Met , and Tmp . Mice treated with Met or Tmp had decreased PR and QTc intervals and increased heart rates compared to peroral VEH ( Fig 4A ) . These results were similar to that obtained by treatment with RSV . Pre- and post-treatment profiles of individual mice are illustrated in Fig 4B ( #1–9 ) . Met decreased the percentage of arrhythmias and conduction disorders among infected mice: while 7/7 ( 100% ) animals treated with vehicle presented sinus arrhythmia , sinoatrial block , intra-atrial/ interatrial block , and/ or second-degree atrioventricular block at 90 dpi , only 4/9 ( 44% , VEH x Met P = 0 . 03 ) among those treated with Met were positive . Treatment with either Met or Tmp also restored a normal LVEF ( Fig 4C ) and Tmp significantly increased stroke volume and cardiac output when compared to peroral VEH . The results obtained with Tmp were similar to that obtained by treatment with RSV . Tmp also significantly reduced right ventricle dilation ( S9 Fig ) . Importantly , Met and Tmp reduced lipid peroxidation of heart samples ( Fig 4D ) , but did not alter heart parasite burden ( Fig 4E ) . Though mimicking the results obtained with RSV using Met or Tmp does not actually demonstrate that RSV acted in the same way the other drugs did , it indicates that AMPK activation and SOD mimetic activity can be exploited as therapeutic strategies in Chagas heart disease . Taken together , these results suggest that reducing ROS is sufficient to improve heart function in CCC , while decreasing parasite burden is not required to improve heart function .
The several attempts to treat CCC with trypanocidal drugs have produced inconsistent results , despite reductions in parasite load . Infected mice have been studied pre- and post-treatment with benznidazole [18]: treatment eliminated the parasite and prevented to a small extent the prolongation of the PR interval over time , but by the end of the study , benznidazole-treated mice offered no improvements over controls . A study of infected rats treated with benznidazole found no improvement of heart function analyzed by catheterism [15] . In human CCC , an attempt to reduce parasite load with benznidazole cured heart disease in some cases [19] , but not in another study [20] . The recent results of the BENEFIT trial of benznidazole at the chronic phase show that despite greatly reducing parasite burden , it does not affect cardiac deterioration [21] . These results reinforce the notion that heart pathology and parasite burden have a loose association at the chronic stage and discourage the trypanocidal strategy against disease . Here , we show that resveratrol improves heart function ( a diagrammatic illustration of its heart function effects in CCC is shown in Fig 5 ) and reduces heart parasite burden . Nevertheless , we believe that instead of acting primarily by reducing parasite burden to improve heart function , resveratrol acted as an antioxidant . Different from resveratrol , trypanocidal drug benznidazole failed to restore heart pumping function . Moreover , both metformin ( AMPK activator ) and tempol ( SOD-mimetic ) improved heart function and decreased lipid peroxidation , but did not change parasite burden . These findings support raising disease tolerance as an effective strategy against CCC , as long as the parasite burden is kept low in order to avoid a rebound . We have previously shown that oxidative stress fuels acute T . cruzi infection in mice [24] , promoting parasitism in heart and macrophages . In that previous study , resveratrol was able to reduce acute parasitemia and macrophage parasite burden . Here we show that at the chronic phase of infection , resveratrol , AMPK-activator metformin and SOD-mimetic tempol reduce lipid peroxidation , a measure of oxidative stress , but only resveratrol reduces heart parasite burden . We believe that at this stage of infection , oxidative stress caused by respiratory burst in macrophages is no longer a significant factor promoting T . cruzi infection in mice . In fact , resveratrol was recently demonstrated to have a direct trypanocidal effect [44] , and therefore any of its host-dependent antioxidant effects can be dispensed with to explain the reduction of heart parasite burden it promoted . Exercise and a healthy diet remain the cornerstones of prevention and treatment of heart disease . Some drugs mimic the effects of diet and exercise by triggering signals that resemble those seen with reduced ATP levels and activating the phosphorylation of AMPK . By doing so , these drugs protect the heart [3] . Here , we show that treatment with one of these drugs , resveratrol , also promoted phosphorylation of AMPK in the hearts of infected mice , suggesting that AMPK activation is likely to be one of the cardioprotective mechanisms in this case [3] . Consistently , metformin , an indirect AMPK activator [45] , mimicked most of the effects of resveratrol on heart electrical function and its effects on ejection fraction . SOD2 , an AMPK-controlled mitochondrial antioxidant enzyme [38] , was found to be significantly increased by treatment with resveratrol and general SOD-mimetic tempol was able to mimic resveratrol’s effects on heart function . These results indicate that the AMPK-pathway and its effects on oxidative stress are a likely target to resveratrol , though not the only one . Future studies are required to define the exact mechanism by which resveratrol promotes its beneficial effects in Chaga’s Disease . Resveratrol is a multitarget drug that besides activating AMPK pathway [4] , interacts directly with 20 proteins [46]: it activates Nrf2 gene [47] , has a direct antioxidant effect and is a PDE4 inhibitor [48] , and it is likely that its heart effects in chronic Chagas disease depend on several of these mechanisms . Our results showed that in infected , vehicle treated mice , there was an increase in phosphorylation of ACC that did not depend on AMPK activation . Although in normal heart resveratrol promotes an increase in phosphorylation of ACC [36] , in this case , it worked to prevent the increase in p-ACC promoted by infection . A similar situation seemed to happen to GLUT4 , a gene controlled by AMPK activation and similarly increased in infected vehicle and resveratrol-treated mice . We do not know the reasons for these surprising findings , but we believe infection activates ACC phosphorylation through a pathway other than AMPK while resveratrol reduces the activity of this pathway by reducing the parasite burden . We also did not expect the effects of resveratrol to be independent of SIRT1 activation , but AMPK is known to be activated by high doses of resveratrol independently of SIRT1 [49] . The association between ROS production and heart disease is well known . Here we show that antioxidants resveratrol and tempol reduced oxidative stress and improved heart function in established Chagas heart disease . Neither AMPK activators nor antioxidants have previously been tried as a therapy for established Chagas heart disease . In a previous study , antioxidant phenyl-tert-butyl-nitrone was administered before infection ( starting at day 0 ) until the chronic phase , preventing the establishment of functional Chagas heart disease in rats [15] . Although that study bears similarities to ours , indicating a role for ROS in heart functional damage , no treatment has been shown to reverse established Chagas heart disease until now . Because trypanocidal therapy is only effective during acute stage but diagnosis usually occurs during the chronic disease stage , strategies to reverse established heart disease are welcome , while new strategies to prevent its progression have few applications . Our data indicate that reduction of oxidative stress likely represents a viable therapeutic strategy in CCC , for which there are currently no effective therapies . The mechanisms by which ROS causes cardiac dysfunction in Chagas heart disease are now unclear . The improvement in cardiac function observed in chronically infected mice after treatment with resveratrol challenges the current paradigm about Chagas disease pathogenesis . It is currently believed that the cumulative tissue damage caused by infection and inflammation breaks the functional structure of the heart , and can only be reversed by replacing cardiac tissue with stem cells . Previous reports show that some combinations of T . cruzi strains and mice fail to alter ECG recordings , while presenting as much inflammatory infiltrates/ tissue disorganization as combinations that do alter ECG recordings . These findings indicate that functional disease does not easily correlate with tissue damage [50] . We showed here that reversal of heart dysfunction was not associated with decreased inflammatory infiltrates , while others found that antioxidant-induced prevention of heart dysfunction did not alter inflammatory infiltrates [15] . We hypothesize that there are two overlapping heart dysfunctions in CCC: one that is merely physiological and easily reversible and another that is structural , involves inflammatory infiltrates , cell death and extensive tissue remodeling . Oxidative stress probably underlies both dysfunctions . Based on our data , we speculate that heart cells react to chronic infection/ inflammation with changes in physiology that lead to high ROS production . The heart disease that follows is somewhat similar to that found in other ROS-related cardiomyopathies . Our results show that despite all the tissue damage found in established Chagas heart disease , the physiological impairment affecting the heart is still reversible and the heart function can be significantly improved , offering a new therapeutic opportunity to millions of patients suffering from Chronic chagasic cardiomyopathy worldwide . Our study has some limitations and in order to overcome them and translate our study into a clinical trial , we still plan to perform a full study of cardiac function after resveratrol treatment , approaching the following questions: ( 1 ) oral treatment with extended bioavailability , using co-administration of glucuronidation inhibitor piperine , [51]; ( 2 ) treatment interruption and analysis after various intervals; ( 3 ) extended treatment ( until 200 dpi ) ; ( 4 ) treatment of chronic chagasic mice infected by other T . cruzi strains . Our results show resveratrol reverses important aspects of this heart dysfunction: shortens QTc , an independent risk factor for sudden cardiac arrest; reverses atrial and atrioventricular conduction disorders , risk factors for cardiovascular mortality; and restores ejection fraction , a major contributor to morbidity [22 , 30] . In addition , resveratrol greatly decreases heart parasite burden , reducing concerns of infection rebound . As resveratrol is considered a food supplement to most health agencies , such as US Food and Drug Administration , we believe that resveratrol is a suitable candidate to human trials in a very near future , targeting B1-B2 clinical stage of CCC .
This study was carried out in strict accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals of the Brazilian National Council of Animal Experimentation ( http://www . cobea . org . br/ ) and Federal Law 11 . 794 ( October 8 , 2008 ) . The institutional Committee for Animal Ethics of UFRJ ( CEUA , Licenses IMPPG029 e IMPPG032 . ) and Fiocruz ( Licenses 004/09 and LW10-14 ) approved all the procedures used in this study . Female or male BALB/c mice ( 5–7 weeks of age ) obtained from the animal facilities ( CECAL ) of the Oswaldo Cruz Foundation ( Fiocruz , Rio de Janeiro , Brazil ) and Universidade de São Paulo , Brazil were kept in a sterile environment under standard conditions ( temperature and relative humidity of approximately 22 ± 2°C and 55 ± 10% , respectively ) and received food and water ad libitum . Mice were individually identified by ear tags . Mice were infected intraperitoneally ( i . p . ) with 102 blood trypomastigote forms of the type I Colombian strain of T . cruzi . Treatments were performed daily for 30 days from the establishment of CCC ( 60 dpi ) by i . p . injection of 15 mg . kg-1 trans-resveratrol ( Sigma , 10% ethanol/PBS ) , vehicle ( 10% ethanol/PBS ) , 5 mg . Kg-1 EX527 ( 0 . 1% DMSO , Sigma ) , or peroral administration of 40 mg . Kg-1 resveratrol ( 10%ethanol-PBS ) , 500 mg . kg-1 metformin ( Merck , dissolved in water ) , 100 mg . kg-1 tempol ( Sigma , dissolved in water ) , benznidazole ( Rochagan , 25 mg/Kg , dissolved in water ) and vehicle ( water or 10%ethanol-PBS ) . Mice were sedated with diazepam ( 10 mg/kg ) and transducers were placed subcutaneously ( DII derivation ) . The traces were recorded for 2 minutes using the digital Power Lab 2/20 or Power Lab 4/35 Systems connected to a bio-amplifier ( PanLab Instruments , Spain ) . The filters were standardized to 0 . 1-100Hz and the traces were analyzed with Scope for Windows ( V3 . 6 . 10 , PanLab instruments ) . Further details are provided in Online Methods . The assessment of P wave duration and incidence of conduction disorders required large numbers of mice to provide a valid statistical analysis . Echo was performed under deep isoflurane anesthesia ( 2% in oxygen ) . Mice were trichotomized in the precordial region using depilatory cream and examined under a 30 Mhz transducer with a Vevo 770 Ultrasound apparatus ( Visual Sonics , Canada ) . The left ventricle ejection fraction ( LVEF ) was calculated using Simpson’s method , chosen because of its fit with CD heart geometry and because it is commonly used to assess CD patients . The area of the left and right ventricles during diastoles and systoles were obtained in B mode using a short axis view at the level of the papillary muscles . Most of the comparisons between means ± SEMs were made using unpaired Student’s t tests ( two groups ) or one-way ANOVA with Newman-Keuls post-test ( multiple groups ) , except for pre versus post analyses , in which we used paired Student’s t tests . The comparison between the incidences of arrhythmia across groups was calculated using Fisher’s exact t test . Differences with a p-value <0 . 05 were considered significant and significant p-values are shown in the figures next to the compared groups . | Protection against functional damage , i . e . disease tolerance , is a successful strategy against pathogens the immune system fails to eliminate . In Chagas disease , Trypanosoma cruzi infects the heart and many years after parasite burden is reduced to a minimum by the immune response , a cardiomyopathy ensues , often accompanied by electrical abnormalities and occasionally progressing to heart failure . Most attempted therapies targeted parasite elimination , failing to protect against progression of heart disease . Here we attempted to treat Chagas heart disease in mice with cardioprotectors/ antioxidants , since the parasite is known to promote tissue damage by oxidative stress . Our results show this strategy can partially reverse functional Chagas heart disease even when started late after disease onset , opening a new perspective for Chagas disease therapy . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"cardiovascular",
"anatomy",
"oxidative",
"stress",
"cardiac",
"ventricles",
"tropical",
"diseases",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"cardiovascular",
"medicine",
"organisms",
"electrocardiography",
"protozoans",
... | 2016 | Resveratrol Reverses Functional Chagas Heart Disease in Mice |
The essential coenzyme nicotinamide adenine dinucleotide ( NAD+ ) plays important roles in metabolic reactions and cell regulation in all organisms . Bacteria , fungi , plants , and animals use different pathways to synthesize NAD+ . Our molecular and genetic data demonstrate that in the unicellular green alga Chlamydomonas NAD+ is synthesized from aspartate ( de novo synthesis ) , as in plants , or nicotinamide , as in mammals ( salvage synthesis ) . The de novo pathway requires five different enzymes: L-aspartate oxidase ( ASO ) , quinolinate synthetase ( QS ) , quinolate phosphoribosyltransferase ( QPT ) , nicotinate/nicotinamide mononucleotide adenylyltransferase ( NMNAT ) , and NAD+ synthetase ( NS ) . Sequence similarity searches , gene isolation and sequencing of mutant loci indicate that mutations in each enzyme result in a nicotinamide-requiring mutant phenotype in the previously isolated nic mutants . We rescued the mutant phenotype by the introduction of BAC DNA ( nic2-1 and nic13-1 ) or plasmids with cloned genes ( nic1-1 and nic15-1 ) into the mutants . NMNAT , which is also in the de novo pathway , and nicotinamide phosphoribosyltransferase ( NAMPT ) constitute the nicotinamide-dependent salvage pathway . A mutation in NAMPT ( npt1-1 ) has no obvious growth defect and is not nicotinamide-dependent . However , double mutant strains with the npt1-1 mutation and any of the nic mutations are inviable . When the de novo pathway is inactive , the salvage pathway is essential to Chlamydomonas for the synthesis of NAD+ . A homolog of the human SIRT6-like gene , SRT2 , is upregulated in the NS mutant , which shows a longer vegetative life span than wild-type cells . Our results suggest that Chlamydomonas is an excellent model system to study NAD+ metabolism and cell longevity .
The coenzyme nicotinamide adenine dinucleotide ( NAD+ ) is an essential enzyme . Electron transfer between NAD+ and its reduced form NADH are essential to cells as they are involved in glycolysis and the citric acid cycle as well as regeneration of ATP from ADP [1] . NAD+ is consumed in several non-redox processes in cells ( see [2] for review ) . NAD+ is a substrate of ADP-ribosyl transferase , which transfers ADP-ribose from NAD+ to ADP-ribose receptors , which are involved in DNA damage responses , transcriptional regulation , chromosome separation and apoptosis . NAD+ is also the target of ADP-ribosyl cyclases , which produce cyclic ADP-ribose that acts in second messenger signaling pathways . NAD+ is a substrate of sirtuins ( SIRT/Sir2 , Silent Information Regulator Two ) , a group of NAD+-dependent deacetylases that remove acetyl groups from lysine residues on histones , microtubules , and other proteins . Thus , NAD+ , via sirtuins , modulates many events . NAD+ synthesis pathways are categorized into either de novo pathways , which start with the amino acid aspartate or tryptophan , or salvage pathways , which start with nicotinamide ( NAM ) or nicotinic acid ( NA ) ( Figure 1 ) . Plants and some bacteria initiate de novo synthesis from aspartate and use two enzymes , L-aspartate oxidase ( ASO ) and quinolinate synthetase ( QS ) , to synthesize quinolinate ( QA ) . Fungi , animals and some bacteria synthesize QA from tryptophan via six enzymes , tryptophan 2 , 3-dioxygenase ( TDO ) /indoleamine 2 , 3-dioxygenase ( IDO ) , arylformamidase ( AFMID ) , kynurenine 3-monooxygenase ( KMO ) , kynureninase ( KYNU ) , and 3-hydroxy-anthranilate 3 , 4-dioxygenase ( 3HAO ) . The three enzymes shared by both de novo pathways , quinolinate phosphoribosyltransferase ( QPT ) ; nicotinate/nicotinamide mononucleotide adenylyltransferase ( NMNAT ) ; and NAD synthetase ( NS ) , are required for the conversion of QA to NAD+ . In the salvage pathways , the starting substrate is usually NA or NAM ( Figure 1 ) . Fungi , plants , and most bacteria , use NAM in a 4-step process involving nicotinamidase ( NAMase ) , nicotinate phosphoribosyltransferase ( NAPRT ) , NMNAT , and NS to synthesize NAD+ . In C . elegans , this is the only known pathway to synthesize NAD+ [3] . On the other hand , in mammals and some bacteria , NAD+ is synthesized via a 2-step enzymatic process and the enzymes involved are nicotinamide phosphoribosyltransferase ( NAMPT ) and NMNAT . The consumption of NAD+ by sirtuin mediated-protein deacetylation results in the production of nicotinamide . Recent studies have linked SIRT proteins to transcriptional gene silencing [4] , DNA break repair [5] , cell cycle regulation [6] , aging [7] , metabolism [8] and apoptosis [9] . In human , seven members of the SIRT protein family , SIRT1-7 , are separated into 5 classes , I-IV , and U [10] . Human SIRT6 , SIRT7 , and some plant SIRT proteins belong to Class IV [10] , [11] . The nuclear-localized SIRT6 is a NAD+-dependent histone deacetylase involved in telomeric chromatin modulation [12] . Deficiency of SIRT6 in mice is correlated with defective DNA repair , genomic instability , age-related degeneration [7] , as well as increased glucose uptake , which is caused by transcriptional upregulation of several glycolytic genes that are normally repressed by SIRT6 [8] . SIRT7 localizes to the nucleolus and is involved in gene regulation of rDNA [13] , [14] . Chlamydomonas reinhardtii , a unicellular green alga , is evolutionarily related to the seed plants and contains a chloroplast [15] , [16] . Additionally , it contains animal specific organelles known as cilia/flagella and centrosomes . As discussed above , NAD+ synthesis pathways are diverse , but enzymes involved at each specific step are conserved in many organisms . Sequence similarity searches indicate that enzymes involved in the aspartate pathway from Arabidopsis , rice , and E . coli are conserved with protein identities ranging from 22% to 70% [17] . With the completion of the Chlamydomonas genome project [15] , it became possible to identify Chlamydomonas homologs involved in the NAD+ synthesis pathways via sequence similarity searches . A group of NAM-requiring mutants ( nic ) was isolated by Eversole that fail to grow well on medium lacking NAM [18] . The mutations also confer sensitivity to 3-acetylpyridine ( 3-AP ) [19] . Eight NAM-requiring strains were originally isolated and six of these mutant strains are still extant . The NIC7 locus was identified in a walk through the mating-type locus and shown to encode a homolog of QS [20] , [21] . We tested whether the remaining NIC loci encode the enzymes of the de novo aspartate NAD+ synthesis pathway . The nic mutant loci define six different loci and map to six different linkage groups ( LG ) [19]: NIC1 maps to LG XV; NIC2 maps to LG II; NIC7 maps to LG VI; NIC11 maps to LG IV; NIC13 maps to LG X; and NIC15 maps to LG XII/XIII [[20] , [ 22]–[25]; see Materials and Methods for linkage group to chromosome translation] . In our study , phenotypic characterization and genetic crosses of nic11 strains obtained from the Chlamydomonas Center indicate that the Nic− phenotype of these nic11 strains ( sensitivity to 3-AP or a growth defect on medium lacking NAM ) can no longer be scored . Therefore , only five mutant strains are used in our study and we find that they encode the five enzymes in the de novo biosynthesis of NAD+ from aspartate .
Wild-type ( CC-124 ) and five different nic mutant cells ( nic1-1 , nic2-1 , nic7-1 , nic13-1 , and nic15-1 ) were tested for their ability to utilize intermediate substrates in different NAD+ biosynthesis pathways ( Figure 2 ) . Wild-type cells show no obvious growth defect on any of the media tested . All the nic mutant strains fail to grow on Sager and Granick rich medium without NAM ( R ) or R medium supplemented with 3-AP , as previously described [19] . These mutants grow well on media supplied with either NAM or NMN , two chemical substances found only in the 2-step salvage biosynthesis pathway of NAD+ . Addition of NA , an intermediate substrate found in the 4-step salvage pathway showed very weak rescue of the Nic− mutant phenotype of the mutants . Addition of 3-HA , which is synthesized in the tryptophan de novo pathway could not rescue the growth defect of any nic strains . NaAD , which acts in both de novo pathways , also does not rescue the Nic− mutant phenotype . The failure to rescue may indicate a failure of Chlamydomonas to transport these metabolites into cells . 3-AP is considered to be a NA analogue , which causes NA deficiency in mice . Injecting animals with NA , NAM , or tryptophan rescues the NA deficiency [26] , [27] . Given that 3-AP causes cell lethality in Chlamydomonas nic mutant cells , we tested whether addition of NAM or NA could rescue the phenotype . Addition of NAM showed weak rescue of the nic7-1 and nic15-1 mutants but not the other mutants ( Figure 2 ) . Addition of NA , NMN , NaAD , or 3-HA does not rescue the lethality conferred by 3-AP in any of the mutants . Katoh et al . showed that Arabidopsis synthesizes NAD+ from aspartate and all five enzymes involved in this pathway have been characterized [28] , [29] . We identified the Chlamydomonas homologs by sequence similarity and linked the genes to corresponding nic mutants via DNA sequencing and complementation with transgenes . The results are summarized in Table 1 . Chlamydomonas ASO gene , which contains 4 exons and encodes a 669 aa protein ( Figure 3A; [15] ) , is ∼63 kb away from the ODA12 gene [30] , which maps to LG XII/XIII [31] . The Chlamydomonas nic15-1 mutant is tightly linked to MAA1 on LGXII/XIII [22] . The nic15-1 strain ( See Materials and Methods ) contains a single nucleotide change C1376T that predicts a S459F change in the predicted protein ( Figure 3A ) . The S459F change falls in a highly conserved region among bacterial and plant ASO proteins and is likely to affect normal function of this protein ( Figure S1 ) . We performed gene complementation with either a BAC ( 32L22 ) or a plasmid containing the full-length ASO gene ( pNIC15a ) . Both transformations produced 3-AP-resistant colonies ( n = 14 for the BAC and n = 24 for the plasmid ) , which indicates that Nic− mutant phenotype was rescued by re-introducing the ASO gene . The nic7-1 mutant maps to a 7 . 9 kb region on LG VI . Transformation of this fragment rescues the 3-AP sensitive phenotype of nic7-1 cells , but the nature of this mutant and the gene structure of NIC7 were not determined [20] . Ferris et al . proposed that NIC7 encodes QS , given the NIC7 gene product displays low similarity to bacterial QS [21] . Using RT-PCR and DNA sequencing , we found that the NIC7 gene contains 15 exons and it shares 63% identity to Arabidopsis QS . Sequencing of the NIC7 coding region reveals a L351Q change in the nic7 -1 mutant ( Figure 3A ) . The amino acid L351 within the quinolinate synthetase domain is conserved among almost all plant QS proteins but not in bacterial proteins ( Figure S2 ) . CNA43 , a cDNA marker mapped to LG II [31] is ∼212kb from the Chlamydomonas QPT gene as determined by the JGI Chlamydomonas v4 . 0 genome assembly . The nic2-1 mutant maps to LG II ( [24]; M . Miller and S . K . Dutcher , unpublished observations ) . RT-PCR and sequencing of the QPT coding region in the nic2-1 mutant strain reveal a single nucleotide deletion at nucleotide 559 that leads to a frameshift . The amino acid sequence changes at Gly187 and generates a premature stop codon at amino acid 240 ( Figure 3A; Figure S3 ) . The mutant protein contains all the conserved QA-binding sites ( Figure S3 , blue reversed triangles ) but the α8-12 helices and β8-12 sheets required to form α/β barrel are missing [32] . Transformation with a BAC clone ( 38P5 ) containing a full-length QPT gene yields 8 independent 3-AP-resistant colonies . The Chlamydomonas NMNAT gene was predicted to contain 4 exons and encode a 234 aa protein [15] . However , RT-PCR , nested PCR and DNA sequencing shows the coding region of Chlamydomonas NMNAT is composed of only 2 exons that encodes a 524 aa protein , as predicted by the GreenGenie2 algorithm [33] . Sequence alignment reveals that Chlamydomonas NMNAT contains extra glycine/proline/glutamine-rich sequences compared to NMNAT proteins from other organisms ( Figure S4 ) . The Chlamydomonas NMNAT gene is ∼262 kb away from the IDA2 gene [34] and maps to LG XV [31] , near where the NIC1 gene maps [25] . Sequencing of the NMNAT genomic DNA from a nic1-1 strain indicates two adjacent nucleotide changes A1406C1407→T1406T1407 result in Q345H , Q346stop ( Figure 3A; Figure S4 ) . These nucleotide changes are likely to generate a truncated NMNAT in the nic1-1 mutant strain . The 3-AP-sensitive nic1-1 mutant phenotype is leaky and reverts at a low frequency of ∼1 spontaneous 3-AP-resistant colony per 108 cells . Therefore , a co-transformation approach was used for gene complementation . BAC DNA ( 10M24 ) or plasmid DNA ( pNIC1-56 ) containing the full-length NMNAT gene was co-transformed with a paromomycin-resistant gene ( APHVIII; [35] ) . A subset of the paromomycin-resistant colonies ( 5/20 for BAC and 2/2 for plasmid transformation ) showed resistance to 3-AP . The Chlamydomonas NS homolog maps between GP220 and GP441 , which are two molecular markers on LG X [31] , where nic13-1 maps [25] . RT-PCR and DNA sequencing from wild-type cells indicate this gene contains 20 exons that encode an 832 aa protein . The intron between exons 2 and 3 is unusually long ( ∼3 . 5 kb ) for a Chlamydomonas gene ( Figure 3A ) . Similar to Chlamydomonas NMNAT , Chlamydomonas NS contains extra sequences not found in other organisms . This insert is rich in alanine residues ( Figure S5 ) . The coding region of NS in nic13-1 was sequenced and a triple nucleotide substitution ( TCC→ATT ) is predicted to produce a S740I change ( Figure 3A ) . The S740 is highly conserved among plants , green algae , and mammals ( Figure S5 ) . Further evidence that this point mutation is responsible for the mutant phenotype is provided by reversion analysis . We reasoned that a single base change of ATT ( I ) to AGT ( S ) would generate an I740S reversion . This change would also create a SfcI site ( TTCTACAGTA ) , which is not present in either wild-type or nic13-1 cells ( Figure 3B ) . Eighteen 3-AP-resistant colonies were isolated following UV mutagenesis of nic13-1 cells . Six of them contained a SfcI site as monitored by PCR and restriction digestion of the product ( Figure 3B ) . Of these , three revertants were randomly selected for sequencing . The ATT→AGT change was confirmed in all three selected revertants . The other 12 revertants were not analyzed . Transformation of nic13-1 cells with BAC ( 10H24 ) produced two 3-AP-resistant colonies and this provides further evidence that the mutation in NS is responsible for the nic13-1 mutant phenotype . Since Chlamydomonas nic mutants can utilize both NAM and NMN ( Figure 2 ) , two metabolites found in the 2-step salvage pathway , we propose that Chlamydomonas uses this pathway to synthesize NAD+ . The 2-step salvage pathway utilizes two enzymes , NAMPT and NMNAT . A NAMPT homolog , which is ∼30% identical to human NAMPT , was identified via sequence similarity search ( Figure S6 ) . RT-PCR and DNA sequencing indicated the Chlamydomonas NAMPT gene ( NPT1 , GenBank HM061641 ) contains 10 exons and encodes a 543 aa protein ( Figure 4A ) in CC-124 wild-type cells . However , we failed to amplify the full-length NPT1 transcript ( ∼2 . 2 kb ) from another wild-type strain ( CC-125 ) ( Figure 4C ) . Further investigation using 3′ RACE and 5′ RACE shows that an insertion in exon 2 of the NPT1 gene is present in the CC-125 strain ( Figure 4A , 4B ) . This region was partially sequenced and the inserted DNA sequence maps to multiple places in the genome and is likely to contain one or more transposable elements . The insertion causes two truncated NPT1 transcripts in the CC-125 strain . The first one is ∼0 . 6 kb long and contains the endogenous NPT1 promoter and ends within ∼100 bp of the inserted DNA ( Figure 4A ) . It is predicted to contain an open reading frame ( ORF ) , which encodes a 127 aa protein . This predicted protein contains the first 60 aa of the conserved phosphoribosyltransferase domain , which is ∼450 aa long . The second transcript is ∼1 . 8 kb long and starts with ∼130 bp of the inserted DNA ( Figure 4A ) . This truncated transcript contains part of exon 2 and the rest of the gene and is predicted to contain an ORF encoding a 435 aa protein . The predicted protein lacks the first 65 aa of the conserved phosphoribosyltransferase domain . Thus , we conclude that the CC-125 strain carries a defective NPT1 and we name the allele npt1-1 ( nicotinamide phosphoribosyltransferase ) . All the nic mutants contain a full-length NPT1 transcript ( Figure 4B , 4C ) . Given that CC-124 and CC-125 strains are considered to be “wild-type” strains in the Chlamydomonas community , we tested whether any other wild-type strains contain this insertion . The CC-124 and CC-125 strains originated from the 137c zygotic isolate by Smith in 1945 [36] . The meiotic progeny from 137c was distributed to Sager in 1953 and to Ebersold in 1955 . Four of the strains we tested , CC-1690 ( 21gr ) , CC-1691 ( 6145c ) , CC-407 ( C8 ) , and CC-408 ( C9 ) , originated from the Sager 1953 branch . The other three strains , CC-503 ( cw92 , used for the genomic sequence by JGI ) , CC-620 ( R3 ) , and CC-621 ( NO ) , as well as CC-124 and CC-125 , all came from the Ebersold branch . Genomic DNA PCR was used to identify the transposon insertion event while cDNA PCR indicated the presence/absence of the NPT1 transcript . Most of the strains have an intact NPT1 gene ( CC-407 , 408 , 503 , 620 , 1690 , and 1691 ) . The CC-621 strain ( upper panel , Figure 4B ) has an insertion in exon 2 of NPT1 , but the insertional sequence is not identical to the CC-125 insertion since the PCR primers that recognize the insertion in CC-125 failed in CC-621 ( lower panel , Figure 4B ) . As expected , the CC-621 strain also does not express the full-length NPT1 transcript ( Figure 4C ) . In contrast to the nic mutant strains , the npt1-1 mutant strains , CC-125 ( npt1-1 ) and CC-621 ( npt1-2 ) , show no obvious growth defect on rich medium or medium supplied with 3-AP ( Figure 2 and data not shown ) . Sequence similarity searches indicate that Chlamydomonas does not contain four of the six enzymes required to synthesize NAD+ from the de novo tryptophan pathway and it is missing a homolog of NAPRT from the 4-step salvage pathway . Chlamydomonas has genes for the IDO and KMO enzymes in the tryptophan pathway and has a NAMase homolog in the 4-step salvage pathway . Given the apparent incompleteness of either of these two pathways , we hypothesized that Chlamydomonas is unable to synthesize NAD+ via the tryptophan pathway or the 4-step salvage pathway ( Figure 1 ) . If the hypothesis that Chlamydomonas contains only the de novo aspartate pathway and the 2-step salvage pathway is correct , then double mutant strains containing the npt1-1 mutation as well as one of the nic mutations should be lethal and fail to grow on medium supplied with NAM . Otherwise , if there were additional NAD+ synthesis pathways available , then npt1-1; nic2-1 or npt1-1; nic15-1 double mutants should survive on NAM medium . We performed crosses between the nic mutants and the npt1-1 mutant strain . Genotypes of the meiotic progeny were scored based on their viability ( NIC ) or inviability ( nic ) on medium containing 3-AP , and the presence ( npt1 ) or absence ( NPT1 ) of the NPT1 transposon insertion , which was tested by genomic DNA PCR . The results are summarized in Table 2 . Out of 198 viable progeny generated from 4 different crosses , no npt1; nic double mutants were recovered on NAM medium . Because addition of NMN rescued the nic mutants and it is downstream of the NAMPT catalyzed step , we expected that npt1; nic double mutants should be viable on medium supplied with NMN . However , we found that no npt1; nic double mutants were isolated out of 148 viable progeny on NMN medium . One potential cause may be the hydrolysis or inefficient uptake of NMN by meiotic progeny compared to vegetative cells . Alternatively , the NIC1 message may not be expressed in meiotic progeny ( Figure 2 ) . Thus , based on the results from these genetic crosses between the nic mutants and the npt1-1 mutant , we conclude that Chlamydomonas synthesizes NAD+ via the de novo aspartate pathway and the 2-step salvage pathway and it is very unlikely that there is additional NAD+ biosynthesis pathway . Previous studies on bacterial and mammalian NAD+ biosynthesis indicate that transcriptional regulation among genes involved in the pathways is common . In Escherichia coli and Salmonella enterica , expression of nadB ( encodes ASO ) and nadA ( encodes QS ) is regulated by nadR , which has NMNAT activity [37] , [38] . In mammals , the circadian expression of NAMPT is partially regulated by SIRT1 , the enzyme that converts NAD+ to NAM , which is the substrate of NAMPT [39] . To investigate whether transcriptional regulation among the NIC genes and NPT1 exists in Chlamydomonas , we measured transcript levels of these genes in nic and npt1-1 mutants by quantitative real-time RT-PCR ( qRT-PCR , Figure 5 ) . Changes were not considered significant unless a gene is >2-fold upregulated or <2-fold downregulated when compared to its expression level in wild-type cells . The first step in the de novo aspartate pathway , which is rate limiting in bacteria , is catalyzed by ASO , encoded by NIC15 . As expected , mutations in the downstream enzymes ( nic1-1 , nic2-1 , and nic13-1 ) result in reduced NIC15 transcript while the npt1-1 mutation causes a 2-fold elevation in NIC15 transcript level . The NIC7 transcript level was not affected in any of the mutants tested . NIC2 and NIC13 transcript levels are upregulated in the npt1-1 mutant and in all the nic mutants except nic2-1 . The NIC1 transcript level is upregulated in the nic mutants but not in npt1 . Finally , the expression level of NPT1 is only upregulated in the nic15-1 mutant strain . Overall , gene expression of the NIC , NPT1 and SRT2 genes shows complicate patterns . No single gene is upregulated or downregulated in all mutants and no single mutant shows clear upregulation or downregulation of all genes involved in a pathway . This result suggests that in addition to regulation at the transcription level , NAD+ biosynthesis may be regulated post-transcriptionally . In studies of yeast , worms , and mammals , upregulation of NAD+-dependent histone deacetylase Sir2/SIRT1 is correlated to longevity [40] , [41] . In rice , RNA interference of OsSRT1 leads to DNA fragmentation and programmed cell death [42] . We wanted to test if SIR2 homologs play a similar role in algae . A sequence similarity search using SIR or SIRT proteins finds two proteins in Chlamydomonas . We named the one most to similar to yeast Sir2 protein SRT2 . RT-PCR and sequencing show that the Chlamydomonas SRT2 gene contains 9 exons ( Figure 6A ) and encodes a 320 aa protein ( GenBank HM061642 ) . Sequence alignment ( Figure S7 ) shows that this protein contains the NAD-dependent catalytic core domain and is closely related to human SIRT6 , SIRT7 , and plant SRT proteins , which are class IV SIRT proteins . The second SIR2-like gene , SRT1 , is most similar to human SIRT4 [11] , which is a mitochondrial protein . This gene encodes a 399 aa protein ( GenBank HM161714 ) and belongs to the class II sirtuin family ( Figure 6A and Figure S7 ) . Since Sir2-like proteins are involved in the enzymatic step of converting NAD+ to NAM , we tested the expression of Chlamydomonas SRT1 and SRT2 by qRT-PCR . The transcript level of SRT1 is extremely low and we could not obtain informative qRT-PCR data . Thus , we focused on SRT2 transcript levels in wild-type , nic and npt1-1 mutants ( Figure 5 ) , and find that SRT2 remained unchanged in all strains except in nic13-1 cells , which show a ∼2 . 5 fold increase . We then tested whether this increase of SRT2 expression in nic13-1 cells affects Chlamydomonas cell longevity . We took advantage of the Chlamydomonas uni3-1 cells , which have a deletion of delta-tubulin . A pedigree analysis of this mutant suggested that the flagellar number is a metric of the mitotic age of cells ( Figure 6B ) . As shown by Dutcher and Trabuco , biflagellate cells are only produced by uni3-1 cells that have undergone at least two cell divisions [43] . Aflagellate cells never produce biflagellate daughters , but a uniflagellate or biflagellate uni3-1 cell produces one aflagellate daughter cell and one biflagellate daughter cell . We suggest that the biflagellate cell is the equivalent of using the mother cell in budding yeast as a marker of generational age . Having two flagella allows a cell to swim effectively to the air-liquid interface of the medium , while an aflagellate or uniflagellate daughter cell sinks to the bottom of the culture tube . The biflagellate daughter cells can then be transferred to a new test tube and of the number of generations that the uni3-1 cells undergoes can be monitored ( Figure 6C ) . As illustrated in Figure 6D , NIC13; uni3-1 biflagellate cells complete 38–40 cell cycles . On the other hand , nic13; uni3-1 cells complete 48–50 cell cycles . This represents a ∼25% increase in reproductive capacity . We assayed a nic2-1; uni3-1 strain since it does not have increased SRT2 levels but has a synthesis defect and found that it completed 37 cell cycles like wild-type cells ( data not shown ) . Therefore , we conclude that NAD+ biosynthesis in Chlamydomonas can affect life span and this might be achieved by alternating the expression level of Chlamydomonas SRT2 .
The essential roles of NAD+ in many metabolic oxidation-reduction reactions are well established . Recent studies link its function to transcriptional regulation [44] , epigenetic regulation [45] , longevity [2] , cell death [46] , neurogeneration [47] , circadian clocks [48] , and signal transduction [49] . Understanding its biosynthetic pathways will facilitate understanding of lifespan extension [50] , disease regulation [51] , drug design [52] , as well as evolution [53] . Recent studies on NAD+ biosynthesis indicate that pathways in different organisms are more diverse than expected . The tryptophan pathway , which was thought to be eukaryotic specific , was identified in several bacteria [54] . The aspartate pathway , which was considered only prokaryotic , is present and essential to Arabidopsis thaliana [28] . An organism may contain all the enzymes required for more than one pathway , but a single pathway is predominantly used . Bacillus subtilis can synthesize NAD+ via aspartate or the 4-step pathway but only genes involved in the conversion from NA to NAD+ are indispensable [55] . Arabidopsis thaliana contains the aspartate pathway and the 4-step pathway . However homozygous ASO and QS mutants , which specifically affect the aspartate pathway , are inviable [28] . In mammals , the enzyme NAMPT , which is the rate-limiting enzyme in the 2-step pathway , is essential even though organisms harbor all the enzymes required to synthesize NAD+ from tryptophan [56] . However in D . melanogaster and C . elegans , there is only one pathway . The de novo NAD+ synthesis pathway is incomplete and they rely on the NAMase-dependent salvage pathway to synthesize NAD+ [53] . Our study indicates that Chlamydomonas can synthesize NAD+ via the aspartate pathway , which is found in land plants and bacteria , or the 2-step salvage pathway , which is found in mammals . This combination in Chlamydomonas makes it a great model for the study of NAD+ biosynthesis . Similar to Arabidopsis , Chlamydomonas contains one copy of each gene that encodes the de novo pathway enzymes and the Chlamydomonas proteins are 51%∼63% identical to Arabidopsis homologs . However , unlike Arabidopsis mutants , which are lethal when homozygous [28] , [57] , the Chlamydomonas nic mutants show conditional lethality as they are rescued by the addition of NAM or NMN to the medium . Thus , the effects of loss of function mutations , which can not be studied in Arabidopsis , can be easily analyzed in Chlamydomonas . In mammals , NAMPT is essential . It is encoded by three different genes and the proteins localize to different cellular compartments . In addition , mammal NAMPT has an extracellular form; both intracellular ( iNAMPT ) and extracellular ( eNAMPT ) forms are involved in NAD+ synthesis and in regulation of insulin secretion in pancreatic β cells [58] . Chlamydomonas contains only one copy of NAMPT ( NPT1 ) . The npt1-1 mutant has no growth defect but none of the nic; npt1-1double mutants are viable ( Table 2 ) . Since Chlamydomonas cells are haploid and easy to maintain , this mutant provides an alternative for screening for NAMPT-blocking drugs . The potential drugs would have no effect on wild-type cells but would be lethal to nic cells . In mammals 3-AP acts as an analog of nicotinic acid and inhibits the 4-step salvage pathway . In Chlamydomonas , 3-AP prevents the rescue of nic mutants by NMN and greatly suppresses the rescue by NAM . The easiest explanation for these results would be that Chlamydomonas has the 4-step salvage pathway and it is active . However , the Chlamydomonas genome has only three of the four enzymes; the genome assembly is missing the key enzyme , NAPRT . It remains a possibility that Chlamydomonas has a NAPRT gene , but it is not present in the assembled genome sequence . Two lines of evidence suggest that a functional NAPRT is not likely to be present in Chlamydomonas . First , using 40 million Illumina transcriptome reads ( 1 . 2 Gb of sequence ) from three independent mRNA preparations , we find evidence for transcription of the first 45 amino acids of NAPRT using a splice aware assembly algorithm , but find no evidence for the transcription of the rest of the gene that contains all of the known active sites needed for function [59] , [60] ( unpublished data ) . Given the high identity of this protein from microalgae to mammals , it is likely either that the rest of the Chlamydomonas NAPRT gene was lost or the gene is not transcribed beyond the first 135 bps of the open reading frame . Second , the genome sequence of Volvox carteri ( http://genome . jgi-psf . org/Volca1/Volca1 . home . html ) , a multicellular green alga that shared a common ancestor with Chlamydomonas around 35 million years ago [61] , also lacks NAPRT . Therefore , we suggest that Chlamydomonas cells do not have a functional copy of NAPRT . It remains an open possibility whether an alternative enzyme without sequence similarity exists in Chlamydomonas . Our study on Chlamydomonas also provides important insights into the evolution of NAD+ biosynthesis ( Figure 7 ) . Through sequence similarity searches , we find that Volvox contains enzymes required for the aspartate de novo pathway and the 2-step salvage pathway . Given the common ancestor , it is not surprising that both of them contain NAMPT , the enzyme that is unique to the 2-step pathway . Two unicellular green microalgae , Ostreococcus lucimarinus and Ostreococcus tauri [62] , [63] , contain enzymes required for the aspartate de novo pathway and the 4-step salvage pathway . Ostreococcus are believed to have diverged from Chlamydomonas around 750 million years ago , ∼250 million years after the separation of chlorophytes ( green algae ) and streptophytes ( seed plants ) [64] . The unicellular choanoflagellate Monosiga brevicollis , which is considered the progenitor to animals and separated from other metazoans more than 600 million years ago [65] , has enzymes found in the tryptophan pathway and the 2-step pathway , as in animals . Chlamydomonas , which has remnants of four pathways , may suggest that an ancestral organism had multiple pathways and that most organisms have retained only a subset . In Arabidopsis , the first three enzymes , ASO , QS , and QPT , are localized to chloroplasts . It is currently unclear where the other two proteins , NMNAT and QS , are localized . When we used Predotar [66] and TargetP [67] for Chlamydomonas protein localization prediction , ASO , QS , and NS are predicted to localize to mitochondria by both programs . NMNAT is predicted to be in the mitochondria by TargetP while Predotar gives no specific location . The localization of QPT is unspecified by either program . The actual localization of Chlamydomonas proteins will require experimental determination . If all Arabidopsis enzymes are localized to chloroplasts while all Chlamydomonas enzymes are not , it would suggest that having NAD+ biosynthesis in plastids happened after the separation of green algae and seed plants . As illustrated in Figure 1 , NMNAT is an essential enzyme utilized by all NAD+ biosynthetic pathways . We observed that NIC1 has a low basal expression level in wild-type cells compared to the other NIC genes , but is upregulated 2∼6 fold in various nic mutants . This upregulation is consistent with the hypothesis that NMNAT is the key enzyme involved in all NAD+ biosynthetic pathways and any mutation along the pathway affects the expression of NIC1 significantly . The nonsense mutation found in nic1-1 cells presumably generates a truncated protein that must be partially functional as we would expect that a null mutant would disrupt both pathways and be lethal like the double nic; npt1-1 mutant strains . The truncated protein has the catalytic motif residue H30 but only one of two substrate binding motif residues ( W98 and not R224 ) [68] , [69] . Through our sequence similarity search , only two of the six homologs in the de novo synthesis pathway starting from tryptophan were identified . Previously , nic1-1 and nic4 mutants were reported to grow on medium supplemented with 3-HA , a metabolite produced in the tryptophan pathway [18] . We find that the growth defect of nic1-1 cannot be rescued by the addition of 3-HA ( Figure 1 ) and this agrees with our finding that NIC1 encodes NMNAT , which acts downstream of 3-HA . In addition , 3-HAO , the enzyme that uses 3-HA as a substrate , is not present in the Chlamydomonas genome sequence . The nic4 mutant is no longer existent in the Chlamydomonas strain collection and was never mapped ( [19] , www . chlamydb . html ) . Therefore , we are unable to test its grow ability on medium provided with 3-HA . Similar to our finding , 3-HA and other intermediates found in the tryptophan pathway fail to rescue nicotinamide requiring mutants in Chlamydomonas eugametos [70] . Recent studies indicate that nicotinamide riboside ( NR ) and nicotinic acid riboside ( NaR ) are NAD+ precursors in yeast and mammalian cells [71]–[73] . Enzymes involved in the NR and NaR salvage pathways include nicotinamide riboside kinase ( NRK1 ) , purine nucleoside phsophorylase ( PNP1 ) , uridine hydrolase ( URH1 ) , and methylthioadenosine phosphorylase ( MEU1 ) . Similarity searches using yeast protein sequences identified only one PNP1-like protein in Chlamydomonas , but none of the other proteins . Thus , it is unlikely that Chlamydomonas contains the NR/NaR salvage pathway . In the study of longevity , several model organisms ( S . cerviseae , C . elegans , D . melanogaster , and mouse ) have been widely used . Caloric restriction leads to extended life span in these organisms , but the mechanisms behind these findings are not well understood . Studies indicate that caloric restriction-mediated longevity links to upregulation of Sir2 in yeast [41] , [74] , flies [75] , and mammals [76] but is independent of Sir2 expression in worms [77] . However , increasing the dosage of SIR2 in C . elegans leads to longer life span [40] . Our observation that mutant cells with a longer life span have increased SRT2 expression suggests a link between Chlamydomonas SIR2-like genes and longevity . It is intriguing that only the nic13-1 mutant strain has increased levels of SRT2 . Since nic13-1 mutants should have increased levels of the intermediate , NaAD , we attempted to ask if exogenous NaAD altered SRT2 levels . Exogenous NaAD failed to rescue upstream mutants , which suggests that it was not effectively imported into cells ( Figure 2 ) . We assayed replicative aging in Chlamydomonas using centriole or basal body age as our marker . In the uni3-1 populations , the biflagellate cells contain a grandmother centriole ( at least three cell cycles old ) and a daughter centriole . We can recover biflagellate cells by virtue of their ability to swim . The cells that are biflagellate represent the oldest cells in the population . We find that wild-type cells fail to divide after 38–40 generations while the nic13-1 mutant continues for at least 10 more cell divisions . We suggest that this aging may include aging of the centrioles . Recent studies on fruit fly germline stem cells [78] and mouse neural progenitor cells [79] indicate that the mother centriole stays with the self-renewing daughter stem cell while the daughter centriole goes with the differentiating daughter cell . As cells age , misorientation of centrioles accumulates and eventually causes cell cycle delay or arrest in mouse neural progenitor cells . Using Chlamydomonas as a model system to study aging , we can further pursue the link between NAD+ metabolism , Sir2-like genes , and centriole aging . Whether overexpression of SIR2 in Chlamydomonas causes extended life span as shown in other organisms needs additional experimentation . A recent study on mammalian SIRT1 indicates that it is involved in regulation of circadian rhythm via transcriptional regulation of several key genes [80] . It is currently unclear whether other SIRT proteins have similar effect on circadian rhythm . Given that synchronous Chlamydomonas cell culture can be easily achieved by alternating light/dark cycles , we foresee Chlamydomonas as a model to explore the effect of SRT2 ( SIRT6-like ) and SRT1 ( SIRT4-like ) on circadian rhythms [81] . In conclusion , the results presented in this work underscore several key advantages of using Chlamydomonas as a model system for further studies of NAD+ metabolism . The Chlamydomonas genome contains a single copy of each of the proteins that make up the plant-specific de novo NAD+ biosynthesis pathway . However , unlike Arabidopsis , which is homozygous lethal for the first three enzymes , all five Chlamydomonas mutants show conditional lethality . Consequently , Chlamydomonas will facilitate future studies on metabolites involved in NAD+ biosynthesis . Chlamydomonas also contains a single copy of the genes in the mammal-specific 2-step NAD+ salvage pathway . The fact that mammals contain multiple isoforms of NAMPT and that this enzyme is essential to viability impede NAMPT-blocking drug studies in mammal-based model systems . As such , NAMPT targeted drug screens using Chlamydomonas avoid the many confounding factors that are inherent in current screening methods . Our centriole aging results demonstrate how Chlamydomonas may be a valuable model organism for future studies in cellular and organelle aging .
Chlamydomonas reinhardtii strains , CC-14 ( nic15-1; mt+ ) , CC-124 ( mt− ) , CC-125 ( mt+ ) , CC-407 ( C8 , mt+ ) , CC-408 ( C9 , mt− ) , CC-503 ( cw92; mt+ ) , CC-599 ( nic1-1; mt+ ) , CC-620 ( R3 , mt+ ) , CC-621 ( NO , mt− ) , CC-864 ( nic13-1; mt+ ) , CC-1079 ( ac12; thi9; nic2-1; mt+ ) , CC-1690 ( 21gr , mt+ ) , CC-1691 ( 6145c , mt− ) , and CC-3657 ( nic2-1; mt+ ) , were obtained from Chlamydomonas Center ( Duke University ) and maintained on solid rich growth ( R ) medium [82] or medium containing 2 µg/ml ( 16 µM ) nicotinamide ( NAM ) . To confirm that the nic mutant strains show the Nic− phenotype , cells were plated on R medium containing 15 µl/l ( 16 . 5 mg/l ) 3-acetylpyridine ( 3-AP ) [20] . The original nic15-1 strain acquired from the Chlamydomonas Center failed to confer sensitivity to 3-AP , which suggests the possibility of a revertant or an extragenic suppressor . A backcross to the wild-type strain produced progeny sensitive to 3-AP , which reveals the presence of an extragenic suppressor in the original stock culture . The 3-AP sensitivity phenotype of the nic2 strain ( CC-3657 ) was difficult to score , so a second strain CC-1079 ( ac12; thi9; nic2-1; mt+ ) was backcrossed to wild-type cells several times to generate an AC12; THI9; nic2-1 strain that confers sensitivity to 3-AP . For the spotting assay , 104 cells were spotted on R medium or R+3-AP medium supplemented with one of the following compounds: 10 µM NAM , 10 µM nicotinamide mononucleotide ( NMN , dissolved in water ) , 10 µM nicotinate adenine dinucleotide ( NaAD , dissolved in water ) , 10 µM nicotinic acid ( NA , dissolved in water ) , or 10 µM 3-hydroxyanthranilate ( 3-HA , dissolved in DMSO ) . The plates were placed under constant light at room temperature for 3 days before pictures were taken . All the reagents were obtained from Sigma ( St . Louis , MO ) . We have changed the linkage group names to chromosome names as specified in [15] . Linkage groups I-XI correspond to chromosomes 1–11 . Linkage group XII/XIII is chromosome 12 , and Linkage group XV is chromosome 14 . Protein sequences of Arabidopsis thaliana ASO , QS , QPT , NMNAT , NS ( listed in Table 1 ) , human NAMPT ( NP_005737 ) , yeast SIR2 ( NP_010242 ) , and human SIRT4 ( NP_036372 ) were used in TBLASTN against JGI ( Joint Genome Institute ) Chlamydomonas reinhardtii genome version 4 . 0 ( JGI v4 . 0 , http://genome . jgi-psf . org/Chlre4/Chlre4 . home . html ) with expected E-values less than or equal to 1E-5 ( 1E-3 for SIR2 and SIRT4 ) . The resultant genes were checked for EST coverage . Genes without full-length EST coverage , QS , NMNAT , NS , NAMPT , SRT1 , and SRT2 , were subjected to exon-intron predictions using GreenGenie 2 ( http://bifrost . wustl . edu/cgi-bin/greengenie2/greenGenie2 ) [33] . The predicted coding regions were used as guidelines in primer design for RT-PCR to amplify the actual coding regions of these genes . Multiple sequence alignments ( MSA ) were color-coded using the online MSA column percentage composition coloring tool , Colorfy ( http://bifrost . wustl . edu/colorfy ) . Colorfy takes as input any standard ALN format MSA ( e . g . default CLUSTAL output ) [83] and outputs the corresponding color-coded MSA . Chlamydomonas total RNA was prepared as previously described [84] . Five µg of total RNA from wild-type cells were used for cDNA synthesis using a 3′ RACE poly ( dT ) -adaptor primer ( Integrated DNA Technologies , Iowa City , IA ) in a 50 µl reaction , which contains 1× RT buffer ( Invitrogen , San Diego , CA ) , 10 mM DTT , 0 . 5 mM dNTP , 0 . 2 µM primer , 40 U of RNaseOUT ( Invitrogen ) , and 200 U of SuperScript II reverse transcriptase ( Invitrogen ) . The reaction was performed according to manufacturer's recommendation ( Invitrogen ) . To remove RNA from the reaction , 2 units of RNase H ( Invitrogen ) were added at the end of reaction and incubated at 37°C for 20 min . Amplification of the NMNAT coding region requires nested PCR due to highly repetitive sequences found in the gene . Five µl cDNA ( 1/10 of the reaction volume ) from above was used in a 50 µl PCR reaction using a 3′ RACE primer and a gene-specific primer ( nic1-3 ) that binds 4 nucleotides downstream of the predicted start codon . The reaction , which contained 1× KlentaqLA buffer ( pH 9 . 2 ) , 0 . 8 mM dNTP , 10% DMSO , 1 mM MgCl2 , 0 . 5 µl KlentaqLA polymerase [85] , was transferred directly from ice to a thermocycler ( Bio-Rad , Hercules , CA ) that was preheated to 93°C . The reaction conditions were: 93°C 5 min , 30 cycles of ( 93°C 15 sec , 53°C 15 sec , and 68°C 5 min ) , and 70°C 10 min . The resultant 2 . 2 kb fragment was used as template for a second round of amplification . A forward primer ( nic1-20 ) that starts 98 nucleotides downstream of the predicted start codon and a reverse primer ( nic1-24 ) that ends at the predicted stop codon were used . The resultant fragment was gel purified and subjected to DNA sequencing . For amplification of other genes , 1 µl cDNA was used in a 20 µl PCR reaction containing 0 . 4 U Phusion DNA polymerase ( Finnzymes , Woburn , MA ) , 1× GC buffer ( Finnzymes ) , 0 . 2 mM dNTP , 3% DMSO , and 0 . 2 µM each of forward and reverse primers . The general reaction condition was 98°C 30 sec , 30 cycles of ( 98°C 10 sec , T°C 20 sec , and 72°C 30∼45 sec ) , and 72°C 10 min . T is the lower Tm of the primers calculated by Finnzymes' Tm calculator . Different sets of primers were used to cover the whole coding region of individual genes . The PCR products were subjected to gel purification and DNA sequencing to identify exon-intron boundaries . A DNA mini-prep protocol was modified [86] and used . Approximately 1×106 cells were resuspended in 0 . 5 ml 1× TEN ( 150 mM NaCl , 10 mM EDTA pH 8 . 0 , 10 mM Tris-HCl pH 8 . 0 ) and pipetted repeatedly until well resuspended . Cells were collected by centrifugation at 13 , 200 rpm for 10 sec in a microcentrifuge ( Hermle Z233 M-2 , Labnet , Woodbridge , NJ ) and the supernatant was discarded . Cells were resuspended with 150 µl chilled water , followed by the addition of 300 µl SDS-EB buffer ( 2% SDS , 100 mM Tris-HCl pH 8 . 0 , 400 mM NaCl , 40 mM EDTA pH8 . 0 ) . DNA was extracted once with 350 µl phenol/chloroform ( 1∶1 ) , followed by a second extraction using 350 µl chloroform . The volume was determined and twice the volume of 100% ethanol was added to precipitate DNA on ice for 30 min . Precipitated DNA was collected by centrifugation at room temperature for 10 min followed by a wash using 70% ethanol . DNA was dried using Savant SpeedVac ( Thermo Scientific , Waltham , MA ) and resuspended in 50 µl water . The concentration of DNA was determined by spectrophotometry at 260 nm ( Eppendorf Biophotometer 6131 , Westbury , NY ) . Approximately 20 ng of genomic DNA was used in PCR and the resultant PCR products were gel-purified and subjected to DNA sequencing . In the nic1-1 cells , the region that carries mutations were amplified by the primer set nic1-10 and nic1-11 . In the nic15-1 cells , the region that contains a point mutation was amplified by nic15-3F and nic15-3R . Chlamydomonas BAC DNA was prepared using Qiagen Plasmid Midiprep kit . To prepare the pNIC15a plasmid , the BAC ( 32L22 ) DNA was digested with XmaI and a 6 . 1 kb fragment was isolated and cloned into a pBlueScript II SK vector ( Stratagene , La Jolla , CA ) . This fragment contains a 1 kb upstream sequence , the full-length NIC15 gene , and a 2 . 5 kb downstream sequence , which is predicted to be part of an unknown zinc finger protein ( protein id 150664 ) . To prepare the pNIC1-56 plasmid , a 7 . 1 kb KpnI fragment from the BAC ( 10M24 ) DNA was cloned into a pBlueScript II SK vector . The plasmid contains a 0 . 7 kb upstream sequence , the full-length NIC1 gene , and a 4 . 7 kb downstream sequence , which is predicted to contain an unknown protein that has a HAD-superfamily hydrolase domain . This protocol is modified from Iomini et al [87] . Chlamydomonas cells were inoculated in 100 ml liquid R medium for three days under continuous illumination with gentle shaking until cells reached a concentration of ∼5×106 cells/ml . Cells were collected by centrifugation and treated with autolysin for 0 . 5 hr at room temperature to remove cell walls [19] . Autolysin-treated cells were chilled on ice for 10 min before collected by centrifugation at 4°C . Cells were gently resuspended on ice in R+100 mM mannitol to the final concentration of ∼4×108 cells/ml . Two hundred fifty µl of cells ( ∼1×108 cells ) were used for transformation with 1 µg of BAC DNA or plasmid DNA with ( the nic1-1 strain ) or without ( the nic15-1 , nic2-1 , and nic13-1 strains ) the addition of 1 µg of pSI103 , which confers resistance to paromomycin [35] , for cotransformation . Cells and DNA were added to an electroporation cuvette ( 4mm gap , Bio-Rad ) and incubated in a 16°C water bath for 5 min before electroporation , which was performed in a Bio-Rad Gene Pulser II with the following setting: 0 . 75 kv , 25 µF , and 50 Ω . Cells were electroporated with one pulse and incubated at room temperature for 10 min before transferring to 50 ml R+100 mM mannitol liquid medium and incubated overnight at room temperature with continuous illumination . Cells were resuspended gently in 1 ml 25% cornstarch in R medium and spread onto 5 R plates with 15 µl/l 3-AP ( nic15-1 , nic2-1 , and nic13-1 cells ) or 5 R plates with 10 µg/ml paromomycin ( nic1-1 cells ) . Colonies appear within 5∼7 days at 25°C . The nic1-1 transformants were tested subsequently on medium with 3-AP . nic13-1 cells were inoculated in 200 ml liquid R medium provided with 16 µM NAM for 4 days until cells reached a density of ∼106 cells/ml . These cells were collected and spread evenly on an R+NAM medium plate . The cells were subjected to UV irradiation at 70 mJoules ( Stratagene UV stratalinker 1800 , Cedar Creek , TX ) and recovered in the dark overnight . The plate was divided into 13 sections and cells were scraped off the plate and spread on 13 R+3-AP plates . 3-AP resistant colonies were observed one week later . Genomic DNA from individual cell lines , wild-type , and nic13-1 cells were prepared as above and a short region was amplified by primers nic13-20F and nic13-3R by Phusion DNA polymerase . The PCR products were subjected to overnight digestion with SfcI at 25°C and separated on a 2% agarose gel . Chlamydomonas total RNA was extracted from ∼108 cells using Qiagen RNeasy Mini Kit ( Qiagen , Valencia , CA ) . Cells were homogenized by passing through a 20-gauge needle fitted to a 1 ml RNase-free syringe 20 times . The lysate was centrifuged and RNA extraction was performed according to manufacturer's recommendation . One microgram of total RNA from each strain was treated with 1 U of RNase-free DNAse I ( Fermentas , Glen Burnie , MD ) at 37°C for 30 min and the reaction was terminated by adding 1 µl of 25 mM EDTA and incubate at 65°C for 10 min . The DNAse I-treated RNA was added into a 20 µl reverse transcription reaction that contains 200 ng random primers ( Invitrogen ) , 1× RT buffer ( Invitrogen ) , 5 mM DTT , 0 . 5 mM dNTP , 20 U of RNaseOUT ( Invitrogen ) , and 100 U of SuperScript III reverse transcriptase ( Invitrogen ) . The reaction was performed according to manufacturer's recommendation ( Invitrogen ) . For real-time PCR , cDNA obtained from above was diluted 1∶10 and 2 µl was used in a 20 µl SYBR Green-PCR reaction [88] which contains 1× homemade PCR buffer ( 10 mM Tris-HCl , pH8 . 8 , 50 mM KCl , 2 mM MgCl2; 0 . 1% Triton X-100 ) ; 1× SYBR Green I mix ( 1× SYBR Green , Molecular Probes; 10 nM Fluorescein , Bio-Rad; 0 . 1% Tween-20; 0 . 1 mg/ml BSA; 5% DMSO ) ; 200 µM dNTP; 0 . 5 µM primers; and 1 . 6 µl TAQ DNA polymerase [89] . The reactions were carried out using a Bio-Rad iCycler iQ under the following conditions: 95°C 3 min , 40 cycles of ( 95°C 10 sec and 57°C 45 sec ) , followed by the melting curve program . The transcript levels of individual genes were standardized by an internal control , CRY1 , which encodes the ribosomal protein S14 [90] . Gene expression was set to 100% in wild-type cells and the relative expression levels in various mutants were plotted as % increasing or decreasing related to transcript levels in wild-type cells . Results represent data from 2 biological replicates . nic13-1 cells were crossed to uni3-1 ( CC-4179 ) cells and the nic13-1; uni3-1 double mutants were identified by 3-AP sensitivity and the presence of cells with 0 , 1 , or 2 flagella . Both NIC13; uni3-1 and nic13-1; uni3-1 cells were inoculated in 20 ml liquid R medium supplied with 16 µM NAM . The top 5 ml of liquid was transferred to a new test tube containing R+NAM every 12 hours . ∼100 cells were plated on R+NAM plates and the fraction of cells that formed colonies was counted under dissecting microscope after 8–10 days . | Nicotinamide adenine dinucleotide ( NAD+ ) is an essential coenzyme . NAD+ is necessary for electron transfer in many metabolic reactions . NAD+ functions as a substrate for several enzymes , one of which is sirtuin , an enzyme involved in gene regulation and aging . NAD+ can be synthesized either from amino acids ( de novo ) or metabolites ( salvage ) . Given the importance of NAD+ , enzymes involved in NAD+ synthesis are targets for drug discovery . In the unicellular green alga Chlamydomonas we investigated both the de novo and salvage NAD+ biosynthetic pathways . Mutations in the plant-like de novo synthesis pathway lead to a nicotinamide-requiring phenotype . We identified an insertional mutation in the first enzyme in the mammal-like salvage pathway; it has no growth defect in cells with an active de novo synthesis pathway but causes lethality when the de novo synthesis pathway is inactive . Coupled with NAD+ biosynthesis , sirtuin is involved in NAD+ consumption . Our study links upregulation of a sirtuin gene with extended life span in the nic13-1 mutant strain , which has a defective de novo synthesis pathway and suggests that Chlamydomonas is an excellent genetic model to study NAD+ metabolism and cell longevity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/gene",
"discovery",
"cell",
"biology/gene",
"expression",
"computational",
"biology/molecular",
"genetics",
"genetics",
"and",
"genomics/gene",
"function",
"genetics",
"and",
"genomics/plant",
"genomes",
"and",
"evolution",
"genetics",
"and",
... | 2010 | Synthesizing and Salvaging NAD+: Lessons Learned from Chlamydomonas reinhardtii |
A balance between excitatory and inhibitory synaptic currents is thought to be important for several aspects of information processing in cortical neurons in vivo , including gain control , bandwidth and receptive field structure . These factors will affect the firing rate of cortical neurons and their reliability , with consequences for their information coding and energy consumption . Yet how balanced synaptic currents contribute to the coding efficiency and energy efficiency of cortical neurons remains unclear . We used single compartment computational models with stochastic voltage-gated ion channels to determine whether synaptic regimes that produce balanced excitatory and inhibitory currents have specific advantages over other input regimes . Specifically , we compared models with only excitatory synaptic inputs to those with equal excitatory and inhibitory conductances , and stronger inhibitory than excitatory conductances ( i . e . approximately balanced synaptic currents ) . Using these models , we show that balanced synaptic currents evoke fewer spikes per second than excitatory inputs alone or equal excitatory and inhibitory conductances . However , spikes evoked by balanced synaptic inputs are more informative ( bits/spike ) , so that spike trains evoked by all three regimes have similar information rates ( bits/s ) . Consequently , because spikes dominate the energy consumption of our computational models , approximately balanced synaptic currents are also more energy efficient than other synaptic regimes . Thus , by producing fewer , more informative spikes approximately balanced synaptic currents in cortical neurons can promote both coding efficiency and energy efficiency .
Cortical neurons receive many thousands of weak ( sub-millivolt ) excitatory synaptic inputs [1] , the majority of which originate from other local or distant neurons within the cortex [2] , [3] . The currents generated by these excitatory inputs are approximately balanced by inhibitory currents [4] , [5] generated by fewer , stronger synaptic inputs from inhibitory interneurons [6] . During ongoing activity in vivo , excitatory and inhibitory currents depolarize the membrane from the resting potential to around −60 mV , slightly below the threshold for spike initiation [7] . For excitatory and inhibitory currents to balance at approximately −60 mV , the inhibitory conductances must be larger than excitatory conductances . Operating this close to threshold , small fluctuations in synaptic inputs can depolarize the neuron sufficiently to trigger spikes , giving rise to highly variable interspike intervals , similar to those expected from a Poisson process [4] , [5] . Depolarizing the membrane with balanced synaptic currents also reduces the membrane time constant , thereby increasing temporal resolution and extending bandwidth [8]–[10] , and alters both the neuron's sensitivity and its working point by changing gain [11]–[14] . Thus , depolarization by balanced excitatory and inhibitory currents affects numerous aspects of information processing in cortical neurons . Cortical information processing accounts for a considerable proportion of the mammalian brain's energy consumption , and cortical energy usage is dominated by synaptic transmission and action potentials [15]–[17] . The cortex's restricted energy budget places limits on the mean spike rate and hence neural processing , suggesting that the cortex may be under strong selective pressure to save energy and increase efficiency [15] , [18] . Balanced synaptic currents increase energy consumption by depolarizing the membrane potential and lowering the input resistance . Consequently , balanced synaptic currents will affect cortical information processing and energy consumption , yet how they do so remains unclear . To assess the impact of balanced synaptic currents on information coding and energy consumption , we compared single-compartment models with stochastic voltage-gated Na+ and K+ channels driven by one of three synaptic input regimes; excitatory inputs only , equal excitatory and inhibitory conductances ( balanced synaptic conductances ) , and stronger inhibitory than excitatory conductances ( balanced synaptic currents ) . By quantifying the performance of these models over a range of synaptic input statistics , we show that balanced inhibitory and excitatory synaptic currents increase both coding efficiency ( bits/spike ) and energy efficiency ( ATP molecules/bit ) in comparison to the other synaptic input regimes . Two factors contributed to the superior efficiency of models with balanced synaptic currents , their firing rates were lower and their spikes more precise . Thus , our models show that balanced synaptic inputs can improve both the coding efficiency ( bits per spike ) and the energy efficiency ( bits per ATP molecule ) of cortical neurons .
We simulated the responses of a 100 µm2 single compartment model containing stochastic voltage-gated Na+ , K+ channels and a non-probabilistic leak conductance , to excitatory synaptic inputs alone ( Figure 1A ) or to combinations of excitatory and inhibitory inputs ( Figure 1B ) . In the limit of large numbers of Poisson synaptic events with small unitary conductances converging on the post-synaptic compartment , the conductances become a Gaussian white noise process ( “the diffusion approximation” ) [19] . For synaptic events with a finite time constant , fluctuations in conductance are represented as an Ornstein-Uhlenbeck ( OU ) process ( see Methods ) [20] , parameterized by the means ( μe , μi ) , the standard deviations ( σe , σi ) , and the time constant of the excitatory and inhibitory synaptic events ( τe , τI were both fixed at 3 . 3 ms ) [20] . The input conductance contrast is the ratio of σ to μ . The mean synaptic conductance depends upon the rate , the unitary synaptic event amplitude , and the exponential decay time constant of synaptic events ( Eq . 5 ) , whilst the contrast is a function of the rate and the decay time constant ( Eq . 7 ) . Therefore , when we increase the conductance contrast we are reducing the frequency with which afferent spikes activate synapses , and when we increase the mean conductance at constant contrast we are increasing event amplitude at constant rate . We modeled three synaptic input regimes . The first was excitation only ( Figure 1C , F , I ) . In the second regime , balanced conductance , the means and standard deviations of the excitatory and inhibitory synaptic conductances were equal , ( Figure 1D , G , J ) . In the third regime , approximately balanced current , the mean excitatory and inhibitory conductances were adjusted ( Figure 1E , H , K ) to produce approximately equal inward and outward currents at the mean sub-threshold membrane potential of approximately −64 mV . In this balanced current regime , μi = 5μe and , because inhibitory and excitatory conductances always had the same contrast , σi = 5σe [7] . All three synaptic regimes evoked action potentials ( APs , spikes ) , the rate of which depended upon the specific regime , and the stimulus mean and contrast ( Figure 1 ) . As expected , increasing the inhibitory input reduced spike rates ( Figure 1I–K ) . We quantified the differences in the spike rates of the models driven by different synaptic input regimes . Within each regime we varied the means of the excitatory and inhibitory inputs at different contrasts ( Figure 2 ) , while keeping μi = μe in the balanced conductance regime , and μi = 5μe in the approximately balanced current regime . At low contrasts ( i . e . high synaptic event rates ) , increasing the mean synaptic conductance in the excitatory regime increases the spike rate from ∼10 spikes/s with minimal input to over 40 spikes/s with 100 μS/cm2 ( Figure 2A , D ) . Adding an inhibitory conductance with the same mean conductance so that model operates in the balanced conductance regime , shifts the curve relating mean synaptic conductance to spike rate down , reducing the maximum spike rate to 30 spikes/s with 100 μS/cm2 ( Figure 2A , E ) . This downward shift reduces sensitivity , yet increases the range over which the compartment can operate . In the approximately balanced current regime , μi = 5μe , increasing the total mean conductance inverts the trend seen in the other two regimes; spike rates decrease from ∼10 spikes/s to ∼2 spikes/s ( Figure 2A , F ) . Next we examined responses to higher contrasts that are produced by larger synaptic events occurring at lower rates . In the excitatory regime the spike rate increases with the mean synaptic conductance , from ∼10 spikes/s with no input to ∼50 spikes/s with 100 μS/cm2 ( Figure 2B , D ) . As with low contrasts , the addition of an inhibitory input with balanced conductance , μi = μe , shifts the rate/conductance curve down , reducing the maximum spike rate to ∼40 spikes/s with 100 μS/cm2 ( Figure 2B , E ) . However , in the balanced current regime , μi = 5μe , increasing the total mean conductance shifts the rate/conductance curve down , reducing the maximum spike rate to ∼25 spikes/s with 100 μS/cm2 ( Figure 2B , F ) . Again , these downward shifts act as a divisive gain control , reducing sensitivity and increasing the range over which the compartment can operate . Thus , by adjusting the amount of inhibition it is possible to tune the responses of the post-synaptic neuron ( Figure 2B , right panel ) . Comparing different contrast levels in the approximately balanced current regime shows that the curve relating spike rate to excitatory conductance becomes steeper at higher contrasts ( Figure 2C ) . Thus , increasing the slope of the F–I curve is not only a property of the intrinsic biophysics but is also strongly dependent upon the input stimulus statistics ( cf . Stemmler and Koch [21]; Figure 2 ) . Differences in the inter-spike intervals of spikes evoked by the three synaptic regimes were quantified using the coefficient of variation ( CV ) ( see Methods ) . Irrespective of the stimulus contrast , excitatory synaptic inputs alone generated spike trains with a high CV when the mean conductance was low ( Figure 3A ) . The addition of inhibitory synaptic inputs of the same mean conductance and contrast increased the CV , indicating greater irregularity in the spike trains , even at high mean conductance levels ( Figure 3B ) . Increasing the inhibitory synaptic inputs to balanced currents , μi = 5μe , increased the CV still further , indicating even greater irregularity in the spike trains ( Figure 3C ) . The CV confounds variation due to the fluctuating synaptic input ( signal ) with noise generated by the stochastic activation of voltage-gated Na+ and K+ channels . Noise is identified by comparing responses to repeated presentations of the same signal and its effects on coding accounted for with information theoretic metrics [22] . For a given stimulus , the total entropy is a measure of the repertoire of spiking patterns that can be produced by the compartment , setting its information capacity [23] . We measured the total entropy by presenting a different conductance waveform on each subsequent trial ( unfrozen noise ) ( see Methods ) . The total spike train entropy generated by excitatory synaptic inputs alone increases with the mean conductance , μe ( Figure 3D ) . The addition of inhibitory synaptic inputs with the same mean and contrast decreases the total entropy ( Figure 3E ) , and entropy decreases still further when the current is approximately balanced by increasing the inhibitory input so that μi = 5μe ( Figure 3F ) . We also presented the same conductance waveform repeatedly ( frozen noise ) to quantify the noise entropy of the responses ( see Methods ) , which is a measure of spike train reproducibility [23] . With purely excitatory inputs of low contrast the noise entropy increases with mean conductance ( Figure 3G ) . The addition of inhibition that balances the excitatory conductance , μi = μe , decreases the noise entropy ( Figure 3H cf . Figure 3G ) , and again noise entropy increases as synaptic conductance increases . Increasing the relative strength of inhibition to approximately balance current , μi = 5μe , greatly reduces noise at all combinations of contrast and mean conductance ( Figure 3I ) , making the spikes more reproducible from trial to trial . The difference between the total and noise entropies determines the mutual information ( MI ) of the spike trains , a direct measure of the amount of information free of assumptions about how the information is represented and what it means [23] . We calculated the MI represented in the spike trains generated by each synaptic input regime ( Figure 4A–C ) . The information rate increases with input contrast when the synaptic inputs are purely excitatory ( Figure 4A ) because increasing contrast increases the signal amplitude , and hence the signal-to-noise ratio ( SNR ) within the model compartment . Incorporating inhibition identical to the excitation ( Figure 4B ) had little effect on the information rates , and they vary with contrast and mean conductance level in the same way . The changes are small because the addition of inhibition reduces the total entropy and the noise entropy by equivalent amounts ( median reduction is 1 . 1 fold ) . When inhibition is increased to approximately balance currents , μi = 5μe , the noise entropy reduces by a factor of 2 . 3 fold ( averaged across the range of contrasts and mean conductance levels ) and the total entropy reduces 1 . 7 fold ( Figure 4C ) . In other words , increased inhibition produces highly irregular spike trains that are precisely timed over trials . This simultaneous yet unequal drop in both total entropy and noise entropy produces a marginally better information encoding – the area of poor encoding ( low bit rate ) increases but there is a steeper rise to a higher bit rate at the highest values of contrast and mean conductance . Hence , more inhibition ( approximately balanced currents ) causes weak signals to perform worse and stronger signals to perform marginally better . Differences in the information rates of spike trains generated by the three synaptic regimes are dependent partly upon the spike rate ( Figure 2D–F ) [24] . However , by dividing the information rate by the corresponding spike rate for each conductance stimulus for a particular synaptic regime it is possible to determine the information encoded by each spike , the coding efficiency ( Figure 4D–F ) . The coding efficiency of spikes evoked by excitation alone or by identical excitation and inhibition was similar; both attained between 0 . 1 and 2 . 4 bits/spike with the higher values being generated by high contrast , low mean stimuli ( Figure 4D , E ) . Increased inhibition not only increases the coding efficiency across the entire stimulus space but also alters the trends so that low mean , low contrast stimuli evoke the most bits/spike ( Figure 4F ) . The higher coding efficiency of the increased inhibition stimuli derives from the increased reliability and precision of the spikes they generate ( Figure 3I ) and emphasizes that although they achieve similar information rates to the other synaptic regimes , they do so despite far lower spike rates . The ion fluxes across the membrane that generate electrical signals and noise in neurons consume energy because the Na+/K+ ATPase must expel Na+ ions and import K+ ions against their concentration gradients , using the energy provided by ATP [16] , [25] , [26] . The ATPase hydrolyzes one ATP molecule to ADP to expel 3 Na+ ions and import 2 K+ ions and this stoichiometry allows one to calculate the energy consumption ( Methods ) from the total fluxes of Na+ and K+ across the membrane [16] . In all three synaptic regimes , the model's energy consumption increased with the excitatory synaptic conductance so that spike trains generated by high mean , high contrast stimuli used the most energy ( Figure 5A–C ) . Comparison among the three synaptic input regimes shows that energy consumption across the entire stimulus space drops as inhibition increases ( Figure 5A–C ) . The total energy consumption of our single compartment model is determined by the currents flowing through the excitatory and inhibitory synaptic conductances , the voltage-gated ion channels that generate the action potentials , and the leak conductance [16] , [27] . We partitioned the energy consumption into these component parts to determine their relative contributions ( see Methods ) ( Figure S1 ) . Under all synaptic regimes , and with all combinations of contrast and mean synaptic conductance , the currents flowing through voltage-gated ion channels during action potentials ( Figure S1A–I ) were the primary energy consumers . This explains why the trends in energy consumption ( Figure 2D–F , 5A–C ) closely resemble those in spike rate ( Figure 2E–F ) . In both the excitation only and balanced conductance regimes , action potentials account for between 85 and 90% of the total energy consumption , and the highest AP consumption occurring with high mean , high contrast stimuli ( Figure S1C , F ) . The majority of the remaining energy is consumed by the leak conductance , between 5–12% , the energy consumed decreasing as the stimulus mean increases ( Figure S1A , D ) . The synaptic currents account for just 2–4% of the total energy consumption , increasing with higher stimulus means ( Figure S1B , E ) . Increasing inhibition to approximately balance the excitatory synaptic current , μi = 5μe , reduces the fraction of the energy consumed by the voltage-activated currents to between 50 and 80% ( Figure S1I ) . These active currents consume the least energy with high mean , low contrast stimuli , the costs rising with increasing contrast or decreasing stimulus mean ( Figure S1I ) . The opposite trend occurs for the synaptic costs , which rise from 2 to 30% of the total energy consumption as the stimulus mean increases and the contrast decreases ( Figure S1H ) . The leak current consumes between 7 and 15% , the highest consumption occurring at low contrasts ( Figure S1 ) . These trends can be explained by the reduced spike rates evoked by low contrast stimuli , especially with high mean stimuli , which cause the spike rate to drop below the spontaneous spike rate ( Figure 2F ) . The energy efficiency ( bits/ATP molecule ) of a spike train can be calculated by dividing the mutual information rate ( bits/s ) by the energy consumed ( ATP molecules/s ) . Increased inhibition generates spike trains that are more efficient than either excitation alone or identical excitation and inhibition irrespective of the stimulus mean and contrast ( Figure 5D–F ) . Increasing both the mean and the contrast of the stimulus produces the highest energy efficiency , up to 5*10−7 bits/ATP molecule for increased inhibition ( Figure 5D–F ) attributable to a drop in spike rate , which reduces total consumption while coding efficiency , the number of bits carried by each spike , increases ( Figure 5D–F ) . We compared the performance of the three synaptic regimes in terms of the net currents that they produce with low and high contrast stimuli . Both the excitation alone and the equal excitation and inhibition regimes generated an increasingly large net inward current as the mean excitatory synaptic conductance increases , irrespective of the stimulus contrast ( Figure 6 ) . However , when the inhibitory conductance is five-fold greater than the excitatory , there is no net current flow ( Figure 6 ) . Comparison of the three regimes shows that balanced synaptic currents generate spike trains with higher mutual information rates ( Figure 6A ) , and lower energy consumption ( Figure 6B ) than either of the regimes that produce a higher net current . Because of the low spike rates generated by balanced synaptic currents , this results in improved metabolic efficiency ( Figure 6C ) , and more information per spike ( Figure 6D ) than the other synaptic input regimes .
We have shown that approximately balanced inhibitory and excitatory synaptic currents increase both coding efficiency and energy efficiency in comparison to two other synaptic input regimes – excitation alone , and balanced excitatory and inhibitory conductances . Key to this improvement in efficiency is a reduction in spike rate and an increase in spike timing precision . The strong inhibitory conductance needed to generate a current that balances the excitatory current produced the lowest spike rates of all the regimes we studied across the entire input stimulus space . This reduction in spike rate is responsible for an overall drop in energy consumption ( ATP molecules/s ) because the voltage-gated currents that generate APs dominate the energy consumption of all the models . In the balanced synaptic current regime , the energy savings from lower spike rates are sufficient to offset the increased costs of the synaptic conductances . Yet , despite generating fewer spikes , the information rates of spike trains generated by balanced synaptic currents in our models are similar to those generated by excitation alone or by balanced excitatory and inhibitory conductances . Thus , balanced synaptic currents increase coding efficiency ( bits/spike ) rather than the information rate ( bits/s ) . By reducing energy consumption and increasing coding efficiency , approximately balanced synaptic currents increase the energy efficiency ( bits/ATP molecule ) of spike trains compared to the other synaptic regimes we modeled . Our conclusions are based upon comparisons among single compartment models that incorporated a well-established model of synaptic input that accounts for the Poisson distribution of spikes in cortical neurons [20] . The model assumes that large numbers of weak synapses are activated individually by afferent spikes that , because they come from a large population of neurons firing with Poisson statistics , are largely uncorrelated [28] . Excitatory and inhibitory synaptic inputs to our models were uncorrelated , noise free , and had identical synaptic time constants . However , within cortical networks excitation and inhibition are continuously synchronized and correlated in strength [29] . Even small differences in the timing of excitation and inhibition can modulate neuronal integration time , forming a selective gate for signal transients that affects information processing [30] . The addition of noise to the time-varying sub-threshold synaptic input of a spiking neuron can increase the regularity ( periodicity ) of spiking ( stochastic resonance ) [31] or , in the absence of time-varying input , additive noise can lead to patterned firing ( coherence resonance ) [32] . In the absence of these effects , the addition of noise to synaptic currents will degrade the quality of the input signal ( SNR ) , decreasing the information rate through an increase in noise entropy [33] . However , noise in the inhibitory and excitatory conductances will be multiplicative rather than additive with consequences for post-synaptic firing rates , information coding and metabolic efficiency [34]–[36] . Post-synaptic inhibitory conductance changes can be phasic or tonic [37]; phasic inhibition supports rhythmic activity in neuronal networks , such as the theta or the gamma oscillations [38] , [39] , whereas tonic inhibition increases conductance affecting signal integration . These specific characteristics have consequences for their effects on neuronal gain control . For example , blockage of tonic inhibition can shift the input-output relationship of cerebellar granule cells to the left ( subtractive gain control ) depending upon the temporal properties of the excitatory conductance [40] , [41] . Random trains of excitatory conductance cause a divisive as well as a subtractive modulation of gain [42] . Although our models encompass a limited set of excitatory and tonic inhibitory input properties that capture qualitatively similar modulation of neuronal gain to that observed empirically , cortical circuits incorporate numerous other combinations of phasic/tonic inhibition and static/random trains of excitation that can modulate gain and affect information processing . We use the simplest possible model of synaptic integration , an electrotonically compact compartment in which synaptic inputs directly drive a membrane containing the minimum set of voltage-gated conductances [43] . Consequently , our models do not account for the complex structure of pyramidal neurons [1] , [3] and the spatial distribution of excitatory and inhibitory inputs [44]–[46]; excitatory inputs are formed mainly on dendritic spines , whereas inhibitory synapses are located primarily on dendritic shafts , the soma and the axon initial segment [47] . Synaptic inputs are shaped and filtered by both passive membrane properties and active conductances in pyramidal neurons [48] that will affect both information processing and energy consumption . The voltage-gated ion channel properties in our models are taken from the squid giant axon because detailed kinetic models exist for these voltage-gated channels [49] . However , the squid action potential is profligate in its energy usage , consuming ∼17-fold more energy than some vertebrate action potentials [27] , suggesting that channels with different kinetics will reduce energy consumption [17] . Our calculations of energy consumption also do not incorporate the presynaptic cost of generating the synaptic conductances . Inhibitory neurons typically have higher firing rates and form more , stronger synaptic connections than excitatory neurons [50] . However , in the cortex , inhibitory neurons are smaller and less numerous than excitatory neurons [2] . This suggests that the pre-synaptic cost of generating inhibitory conductances is lower than generating excitatory conductances and , indeed , cortical energy budgets have ignored the cost of inhibition entirely [16] , [27] . Yet because our analysis is basic , it reveals some biophysical principles of efficient coding . In our model , balanced inhibitory and excitatory currents increase coding efficiency by reducing the number of action potentials and increasing their spike timing precision in the face of channel noise . This sparsening of the output spike train is due to the strong inhibitory conductance needed to generate a current that balances the excitatory current . Sparser codes translate into fewer spikes or the activation of fewer neurons in a network , reducing redundancy [51] . Our work shows that such temporal sparseness [52] is produced by an increase in inhibition that makes the neuron more efficient by increasing the information ( bits ) per spike . A reduction in spike rate also tends to increase the information per spike because spikes become more surprising [24] . Increased spike timing precision is a consequence of a faster membrane time constant and larger changes in conductance ratios creating a faster-rising voltage slope , which again increases the bits per spike . Neurons may reach high firing rates , thereby incurring a heavy metabolic cost , but they can do so only momentarily . Thus , our model demonstrates that inhibition can improve efficiency by facilitating efficient sparser codes by acting on fundamental determinants of coding efficiency . By increasing the information per spike and reducing the spike rate balanced synaptic currents maximize information rate within a limited energy budget . This is particularly important when considered in the context of cortical energy budgets , which limit average firing rates to ∼7 Hz [27] in rat grey matter and probably to even lower rates in humans . Fewer , more informative action potentials can save energy not just in a single neuron but throughout the cortical network [18] , [53] , by ensuring that synapses activate only to transmit signals from more informative spikes , thereby increasing their efficiency with which they are used . A single cortical neuron makes recurrent excitatory synaptic connections with many other cortical neurons , of which about 85% of the synapses are with other excitatory neurons [2] , [3] , [55] , [56] . Despite these synaptic connections being weak [1] , spiking activity can propagate through cortical networks [57] . Indeed , even a single additional spike can lead to a large number of extra spikes in downstream neurons [58] . Thus , even small changes in spike rate can inflate energy costs by evoking additional spikes in post-synaptic neurons . The role of balanced synaptic currents appears to be to allow cortical neurons to process information with low numbers of precise spikes . This is only possible if neurons have fast membrane time constants , sit close to the spike initiation threshold , and depolarize rapidly to conductance changes to produce spikes . These features inflate energy costs suggesting that a low cost resting state that is separated from a high cost ‘active’ state would be advantageous . It seems likely that the cortex has been under considerable pressure to reduce energy consumption whilst retaining the ability to respond rapidly and precisely . Balanced inhibition/excitation appears to be an answer to this problem . When not in ‘active’ use , cortical neurons can sit far from rest with slow membrane time constants incurring relatively low energy costs . When active the balanced synaptic currents depolarize and speed up the cortical neurons allowing them to respond rapidly to synaptic inputs . Thus , balanced synaptic currents effectively uncouple resting and active states in terms of energy use , saving energy when neurons are at rest . We have made a basic general model that reveals that current balanced excitation and inhibition can increase coding efficiency , improving the statistics of spike trains by increasing signal entropy and reducing noise entropy . Energy efficiency also improves due to a reduction in spike rate . This suggests that despite their extra cost , inhibitory synapses will increase the energy efficiency of circuits performing a wide variety of functions by making spikes more informative .
We simulated single compartment models containing stochastic voltage-gated ion channels , the properties of which were based on the original Hodgkin-Huxley model of a squid axon [49] , [59] . The model contained transient voltage-gated Na+ channels and delayed rectifier voltage-gated K+ channels along with a non-probabilistic voltage independent leak conductance . The dynamics of the membrane potential was governed by the following current balance equation: ( 1 ) where Cm is the membrane capacitance , gNa , gK and gLeak are the conductances of the Na+ , K+ and leak channels , respectively . Ej are the reversal potentials of these conductances , where . The variables m , h and n follow first order kinetics of the form , where is the steady-state activation or inactivation function and is the voltage-dependent time constant . The single compartment model is driven either by an excitatory conductance , , or in addition to an inhibitory conductance , . The exact forms of conductance fluctuation that give rise to the synaptic currents are described in the next section . In our simulations the synaptic conductances were modeled to be noise-free . We model the source of the synaptic conductance as a large number of weak synaptic inputs , each driven randomly and independently , as if by spikes from one unique neuron . This diffusion approximation [20] delivers a white noise synaptic current when the post-synaptic response is instantaneous , and pink noise when the post-synaptic response lasts for a finite time . For the diffusion approximation , we used the conductance model of Destexhe et al . [20] and define the synaptic conductances as , ( 2 ) where is the time-dependent excitatory conductance , is the time constant that defines the decay time of the synaptic activation in response to Poisson distributed spike trains , and is the diffusion coefficient of the noise process while is a zero mean and unit standard deviation Gaussian noise process . was set to 0 mV and was fixed at 3 . 3 ms . The inhibitory conductance trace was generated by an identical yet independent differential equation , differing only in the choice of which was set to −75 mV . The conductances were modeled as an Ornstein-Uhlenbeck ( OU ) process . The OU process is a model for a large number of randomly activated synaptic inputs impinging on the single compartment , where each input is simply approximated using a single exponentially decaying conductance . The conductances generated using an OU process have approximately a Gaussian distribution with a Lorentzian power spectrum . Because of this Gaussian distribution , the differential equation can be written as a difference equation , which is independent of step size Δ , ( 3 ) is the amplitude of the noise such that , ( 4 ) The mean synaptic conductance μ ( in Siemens ) , depends upon the rate , R Hz , the unitary synaptic event amplitude , A Siemens , and the exponential decay time constant τ ( in seconds ) of synaptic events ( 5 ) the standard deviation of synaptic conductance , σ , is given by ( 6 ) and the conductance contrast , is ( 7 ) Note that the stimulus contrast only depends on the event rate , R , and the decay time constant , τ , which in this study is fixed . Thus , when we increase the conductance contrast we are decreasing the event rate ( i . e . reducing the frequency with which afferent spikes activate synapses ) , and when we increase the mean conductance at constant contrast we are increasing event amplitude at constant rate . The stimulus was presented for 1 second and each set of simulations consisted of 60 such trials . All Gaussian random numbers were generated using the Marsaglia's Ziggurat algorithm; uniform random numbers were generated using Mersenne Twister algorithm . Deterministic equations were integrated using the Euler-algorithm while stochastic differential equations were integrated using the Euler-Maruyama method; both with a step size of 10 µs . Parameter values are given in Table S1 . Markov state transitions for the voltage-gated ion channels are modeled after the channel noise formulation in Refs . [49] , [60] . We used the “direct method” to measure the entropy of the responses [61] , [62] , which compares different spike trains without reference to the stimulus parameters . The total entropy sets the information capacity for the spike train . The noise entropy measures the variability of the spike train across trials . These quantities were dependent upon the temporal resolution with which the spikes were sampled , ( 1 ms ) , and the size of time window , T . We presented either a different conductance trace in each subsequent trial ( unfrozen noise ) to calculate the total entropy , or the same conductance trace in each subsequent trial ( frozen noise ) to calculate the noise entropy . We divided the spike train to form K-letter words ( K = 2 , 4 , 6 , 8 , 12 , 16 , 24 , 32 , 48 or 64 ) , where . We used the responses from the unfrozen noise presentations ( 60 trials each of 1 second ) to estimate the probability of occurrence of particular word , . The total entropy was estimated as , ( 8 ) We estimated the probability distribution of each word at the beginning of each work at time t to obtain the conditional probability . Entropy estimates were then calculated from these distributions and the average of the distributions at the different starting times t was computed to give the noise entropy ( 60 trials each of 1 second ) as , ( 9 ) where , indicates average over time . The mutual information was then computed as , ( 10 ) The total entropy and the conditional noise entropy diverge in the limit , their difference converges to the true finite information rate in this limit 61 , 62] . Therefore , we used bias correction methods to reduce the effect of sampling errors [63] . Using , we varied the spike trains to form words of different lengths . Using these entropy estimates , we extrapolated to infinite word length from the four most linear values of the plot of entropy and the inverse word length . The energy consumption of each compartmental model is determined by the number of ATP molecules expended per second , averaged over 60 trials of 1 second each . The Na+/K+ pump hydrolyses one ATP molecule for three Na+ ions extruded and two K+ ions imported [26] , [64] . Assuming that the two main charge carriers in a cell are due to Na+ and K+ we divided the excitatory , inhibitory and leak conductances into separate pools of Na+/K+ permeable conductances . We then determined the total K+ permeable current and added it to the delayed rectifier K+ current . We computed the number of K+ ions by integrating the area under the total K+ current curve for the duration of stimulus presentation . Finally , we calculated the number of ATP molecules used by multiplying the total K+ charge by , where is Avogadro's constant and F is Faraday's constant . Pre-synaptic costs ( transmitting an AP to the pre-synaptic terminal , transmitter release and recycling ) are not included in our analysis . The presynaptic costs of calcium entry and transmitter release and recycling are approximately one fifth the cost of post-synaptic current [16] , [65] . | The adult human brain consumes more than 20% of the resting metabolism . With ∼19–23 billion neurons , the cerebral cortex consumes much of this energy , mainly to restore ion gradients across membranes for electrical signaling . Even small increases in the average spike rate of cortical neurons could cause the cortex to exceed the energy available for the whole brain . Consequently , the cortex is likely to be under considerable selective pressure to reduce spike rates but , given its important roles in behavior , to maintain information processing . Numerous experimental studies have shown that excitatory and inhibitory synaptic currents are balanced in cortical neurons . Could this feature of cortical neurons contribute to their efficiency ? We tested this by making comparisons among computational models with different amounts of inhibition and excitation: excitation only , equal excitation and inhibition ( balanced synaptic conductances ) , and more inhibition than excitation ( balanced synaptic currents ) . Our simulations show that computational models with balanced synaptic currents have similar information rates to the other regimes but achieve this with fewer , more informative spikes that consume less energy . Therefore , in comparison to other synaptic regimes , balanced synaptic currents have the highest coding efficiency and the highest energy efficiency . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Balanced Excitatory and Inhibitory Synaptic Currents Promote Efficient Coding and Metabolic Efficiency |
The genome-scale models of metabolic networks have been broadly applied in phenotype prediction , evolutionary reconstruction , community functional analysis , and metabolic engineering . Despite the development of tools that support individual steps along the modeling procedure , it is still difficult to associate mathematical simulation results with the annotation and biological interpretation of metabolic models . In order to solve this problem , here we developed a Portable System for the Analysis of Metabolic Models ( PSAMM ) , a new open-source software package that supports the integration of heterogeneous metadata in model annotations and provides a user-friendly interface for the analysis of metabolic models . PSAMM is independent of paid software environments like MATLAB , and all its dependencies are freely available for academic users . Compared to existing tools , PSAMM significantly reduced the running time of constraint-based analysis and enabled flexible settings of simulation parameters using simple one-line commands . The integration of heterogeneous , model-specific annotation information in PSAMM is achieved with a novel format of YAML-based model representation , which has several advantages , such as providing a modular organization of model components and simulation settings , enabling model version tracking , and permitting the integration of multiple simulation problems . PSAMM also includes a number of quality checking procedures to examine stoichiometric balance and to identify blocked reactions . Applying PSAMM to 57 models collected from current literature , we demonstrated how the software can be used for managing and simulating metabolic models . We identified a number of common inconsistencies in existing models and constructed an updated model repository to document the resolution of these inconsistencies .
The GEnome-scale Models ( GEMs ) of metabolic networks have broad applications in biological research and engineering [1] . Models have been developed for organisms of all three kingdoms of life [2–5] and have been used to simulate a wide variety of metabolic processes , such as photo- and chemo-autotrophic carbon fixation [6 , 7] , fermentation [8] , and the production of specific organic compounds [9] . GEMs can be applied in theoretical research to predict gene essentiality [10 , 11] , simulate the thermo-tolerance of bacterial strains [12] , and study the structural and functional evolution of metabolic networks [13] . They can also be used in practical studies to identify drug targets [14 , 15] , illustrate the mechanism of human diseases [16] , and to optimize the production of compounds of industrial significance [17–19] . By connecting genome annotations with the mathematical simulation of reaction networks , GEMs are particularly applicable for integrating heterogeneous datasets from high-throughput studies [20] , such as the profiles of transcriptional regulation [21 , 22] and measurement of carbon isotope labeling [23] . Specialized MATLAB toolboxes have been released over the past two decades to support the mathematical simulations of metabolic networks . The COBRA Toolbox is a collection of widely used , open source tools . It includes diverse implementations of constraint-based modeling algorithms and has attracted a large number of user contributions from the modeling community [24] . This toolbox , however , is restricted to the MATLAB environment and requires users to maintain paid licenses from MathWorks Inc . The COBRApy software is a more recent implementation of the COBRA Toolbox functions using the Python programming language and Jython , a Java implementation of Python [25] . Like the COBRA Toolbox , COBRApy is released as a toolbox rather than a software package . Therefore , knowledge about the Python programming language is required for users to efficiently set up operations in COBRApy . Other tools have been developed to support the annotation and visualization of metabolic networks . ModelSEED is a web-based platform that supports automated reconstruction of metabolic models from genome annotations [26 , 27] . It links protein functions with an internal reaction database and is associated with the SEED functional annotation database and the RAST genome annotation pipeline [28] . In contrast to the COBRA Toolbox , this platform is focused on the reconstruction instead of the mathematical simulation of GEMs . The automated pipeline of ModelSEED permits direct construction of a draft model from genome annotation . However , the draft model still requires extensive manual curations , and manually editing the draft models is not currently supported [27] . The RAVEN Toolbox supports semi-automatic reconstruction and visualization of genome-scale models [29] . It uses information from the KEGG database [30] and , similar to the COBRA Toolbox , can only be applied under the MATLAB environment . Finally , Pathway Tools is another software package that supports pathway annotation and visualization [31] . It uses pathway information from the MetaCyc database [32] and has recently been extended to include functions for flux balance analysis . Unlike the open source software mentioned above , Pathway Tools is released under a restricted license agreement , but it is free to use for academic users . Despite the technological advances in supporting individual steps of network reconstruction and mathematical simulation , challenges remain in maintaining the quality of metabolic reconstructions and in associating the representation of mathematical problems with the annotation and biological interpretation of metabolic models . Human intervention and iterative manual examinations are required in various steps , such as defining biomass functions , assigning reaction directionality , establishing boundary conditions , and filling novel reaction gaps [26 , 33] . Frequently , these processes are carried out either manually or using customized tools that were developed by individual model curators . However , due to the absence of standardized procedures for quality checking and version tracking of iterative model annotations , the curation and manual editing of metabolic models are prone to introducing inconsistencies [34] . These inconsistencies may lead to false predictions of model viability , and as demonstrated in a recent study , may hinder the evaluation of new modeling approaches [35–37] . Another problem that prevents the effective curation and consistency checking of GEMs is the disassociations between mathematical and biological representations of model metadata . Despite a broad adoption of the SBML format , it is not designed to incorporate detailed annotations of the genes , reactions , compounds , and pathways . Two strategies have been implemented to help address this limitation . The first is to use user-defined , model-specific or toolbox-specific tags in SBML files ( e . g . the <notes> tag in COBRA-compliant SBML format ) . These tags are not part of the standard SBML specifications and are not recognized by the standardized SBML parsers ( e . g . libSBML [38] ) . The second strategy employs tables of annotation data . These tables are customized to represent specific models , but they are disconnected from the simulation problems defined in SBML files , and an automated parsing of these tables is impossible due to the lack of convention in table organizations . In either case , the mathematical representation of simulation problems is separated from the biological annotation of model components . As a result , the consistency between genome annotation and the definition of simulation settings have to be maintained by individual model curators . This again is prone to introducing misrepresentations in GEMs . An additional disadvantage of the SBML format is the lack of modularity in network representations . The SBML definition of GEMs is composed of two major lists: listOfSpecies , which defines all of the metabolites in a GEM; and listOfReactions , which defines all of the reactions and their simulation settings in a single modeling state , including flux bounds , flux values , and objective coefficients . Since the definition of simulation settings interleaves with the definition of static reaction features , SBML does not support reusing the static model definitions . Anytime a new simulation is released on the same base model , information about the static features of metabolites and reactions will need to be duplicated in a new SBML file . Hence , the SBML format is not compatible with a modular organization of modeling components because it mixes the static model properties with the dynamic simulation settings . Here , we present a Portable System for the Analysis of Metabolic Models ( PSAMM ) , an open source software package that was implemented to support the iterative curation of GEMs by connecting model annotations with mathematical simulations . A novel model format was developed in PSAMM using the YAML language to integrate detailed annotations of model components and to provide flexibilities in both the data format ( i . e . support direct referencing of annotation tables or the YAML-based list format ) and the data content ( i . e . support parsing of both standardized and model-specific data fields ) . The new YAML format assembles model definition and simulation settings into independent and reusable modules ( S1 Text ) . By reducing the amount of characters used for data structure definition , it streamlines the content of the real data and enables tracking and management of GEMs with conventional line-based version control systems like Git [39] . Applying PSAMM , we constructed an updated repository of 57 published models using the modular representation of the YAML format , and we further demonstrated the importance of model formatting and consistency checking through identifying and correcting potential inconsistencies in existing GEMs .
PSAMM was developed as an integrated software package that supports calling simulation functions using simple one-line commands . This distinguishes PSAMM from existing metabolic modeling toolboxes , which require the users to go through individual steps of model loading , linear programming solver selection , model optimization , and simulation results mapping . An example of PSAMM commands is given ( S1 Fig ) , in which a thermodynamically constrained flux balance analysis ( tFBA ) was performed through a simple one-line command . The result of the analysis contains metadata about the model name , model version , objective function , reaction fluxes , and reaction equations . In contrast , it takes three steps to set up a tFBA problem in COBRA , and the output is presented in a complex data structure that is dissociated from the biological interpretations of the model . The PSAMM implementations were also much more efficient than the COBRA toolbox functions . While it took only 7 . 15 seconds to solve a tFBA problem in PSAMM , the same process on the same model took 194 . 73 seconds in COBRA . A comparison of the functions in the PSAMM software package with the functions in the COBRA and RAVEN toolboxes is provided in the S1 Table . Besides the availability of diverse functions and implementations of model import/export , model checking , and constraint-based analysis , a unique feature of PSAMM is the availability of easily accessible help information that assists the users in selecting program functions and parameter values . This feature is enabled because PSAMM integrates the various functions under a common framework of two universal programs , psamm-model and psamm-import . The help information can be accessed in PSAMM by adding the -h or --help option following any program functions listed in the S1 Table . In contrast , RAVEN and COBRA do not provide an integrated interface to the available functions but instead require the users to know the names of individual functions in order to retrieve help information . Moreover , to systematically evaluate the efficiency of PSAMM implementations , we recorded the running time of common operations in PSAMM and COBRA on the Escherichia coli model iJO1366 [40] ( Materials and Methods ) . The results suggested that PSAMM is overall much more efficient than COBRA ( Fig 1 ) . For the computationally intensive functions like flux variable analysis ( FVA ) , robustness analysis ( Robustness ) , and thermodynamically constrained flux balance analysis ( tFBA ) , PSAMM ranges from 9 to 25 times faster than implementations in the COBRA toolbox . For the FBA function , PSAMM appears to have an overhead of about 3 seconds on top of the time required for the problem-solving step ( Fig 1 inset ) . This overhead is for the reading of YAML model files , which is also included in calculating the running times of FVA , Robustness , and tFBA , but solving FBA problem in PSAMM ( 0 . 03 seconds ) is slightly more efficient than the same step in COBRA ( 0 . 08 seconds ) . The running time for RAVEN was not plotted in Fig 1 because many of the functions are unavailable in the RAVEN toolbox ( S1 Table ) . A detailed record of the running times for available functions in COBRA , RAVEN , and PSAMM is provided in the S2 Table . Overall , PSAMM is more efficient than the COBRA and RAVEN toolboxes in carrying out constraint-based analysis of GEMs . The PSAMM software package includes five main components: user interface , model input/output , internal model representations , model checking/simulation , and linear programming utilities ( Fig 2 ) . These components are interconnected with one another to form the internal workflow of model importing and model optimizations in PSAMM . Below , we provided a detailed description of these functions . A new model format was implemented in PSAMM based on YAML , a serialization language that is designed to support both computational parsing and human readability of complex data structures [50] . The YAML format has several new features that distinguish it from the SBML format of constraint-based metabolic models . First , while SBML strictly couples the model definition with a single simulation condition , YAML allows the users to freely combine model definitions with simulation setups in a modular form . Hence , a model component can be reused in multiple simulation setups simply by using the include function ( Fig 3 and S1 Text ) . In contrast , to publish the simulation results of multiple modeling conditions in the SBML format , one has to build duplicated instances of the entire model definition to introduce variability in simulation setups . Second , while SBML is suboptimal in working with the widely-used version control systems like Git [51] , the YAML format is fully compatible and supports tracking of model edits with these systems . This is because YAML , by applying line breaks and whitespaces as meaningful structural information , can be tracked in a way similar to programming scripts using the longest common sequence ( LCS ) algorithm that is commonly used by Git [52] . In contrast , SBML does not have a convention of breaking lines , and therefore the LCS line-based version tracking will have difficulties in pinpointing the exact location of changes in the SBML files . Despite the development of new algorithms to support the comparison of SBML model files , the efficient and reliable tracking of changes in the SBML format is still challenging [34 , 51] . Third , while SBML relies on the verbose representation of markup tags , the YAML format minimizes the amount of characters used for marking up the structure of the data to streamline the integration of complex data ( Fig 3 and S1 Text ) . Finally , while the parsing of SBML files requires knowledge of predefined markup tags to capture specific data fields , the parsing of YAML files can be achieved by the recognition of line-based data structure rather than relying on the identification of specific markup tags . For example , some features like the compound formula and reaction EC numbers are not an integral part of standardized SBML specifications [41] , so they will be invisible to standard SBML parsers ( e . g . libSBML [38] or JSBML [53] ) even when they are included in a SBML file with user-defined markup tags . However , the same information can be easily represented in YAML files . A standard YAML parser will have no problem recognizing user-defined , non-standard data domains and subsequently integrate it into the presentation of metabolic models . Overall , the YAML format is an optimized solution for connecting manual curation with the mathematical simulation of GEMs . The PSAMM YAML-based representation of GEMs includes the model name , biomass function , compound/reaction database ( s ) , model reaction lists , reaction flux limits , and growth media . A flexible infrastructure was designed that permits the integration of model definitions either within a centralized file ( e . g . model . yaml ) or in additional files using the include function ( Fig 3a ) . In an example shown in Fig 3a , the reaction database was divided into multiple files based on pathway classification , and both the YAML format ( e . g . glycolysis . yaml and biomass . yaml ) and the tab-delimited table format ( e . g . tca-cycle . tsv ) were illustrated . Similarly , the compound annotations can also be represented as YAML files ( e . g . compound . yaml ) or tab-delimited tables . The reaction and compound files can be used to define a broader database of all possible reactions with additional entries beyond the scope of a given model . For example , when running the gapfill or fastgapfill functions , users may include a broader reaction database that contains enzymatic reactions that can be potentially applied to close network gaps . In such cases , the model file ( e . g . model_rxn . tsv ) is used to identify a subset of reactions that are included in the model definition . The other reactions in the broader database will be ignored in model simulations and will only be probed in gapfill or fastgapfill to predict potential gap filling reactions . Finally , to define the constraints for metabolic simulations , the media file ( e . g . media . yaml ) is used to identify the exchange reactions , and the limits file ( e . g . limits . yaml ) is used to set the flux limits for internal reactions . The limits file is defined only when the flux bounds of an internal reaction deviate from the defaults in PSAMM , which automatically assigns a conventional boundary of [−x , x] for reversible reactions and [0 , x] for irreversible reactions , where the value x is assigned using an option named default_flux_limit in the model . yaml file . When default_flux_limit is not assigned , PSAMM will use a default value of x = 1000 to define reaction bounds ( S1 Text ) . This feature permits further streamlining of the YAML format to reveal model specific simulation setups . Although PSAMM supports references to model files with user-defined file names that are saved in distributed locations in the file system , it is recommended that users save all files of an individual model within a dedicated directory to facilitate model organization and maintenance . The users may also choose to combine all the model information into a single YAML file by replacing the include functions with the actual content of the data files . However , the include functions are recommended to maintain modularity of model definition and simulation settings in the YAML representation . Using the YAML format , annotations of compound and reaction features can be further divided into multiple files that represent logical divisions of cellular compartments or metabolic pathways . Moreover , The annotations can be represented in either a YAML format or a tab-delimited table format . These flexibilities can enable the manual curation of model files and make the model components portable among different simulation conditions . The line-based text representation of YAML files also makes it compatible with broadly-used version control systems like Git , and the YAML format is especially efficient at tracking changes in large reaction equations , such as the biomass function and the protein , RNA , and lipid synthesis functions ( Fig 3b ) . In contrast , the Excel format is not compatible with version control due to the binary form , and the lack of a standard model definition in Excel files also detracts from its suitability as a GEM format . The SBML files , in spite of a well-defined data structure , do not use line breaks or white spaces as a part of the data structure and hence are suboptimal in working with Git [34] . For example , each list item in the SBML file is frequently represented as a single line that contains multiple data entries describing this item . This will render the version control in Git ineffective because changes on any one entry will cause the version tracking system to highlight an entire line of multiple data entries , and it is difficult for a user to pinpoint the exact location of the change within a long line of data . Moreover , since the markup tags are included as a part of the data presentation , changes in the markup tag definitions during the periodical updates of the SBML specifications will show up in the tracking even though there is no change in the model data . To test the application of PSAMM in supporting model annotations , we applied the software to 57 published GEMs collected from current literature or open-source model releases ( S3 Table ) . These included the RECON2 . 04 , which was released in the format of a MATLAB data file . All models were first converted to the PSAMM-specific YAML format and then analyzed individually using the stoichiometric and flux checking functions . FBA simulations were also performed using designated biomass function to verify if the published descriptions of model viability can be replicated . When inconsistencies occurred , the causes were identified by manually inspecting the corresponding models . Overall , PSAMM revealed several types of inconsistencies among existing models . These included inconsistencies in the syntax of SBML files , the annotation contents of Excel tables , the stoichiometric balance of metabolic reactions , and the connectivity of metabolic reactions . Below , we provide an overview of the existing inconsistencies identified by PSAMM .
The reconstruction and analysis of GEMs have broad applications in understanding genotype-phenotype connections , studying evolutionary processes , and predicting the outcomes of metabolic engineering [13 , 81] . Despite the development of tools that support individual steps along the modeling procedure [24 , 26 , 29] , it is still challenging to associate mathematical simulation with the biological interpretation of GEMs . Often times , model files need to be curated iteratively to set up exchange reactions , assign reaction directionality , define biomass functions , and fill reaction gaps , and the results of mathematical simulations need to be mapped to the annotated data of model definitions . However , due to the absence of integrated data representations and the lack of standardized model checking procedures , the curation of GEMs are prone to introducing inconsistencies , which in turn may lead to errors in modeling and in interpreting the biological meaning of modeling outcomes . The PSAMM software package was developed to solve the above challenges by integrating the annotation of metabolic pathways with the consistency checking and the constraint-based analyses of mathematical models ( Fig 2 ) . It offers a novel YAML format of model representation that integrates the heterogeneous annotation of model components with the formulation of mathematical simulation problems . PSAMM takes advantage of several useful features commonly supported in the Python programming language: ( 1 ) it is highly portable and can be easily installed using the pip package management system; ( 2 ) it is open source and does not rely on paid software like MATLAB; ( 3 ) it has extensive library support that permits both matrix-based operations and dictionary-based feature representations; ( 4 ) it is an integrated package that supports calling simulation functions using simple one-line commands ( S1 Fig ) and permits adjustments to modeling parameters using command options ( S1 Table ) . PSAMM supports the import and export of diverse model formats and is applicable for the curation of draft models generated from model reconstruction pipelines like ModelSEED [26] , RAVEN [29] , and Pathway Tools [31] . It provides extensive help information ( e . g . through the command-line -h or --help options ) and an online tutorial that demonstrates the main functions and workflow of PSAMM ( https://psamm . readthedocs . org/en/latest/tutorial . html ) . Additionally , the PSAMM functionality is easily expandable by the user community through an Application Programming Interface ( API ) . PSAMM integrates heterogeneous metadata in model annotations with the simulation of metabolic networks , so that users can perform model curation and mathematical simulations under the integrated framework of a unified software package . While COBRA and RAVEN are released as toolboxes and are dependent on the MATLAB software , PSAMM is an independent software package that is freely available for all academic users . Calling the simulation functions is largely simplified in PSAMM because the individual steps of model reading , solver selection , model simulation , and simulation results mapping are integrated into a single one-line command . In contrast , setting up a simulation problem in COBRA would require calling multiple functions in the MATLAB interface ( S1 Fig ) . Moreover , PSAMM allows users to document their modifications to the model definition as well as the simulation conditions during model curations . In contrast , changes made through the MATLAB interface will be lost when the computing session is closed . Even if users write out their changed models into an SBML file , it will be difficult for the users to track the exact location of the changes due to the incompatibility of SBML files with line-based version control systems [51 , 82] . Finally , PSAMM is configured to perform with high efficiency a number of computationally intensive simulations , such as tFBA , FVA , robustness analysis and random minimal networks ( Fig 1 and S2 Table ) . As an example , we recorded the running time of robustness analysis on iJO1366 , which is one of the larger models in the collection we analyzed . Using a single computing thread , the median runtime of robustness analysis with 1000 steps on the EX_o2 ( e ) exchange reaction using Cplex solver was 4 . 07 seconds in PSAMM ( including the reading of model files ) , whereas COBRA takes 100 . 7 seconds under the same settings for just solving the simulation problem . This is because PSAMM modifies the simulation problem directly through an optimized LP utility interface ( Fig 2 ) , while COBRA requires the simulation problem to be redefined for every step of the robustness analysis . The PSAMM YAML-based model representation provides additional support to model curation by offering a human-readable interface for the organization of heterogeneous data ( Fig 3 and S1 Text ) . The YAML format incorporates information about both model annotations ( e . g . in the reactions , compounds , and model definitions ) and model simulations ( e . g . in the biomass , media , and limits definitions ) , and it supports a flexible interface for storing annotation information either within a centralized model file or as independent feature files . The YAML format streamlines model representation by minimizing the use of markup tags . The use of line breaks and whitespaces as markups of data structure not only enables the parsing and association of user-defined , model-specific model annotations , but also enhances the compatibility of YAML with line-based version control systems . This compatibility with version tracking is especially useful in collaborative projects that involve multiple curators working on the same model since modifications made by different curators can be documented in parallel and later reconciled and merged into a new model release that is more comprehensive than what could be achieved by a single curator . Additionally , YAML supports the modular representation of diverse model components and the free combination of model definitions with diverse simulation conditions ( Fig 5 ) . Hence , instead of being restricted to a single simulation condition , YAML provides additional flexibilities for the documentation of multiple modeling conditions in a single model release . Applying PSAMM to existing GEMs in public literature , a number of inconsistencies were identified . These included inconsistencies in the formatting of model files ( e . g . SBML or Excel files ) , the definition of model stoichiometry , and the presence of blocked reactions . In general , the representation of SBML and Excel files varies among different GEMs both in the formatting of model annotations and in the completeness of information provided . For example , only a small fraction of the collected SBML models included complete metadata regarding protein-coding genes , functional annotations , Enzyme Commission ( EC ) numbers , and subsystem classifications . This was largely due to the lack of standardized definition of such information in SBML specifications and to some extent has prevented the SBML format from integrating heterogeneous annotation data . The Excel files , although frequently used by model curators for the annotation of pathway information , also lacks standardization and data integration . These limitations have caused a number of common inconsistencies in the representation of both SBML and Excel models . Using the annotation framework of PSAMM and the integrated YAML representation , we have corrected these inconsistencies and documented the changes in an online Git repository at https://github . com/zhanglab/psamm-model-collection ( S2–S5 Text ) . To ensure that the GEMs correctly represented the function and connectivity of metabolic networks , the PSAMM stoichiometric and flux consistency checking were applied to search for unbalanced or disconnected reactions . Surprisingly , only 30% of the analyzed models were stoichiometrically balanced , and the unbalanced reactions can be attributed to the inconsistent representation of compounds in different reaction equations . The analysis of flux consistency revealed that while the central metabolic pathways contain only a small fraction of blocked reactions across all GEMs , the pathways of lipid metabolism , glycan metabolism , and cofactors and vitamins metabolism were largely variable among different GEMs and were frequently disconnected from biomass production . Interestingly , the pathway distribution of metabolic reactions varied among different GEMs ( S2 Fig ) . In the successive models of certain organisms ( e . g . in E . coli ) , it appeared that the initial reconstructions have focused on central metabolic pathways , while the later iterations have contributed to the inclusion and completion of peripheral pathways like the transport reactions and the glycan and lipid metabolism . While it is not uncommon to have blocked reactions in GEMs due to the limited knowledge in current literature about certain metabolic pathways , the analysis of flux inconsistency should be a standard step especially in interpreting the biological significance of gene or reaction deletion simulations . Since the blocked reactions do not contribute to the production of biomass , they will always be predicted as non-essential reactions . However , such prediction is not biologically meaningful because it reflects an artifact caused by the disconnection of metabolic pathways .
A list of functions ( marked in S1 Table ) was examined in the PSAMM package as well as the COBRA and RAVEN toolboxes to compare the efficiency of the different tools . The simulations were carried out on the model iJO1366 [40] with the Cplex linear programming solver for PSAMM ( version 0 . 17 ) and COBRA ( version 12 . 6 , IBM academic release for PSAMM and TOMLAB release for COBRA ) , and with the MOSEK solver for RAVEN ( version 7 . 1 . 0 . 36 ) . The specific simulation parameters were set according to the conditions specified for each function in the S2 Table . Outputs from different tools were manually examined to ensure that the same results were generated form the same simulation settings . The running time was recorded in a CentOS operating system with a single processor core allocated to each program ( the solver was not restricted in this way and was able to use up to 20 cores ) . Each simulation was carried out at least seven times , from which the median was plotted in Fig 1 with the 75th and the 25th percentile values as the upper and the lower limits of the error bar . PSAMM has the following dependencies , some of which are required and some of which are optional but limit the functionality of PSAMM in their absence . The dependencies PyYAML , xlrd , xlsxwriter , and NumPy can be automatically installed through the Python package manager pip . The linear programming solvers need to be installed by the user following instructions from corresponding releases . A list of 56 GEMs was collected from current literature and a public model collection [84] as SBML files . Additionally , the RECON2 . 04 model was downloaded in the format of a MATLAB data file from a public release at https://vmh . uni . lu/#downloadview . Compound and reaction features of nine models were downloaded in the Microsoft Excel format ( S3 Table ) . These models were used to test the application of PSAMM for model management and model curation . For constraint-based simulations , a biomass reaction was identified for each model by examining the objective coefficient defined under the SBML “kineticLaw” section . The reaction with a non-zero objective coefficient was treated as the biomass reaction . If the objective function was not specified in SBML file , the reaction identifiers were searched and the ones containing the string “biomass” were marked as biomass reactions . Information was also obtained from original publications of the collected models to identify biomass reactions and to select the main biomass reaction when more than one was present . The biomass reactions of each model were listed in S3 Table . The stoichiometric consistency check was implemented in the PSAMM masscheck function , which supports both compound-based and reaction-based checking of the stoichiometric consistency . By default , the masscheck function excludes the exchange and biomass reactions from consideration because by design , stoichiometric balance is not required in these reactions . The compound-based stoichiometric consistency checking was implemented based on Thiele et al . [49] and was formulated as: Maximize Σi zi Subject to This problem looks for compounds that failed to obtain a positive mass under optimizing conditions . By maximizing the values in z up to 1 for as many compounds as possible , it forced the mass values ( m ) to be at least 1 for the same compounds . The sum of the values in z can be used to approximate the number of compounds that were properly balanced . Although the compound-based checking was useful for identifying compounds that may cause stoichiometric inconsistency , it provides no insights into the stoichiometric balance of individual reactions . Therefore , PSAMM implements a new algorithm to directly search for unbalanced reactions . This approach is performed using an LP problem that was modified from the above stoichiometric consistency check . A mass residual variable ( r ) is introduced into every reaction and bounded by the residual bound variable ( z ) such that rj ∈[−zj;zj] . The residual bound variables are then minimized , and the reactions that carry a non-zero residual value rj ( i . e . bounded by a positive zj ) are reported as a candidate of inconsistency . The mathematical formulation of this approach is described in the following LP problem: Maximize Σj zj Subject to This approach can be applied iteratively to assist with the correction of stoichiometric inconsistencies in metabolic models . The residual variables indicate a mass that is missing from identified reactions , and the sign of the residual can be used to determine whether the left or right side of an equation has a missing mass . Additionally , PSAMM provides a “--checked” option in the masscheck function , which allows fixing certain residual values at zero when the corresponding reactions have been confirmed to be balanced , making the program able to converge on the remaining set of unbalanced reactions . When the stoichiometric consistency check was performed on the collection of models , the biomass and exchange reactions were excluded from consideration and the requirement of a positive mass was removed from compounds that were used to model photons or electrons . The flux consistency check was implemented in the PSAMM fluxcheck function . A flux inconsistent reaction is a reaction that cannot take a non-zero flux under any simulation conditions . In other words , given the stoichiometric matrix S , the vector of fluxes v , and the reaction j , if no solution to Sv = 0 exists , where vj ≠ 0 the reaction is considered to be flux inconsistent . Conversely , if a solution exists the reaction is considered to be flux consistent . A consistent model is a model that only contains consistent reactions [85] . The flux consistency check in PSAMM was implemented based on two independent approaches . The first approach was based on the definitions in [86] . This was achieved by applying a modified version of the flux variability analysis ( FVA ) on each reaction while using default constraints on other reactions . For reversible reactions , the flux is first maximized and then minimized , and for irreversible reactions , only the maximization is necessary . A significant speedup was possible in PSAMM ( S2 Table ) by avoiding regenerating new problem definitions throughout the procedure . In previous implementations ( e . g . COBRA ) the LP problem was regenerated for optimizing each reaction . However , in PSAMM the LP problem definition was instead reused . This was feasible since the LP problems in this approach are all equivalent except for using a different objective function for each reaction . Reactions that carried non-zero fluxes were considered to be flux consistent , and vice versa . Alternatively , PSAMM provides another way of checking reaction flux consistency based on the fast consistency check ( FASTCC ) algorithm , which was designed to provide a more efficient solution in evaluating reaction flux consistency [87] . When analyzing the models in the collection for flux inconsistencies , the constraints on the flux limits of the exchange reactions were removed prior to the analysis to eliminate the influences of external settings and to provide a lower estimate for the fraction of inconsistent reactions ( Fig 4 ) . | The broad application of genome-scale metabolic modeling has made it a useful technique for tackling fundamental questions in biological research and engineering . Today over 100 models have been constructed for organisms that carry out a diverse array of metabolic activities spanning all three kingdoms of life . These models , however , have been curated independently following different conventions . The maintenance of model consistency has been challenging due to the lack of consensus in model representation and the absence of integrated modeling software for associating mathematical simulations with the annotation and biological interpretation of metabolic models . To solve this problem , we developed a new software package , PSAMM , and a new model format that incorporates heterogeneous , model-specific annotation information into modular representations of model definitions and simulation settings . PSAMM provides significant advances in standardizing the workflow of model annotation and consistency checking . Compared to existing tools , PSAMM supports more flexible configurations and is more efficient in running constraint-based simulations . All functions of PSAMM are freely available for academic users and can be downloaded from a public Git repository ( https://zhanglab . github . io/psamm/ ) under the GNU General Public License . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"xenobiotic",
"metabolism",
"metabolic",
"networks",
"mathematical",
"models",
"simulation",
"and",
"modeling",
"network",
"analysis",
"exchange",
"reactions",
"lipid",
"metabolism",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
... | 2016 | PSAMM: A Portable System for the Analysis of Metabolic Models |
Despite causing considerable damage to host tissue during the onset of parasitism , nematodes establish remarkably persistent infections in both animals and plants . It is thought that an elaborate repertoire of effector proteins in nematode secretions suppresses damage-triggered immune responses of the host . However , the nature and mode of action of most immunomodulatory compounds in nematode secretions are not well understood . Here , we show that venom allergen-like proteins of plant-parasitic nematodes selectively suppress host immunity mediated by surface-localized immune receptors . Venom allergen-like proteins are uniquely conserved in secretions of all animal- and plant-parasitic nematodes studied to date , but their role during the onset of parasitism has thus far remained elusive . Knocking-down the expression of the venom allergen-like protein Gr-VAP1 severely hampered the infectivity of the potato cyst nematode Globodera rostochiensis . By contrast , heterologous expression of Gr-VAP1 and two other venom allergen-like proteins from the beet cyst nematode Heterodera schachtii in plants resulted in the loss of basal immunity to multiple unrelated pathogens . The modulation of basal immunity by ectopic venom allergen-like proteins in Arabidopsis thaliana involved extracellular protease-based host defenses and non-photochemical quenching in chloroplasts . Non-photochemical quenching regulates the initiation of the defense-related programmed cell death , the onset of which was commonly suppressed by venom allergen-like proteins from G . rostochiensis , H . schachtii , and the root-knot nematode Meloidogyne incognita . Surprisingly , these venom allergen-like proteins only affected the programmed cell death mediated by surface-localized immune receptors . Furthermore , the delivery of venom allergen-like proteins into host tissue coincides with the enzymatic breakdown of plant cell walls by migratory nematodes . We , therefore , conclude that parasitic nematodes most likely utilize venom allergen-like proteins to suppress the activation of defenses by immunogenic breakdown products in damaged host tissue .
Soil-borne plant-parasitic nematodes are major constraints on global food security , as they reduce the annual yield of food crops by approximately 10 percent [1] , [2] . This figure is a global average and may therefore be somewhat misleading . In areas where people depend on local cultivation of staple crops the effect of these microscopic roundworms can be devastating . The impact of plant-parasitic nematodes on food production provides plant breeders with a strong incentive to better exploit genetic variation in resistance to nematodes in crop cultivars . However , this requires knowledge of the mechanisms underlying the activation and suppression of plant innate immunity by plant-parasitic nematodes , an area which is currently underexplored [3] , [4] . Plants utilize pattern recognition receptors belonging to the receptor-like kinase ( RLK ) /Pelle superfamily to detect extracellular microbes or their actions in the apoplast ( i . e . the extracellular matrix; [5] , [6] ) . The recognition of immunogenic microbe- and damage-associated molecular patterns by receptor-like kinases activates intracellular immune signaling pathways , resulting in a wide range of structural and chemical defenses [7] , [8] . Several members of the RLK/Pelle superfamily in plants lack a cytoplasmic kinase domain , while they are nonetheless able to activate immune responses to pathogens ( e . g . Cf-proteins in tomato; [9]–[11] ) . The activity of these so-called receptor-like proteins requires mediation by other transmembrane proteins , or cytoplasmic membrane-associated kinases , that function as co-factors within multimeric receptor complexes [12] , [13] . At present , little evidence is available on the role of surface-localized pattern recognition receptors in immunity to parasitic nematodes in plants . Recently , we showed that the receptor-like protein Cf-2 in tomato mediates dual disease resistance by guarding a common virulence target of a nematode and a fungus [14] . Perturbations of the apoplastic papain-like cysteine protease Rcr3pim by two unrelated effectors from the leaf mold fungus Cladosporium fulvum and from the root parasitic nematode Globodera rostochiensis activate Cf-2-mediated disease resistance . The function of Rcr3pim , or any of its close homologs in tomato , has not yet been resolved [15]–[17] . Tomato plants harboring the Rcr3pim allele , but not the receptor Cf-2 , are far more susceptible to infections by G . rostochiensis than tomato plants lacking Rcr3pim [14] . Apoplastic Rcr3pim is a molecular target of the venom allergen-like protein Gr-VAP1 of G . rostochiensis , which is secreted by infective juveniles during the onset of parasitism . However , the role of this venom allergen-like protein ( VAP ) , or its interaction with Rcr3pim , in nematode virulence is not clear . Venom allergen-like proteins constitute a monophyletic clade of cysteine-rich secretory proteins within the Sperm Coating Protein/Tpx-1/Ag-5/Pr-1/Sc-7 ( SCP/TAPS ) superfamily ( [18] , [19] ) . Members of this clade show similarity to venom allergen 5 from vespid wasps , pathogenesis-related protein PR-1 from plants , brain tumor specific proteins in humans , and a wide range of other secreted proteins ( reviewed in [18] ) . Venom allergen-like proteins have been identified in all animal- and plant-parasitic nematodes studied to date [19] , [20] . They are even the most abundantly secreted proteins during the onset of parasitism of some animal-parasitic nematodes [19] , [21]–[24] . In spite of their conservation , abundance , and strong association with the onset of parasitism , little is currently known of the function of venom allergen-like protein in nematode infections in plants and in animals [19] . Sedentary plant-parasitic nematodes , such as cyst nematodes ( genera Globodera and Heterodera ) and root-knot nematodes ( genus Meloidogyne ) , deliver effectors into the apoplast and cytoplasm of host cells to induce the formation of a permanent feeding structure [25]–[27] . The permanent feeding structure is the sole source of plant nutrients for sedentary nematodes throughout their life [28] . Besides altering host cell metabolism and function , sedentary nematodes also use effectors to modulate host immunity [3] , [27] . Specific immunity to nematodes in host plants often involves a programmed cell death in or around permanent feeding structures , resulting in the developmental arrest of feeding juveniles [3] . The most advanced sedentary nematodes deliver effectors into the cytoplasm of host cells to suppress the defense-related programmed cell death mediated by intracellular immune receptors [29] , [30] . However , sedentary nematodes are extracellular parasites and their prolonged contact with surrounding host cells makes them also vulnerable to detection by surface-localized pattern recognition receptors [14] . Recent discoveries with the root-knot nematode M . incognita suggest that sedentary plant-parasitic nematodes may have adapted to this by evolving a separate set of apoplastic effectors to further control host immunity [31] , [32] . The unique conservation of venom allergen-like proteins in secretions of animal- and plant-parasitic nematodes might point to a common activity of these effector proteins in the extracellular matrix of animal and plant cells . However , the composition , structure , and function of the extracellular matrix of animal and plant cells are fundamentally different [33] , [34] . A possible exception might be that both in animals and plants nematodes encounter an innate immune system that relies on the surveillance of the extracellular matrix by surface-localized pattern recognition receptors [35] , [36] . Earlier work has demonstrated that a secreted venom allergen-like protein from the animal-parasitic nematode Necator americanus acts in vitro as an antagonistic ligand of the integrin complement receptor 3 , a pattern recognition receptor on the surface of human neutrophils [37]–[40] . This observation led us to investigate if venom allergen-like proteins of plant-parasitic nematodes similarly interfere with the functioning of surface-localized immune receptors in plants . To address this question , we first analyzed if venom allergen-like proteins are important for the onset of parasitism by silencing the expression of Gr-VAP1 in infective juveniles of G . rostochiensis . Next , we analyzed the effect of ectopic venom allergen-like proteins in transgenic plants on susceptibility to nematodes , and diverse plant pathogenic fungi , oomycetes , and bacteria . Based on the response of these plants to the immunogenic epitope flg22 from bacterial flagellin [41] , we concluded that the venom allergen-like proteins suppress basal plant defenses to biotic stresses . We further provide evidence that the breakdown of basal immunity by ectopic venom allergen-like proteins involves a plant cell wall-associated subtilisin-like serine protease , not previously linked to defense regulation in plants . Remarkably , our data also suggest that cyst nematodes exploit a trade-off mechanism between resistance to biotic and abiotic stress to ward off host defense responses .
To investigate whether venom allergen-like proteins are required for the onset of parasitism by sedentary plant-parasitic nematodes , we soaked infective juveniles of G . rostochiensis in double-stranded RNA , matching 820 base pairs of the Gr-VAP1 transcript sequence . Reverse transcription PCR on nematodes , soaked in Gr-VAP1-specific dsRNA , showed a significant reduction in Gr-VAP1 transcript levels , whereas the control treatment with dsRNA matching the NAU gene from Drosophila melanogaster did not alter Gr-VAP1 expression ( Fig . 1A ) . Next , susceptible tomato plants ( Solanum lycopersicum , cultivar Moneymaker ) were challenged with the dsRNA-treated infective juveniles , and monitored for nematode infections for 7 days post inoculation . Treatment with Gr-VAP1-specific dsRNA significantly reduced the number of nematodes inside tomato roots compared to the treatment with NAU-specific dsRNA ( Fig . 1B ) . We therefore concluded that the apoplastic venom allergen-like protein Gr-VAP1 is required for the establishment of successful infections by G . rostochiensis during the onset of parasitism . Both infective second-stage juveniles and adult males of G . rostochiensis migrate through host tissues , which causes significant damage to host cells . By contrast , intermediate juvenile stages and adult females are immobile , and thus induce little damage . To determine whether the expression of Gr-VAP1 in G . rostochiensis coincides with either migration or sedentarism in host plants , we used semi-quantitative reverse transcription PCR on nematodes isolated from infected potato roots at different time points prior to and post host invasion ( Fig . 2 ) . Gr-VAP1 was highly expressed in infective second stage juveniles during the onset of parasitism . Thereafter , the level of Gr-VAP1 expression declined in successive sedentary juvenile stages inside host roots to total absence in sedentary adult females . However , the expression of Gr-VAP1 was raised again in migratory adult males . We therefore concluded that the temporal expression of Gr-VAP1 in G . rostochiensis coincides with nematode migration inside host plants . To examine whether Gr-VAP1 affects the susceptibility of host plants to G . rostochiensis , we generated transgenic potato plants ectopically overexpressing Gr-VAP1 . Two randomly selected independent transgenic lines without any visible anomalies in shoots and roots were challenged with infective juveniles of G . rostochiensis . Six weeks after inoculation the number of adult females in plants expressing Gr-VAP1 was significantly higher than in the corresponding empty vector control plants ( Fig . 3A ) . To confirm that the altered nematode susceptibility correlates with Gr-VAP1 expression , we used a real-time quantitative reverse transcription PCR on the two potato lines expressing Gr-VAP1 . Transgenic line Gr-VAP1-A , that showed the highest nematode susceptibility , had a 7 . 9-fold higher expression of Gr-VAP1 than transgenic line Gr-VAP1-B . We therefore concluded that ectopic Gr-VAP1 enhances the susceptibility of potato plants to G . rostochiensis . Like the effector Avr2 from C . fulvum [42] , we expected Gr-VAP1 to interact with other extracellular papain-like cysteine proteases in different host plant species of G . rostochiensis . We used DCG-04 activity profiling to demonstrate that Gr-VAP1 also perturbs the active site of the apoplastic papain-like cysteine protease C14tub from potato ( S . tuberosum ) ( Fig . 3B ) . By contrast , Gr-VAP1 does not interfere with the binding of fluorescent DCG-04 to apoplastic C14lyc from tomato ( S . lycopersicum ) . For this experiment , Gr-VAP1 and C14tub/lyc were separately produced in the apoplast of agroinfiltrated leaves of Nicotiana benthamiana . Protease activity was subsequently determined by the binding of fluorescent DCG-04 to C14tub and C14lyc in the presence of Gr-VAP1 on gels of mixtures of isolated apoplastic fluids . The experiment was repeated three times , and each attempt resulted in significantly less binding of the fluorescent probe to C14tub , but not to C14lyc ( S1 Figure ) . Only C14tub from potato is under strong diversifying selection , because of which it is thought to be involved defenses [43] . We therefore concluded that the enhanced susceptibility by ectopic Gr-VAP1 in potato most likely involves the perturbation of C14tub and perhaps other apoplastic papain-like cysteine proteases . Arabidopsis thaliana is a far better model to study the molecular changes induced by venom allergen-like proteins in plants than either potato or tomato . However , G . rostochiensis is not able to establish infections in A . thaliana . We therefore continued our investigations with two homologous venom allergen-like proteins from the beet cyst nematode Heterodera schachtii , which is a parasite of A . thaliana . These two venom allergen-like proteins are formally designated as Nem-Hsc-SCP/TAPS-1A and Nem-Hsc-SCP/TAPS-2A [18] , but for the remainder of this paper they are referred to as Hs-VAP1 and Hs-VAP2 . Hs-VAP1 is 81 . 4 percent identical to Gr-VAP1 , while Hs-VAP2 shows only 34 . 8 percent sequence identity to Gr-VAP1 . In comparison , a previously reported venom allergen-like protein from the root-knot nematode Meloidogyne incognita ( hereafter named Mi-VAP1; [44] ) shows about 28 . 6% identity to Gr-VAP1 and Hs-VAP1 , while it is for 33 . 9% of its sequence identical to Hs-VAP2 ( S2 Figure ) . To investigate if ectopic venom allergen-like proteins from H . schachtii and G . rostochiensis alter the susceptibility of A . thaliana to cyst nematodes , we generated transgenic plants overexpressing Gr-VAP1 , Hs-VAP1 , and Hs-VAP2 , including their native signal peptides for secretion . We challenged two independent single insertion lines , without visible anomalies in shoots and roots , of each construct with infective juveniles of H . schachtii . Twenty-eight days after inoculation the number of females per plant in plants expressing Gr-VAP1 , Hs-VAP1 , and Hs-VAP2 was significantly higher than in the corresponding transgenic empty vector line or in the wild type Col-0 plants ( Fig . 4A ) . We therefore concluded that venom allergen-like proteins from two unrelated cyst nematodes commonly enhance the susceptibility of different plant species to nematode infections . Next , we reasoned that if the VAP-enhanced susceptibility of the transgenic Arabidopsis lines to cyst nematodes involves modulation of the innate immunity , these lines might also be more susceptible to entirely unrelated plant pathogens . To test this , we analyzed the transgenic Arabidopsis lines overexpressing Hs-VAP1 and Hs-VAP2 for their susceptibility towards Botrytis cinerea , Plectosphaerella cucumerina , a virulent and a non-virulent isolate of Phytophthora brassicae , Alternaria brassicicola , Verticillium dahliae , and Pseudomonas syringae pv . tomato ( Fig . 4B , and S3 Figure ) . The overexpression of both Hs-VAP1 and Hs-VAP2 significantly increased the severity of the grey mold symptoms caused by B . cinerea in the transgenic Arabidopsis plants ( S4A Figure ) . Similarly , both Hs-VAP1 and Hs-VAP2 significantly increased susceptibility of Arabidopsis plants to infections by P . syringae pv . tomato ( S3A Figure ) . Only Arabidopsis plants overexpressing Hs-VAP1 showed larger necrotic lesions following the inoculation with the fungus P . cucumerina ( S4B Figure ) . By contrast , the oomycete P . brassicae ( isolate CBS686 . 95 ) only caused faster developing and larger lesions on transgenic Arabidopsis expressing Hs-VAP2 ( S4C Figure ) . Surprisingly , the P . brassicae isolate HH , which is not virulent on wild type A . thaliana Col-0 , was able to colonize transgenic A . thaliana lines expressing Hs-VAP2 ( S4D Figure ) . However , neither Hs-VAP1 nor Hs-VAP2 altered the susceptibility of A . thaliana towards the fungal pathogens A . brassicicola or V . dahliae , both of which do not cause expanding lesions in Arabidopsis ecotype Col-0 . Altogether , our data suggests that ectopic Hs-VAP1 and Hs-VAP2 , albeit differently , modulate basal innate immunity of A . thaliana toward multiple , but not all , plant pathogens . Furthermore , ectopic VAPs specifically altered the susceptibility of Arabidopsis to pathogenic microbes that typically cause expanding lesions in their necrotrophic phase ( i . e . B . cinerea , P . cucumerina , P . brassicae , and P . syringae pv . tomato ) . The flagella of P . syringae pv . tomato harbor an immunogenic epitope of twenty two amino acids ( flg22 ) that is recognized as a pathogen-associated molecular pattern in Arabidopsis [45] . Prolonged exposure to flg22 elicits a persistent basal immune response in seedlings of Arabidopsis Col-0 plants , which occurs at the expense of plant growth [45] . We used this phenotype to confirm that ectopic venom allergen-like proteins undermine basal immunity in our transgenic Arabidopsis lines . As expected , treatment with flg22 significantly inhibited seedling growth in both wild-type Arabidopsis and in our transgenic lines harboring the empty expression vector ( Fig . 4C ) . By contrast , the expression of Hs-VAP1 and Hs-VAP2 in Arabidopsis seedlings largely abrogated this growth inhibition by flg22 ( Fig 4C; S5 Figure ) . Remarkably , the leaves of the transgenic plants overexpressing Hs-VAP1 and Hs-VAP2 also remained much greener as compared to the leaves of wild type Col-0 and empty vector control plants following the treatment with flg22 . As the perception of flg22 in Arabidopsis is mediated by the extracellular pattern recognition receptor FLS2 [45] , we concluded that ectopic venom allergen-like proteins most likely modulate the activation of basal immunity by surface-localized immune receptors . The apoplastic effectors Avr2 of C . fulvum and Gr-VAP1 of G . rostochiensis commonly inhibit the extracellular papain-like protease Rcr3pim in tomato [14] . Although C . fulvum is not a pathogen of Arabidopsis either , ectopic Avr2 has been shown to interact with multiple extracellular papain-like cysteine proteases of Arabidopsis [42] . To investigate if heterologous expression of Avr2 through its interactions with extracellular papain-like cysteine proteases also affects susceptibility of Arabidopsis to nematode infections , we challenged transgenic Arabidopsis plants stably overexpressing Avr2 with H . schachtii . Four weeks post inoculation the number of adult females of H . schachtii per root was significantly higher in plants overexpressing Avr2 than in the corresponding wild type Arabidopsis plants ( Fig . 5A ) . This data shows that the inhibition of extracellular papain-like cysteine proteases by promiscuous effectors from different non-adapted plant attackers ( i . e . Gr-VAP1 and Avr2 ) undermines basal immunity in Arabidopsis . To further confirm the importance of extracellular papain-like cysteine proteases in basal immunity to nematode infections , we challenged the homozygous knockout mutants pap-1 , pap-4 , and pap-5 of Arabidopsis with H . schachtii . Members of the pap gene family are the closest homologs of Rcr3pim in Arabidopsis [14] . The loss of functional pap genes in all three mutant Arabidopsis lines resulted in significantly enhanced susceptibility to H . schachtii ( Fig . 5B ) . We therefore conclude that conserved extracellular protease-based immune signaling networks most likely regulate basal immunity to plant-parasitic nematodes in multiple unrelated plants . To better understand the molecular basis of the suppression of basal immunity by venom allergen-like proteins , we analyzed the transcriptomes of Arabidopsis lines expressing Hs-VAP1 , Hs-VAP2 , and the corresponding transgenic empty vector control plants . In total , the expression of 1294 genes was significantly down-regulated , while 535 genes were significantly up-regulated in the Arabidopsis lines overexpressing either Hs-VAP1 or Hs-VAP2 ( False Discovery Rate <0 . 05 ) ( S6A Figure and S6B Figure ) . More than sixty percent of the Arabidopsis genes strongly down-regulated by ectopic Hs-VAP1 and Hs-VAP2 encode a protein that is either predicted to be extracellular or localized to the plasma membrane ( S4C Figure; S1 Table and S2 Table ) . By contrast , the predicted subcellular location of the products of the genes strongly up-regulated by either Hs-VAP1 or Hs-VAP2 are more evenly distributed over different cellular compartments ( S3 Table and S4 Figure ) . To resolve specific pathways particularly affected by the overexpression of Hs-VAP1 and Hs-VAP2 in A . thaliana , we subjected all differentially expressed genes to a KEGG pathway gene set enrichment analysis [46] , [47] . The KEGG pathway most significantly altered by both Hs-VAP1 and Hs-VAP2 in Arabidopsis is named ‘Plant-pathogen interactions' ( KO04626; S5 Table; FDR <10-12 ) . The vast majority of these Arabidopsis genes currently assigned to this pathway have been associated with innate immunity to plant pathogens [48] . We therefore concluded that the overexpression of venom allergen-like proteins in Arabidopsis particularly affects molecular components in immune signaling pathways . To further investigate the expression of specific genes associated with the loss of immunity in the transgenic Arabidopsis plants overexpressing Hs-VAP1 and Hs-VAP2 , we first focused on the most down-regulated genes ( Table 1 ) . The annotations of many of the most down-regulated genes point to an involvement of plant cell wall-associated processes such as modification ( e . g . plant invertase/pectin methylesterase inhibitor family , and glycosyl hydrolases ) , signaling ( e . g . proline-rich extension-like receptor kinases ) , and protein processing ( e . g . subtilisin-like serine proteases ) . An exceptionally strong down-regulation was observed for gene locus AT4G21630 in Hs-VAP1-overexpressing Arabidopsis plants ( Log2 fold change = −32 . 0 ) . Albeit less , AT4G21630 was also strongly down-regulated in the transgenic Arabidopsis lines overexpressing Hs-VAP2 . AT4G21630 encodes a putative plant cell wall-associated subtilase-like serine protease ( i . e . AtSBT3 . 14; [49] ) . The role of AtSBT3 . 14 in Arabidopsis is not known , but a closely related homolog from the same subtilase subfamily , named AtSBT3 . 3 , acts as an extracellular molecular switch in priming of defense responses [50] . To investigate if AtSTB3 . 14 is required for basal immunity of Arabidopsis to the cyst nematodes , we challenged a homozygous knockout mutant line with H . schachtii . Four weeks post inoculation the number of adult females was almost twice as high in the Atsbt3 . 14 knock-mutant , as compared to the corresponding Col-0 wild type Arabidopsis plants ( Fig . 6 ) . We therefore concluded that the strong down-regulation of AtSBT3 . 14 most likely contributes to the loss of basal immunity in Arabidopsis plants overexpressing Hs-VAP1 and Hs-VAP2 . Next , we focused on four of the most up-regulated transcripts in the transgenic Arabidopsis plant overexpressing Hs-VAP1 and Hs-VAP2 ( Log2 fold change>30 . 9; Table 2 ) . Two of these transcripts derive from gene loci encoding unknown proteins ( i . e . AT1G44608 and AT1G44542 ) . However , the two other transcripts are splice variants from the same and most up-regulated gene in both Hs-VAP1 and Hs-VAP2 overexpressing plants ( i . e . AT1G44575 , or NPQ4 ) . NPQ4 encodes the chlorophyll-associated Photosystem II subunit S protein ( PsbS ) , which is involved in non-photochemical quenching of excess excitation energy [51] . Recently , it was shown that the Arabidopsis knockout mutant npq4-1 lacking PsbS displays an enhanced response to flg22 [52] . The npq4-1 knockout mutant is also less attractive to herbivorous insects than the corresponding wild type Arabidopsis plants [53] . We used the npq4-1 knockout mutant to demonstrate that the lack of PsbS enhances immunity of Arabidopsis to H . schachtii ( Fig . 6 ) . We therefore conclude that the constitutively enhanced expression of NPQ4 by ectopic Hs-VAP1 and Hs-VAP2 in transgenic Arabidopsis most likely undermines their ability to mount an adequate immune response . The increase of non-photochemical quenching capacity may block the activation of singlet oxygen-dependent programmed cell death [54] , [55] . To investigate if the venom allergen-like proteins are able to suppress programmed cell death , we transiently co-expressed several inducers of cell death and nematode VAPs in leaves of Nicotiana benthamiana ( S6 Table; Fig . 7 ) . Because N . benthamiana is not a host of G . rostochiensis or H . schachtii , we also included the venom allergen-like protein Mi-VAP1 from the polyphagous M . incognita in these cell death suppression assays . Both Mi-VAP1 and Hs-VAP1 consistently suppressed the cell death induced by the Phytophthora infestans elicitin INF1 ( Fig . 7A ) . Both Mi-VAP1 and Hs-VAP1 also suppressed the cell death induced by extracellular receptor protein Cf-4 from tomato and its cognate elicitor Avr4 from C . fulvum ( Fig . 7B ) . All tested VAPs similarly suppressed the cell death induced by the extracellular receptor-like protein Cf-9 from tomato and its cognate elicitor Avr9 from C . fulvum ( Fig . 7C ) . Surprisingly , none of the venom allergen-like proteins suppressed the cell death responses induced by several cytoplasmic immune receptors and their cognate elicitors ( e . g . Rx1 , Gpa2 , R3a , Blb2 ) . To confirm that the ectopic VAPs harboring their native signal peptide for secretion are indeed secreted to the apoplast in planta , we analyzed apoplastic fluids isolated from agroinfiltrated leaf of N . benthamiana on western blots using antiserum towards an additional carboxyl terminal FLAG affinity tag on the proteins ( S7 Figure ) . Taken together , we conclude that apoplastic venom allergen-like proteins selectively suppress the activation of the programmed cell death by surface-localized immune receptors .
Since their first identification in the canine hookworm Ancylostoma caninum ( Ac-ASP; [21] ) and the root-knot nematode M . incognita ( Mi-MSP1/VAP1; [44] ) venom allergen-like proteins are thought to be crucial for the onset of parasitism of nematodes in animals and plants . However , in spite of their conservation , relative abundance in nematode secretions , and strong association with the onset of parasitism , the role of venom allergen-like proteins in host-parasite interactions has so far remained elusive [18] , [19] . To date , the only available functional data on the role of a venom allergen-like protein in secretions of parasitic nematodes point to a perturbation of a complement receptor on human immune cells [37]–[40] . Here , we demonstrate that plant-parasitic nematodes most likely deliver venom allergen-like proteins into the apoplast of host cells to suppress basal immunity mediated by surface-localized immune receptors . Several lines of evidence in our data suggest that plant-parasitic nematodes may use venom allergen-like proteins to modulate immune responses activated by tissue damage caused by migratory nematodes inside host plants . First , the expression of Gr-VAP1 in G . rostochiensis coincides with host invasion by infective juveniles and intracellular migration by adult males inside plants ( Fig . 2 ) . Both host invasion and migration by parasitic nematodes involves the enzymatic breakdown of plant cell walls , resulting in extensive damage to host tissue [56] . More importantly , Gr-VAP1 is secreted at the same time and from the same pharyngeal glands in G . rostochiensis as an elaborate set of plant cell wall-degrading enzymes [14] . The transient knockdown of these plant cell wall-degrading enzymes in infective juveniles of G . rostochiensis inhibits the onset of parasitism [56] . Similarly , the transient knockdown of Gr-VAP1 expression during host invasion significantly reduced the number of infective juveniles inside susceptible tomato plants ( Fig . 1 ) , showing that this venom allergen-like protein is also a critical factor during the onset of parasitism . Although we have not formally shown that plant-parasitic nematodes deliver VAPs into the apoplast of host cells in planta , our previous work demonstrates that this is nonetheless most likely the case [14] . More specifically , we have shown that Gr-VAP1 associates with and perturbs apoplastic Rcr3pim in tomato , and that this perturbation specifically activates nematode resistance and programmed cell death mediated by the extracellular receptor-like protein Cf-2 . Furthermore , tomato plants that harbor apoplastic Rcr3pim , but not Cf-2 , are almost twice as susceptible to nematode infections than tomato plants with allelic variants of apoplastic Rcr3 to which Gr-VAP1 is unable to bind or tomato plants that have no functional Rcr3 at all . Altogether , we conclude that venom allergen-like proteins may specifically function as apoplastic suppressors of immune responses triggered by plant cell wall fragments released by the enzymatic breakdown of plant cell walls during nematode migration inside host plants [57] , [58] . Little work has been done on the importance of surface-localized pattern recognition receptors mediating damage-triggered immunity in nematode-plant interactions . Transcriptome analyses of nematode-infected roots suggest that plant cell wall-associated legume-like lectin receptor kinases might be involved in basal immunity to H . schachtii in Arabidopsis [59] , but further research is needed to corroborate this . Interestingly , both Hs-VAP1 and Hs-VAP2 significantly down-regulate the expression of five proline-rich extensin-like receptor kinases ( i . e . At4G34440 , AtPERK5; At3G18810 , AtPERK6; At1G49270 , AtPERK7; At1G10620 , AtPERK11; S1 Table and S2 Table ) ) . These PERKs belong to a family of fifteen predicted transmembrane receptor-like kinases in Arabidopsis . The extracellular domain in PERKs shares similarity with plant cell wall-associated extensin proteins [60] , but the biological function of most members of the AtPERK family is unknown . However , the expression of BnPERK1 from Brassica napus is rapidly induced following wounding , because of which it is thought to mediate early events in defense responses to cell wall damage by invading plant pathogens [61] . The modulation of basal immunity by venom allergen-like proteins in plants most likely involves at least two different classes of extracellular proteases . The first class of extracellular proteases regulating basal immunity to nematode infections in plants concerns the papain-like cysteine proteases . The inhibition of the extracellular papain-like cysteine protease Rcr3pim from S . pimpinellifolium by Gr-VAP1 results in enhanced susceptibility of tomato plants to G . rostochiensis [14] . Here , we showed that Gr-VAP1 also perturbs the extracellular papain-like cysteine protease C14tub from potato ( S . tuberosum; Fig . 3B ) , while ectopic Gr-VAP1 significantly increased the susceptibility of potato plants to G . rostochiensis ( Fig . 3A ) . Remarkably , ectopic Gr-VAP1 also suppressed basal immunity of Arabidopsis , even though this plant species is not a host of G . rostochiensis . However , a similar phenomenon has been observed with the apoplastic effector Avr2 of the C . fulvum , which acts as an inhibitor of Rcr3pim and several other extracellular papain-like cysteine proteases in tomato [15] , [16] , [42] , [62] . Although A . thaliana is not a host of C . fulvum either , ectopic Avr2 nonetheless interacts with multiple papain-like cysteine proteases required for basal defense in Arabidopsis [42] . The inhibition of apoplastic papain-like cysteine proteases by ectopic Avr2 also suppresses immunity of Arabidopsis to H . schachtii ( Fig . 5A ) . Similarly , the lack of three papain-like cysteine proteases most related to Rcr3pim in Arabidopsis mutants ( i . e . pap1 , pap4 , and pap5 ) suppresses immunity to H . schachtii ( Fig . 5B ) Altogether , these findings position the inhibition of extracellular papain-like cysteine proteases by apoplastic effectors as an important regulatory process in plant innate immunity to cyst nematodes . The second class of extracellular proteases most likely involved in the suppression of basal immunity by venom allergen-like proteins concerns subtilisin-like serine proteases . The transcript most down-regulated by ectopic Hs-VAP1 in Arabidopsis encodes the plant cell wall-associated subtilisin-like serine protease AtSBT3 . 14 ( S1 Table; AT4G21630; [49] , [63] ) . AtSBT3 . 14 is also down-regulated by ectopic Hs-VAP2 , albeit to a lesser extent . The Arabidopsis SBT family comprises 56 members , most of which are still uncharacterized . AtSBT3 . 14 is specifically expressed in roots , siliques , and dry seed of A . thaliana , but its function is not known [63] . AtSBT3 . 14 belongs to the same subfamily as AtSBT3 . 3 , which functions as an extracellular molecular switch in the priming of defense responses in Arabidopsis [50] . T-DNA insertion knockout mutations in the AtSBT3 . 3 gene compromise innate immunity of Arabidopsis . The loss of immunity of Arabidopsis mutants lacking a functional AtSBT3 . 14 gene to H . schachtii suggests that this plant cell wall-associated subtilase may also act as an extracellular regulator of basal innate immunity ( Fig . 6 ) . Most of the genes differentially regulated by the overexpression of venom allergen-like proteins in Arabidopsis are typically associated with innate immunity and plant cell wall-associated processes . A notable exception to this is NPQ4 , which was the most up-regulated gene in both Hs-VAP1 and Hs-VAP2 overexpressing plants . The PsbS protein encoded by NPQ4 is involved in non-photochemical fluorescence quenching in the thylakoid membranes of chloroplasts [51] , [64] . PsbS functions as photo protectant by mediating the thermal dissipation of excess excitation energy of singlet chlorophyll . Saturation of the electron transport chain in the photosystem II by excess light can lead to the accumulation of excited chlorophyll , which when insufficiently quenched by PsbS transfers its energy to oxygen to form highly reactive singlet oxygen [65] . The non-photochemical quenching capacity in chloroplasts thus regulates the generation of singlet oxygen [54] . Singlet oxygen can be the cause of oxidative damage , but on the other hand it is also involved in the peroxidation of lipids into oxylipin hormones ( e . g . jasmonic acid; [54] ) and in the onset of programmed cell death [55] , [66] . It is for this duality that PsbS is thought to play a key role in the trade-off between the ability to protect against abiotic photo-oxidative stress and the ability to mount effective redox-dependent immune responses to biotic invaders [54] . Our data shows that ectopic Hs-VAP1 and Hs-VAP2 suppress innate immune responses in Arabidopsis , at least partly , through their regulation of PsbS . As PsbS is a rate-limiting factor in non-photochemical quenching of excited singlet chlorophyll [67] , the more than 30-fold increase in the expression of NPQ4 by ectopic venom allergen-like proteins in the transgenic Arabidopsis most likely reduces the formation of singlet oxygen under biotic stress [54] . As a consequence , the constitutive augmentation of the non-photochemical quenching capacity by elevated levels of PsbS will probably also affect the production of oxylipin hormones ( i . e . jasmonic acid , and its precursors ) and the signaling of programmed cell death in response to biotic stress [54] . By contrast , PsbS-deficient npq4-1 mutant Arabidopsis plants show an enhanced production of jasmonic acid in response to herbivory by feeding insects [68] . We used the same Arabidopsis mutant line to demonstrate that PsbS-deficient plants are immune to infections by H . schachtii ( Fig . 6 ) . This finding shows that a functional PsbS protein is required for virulence of H . schachtii in Arabidopsis , possibly for down-regulating oxylipin hormone signaling or other singlet oxygen-dependent immune responses . While a PsbS-centered model may offer a plausible explanation for the suppression of immune responses by ectopic venom allergen-like proteins in leafs , the biological relevance of PsbS as regulator immunity to parasitic nematodes in roots is more puzzling . PsbS is localized in chloroplasts , which mainly occur in aerial plant parts that are exposed to light but not in roots . However , the Arabidopsis plants used in nematode infection assays are routinely cultured in vitro on translucent media in a light/dark cycle to monitor the infection over time . It has been shown before that permanent feeding structures formed by H . schachtii under these circumstances harbor chloroplasts [69] . Importantly , it has also been shown that NPQ4 is strongly up-regulated in permanent feeding structures of H . schachtii as compared to other root cells in the elongation zone of Arabidopsis plants kept in a light/dark cycle [70] . One could argue that the occurrence of chloroplasts and the expression of NPQ4 in nematode-induced feeding structures in roots are artifacts caused by the unnatural exposure of the roots to light . However , others have shown that permanent feeding structures of H . schachtii in Arabidopsis roots that are kept in the dark also harbor plastids with similar fluorescence spectra as chloroplasts [70] . More research is therefore needed to further investigate the nature and functions of PsbS in plastids of nematode-infected roots . Taken together , we conclude that non-photochemical quenching capacity is at least partly responsible for regulating innate immunity of roots to infection by H . schachtii under our experimental conditions . In conclusion , plants monitor the integrity of their cell walls with specific surface-localized pattern recognition receptors [71] , [72] . The recognition of fragments of plant cell walls can elicit strong basal defense responses that counteract further invasion by microbial invaders [73] , [74] . As plant-parasitic nematodes cause significant damage to plant cell walls during their migration inside host plants , they could evidently benefit from effectors that suppress immunity triggered by fragments from damaged plant cell walls . Our data allows for a model in which apoplastic venom allergen-like proteins of plant-parasitic nematodes suppress host defenses activated by the detection of fragments of plant cell walls released by migrating nematodes . This model could also explain why ectopic VAPs particularly affect the susceptibility of Arabidopsis to diverse unrelated lesion-forming plant pathogens , the symptoms of which also involve significant plant cell wall modifications . So far , we have identified three components of the molecular mechanisms that are most likely underlying the suppression of plant innate immunity by apoplastic venom allergen-like proteins ( i . e . extracellular papain-like cysteine proteases , cell wall-associated subtilisin-like serine protease , and the chlorophyll-associated Photosystem II subunit S protein ) . As our mutant analyses showed , each of these components separately has a major impact on immunity to cyst nematodes in Arabidopsis . An important question that needs further research is if all three components are part of a single signaling pathway that spans different subcellular compartments . This might not be the case , because promiscuous effectors like Gr-VAP and Avr2 can interact with multiple apoplastic papain-like cysteine proteases , each of which may feed into different signaling pathways .
Gr-VAP1 expression in preparasitic second stage juveniles ( ppJ2s ) of G . rostochiensis was knocked-down by soaking nematodes in double-stranded ( ds ) RNA matching 820 base pairs of the Gr-VAP1 coding sequence as described by Chen et al [75] and Rehman et al [56] . Briefly , a cDNA fragment was PCR-amplified with the primers Gr-VAP1-RNAiFW and Gr-VAP1-RNAiR ( S7 Table ) using Gr-VAP1 cDNA as template . The amplified cDNA fragment of Gr-VAP1 was subsequently used as template for generating dsRNA in vitro using the Megascript RNAi kit ( Ambion , Cambridgeshire , UK ) . Double-stranded RNA matching the sequence of the Nautilus gene from Drosophila melanogaster ( Genbank accession number M68897 ) was used as control treatment . RNA interference was induced in nematodes by soaking approximately 15 , 000 freshly hatched ppJ2s of G . rostochiensis Ro1 Mierenbos in a 1 mg/ml dsRNA solution , including 50 mM octopamine , 3 mM spermidine , and 0 . 05% gelatin . The treatments were done in duplo so that 15 , 000 juveniles could be processed further for infectivity assay on tomato seedlings and 15 , 000 juveniles could be used for semi-quantitative reverse transcription ( RT ) -PCR . To test the effect of RNA interference on infectivity of G . rostochiensis , we inoculated plates with five two-week old tomato seedlings ( cultivar Moneymaker ) on Gamborg B5 medium with 400 dsRNA-treated ppJ2s [56] . For each treatment a total number of 10 plates was inoculated with dsRNA-treated ppJ2s . The plants were grown at 24°C and light/dark cycles of 16 h/8 h . Seven days post inoculation , the roots were stained with acid fuchsin , destained using acidified glycerol , and the number of nematode per root was determined using a dissection microscope . The means of numbers of nematodes per plant were tested for significant differences in a one-way ANOVA . To analyze Gr-VAP1 expression after dsRNA treatment , total RNA was extracted from dsRNA-treated ppJ2s using the RNeasy Mini kit ( Qiagen , Hilden , Germany ) . RT-PCR was done following the protocol of the SuperScript™ III One-Step RT-PCR System ( Invitrogen ) using the primer Gr-VAP1-sRTFw and Gr-VAP1-sRTRv ( S7 Table ) , which were designed outside the region targeted by the dsRNA . The expression of the 60S acidic ribosomal protein-encoding gene ( Genbank accession number BM354715 . 1 ) was analyzed with primers 60S-RTFw and 60S-RTRv ( S7 Table ) as a reference for constitutive expression . Aliquots of the PCR solutions were visualized on ethidium bromide stained 1% agarose gel after 28 cycles . Fluorescent activity based protease profiling was used to test whether Gr-VAP1 perturbs the active site of cysteine proteases . The cysteine proteases C14tub of S . tuberosum and C14lyc of S . lycopersicum were transiently overexpressed in apoplastic fluids of N . benthamiana leaves following agroinfiltration [62] . Twenty-five to fifty microliters of apoplastic fluid was incubated with either 100 nM of P . pastoris produced Avr2 , 100 nM cystatin from chicken egg-white ( Sigma-Aldrich ) , or 300 nM of Gr-VAP1 isolated from apoplastic fluids of agroinfiltrated N . benthamiana leaves ( see below ) in 50 mM sodium acetate ( pH 5 . 5 ) and 100 µM DTT . To label the available active sites in these cysteine proteases , the proteins were subsequently incubated for 5 h with 1 µM of fluorescent DCG-04-TMR [76] . Fluorescent labeled proteins were separated in 12% Bis-Tris gels ( Invitrogen ) , which were subsequently analyzed using a fluorescent imager scanner ( Molecular Imager FX , Bio-Rad , Hercules , CA , USA ) . Labelling densities were quantified in triplicates using the computer software Quantity One 4 . 6 . 9 ( Bio-Rad Laboratories ) . To study the expression of Gr-VAP1 at different time points before and after inoculation , we used semi-quantitative RT-PCR as described above . Messenger RNA extraction and cDNA synthesis was conducted on parasitic second , third , and fourth stage juveniles and the adult males and females isolated from roots of susceptible potato ( cultivar Bintje ) at 13 , 19 , 23 , 27 , and 34 days post inoculation respectively . Gr-VAP1 expression in these developmental stages was examined by targeting a gene specific fragment of 146 base pairs of Gr-VAP1 with primers Gr-VAP1-RTFw and Gr-VAP1-RTRv ( S7 Table ) . The constitutively expressed cAMP-dependent protein kinase ( Gr-cAMP; GenBank accession number BM343563 ) was PCR amplified with the primers cAMP-RTFw and cAMP-RTRv ( S7 Table ) as a reference . We included reactions without reverse transcriptase to test for contaminating genomic DNA of the nematodes , while non-infected potato roots were included to check for non-specific amplification of host-derived cDNA . Transgenic potato plants ( Line V; genotype 6487-9 ) overexpressing Gr-VAP1 in the apoplast were generated as described by Postma et al [30] . Briefly , potato stem pieces were incubated for 10 minutes with a suspension of Agrobacterium tumefaciens strain AGL1 carrying the Gr-VAP1 cDNA sequence , including its native signal peptide for secretion , in pMDC32 [14] , [77] . Transformant callus was selected on ZCVK medium ( MS20 medium , 8 g/l plant agar , 1 mg/l zeatin , 100 mg/l kanamycin , 200 mg/l cefotaxim , 200 mg/l vancomycin; pH 5 . 8 ) . The introgression of Gr-VAP1 insert was checked by PCR on genomic DNA extracted from plant leaves using the DNeasy Plant Mini Kit ( Qiagen ) . The expression of Gr-VAP1 was checked by quantitative PCR ( qPCR ) using the primers qGrVAP1-Fw and qGrVAP1-Rv ( S7 Table ) on RNA extracted from leaves using the RNeasy Plant Mini Kit ( Qiagen ) . The constitutively expressed actin was amplified with the primers StActinF and StActinR ( S7 Table ) as a reference . qPCR was performed using Absolute QPCR SYBR Green Mix ( Thermo Fisher Scientific ) with amplification of 15 min at 95°C , followed by 35 cycles of 30 s at 95°C , 30 s at 63°C and 30 s at 72°C . Dried cysts of G . rostochiensis pathotype Ro1-Mierenbos were soaked in potato root diffusate on a 100-µm sieve to collect ppJ2s [78] . Remnants of roots and other debris were removed from suspensions of freshly hatched ppJ2s using centrifugation in sucrose gradient . Prior to inoculation potato plants the ppJ2s were surface sterilized , and resuspended in sterile 0 . 7% ( w/v ) solution of Gelrite ( Duchefa ) as previously described [30] . Approximately , 200 ppJ2s were inoculated onto 3-week-old in vitro-grown plants potato plant . Adult females per plant were counted 6 to 8 weeks after inoculation . Two independently transformed potato lines were used in these experiments . The infection assays were repeated at least 3 times . To identify and clone venom allergen-like proteins from the beet cyst nematode H . schachtii , we first queried the expressed sequence tag database at Genbank using the sequence of Gr-VAP1 as query . Four cDNA library clones , from which matching expressed sequence tags derived , were acquired from “The Washington University Nematode EST Project”[79] . Re-sequencing of library insert in these clones , with the primers M13Fw and M13Rv ( S7 Table ) , resulted in the identification of two full-length cDNA sequences encoding novel venom allergen-like proteins named Hs-VAP1 and Hs-VAP2 . The cDNA sequences encoding the complete open reading frames of Hs-VAP1 and Hs-VAP2 , including native signal peptides for secretion , were PCR- amplified with gene specific primers Hs-VAP1-GWFw , HsVAP1-GwRv , HsVAP2-GWFw , and HsVAP2-GwRv ( S7 Table ) and cloned into the entry vector pENT/D-TOPO ( Invitrogen ) . The inserts in these entry vectors were subcloned into the binary plasmids pMDC32 for stable plant transformation [77] or pGWB411 [80] for transient expression , using Gateway technology ( Invitrogen ) . To generate transgenic Arabidopsis lines constitutively overexpressing venom allergen-like proteins in apoplast , we transformed A . thaliana Columbia 0 with A . tumefaciens strain GV3101 carrying constructs of Gr-VAP1 , Hs-VAP1 , and Hs-VAP2 in pMDC32 , and pMDC32 without insert , using the floral dip method [81] . Primary transformants were selected on agar with 50 µg/ml kanamycin after which the plants were transferred to soil to produce seeds . Several independent homozygous single insertion lines were selected , and T3 and T4 generations were used for infection assays ( see below ) . The introgression of Hs-VAP1 and Hs-VAP2 was checked by PCR on genomic DNA extracted from seedlings using the DNeasy Plant Mini Kit ( Qiagen ) . The expression of the transgenes was checked by qPCR using the primers qHsVAP1-F , qHsVAP1-R , qHsVAP2-F , and qHsVAP2-R ( S7 Table ) on RNA extracted from seedlings using the RNeasy Plant Mini Kit ( Qiagen ) . The clathrin adaptor complex medium subunit family protein ( At5g46630 ) was amplified with the primers AtClathrinF and AtClathrinR ( S7 Table ) as a reference . qPCR was performed using Absolute QPCR SYBR Green Mix ( Thermo Fisher Scientific ) with amplification of 15 min at 95°C , followed by 35 cycles of 30 s at 95°C , 30 s at 60°C and 30 s at 72°C . Seeds of the homozygous transgenic T-DNA insertion mutants of the cysteine proteases PAP1 ( At2g34080 ) , PAP4 ( At2g27420 ) , PAP5 ( At3g49340 ) , the serine protease SBT3 . 13 ( At4g21630 ) , and the chlorophyll-associated Photosystem II subunit S ( At1g44575 ) were obtained from the SALK homozygote T-DNA collection . The mutant plants were propagated under standard greenhouse conditions of a 16-h/8-h light/dark regime and 60% relative humidity . Seeds from transgenic Arabidopsis and wild-type A . thaliana Col-0 were vapor sterilized and planted in 12-well cell culture plates ( Greiner bio-one ) containing modified Knop's medium [82] . Plants were grown at 24°C under 16-h-light/8-h-dark conditions . Two-week-old seedlings were inoculated with ∼250 surface-sterilized ppJ2s of H . schachtii [83] . Two and four weeks after inoculation , the number of female J4s of H . schachtii was counted by visual inspection . The statistical significance of the pairwise differences between plant genotypes and the empty pMDC32 vector control and the wild type Arabidopsis was assessed with a one-way ANOVA . The susceptibility of the Arabidopsis plants to infections by B . cinerea , P . cucumerina , A . brassicicola , and P . brassicae was determined on 4-week-old soil-grown plants [42] , [84] , [85] . Briefly , for B . cinerea , P . cucumerina , A . brassicicola , plants were drop inoculated by placing two 4-µl drops of conidial suspension ( 5×105 conidia/ml ) on each leaf . Plants were incubated at 20°C , 100% relative humidity , and a 16-h/8-h light/dark regime . Arabidopsis was inoculated with P . brassicae by placing 5-mm-diameter mycelial plugs of a 2-week-old P . brassicae agar plate culture onto leaves . Subsequently , the plants were incubated at 16°C , 100% relative humidity , and a 16-h/8-h light/dark regime . After two days the mycelial plugs were removed from the leaves . Disease progression for these pathogens was scored at regular intervals , and representative pictures were taken at 4 days after inoculation . The statistical significance of the pairwise differences between plant genotypes and the empty pMDC32 vector control was assessed with a one-way ANOVA . For inoculation of Arabidopsis with V . dahliae , 2-week-old soil-grown plants were uprooted and inoculated by dipping the roots for 2 min in a conidial suspension ( 106 conidia/ml ) . After replanting in soil , plants were incubated at standard greenhouse conditions of a 16-h/8-h light/dark regime and 60% relative humidity . Disease progression was monitored until 25 days after inoculation . The statistical significance of the pairwise differences between plant genotypes and the empty pMDC32 vector control was assessed with a one-way ANOVA . All infection assays were performed at least 2 times . The susceptibility of the Arabidopsis plants to infections by P . syringae pv . tomato DC3000 was determined on 2-week-old Arabidopsis seedlings as previously described [86] . Briefly , three 2-µl drops of a cell suspension of Pst at 109 CFU/ml , in 10 mM MgSO4 supplemented with 0 . 01% ( v/v ) Silwet L77 , was inoculated on the two most expanded and in the center of the leaf rosette . Inoculated plants were subsequently incubated at 21°C , 100% relative humidity , and a 16-h/8-h light/dark regime . Disease severity was scored 3 days after challenge inoculation . Colonization levels of the bacteria were determined with the method described by Pieterse et al . [87] . The statistical significance of the pairwise differences between plant genotypes and the empty pMDC32 vector control was assessed with a one-way ANOVA . Arabidopsis growth inhibition assays were performed as described elsewhere [88] . Briefly , seedlings were grown for 5 days on MS agar plates , supplemented with 1% w/v sucrose and 0 . 8% agar . Subsequently , seedlings were transferred to liquid MS medium supplied with 10 µM of the flg22 ( QRLSTGSRINSAKDDAAGLQIA ) synthetic peptide . One seedling was placed on 400 µl of medium in wells of 24-well-plates . The effect of treatment with the flg22 peptide on the growth of transgenic and wild type Arabidopsis ( Col-0 ) seedlings was analyzed after 7 days by measuring root length . Statistical significance of the difference between plant genotypes was assessed with a one-way ANOVA . Two-weeks-old transgenic Arabidopsis plants , grown under the same conditions as for the infection with H . schachtii , were collected , flash-frozen in liquid nitrogen and total RNA was extracted with the Maxwell® 16 LEV simplyRNA purification kit ( Promega ) . cDNA synthesis , library preparation ( 200-bp inserts ) , and Illumina sequencing ( 90-bp paired-end reads ) was performed at BGI ( Hong-Kong ) . Reads were mapped to the Arabidopsis genome ( tair10 ) using TopHat and transformed into a count per gene per sample by using the BEDTools suite ( function coverageBed ) . The edgeR [89] method was used to analyze differentially expressed genes ( DEGs ) between groups . DEGs were mapped to Gene Ontology ( GO ) terms in the database ( http://www . geneontology . org/ ) , and gene numbers were calculated for every term using an ultra-geometric test to find significantly enriched GO terms in DEGs . Calculated p-value went through a Bonferroni Correction , taking corrected p-value ≤0 . 05 as a threshold . KEGG pathway enrichment analysis was used to identify significantly enriched metabolic pathways or signal transduction pathways in DEGs comparing with the whole genome background . Subcellular localization was determined for all DEGs using the SUBcellular localization database for Arabidopsis proteins [90] . The suppression of programmed cell death in leaves of N . benthamiana was assessed by using Gr-VAP1 , Hs-VAP1 , Hs-VAP2 , and Mi-VAP1 ( including their native signal peptide for secretion ) subcloned into pGWB411 [80] . The Mi-VAP1 construct was synthetized at GeneArt ( Life Technologies ) based on the sequence in Genbank ( accession AAD01511 . 1 ) . All constructs were transferred to Agrobacterium tumefaciens GV3101 , and used for agroinfiltration in leaves of the N . benthamiana . Empty pGWB411 and plasmids carrying GFP in pGWB411 [80] were used as controls to assess the non-specific suppression of programmed cell death by agroinfiltration . The transient co-expression by agroinfiltration of several pairs of resistance genes and cognate elicitors was used to induce programmed cell death in leaves of N . benthamiana ( S6 Table; [30] ) . A . tumefaciens harboring individual binary vectors was grown at 28°C in liquid yeast extract peptone medium with appropriate antibiotics for 16 h . The bacteria were spun down and resuspended in infiltration medium to an optical density at 600 nm ( OD600 ) of 0 . 1 [30] . Agroinfiltration was done the abaxial side of the leaves of N . benthamiana using a 1 ml syringe . Co-infiltration of different constructs was performed by mixing equal volumes of the bacterial suspensions to a final optical density of 0 . 3 . Agroinfiltrated leaves were monitored for up to 7 d for cell death symptoms . To assess whether the VAPs harboring their native signal peptide for secretion ( in pGWB411 ) were indeed secreted to the apoplast of agroinfiltrated leaves of N . benthamiana , we isolated apoplastic fluids by vacuum-infiltrating ice-cold extraction buffer ( 50 mM phosphate-buffered saline pH = 7 . 4 , 100 mM NaCl , and 0 . 1% v/v Tween-20 ) for 10 min . Infiltrated leaves were surface dried and placed in a 10-ml syringe hanging in a 50 ml tube and centrifuged at 2000 g for 10 min at 4°C . Apoplastic fluids were subsequently separated under reducing conditions by SDS-PAGE on a 12% Bis-Tris gel and transferred to an Invitrolon™ PVDF membrane ( Life Technologies ) . For visualization of VAPs on western blots , we used a mouse monoclonal ANTI-FLAG® M2-Peroxidase ( HRP ) antibody to detect the FLAG tag at the carboxyl terminus of the recombinant proteins . Pictures were taken using the G:BOX Chemi System device ( SynGene ) . | Plant-parasitic nematodes have a major impact on global food security , as they reduce the annual yield of food crops by approximately 10 percent . For decades , the application of non-selective toxic chemicals to infested soils controlled outbreaks of plant-parasitic nematodes . The recent bans on most of these chemicals has redirected attention towards a wider use of basal , broad-spectrum immunity to nematodes in crop cultivars . However , it is currently not known if this most ancient layer of immunity affects host invasion by plant-parasitic nematodes at all . Basal immunity in plants relies on the detection of molecular patterns uniquely associated with infections in the apoplast by surface-localized receptors . Here , we demonstrate that venom allergen-like proteins in secretions of soil-borne cyst nematodes suppress immune responses mediated by surface-localized pattern recognition receptors . Migratory stages of cyst nematodes most likely deliver venom allergen-like proteins together with a range of plant cell wall-degrading enzymes into the apoplast of host cells . We therefore conclude that these nematodes most likely secrete venom allergen-like proteins to modulate host responses triggered by the release of immunogenic fragments of damaged plant cell walls . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immune",
"system",
"proteins",
"biochemistry",
"plant",
"science",
"proteins",
"plant",
"pathogens",
"plant",
"pathology",
"biology",
"and",
"life",
"sciences",
"immunology",
"nematology",
"zoology",
"immunomodulation",
"agriculture"
] | 2014 | Apoplastic Venom Allergen-like Proteins of Cyst Nematodes Modulate the Activation of Basal Plant Innate Immunity by Cell Surface Receptors |
The rapid accumulation of biological networks poses new challenges and calls for powerful integrative analysis tools . Most existing methods capable of simultaneously analyzing a large number of networks were primarily designed for unweighted networks , and cannot easily be extended to weighted networks . However , it is known that transforming weighted into unweighted networks by dichotomizing the edges of weighted networks with a threshold generally leads to information loss . We have developed a novel , tensor-based computational framework for mining recurrent heavy subgraphs in a large set of massive weighted networks . Specifically , we formulate the recurrent heavy subgraph identification problem as a heavy 3D subtensor discovery problem with sparse constraints . We describe an effective approach to solving this problem by designing a multi-stage , convex relaxation protocol , and a non-uniform edge sampling technique . We applied our method to 130 co-expression networks , and identified 11 , 394 recurrent heavy subgraphs , grouped into 2 , 810 families . We demonstrated that the identified subgraphs represent meaningful biological modules by validating against a large set of compiled biological knowledge bases . We also showed that the likelihood for a heavy subgraph to be meaningful increases significantly with its recurrence in multiple networks , highlighting the importance of the integrative approach to biological network analysis . Moreover , our approach based on weighted graphs detects many patterns that would be overlooked using unweighted graphs . In addition , we identified a large number of modules that occur predominately under specific phenotypes . This analysis resulted in a genome-wide mapping of gene network modules onto the phenome . Finally , by comparing module activities across many datasets , we discovered high-order dynamic cooperativeness in protein complex networks and transcriptional regulatory networks .
The advancement of high-throughput technology has resulted in the accumulation of a wealth of data on biological networks . Co-expression networks , protein interaction networks , metabolic networks , genetic interaction networks , and transcription regulatory networks are continuously being generated for a wide range of organisms under various conditions . Thanks to this great opportunity , network biology is rapidly emerging as a discipline in its own right [1] , [2] . Thus far , most computational methods have focused on the analysis of individual biological networks , but in many cases a single network is insufficient to discover patterns with multiple facets and subtle signals . There is an urgent need for methods supporting the integrative analysis of multiple biological networks . The analysis of multiple networks can be classified into two categories: ( 1 ) those studying conservations and evolvements of multiple networks between different species [3]–[8] , and ( 2 ) those identifying shared network modules or variations of modules across multiple networks of the same species but under different conditions [9]–[15] . The two types of problems face different challenges . Cross-species network comparisons are typically carried out on tens of networks , with the bottleneck being the graph isomorphism problem caused by the possible many-to-many ortholog mapping; while the network comparison within the same species deal with hundreds of networks simultaneously , and their principal challenge is the large search space . In this paper , we will focus on the latter problem . The analysis of multiple networks from the same species under different conditions has recently been addressed by ourselves and others with a series of heuristic data mining algorithms [9]–[14] . While useful , these methods still face two major limitations . ( 1 ) The general strategy of their searching heuristics is a stepwise reduction of the large search space , where each step involves one or more arbitrary cutoffs in addition to the initial cutoff that transforms continuous measurements ( e . g . expression correlations ) into unweighted edges . The ad hoc nature of these cutoffs has been a major criticism directed at this body of work [9]–[13] . ( 2 ) The cited algorithms cannot be easily extended to weighted networks . Most graph-based approaches to analyzing multiple networks are restricted to unweighted networks , and weighted networks are often perceived as harder to analyze [16] . However , weighted networks are obviously more informative than their unweighted counterparts . Transforming weighted networks into unweighted networks by dichotomizing weighted edges with a threshold obviously leads to information loss [17] , and if there is no reasonable way to choose the threshold , this loss cannot be controlled . This paper presents a new method of analyzing multiple networks that overcomes both of these issues . Generally speaking , a network of vertices can be represented as an adjacency matrix , where each element is the weight of the edge between vertices and . A number of numerical methods for matrix computation have been elegantly applied to network analysis , for example graph clustering [18]–[21] and pathway analysis [22] , [23] . In light of these successful applications , we propose a tensor-based computational framework capable of analyzing many weighted and unweighted massive networks . Although tensor computation has been applied in the fields of psychometrics [24] , [25] , image processing and computer vision [26] , [27] , chemometrics [28] , and social network analysis [29] , [30] , it has been explored only recently in large-scale data mining [31]–[35] and bioinformatics [36] , [37] . Simply put , a tensor is a multi-dimensional array and a matrix is a 2nd-order tensor . Given networks with the same vertices but different topologies , we can represent the whole system as a 3rd-order tensor ( see an example in Figure 1 ) . Each element is the weight of the edge between vertices and in the th network . By representing a set of networks in this fashion , we gain access to a wealth of numerical methods – in particular continuous optimization methods . In fact , reformulating discrete problems as continuous optimization problems is a long-standing tradition in graph theory . There have been many successful examples , such as using a Hopfield neural network for the traveling salesman problem [38] and applying the Motzkin–Straus theorem to solve the clique-finding problem [39] . Moreover , when a graph pattern mining problem is transformed into a continuous optimization problem , it becomes easy to incorporate constraints representing prior knowledge . Finally , advanced continuous optimization techniques require very few ad hoc parameters , in contrast with most heuristic graph algorithms . In this paper , we develop a tensor-based computational framework to identify recurrent heavy subgraphs ( RHSs ) in multiple weighted networks . A heavy subgraph ( HS ) is a subset of heavily interconnected nodes in a single network . We define a RHS as a HS that appears in a subset of multiple networks . The nodes of a RHS are the same in each occurrence , but the edge weights may vary between networks . Although the discovery of heavy subgraphs in a single biological network can reveal functional and transcriptional modules [40]–[42] , such results often contain false positives . Extending the search to a RHS is a promising way to enhance signal noise separation . Intuitively , any set of genes forming a RHS in multiple datasets generated under different conditions is more likely to represent a functional and transcriptional module than the genes in a single occurrence of a HS . We will use co-expression networks as examples due to their wide availability , but the tensor method described in this paper is applicable to any type of genome-wide networks . The concept of a RHS can be explained using the language of tensors , as shown in Figure 1 . Given microarray datasets , we model each dataset with a co-expression network . Each node represents one gene , and each edge's weight is the estimated co-expression correlation of the two genes . We then “stack” the collection of co-expression networks into a three-dimensional array such that each slice represents the adjacency matrix of one network . This array is a third-order tensor with dimensions genegenenetwork . A RHS intuitively corresponds to a heavy region of the tensor ( a heavy subtensor ) . The RHS can be found by reordering the tensor so that the heaviest subtensor moves toward the top-left corner . The subtensor in the top-left corner can then be expanded outwards from the left-top corner until the RHS reaches its optimal size . We applied our tensor algorithm to 130 weighted co-expression networks derived from human microarray datasets . We identified an atlas of functional and transcriptional modules and validated them against a large set of biological knowledge bases including Gene Ontology annotations , KEGG pathways , 191 Encode genome-wide ChIP-seq profiles , and 109 Chip-chip datasets . The likelihood for a heavy subgraph to be biologically meaningful increases significantly with its recurrence , highlighting the importance of the integrative approach . Moreover , our approach based on weighted graphs detected many patterns that would have been overlooked if we were analyzing unweighted graphs . In addition , we identified many modules that occur predominately under a specific type of phenotypes . Thus , we were able to create a genome-wide mapping of gene network modules onto the phenome . Finally , based on module activities across multiple datasets , we used a high-order analysis approach to reveal the dynamic cooperativeness in protein complex networks and transcription regulatory networks .
The choice of vector norms has a significant impact on the outcome of the optimization . The norm of a vector is typically defined in the form , where . The symbol , called the “-vector norm” , refers to this formula for the given value of . In general , the norm leads to sparse solutions where only a few components of the membership vectors are significantly different from zero [43] . The norm generally gives a “smooth” solution where the elements of the optimized vector are approximately equal . Details of these vector norms refer to Text S1 . A RHS is a subset of genes that are heavily connected to each other in as many networks as possible . These requirements can be encoded as follows . ( 1 ) A subset of values in each gene membership vector should be significantly non-zero and close to each other , while the rest are close to zero . To this end , we consider the mixed norm ( ) for . Since favors sparse vectors and favors uniform vectors , a suitable choice of should yield vectors with a few similar , non-zero elements and many elements that are close to zero . In practice , we approximate with the mixed norm , where . ( 2 ) As many network membership values as possible should be non-zero and close to each other . As discussed above , this is the typical outcome of optimization using the norm . In practice , we approximate with where for . Therefore , the vector norms and are specified as follows , ( 3 ) We performed simulation studies to determine suitable values for the parameters , , and by applying our tensor method to collections of random weighted networks . In subsets of these networks , we randomly placed RHSs of varying size , occurrence , and heaviness . We then tried different combinations of , , and , and adopted the combination ( , , and ) that led to the discovery of the most RHSs . More details on these simulations are provided in Text S1 . Since the vector norm is non-convex , our tensor framework requires an optimization method that can deal with non-convex constraints . While the global optimum of a convex problem can be easily computed , the quality of the optimum discovered for a non-convex problem depends heavily on the numerical procedure . Standard numerical techniques such as gradient descent converge to a local minimum of the solution space , and different procedures often find different local minima . Considering the fact that most sparse constraints are non-convex , it is important to find a theoretically justified numerical procedure . To design the optimization protocol , we use our previously developed framework known as Multi-Stage Convex Relaxation ( MSCR ) [43] , [44] . MSCR has good numerical properties for non-convex optimization problems [43] , [44] . In this context , concave duality is used to construct a sequence of convex relaxations that give increasingly accurate approximations to the original non-convex problem . We approximate the sparse constraint function by the convex function , where is a specific convex function ( ) and is the concave dual of the function ( defined as ) . In practice , is an effective choice as the convex upperbound of . The vector contains coefficients that will be automatically generated during the optimization process . After each optimization , the new coefficient vector yields a convex function that more closely approximates the original non-convex function . The solution of our tensor formulation Eq . ( 2 ) is a stationary point of the following regularized optimization problem: ( 4 ) where and are Lagrange multipliers . By exploiting the concave duality of , we can substitute for . Therefore , Eq . ( 4 ) can be rewritten as ( 5 ) We solve Eq . ( 5 ) by repeatedly applying the following two steps: The following box ( see Box 1 ) presents our two-stage protocol to solve the regularized form of Eq . ( 2 ) . The procedure can be regarded as a generalization of concave-convex programming [45] , which takes . By repeatedly refining the parameters in , we can obtain better and better convex relaxations leading to a solution superior to that of the initial convex relaxation with . The initial values of and could be uniform , randomly chosen , or taken from prior knowledge . In practice , by choosing an appropriate solver for Step 1 , the complexity of MSCR is linear with respect to the total number of edges in the tensor . For a detailed description of the optimization algorithm and procedure , see Text S1 . The RHSs can be intuitively obtained by including those genes and networks with large membership values . In practice , a pair of gene and network membership vectors and , i . e . , the solution of Eq . ( 2 ) , can result in multiple RHSs whose “heaviness” is greater than a specified value ( i . e . , a threshold ) . Here , the “heaviness” of a RHS is defined as the average weight of all edges in the RHS . In particular , the genes and networks are sorted in decreasing order of their membership values in and . As illustrated by the example in Figure 2A–C , the more top-ranking genes are included in the RHS , the less networks the RHS recurs in; and vice versa . Such overlapping structure is like a tower as shown in Figure 2D . We refer to a group of overlapping RHSs that is obtained from the same pair of and as a RHS family . To compress the redundant information , we build the representative RHSs for a RHS family as following: ( 1 ) if a RHS family contains multiple RHSs , the representatives are its two “extreme” RHSs: the RHS with the minimal number of genes ( e . g . , ) and as maximal recurrence as possible , and the RHS with the minimal number of networks ( e . g . , ) and as maximal number of genes as possible; ( 2 ) if a RHS family has only one RHS , it is the representative RHS . After discovering the representative RHSs in this manner , we can mask their edges in the networks where they recur with zero weights and optimize Eq . ( 2 ) again to search for the next heaviest RHS . The source code of the algorithm is available at our Supplementary Website http://zhoulab . usc . edu/tensor/ . This software is implemented by ANSI C and can be readily compiled and used in both Windows and Unix platforms . Even though the MSCR method is efficient , its computation time can still be long for large sets of networks with many edges . In such cases , edge sampling can provide an efficient approximation to many graph problems [46] , [47] . From the perspective of matrix or tensor computation , such sampling methods can be also viewed as matrix/tensor sparsification [48] . As RHS patterns predominately contain edges with large weights , we designed a non-uniform sampling method that preferentially selects edges with large weights . Specifically , each edge is sampled with probability : ( 6 ) where , and are constants that control the number of sampled edges . Note that Eq . ( 6 ) always samples edges with weights . It selects an edge of weight with probability proportional to the power of the weight . We choose , , and as a reasonable tradeoff between computational efficiency and the quality of the sampled tensor . To correct the bias caused by this sampling method , the weight of each edge is corrected by its relative probability: . The expected weight of the sampled network , , is therefore equal to the weight of the original network . However , in practice , when the adjusted edge weight ( but the original edge weight ) , we enforced it to be for avoiding too large edge weights . The overall edge sampling procedure adopts the simple random-sampling based single-pass sparsification procedure introduced in [48] . Details of the edge sampling procedure is provided in Text S1 . After edge sampling , the procedure described above will use the corrected tensor instead of the original tensor . We selected every microarray dataset from NCBI's Gene Expression Omnibus that met the following criteria: all samples were of human origin; the dataset had at least 20 samples to guarantee robust estimates of the expression correlations; and the platform was either GPL91 ( corresponding to Affymetrix HG-U95A ) , GPL96 ( Affymetrix HG-U133A ) , GPL570 ( HG-U133_Plus_2 ) , or GPL571 ( HG-U133A_2 ) . We averaged expression values for probe that map to the same gene within a dataset . The 130 datasets that met these criteria on 28 January 2008 were used for the analysis described herein . Details are available at http://zhoulab . usc . edu/tensor/ ) . We applied our methods to these 130 microarray datasets . Each microarray dataset is modeled as a co-expression network wherein each node represents a unique gene and each edge weight represents the strength of co-expression of two genes . To determine the weights , we first compute the expression correlation between two genes as the leave-one-out Pearson correlation coefficient estimate [49] . The resulting correlation estimate is conservative and sensitive to similarities in the expression patterns , yet robust to single experimental outliers . To make the correlation estimates comparable across datasets , we then applied Fisher's z transform [50] . Given a correlation estimate , Fisher's transformation score is calculated as . Because we observed the distributions of -scores to vary from dataset to dataset , we standardized the -scores to enforce zero mean and unit variance in each dataset [11] . Then , the “normalized” correlations are obtained by inverting the -score . Finally , the absolute value of is used as the edge weight of co-expression networks . Details is provided in Text S1 . In the other applications where networks contain negative edge weights , their edge weights can be transformed to be non-negative through translation , scaling or other transformation methods .
After applying our method to 130 microarray datasets generated under various experimental conditions , we identified 11 , 394 RHSs . Each RHS contains 5 member genes , appears in 5 networks , and has a “heaviness” ( defined as the average weight of its edges in networks where the RHS appears ) 0 . 4 . The average size of these patterns is 8 . 5 genes , and the average recurrence is 10 . 1 networks . The identified RHSs can be organized into 2 , 810 families with 4 , 327 representative RHSs , which we refer to in the following analysis . To assess the statistical significance of the identified RHSs , we applied our method to 130 random networks ( each of which is generated from one of the 130 weighted networks by the edge randomization method ) to identify RHSs with genes , networks and “heaviness” . We repeated this process 100 times . None of RHSs were identified in any of the 100 times . When the minimum recurrence is 4 and other criteria remain unchanged , only 3 RHSs were found ( Detail is provided in Text S1 ) . To assess the biological significance of the identified RHSs , we evaluate the extent to which these RHSs represent functional modules , transcriptional regulatory modules , and protein complexes . Our microarray data collection covers a wide range of phenotypic conditions , especially most of all , many different types of cancers ( cancers accounts for 46% of the datasets ) . If an RHS is activated repeatedly only under one type of phenotypic condition , then it is likely to contribute specifically to the molecular basis of the phenotype . It is known that phenotypes are determined not only by genes , but also by the underlying structure of genetic networks . While traditional genetic studies have sought to associate single genes with a particular phenotypic trait , identifying phenotype-specific network modules has been a challenge of network biology . Below we show that a large number of the RHSs identified by our method are indeed phenotype-specific . First , we determined the phenotypic context of a microarray dataset by mapping the Medical Subject Headings ( MeSH ) of its PubMed record to UMLS concepts . We used the MetaMap Transfer tool provided by the UMLS [55] for this purpose . UMLS is the largest available compendium of biomedical vocabularies , spanning approximately one million interrelated concepts . It includes diseases , treatments , and phenotypic concepts at several levels of resolution ( molecules , cells , tissues , and whole organisms ) . We annotated each microarray dataset with matching UMLS concepts and all of their ancestor concepts . For each RHS , we evaluated phenotype specificity by computing the hypergeometric enrichment of specific UMLS concepts present in those datasets where the RHS occurs . If the -value , we consider the RHS module is significantly phenotype-specific . 5 . 62% of RHSs show phenotype-specific activation patterns , compared to 0 . 14% of randomly generated RHSs . The most frequently enriched phenotype concepts are related to cancer . For example , the most prevalent concepts are “Leukemia , Myelocytic , Acute” ( enriched in 1 . 8% of modules ) and “Neoplasms , Neuroepithelial” ( 1 . 3% ) . Among non-cancer concepts , the most frequent are “Respiratory Tract Diseases” ( enriched in 0 . 2% of modules ) , “Bone Marrow Diseases” ( 0 . 2% ) and “Lung diseases” ( 0 . 1% ) . Below we illustrate two examples of phenotype-specific modules . Figure 7A shows a 7-gene module ( CCNB1 , POLE2 , CDC2 , PTTG1 , RNASEH2A , CDKN3 , MCM4 ) that is active in 21 datasets . Twelve of the 21 datasets are related to cancer , and three relate to the study of Glioma ( GDS1975 , GDS1815 , GDS1962 ) ( -value = 0 . 075 ) . Interestingly , four out of the seven genes are known to be associated with Glioma . CCNB1 and CDC2 play important roles in the proliferation of Glioma cells [56] , the expression of PTTG1 is correlated with poor prognosis in Glioma patients [57] , and aberrant splicing of CDKN3 increases proliferation and migration in Glioma cells [58] . This knowledge confirms our prediction of the module's strong association with Glioma . This module is enriched in genes from the cell cycle pathway ( CCNB1 , CDC2 , PTTG1 , and MCM4; -value = 1 . 08E-3 ) . Figure 7B shows a 5-gene module ( COL3A1 , COL1A2 , COL5A2 , VCAN , THY1 ) that is active in 22 datasets . Four of these datasets study expression in muscle tissue ( GDS914 , GDS563 , GDS268 , GDS2055 ) ( -value = 0 . 03 ) . This module contains 3 genes ( COL3A1 , COL1A2 , COL5A2 ) annotated with fibrillar collagen ( -value = 8 . 41E-4 ) , a major component of muscle ( especially cardiac skeleton ) . Furthermore , COL1A2 and VCAN are targeted by neuron-restrictive silencer factor ( NRSF ) . Notably , [59] has reported that the NRSF maintains normal cardiac structure and function and regulates the fetal cardiac gene program . In addition , VCAN plays a role in conditions such as wound healing and tissue remodeling in the infracted heart [60] . Four out of five genes in the module are associated with muscle , providing strong evidence for phenotype specificity . The discovery of RHS modules spanning a variety of experimental or disease conditions enabled us to investigate high-order coordination among those modules . We applied our previously proposed second-order analysis to study cooperativity among the protein complexes[49] . We define the first-order expression analysis as the extraction of patterns from one microarray data set , and the second-order expression analysis as a study of the correlated occurrences of those patterns ( e . g . heavy subgraph recurrence ) across multiple data sets . Here , for each identified RHS , we constructed a vector of length storing its heaviness factors in the data sets . This vector is a profile of the module's first-order average expression correlations , and can be interpreted as the activity profile of the module in different datasets . To quantify the cooperativity between two modules , we calculated the correlation between their first-order expression correlation profiles . It is defined as the second-order expression correlation of the two modules . Figure 8 shows a cooperativity map of all protein complexes represented by the RHSs that have high ( ) second-order correlations with at least one other protein complexes . The most striking feature of this map is a large and very heavily interconnected subnetwork of 32 complexes , all involved in the cell cycle . Within this subnetwork , 17 complexes ( including CDC2_Complex , CCNB2_CDC2_Complex , CDK4_Complex , Chromosomal_Passenger_Complex , and Emerin_Complex_24 ) form a tight core wherein each complex has strong second-order correlations ( 0 . 95 ) with all others in the core . This structure highlights the strict transcription regulation of cell cycle processes . Two other prominent dense subnetworks contain protein complexes involved in the respiratory chain and those in translation ( e . g . the ribosomal complex , the NOP56 associated pre-RNA complex , and the TRBP complex associated with miRNA dicing ) . Another loosely coupled subnetwork contains protein complexes mainly involved in transcription and post-transcriptional modification , including the participating members of CDC5L complex ( pre-mRNA splicing ) , CF IIAm complex ( pre-mRNA cleavage ) , SNF2h-cohesion-NuRD complex ( chromatin remodeling ) , DA complex ( transcription activation ) , and the large drosha complex ( primary miRNA processing ) , revealing the tight coupling between transcription and post-transcriptional processes . Numerous protein complexes ( e . g . CEN complex , FIB-associated complex , and CCT complex ) connect these dominant subnetworks or supercomplexes into an integrated network . Thus , our approach not only provides a comprehensive catalogue of modules that are likely to represent protein complexes , but also the very first systematic view of how protein complexes dynamically coordinate to carry out major cellular functions . That is , by integrating data generated under a variety of conditions , we have gained a glimpse into the activity organization chart of the proteome . The same principle can be applied to uncover the cooperativity among the transcription modules , thereby reconstructing transcriptional networks . The RHS discovery resulted in an atlas of transcription modules activated under different conditions . Each transcription module can be regulated by one or more transcription factors . Intuitively , if two transcription modules form or do not form two co-expression clusters always under the same set of conditions ( that is , in the same data sets ) , it in fact suggests that their respective transcription factors are active or inactive simultaneously . The cooperativity between two sets of transcription factors can again be quantified using second-order expression correlation , since the the activity of a transcription factor can be assessed by the tightness of co-expression among the genes it regulates , i . e . , the first-order profiles of the corresponding RHS module . We focus on the 57 transcription factors with enriched targets in our modules . Among these TFs , we identified 25 TF pairs , each of which regulate two distinct modules with second-order correlations greater than 0 . 7 . We traced the potential sources of cooperativity in these pairs using genome-wide TF binding data and protein-protein interaction data [61] . Given two modules controlled respectively by transcription factors TF1 and TF2 , which for simplicity are assumed to be individuals instead of sets of transcription factors , there are at least three types of possible direct causes of cooperativity between TF1 and TF2 ( Figure 9A ) : the expressions of TF1 and TF2 are activated by a common transcription factor TF3 ( a type I transcription network ) , or TF1 activates the expression of TF2 ( a type II transcription network ) , or TF1 and TF2 interact at the protein level ( a type III transcription network ) . In the special case where a module pair shares the majority of common genes , the cooperativity between TF1 and TF2 is known to be combinatorial control . Note that these three types of transcription networks are certainly only a few of the many possibilities . We identified 33 transcription networks , among which 10 networks are of Type I , 19 are of Type II , and 4 are of Type III . These transcription networks interconnect to form a partial cellular regulatory network ( Figure 9 ) . Four networks are involved in the cell cycle: the Type I network involving SREBP1 and TAF1/E2F4 , the two Type II networks involving STAT1 and E2F4 as well as SP1 and NFYA , and the Type III network involving ELF1 and SP1 . The roles of these networks are supported by the independent evidence of cooperative roles of those transcription factors reported in the literature [62]–[65] . Other transcription networks participate in translational elongation , rRNA processing , RNA splicing , DNA replication , DNA packaging , electron transport , etc . Notably , our reconstructed transcriptional regulatory network includes 35 modules that represent protein complexes , which provides a mechanistic explanation for the correlated activities of those protein complexes , as shown in Figure 8 . For example , cooperativity between the chromosome passenger complex CPC and the MCM complex ( see Figure 9B ) can be attributed to the Type II networks between their regulators E2F4 and NFY . This is consistent with previous evidences on the synergistic activities between the two transcription factors [66] . Strikingly , the protein complexes in the ribosome that participate in the translational elongation are regulated by a network of intertwined transcription networks . This highlights the regulatory complexity of the translation process , an impressive feat given that the TFs used in this study represent only a very small fraction of the TF repertoire .
We have developed a novel tensor-based approach to identify recurrent heavy subgraphs in many massive weighted networks . This is the first method suitable for pattern discovery in large databases of many weighted biological networks . We applied the method to 130 co-expression networks , and identified a large number of functional and transcriptional modules . We show that the likelihood for a heavy subgraph to be meaningful increases significantly with its recurrence in multiple networks , highlighting the importance of the integrative approach for network analysis . By analyzing databases of networks derived from a wide range of experimental conditions , we can also study the high-order dynamic coordination of modules , a task that can be hardly addressed using only a single network . In addition , the phenotype information associated with gene expression datasets provides opportunities to perform systematic genotype-phenotype mapping [14] , [67] . Among our identified modules , many have been shown to be phenotype-specific . While weighted networks are often perceived as harder to analyze than their unweighted counterparts , we show that many patterns are overlooked if using the unweighted networks . Although currently unweighted networks ( protein-protein interaction network , genetic interaction network , and metabolic network , etc . ) still dominate biological studies , rapidly evolving genomics technology will soon be able to provide quantitative assessments of those interactions , thus resulting in accumulated weighted networks . Our method is well positioned to respond to the emerging challenges of network biology . | To study complex cellular networks , we need to consider their dynamic topologies under many different experimental or physiological conditions . Integrative analysis over large numbers of massive biological networks thus emerges as a new challenge in data mining . Recently , we and others have proposed several algorithms for recurrent pattern mining across many ( ) biological networks ( with the main focus on unweighted networks ) . However , thus far no algorithms have been specifically designed to mine recurrent patterns across a large collection of weighted massive networks . In this paper , we propose a computational framework to identify recurrent heavy subgraphs from many weighted large networks . By applying our method to 130 co-expression networks , we identified an atlas of modules that are highly likely to represent functional modules , transcriptional modules , and protein complexes . Many of these modules would be overlooked with unweighted networks analysis . Furthermore , many of the identified modules constituted signatures of specific phenotypes . Finally , we demonstrated that our results facilitate the study of high-order dynamic coordination in protein complex networks and transcriptional regulatory networks . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"computational",
"biology"
] | 2011 | Integrative Analysis of Many Weighted Co-Expression Networks Using Tensor Computation |
Dengue is one of the most important infectious diseases of humans and has spread throughout much of the tropical and subtropical world . Despite this widespread dispersal , the determinants of dengue transmission in endemic populations are not well understood , although essential for virus control . To address this issue we performed a phylogeographic analysis of 751 complete genome sequences of dengue 1 virus ( DENV-1 ) sampled from both rural ( Dong Thap ) and urban ( Ho Chi Minh City ) populations in southern Viet Nam during the period 2003–2008 . We show that DENV-1 in Viet Nam exhibits strong spatial clustering , with likely importation from Cambodia on multiple occasions . Notably , multiple lineages of DENV-1 co-circulated in Ho Chi Minh City . That these lineages emerged at approximately the same time and dispersed over similar spatial regions suggests that they are of broadly equivalent fitness . We also observed an important relationship between the density of the human host population and the dispersion rate of dengue , such that DENV-1 tends to move from urban to rural populations , and that densely populated regions within Ho Chi Minh City act as major transmission foci . Despite these fluid dynamics , the dispersion rates of DENV-1 are relatively low , particularly in Ho Chi Minh City where the virus moves less than an average of 20 km/year . These low rates suggest a major role for mosquito-mediated dispersal , such that DENV-1 does not need to move great distances to infect a new host when there are abundant susceptibles , and imply that control measures should be directed toward the most densely populated urban environments .
Dengue is the most important mosquito-borne viral disease of humans , annually responsible for approximately 40 million cases and some 20 , 000 deaths in tropical and subtropical regions [1] . Dengue is caused by one of four single-stranded positive-sense RNA viruses ( DENV-1 to DENV-4 , also referred to as serotypes ) of the genus Flavivirus ( family Flaviviridae ) . Despite the large burden of dengue disease , and considerable research effort , there are currently no licensed vaccines or specific therapies . The challenge of effective and safe dengue vaccination is increased by the possibility that imperfect cross-protective vaccination could enhance DENV infection , or even virulence [2] , and that lineages within individual DEN viruses , particularly different ‘genotypes’ , may also differ in antigenicity [3]–[6] . In addition , the population dynamics of DENV within individual localities are complex , involving the birth-and-death of viral lineages that may also differ in both virulence and fitness [7]–[13] , as well as the intricate patterns of gene flow , at both the local and international scales [7] , [14] , [15] . DENV transmission among humans is largely caused by the urban adapted and anthropophillic Aedes aegypti mosquito . Spatial and temporal patterns of dengue prevalence are likely driven by multiple factors including the immune status of human hosts [16] , their age [17] , [18] , virus traits [13] , [19] , [20] , the mosquito vector , and environmental variables including aspects of climate such as levels of precipitation [21] , [22] . Human movement must also be an important , but poorly understood , contributor to viral transmission dynamics , and is obviously responsible for the increasingly widespread and complex distribution of the four DEN viruses at the global scale . On a local scale , how much DENV transmission within a specific population is due to the local movement of infected human hosts rather than of mosquitoes is unclear . Understanding the spatial and temporal dynamics of dengue transmission in endemic dengue populations is therefore central to the rational deployment of vector control activities and the design of intervention strategies . In this respect it is critical to determine the spatial structure of DENV within endemic populations , the rate at which DENV lineages diffuse through space ( particularly in the face of a partially immune population ) , whether specific lineages are spreading more rapidly than others and indicative of enhanced fitness , and the likely contribution of mosquitoes and humans to local transmission patterns . To address these questions we employed a fine-scale molecular approach to characterize the virus population dynamics of a recent DENV-1 outbreak in southern Viet Nam , a region of high dengue endemicity . Between 2006–2008 the estimated incidence of DENV-1 infection in the southern twenty provinces of Viet Nam ranged from 86–190 cases/100 , 000 [13] , markedly higher than during the preceding six-year period when it ranged from 1–28 cases/100 , 000 . The causes of this increased incidence are unknown . To determine the patterns and dynamics of dengue transmission we utilized an expansive data set of DENV-1 whole genome sequences sampled prior to and during the peak in DENV-1 prevalence over a period of six years ( 2003–2008 ) . We inferred the dynamics of viral transmission within individual communities , between communities , and between neighboring countries , using recently developed Bayesian phylogenetic methods that utilize both the temporal and spatial information of the sampled sequences . Uniquely , these time-calibrated phylogenetic methods provide the ability to reveal the complex interplay of spatial , genetic and epidemiological dynamics at the local , regional and global scales , and have the ability to consider individual viral lineages , whereas epidemiological approaches based on the analysis of incidence data are at best only able to distinguish among the four DEN viruses .
We determined the consensus DENV-1 genome sequence ( minimum sequence from nt 70–10 , 400 ) in acute plasma samples collected from 751 hospitalized patients in urban Ho Chi Minh City ( HCMC ) ( n = 575 sampled between 2003–2008 ) and rural Dong Thap Province in the Mekong Delta region ( n = 176 sampled between 2006–2007 ) . The majority of viruses were sampled from 2006 to 2008 during which DENV-1 was the most prevalent serotype in circulation ( Figure S1 ) . To determine the evolutionary relationships of DENV-1 in Viet Nam in the context of surrounding countries we analyzed the envelope ( E ) gene sequences from these locations ( Figure 1A ) . The 751 DENV-1 sequences sampled from Viet Nam fell into one of five clades within the broader Genotype I cluster of viruses [23] . Four of the five clades consistently clustered within the diversity of Cambodian viruses with good support ( posterior probability ranging from 0 . 81 to 1 . 0 ) . This phylogeographic evidence , coupled with Cambodia and Viet Nam's shared border , is compatible with Cambodia acting as the major source of Vietnamese DENV-1 . A caveat to this is the lack of contemporaneous DENV-1 sequences from nearby Thailand , which has previously been shown to harbor substantial DENV diversity and importation into Viet Nam [13] . Clearly , wider sampling in both time and space is needed to test this hypothesis . The majority of the clades largely comprised viruses from HCMC , with the exception of clade 1 , which was found to be Dong Thap dominant . The timing of these inferred introductions were gauged from the age of the most recent common ancestor ( TMRCA ) of each clade ( Table 1 ) . The period in which these different viral clades emerged in southern Viet Nam ranged from late 2001 to mid-2005 . Apart from clade 1 , which was found to be the most recent introduction , the mean ages of clades 2–5 did not differ significantly , suggesting that different viral lineages were imported over short or similar time-scales , and then co-circulated . These clades were chosen for more detailed phylogeographic analysis . Finally , genome-wide rates of nucleotide substitution – at ∼1×10−3 nucleotide substitutions per site , per year ( Table 1 ) – were the same among clades and highly consistent with those previously determined for DENV [24] , [25] . For the clades identified as being within Viet Nam , a discrete spatial model [26] was employed to reveal the migration between the sampling locations . The results are shown in Figure 1B , in which branches are colored by the most probable state location . In four of the five clades HCMC was the most likely viral source , with viruses exported to the rural area of Dong Thap . The non-HCMC isolates in these clades were interspersed among the HCMC sampled isolates , which strongly suggested that the DENV-1 epidemics in southern Viet Nam mainly emerged first in HCMC . The exception was clade 1 , which was dominated by Dong Thap viruses and where Dong Thap was inferred to be the most likely place of origin . Moreover , the HCMC viruses in clade 1 did not form a monophyletic group , supporting the view that clade 1 viruses were imported into HCMC on multiple occasions from Dong Thap . To determine whether the viral migration rates varied between urban and rural epidemics , we compared the spatial dynamics between clades 1 and 4 ( Table 2 ) . When focusing on the number of transitions from the inferred source location , a symmetrical pattern was observed between the two clades . For instance , the transmission rate between HCMC and Dong Thap was higher in the HCMC dominant clade 4 , while for the reverse direction ( Dong Thap to HCMC ) it was greater in Dong Thap dominant clade 1 . Hence , once a virus became established in a location , rural or urban , the rate of viral exportation was found to be greater than the rate of viral importation . The geographical coordinates of the patient's residential address in HCMC ( n = 381 ) or Dong Thap Province ( n = 175 ) was known for 556 cases and this information was employed to reconstruct the fine-scaled dispersion of the individual viral lineages within the sampling areas using a continuous spatial diffusion model with non-homogenous dispersion rates [27] . The average viral dispersion rate ( km/year ) was calculated for each clade , and separately for HCMC and/or Dong Thap data subsets , as if the epidemic in these regions derived from a single introduction ( Table 2 ) . We define virus dispersion rate as a measure of how quickly a virus lineage spreads geographically , given the inferred root location and final sampling locations . Even though we only had one estimate of the average dispersion rate of DENV-1 in Dong Thap , a clear disparity was observed when compared to the rates from HCMC lineages ( Table 3 ) . Specifically , the viral lineages from clade 1 in Dong Thap spread approximately 2–3 times faster than any lineage from HCMC . This is indicative of a fundamental difference in the epidemiological dynamics of DENV-1 in the two areas . A further dissection of the dispersion rates through time in HCMC ( clades 2 , 4 and 5 ) and Dong Thap ( clade 1 ) revealed interesting patterns in the rate of viral spread in the two locations . In HCMC ( Figure 2A and B ) , the monthly incidence of DENV-1 showed a similar trend as in Dong Thap , with corresponding regular fluctuations and an increasing overall trend . However , there was no clear association between genetic diversity , incidence , and dispersion rate observed in the urban environment demonstrated by the roughly horizontal relationship in Figure 2B and the overlapping 95% HPD ( highest posterior density ) intervals . Hence , although the DENV-1 clades were introduced independently into HCMC , they had spread at similar and effectively constant rates . For Dong Thap , clade 1 was the only one clearly derived from a distinct single importation and of a sufficient size for analysis . The dispersion rate of DENV-1 appeared to be associated with the fluctuations in genetic diversity and monthly incidence in Dong Thap ( Figure 2C and D ) . The two peaks in relative genetic diversity of clade 1 in Dong Thap coincided with the two major peaks in the monthly incidence , indicating that DENV-1 epidemic in Dong Thap is largely driven by this lineage . To investigate whether these dispersion rate estimates in HCMC were simply a reflection of the geographic constraint of our samples , they were re-estimated by randomizing the tip location for each clade ( Table 4 ) . The results indicated what the maximum dispersion rate could be given the sampled locations , which were found to be 2–3 times greater than the empirical estimates , with wide HPD intervals ( Table 4 ) . The spatial reconstruction of the viral spread at different stages of the epidemics showed that these viral lineages had co-circulated in the same place at the same time ( Figure 3 ) . This observation is of fundamental importance as it suggests that the number of susceptible hosts to DENV-1 had not been saturated in HCMC , and could potentially have supported additional DENV-1 lineages in this area . To determine whether transmission routes within HCMC varied according to population density , we employed a non-reversible discrete phylogeography model applied to district level data . Importantly , the more densely populated inner city districts ( above 30 , 000 people per km2 ) were found to contribute significantly to DENV-1 transmission compared to the suburban districts ( Figure 4 ) . Moreover , the most densely populated region , District 5 , had the highest number of connections , providing compelling evidence that this area might be a major hub in the city .
At the scale of South-East Asia , the observation that there is a strong clustering by country indicates that there is a far higher level of DENV-1 gene flow within than between countries . Such a phylogeographic pattern is compatible with relatively short transmission distances for DENV as a whole , including that meditated by mosquitoes . This rather limited spatial movement also sits in marked contrast to that observed in respiratory borne pathogens such as influenza , where there is relatively little clustering by place of isolation even on a global scale [28] . Each of the five clades of DENV-1 we identify has a very recent common ancestry , dating only shortly before the appearance of that clade . Given that dengue is endemic in southern Viet Nam , with DENV-1 circulating there for at least 23 years [29] , such recent common ancestry suggests that there is a rapid and continual turnover of viral lineages , as has been increasingly described for this and other DEN viruses [8]–[11] , [30] , [31] . Less clear is whether these instances of lineage turnover are due to fitness differences between the lineages in question , such that natural selection is preferentially able to favor one lineage over another , or whether there is simply a stochastic die-off . That the three major clades we detect in HCMC co-circulate in the same spatial region with overlapping ranges , and possess broadly equivalent levels of relative genetic diversity , suggests that they are of similar fitness and hence that there is little , if any , competition between them . Consistent with this , we did not observe differences in early plasma viremia levels between patients infected with viruses belonging to the different clades ( Figure S2 ) . Indeed , we suggest below that HCMC is likely characterized by a large number of susceptible hosts , which would in turn reduce the extent of selective competition between lineages . More generally , these results indicate that although a specific viral serotype may appear to be endemic in a specific geographic region for an extended period , this does not mean that the same viral clades are involved throughout this period . A striking result from this study is that the ‘virus dispersion rates’ we estimate appear to be very low , and particularly in HCMC where mean rates were universally <20 km/year . Such low rates are especially noteworthy given the rapidity and geographic scale with which DENV-1 re-emerged as the dominant serotype in southern Viet Nam [13] . We therefore interpret these low rates to mean that urban centers like HCMC are characterized by sufficiently high numbers of susceptible hosts such that the virus does not have to move very far to infect a new host . Such a notion is supported by the fact that higher virus dispersion rates are observed in Dong Thap , which is characterized by an approximately ten-fold lower population density ( 495 persons/km2 ) relative to HCMC ( 3024 persons/km2 ) , although more estimates are clearly needed from this locality . In addition , the highest levels of viral movement were found in and out of the most densely populated region of HCMC ( District 5 ) , suggesting that this well-connected locality acts as a focal point for dengue dispersion within the city . Hence , it is not that DENV-1 moves slowly at a spatial scale in HCMC , but rather that it does not have to move far geographically to continue its transmission . Although our sample of genome sequences is biased toward those from HCMC , our analysis indicates that DENV-1 generally diffuses from HCMC to Dong Thap . Again , this observation is suggestive of a gravity model of viral transmission , in which spatial diffusion occurs over a gradient of population density , and is compatible with our observation that dispersion rates are associated with the numbers of susceptible hosts . A similar gravity-dependent pattern of virus dispersion was recently suggested for DENV-2 in Viet Nam [14] , although the use of a strictly reversible phylogeographic model in that case meant that directionality could not be ascertained with certainty . Combined , these studies strongly suggest that the density of the human host population plays a fundamental role in determining the transmission dynamics of endemic dengue . Typically , adult A . aegypti mosquitoes travel short distances of less than ∼100 m during their average life-span of a few weeks [32]–[34] . The very short distances traveled by DENV-1 , particularly in HCMC is consistent with mosquitoes , rather than humans , being responsible for the majority of the spatial spread in HCMC , which is again in part a function of the high density of susceptible hosts . A similarly limited movement of dengue has been reported by recent studies that focused on smaller geographic areas , reflecting the restricted spatial range of mosquito vectors , and corroborating the highly focal pattern of DENV transmission observed in HCMC [15] , [35] . It is also notable that the geographical range of the three major clades in HCMC changed little from 2003–2008 . As such , the full geographic range of these clades is established very early on as the virus is able to spread rapidly through a susceptible host population . Upon the introduction of a new dengue serotype into Iquitos , Peru , it was noted that early-confirmed cases were scattered throughout the city , suggesting a rapid establishment of the virus when entering a completely naïve population [36] . This observation gives added weight to our conclusion that the dispersion rates of DENV-1 in southern Viet Nam are largely a function of the availability of susceptible hosts . These results have a number of important implications for the future control of dengue . Most generally , that DENV tends to spread relatively slowly on a spatial scale ( such that DENV phylogenies exhibit a strong spatial structure both nationally and internationally ) suggests that any future vaccine escape or drug resistance mutations would also spread relatively slowly . In addition , that the dispersion rates of DENV appear to largely reflect the density of human host population , including movement from Ho Chi Minh City to Dong Thap , suggests that future control measures , including mosquito spraying , should be directed toward the densest host populations and preferentially to urban over rural areas .
The dengue patients from whom DENV whole genome sequences were determined were enrolled in one of two prospective studies at the Hospital for Tropical Diseases in Ho Chi Minh City , Viet Nam or at Dong Thap Hospital , Dong Thap Province , Viet Nam . The median age of these patients was 12 years ( interquartile range 7–17 years ) and 51% were male . Serological investigations ( IgM and IgG capture ELISAs ) were performed using paired plasma samples using methods described previously [37] . DENV serotype and viraemia levels were determined using an internally-controlled real-time RT-PCR assay that has been described previously [38] . Viral genomes were sequenced using the Broad Institute's capillary sequencing ( Applied Biosystems ) directed amplification viral sequencing pipeline http://www . broadinstitute . org/scientific-community/science/projects/viral-genomics-initiative ) . This sequencing effort was part of the Broad Institute's Genome Resources in Dengue Consortium ( GRID ) project . Viral RNA was isolated from diagnostic plasma samples ( QIAmp viral RNA mini kit , Qiagen ) and the RNA genome reverse transcribed to cDNA with superscript III reverse transcriptase ( Invitrogen ) , random hexamers ( Roche ) and a specific oligonucleotide targeting the 3′ end of the target genome sequences ( nt 10868 to 10890 , AGAACCTGTTGATTCAACAGCAC ) . cDNA was then amplified using a high fidelity DNA polymerase ( pfu Ultra II , Stratagene ) and a pool of specific primers to produce 14 overlapping amplicons of 1 . 5 to 2 kb in size for a physical coverage of 2-fold across the target genome ( nt 40 to 10649 ) . Amplicons were then sequenced in the forward and reverse direction using primer panels consisting of 96 specific primer pairs , tailed with M13 forward and reverse primer sequence , that produce 500–700 bp amplicons from the target viral genome . Amplicons were then sequenced in the forward and reverse direction using M13 primer . Total coverage delivered post amplification and sequencing was 8-fold . Resulting sequence reads were assembled de novo using the Broad Institute's AV454 assembly algorithm ( Henn et al . 2011 . in review ) and a reference-based annotation algorithm . All whole genome sequences newly determined here have been deposited in GenBank and assigned accession numbers ( Table S1 ) . A data set of DENV-1 sequences was collated to include isolates from countries in Southeast Asia that were likely linked to Viet Nam via migration . An alignment of the envelope ( E ) gene ( 1485 nt ) was assembled for the Southeast Asian and Vietnamese isolates ( n = 134 and 751 , respectively ) to include the broadest range of locations . An initial neighbor-joining tree was constructed in PAUP* [39] , using a HKY85 nucleotide substitution model with gamma-distributed rates . This allowed us to make an initial identification of the major clades of DENV-1 in Viet Nam . These Vietnamese isolates were then subsampled ( n = 101 ) to explore their phylogeography in context of the South East Asian isolates . Isolation dates for the South East Asia data set were obtained from GenBank annotations and via personal communication . Where specific dates were not available in terms of day and month , a mid-point of the year of isolation was used . The spatial dynamics of DENV-1 in Southeast Asia were investigated with a discrete diffusion model [26] using Bayesian Monte Carlo Markov Chain ( MCMC ) method implemented in BEAST [40] . The phylogeography analysis was executed with a codon-structured SDR06 substitution model [41] , a relaxed uncorrelated lognormal clock [42] and a Gaussian Markov Random Field ( GMRF ) coalescent prior [43] over the unknown phylogeny . The discrete diffusion model used the country of isolation of the sampled sequences to reconstruct the ancestral location states of the internal nodes from the posterior time-scaled tree distribution . The MCMC was run for 50 million generations , sampling every 5000th state , and executed multiple times to ensure adequate mixing and stationarity had been achieved . Major clades of Vietnamese DENV-1 identified from the broad-scale South East Asian analysis were selected for further study to examine the spatial and temporal variation in Viet Nam . In clades with appreciable numbers of sequences from Dong Thap and HCMC , isolates from these locations were analyzed independently to gauge the regional variation in viral transmission patterns . For the fine-scale analysis , a continuous diffusion model based on a lognormal relaxed random walk [27] was employed to model the DENV-1 spatial dynamics in Viet Nam . For each isolate , the specific sample date and location information in terms of the longitude and latitude of the patient's household were used . Isolates that were identical in sample date and location information were down-sampled so as to reduce the potentially biasing effect of over-sampling of epidemiologically-linked cases . The MCMC runs were evaluated as previously described , and the chain lengths ranged from 50 to 100 million generations , and were sampled regularly to yield 10 , 000 trees from the posterior distribution . The viral dispersion rates ( km/yr ) for each data set were calculated across the tree ( i . e . total straight-line distance travelled divided by the total time ) and biannually to consider the spatial heterogeneity in a time-scaled framework . Plots of relative genetic diversity over time were reconstructed using the GMRF coalescent prior to reveal the association between the genetic diversity of each group in terms of their evolutionary history [43] . Further discrete phylogeography analyses were performed with the robust counting method [44] , [45] to determine the extent of viral migration between Dong Thap and HCMC and whether this varied when the lineage originated in a rural or urban area . In this case , the discrete states were represented by either the isolate being sampled from HCMC , Dong-Thap or neither ( non-Dong Thap or HCMC ) . For the limiting case of a freely mixing ( non-spatially structured ) epidemic in HCMC , dispersion rates were estimated whilst randomizing the tip locations during the tree proposal in the MCMC , whilst co-estimating the rates for each independent lineage and the joint DENV-1 diffusion rate . To determine the viral transmission network within HCMC , a non-reversible discrete phylogeography model was applied to all the HCMC isolates , using the district of isolation for the discrete states . The analysis was performed and evaluated as described above with the addition of implementing Bayesian Stochastic Search Variable selection ( BSSVS ) to identify significant transition rates between locations [26] . The transition rates supported by a Bayes factor of at least 3 were examined further by looking at the number of in-degree and out-degree per district . The number of connections was normalized by the number of samples from the source location in order to reduce the bias from under-represented locations in our data set . Patients ( or their parents/guardians ) gave written informed consent to participate in each of the studies . The study protocols were approved by the Hospital for Tropical Diseases and the Oxford University Tropical Research Ethical Committee . | Although dengue is a major cause of morbidity in many tropical and subtropical regions of the world , little is known about how the causative virus ( dengue virus , DENV ) spreads through endemic populations . To address this issue we undertook a phylogeny-based analysis of 751 complete genome sequences of DENV-1 sampled from patients in southern Vietnam during 2003–2008 . We show that multiple viral lineages co-circulate within the urban area of Ho Chi Minh City ( HCMC ) , and spread at approximately equivalent rates through overlapping geographical areas , suggesting that they are of equivalent fitness . We also observed that DENV-1 within HCMC tended to disperse from more to less densely populated regions , and that this city was the source population for DENV-1 in the rural area of Dong Thap . Despite the high prevalence of DENV-1 in southern Vietnam , viral dispersion rates were relatively low , especially in HCMC where they averaged less then 20 km/year . Such a low rate is consistent with predominantly mosquito-borne spatial dispersal of DENV-1 in this urban setting containing a large number of susceptibles . Together , these results suggest that dengue control measures such as insecticide spraying should be directed toward the most densely populated regions of localities where the virus is endemic . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"sequence",
"analysis",
"phylogenetics",
"emergence",
"evolutionary",
"biology",
"population",
"genetics",
"biology",
"computational",
"biology",
"gene",
"flow",
"evolutionary",
"systematics",
"evolutionary",
"processes",
"evolutionary",
"genetics"
] | 2011 | Endemic Dengue Associated with the Co-Circulation of Multiple Viral Lineages and Localized Density-Dependent Transmission |
We perform a large-scale study of intrinsically disordered regions in proteins and protein complexes using a non-redundant set of hundreds of different protein complexes . In accordance with the conventional view that folding and binding are coupled , in many of our cases the disorder-to-order transition occurs upon complex formation and can be localized to binding interfaces . Moreover , analysis of disorder in protein complexes depicts a significant fraction of intrinsically disordered regions , with up to one third of all residues being disordered . We find that the disorder in homodimers , especially in symmetrical homodimers , is significantly higher than in heterodimers and offer an explanation for this interesting phenomenon . We argue that the mechanisms of regulation of binding specificity through disordered regions in complexes can be as common as for unbound monomeric proteins . The fascinating diversity of roles of disordered regions in various biological processes and protein oligomeric forms shown in our study may be a subject of future endeavors in this area .
Many proteins and protein regions have been shown to be intrinsically disordered under native conditions; namely , they contain no or very little well-defined structure [1]–[6] . Intrinsically disordered proteins ( IDPs ) have been found in a wide scope of organisms and their disorder content was shown to increase with organism complexity [7]–[11] . Comparative analysis of the functional roles of disordered proteins suggest that they are predominantly located in the cell nucleus; are involved in transcription regulation and cell signaling; and also can be associated with the processes of cell cycle control , endocytosis , replication and biogenesis of cytoskeleton [10] , [12] . IDPs have certain properties and functions that distinguish them from proteins with well-defined structures . 1 ) IDPs have no unique three-dimensional structure in an isolated state but can fold upon binding to their interaction partners [1] , [4] , [13]–[18] . 2 ) Conformational changes upon binding in proteins with unstructured regions are much larger than those in structured proteins [1] . 3 ) The conformations of disordered regions in a protein complex are determined not only by the amino acid sequences but also by the interacting partners [1] , [19] . 4 ) IDPs can have many different functions and can bind to many different partners using the same or different interfaces [20] . 5 ) IDPs can accommodate larger interfaces on smaller scaffolds compared to proteins with well-defined structure [14] , [21] , [22] . 6 ) IDPs typically have an amino acid composition of low aromatic content and high net charge as well as low sequence complexity and high flexibility [2] , [10] , [23] . 7 ) Intrinsic disorder provides for a rapid degradation of unfolded proteins , thereby enabling a rapid response to changes in protein concentration ( regulation through degradation ) [24] . 8 ) Finally , intrinsic disorder offers an elegant mechanism of regulation through post-translational modifications for many cellular processes [20] , [25] . Predictions of disorder in proteins take into account the characteristic features of unstructured proteins and have been shown to be rather successful , especially in the case of large regions . According to the results of CASP7 ( 7th Community-Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction ) , the best prediction groups successfully identified 50–70% of the disordered residues with false positive rates from 3% to 16% [26] . Prediction methods aim to identify disordered regions through the analysis of amino acid sequences using mainly the physico-chemical properties of the amino acids [23] , [27]–[36] or evolutionary conservation [12] , [37]–[39] . As protein interactions are crucial for protein function ( [40] , references within ) , the biological role of disordered proteins should also be studied in this context . Indeed , folding of disordered proteins into ordered structures may occur upon binding to their specific partners [1] , [4] , [13]–[17] which may allow disordered regions to structurally accommodate multiple interaction partners with high specificity and low affinity [1] , [41]–[43] . Moreover , it has been shown that the binding mechanism , whether binding occurs between folded or unfolded chains , depends on the structural characteristics , interface properties , and degree of minimal frustration of monomers [21] , [44] . Binding through unfolded or partially unfolded intermediates can provide a kinetic advantage through the “fly-casting” mechanism [19] . According to this mechanism a dimensionality reduction occurs when the folding of a disordered protein is coupled with binding , thereby speeding up the search for specific targets . A database of continuous protein fragments ( Molecular Recognition Features or MORFs ) has been compiled from the Protein Data Bank to include short protein chains ( with fewer than 70 residues ) bound to larger proteins [45] , [46] . It has been argued that MORFs participate in the coupling of binding and folding , a hypothesis that was supported by the analysis of the composition and predicted disorder of MORF segments . As a result of studying the subtle structural differences of the same proteins in different conditions and functional states , many so-called “dual personality” protein segments were found able to exist in both ordered and disordered states [47] . There is a continuous range between completely structured and completely disordered proteins in which intermediate cases are rather common [24]: proteins that are disordered but compact , multi-domain proteins with disordered linkers , and ordered proteins with some local disorder . Examples of proteins with intrinsically disordered regions which exhibit coupling between folding and binding have been described in the literature previously [1] , [4] , [13]–[18] . Nevertheless , the universality of this phenomenon and functional importance of many disordered regions remains unclear . The question can be expanded further to how much intrinsic disorder do protein complexes contain and what is its functional importance ? To answer these questions we examine observed and predicted disorder in protein complexes and unbound proteins using a large-scale dataset of protein structures . The atomic details of structures and the conserved binding mode analysis introduced earlier [48] allow us to monitor changes happening on or near interaction interfaces and to infer their functional importance .
Figure 1 presents a flowchart of the assembly of the dataset . From the Protein Data Bank ( PDB ) [49] we selected X-ray structures with resolution better than 3Å . We assigned domains from the Conserved Domain Database ( CDD ) [50] on each protein structure chain using RPS-BLAST [51] with default parameters ( E-value≤0 . 01 ) . As we focus on protein-protein interactions ( interactions between different protein chains ) we ensured that each chain has only one CDD domain which covers at least 70% of the full chain sequence . Among overlapping domain assignments , the domain with the longest footprint was chosen where the footprint region extends from the first to the last residue in the alignment mapping a CDD family to a given chain . Once CDD families are assigned , we identify all interacting chains within a PDB entry . Two chains qualify as interacting if they have at least 5 residue-residue contacts . A contact takes place between a residue from one chain and a residue from the other when the distance between any non-hydrogen atom of one residue is within 6 Å of any non-hydrogen atom of the other residue . The set of residues which make contacts between the chains form the interface . To ensure that interactions are biological and not spurious , such as from crystal packing , we remove interactions that are not confirmed with additional instances of the same family pair interacting in the same orientation , so-called Conserved Binding Modes ( CBM ) [48] . These CBMs are defined using structural alignments between different structural instances of the same interacting family pair to confirm overlap of at least 50% of interface residue positions ( Figure 2 ) . Two definitions of conserved binding modes ( CBMs ) have been used: in one case confirmation of a binding mode can occur only between different non-redundant structures; in the other case recurrent interactions might occur within one structure . We refer to a dimer of interacting chains with a distinct CBM as a “complex” although it includes only pairwise interactions and several such “complexes” can be found in one PDB entry . While analyzing disorder in dimer complexes , we also compare their disorder content with the fraction disorder of the protein in a monomeric state ( Figure 1 ) . Monomer and complex chains ( as defined in PDB ) corresponding to the same domain family were aligned to ensure 100% sequence identity in the non-gapped alignment . Their alignment was extended beyond the CDD footprint region as far as possible . In 95% of all cases the alignment was extended to include the entire shorter chain and in 75% of cases the alignment was extended to include both entire chains from monomer and complex structures ( within 1–2 residues from both ends ) . The alignments are more extensive than footprint regions and cover footprint regions plus C- and N- terminal sequence regions which often do not have coordinates . Biological unit assignments were taken from the PDB asymmetric unit ( ASU ) assignments and from PISA predictions of multimeric states ( which are based on calculation of stability of multimeric states inferred from the crystalline state ) [52] . We cannot directly investigate the disorder on the interfaces in complexes as complexes are defined through residue contacts so those interface residue coordinates must be present in PDB files ( see definitions of disorder below ) . As shown in Figure 2 , disorder on the interfaces can be inferred by exploiting monomeric states of proteins , using their alignment to map the interface region from a complex onto the monomers . Given the overall numbers of disordered and non-disordered residues in the alignment , the number of residues on the mapped interface and the number of disordered residues on the interface , we can estimate the probability of observing a given number ( or higher ) of disordered residues on the mapped interfaces purely by chance . Using the binomial test we calculated p-values for all complexes with at least five disordered residues in the footprint or aligned regions and at least one disordered residue on the mapped interface ( altogether there are 55 complexes for which interface p-values can be calculated ) . After excluding those cases where interfaces are entirely outside of the alignment , our data set contained 4 , 884 dimer complexes and 418 unique monomer structures . Since multiple protein chains can be found in the same PDB entry ( on average four chains per PDB entry from our test set ) and these chains may belong to the same family , we performed an averaging of all observed quantities over the members of the family and conserved binding modes . Namely , as shown in Figure 2 , disorder content observed in family type X was averaged over all instances ( structures ) of family X interacting with family type Y through a specific CBM . Hereafter we refer to them as “CBM interactions” or merely “interactions” . Overall , we ended up with 588 CBM interactions ( “test588” ) . To compare disorder content in monomeric and complex states we used the more strict definitions for both binding modes and oligomerization states ( see previous section ) . If we use the more strict CBM definitions and restrict the monomeric states by PISA ( those structures which are monomeric in ASU are also predicted to be monomeric by PISA ) the set is reduced to 149 interactions ( “test149” ) . Also , for each protein used in our test set we retrieve the Gene Ontology ( GO ) functional annotations [53] . All structures , protein families , disorder content , GO functional annotations and other relevant information are provided in the Supporting Information . Disordered regions were defined as those regions with missing coordinates in X-ray-resolved structures . This is the most direct way to observe intrinsically disordered regions although largely disordered proteins may be underrepresented in PDB because of the difficulties in their crystallization [5] . Disordered regions were also predicted as those with low packing density using the FoldUnfold described previously [31] , [32] . Some advantages of the FoldUnfold method are that the program was not trained on the missing coordinates in PDB and that it reports a very high specificity ( small number of false positives ) . Its performance has been shown to be comparable to other disorder prediction methods [31] , [54] . ( See also Table S2 ) . According to FoldUnfold , an average packing density observed in structures was computed for each of the 20 amino acid residues . These values were considered to be the expected packing density for the same type of residues in a query protein ( with or without known structure ) . Using a sliding window of 11 residues , the center residue of each window is predicted to be disordered if the mean packing density of the window falls below a threshold . We performed disorder predictions for all proteins in our data set . To differentiate between ordered regions ( hinge-like movements or “wobbly” domains , for example ) with missing PDB coordinates and true disordered regions , we annotated those regions which are both predicted to be disordered and at the same time have missing coordinates in PDB . They will be referred hereafter as “confirmed disordered regions” . To quantify the disorder content , we calculated the “fraction disorder” as a ratio of the number of residues in disordered regions and the number of residues in the footprint or aligned regions . To see all computed values of fraction disorder consult Dataset S1 ( missing coordinate definition ) and Dataset S2 ( confirmed disordered regions ) .
Analysis of fraction disorder in different families shows that one quarter of our test complexes do not have any disorder while others can have as much as one third of their residues in the disordered state ( Figure 3 ) . The three quarters of complexes with non-zero disorder have on average 4 . 3% disorder in the aligned regions and about 1 . 6% in the footprint regions . Confirmed disordered regions have similar disorder content for pairs with non-zero disorder and drops to about 1% if all 588 interactions are included . The reason is that disordered regions with missing coordinates sometimes do not overlap with the predicted disordered regions . There are also families that exhibit rather wide variation in fraction disorder among different members of these families ( a ratio of standard deviation over the mean value of fraction disorder is greater than 1 ) ; they constitute 13% of all cases . Table 1 shows several cases of complexes with disorder that were confirmed by experimental studies to be functional . Proteins from these families are found to function in dimer , tetramer and other oligomeric states . Their disordered regions play important roles in regulating the specificity of interactions between the dimer complexes and their interacting partners , in establishing the links between different residues upon allosteric regulation , and possibly in kinetics . In this table we highlight the generality of this phenomenon for many different proteins including enzymes , chaperones and others . As can be seen from this table , all cases ( except for the last one ) constitute homodimer complexes and , as will be shown in the next section , homodimers have a tendency to contain larger fractions of disordered regions compared to heterodimers . References for Table 1 can be found in Table S1 ( a ) . Here we describe in detail one example from the table: a complex of heat shock protein hsp31 which has chaperone activity and functions as a homodimer in solution ( 1PV2 [55] ) ( Figure 4 ) . The complex contains four dimers in a triclinic cell exhibiting a conserved symmetrical homodimer binding mode . Structures of the homodimers show significant fraction disorder of about 8–9% in both aligned and footprint regions . Disordered regions D2 and D3 are found at positions 27–49 and 109–115 and part of the first and the entire second region are also predicted to be disordered by the sequence-based method [32] . These regions have particular functional importance as they are located close to the dimer interface and at high temperatures become disordered and expose a large hydrophobic interface area that helps in binding to client proteins [55] . When the temperature decreases , D2 and D3 lock in certain conformations and facilitate the removal of the client protein from the hydrophobic patch . We performed an analysis separating all interacting pairs from our test set into homo- ( 535 complexes ) and heterodimers ( 53 complexes ) , where both chains in a pair are classified as belonging to the same or different families respectively . Similarly , the prevalence of homodimers over heterodimers in a cell was reported previously [56] . All homodimers were separated into symmetrical and non-symmetrical classes ( “isologous” and “heterologous” according to [57] ) . We define symmetrical homodimers as those that use more than 80% of the same surface in both subunits for binding ( 316 complexes ) ; all other homodimer arrangements were defined as non-symmetrical ( 266 complexes ) . Some homodimer families have structures belonging to both symmetrical and non-symmetrical classes ( near the 80% cutoff ) but such cases are rare . Eleven families form both homo- and heterodimers . The majority of such cases are examples of larger complexes where the same protein participates in homo- and hetero-interactions within the same complex . Figure 5 shows average fraction disorder in different classes of homo- and heterodimers . As can be seen from this figure , fraction disorder in complexes decreases as the interaction interface deviates more from being a symmetrical homodimer interface . Fraction disorder in heterodimers is almost two times smaller compared to symmetrical homodimers and the difference is statistically significant ( p-value<0 . 001 ) . The observed trend for hetero- and non-symmetrical homo-complexes to contain smaller disordered regions was confirmed by the disorder prediction analysis , although the trend is not as pronounced for predicted disorder in aligned regions . We did not find significant differences in fraction disorder between homo- and heterodimers for proteins that participate in homo- and hetero-interactions within the same complex . In studying disorder in protein complexes , we can use the monomer states of the proteins as references . First we would like to check whether the disorder-to-order transition may occur upon binding; and second , to analyze if this transition happens on binding interfaces . In this section we compared fraction disorder of proteins in their monomer and complex states . By definition , binding interfaces should involve only residues with coordinates and therefore can introduce bias toward ordered regions in the complexes ( complexes with the entire interface disordered are not considered in the analysis ) . Therefore , for fair comparison between monomers and complexes we subtracted the number of disordered residues in a monomer which are mapped onto interfaces in a complex from the overall number of disordered residues in a monomer . Figure 6 shows fraction disorder in aligned regions for monomer and complex structures of the same interaction using the “test588” and “test149” sets . As can be seen from this figure , there exist three types of behavior: cases with higher fraction disorder in a monomer compared to the complex , cases with higher fraction disorder in a complex and , finally , those interactions with no preference towards disordered or ordered states in a monomer or a complex . It should be mentioned that different ways of averaging over structures or using confirmed disorder regions does not change the overall result , namely , that there are three groups and that the sizes of the first and second groups are comparable . While in the previous section we focused on the disordered regions spanning the whole aligned or footprint regions , here we will focus on disorder in the interface regions . Since the interface in complexes is ordered by definition , we looked at disordered regions in monomers which are aligned to the interface region of the same protein in a complex . The monomer reference state gives us an opportunity to analyze the disorder in the regions of a monomer which form the interface upon binding . We found that the mapped ( inferred ) interface regions can be up to 50% disordered in a monomer and for 42% of the complexes ( 23 out of 55 complexes for which p-values can be calculated , see Methods ) , there is a statistically significant bias toward the disorder on inferred interface regions with p-values of less than 0 . 05 . We observed similar fractions of cases with significant disorder on inferred interfaces if we use confirmed disorder regions ( see Methods ) . Additional restriction of monomeric states by PISA indicates 75% of the cases have significant disorder on interfaces ( 9 out of 12 complexes from “test149” used for p-value calculation ) . Several cases with significant disorder on inferred interfaces are listed in Table 2 ( and in Table S1 ( b ) to include references ) . Their disordered regions predicted by FoldUnfold and by five other methods are highlighted in Table S2 . Figure 7 shows one example of ubiquitin C-terminal hydrolase in two states: monomeric ( 1UCH [58] ) and in complex ( 1XD3 [59] ) with ubiquitin vinylmethylester , a ubiquitin-based active site-directed probe . Ubiquitin C-terminal hydrolase catalyzes the hydrolysis of the isopeptide linkage between the C-terminal glycine of ubiquitin and a lysine of the target polypeptide . The structure of the free form of this enzyme has 4–6% fraction disorder in footprint and aligned regions compared to only 0–0 . 9% fraction disorder in the complex with ubiquitin . The disordered region in 1UCH constitutes a 20 residue loop ( 147–166 ) which is also predicted to be disordered ( region 150–164 ) by the sequence-based method [32] . This disordered loop is positioned just over the active site cleft and becomes ordered upon binding to ubiquitin vinylmethylester . The interaction interface mapped from complex structure to monomer shows that 30% of the interface is disordered in a monomer ( binomial p-value<10−8 ) which points to the coupling between folding and binding . It was suggested earlier that this disordered loop might prevent access to the active site for larger substrates and affect substrate specificity as larger substrates could only be accommodated in the active site by peeling away this loop from the active site cleft [58] , [59] .
Our large-scale study of disordered regions in proteins and protein complexes underscores a fascinating diversity among the biological processes that make use of protein disorder . Analysis of GO functional annotations of complexes reveals a variety of categories where intrinsic disorder can play an important functional role , the most frequent of them being nucleic acid binding proteins , enzymes , ATP binding proteins , receptor binding proteins and other ligand binding proteins ( see Dataset S3 ) . In addition to well-documented cases of signaling and transcription related proteins , we detect and describe intrinsic disorder in a large variety of enzymes and other proteins . In accordance with the conventional view that folding of disordered regions occurs upon binding to the interaction partners , we find many such cases in our analysis where ordering occurs upon complex formation . Moreover , we investigated the details of protein interaction interfaces and deduced changes occurring on the interfaces in disorder-to-order transitions . We find that in 42–75% of interactions ( for which statistical significance could be estimated ) , there is evidence that disorder-to-order transition occurs on binding interfaces . Many complexes in our dataset have significant amounts of intrinsic disorder . The role of disordered regions in complexes has been analyzed in several previous studies on smaller test sets [22] , [60] . In our study we find as many cases with disorder in complexes as the number of instances of disorder-to-order transition upon binding . This is a rather unusual result as many such cases until recently were largely overlooked . It has been proposed that disordered regions can be energetically beneficial in proteins and their complexes due to a number of reasons: they can provide an increase in backbone conformational entropy upon ligand binding , can accommodate sites for post-translational modifications , and can provide interfaces for binding other partners [6] , [22] , [60]–[65] . In addition , the formation of complexes of proteins containing functionally important disordered regions can help to increase their stability ( entropy-driven complexation , see the last section ) and prevent their degradation . Many proteins perform their functions while interacting with each other in larger complexes . We argue that intrinsic disorder in complexes may play an important functional role in regulating the specificity of interactions between the dimer complexes and their interacting partners , in establishing the links between different residues upon allosteric regulation , and in possibly influencing the kinetics . For example , the mechanisms of regulation of binding specificity through disordered regions in complexes can be as common as for unbound proteins: controlling the exposure of the dimer interface or nearby regions for potential binding targets , or providing specific binding for substrates of certain sizes . The former mechanism has been recently investigated in the stable symmetrical homodimers , UmuD2 and UmuD2′ , which lack secondary structure and might lock the disordered regions in conformations that facilitate further binding of other proteins [66] . In addition , the formalism of flexible folding and mechanism of the “conformational selection” model [19] , [67]–[72] can be expanded to include the binding between protein complexes and their interacting partners . Interestingly , we find that the disorder content in homodimers , especially in symmetrical homodimers , is significantly higher than in heterodimers . Indeed , many soluble and membrane-bound proteins form homo-oligomeric complexes in a cell and oligomerization can generate new binding sites at dimer interfaces to increase specificity and diversity in the formation of complexes . Indeed , intrinsic disorder in homodimers might have more pronounced functional importance compared to the disorder in heterodimeric complexes . Symmetrical arrangements in homodimers might be crucial to keep functional disordered regions close together in space to form joint binding interfaces or to form near-interface regions to regulate the accessibility of the binding partner . Moreover , from the energetic point of view , symmetrical homodimers have an advantage over non-symmetrical arrangements [73] , [74]; at the same time , self-interactions between disordered parts in homodimers can be of evolutionary and functional importance [66] , [75] . Another explanation comes from thermodynamics considerations . Entropy of complexation gives an important contribution to the complex stability and drives macromolecular complexes to less symmetric states . Any rearrangement of monomers that decrease complex symmetry would therefore result in a more stable complex ( see Eq . 20 in [52] ) . The presence of disordered regions in the symmetrical homodimers will make the protomers asymmetric and change the symmetry number γ from 2 to 1 ( two-fold symmetry to asymmetry ) and make a favorable contribution to the free energy . At the same time disordered regions should not affect symmetry numbers in cases of heterodimers or non-symmetrical homodimers ( they are asymmetric by default ) and will not change their stability . Ultimately , the interplay between the binding energy and entropy contributions is important and it is not unrealistic that the entropy-driven disordered complex formation can be realized in some cases . It is difficult to systematically account for all factors which influence the fraction disorder in proteins . The amount of disorder in crystals depends in general on crystallization conditions and crystal packing parameters . The balance between order and disorder is rather subtle and is difficult to detect but the evidence pointing to the tremendous importance of intrinsic disorder in a large variety of cellular processes is accumulating and merits further study . | Traditionally , protein structure is believed to determine function . Recently , it was observed that many proteins contain regions without well-defined structure ( intrinsically disordered regions ) , including a large fraction of eukaryotic proteins . Intrinsic disorder has been associated with particular functions including cell regulation; signaling; and protein , DNA , and ligand binding . Many proteins are intrinsically disordered in native form and fold upon binding , following the conventional paradigm . Accordingly , disorder in a protein may facilitate binding to multiple partners . However , in some cases disorder has also been found in the bound state . To gain clearer insight into the functional importance of disorder regions in protein complexes , we perform a large-scale analysis of disorder using protein structures in complex and in unbound forms . We show that disorder in protein complexes is rather common and pinpoint changes that occur upon protein binding at interaction interfaces . By illustrating a variety of functional roles for disorder in specific proteins , we emphasize the versatility and importance of this phenomenon . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"biophysics/structural",
"genomics",
"biophysics/protein",
"folding",
"biophysics/biomacromolecule-ligand",
"interactions",
"evolutionary",
"biology/bioinformatics",
"computational",
"biology/systems",
"biology"
] | 2009 | Intrinsic Disorder in Protein Interactions: Insights From a Comprehensive Structural Analysis |
Melioidosis , caused by Burkholderia pseudomallei , is an endemic disease in Bangladesh . No systematic study has yet been done to detect the environmental source of the organism and its true extent in Bangladesh . The present study attempted to isolate B . pseudomallei in soil samples and to determine its seroprevalence in several districts in Bangladesh . Soil samples were collected from rural areas of four districts of Bangladesh from where culture confirmed melioidosis cases were detected earlier . Multiple soil samples , collected from 5–7 sampling points of 3–5 sites of each district , were cultured in Ashdown selective media . Suspected colonies of B . pseudomallei were identified by biochemical and serological test , and by polymerase chain reaction ( PCR ) using 16s rRNA specific primers . Blood samples were collected from 940 healthy individuals of four districts to determine anti- B . pseudomallei IgG antibody levels by indirect enzyme linked immunosorbent assay ( ELISA ) using sonicated crude antigen . Out of 179 soil samples , B . pseudomallei was isolated from two samples of Gazipur district which is located 58 km north of capital Dhaka city . Both the isolates were phenotypically identical , arabinose negative and showed specific 550bp band in PCR . Out of 940 blood samples , anti- B . pseudomallei IgG antibody , higher than the cut-off value ( >0 . 8 ) , was detected in 21 . 5% individuals . Seropositivity rate was 22 . 6%-30 . 8% in three districts from where melioidosis cases were detected earlier , compared to 9 . 8% in a district where no melioidosis case was either detected or reported ( p<0 . 01 ) . Seropositivity increased with the advancement of age from 5 . 3% to 30 . 4% among individuals aged 1–10 years and > 50 years respectively . The seropositivity rates were 26 . 0% and 20 . 6% in male and female respectively , while it was 20–27% among different occupational groups . No significant association was observed with gender ( χ2 = 3 . 441 , p = 0 . 064 ) or any occupational group ( χ2 = 3 . 835 , p = 0 . 280 ) . This is the first study demonstrating the presence of B . pseudomallei in the environmental ( soil ) samples of Bangladesh . It also suggested that a large proportion of people , residing in these districts , were exposed to the organism .
Melioidosis is an endemic disease of public health and clinical importance in tropical and subtropical regions of the world [1] . It is caused by a Gram negative saprophytic bacterium called Burkholderia pseudomallei . Infection occurs through skin and by inhalation when susceptible individuals are exposed to contaminated water and soil [2 , 3] . Melioidosis accounts for about 20% of all community-acquired septicemias in north-eastern Thailand and 2000 to 3000 new cases are diagnosed every year [4 , 5] . Multiple cases have also been reported from India and several countries of South-East Asia , the Middle East , Africa and South America [6] . In 1964 , melioidosis was reported in a foreign sailor who was travelling through Bangladesh [7] . However , the first case of melioidosis in Bangladesh was diagnosed in a native Bangladeshi infant in 1988 [8] . Later on several cases of melioidosis were reported up to 2014 [9] . From 1991 to 1999 , five cases were detected in United Kingdom ( UK ) among Bangladeshi people who immigrated to UK from the Sylhet region ( a northeastern district ) of Bangladesh [10–12] . In 2001 , we reported the second culture confirmed suppurative melioidosis case in a 48 years old diabetic patient who came from Sherpur district of Bangladesh [13] . The district is located about 140 km north of capital Dhaka . Later on , at least 10 cases were detected among the diabetic patients at Bangladesh Institute of Research and Rehabilitation in Diabetes , Endocrine and Metabolic Disorders ( BIRDEM ) Hospital , Dhaka from 2009 to 2014 [9] . Analyses of the reported cases strongly indicate that the disease is potentially endemic in ten districts of Bangladesh particularly in northern and northeastern parts of the country . Recently in 2013 , melioidosis endemic countries of the world have been categorized into ‘definite’ and ‘probable’ country based on the presence of B . pseudomallei in humans and in the environment in the respective countries [6] . According to the above categorization , Bangladesh falls into ‘probable’ category of country as the presence of the organism in the environment has not yet been identified or reported even though several culture-confirmed melioidosis cases have been detected . Probability of the presence of B . pseudomallei in soil and water of Bangladesh is very high as the climatic condition of the country is favorable for its growth in the environment . Therefore , isolation and identification of B . pseudomallei from environmental samples ( e . g . soil or water ) is important to determine the source of the organism of melioidosis cases in the country . The true extent of the disease in Bangladesh is not known , as this disease is not familiar to most of the physicians and microbiologists of the country . Seroepidemiological studies showed that 80% of children in north-eastern Thailand were positive for antibodies against B . pseudomallei by the age of 4 years [1] . In Malaysia , reported seroprevalence in healthy individuals was 17–22% among rice farmers and 26% in blood donors [14] . In north Australia 0 . 6 to 16% of children had evidence of infection by B . pseudomallei [15] . A hospital based serological survey in Bangladesh reported 28 . 9% seropositive rate for B . pseudomallei antibody among patients attending several tertiary care hospitals for unrelated ailments . The study , however , used a very low cut off titer ( 1:10 ) of indirect haemagglutination assay ( IHA ) for defining seropositive cases without considering the presence of cross reactive background antibody among the local population . The study did not investigate the possible source of the organism [16] . In view of the above , detection of B . pseudomallei in the soil samples and determination of anti-pseudomallei antibody in healthy population would help to establish the environmental source of the organism as well as the extent of its exposure in Bangladesh . So far , no systematic study has been done to find out the presence of organisms in environmental samples of Bangladesh . Therefore , the present study was designed for detection of B . pseudomallei by culture and molecular method from soil as well as to determine the extent of exposure by detecting antibodies to B . pseudomallei among the healthy population of four districts of Bangladesh .
The Ethical Review Committee ( ERC ) of the Diabetic Association of Bangladesh ( BADAS ) has approved the study . Ibrahim Medical College ( IMC ) is an institution under the BADAS and its ERC is the approval body for research protocols of IMC . Informed written consent was obtained from all adult participants ( age 18 years and above ) and from the parents/guardians of all children ( age up to 17 years ) prior to collection of blood samples and demographic data . Soil samples were collected from rural areas of four districts of Bangladesh with diagnosed melioidosis cases [9] . Three districts namely Mymensingh , Sylhet and Gazipur are situated in the north and northeast of capital Dhaka city while one district ( Narayangange ) is located south of Dhaka city . The locations of soil sampling district and their distance from capital Dhaka is shown in Fig 1 . In each district , 3–5 sites were selected for collection of soil samples . Each sampling site was about 5 km apart from the next sampling sites . At each site 5-7sampling points were identified which were about 30 meters apart from each other . The preferred collection site was moist area within a rice field . Approximately 200 g soil was taken from each point from a depth of about 20–30 cm using a shell augur disinfected with 70% alcohol in between soil collection . Collected soil was placed into a sterile plastic bag and sealed with rubber band to prevent moisture loss and was transported to the laboratory as soon as possible . All the soil samples were collected and processed for culture from June to September 2011 . Soil samples were processed for culture as described by Brook et al [17] . Twenty grams of soil were mixed with 40 ml sterile distilled water and the suspension was shaken vigorously for one minute and allowed to settle for 5–10 minutes . The supernatant fluid was collected . For enrichment , 1 ml of supernatant fluid was inoculated into 9 ml of modified Ashdown’s selective enrichment broth ( ASB ) and incubated at 37°C for 48 hours [18] . After enrichment , 10 μl of broth was streaked onto modified Ashdown`s selective agar ( ASA ) medium . The plates were incubated for 48–72 hours to allow typical colonies to grow . Purple colored dry , wrinkled and oxidase positive colonies were then sub cultured on MacConkey`s agar medium and incubated at 42°C . The organisms which grew on MacConkey`s agar medium at 42°C were identified as B . pseudomallei by typical colony morphology , Gram staining ( bipolar staining ) , motility , biochemical tests ( including API 20NE ) , arabinose assimilation and resistance to colistin and aminoglycoside [19] . Monoclonal antibody based latex agglutination test ( Melioidosis Research Center , Khon Kaen , Thailand ) was performed for the final identification and confirmation of the suspected colonies of B . pseudomallei . Serologically confirmed B . pseudomallei isolates from soil samples were further confirmed by PCR using specific primers constructed from 16s rRNA region of B . pseudomallei [16] . The primers were constructed from 16s rRNA region of B . pseudomallei to amplify a fragment of 550 bp in length . Primers were—PPM3 forward primer ( 5`AATCATTCTGGCTAATACCCG 3` ) and PPM4 reverse primer ( 5`CGGTTCTCTTTCGAGCTCG 3` ) . Total genomic DNA was prepared by RealLine DNA Extraction Sample Kit” ( BIORON Diagnostics GmbH , Germany ) . Briefly , 500 μl preheated ( at 56°C ) lysis reagent with sorbent was added to 100 μl bacterial suspension . The tube was vortexed for 10 seconds followed by incubation in thermo shaker for 10 minutes at 1300 rpm . DNA/RNA solution was added to the tube and again vortexed and centrifuged at 13000 x g for 5 minutes . Supernatant was discarded and Wash Solution No 1 was added to the tube , vortexed and centrifuged at 13000 x g for 5 minutes . Supernatant was again removed and 300 μl Wash Solution No 2 was added to the tube , vortexed and centrifuged at 13000 x g for 5 minutes . Supernatant was discarded and the pellet was dried by opening the cap for 2–3 minutes at room temperature . 200 μl specimen diluents was added to the air dried tube and vortexed vigorously for 10 seconds followed by incubation at thermo shaker for 10 minutes at 56°C at 1300 rpm and then centrifuged at 13000 x g for 1 minute . Supernatant containing DNA was finally collected and stored at -20°C until used . PCR amplification was carried out in a 25 μl final volume containing 2 . 0 μl DNA , 2 . 5 μl 1 x PCR buffer , 1 . 5 mM MgCl2 , 25 μM of each dNTP , 10 pM of each primer , and 1 . 25 unit of Taq DNA polymerase enzyme . Samples were subjected to initial denaturation at 94°C for 2 minutes followed by denaturation at 94°C for 60 sec , primer annealing at 55°C for 60 sec and extension at 72°C for 90 sec . Final extension was for 10 minutes at 72°C . Amplification was performed in Master Cycler ( Eppendorf ) programmed for 35 cycles . Amplified PCR product was analyzed by electrophoresis in 1 . 5% agarose gel containing ethidium bromide ( 0 . 5 μg/ml ) in TBE buffer ( 0 . 04 M Tris acetate , 0 . 001 M EDTA , ( pH 8 . 6 ) and photographed under UV illumination . The bands were compared to the band obtained with a positive B . pseudomallei DNA control . In all assays , DNA from known B . pseudomallei was included as positive control . A tube without DNA served as no template DNA control . All suspected soil and clinical B . pseudomallei isolates from the present and previous study [9] were sent to the Emerging Pathogens Institute ( EPI ) , University of Florida , USA for species identification using type III secretion system ( TTS1 ) assay [20] Relatives or attendants of patients attending the rural healthcare facilities of the four districts namely Mymensingh , Sylhet , Narayangange and Kishoregange were recruited for determining the anti- B . pseudomallei antibodies ( Fig 1 ) . Blood samples were collected from 940 healthy individuals with no history of fever , persistent cough , wasting or suppurative lesion . Age , sex and socio-economic conditions were recorded . In order to determine the cut off optical density ( OD ) value of ELISA test , 51 healthy newborn babies of Dhaka city were enrolled in the study . About 1–2 ml of venous blood was collected from each individual with proper aseptic technique . Serum samples from 10 culture confirmed melioidosis cases admitted at BIRDEM Hospital were included in this study as positive controls . Serum anti- B . pseudomallei IgG antibody was determined by an indirect ELISA as described by Voller et al [21] . To prepare sonicated antigen , 50 ml of Trypticase Soya Broth ( TSB ) was inoculated with pure colonies of B . pseudomallei USM strain and incubated overnight at 37°C . Organisms were harvested by centrifugation for 30 minutes at 4000 x g at 10°C . Pellets were suspended with 3 ml of 25 mM Tris-HCL ( pH 7 . 4 ) and washed three times with Tris-HCL for 30 minutes at 4000 x g at 10°C . Deposited pellet , suspended in 5 ml of ice-cold Tris-HCL , was sonicated at 40W for 8 minutes in each pulse inside the assigned biosafety cabinet . Sonicated bacterial suspension was then centrifuged at 5000 x g at 10°C for 30 minutes . After centrifugation , the supernatant containing the bacterial proteins was collected and its protein concentration was determined . The 96 well EIA plate ( Linbro , USA ) was coated with sonicated antigen 10 μg/ml in 0 . 5 M carbonate/bicarbonate buffer ( pH 9 . 6 ) . To each well 100 μl volume of coating buffer was added and incubated overnight at 4°C . The plate was washed three times with PBS-0 . 05% Tween 20 ( PBS-T , pH 7 . 4 ) ) and blocked by incubating for 2 hrs with PBS-T containing 2% BSA at 37°C . The plate was then washed three times with PBS-T . A volume of 100 μl serum ( 1:1600 dilutions ) sample was added into each well and incubated for 4 hours at 37°C . After washing with PBST three times , 100 μl of horseradish peroxidase conjugated anti-human IgG antibodies ( 1:4000 ) was added and incubated at 37°C for 2 hours . After washing three times with PBST , 50 μl of TMB substrate was added to each well and incubated at room temperature for 30 minutes in dark . Then 50 μl of 1 M sulfuric acid was added in each well . The colour developed was measured by EIA plate reader ( Human ELISA Reader ) at 450 nm . Optimum concentration of the antigen ( 10 μg/ml ) and serum dilution ( 1:1600 ) was predetermined by checkerboard titrations . A cut off OD values for anti- B . pseudomallei IgG antibody was determined to find out the exposure rate to B . pseudomallei in the study population . ELISA was performed with sera from 51 healthy newborn babies of Dhaka city who were presumed not to be exposed to B . pseudomallei . The mean OD + 3xSD of these sera were taken as cut-off OD value to determine the exposure rate . The mean OD±SD of the 51 healthy newborn babies were 0 . 2±0 . 2 . Therefore , the calculated cut-off OD value was 0 . 8 ( 0 . 2+3x0 . 2 ) . Any sample showing OD above this cut-off value of 0 . 8 was considered positive and referred to as exposed to B . pseudomallei infection . The mean OD value of ten culture positive cases ( positive control ) was 2 . 26±0 . 2 . To determine the specificity of anti- B . pseudomallei IgG by ELISA , a sub-set of 24 known positive serum samples were adsorbed with whole cell killed Pseudomonas aeruginosa and B . pseudomallei USM strain ( 1x 108 organisms/ml ) by incubating overnight at 4°C . The adsorbed serum samples were centrifuged at 10 , 000 x g for 5 minutes to remove the bacteria . Anti- B . pseudomallei IgG antibody was then determined in adsorbed serum by ELISA as described above . Decline of antibody concentration in terms of OD values after adsorption with B . pseudomallei indicated presence of specific antibody to B . pseudomallei in serum samples while decline with P . aeruginosa indicated antibodies cross reacting to pseudomonas antigens . The adsorption assay showed that mean antibody level of the positive sera reduced significantly , after adsorption with B . pseudomallei compared to pre-adsorbed value from OD 1 . 1 to 0 . 6 ( Fig 2 ) . The mean OD value decreased below the cut off OD of 0 . 8 following absorption . But , the OD value decreased insignificantly from 1 . 1 to 0 . 9 following adsorption with P . aeruginosa
Total 179 samples from four districts with diagnosed melioidosis cases were tested for the presence of B . pseudomallei . Out of 179 soil samples , 87 yielded growth of oxidase positive non-fermenting , aminoglycoside and colistin resistant suspected colonies on the Ashdown selective media after enrichment at 42°C . Out of these suspected isolates , only two isolates ( K23 and K35 ) were identified as B . pseudomallei ( Table 1 ) . These two isolates were finally confirmed as B . pseudomallei by specific monoclonal anti-sera to B . pseudomallei and polymerase chain reaction ( Fig 3 PCR gel ) . These two soil and all the clinical isolates ( ten ) were positive for TTS1 ( S1 Table and S1 Fig ) . Both the isolates were arabinose negative suggesting that they were not B . thailandensis . The two soil samples that yielded growth of B . pseudomallei were collected from paddy field of Gazipur district . No other soil samples from any location yielded growth of B . pseudomallei . The remaining suspected isolates were identified as 12 different organisms ( S2 Table ) . Total 940 blood samples were collected from apparently healthy individuals residing in four districts of Bangladesh . Out of total 940 healthy subjects , anti- B . pseudomallei IgG antibody higher than the cut-off value ( >0 . 8 ) were detected in 203 individuals ( 21 . 5% ) . Highest positive result was obtained from Mymensingh district ( 30 . 8% ) while the rate was only 9 . 8% in Kishoregange district . The detail district wise rate of sero-positivity is shown in Table 2 . The age distribution of the seropositive cases showed that the maximum number ( 30 . 4% ) of positive cases belonged to > 50 years age group while the lowest rate was among the 1–10 years age group ( 5 . 3%; Table 3 ) . The seropositivity rate of anti- B . pseudomallei IgG antibody in male and female population and among different occupational groups ranged from 20% to 27% ( Table 4 ) . No significant association was observed .
In Bangladesh , melioidosis has been infrequently detected for last the 25 years but no systematic epidemiologic information regarding its true magnitude and source is available . Isolation of the organism , B . pseudomallei , from clinical specimens indicates that the organism is present in our environment . However , its actual source has never been identified . The present study has been designed to isolate B . pseudomallei from the soil as well as to determine the seroprevalence among an apparently healthy population residing in four northern and northeastern districts of Bangladesh from where melioidosis cases were diagnosed previously . In the first phase of the study , we aimed to find out the presence of B . pseudomallei in the soil samples from the melioidosis endemic districts of Bangladesh . This strategy of selecting four endemic districts to detect B . pseudomallei from environmental sources has been previously used throughout South-East Asia and northern Australia [22] . Out of 179 soil samples , only two soil samples ( K23 and K35 ) from paddy fields of Gazipur district yielded growth of B . pseudomallei . Our failure to isolate B . pseudomallei from soil samples from other sites could be due to soil condition at the time of sample collection and limited number of samples . The presence of B . pseudomallei is influenced by low bacterial density , rain fall , load of organic materials and oxygen contents of the soil [23] . It is to be noted that the collection and culture of soil in our study was carried out from June to September 2011 which encompasses summer and rainy seasons . Moreover , if we could use the method of soil culture as outlined in international consensus guidelines of 2013 [6] then the rate of isolation of the organism might be higher than the observed result . The isolation of B . pseudomallei from the soil samples in one district of Bangladesh indicates for the first time that the organism might be present in the local environment and is the source of human infections . However , we could not ascertain whether our soil and previous clinical isolates were similar and whether the organisms from the soil reservoir were the source of clinical infections . If we could perform multi locus sequence typing ( MLST ) or sequencing of the genome of the isolates then we could confirm that the soil organisms might be the source of the clinical infections in Bangladesh . But , the presence of the organism in the soil strongly indicates that it could be a potential reservoir . So far 18 countries of the world had been designated as definitive country for melioidosis based on the presence of culture confirmed B . pseudomallei in clinical case as well as in the environmental samples namely soil , water , etc of the locality [6 , 24] . Our finding of B . pseudomallei from the soil of one of the districts finally confirms that Bangladesh too is a definite endemic country for melioidosis . In the second phase of the study , blood samples were collected from healthy population of four districts to determine the seroprevalence of B . pseudomallei infection . We have used ELISA assay using crude sonicated antigen to determine the presence of anti- B . pseudomallei IgG antibody . The rate of serological positive cases for anti- B . pseudomallei IgG antibody ranged from 22%-31% among study population of three districts where melioidosis cases were detected earlier . The seropositive rate was highest in Mymensingh ( 30 . 8% ) district while it was lowest ( 9 . 2% ) in Kishoregange district where no case of melioidosis has yet been diagnosed or reported . The serological findings suggest that there is a wide exposure of people to B . pseudomallei present in the soil or other environmental sources . Seroprevalence rates may vary widely according to the region within the endemic country [25] . Our overall seroprevalence rate was almost similar to rates reported from other countries of the region [26] . In 2012 , a hospital based serological study in Bangladesh has reported sero-positive rate of 28 . 9% for B . pseudomallei antibody among various patients attending selective tertiary care hospitals of the Bangladesh [16] . The study used a very low cut off titer ( 1:10 ) of indirect haemagglutination assay ( IHA ) for defining seropositive cases . The use of the IHA in sero-epidemiological study is problematic in endemic areas , particularly where rates of background seropositivity may be high [26] , presumably due to subclinical exposure to organisms related to B . pseudomallei [27 , 28] . It is important to note from our adsorption study that some degree of non-specific cross reactive antibody was present in the serum which reacted with crude whole cell antigen used in the ELISA as was seen by reduction of OD value following adsorption with P . aeruginosa antigen . This non-specific reactivity might contribute to higher rate of seropositivity . Therefore , we assume that the seropositivity rate could be a little lower if more defined and specific antigen from B . pseudomallei was used . Such specific antigen would eliminate the cross reaction with background antibody to related species of B . pseudomallei like B . thailandensis or B . cepacia . In our study , seropositivity rate for B . pseudomallei antibody significantly increased with advancement of age . Seropositivity was highest ( 30 . 4% ) among individuals more than 50 years of age . Lowest rate ( 5 . 3% ) was observed in 1–10 years age group . The result suggests that the chance of exposure to B . pseudomallei increases with age . Most of the people residing in the rural settings are involved in agricultural work and frequently come in contact with soil , mud and contaminated water , as they grow older . The rate of seroprevalence among male and female population varies in different studies [29 , 30] . However , we found no significant association of seropositivity with gender or any particular occupation . Probably our adult inhabitants in rural areas are equally exposed to the environment . The present study has for the first time identified the presence of B . pseudomallei in the soil samples of Bangladesh . The study has also demonstrated that a large proportion of people residing in four districts are exposed to the organism and have a potential for developing overt diseases during their lifetime . | Melioidosis , caused by B . pseudomallei , can be a fatal disease if not treated with appropriate antibiotics . The organism is mainly present in soil and water in endemic areas , and people become infected through skin inoculation , inhalation or ingestion . The disease has been sporadically detected in Bangladesh over last several decades . However , its actual prevalence in Bangladesh is largely unknown due to the lack of systematic study and awareness of the medical community about the disease and the organism . In order to address the issue , we have undertaken this study to assess the presence of the organism in the soil as well as its magnitude of exposure among the people of selected areas of the country . The study revealed the presence of B . pseudomallei in the soil and its exposure among the people of different areas . The information would increase the awareness of the medical community for prevention and correct diagnosis of the disease . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2016 | Burkholderia pseudomallei: Its Detection in Soil and Seroprevalence in Bangladesh |
Stochastic simulations are one of the cornerstones of the analysis of dynamical processes on complex networks , and are often the only accessible way to explore their behavior . The development of fast algorithms is paramount to allow large-scale simulations . The Gillespie algorithm can be used for fast simulation of stochastic processes , and variants of it have been applied to simulate dynamical processes on static networks . However , its adaptation to temporal networks remains non-trivial . We here present a temporal Gillespie algorithm that solves this problem . Our method is applicable to general Poisson ( constant-rate ) processes on temporal networks , stochastically exact , and up to multiple orders of magnitude faster than traditional simulation schemes based on rejection sampling . We also show how it can be extended to simulate non-Markovian processes . The algorithm is easily applicable in practice , and as an illustration we detail how to simulate both Poissonian and non-Markovian models of epidemic spreading . Namely , we provide pseudocode and its implementation in C++ for simulating the paradigmatic Susceptible-Infected-Susceptible and Susceptible-Infected-Recovered models and a Susceptible-Infected-Recovered model with non-constant recovery rates . For empirical networks , the temporal Gillespie algorithm is here typically from 10 to 100 times faster than rejection sampling .
Networks have emerged as a natural description of complex systems and their dynamics [1] , notably in the case of spreading phenomena , such as social contagion , rumor and information spreading , or epidemics [1–3] . The dynamics of contagion processes occurring on a network are usually complex , and analytical results are attainable only in special cases [3 , 4] . Furthermore , such results almost systematically involve approximations [3 , 4] . Numerical studies based on stochastic simulations are therefore necessary , both to verify analytical approximations , and to study the majority of cases for which no analytical results exist . The development of fast algorithms is thus important for the characterization of contagion phenomena , and for large-scale applications such as simulations of world-wide epidemics [2 , 5] . The Doob-Gillespie algorithm [6–11] ( also known as Gillespie’s Stochastic Simulation Algorithm—SSA or Gillespie’s direct method ) , originally proposed by David Kendall in 1950 for simulating birth-death processes and made popular by Daniel Gillespie in 1976 for the simulation of coupled chemical reactions , offers an elegant way to speed up such simulations by doing away with the many rejected trials of traditional Monte Carlo methods . Instead of checking at each time-step if each possible reaction takes place , as rejection sampling algorithms do , the Gillespie algorithm draws directly the time elapsed until the next reaction takes place and what reaction takes place at that time . It is readily adapted to the simulation of Poisson processes on static networks [12–16] and can be generalized to non-Markovian processes [17] . Systems in which spreading processes take place , e . g . , social , technological , infrastructural , or ecological systems , are not static though . Individuals create and break contacts at time-scales comparable to the time-scales of such processes [18–20] , and the dynamics of the networks themselves thus profoundly affect dynamical processes taking place on top of them [21–27] . This means that one needs to take the network’s dynamics into account , e . g . , by representing it as a time-varying network ( also known as a time-varying graph , temporal network , or dynamical network ) [28] . The dynamical nature of time-varying networks makes the adaptation of the Gillespie algorithm to such systems non-trivial . The main difficulty in adapting the Gillespie algorithm to time-varying networks is taking into account the variation of the set of possible transitions and of their rates at each time step . We show that by normalizing time by the instantaneous cumulative transition rate , we can construct a temporal Gillespie algorithm that is applicable to Poisson ( constant rate ) processes on time-varying networks . We give pseudocode and C++ implementations for its application to simulate the paradigmatic Susceptible-Infected-Susceptible ( SIS ) and Susceptible-Infected-Recovered ( SIR ) models of epidemic spreading , for both homogeneous and heterogeneous [29] populations . We verify the accuracy of the temporal Gillespie algorithm numerically by comparison with a classical rejection sampling algorithm , and we show that it is up to ∼ 500 times faster for the processes and the parameter ranges investigated here . While Poissonian models are of widespread use , real contagion phenomena show memory effects , i . e . , they are non-Markovian . Notably , for realistic infectious diseases , the rate at which an infected individual recovers is not constant over time [30 , 31] . Social contagion may also show memory effects , e . g . , one may be more ( or less ) prone to adopt an idea the more times one has been exposed to it . To treat this larger class of models , we show how the temporal Gillespie algorithm can be extended to non-Markovian processes . We give in particular an algorithm for simulating SIR epidemic models with non-constant recovery rates .
We define in this section the type of stochastic processes for which the temporal Gillespie algorithm can be applied . At the time of writing , the main domain of application of the algorithm is the class of compartmental models for contagion processes on time-varying networks , which we introduce below . For definiteness , algorithms detailing the application of the temporal Gillespie algorithm will concern this class of stochastic processes . In general , we consider a system whose dynamics can be described by a set of stochastic transition events . We assume that the system can be modeled at any point in time by a set , Ω ( t ) , of M ( t ) independent stochastic processes m , which we term transition processes; the rate at which the transition m takes place is denoted λm . The set Ω ( t ) thus defines the possible transition events at time t and in general changes over time , depending on both external factors and the evolution of the system itself; the number of possible transitions , M ( t ) , thus also generally changes over time , while λm may or may not vary over time . For the classic “static” Gillespie algorithm to be applicable , Ω ( t ) is allowed to change only when a transition ( or chemical reaction in the context of Gillespie’s original article ) takes place . For processes taking place on time-varying networks , the medium of the process—the network—also changes with time . As these changes may allow or forbid transitions , Ω ( t ) is not only modified by every reaction , but also by every change in the network . This is the domain of the temporal Gillespie algorithm , which only requires that the number of points in which Ω ( t ) changes be finite over a finite time-interval [32] . The assumption that the transition processes are independent is essential to the validity of the Gillespie algorithm , as it allows the calculation of the distribution of waiting times between consecutive transitions . This assumption is not overly restrictive , as it only requires a transition process to be independent of the evolution of the other simultaneous transition processes . A transition process may depend on all earlier transitions , and the current and past states of all nodes . As such , Gillespie algorithms can notably be applied to models of cooperative infections and other non-linear processes such as threshold models [17] , and has even been applied to model the dynamics of ant battles [33] . A straightforward way to simulate a stochastic process is to use a rejection sampling algorithm , akin to the classical Metropolis algorithm . Here one divides the time-axis in small time-steps Δt , where Δt should be chosen sufficiently small such that this discretization does not influence the outcome of the process significantly; on time-varying networks , the choice of Δt often comes naturally as the time-resolution at which the network is measured or simulated ( Fig 1A ) . At each time-step t = 0 , Δt , 2Δt , … , we test whether each possible transition m ∈ Ω ( t ) takes place or not . In practice this is done by drawing a random number rm that is uniformly distributed on [0 , 1 ) for each m and comparing it to λmΔt: if rm < λmΔt the reaction takes place , if rm ≥ λmΔt nothing happens [Fig 2 ( Transitions ) ] . ( Note that one should technically compare rm to 1 − exp ( λmΔt ) to ensure that λm defines a proper transition rate for finite Δt . However , the two procedures are equivalent in the limit Δt → 0 . ) From the design of the rejection sampling algorithm we see that the proportion of trials that are rejected is equal to a weighted average over {1 − λmΔt}m . Thus , since we require λmΔt ≪ 1 in order to avoid discretization errors , the vast majority of trials are rejected and the rejection sampling algorithm is computationally inefficient . The Gillespie algorithm lets us perform stochastically exact Monte Carlo simulations without having to reject trials . For Poisson processes on static networks , it works by recognizing that the waiting time between two consecutive transitions is exponentially distributed , and that each transition happens with a probability that is proportional to its rate . Specifically , the ( survival ) probability that the transition m has not taken place after a time τ since the last transition event is S m ( τ ) = e - λ m τ . ( 1 ) Since each transition takes place independently , the probability that no event takes place during the interval τ since the last event is S ( τ ) = ∏ m S m ( τ ) = e - Λ τ , ( 2 ) where Λ = ∑ m = 1 M λ m is the cumulative transition rate . The above result is obtained by using the fact that while Ω and M do depend on t , they only change when an event takes place and not in-between . They can thus be treated as constant for the purpose of calculating the waiting time between events . The distribution of the waiting times τ is then given by the probability density p ( τ ) = Λe − Λτ , while the probability density for the reaction m being the next reaction that takes place and that it takes place after exactly time τ is equal to pm ( τ ) = λm e − Λτ The static Gillespie algorithm thus consists in drawing the waiting time τ∼ Exp ( Λ ) until the next transition and then drawing which transition m takes place with probability πm = λm/Λ . [Here τ∼ Exp ( Λ ) is short for: τ is exponentially distributed with rate Λ . ] For processes taking place on time-varying networks however , the set of transition process , Ω ( t ) , changes with time independently of the transition events , e . g . , for the case of an SIR process nodes may become infected only when in contact with an infected individual ( Fig 1A ) . This means that the survival probability does not reduce to a simple exponential as in Eq ( 1 ) ; it is instead given by S m ( τ ; t * ) = exp ( - ∫ t * t * * I m ( t ) λ m d t ) , ( 3 ) where t* is the time at which the last transition took place , t** = t* + τ is the time when the next transition takes place , and Im ( t ) is an indicator function that is equal to one when the process m may take place , e . g . , when two given nodes are in contact , and zero when m may not take place . The meaning of Im is exemplified in Fig 1A: the node i may be infected by the infectious node j only when the two nodes are in contact; if we let m denote this transition process , Im ( t ) is then one for t = Δt , 3Δt , 4Δt and zero for t = 0 , 2Δt . Note that for processes taking place on adaptive time-varying networks , whose changes only depend on the process itself , Im ( t ) only changes when a transition takes place and Eq ( 3 ) reduces to Eq ( 1 ) . This means that from the point of view of the algorithm , such networks are effectively static and the classic “static” Gillespie algorithm may simply be used there [14 , 16] . We now consider the general case where Ω ( t ) may change independently of the processes evolving on the network ( described in Sec . 1: “Stochastic processes on time-varying networks” ) . Using , as in the previous section , that transition processes are independent , we can write the probability that no event takes place during an interval τ ( the waiting time survival function ) : S ( τ ; t * ) = ∏ m ∈ Ω S m ( τ ; t * ) = exp ( - ∑ m ∈ Ω ∫ t * t * * I m ( t ) λ m d t ) , ( 4 ) where Ω denotes the set of all possible transitions ( transition processes ) on the interval between two transition events , ( t* , t**] , i . e . , Ω is the union over Ω ( t ) for t ∈ ( t* , t**] , and M is the total number of transition processes on the same interval ( the size of Ω ) . We switch the sum and the integral in Eq ( 4 ) to obtain S ( τ ; t * ) = exp ( - ∫ t * t * * ∑ m ∈ Ω I m ( t ) λ m d t ) . ( 5 ) Finally , using that Im ( t ) = 0 for all m ∉ Ω ( t ) , we may write S ( τ ; t * ) = exp ( - ∫ t * t * * Λ ( t ) d t ) , ( 6 ) where Λ ( t ) = ∑ m ∈ Ω ( t ) λ m ( 7 ) is the cumulative transition rate at time t . The dynamics of empirical time-varying networks is highly intermittent and we cannot describe Ω ( t ) analytically . This means that we cannot perform the integral of Eq ( 6 ) to find the waiting time distribution directly . We may instead normalize time by the instantaneous cumulative transition rate , Λ ( t ) : We define a unitless normalized waiting time between two consecutive transitions , τ′ , as τ ′ = L ( t * * ; t * ) = ∫ t * t * * Λ ( t ) d t , ( 8 ) i . e . , equal to the cumulative transition rate integrated over ( t* , t**] . The survival function of τ′ has the following simple form: S ( τ ′ ) = exp ( - τ ′ ) . ( 9 ) The time t** when a new transition takes place is given implicitly by L ( t * * ; t * ) = τ ′ , while the probability that m is the transition that takes place at time t = t** is given by: π m ( t ) = I m ( t ) λ m / Λ ( t ) . ( 10 ) This lets us define a Gillespie-type algorithm for time-varying networks by first drawing a normalized waiting time τ′ until the next event from a standard exponential distribution [i . e . with unit rate , τ′ ∼ Exp ( 1 ) ] , and second , solving L ( t ; t * ) = τ ′ numerically to find t** . In practice , since Λ ( t ) only changes when a transition takes place or at tn = nΔt with n ∈ N , we need only compare τ′ to L ( t n + 1 ; t * ) = ( t n * + 1 - t * ) Λ ( t * ) + Δ t ∑ i = n * + 1 n Λ ( t i ) , ( 11 ) for each time-step n ( Fig 2A–2C ) . Here n* is the time-step during which the last transition took place , and Λ ( t* ) is the cumulative transition rate at t* , immediately after the last transition has taken place . The first term of Eq ( 11 ) is the cumulative transition rate integrated over the remainder of the n*th time-step left after the last transition; the second term is equal to L ( t n + 1 ; t n * + 1 ) . A new transition takes place during the time-step n** where L ( t n * * + 1 ; t * ) ≥ τ ′ ( Fig 2D ) ; the precise time of this new transition is t * * = t n * * + τ ′ - L ( t n * * ; t * ) Λ ( t n * * ) ; ( 12 ) the reaction m that takes place is drawn with probability given by Eq ( 10 ) ( Fig 2D ) . We then update Ω and Λ to Ω ( t** ) and Λ ( t** ) ( Fig 2E ) , draw a new waiting time , τ′ ∼ Exp ( 1 ) , and reiterate the above procedure ( Fig 2F ) . The algorithm can be implemented for contagion processes on time-varying networks as follows ( see Methods for pseudocode for specific contagion models and S1 Files for implementation in C++ ) : By construction , the above procedure produces realizations of a stochastic process for which the waiting times for each transition follow exactly their correct distributions . The temporal Gillespie algorithm is thus what we term stochastically exact: all distributions and moments of a stochastic process evolving on a time-varying network obtained through Monte Carlo simulations converge to their exact values . Rejection based sampling algorithms are stochastically exact only in the limit λmΔt → 0 . A large literature exists on the related problem of simulating coupled chemical reactions under externally changing conditions ( e . g . , time-varying temperature or volume ) [35–40] . Most of these methods consider only external perturbations that can be described by an analytical expression . In this case the problem reduces to that of defining a static , yet non-Markovian , algorithm . Some methods , and notably the modified next reaction method developed by Anderson [37] , can be adapted to a completely general form of the external driving and thus , in principle , to simulate dynamical processes taking place on time-varying networks . These methods are based on a scheme that is conceptually similar to Gillespie’s direct algorithm , the next reaction method , proposed by Gibson and Bruck [35] . The next reaction method draws a waiting time for each reaction individually and chooses the next reaction that happens as the one with the shortest corresponding waiting time . It then updates the remaining waiting times , draws new waiting times ( if applicable ) , and reiterates . To generalize the next reaction method to processes with non-exponential waiting times , Anderson introduced the concept of the internal time for each transition process [37] . In the notation used in the present article it is defined as T m ( t ) = ∫ 0 t I m ( t ) λ m d t and is thus equivalent to the normalized time , L ( t , 0 ) , only for an individual transition process . By construction , the next reaction method needs to draw only one random number per transition event , where the Gillespie algorithm draws two . However , this reduction in the number of required random variables comes at a price: one must draw a random number for each individual transition process and keep track of , compare , and update each of the individual waiting times . For chemical reactions , where the number of different chemical reactions is small ( it scales with the number of chemical species ) , this tradeoff favors the next reaction method . However , for contagion processes on networks , each individual is unique ( if not intrinsically , at least due to its position in the network ) . The number transition processes thus scales with the number of nodes and contacts , which favors the Gillespie algorithm as it does not need to keep track of each of them individually [17] . On time-varying networks ( or for time-varying external driving ) one must furthermore update relevant internal times each time the network structure ( external conditions ) changes in the next reaction method . Chemically reacting systems are usually close to being adiabatic , i . e . , the external driving changes slowly compared to the time-scales of chemical reactions . Thus , the additional overhead related to updating individual internal times is practically negligible . However , the dynamics of temporal networks is highly intermittent and the time-scale of network change is typically smaller than the time-scales of relevant dynamical processes . Here one must thus update the internal times many times between each transition event , inducing a substantial overhead . Since the temporal Gillespie algorithm operates with a single global normalized waiting time , it handles these updates more efficiently . Finally , the modified next reaction method may in principle be extended to non-Markovian processes taking place on time-varying networks ( as treated in Sec . 6: “Non-Markovian processes” using the temporal Gillespie algorithm ) . However , such an approach would , for each single transition , require solving numerically Eq . ( 13 ) of [37] for the internal waiting time of each individual transition process , taking into account the time-varying network structure , finding the shortest corresponding waiting time in real time , and then updating the internal waiting times of all the other reactions , rendering the next reaction method even more inefficient in this general case . For real-world contagion processes , transition rates are typically not constant but in general depend on the history of the process [30 , 31] . Such processes are termed non-Markovian . The survival probability for a single non-Markovian transition process taking place on a time-varying network is given by: S m ( τ ; F t ( m ) ) = exp ( - ∫ t * t * * I m ( t ) λ m ( t ; F t ( m ) ) d t ) . ( 15 ) Here F t ( m ) is a filtration for the process m , i . e . , all information relevant to the transition process available up to and including time t; typically , F t ( m ) will be its starting time and relevant contacts that have taken place since . As above , t* is the time of the last transition and t** = t* + τ is the time of the next . [Note that since λm now depends explicitly on t , we may absorb Im in λm; however , to underscore the analogy with the Poissonian case , we keep the factor Im explicitly in Eq ( 15 ) . ] We use again that the transition processes are independent , to write the waiting time survival probability: S ( τ ; F t ) = exp ( - ∫ t * t * * Λ ( t ; F t ) d t ) , ( 16 ) with Λ ( t ; F t ) = ∑ m ∈ Ω ( t ) λ m ( t ; F t ( m ) ) , ( 17 ) and where F t is the union over F t ( m ) for m ∈ Ω . For a static network , Eq ( 6 ) reduces to the result found in [17] . This can be seen by noting that M ( t ) = M and Ω ( t ) = Ω are then constant , and thus that λ m ( t ; F t ( m ) ) = - [ d S m ( t ; F t ( m ) ) / d t ] / S m ( t ; F t ( m ) ) = d { ln [ 1 / S m ( t ; F t ( m ) ) ] } / d t and S m ( t ; F t ( m ) ) = S m ( t + t m ; F t ( m ) ) / S m ( t m ; F t ( m ) ) , yielding directly Eq . ( 7 ) of [17] . As in the Poissonian case ( Sec . 4: “Temporal Gillespie algorithm” ) we define the normalized waiting time , τ′ , as τ ′ = L ( t * * ; t * , F t ) = ∫ t * t * * Λ ( t ; F t ) d t . ( 18 ) This gives us the same simple form as above for the survival function of the normalized waiting time , τ′ , S ( τ ′ ) = exp ( - τ ′ ) , ( 19 ) and the probability that m is the transition that takes place at t = t** , π m ( t ; F t ) = I m ( t ) λ m ( t ; F t ( m ) ) Λ ( t ; F t ) . ( 20 ) Until now our approach and results are entirely equivalent to the Poissonian case considered above . However , since λm ( t ) in general depend continuously on time , the transition time t** is not simply found by linear interpolation as in Eq ( 12 ) . Instead , one would need to solve the implicit equation L ( t * * ; t * ) = τ ′ numerically to find t** exactly . To keep things simple and speed up calculations , we may approximate Λ ( t ) as constant over a time-step . This assumes that ΔΛ ( t ) Δt ≪ 1 , where ΔΛ ( t ) is the change of Λ ( t ) during a single time-step . It is a more lenient assumption than the assumption that Λ ( t ) Δt ≪ 1 which rejection sampling relies on , as can be seen by noting that in general ΔΛ ( t ) /Λ ( t ) ≪ 1 . The same assumption also lets us calculate L ( t n + 1 ; t * ) as in the Poissonian case: L ( t n + 1 ; t * , F t ) = ( t n * + 1 - t * ) Λ ( t * ) + Δ t ∑ i = n * + 1 n Λ ( t i ; F t ) , ( 21 ) and the time , t** , at which the next transition takes place: t * * = t n * * + τ ′ - L ( t n * * ; t * , F t ) Λ ( t n * * ; F t ) . ( 22 ) Using the above equations , we can now construct a temporal Gillespie algorithm for non-Markovian processes . This algorithm updates all λm ( t ) that depend on time at each time-step , where for the Poissonian case we only had to initialize new processes , i . e . , contact-dependent processes ( type b and c , Sec . 1: “Stochastic processes on time-varying networks” ) . This means the algorithm is only roughly a factor two faster than rejection sampling [compare dotted lines ( ϵ = 0 ) in Fig 6] . To speed up the algorithm , we may employ a first-order cumulant expansion of S ( τ ; F t ) around τ = 0 , as proposed in [17 , 38] for static non-Markovian Gillespie algorithms . It consists in approximating λ m ( t ; F t ( m ) ) by the constant λ m ( t * ; F t ( m ) ) for t* < t < t** and gives a considerable speed increase of the algorithm [full ( ϵ → ∞ ) in Fig 6] . However , the approximation is only valid when M ( t ) ≫ 1 [43] , which is not always the case for contagion processes . Notably , at the beginning and end of an SIR process , and near the epidemic threshold for an SIS process , M is small and the approximation breaks down; the approximate algorithm for example overestimates the peak number of infected nodes in a SIR process with recovery rates that increase over time [compare full black line ( ϵ → ∞ ) to the quasi-exact full red line ( ϵ = 0 ) in Fig 7A] . An intermediate approach , which works when the number of transition processes is small , but is not too slow to be of practical relevance , is needed . We propose one such approach below [44] .
We have presented a fast temporal Gillespie algorithm for simulating stochastic processes on time-varying networks . The temporal Gillespie algorithm is up to multiple orders of magnitude faster than current algorithms for simulating stochastic processes on time-varying networks . For Poisson ( constant-rate ) processes , where it is stochastically exact , its application is particularly simple . The algorithm is also applicable to non-Markovian processes , where a control parameter lets one choose the desired accuracy and performance in terms of simulation speed . We have shown how to apply it to compartmental models of contagion in human contact networks . The scope of the temporal Gillespie algorithm is more general than this , however , and it may be applied e . g . to diffusion-like processes or systems for which a network description is not appropriate .
Tables 2 and 3 list the notation used in the manuscript . Table 2 gives notation pertaining to the temporal Gillespie algorithm , and Table 3 lists notation pertaining to time-varying networks and compartmental contagion processes . All simulations for comparing the speed of algorithms were performed sequentially on a HP EliteBook Folio 9470m with a dual-core ( 4 threads ) Intel Core i7-3687U CPU @ 2 . 10 GHz . The system had 8 GB 1 600 MHz DDR3 SDRAM and a 256 GB mSATA Solid State Drive . Code was compiled with TDM GCC 64 bit using g++ with the optimization option -O2 . Speedtests were also performed using -O3 and -Ofast , but -O2 gave the fastest results , both for rejection sampling and temporal Gillespie algorithms . For SIR processes simulations were run until I = 0; for SIS processes simulations were run for 20/ ( μΔt ) time-steps ( as in Fig 3 ) or until I = 0 , whichever happened first . Between 100 and 10 000 independent realizations were performed for each data point depending on μΔt ( 100 for low μΔt and 10 000 for high μΔt ) . For simulations on empirical contact data , data sets were looped if necessary . We here give pseudocode for the application of the temporal Gillespie algorithm to three specific models: the first subsection treats the Poissonian SIR process , the second treats the Poissonian SIS process , and the third treats a non-Markovian SIR process with recovery times following a general distribution . We assume in the following that the time-varying network is represented by a list of lists of individual contacts taking place during each time-step . An individual contact , termed contact , is represented by a tuple of nodes , i and j . The list contactLists[t] gives the contacts taking place during a single time-step , t , for t = 0 , 1 , … , T_simulation-1 , where T_simulation is the desired number of time-steps to simulate . The state of each node is given by the vector x , where the entry x[i] ∈ {S , I , R} gives the state of node i . As one may always normalize time by the duration of a time-step , Δt , we have in the following set Δt = 1 . Note that beta and mu in the code then corresponds to βΔt and μΔt , respectively . When simulations are carried out on data which are looped due to their finite length , the speed of the temporal Gillespie algorithm may be further increased for processes with an absorbing state , such as the SIR process , by removing obsolete contacts to nodes that have entered such a state . Pseudocode 3 shows pseudocode for removing obsolete contacts; its replaces lines 11 to 19 of Pseudocode 1 . Pseudocode 3: Pseudocode for counting possible S → I transitions with removal of outdated contacts . C++ code is given in S1 Files . 01 CLEAR m_SI //S nodes in contact with I nodes 02 FOR contact in contactLists[t] 03 ( i , j ) = contact 04 IF x[i]==S 05 IF x[j]==I 06 APPEND i to m_SI 07 ELSE IF x[j]==R //remove if x[j]==R 08 REMOVE contact from contactLists[t] 09 ENDIF 10 ELSE IF x[i]==I 11 IF x[j]==S 12 APPEND j to m_SI 13 ELSE //remove if ( x[i] , x[j] ) ==I or x[i]==R 14 REMOVE contact from contactLists[t] 15 ENDIF 16 ELSE //remove if x[i]==R 17 REMOVE contact from contactLists[t] 18 ENDIF 19 ENDFOR | When studying how e . g . diseases spread in a population , intermittent contacts taking place between individuals—through which the infection spreads—are best described by a time-varying network . This object captures both their complex structure and dynamics , which crucially affect spreading in the population . The dynamical process in question is then usually studied by simulating it on the time-varying network representing the population . Such simulations are usually time-consuming , especially when they require exploration of different parameter values . We here show how to adapt an algorithm originally proposed in 1976 to simulate chemical reactions—the Gillespie algorithm—to speed up such simulations . Instead of checking at each time-step if each possible reaction takes place , as traditional rejection sampling algorithms do , the Gillespie algorithm determines what reaction takes place next and at what time . This offers a substantial speed gain by doing away with the many rejected trials of the traditional methods , with the added benefit of giving stochastically exact results . In practice this new temporal Gillespie algorithm is tens to hundreds of times faster than the current state-of-the-art , opening up for thorough characterization of spreading phenomena and fast large-scale applications such as the simulation of city- or world-wide epidemics . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Temporal Gillespie Algorithm: Fast Simulation of Contagion Processes on Time-Varying Networks |
Thrombocytopenia is a major side-effect of cytotoxic cancer therapies . The aim of precision medicine is to develop therapy modifications accounting for the individual’s risk . To solve this task , we develop an individualized bio-mechanistic model of the dynamics of bone marrow thrombopoiesis , circulating platelets and therapy effects thereon . Comprehensive biological knowledge regarding cell differentiation , amplification , apoptosis rates , transition times and corresponding regulations are translated into ordinary differential equations . A model of osteoblast/osteoclast interactions was incorporated to mechanistically describe bone marrow support of quiescent cell stages . Thrombopoietin ( TPO ) as a major regulator is explicitly modelled including pharmacokinetics and–dynamics of TPO injections . Effects of cytotoxic drugs are modelled by transient depletions of proliferating cells . To calibrate the model , we used population data from the literature and close-meshed individual data of N = 135 high-grade non-Hodgkin’s lymphoma patients treated with CHOP-like chemotherapies . To limit the number of free parameters , several parsimony assumptions were derived from biological data and tested via Likelihood methods . Heterogeneity of patients was explained by a few model parameters . The over-fitting issue of individual parameter estimation was successfully dealt with a virtual participation of each patient in population-based experiments . The model qualitatively and quantitatively explains a number of biological observations such as the role of osteoblasts in explaining long-term toxic effects , megakaryocyte-mediated feedback on stem cells , bi-phasic stimulation of thrombopoiesis by TPO , dynamics of megakaryocyte ploidies and non-exponential platelet degradation . Almost all individual time series could be described with high precision . We demonstrated how the model can be used to provide predictions regarding individual therapy adaptations . We propose a mechanistic thrombopoiesis model of unprecedented comprehensiveness in both , biological mechanisms considered and experimental data sets explained . Our innovative method of parameter estimation allows robust determinations of individual parameter settings facilitating the development of individual treatment adaptations during chemotherapy .
Reduced platelet counts , called thrombocytopenia , is a major dose-limiting side effect of many dose-intense cancer chemotherapies [1 , 2] . Understanding thrombopoiesis during cytotoxic chemotherapy is crucial for its amelioration by chemotherapy dose adjustments , therapy postponement , platelet transfusion or growth factor applications such as thrombopoietin ( TPO ) . However , this is a non-trivial task since thrombocytopenia risk depends on several therapy-based and individual factors such as dosing and timing of the cytotoxic drugs , application of platelet concentrates , age , sex and individual chemosensitivity [1 , 3] . A major challenge of precision medicine is to take all of these factors into account for optimal risk management . Otherwise , possibly small groups of patients with particularly worse outcome or side-effects impose therapy constrains for large patient collectives . Due to the large number of factors influencing therapy outcome and side-effects , we hypothesize that comprehensive models are required to support the concept of precision medicine . In the present paper , we construct a comprehensive biomathematical model of human thrombopoiesis under chemotherapy , which allows prediction of time courses at an individual level for the first time . Our model is based on an earlier proposed model of human thrombopoiesis under chemotherapy [4] constructed to explain median time courses of patients . We call this model the ‘former model’ and propose a refinement here; based on the assumption of heterogeneity of a few model parameters rather than mechanistic differences between patients . Required additional model assumptions and corresponding adaptations of equations will be presented and discussed in detail . Some emphasis is placed on parametrizing the model on the basis of available time series data of patients under therapy . We propose a Bayesian approach to include both , individual data and population data into the model fitting . To avoid over-fitting , we successfully addressed the problem of parameter identifiability , i . e . model selection is based on a generic , data-driven tradeoff between parsimony and precision . Individualized model predictions require a detailed clinical data base . We used data of patients treated in the framework of randomized clinical trials of the German non-Hodgkin’s lymphoma study group guaranteeing high quality of individual patient data . These comprise therapy adaptations , supportive treatments , and most importantly , closely meshed time series of blood counts . Hence , we do not only consider heterogeneity in individual model parameters but also heterogeneity of treatment for the first time .
Ethics approval and consent to participate: Data were obtained from studies of the German High-Grade Non-Hodgkin’s Lymphoma Study Group . All patients had given informed consent and studieswere approved by responsible ethics committees and were carried out in accordance with the principles of good clinical practice and the declaration of Helsinki . Details on ethics committees and reference numbers can be found in the respective publications of the studies used for our modelling [5 , 6] . The model proposed in the present paper is a modified and improved version of our previous work [4] , which we briefly summarize here . This ordinary differential equations ( ODE ) model describes the dynamics of concatenated cell compartments of stem cells , colony-forming units of megakaryocytes , megakaryocytes and platelets in spleen and circulation . It contains several feedback loops where TPO is a major mediator . The model already considered the effect of cytotoxic chemotherapy by assuming transient depletion of bone marrow cell compartments after application . The principle structure is displayed in the figure in S1 Appendix . We briefly present the major features of the model . Then , necessary adaptations are described in detail: In our model , we also adopted the formalism developed in [14] regarding modelling of delayed transitions between compartments . Briefly , this is achieved by introducing a number of concatenated sub-compartments with first-order transitions mimicking a Gamma-distributed overall transition rate . Our former model correctly described median time courses of platelets of patients treated with chemotherapy . Individual deviations from the standard therapy were neglected and heterogeneity of patient responses was not considered so far . To remove these restrictions , we present an improved version of the model ( Fig 1 , description of model compartments , see Table 1 ) . This however required a number of adaptations of model hypotheses explained and motivated in the following: In this section , we derive all model equations . Corresponding parameters , their values and procedures for estimation are given in tables in S2 Appendix . Simulation results of the model are compared with clinical data in order to verify the model assumptions and to estimate parameter settings . Data were either taken from the literature or are provided by the German non-Hodgkin’s lymphoma trial group ( PI: Michael Pfreundschuh ) . These data comprise individual therapy settings and time series data of platelets . Here we briefly present the literature data used for our model development . Typically , only averaged data are available . Data of Harker et al [30] . Single doses of 3 μg/kg of pegylated TPO were injected subcutaneously into 16 healthy subjects . Total TPO concentrations and platelet counts were measured daily between day 0 and day 28 . MKC counts and percentages of MKC ploidies were determined at days 0 , 7 , 11 and 17 after TPO injection . Data of Hanson and Slichter 1985 [26] . Autologous 51Cr-labeled platelets were transfused to 16 normal subjects and 27 patients with stable , untreated thrombocytopenia secondary to bone marrow failure . Platelet counts range between 12 , 000 and 70 , 000/μL . Dynamics of labeled platelets were determined during 5 days after injection . Compared to normal subjects , platelet life span was slightly reduced in patients with platelet counts in between 50 , 000 to 100 , 000/μL but was markedly reduced for patients with platelet counts below 50 , 000/μL . Data of Li et al [21] . Ten cancer patients received four chemotherapy cycles in median ( range 3–6 cycles ) . Their bone marrow niches have been examined before and after treatment . The number of osteoblasts per bone surface was markedly reduced after chemotherapy . Data of Engel et al [39] . Three patients with Hodgkin’s or aggressive non-Hodgkin lymphomas were considered in this study . Patients received intensified multi-cycle poly-chemotherapies . Close-meshed time series data of endogenous TPO and platelet counts were determined . All three patients showed clear long-term effects of the therapy: Data of NHL-B study [1 , 5 , 6] . The NHL-B study is a randomized clinical trial of elderly patients with aggressive non-Hodgkin’s lymphoma . Patients were randomized to one of the four arms 6xCHOP or 6xCHOEP with either 14 or 21 days of cycle duration . Schedules with 14 day cycle duration were supported by G-CSF injections . Thrombopenia was treated with platelet transfusions , postponement of therapy or reduction in chemotherapy dose . Close meshed time series data of blood cell counts are available for these patients as well as individual information regarding the course of the therapy ( i . e . individual risk factors , dosing of drugs and growth factors , therapy delays , supportive care , outcome ) . We selected 135 from 1600 patients , whose platelets counts were measured during 4 or more cycles with at least 5 measurements per cycle to obtain sufficiently detailed individual time series data . Our ODE model ( 1–34 ) was implemented in Matlab and numerically solved with the 15s solver [40 , 41] . Simulation settings of specific therapy scenarios are explained in S3 Appendix . Parameter estimation is based on the optimization of the agreement of model and data as measured by fitness functions . Only a few parameters were assumed to differ between individuals explaining patient heterogeneity . These parameters are determined by optimizing individual fitness functions . In order to overcome the overfitting problem , each individual fitness function includes both , individual data as well as averaged data from the literature ( called biological data in the following ) . This is equivalent to a virtual participation of a patient in the literature studies from which the data were retrieved . Details of the parameter estimation procedure and measures to avoid model overfitting are explained in detail in S13 Appendix . We performed a step-wise fitting process starting with Engel et al . data and including literature data . Finally , individual time course of patients from the NHL-B study are fitted . We performed a stepwise fitting procedure to estimate the parameters of our model: All parameters are described in tables from S2 Appendix . Further details of the fitting process can be found in S14 Appendix .
Here , we present the agreement of our model with biological data and the individual patient data of Engel et al . First , we consider the three patients studied in Engel et al . Comparisons of model and data can be found in Fig 3 . We observed a good agreement of the model with all individual time series of TPO and platelets . In particular , the long-term platelet decrease accompanied with stronger long-term TPO increase could be explained by the model . During parameter fitting , it is assumed that these three patients virtually participated in the study of Li et al . [21] , which measured relative counts of osteoblast prior to and after chemotherapy . Results for the three patients are shown in Fig 4C . All patients are in agreement with these data . As described in the methods section , we assumed that the three patients ( virtually ) also took part in the studies of Hanson et al and Harker et al . Results are shown in Fig 4A and 4B , respectively . Our model successfully explains the non-exponential dynamics of platelet elimination observed in Hanson et al . As the initial platelet level becomes smaller , platelet degradation curve becomes more concave and the half-life reduced from 3 . 5–4 . 5 days to 2 days . Patient 3 was estimated to have shorter transit time of platelets than normal patients , which explains the rapid chemotherapy-induced oscillations . In summary , all three individual parameter sets resulted in a good fit of the dynamics of platelets count , total TPO and MKC ploidies for most of the data points from Harker et al study . Only counts of MKC ploidy 64 are slightly underestimated at day 7 and the total count of MKC is slightly underestimated at day 11 . However , since standard errors are large , almost all simulations are within the 95% confidence interval of the average . Table 4 in S15 Appendix shows a good agreement of data and the simulated steady state distributions of MKC of different ploidies . Most of the parameters showed good identifiability . The estimated population and individual parameters values and their relative standard errors are shown in Tables 1 and 2 in S15 Appendix . In S17 Appendix , we present parameters with poorer identifiability . Situations in which poor identifiability corresponds to correlated parameter estimates are summarized in Table 3 in S15 Appendix . We present only the strongest of these correlations having respective absolute values larger than 0 . 8 . Procarbazine PD effect was unidentifiable , and thus , assumed to be zero . We compared distributions of the individual parameter estimates of NHL-B with the corresponding priors derived from Engel et al . No significant differences were detected ( see S18 Appendix ) . Here we discuss the parameters assumed to be heterogeneous between patients ( with IIV ) and compare corresponding estimates for the three patients of Engel et al . A total of 9 parameters were assumed to express IIV . Of those , two parameters are related to chemotherapy namely pdcyclo ( toxicity of cyclophosphamide ) , and dOsteoloss ( elimination rate of dormant stem cells and megakaryocytes due to loss of supporting osteoblasts ) . pdcyclo is the only individual parameter of a direct effect of chemotherapy on active precursors . Consequently , its value basically influences the depth of the nadir during the first treatment cycle . Patient 2 has the largest pdcyclo value corresponding to the deepest platelet nadir at the first chemotherapy cycle . The parameter dOsteoloss determines whether dormant cells are preferentially mobilized to the respective active compartments ( small values ) or die ( large values ) when supporting osteoblasts are eliminated by chemotherapy . Activation of dormant stem cells and MKC contribute to the delayed peak of platelets about 10 days after chemotherapy application . This peak is also influenced by the strength of feedback , in particular by sensitivity parameter bS_act ( self-renewal probability of active stem cells ) for which we also assume IIV . All patients have small bS_act values indicating weak feedback on self-renewal of stem cells resulting for example in the absence of the predicted increase in total megakaryocytes counts during days 0–17 after a hypothetical stimulation with pegylated TPO ( see agreement with Harker data ) . However these average data have large standard deviation limiting their informative value . Patient 1 has the largest value of bS_act as well as the smallest value of dOsteoloss resulting in the largest platelet peaks during recovery . Patient 3 has low recovery peaks due to the estimated strong osteoblast reduction ( large dOsteoloss ) , large PD effect pdcyclo and smallest value for bS_act . The above considerations show that a complex interplay between parameters pdcyclo , bS_act and dOsteoloss determines the depth of platelet nadirs , dynamics of platelet recoveries and severity of cumulative toxicity during multicycle therapy . We simulated 135 patients of the NHL-B study separately , considering individual therapy adaptations such as dose reduction , therapy postponement or application of platelet concentrates . Likewise , we determined individual parameter estimates of the subset of parameters for which we assume patient heterogeneity . Ten parameters were assumed to show IIV for this patient population . All population-based parameters determined in the previous fitting steps were kept constant . In Fig 5 we show the agreement of model and data for 9 selected individuals of the 135 patients . All other patients are presented in figures of S16 Appendix . We obtained a good agreement for most of the patients . The individual parameter estimates are presented in the Table 1 in S16 Appendix , their relative standard errors as well as residual errors are shown in the Table 2 in S16 Appendix . No general overfitting was observed . We observed inferior fitting results for a few patients with small initial toxicity ( e . g . patient numbers 2 , 52 and 313 ) . In these cases , platelet dynamics were often noisy , i . e . similar to dynamics of healthy untreated patients from Schulthess et al study [44] . In two cases , unexpected jumps in platelet dynamics spoiled the fits ( patients 1693 , 1696 ) . Errors within clinical records such as missing information regarding platelet transfusions cannot be excluded in these cases . In Fig 6A , 6B and 6C we show how the resulting 135 individual parameter sets reproduce the averaged data from Hanson et al , Harker et al and Li et al studies with a precision similar to that of the three patients of Engel et al . Standard deviations of the simulated biological data are often smaller than those of the observed data except for relative osteoblasts from Li et al study , whose respective virtual fits are more variable than the observed ones . We used our model to simulate possible effects of a shift of the last chemotherapy cycle for patients , who developed significant thrombocytopenia at the late stages of the treatment . Fig 7A , 7B , 7C and 7D show platelet dynamics corresponding to patients 15 , 20 , 677 and 1463 respectively . For all patients we simulated treatment shifts of -5 ( an earlier start of the last cycle ) , 0 , 5 , 10 days postponement , and finally , omission of the last cycle with subsequent tree-months follow up . All simulations show that it takes nearly two-four months after the last chemotherapy application to damp platelet oscillations sufficiently so that sever thrombopenia grades 3 or 4 is no longer observed . Interestingly , therapy postponement or earlier starts did not always result in ameliorated thrombopenia compared to the original schedule ( patients 20 , 677 and 1463 ) . Sensitivity of patients regarding change of treatment schedule is predicted to be highly variable from almost no effect ( patient 677 ) to strong nadir differences and respective thrombopenia grades ( patients 15 , 20 and 1463 ) .
Dose-intense cytotoxic chemotherapies improved the outcome of several cancer entities [2 , 5 , 6] but is limited by the general toxicity of the drugs . However , this toxic response is highly heterogeneous between patients so that general therapy constrains are caused by a possibly small subset of patients with high risk . It is a major goal of precision medicine to identify these patients early and to introduce chemotherapeutic regimen adapted to individual risks . Here we study thrombocytopenia in the context of the treatment of aggressive non-Hodgkin’s lymphoma where it is frequently dose-limiting [1 , 3] . Current statistical risk models [1 , 3] have low precision since even in the lowest risk groups a significant amount of patients develop high toxicity , i . e . the statistical model cannot unambiguously identify the group of patients at high risk . Therefore , there is an urgent need to develop individualized mechanistic models of thrombocytopenia , which not only can explain long-term effects of multi-cyclic poly-chemotherapy but can also update their predictions based on available patient data . This is a major requirement for establishing model-based individual treatment adaptations such as individualized supportive treatments by platelet transfusions or growth factor applications or by postponement or dose-reduction of chemotherapy . Several other complex mechanistic and semi-mechanistic models of animal and human thrombopoiesis were proposed in the past [4 , 24 , 25 , 45–50] . So far , only a few of them include chemotherapy applications [4 , 47 , 48] and only one of them assumes inter-individual variability of parameters [47] . This model includes biologically well-established TPO-mediated feedbacks , although without detailed description of dynamics of megakaryocytes of different ploidies . Moreover , this model does not consider dynamics of the early precursors ( blast and stem cells ) nor long-range effects of multi-cyclic chemotherapy . A major goal of our model development is to allow model-based predictions regarding individual therapy adaptations . To serve similar purposes in the case of granulopoiesis , a model-based dose adaptation tool has been proposed earlier [51 , 52] . This tool is based on a simplistic pharmacodynamical hematopoiesis model of Friberg and Karlsson [42 , 43] assuming inter-individual variability of parameters . This model was applied to different hematopoietic cell lines subjected to chemotherapy including thrombopoiesis [49 , 50] . The model of Friberg and Karlsson assumes one proliferation compartment vulnerable to chemotherapy , three equal transit compartments and one circulating compartment , imposing a single negative feedback on the first compartment . This simplicity enables straightforward clinical data fitting for different hematopoiesis processes under chemotherapies . On the other hand , the oversimplification could obscure the connection of individual model parameters with underlying biological mechanisms such as TPO action on dynamics of maturating components CM and MKC . The model ignores the fact that chemotherapy affects all replicating compartments . Known biological feedback mechanisms of thrombopoiesis are not considered [32] . Moreover , the model is typically applied to single clinical data sets without validation of the parameter settings on the basis of other data . By our model proposal , we aim at improving this situation by building our model on biologically plausible assumptions and for several clinical and biological data sets in parallel . To develop a reliable , comprehensive and individualized mechanistic model of thrombopoiesis under multi-cyclic poly-chemotherapy , we comprehensively revised our former model [4] . We introduced new model compartments and feedbacks in order to describe features not covered previously . The data on dynamics of megakaryocyte ploidy grades during TPO stimulation [30] allowed us to model complex parallel maturation of platelets in a much more mechanistic way as has been done so far [25] . We introduced a dormant stem cells compartment established earlier by agent-based models [15 , 16] but modelled here in a simpler ODE form . We modelled osteoblast support of dormant stem cells and dormant megakaryocytes . For this purpose , we integrated a mathematical model proposed by Komarova et al [20] describing the interaction of osteoblasts and osteoclasts . Chemotherapy effects were added to this model . This also allowed us to describe osteoblast dynamics during multi-cyclic chemotherapy measured by Li et al . [21] . Combined with an indirect elimination of dormant stem cells and MKCs due to lack of support , this gives a mechanistic explanation of frequently observed cumulative or late time toxicity which is not covered by statistical risk models [3] . We also integrated a non- exponential model of platelet degradation proposed by Hersh et al . [38] and revised it here . We included a number of new biological evidence and considered additional experimental data to improve our model . For this purpose , we established and applied an innovative way of parameter fitting for our individualized model . At this , available population data and individual time courses are combined by a Bayesian approach assuming that an individual virtually participated in all experiments for which population data are available . This kind of simultaneous consideration of individual and biological ( prior ) data exploits all available information in a more reliable and complete way than separate estimations of groups of parameters on limited data sets . A caveat of this approach is , however , that one needs to assume comparability of patient collectives across different studies . Our approach is novel in the field since most of PK/PD modelers do not use prior information from other studies but fit exclusively clinical data of interest [42 , 49 , 50] . Heterogeneity is then addressed by mixed effects modeling [53] where parameters estimation is based on likelihood maximization for the entire population . In this case , assessment of algorithm’s convergence and overfitting are controlled exclusively for population parameters determining the distributions of individual parameters . Consequently , mixed effects modeling derives individual parameter estimates as a by-product implying high probability of insufficient fitting quality for a significant number of subjects . Moreover , pre-assumptions on the parameter distributions could spoil individual fits as well . This limits the usefulness of these models to develop individualized therapies . In contrast , our approach maximizes individual fitting precision without making any pre-assumptions on the underlying parameter distributions . We controlled convergence of fitting algorithm and reported standard errors on the individual level . We believe that such an individualized control of goodness of fit is much more appropriate for the purpose of individualized treatment management . Our model has 102 parameters in total of which only 31 are estimated . Four parameters were directly be taken from the Komarova model [20] , 28 are fixed and 28 reduced through parsimony assumptions , as described in the Table 4 of the S2 Appendix . We took 11 PK parameters as well as structural assumptions for PK models of etoposide , cyclophosphamide , doxorubicin and procarbazine from other studies [57 , 58] as described in S3 Appendix and Table 7 of S2 Appendix . Most of the fixed or reduced parameters correspond to chemotherapy effects or behavior of megakaryocytes of different ploidies . This can be explained by the lack of detailed data , e . g . regarding relative cytotoxic contribution of the chemotherapy drugs applied . To improve this situation , separate as well as joint individualized in-vitro studies of cytotoxic effects of cyclophosphamide , doxorubicine , etoposide and procarbazine would be helpful such as those proposed in Zeuner et al [8] . For the sake of parsimony , we did not consider potential chemotherapy drug interactions . Such interactions are not uncommon ( e . g . carboplatin and paclitaxel [54] ) . However , for the drugs considered in the present study , no interactions could be detected based on the available data . Drug combinations of different doses of cyclophosphamide and doxorubicine would be required to unravel any interaction effects . Moreover , for deeper understanding of the TPO-mediated regulation of megakaryocytes , one would require observations of megakaryocytes in analogy to Zeuner et al [8 , 22–24] . A caveat is , however , that this group studied cord blood ( CB ) megakaryocytes which differ significantly from bone marrow megakaryocytes by much less ploidy [24] and by their ability to frequently undergo mitoses in polyploidy state [22] . These facts leave open the question to which extend CB megakaryocytes observations could be applied to the modeling of BM megakaryocytes . Another open question is to which extent thrombopoiesis in young healthy subjects as studied in Harker et al [30] can be compared to that of elderly patients with aggressive non-Hodgkin’s lymphoma studied in NHL-B study [1 , 5 , 6] . It is conceivable , for example , that the lack of MKC peak at day 11 after TPO injection observed in the simulations of patients from Engel et al and NHL-B ( Figs 4B and 6B ) can be attributed to reduced TPO responsiveness of elderly patients compared to young healthy subjects . This issue can only be resolved by studying TPO and megakaryocyte dynamics in chemotherapy-treated elderly patients , which , however , is problematic from the ethical point of view . To derive precise estimates of individual parameters , closely meshed time series data of patients under therapy are required including full information regarding therapeutic interventions ( dosing and timing of cytotoxic drugs , application of platelet concentrates or growth factors ) . Our major resource of individual therapy data , the NHL-B study , only partly fulfils this requirement since from the 1600 available patients only 8 . 4% met our inclusion criteria regarding data quality ( participation in at least 4 chemotherapy cycles with at least five measurements per cycles ) . Moreover , only numbers but not exact time points of platelet transfusions within cycles were documented . For our modelling purpose , we assumed that transfusions occurred immediately after the smallest value and prior to an abrupt platelet increase . To evaluate the quality of our data fitting procedure , we compared the residual errors of subjects with natural fluctuations measured in healthy subjects [44] . Residual errors appeared to be about twice the natural fluctuations implying existence of some effects unexplained by our model . Especially individual platelet dynamics with high irregularity despite of regular ( cyclic ) treatment are difficult to explain by our model . Irregular time series are more often observed in the 14 day regimen compared to the 21 day regimen suggesting a possible interaction with G-CSF treatment . Indeed , G-CSF can induce thrombocytopenia in healthy patients [55] . On the other hand , it is well known that G-CSF stimulates stem cells to start differentiation to blast cells [56] , which are precursors for both thrombocytes and granulocytes . Thus , G-CSF has both , stimulatory and inhibitory effects on thrombopoiesis which might be relevant to explain irregular behavior . Other reasons of inferior agreement of model and data could be bleeding events , platelet consumption by infection episodes or platelet destruction by additional medications . However , in general our model predictions fitted well to the population based data and to the vast majority of individual patient data , also covering long range effects of chemotherapy . As a possible field of clinical application , our simulations showed a strong impact of shifting the start of the next treatment cycle on the resulting thrombocytopenia . This demonstrates the importance of individual next-cycle management for chemotherapy treated patients . We currently develop a medical tool supporting dosing and timing adaptations of chemotherapies in dependence on the individual therapy response . A prototype can be found elsewhere ( https://www . health-atlas . de/shiny-public/apps/thrombopenia/ ) . It revealed that two cycles are sufficient to derive individual parameter estimates for 21-day schedules while three cycles are required for 14-day schedules . In summary , we propose a mechanistic thrombopoiesis model of unprecedented comprehensiveness in both , biological mechanisms considered and experimental data sets explained . Our innovative method of parameter estimation allows robust determinations of individual parameter settings facilitating the development of individual treatment adaptations in the course of cytotoxic chemotherapy as an ultimate goal of systems-medicine . | Chemotherapy is ubiquitously used to treat cancer diseases . Due to general toxicity of the drugs , chemotherapy results in a number of side effects especially with respect to blood formation . Here we study the loss of platelets during chemotherapy which is dose limiting in many situations . However , this side-effect greatly varies between patients with respect to both , severity and necessity of clinical countermeasures . We therefore developed a mathematical model to predict the time course of platelets of patients under chemotherapy and to propose possible treatment adaptations in cases of intolerable toxicity . The model is based on available biological knowledge and data of platelet formation and therapeutic effects thereon . As a major result , we could describe individual time series data of 135 patients under chemotherapy . Conversely , the model can be used to make predictions regarding alternative therapy schedules such as postponement of therapy or chemotherapy dose reductions . Our model is intended to support clinical decision making on an individual patient level . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"population",
"genetics",
"ploidy",
"toxicology",
"bone",
"marrow",
"cells",
"toxicity",
"osteoblasts",
"pharmaceutics",
"connective",
"tissue",
... | 2019 | Modeling individual time courses of thrombopoiesis during multi-cyclic chemotherapy |
Cleft palate is among the most common birth defects in humans . Previous studies have shown that Shh signaling plays critical roles in palate development and regulates expression of several members of the forkhead-box ( Fox ) family transcription factors , including Foxf1 and Foxf2 , in the facial primordia . Although cleft palate has been reported in mice deficient in Foxf2 , whether Foxf2 plays an intrinsic role in and how Foxf2 regulates palate development remain to be elucidated . Using Cre/loxP-mediated tissue-specific gene inactivation in mice , we show that Foxf2 is required in the neural crest-derived palatal mesenchyme for normal palatogenesis . We found that Foxf2 mutant embryos exhibit altered patterns of expression of Shh , Ptch1 , and Shox2 in the developing palatal shelves . Through RNA-seq analysis , we identified over 150 genes whose expression was significantly up- or down-regulated in the palatal mesenchyme in Foxf2-/- mutant embryos in comparison with control littermates . Whole mount in situ hybridization analysis revealed that the Foxf2 mutant embryos exhibit strikingly corresponding patterns of ectopic Fgf18 expression in the palatal mesenchyme and concomitant loss of Shh expression in the palatal epithelium in specific subdomains of the palatal shelves that correlate with where Foxf2 , but not Foxf1 , is expressed during normal palatogenesis . Furthermore , tissue specific inactivation of both Foxf1 and Foxf2 in the early neural crest cells resulted in ectopic activation of Fgf18 expression throughout the palatal mesenchyme and dramatic loss of Shh expression throughout the palatal epithelium . Addition of exogenous Fgf18 protein to cultured palatal explants inhibited Shh expression in the palatal epithelium . Together , these data reveal a novel Shh-Foxf-Fgf18-Shh circuit in the palate development molecular network , in which Foxf1 and Foxf2 regulate palatal shelf growth downstream of Shh signaling , at least in part , by repressing Fgf18 expression in the palatal mesenchyme to ensure maintenance of Shh expression in the palatal epithelium .
The mammalian secondary palate develops from the oral side of the embryonic maxillary processes as a pair of outgrowths , which initially grow vertically to form the palatal shelves flanking the developing tongue . As development proceeds , the palatal shelves reorient to the horizontal position above the dorsum of the tongue , grow towards and subsequently fuse with each other at the midline to form the roof of the oral cavity . Genetic or environmental perturbations of any of these developmental processes , including palatal shelf growth , elevation and fusion , cause cleft palate , one of the most common congenital birth defects in humans [1–4] . Previous studies have shown that palatal shelf growth is regulated by reciprocal signaling interactions between the epithelium and the underlying neural crest-derived mesenchyme . At the onset of the palatal outgrowth , the secreted signaling molecule Sonic hedgehog ( Shh ) is expressed in the oral epithelium [5] . Shh is a mitogen and promotes cell proliferation in many embryonic and adult tissues [6] . Explant culture assays indicate that exogenous Shh protein stimulates palatal mesenchyme proliferation [7 , 8] . Tissue-specific inactivation of the Smoothened ( Smo ) gene , which encodes a transmembrane protein required for transduction of the Shh signaling , in the early cranial neural crest cells resulted in complete absence of secondary palate structures in the Smoc/n;Wnt1-Cre mutant mice [9] . Moreover , tissue-specific inactivation of Shh in the oral epithelium or Smo in the early palatal mesenchyme resulted in defects in palatal shelf growth in the mutant mouse embryos [3 , 8 , 10] . Whereas the mechanism by which Shh signaling regulates palatal mesenchyme cell proliferation is incompletely understood , Shh signaling regulates palatal epithelial cell proliferation indirectly through , at least in part , activation of the fibroblast growth factor Fgf10 in the palatal mesenchyme [8] . Remarkably , in addition to regulating palatal epithelial cell proliferation , both Fgf10 and its epithelial receptor Fgfr2b are required for maintenance of Shh expression in the developing palatal epithelium [8] . Thus , Shh and Fgf10 signaling pathways function in a positive feedback loop to control palatal shelf growth [10] . In addition to its interaction with Fgf10 signaling , Shh signaling has also been shown to cooperate with Bmp signaling to regulate palatal shelf growth . In palatal explant culture assays , exogenous Shh protein induces Bmp2 mRNA expression [7] . Tissue-specific inactivation of Smo in the palatal mesenchyme caused down-regulation of Bmp2 expression in the anterior palatal mesenchyme [10] , indicating that Shh signaling is required for maintenance of Bmp2 expression during normal palatogenesis . Bmp signaling plays a critical role in anterior palatal shelf growth , as targeted deletion of Bmpr1a , encoding a type I receptor for Bmp ligands , in either the early neural crest or in the early palatal mesenchyme resulted in cleft of the anterior palate [11 , 12] . Moreover , mice lacking the homeobox transcription factor Msx1 exhibit complete cleft palate that could be rescued by transgenic expression of Bmp4 driven by the Msx1 gene promoter [7] . During palatal shelf growth , the Msx1-/- mutant mouse embryos showed reduced expression of Bmp4 in the anterior palatal mesenchyme as well as reduced expression of Shh in the anterior palatal epithelium , in comparison with wildtype embryos [7] . Transgenic Bmp4 expression in the anterior palatal mesenchyme was sufficient to restore Shh expression in the anterior palatal epithelium in the Msx1-/- embryos , suggesting that Bmp4 acts downstream of Msx1 in the anterior palatal mesenchyme to maintain Shh expression in the anterior palatal epithelium [7] . The forkhead-box ( Fox ) family proteins form a large family of DNA-binding transcription factors [13 , 14] . Through comparative transcriptional profiling of E10 . 5 embryonic head tissues of Shh mutant and control mouse embryos , Jeong et al . ( 2004 ) found that expression of several Fox family genes , including Foxc2 , Foxd1 , Foxd2 , Foxf1 , and Foxf2 , in the neural crest derived facial mesenchyme was dependent on Shh signaling and suggested that these Fox family transcription factors might be key mediators of Hh pathway function in craniofacial development [9] . Both Foxf1 and Foxf2 are expressed in the developing palatal mesenchyme in wildtype mouse embryos and tissue-specific deletion of Smo also caused significant reduction in expression of Foxf1 and Foxf2 in the palatal mesenchyme [10 , 15] . Foxf1 and Foxf2 display highly conserved amino acid sequences in the Forkhead DNA binding domain ( 100% identical in mouse and 97% identical in human FOXF subfamily ) [16–18] . Whereas mouse embryos lacking Foxf1 function die during midgestation due to severe defects in the extraembryonic mesoderm [19] , mice lacking Foxf2 die shortly after birth with a cleft palate phenotype [15] . Mutations in FOXF2 have also been associated with cleft palate in humans [20] . However , how Foxf2 regulates palate development remains to be elucidated . In this study , we show that Foxf1 and Foxf2 exhibit partially overlapping patterns of expression during palate development , with Foxf2 expressed more broadly than Foxf1 along the anterior-posterior axis of the developing palatal shelves . By using Cre/loxP mediated conditional gene inactivation , we demonstrate that the Foxf1 and Foxf2 transcription factors act partly redundantly to control palatal shelf growth through a novel Foxf-Fgf18-Shh regulatory circuit .
Wang et al . ( 2003 ) reported that Foxf2-deficient mice die shortly after birth and exhibit cleft palate . Although Wang et al . ( 2003 ) detected Foxf2 mRNA expression in the developing palatal shelves in wildtype mouse embryos , they found that Foxf2 mRNAs were more abundantly expressed in the muscle layers of the developing tongue and suggested that the cleft palate defect in Foxf2-/- mice might be secondary to defects in tongue movement because the tongue did not properly descend in the mutant embryos [15] . To determine whether Foxf2 plays an intrinsic role in palate development , we generated mice with tissue-specific inactivation of Foxf2 in the cranial neural crest lineage or in the developing palatal mesenchyme using mice carrying a loxP-flanked Foxf2 conditional allele ( Foxf2c ) [21] . The Wnt1-Cre transgenic mice have been shown to exhibit Cre recombinase activity in the premigratory neural crest cells that give rise to most of the non-muscle mesenchyme in the craniofacial tissues [22–24] . On the other hand , the Osr2IresCre/+ mice exhibit highly specific expression of the Cre recombinase in the Osr2-expressing palatal mesenchyme cells and with no Cre activity in the muscle cells of the developing tongue [10 , 25] . We found that both Foxf2c/-Wnt1-Cre and Foxf2c/-Osr2IresCre/+ mice have complete penetrance of cleft palate and many of the mutant embryos showed failure of palatal shelf elevation , similar to the cleft palate phenotype in mice with constitutive inactivation of the Foxf2 gene ( Foxf2-/- ) , ( Fig 1 ) . These results suggest that Foxf2 function is required in the palatal mesenchyme for normal palatogenesis . To investigate whether Foxf2 function is required for palatal shelf growth , we performed BrdU incorporation assays in E13 . 5 embryos . Since the developing palatal shelves exhibit morphological and molecular heterogeneity along the anterior-posterior and oral-nasal axes [1] , we analyzed the BrdU-labeling index separately for the oral and nasal halves of the palatal shelves in each of the anterior , middle , and posterior sub-regions , with the middle region corresponding to that flanked by the maxillary first molar tooth buds ( Fig 2A–2F ) . We found that Foxf2-/- mutant embryos exhibited significant reduction in cell proliferation in the anterior region , nasal half of the middle region , and the posterior region of the palatal mesenchyme ( Fig 2G ) . These data confirm an intrinsic role for Foxf2 in palate development . To gain insight into the molecular mechanisms mediating Foxf2 function in palate development , we analyzed whether the expression patterns of a number of genes known to play critical roles in palate development were altered in the Foxf2-/- mutant palatal shelves . Shh signaling has been shown to regulate palatal shelf growth and Shh mRNA expression marks the palatal rugae , the epithelial ridges that form in specific spatiotemporal patterns on the oral surface of the palatal shelves during palatal outgrowth [10 , 26 , 27] . Thus , the whole mount Shh mRNA expression pattern has been used as a valuable molecular marker for analysis of palatal shelf growth or patterning defects in mutant mouse studies [12 , 27–29] . In comparison with the Shh mRNA expression pattern in the palatal epithelium in wildtype embryos ( Fig 3A and 3E ) , the Foxf2-/- mutant embryos exhibited specific loss of the most anterior Shh mRNA expression domain that corresponds to Ruga-3 ( Fig 3B and 3F ) at E13 . 5 and E14 . 5 . Shh mRNA expression in the posterior palatal epithelium was also dramatically downregulated in the Foxf2-/- mutant embryos in comparison with the wildtype littermates ( Fig 3A , 3B , 3E and 3F ) . Corresponding to the region-specific loss of Shh mRNA expression , expression of Ptch1 , a well-known direct transcriptional target of canonical Hedgehog signaling [30–33] , was specifically downregulated in the Ruga-3 region as well as in the posterior palate in the Foxf2-/- mutant embryos in comparison with the wildtype littermates ( Fig 3C , 3D , 3G and 3H ) . We examined the expression patterns of Shox2 and Barx1 , which mark the anterior and posterior halves of the developing palatal mesenchyme , respectively [27 , 34 , 35] . The level of Shox2 mRNA expression was reduced , especially in the anterior-most region of the palatal shelves , in Foxf2-/- mutant embryos in comparison with the wildtype littermates ( Fig 3I and 3J ) . In contrast , the pattern and levels of expression of Barx1 in the developing palatal shelves were not obviously altered in Foxf2-/- mutant embryos in comparison with the wildtype littermates ( Fig 3K and 3L ) . Previous studies showed that the Msx1-Bmp4 and Fgf10 signaling pathways regulate palatal shelf growth as well as Shh mRNA expression in the palatal epithelium [7 , 8 , 28] . We examined whether palatal expression of these genes was affected by Foxf2 deficiency . By in situ hybridization analyses , however , we didn’t detect obvious differences in expression of Msx1 , Bmp4 , and Fgf10 , respectively , in the developing palatal shelves between Foxf2-/- mutant and wildtype littermates ( S1A–S1J Fig ) . Moreover , we performed quantitative RT-PCR analysis of manually microdissected palatal shelves from E13 . 5 wildtype , Foxf2+/- , and Foxf2-/- littermates but did not detect any significant differences in expression levels of Bmp4 , Fgf10 , and Msx1 mRNAs , respectively , between the samples of different genotypes . To gain a better understanding of the molecular mechanisms mediating Foxf2 function in palate development , we compared the transcriptome expression profiles of E13 . 5 Foxf2-/- and control palatal mesenchyme by using RNA-seq analysis . To facilitate RNA-seq analysis of palatal mesenchyme cells , and since it has been shown that Foxf2 and Osr2 are expressed in a similar oral-to-nasal gradient pattern in the developing palatal mesenchyme at E13 . 5 [10] , we took advantage of the Osr2RFP/+ knockin mice for isolation of palatal mesenchyme cells using fluorescence-activated cell sorting ( FACS ) . We crossed Foxf2+/- mice with Osr2RFP/+ mice to generate the Foxf2+/-Osr2RFP/+ mice and then set up timed mating of the Foxf2+/- female mice with Foxf2+/-Osr2RFP/+ male mice . We verified that RFP expression in the developing palate was not affected by Foxf2-deficiency ( Fig 4A and 4B ) . We harvested embryos at E13 . 5 , microdissected the palatal shelves from each RFP-positive embryo , and isolated the RFP-positive palatal mesenchyme cells by FACS . Following identification of the embryo genotypes , we performed RNA-seq analysis of the FACS-isolated palatal mesenchyme from Foxf2-/-Osr2RFP/+ , Foxf2+/-Osr2RFP/+ and Foxf2+/+Osr2RFP/+embryos , respectively . Differential expression analysis of the RNA-seq data identified 155 genes whose expression was up- or down-regulated by more than 1 . 5-fold in the Foxf2-/-Osr2RFP/+ palatal mesenchyme in comparison with both Foxf2+/-Osr2RFP/+ and Foxf2+/+Osr2RFP/+ samples ( S1 Table ) . Among these , Fgf18 , which encodes a member of the fibroblast growth factor family ligands , was up-regulated by more than 2-fold in Foxf2-/-Osr2RFP/+ mutant palatal mesenchyme compared with the control littermates ( S1 Table ) . Subsequent quantitative real-time RT-PCR analysis validated the significantly increased expression of Fgf18 in the Foxf2-/- mutant palatal mesenchyme ( Fig 4C ) . Consistent with results from whole mount in situ hybridization , expression of Shox2 was significantly decreased whereas expression of Osr2 and Barx1 were not significantly changed in the Foxf2-/- palatal mesenchyme in comparison with the control samples ( Fig 4C ) . We further compared the patterns of Fgf18 expression in the Foxf2-/- and littermate control embryos by whole mount in situ hybridization . Strikingly , we found that Fgf18 mRNAs were ectopically expressed in specific anterior and posterior sub-regions of the developing palatal shelves in the Foxf2-/- mutant embryos ( Fig 5A and 5B ) that correspond to where expression of both Shh and Ptch1 was dramatically downregulated in the Foxf2-/- mutant embryos in comparison with the control embryos ( compare Fig 5A and 5B with Fig 3A–3H ) . Further in situ hybridization analysis of serial coronal sections through the E13 . 5 palatal shelves confirmed ectopic expression of Fgf18 mRNAs in the palatal mesenchyme in the specific anterior and posterior regions in the Foxf2-/- mutant embryos while Fgf18 mRNA expression in the wildtype littermates was restricted to the mesenchyme cells at the hinge region of the palatal shelves ( Fig 5C–5F ) . To understand why the Fox2-/- mutant embryos exhibit region-specific changes in gene expression profiles along the anterior-posterior axis of the developing palatal shelves , we analyzed and compared the patterns of expression of Foxf2 and Foxf1 during palate development . From E12 . 5 to E13 . 5 , Foxf1 mRNA expression was restricted to the middle region of developing palatal shelves , with the strongest level of expression detected in the molar tooth germs ( Fig 6A and 6C ) . In contrast , Foxf2 mRNA expression was detected throughout the anterior-posterior axis of the developing palatal shelves , with the posterior region of the palatal shelves exhibiting higher levels of expression than the anterior region at both E12 . 5 and E13 . 5 ( Fig 6B and 6D ) . We further analyzed the distribution of the Foxf1 and Foxf2 proteins in the developing palatal shelves by immunofluorescent detection in Osr2RFP/+ knockin mouse embryos , which express the RFP reporter from the endogenous Osr2 locus that exhibits an oral-to-nasal gradient pattern of expression in the developing palatal mesenchyme [36] . At E13 . 5 , Foxf1 protein is expressed at moderate levels in the mid-anterior region of the palatal mesenchyme while it is expressed very weakly in the anterior-most region and absent in the posterior region of the developing palatal shelves ( Fig 6E–6H ) . Foxf2 protein is distributed throughout the anterior-posterior axis of the palatal mesenchyme , but with different patterns along the oral-nasal axis in the anterior versus posterior regions ( Fig 6I–6L ) . In the anterior region up to the level of the molar tooth germs , Foxf2 protein distribution exhibited an oral-to-nasal gradient , with highest levels in the mesenchyme immediately underneath the palatal epithelium at the oral side ( Fig 6I , 6J and 6K ) . In the palatal region posterior to the molar tooth germs , Foxf2 protein is expressed at high levels throughout the oral-nasal axis of the palatal mesenchyme ( Fig 6L ) . Neither Foxf1 nor Foxf2 protein was detected in the palatal epithelium . These data indicate that expression of Foxf1 and of Foxf2 are differentially regulated during palate development . Remarkably , the regions where expression of Fgf18 , Shh , and Ptch1 is significantly altered in the palatal shelves in Foxf2-/- embryos , compared with the wildtype littermates , correspond to the palatal regions where Foxf2 , but not Foxf1 , is expressed during palate development in wildtype embryos , suggesting that Foxf1 might complement Foxf2 function in the middle region of the developing palatal shelves in Foxf2-/- mutant embryos . Since Foxf1-/- mutant mouse embryos die during midgestation prior to palate morphogenesis [19] , which prevents a direct analysis of the role of Foxf1 in palate development in these mice , we generated and analyzed Foxf1c/cWnt1-Cre mouse embryos in which Foxf1 is tissue-specifically inactivated in the neural crest derived craniofacial mesenchyme . Whereas Foxf1c/+Wnt1-Cre mice appear normal , Foxf1c/cWnt1-Cre mice are born with cleft palate ( S2A and S2B Fig ) . Histological analysis of E16 . 5 embryos showed that the Foxf1c/cWnt1-Cre mutant embryos exhibit failure of palatal shelf elevation ( S2C and S2D Fig ) . However , in contrast to the Foxf2-/- mutant embryos , the patterns of expression of Fgf18 and Shh mRNAs in the developing palatal shelves were not significantly altered in the Foxf1c/cWnt1-Cre mutant embryos in comparison with the control littermates ( S3A–S3D Fig ) . To directly test the hypothesis that the region-specific molecular effects of Foxf2 deficiency on palate development is due to functional complementation by Foxf1 , we generated Foxf1c/cFoxf2c/cWnt1-Cre embryos and examined the patterns of Fgf18 and Shh expression by whole mount in situ hybridization . Remarkably , although the Foxf1c/cFoxf2c/cWnt1-Cre embryos have smaller palatal shelves compared with the control littermates , they exhibit ectopic Fgf18 expression specifically throughout the palatal shelves at E12 . 5 and E13 . 5 , compared with the highly restricted pattern of Fgf18 mRNA expression in control embryos ( Fig 7A–7D ) . Analysis by in situ hybridization of serial coronal sections confirmed ectopic expression of Fgf18 mRNAs from anterior to posterior regions of the palatal mesenchyme in the Foxf1c/cFoxf2c/cWnt1-Cre mutant embryos at E12 . 5 ( S4 Fig ) . Furthermore , we found that Shh mRNA expression in the palatal epithelium was dramatically lost throughout the anterior-posterior axis of the palatal shelves in Foxf1c/cFoxf2c/cWnt1-Cre mutant embryos in comparison with the control littermates ( Fig 7E and 7F ) . In addition , although the palatal shelves in the Foxf1c/cFoxf2c/cWnt1-Cre embryos still exhibited Shox2 expression in the anterior region and Barx1 in the posterior region , both the domain and level of Shox2 expression were dramatically reduced in comparison with that in the control littermates ( S5 Fig ) . These results suggest that Foxf1 and Foxf2 act partly redundantly to regulate Fgf18 and Shox2 mRNA expression in the palatal mesenchyme and act indirectly to maintain Shh expression in the palatal epithelium during early palate development . We next investigated whether the ectopic Fgf18 expression in the palatal shelves could account for loss of Shh expression in the palatal epithelium in the Foxf2-/- and Foxf1c/cFoxf2c/cWnt1-Cre mutant embryos . We placed Fgf18-soaked beads on the oral side of E13 . 0 palatal shelves in explant culture and examined effects of Fgf18 protein on Shh expression in the palatal epithelium after 24 hours of culture . As shown in Fig 8 , application of Fgf18 protein caused a dramatic down-regulation of Shh gene expression in the palatal epithelium whereas the BSA-soaked beads did not have any obvious effect , as detected by in situ hybridization of Shh mRNA expression ( Fig 8A ) and by detection of GFP reporter expressed from the ShhGFP allele in ShhGFP/+ embryonic palates ( Fig 8B ) . These results , together with the corresponding domains of ectopic Fgf18 expression in the palatal mesenchyme and of loss of Shh expression in the palatal epithelium in the Foxf2-/- and Foxf1c/cFoxf2c/cWnt1-Cre mutant embryos , respectively , indicate that Foxf1 and Foxf2 regulate palatogenesis , at least in part , through repressing Fgf18 expression to maintain Shh signaling to stimulate palatal shelf growth .
Although expression of both Foxf1 and Foxf2 mRNAs in the developing craniofacial and palatal mesenchyme depends on Shh signaling [9 , 10] , we detected Foxf2 mRNA expression throughout the anterior-posterior axis of the developing palatal mesenchyme but Foxf1 mRNA expression is absent from specific anterior and posterior sub-regions of the developing palatal shelves , suggesting that other molecular pathways converge with the Shh signaling pathway to differentially regulate Foxf1 and Foxf2 expression during palate development . Consistent with this hypothesis , Hoffmann et al . ( 2014 ) recently identified a Foxf1 cis-regulatory element that bound both Gli1 and the T-box transcription factor Tbx5 in the developing heart tissues and that Gli1 and Tbx5 synergistically activated transcription from this cis-regulatory element [41] . In embryonic lung explant culture assays , Foxf1 mRNA expression in the lung mesenchyme was shown to be induced by Shh and repressed by Bmp4 [44] . We recently reported that Bmp4 is expressed in an anterior and a posterior subdomain of the developing palatal mesenchyme [29] . Future studies will determine whether Foxf1 expression during palate development is directly and antagonistically regulated by Shh and Bmp4 signaling . Wang et al ( 2003 ) first showed cleft palate defect in the Foxf2-/- mutant mice , but suggested that the cleft palate phenotype might be secondary to defects in tongue muscle development because Foxf2 mRNAs are abundantly expressed in the muscle layers of the developing tongue in wildtype mouse embryos , and because they did not detect a significant difference in cell proliferation in the developing palate at E13 . 5 and E15 . 5 using BrdU labeling . Our analysis of BrdU labeling of the palatal mesenchyme detected significant reduction in palatal mesenchyme proliferation at E13 . 5 , at the peak of palatal shelf growth . The discrepancy in these findings are most likely due to differences in the BrdU labeling procedure and in data analysis . We injected the pregnant mice intraperitonially with 50 μg/g body weight BrdU and harvested the embryos one hour later . This procedure resulted in labeling of up to 40% of the palatal mesenchyme cells in E13 . 5 wildtype embryos . In contrast , Wang et al . ( 2003 ) injected the pregnant mice intraperitonially with 100 μg/g body weight and sacrificed the injected animals 2 hours later . At E13 . 5 , this procedure might have saturated labeling of the palatal mesenchyme cells . In addition , since the developing palatal shelves have both morphological and molecular heterogeneity [1] , we recorded the percentage of BrdU-labeled palatal mesenchyme cells on serial sections throughout the palatal shelves and analyzed the data separately in the six regions along the anterior-posterior and oral-nasal axes . It is not clear how Wang et al . analyzed the BrdU-labeling data . Given the heterogeneity of the palatal mesenchyme and our finding that the molecular effects of Foxf2-deficeincy on the developing palatal shelves are most pronounced in the anterior and posterior domains where Foxf2 , but not Foxf1 , is expressed during normal palatogenesis , it is possible that simple analysis of the BrdU-labeling index in the middle region of palatal shelves would not find significant differences in the mutant and control embryos . Indeed , our data show that cell proliferation index is not significantly different in the mid-oral portion of the developing palatal shelves in the Foxf2-/- and control littermates ( Fig 2G ) . The differential effects of loss of Foxf2 function on palatal cell proliferation along the anterior-posterior axis is mostly likely due to partial functional compensation by Foxf1 in the middle region of the developing palatal shelves . The fact that the Foxf1c/cFoxf2c/cWnt1-Cre compound mutant embryos exhibit only rudimentary palatal shelves supports the conclusion that Foxf1 and Foxf2 act partly redundantly to control palatal shelf growth . Our conclusion that Foxf2 plays an intrinsic role in palate development is also supported by the molecular effects of Foxf2-deficiency on palatal gene expression . The domain-specific changes in Fgf18 and Shh expression in the developing palatal shelves in the Foxf2-/- mutant embryos correlated with the lack of Foxf1 expression in the anterior and posterior regions of the palatal mesenchyme . Moreover , we found that Foxf1c/cFoxf2c/cWnt1-Cre embryos exhibit ectopic Fgf18 mRNA expression specifically throughout the palatal mesenchyme and loss of Shh mRNA expression throughout the palatal epithelium . Thus , similar to their functions in gut and heart development in which Foxf1 and Foxf2 exhibit synergistic effects [41 , 43] , Foxf1 and Foxf2 could partly complement for each other’s function during palate development in the regions where they exhibit overlapping expression . Whereas the cell proliferation and molecular studies clearly demonstrate an intrinsic role for Foxf2 in palate development , both Foxf2c/-Wnt1-Cre and Foxf2c/-Osr2IresCre/+ embryos exhibit defects in palatal shelf elevation and abnormal tongue shape ( Fig 1G and 1H ) . Whereas the Foxf2c/-Wnt1-Cre embryos are expected to have loss of Foxf2 function throughout the neural crest-derived craniofacial mesenchyme , including the non-muscle connective tissues in the tongue , which could cause a primary defect in tongue development , the Osr2IresCre/+ embryos exhibit only limited Cre activity in a subset of tongue mesenchyme cells directly underlying the tongue epithelium [10 , 25] . However , it is possible that loss of Foxf2 function in the small population of tongue mesenchyme cells directly underlying the epithelium also perturbs epithelial-mesenchymal interactions during tongue development . Thus , whether disruption of Foxf2 function in the developing tongue could secondarily affect palatal shelf elevation remains to be investigated by generation and analysis of mice with tissue-specific inactivation of Foxf2 in the developing tongue . Our RNA-seq analysis of Foxf2-/- and control embryonic palatal mesenchyme revealed that Foxf2-deficiency significant affected the expression of over 150 genes in the developing palate . Our in situ hybridization analysis revealed the striking pattern of ectopic Fgf18 expression in the Foxf2-/- mutant palate , which correlated well with the unique domains of Foxf2 , but not Foxf1 , expression in the developing palatal shelves in wildtype embryos . This , together with the data that Fgf18 is expressed throughout the palatal mesenchyme in the Foxf1c/cFoxf2c/cWnt1-Cre embryos suggests that Fgf18 is a direct target gene repressed by the Foxf transcription factors . Although Shh expression in the palatal epithelium also exhibits domain-specific loss in the Foxf2-/- mutant embryos , the loss of Shh expression in the palatal epithelium in the Foxf1c/cFoxf2c/cWnt1-Cre mutant embryos , in which Foxf1 and Foxf2 are specifically inactivated in the mesenchyme , indicate that the Foxf transcription factors indirectly regulate Shh expression . Together with the findings that exogenous Fgf18 protein inhibited Shh expression in palatal explant culture and our previously reported data that expression of Foxf1 and Foxf2 in the palatal mesenchyme depends on Shh signaling [10] , these results identify a novel negative feedback loop controlling Shh expression during palate development . Fgf18 belongs to the fibroblast growth factor family of ligands , which consists of 22 members and signal through alternatively spliced forms of tyrosine kinase receptors encoded by four distinct genes , Fgfr1 , Fgfr2 , Fgfr3 , and Fgfr4 [45 , 46] . Previous studies have implicated Fgf7 and Fgf10 in the regulation of Shh expression during palate development [8 , 28] . Whereas Fgf7 and Fgf10 share high amino acid sequence homology and both signal exclusively through the Fgfr2b receptor in epithelial tissues , they elicit distinct and sometime opposite cellular responses in developmental tissues as well as in cell culture assays [8 , 28 , 47 , 48] . During palate development , Fgf7 and Fgf10 exhibit complementary expression patterns in the developing palatal mesenchyme , with Fgf7 mRNAs preferentially expressed in the nasal side and Fgf10 mRNAs restricted to the oral side of the palatal mesenchyme [8 , 28] . Mice lacking Fgf10 or Fgfr2b exhibit cleft palate , with loss of Shh expression in the developing palatal epithelium [8] . In palatal explant culture assays , exogenous Fgf10 protein induced , whereas exogenous Fgf7 protein repressed , Shh mRNA expression , suggesting that Fgf7 antagonizes Fgf10 function to restrict Shh expression to the oral side of the palatal epithelium [8 , 28] . Although the mechanism underlying the opposite effects of Fgf7 and Fgf10 on Shh expression during palate development is not known , Francavilla et al . ( 2013 ) recently reported that Fgf10 specifically induced rapid phosphorylation of the tyrosine ( Y ) -734 residue on Fgfr2b , which led to the receptor recycling and enhanced and prolonged Fgfr signaling , whereas Fgf7 led to rapid degradation of the receptors [48] . It is plausible that inhibition of Shh expression in the palatal explant by exogenous Fgf7 could be mediated by Fgf7-induced Fgfr2b degradation , as Fgfr2b function is required for maintenance of Shh expression in the palatal epithelium [8] . It is possible that the ectopically expressed Fgf18 in the Foxf2-/- and Foxf1c/cFoxf2c/cWnt1-Cre mutant palatal mesenchyme might also cause reduction in Shh expression in the palatal epithelium by inducing rapid Fgfr2b degradation . However , in vitro studies and prediction from crystal structures suggested that Fgf18 lacks affinity for Fgfr2b [49 , 50] . On the other hand , Fgf18 has been shown to bind Fgfr3c and the cysteine-rich Fgf receptor [51 , 52] . The detailed molecular mechanism involving Fgf18-mediated regulation of Shh expression during palate development requires further investigation . Interestingly , mice lacking Fgf18 function exhibit high penetrance of cleft palate [53 , 54] . Moreover , genome-wide association studies of cleft lip and palate in humans have shown disease association with the FGF18 locus [55] . Thus , further investigation of the role and molecular mechanisms involving Fgf18 in palate development will directly improve our understanding of the genetic basis and molecular mechanisms of cleft palate pathogenesis in humans .
The Foxf1c/c , Foxf2c/c , Wnt1-Cre and Osr2IresCre/+ mice have been described previously [21 , 22 , 25 , 56] . Osr2RFP/+ ( JAX stock #010986 ) and ShhGFP/+ ( JAX stock #005622 ) mice were obtained from the Jackson Laboratory . The Osr2IresCre and Osr2RFP/+ mice were maintained by crossing to C57BL/6J mice . The Wnt1Cre mice were maintained by crossing with CD1 ( Charles river ) females . Foxf1c/c and Foxf2c/c mice were maintained by intercrossing homozygotes . Noon of the day a vaginal plug was identified was designated as embryonic day ( E ) 0 . 5 . This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals by the National Institutes of Health . The animal use protocol was approved by the Institutional Animal Care and Use Committee of Cincinnati Children’s Hospital Medical Center ( Permit Number IACUC2013-0036 ) . For histological analysis , embryos were dissected at desired stages from timed pregnant mice , fixed in 4% paraformaldehyde ( PFA ) , dehydrated through an ethanol series , embedded in paraffin , sectioned at 7μm thickness , and stained with alcian blue followed by hematoxylin and eosin . Immunofluorescent staining of paraffin sections was performed following standard protocols . Antibodies used are: rabbit anti-RFP ( MBL international , PM005 ) , Goat anti-Foxf1 ( R&D , AF4798 ) , and sheep anti-Foxf2 ( R&D , AF6988 ) . To determine cell proliferative activity in the developing palatal shelves , timed pregnant mice were injected intraperitoneally with BrdU ( 5 mg/ml stock solution , 10 μl/g body weight ) ( Sigma-Aldrich ) . Embryos were harvested 1 hour after injection , fixed with 4% PFA , paraffin embedded and sectioned at 7 μm . The BrdU-labeling index was defined as the number of BrdU-positive nuclei relative to total nuclei , which was co-stained by DAPI . The cell proliferation data were recorded from seven sections from each of the anterior , middle and posterior regions of each palatal shelf , and also analyzed separately for the oral and nasal halves of each region of the palatal shelves in each embryo . Data from two independent litters , each containing two wildtype and two Foxf2-/- embryos , were used for statistical analysis . Whole mount and section in situ hybridization was performed as previously described [36 , 57] . At least three embryos of each genotype were hybridized to each probe and only probes that detected consistent patterns of expression in all samples were considered as valid results . Palatal shelf explant culture and bead implantation experiments were carried out using a previously described protocol with minor modification [28] . Briefly , Timed pregnant mice were sacrificed on post-coital day 13 . 0 ( E13 . 0 ) . The embryonic maxillary processes with the secondary palatal shelves were manually microdissected and cultured in BGJb medium supplementary with 10 U/ml penicillin/ streptomycin ( Invitrogen ) , 50 mM transferrin ( Sigma ) and 150 μg/ml ascorbic acid ( Sigma ) . For bead implantation , Affi-Gel blue agarose beads ( BioRad ) were soaked in recombinant Fgf18 proteins ( 1mg/ml , Peprotech ) , or BSA ( 1mg/ml ) as control . Tissues were harvested after 24 hours of culture at 37°C at an atmosphere of 5% CO2 and 100% humidity and fixed in 4% paraformaldehyde for whole mount in situ hybridization experiments . The palatal shelves of E13 . 5 embryos from Foxf2+/- females crossed with Foxf2+/- Osr2RFP/+ males were manually microdissected and digested with trypsin-EDTA solution ( Invitrogen ) at 37°C for 4 minutes . After inactivation of trypsin with DMEM containing 10% FBS , cells were dissociated by pipetting . The dissociated palatal cells were resuspended in PBS with 2% FBS and 10 mM EDTA , and filtered through a 40 μm nylon cell strainer ( BD Falcon , 352340 ) . RFP+ cells were isolated using BD FACSAria II . FACS-isolated RFP+ palatal mesenchyme cells from two E13 . 5 Foxf2-/-Osr2RFP/+ embryos and one Foxf2+/-Osr2RFP/+ and one Osr2RFP/+ littermates were used for RNA-seq experiment . Foxf2+/-Osr2RFP/+ and Osr2RFP/+ samples were used as controls . Sequencing libraries were generated by using Illumina Nextera DNA Sample Prep kit and sequenced using Illumina HisEq 2000 . Sequence reads were mapped to the reference mouse genome ( mm9 ) using Bowtie . Single-end reads were aligned using Tophat . RNA-seq data were then analyzed using Strand NGS software , with the reads per kilobase exon per million mapped sequences value calculated for each RefSeq gene for relative levels of gene expression . For analyses of differential expression , the fold-change cutoff was set at 1 . 5 or higher . P value less than 0 . 05 from the Audic Claverie test was considered statistically significant , with Benjamini–Hochberg false discovery rate multiple testing correction [58] . The original RNA-seq data files from this study have been deposited into the National Center for Biotechnology Information Gene Expression Omnibus ( NCBI GEO ) database ( accession number GSE67015 ) . First-strand cDNAs were synthesized using SuperScript First-Strand Synthesis System ( Invitrogen , 11904–018 ) . Primers for specific transcripts were designed for real-time RT-PCR ( SYBR ) . β-Actin was used as internal control in each reaction . Real-time PCR was performed using a Bio-Rad CFX96 Real-Time System using conditions recommended by the manufacturer . Each reaction was performed in duplicate . The quantity of each mRNA was first determined using a standard curve method and normalized to the internal control . The primers used for real-time RT-PCR are listed in S2 Table . All results were presented as mean ± SEM . All statistical analyses were done using Excel software . Two-tailed Student’s t tests were used for comparisons between two groups . P value less than 0 . 05 was considered significant . | Cleft lip and/or cleft palate ( CL/P ) are among the most common birth defects in humans , occurring at a frequency of about 1 in 500–2500 live births . The etiology and pathogenesis of CL/P are complex and poorly understood . Generation and analysis of mice carrying targeted null and conditional mutations in many genes have revealed that functional disruption of each of more than 100 genes could cause cleft palate . However , how these genes work together to regulate palate development is not well understood . In this study , we identify a novel molecular circuit consisting of two critical molecular pathways , the fibroblast growth factor ( FGF ) and Sonic hedgehog ( SHH ) signaling pathways , and the Forkhead family transcription factors Foxf1 and Foxf2 , mediating reciprocal epithelial-mesenchymal signaling interactions that control palatogenesis . As mutations affecting each of multiple components of both the FGF and SHH signaling pathways have been associated with CL/P in humans , our results provide significant new insight into the mechanisms regulating palatogenesis and cleft palate pathogenesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2016 | A Shh-Foxf-Fgf18-Shh Molecular Circuit Regulating Palate Development |
The interferon inducible protein , BST-2 ( or , tetherin ) , plays an important role in the innate antiviral defense system by inhibiting the release of many enveloped viruses . Consequently , viruses have evolved strategies to counteract the anti-viral activity of this protein . While the mechanisms by which BST-2 prevents viral dissemination have been defined , less is known about how this protein shapes the early viral distribution and immunological defense against pathogens during the establishment of persistence . Using the lymphocytic choriomeningitis virus ( LCMV ) model of infection , we sought insights into how the in vitro antiviral activity of this protein compared to the immunological defense mounted in vivo . We observed that BST-2 modestly reduced production of virion particles from cultured cells , which was associated with the ability of BST-2 to interfere with the virus budding process mediated by the LCMV Z protein . Moreover , LCMV does not encode a BST-2 antagonist , and viral propagation was not significantly restricted in cells that constitutively expressed BST-2 . In contrast to this very modest effect in cultured cells , BST-2 played a crucial role in controlling LCMV in vivo . In BST-2 deficient mice , a persistent strain of LCMV was no longer confined to the splenic marginal zone at early times post-infection , which resulted in an altered distribution of LCMV-specific T cells , reduced T cell proliferation / function , delayed viral control in the serum , and persistence in the brain . These data demonstrate that BST-2 is important in shaping the anatomical distribution and adaptive immune response against a persistent viral infection in vivo .
Arenaviruses are enveloped viruses with a bi-segmented , negative strand RNA genome and a life cycle restricted to the cell cytoplasm [1] . Each genome segment uses an ambisense coding strategy to direct the expression of two proteins in opposite orientation and separated by a non-coding intergenic region . The large segment encodes for the RNA-dependent-RNA-polymerase ( L ) and the matrix protein ( Z ) that mediates viral assembly and budding [2–4] . The small segment ( S; 3 . 5 kb ) encodes the glycoprotein ( GP ) precursor , GPC , and the viral nucleoprotein ( NP ) . GPC is co-translationally cleaved by signal peptidase to produce a stable 58 amino acid Stable Signal Peptide ( SSP ) and GPC that is post-translationally processed by the cellular Site 1 Protease to yield the two mature virion glycoproteins ( GP1 and GP2 ) that together with SSP form the GP complex involved in receptor binding and virus cell entry . GP1 mediates virion attachment to the cell surface followed by cell entry via receptor-mediated endocytosis , whereas GP2 is responsible for the pH-dependent fusion event in the acidic environment of the endosome to complete the virus cell entry process and release of virus ribonucleoprotein into the cell cytoplasm to initiate transcription and replication of the viral genome [1] . The L polymerase and NP are the minimal trans-acting factors required for virus RNA replication and gene transcription [5] , whereas production of infectious particles also requires GP and Z [2] . Several arenaviruses cause hemorrhagic fevers in humans [1] . Thus , Lassa ( LASV ) and Junin ( JUNV ) viruses , the causative agents of Lassa fever and Argentine hemorrhagic fever [6] , respectively , pose important public health concerns within their endemic regions of West Africa ( LASV ) and Argentina ( JUNV ) . Notably , increased travel has resulted in importation of Lassa fever cases into non-endemic regions [7] . Moreover , evidence indicates that the globally distributed prototypic arenavirus lymphocytic choriomeningitis virus ( LCMV ) is a neglected human pathogen of clinical significance , especially in congenital viral infections [8] . Besides their impact on public health , arenaviruses also pose a credible bioterrorism threat , and six of them , including LASV and JUNV , have been classified as Category A agents . Concerns about arenavirus infections of humans are exacerbated by a limited availability of arenavirus countermeasures . The live attenuated Candid#1 strain of JUNV was shown to be an effective vaccine against Argentine hemorrhagic fever [9] , but Candid#1 is licensed exclusively in Argentina and does not protect against Lassa fever . Likewise , current anti-arenaviral drug therapy is restricted to an off label use of the nucleoside analogue ribavirin , which is only partially effective and associated with significant side effects [10] . Therefore , there is an unmet need to identify novel compounds that could be developed into FDA-approved antiviral drugs to combat human pathogenic arenaviruses—a task that would be facilitated by a better understanding of virus-host interactions that regulate different steps of the arenavirus life cycle . BST-2 ( a . k . a . tetherin , CD317 , or HM1 . 24 ) is a type I interferon ( IFN-I ) -inducible cellular protein that was initially identified as an inhibitor of HIV-1 release and was subsequently shown to inhibit cell release of a wide range of enveloped viruses including LASV [11 , 12] . Several viral gene products , including HIV-1 Vpu and Ebola virus GP [13–16] , were demonstrated to counteract the antiviral activity of BST-2 . Here , we sought insights into whether BST-2 interferes with the release and propagation of the prototypic arenavirus , LCMV , both in cultured cells and during the establishment of a persistent viral infection in mice . Our results demonstrate how a relatively modest antiviral effect in vitro can translate into a large impact on antiviral immunity in vivo .
All mice in this study were handled in accordance with the guidelines set forth by the NIH Animal Care and Use Committee . C57BL/6J ( B6 ) mice were purchased from The Jackson Laboratory . BST-2 KO [17] ( provided by Dr . Marco Colonna , Washington University ) , P14 [18] , Thy1 . 1+ P14 , mOrange+ P14 [19] , SMARTA [20] , and CD45 . 1+ SMARTA , ( all on a pure B6 background ) were bred and maintained under specific pathogen–free conditions at the National Institutes of Health ( NIH ) . Plasmids pC-LCMV-Z-FLAG , pTeth-FL , pGFP , pCEboZVP40 [21–25] , as well as p-T7 , pMG-CAT , pCAGGS-NP , and pCAGGS-L [5 , 26–29] have been described . Plasmid psiCHECK2 was purchased from Promega ( Madison , WI ) . The mouse monoclonal antibody ( MAb ) to the FLAG epitope was purchased from SIGMA ( M2 , St . Louis , MO ) . The mouse MAb to LCMV NP has been described [30] . Anti-human BST-2 polyclonal antibody was provided from NIH AIDS Reagent Program ( Catalog number: 11721 , provided by Drs . Klaus Strebel and Amy Andrew ) [31] . Anti-Ebola virus VP40 polyclonal antibody was described previously [23] . HeLa-pLKO ( control HeLa cell line ) , HeLa-TKD ( BST-2 stably knocked down HeLa cell line ) , 293T , BHK-21 , VeroE6 , Huh7 . 5 . 1 and Huh7 . 5 . 1/BST2 ( Huh7 . 5 . 1 constitutively expressing high levels of BST-2 ) cells were grown in Dulbecco’s modified Eagle’s medium ( DMEM; Invitrogen , Carlsbad , CA ) containing 10% fetal bovine serum ( FBS ) , 1% penicillin and streptomycin . To generate Huh7 . 5 . 1/BST2 cells , we cloned the human BST-2 gene into a lentiviral vector that expressed also , via an IRES sequence , the gene of resistance to puromycin . We produced VSV-G pseudotyped lentiviral particles expressing BST-2 by co-transfecting HEK293T cells with HIV-1 packaging plasmids and the lentiviral vector expressing BST-2 . At 48 hours post-transfection , we collected and concentrated lentiviral particles present in tissue culture supernatant . We used dilutions of the prep of BST-2 expressing lentiviral particles to infect Huh 7 . 5 . 1 cells to determine the dilution that resulted in 50% of cells expressing BST-2 at 72 hours post-transduction . This dilution was used to transduce Huh 7 . 5 . 1 cells followed by selection in the presence of puromycin ( 8 μg/ml ) . BST-2 expression in puromycin selected cells was determined by Western blot using an antibody to BST-2 . Anti-actin mouse monoclonal antibody ( AC-40 , Sigma ) was used to detect actin . Lentivirus transduced cells expressing BST-2 were further selected by FACS to select a population of cells expressing high levels of BST-2 , called Huh-7 . 5 . 1/BST-2 . HeLa-pLKO , HeLa-TKD and Huh7 . 5 . 1/BST2 cells were maintained in the presence of puromycin ( 8 μg/ml ) . 2 . 5x105 293T cells were transfected with 0 . 2 μg of pC-LCMV-Z-FLAG using LT-1 ( 3 μl LT-1/μg DNA , Mirus ) . At 48 hrs post-transfection , VLP-containing tissue culture supernatants ( TCS ) and cells were collected . After clarification from cell debris ( 1 , 500 x g; 5 min ) , VLPs were collected by ultracentrifugation ( 100 , 000 x g; 30 min at 4°C ) through a 20% sucrose cushion . Cells and VLPs were re-suspended in lysis buffer ( 1% NP-40 , 50 mM Tris-HCl [pH 8 . 0] , 62 . 5 mM EDTA , 0 . 4% sodium deoxycholate ) and analyzed by Western blot . For Ebola virus ( EBOV ) VP40-mediated VLPs , 293T cells were transfected with 0 . 1 μg of pEboZVP40 for 5 h , followed by infection with rLCMV/Z-FLAG ( moi = 5 ) , or mock-infected . At 16 hrs p . i . , VLP-containing TCS and cells were collected . After clarification from cell debris ( 1 , 500 x g; 5 min ) , VLPs were collected by ultracentrifugation ( 345 , 000 x g for Ebola VLPs; 100 , 000 x g for LCM VLPs; 30 min at 4°C ) through a 20% sucrose cushion . Cell lysates and VLPs were prepared and analyzed by WB . Bands corresponding to Z and VP40 were quantified using Multi Gauge software ( Ver2 . 0 , Fuji Film ) . Cell lysates or VLP samples prepared as described above , were resolved by SDS-PAGE followed by Western blot using the indicated antibodies . Flag-tagged Z and BST2 proteins were detected with an anti-Flag mouse monoclonal antibody , and antigen-antibody complexes revealed using an HRP-conjugated anti-mouse IgG antibody . Endogenous BST-2 was detected using a rabbit polyclonal anti-BST2 serum , followed by an HRP-conjugated anti-rabbit IgG antibody . Ebola virus VP40 was detected using a rabbit polyclonal antibody to VP40 , followed by an HRP-conjugated anti-rabbit IgG antibody . Actin , used as loading control , was detected using a mouse monoclonal antibody to actin , followed by an HRP-conjugated anti-mouse IgG antibody . ECL prime ( GE healthcare ) and LAS3000 ( GE healthcare ) were used to detect labeled proteins . Total cellular RNA was isolated using TRI Reagent ( Molecular Research Center , Inc , Cincinnati , OH ) per the manufacturer’s instructions , and isolated RNA was analyzed by Northern blot hybridization . Briefly , RNA samples were fractionated by 2 . 2 M formaldehyde-agarose ( 1 . 2% ) gel electrophoresis followed by transfer ( 4 hrs ) in 20 X SSC of the RNA to a Magnagraph ( 0 . 22 μm ) membrane using the rapid downward transfer system ( TurboBlotter ) . Membrane bound RNA was crosslinked by exposure to UV and the membrane was hybridized to a 32P-labeled strand specific probe to the MG-derived CAT mRNA . The Armstrong ( Arm ) and Clone 13 ( Cl-13 ) strains of LCMV [32 , 33] as well as tri-segment recombinant LCMV ( 3rLCMV/GFP ) [21 , 34 , 35] and rLCMV/Z-FLAG [35] have been described . Mice were infected intravenously with 2x106 PFU of LCMV Arm or Cl-13 . LCMV titers were determined using an immunofocus assay [36] or by plaque assay using Vero cells . For the immunofocus assay , 10-fold serial virus dilutions were used to infect Vero cell monolayers in a 96-well plate , and at 20 hrs p . i . , cells were fixed with 4% paraformaldehyde ( PFA ) in PBS . After cell permeabilization by treatment with 0 . 3% Triton X-100 in PBS containing 3% BSA , cells expressing viral antigen were stained by using a mouse MAb to LCMV NP and an Alexa Fluor 568-labeled anti-mouse second antibody ( Molecular Probes , Eugene , OR ) . For the plaque assay , 10-fold serial virus dilutions were used to infect Vero cells in M-24 well plates . After 90 min adsorption , the virus inoculum was removed , and cell monolayers were washed once and overlaid with DMEM containing 0 . 5% agarose , 1% Glutamine , 1% penicillin and streptomycin , and 0 . 7% FBS . After six days incubation , cells were fixed with 4% PFA and plaques were visualized by crystal violet staining . Titers of vesicular stomatitis Indiana virus ( VSV ) were determined by plaque assay using BHK-21 cells or Vero cells . 293T cells were seeded ( 4 . 5 x 105/well ) on M-12 well plates and the following day transfected with p-T7 , pMG-CAT , pCAGGS-NP , pCAGGS-L , and either pTeth-FL or pGFP under the conditions described [5 , 26–29] . At 24 hrs post-transfection cell lysates were prepared to determine levels of CAT protein by ELISA using a CAT ELISA kit ( Roche , Basel , Switzerland ) and using an ELISA reader ( SPECTRA max plus 384 , Molecular Devices , Sunnyvale , CA ) to determine the absorbance ( 405 nm for samples , 490 nm for the reference ) . 293T cells were seeded ( 5x104/well ) on 96 well plate and transfected with psiCHECK2 ( 100 ng/well ) and indicated amount of pTeth-FL using LT-1 . Total plasmid volume was adjusted with pcDNFL ( empty ) vector . At 48 hrs post-transfection cell lysates were prepared for Dual-Glo Reporter Assay ( Promega ) to detect both Firefly luciferase ( Fluc ) and Renilla Luciferase ( Rluc ) , respectively , according to the manufacturer’s recommendations . Values of luciferase activity were measured using a luminometer ( Centro LB 960 , Berthold technologies , Bad Wildbad , Germany ) . Luciferase activity values were normalized assigning 1 . 0 to those obtained with lysates from vehicle ( DMSO ) -treated cells . Cell viabilities of increasing amounts of exogenous BST-2 expression in 293T cells were assessed using the CellTiter-Glo Luminescent Cell Viability Assay ( Promega ) . The assay was performed according to the manufacturer’s recommendations and readings were obtained using a luminometer ( Centro LB 960 ) . Cell viability of control plasmid transfected cells was normalized and set at 1 . 0 . HeLa cells infected with rLCMV/Z-FLAG were fixed using 4% PFA at 24 or 48 hrs p . i . and permeabilized by treatment with 0 . 3% Triton X-100 in PBS containing 3% BSA . Cells were stained using a mouse MAb to FLAG ( M2 ) as the 1st antibody for 2 hrs at room temperature ( RT ) , followed by Alexa Fluor 568-labeled anti-mouse second antibody ( Molecular Probes ) . To detect endogenous intracellular BST-2 , cells were reacted with a rabbit polyclonal serum to BST-2 [31] for 2 hrs at RT , followed by an Alexa 488-conjugated anti-rabbit antibody for 2 hrs at RT . Cell nuclei were identified by DAPI staining . Samples were examined by laser confocal microscopy ( LSM780 ELYRA system; Carl Zeiss , Oberkochen , Germany ) . Anesthetized mice received an intracardiac perfusion with PBS or saline to remove all contaminating blood lymphocytes . Single cell suspensions from spleen were prepared by mechanical disruption through a 100-μm strainer followed by red blood cell lysis with ammonium chloride buffer ( 0 . 017 M Tris-HCl and 0 . 14 M NH4Cl , pH 7 . 2 ) . For experiments involving transfer of Thy1 . 1+ P14 , mOrange+ P14 , and CD45 . 1+ SMARTA T cells , the respective populations were purified from the spleens of naive transgenic mice using negative selection kits ( STEMCELL Technologies ) . For co-transfer experiments , naive recipient mice were seeded i . v . with 2 , 000 Thy1 . 1+ P14 and 2 , 000 CD45 . 1+ SMARTA T cells . For experiments to determine the anatomical distribution of antiviral CD8+ T cells , mice were seeded i . v . with 10 , 000 mOrange+ P14 . Mice for both experimental lines were infected one day later with LCMV Cl-13 . To detect cell surface expression of BST-2 in vitro , cells were washed with PBS once and treated with Accutase ( AT104; Innovative Cell Technologies , Inc . , San Diego , CA ) . Cells were collected in PBS containing 1% FBS and separated into two tubes for PE-conjugated mouse monoclonal control antibody ( MOPC-21; BioLegend , San Diego , CA ) or PE-conjugated mouse monoclonal anti-BST-2 antibody ( RS38E; BioLegend ) staining . After antibody staining , cells were fixed with 2% PFA for 15 min at RT and analyzed by FACS ( Caliber , BD , San Jose , CA ) . For flow cytometry experiments involving splenocytes , cells were incubated for 20 min on ice with cocktails of mAbs in PBS containing 2% FBS . Before staining , all cell preparations were incubated with 3 . 3 μg/ml rat anti–mouse CD16/32 ( Fc receptor block; BD ) and 1:50 whole mouse IgG ( Jackson ImmunoResearch Laboratories , Inc . ) for 10 min on ice to reduce unspecific antibody binding . Dead cells were excluded from the analysis by using the LIVE/ DEAD fixable Blue Cell Staining kit ( ThermoFisher Scientific ) . The following antibodies were obtained from BioLegend ( BL ) , BD , or eBioscience ( eB ) : B220 PE ( RA3-6B2; BD ) , BST-2 Pacific Blue ( 927; BL ) , CD11b BV605 ( M1/70; BL ) , CD11c PE/Cy7 ( N418; BL ) , CD19 APC-Cy7 ( 6D5; BL ) , CD4 APC ( RM4-5; BL ) , CD4 BV605 ( RM4-5; BL ) , CD45 . 1 PECy7 ( A20; BL ) , CD45 . 2 AF700 ( 104; BL ) , CD45 . 2 FITC ( 104; BD ) , CD45 . 2 BV421 ( 104; BL ) , CD8 BV510 ( 53–6 . 7; BL ) , CD8 FITC ( 53–6 . 7; BL ) , Thy1 . 2 AF700 ( 30-H12; BL ) , Thy1 . 1 PE ( OX-7; BD ) , and Thy1 . 1 PECy7 ( HIS51; eB ) . To detect intracellular cytokine production , single cell suspensions were surface stained as described above , treated with cytofix/cytoperm ( BD ) , and then stained intracellularly with anti-IFN-γ PE ( XMG1 . 2; BD ) , TNF-α FITC ( MP6-XT22; BD ) , and IL-2 APC ( JES6-5H4; BD ) . Samples were acquired using an LSRII flow cytometer ( BD ) , and data were analyzed using FlowJo software version 10 . 0 . 7 ( Tree Star ) . Two million splenocytes were plated in 96-well round bottom plates in RPMI complete media ( RPMI; 10% FBS , 1% penicillin/streptomycin , 1% L-glutamine , 1% HEPES , 1% nonessential amino acids , 1% sodium pyruvate , 50 μM 2-mercaptoethanol , 1 μg/ml of Brefeldin A ) with 2 μg/ml GP33-41 peptide ( KAVYNFATC; Anaspec ) [37] or 4 μg/ml GP61-80 peptide ( GLNGPDIYKGVYQFKSVEFD; Anaspec ) [38] at 37°C for five hours . T cell proliferation was measured by carboxyfluorescein diacetate N-succinimidyl ester ( CFSE; Molecular Probes ) dilution . Naïve Thy1 . 1+ P14 cells were incubated for 10 min at 37°C in PBS containing 0 . 1% BSA and 5 μM CFSE . Following a wash , 5x105 labeled T cells were injected i . v . into naïve mice , which were then infected with LCMV . For immunohistochemical experiments , mice were perfused with saline or 4% paraformaldehyde ( PFA ) in PBS . The latter was used for experiments involving fluorescent protein expressing transgenic P14 cells . Spleens extracted from PFA-perfused mice were incubated overnight in 4% PFA in PBS and then for an additional 24 hr in a 30% sucrose solution . Tissues were then frozen on dry ice in Tissue-tek optimal cutting temperature medium ( Thermo Fisher Scientific ) . 6-μm cryosections were fixed for 10 min with 4% PFA , washed three times with PBS , blocked with avidin-biotin blocking kit ( Vector Laboratories ) per the manufacturer’s instructions , and stained with primary antibodies overnight at 4°C in PBS containing 2% FBS . The following were used as primary antibodies: rat anti-LCMV ( 1:1 , 000; clone VL-4; Bio X Cell ) , polyclonal anti-laminin ( 1:500; Abcam ) , biotinylated anti-BST-2 ( 1:500; clone 927; Biolegend ) , and anti-CD169-FITC ( 1:400; clone 3D6 . 112; Serotec ) . After washing three times with PBS , tissue sections were stained with species-specific fluorescently conjugated secondary antibodies ( 1:400; Jackson ImmunoResearch Laboratories , Inc . ) or streptavidin rhodamine red X ( 1:400; Jackson ImmunoResearch Laboratories , Inc . ) for 1 hr at RT , washed , and costained with 10 ng/ml DAPI ( Sigma-Aldrich ) to label cell nuclei . All slides were mounted with Vectashield ( Vector Laboratories ) , and fluorescent images were acquired using a FV1200 confocal microscope ( Olympus ) equipped with an automated xyz stage , six laser lines ( 405 , 458 , 488 , 515 , 559 , and 635 nm ) , and 4 , 10 , 20 , and 40x objectives . In Fig 4D , the percentage of LCMV+ pixels in the splenic white versus red pulp were calculated by segmenting out the signal corresponding to LCMV staining . The segmented images were then used to calculate the total number of white vs . red pulp LCMV+ pixels . The LCMV+ pixel counts were divided by the white vs . red pulp pixel areas and multiplied by a hundred to generate percentages of tissue occupied by LCMV . In Fig 6B , the percentage of P14 cells in the white vs . red pulp was calculated by identifying fluorescently labeled cells in splenic images and then calculating the proportion that localized to these distinct anatomical regions .
To assess whether cellular endogenous levels of BST-2 could restrict cell release , and thereby propagation , of LCMV progeny in vitro , we examined whether RNAi-mediated knock-down of BST-2 in HeLa cells , known to express high levels of BST-2 constitutively [31] , affected cell release of LCMV particles . For this we transduced HeLa cells with a lentivirus expressing the pac gene that mediates resistance to puromycin and shRNAs to either BST-2 ( HeLa-TKD ) , or a control shRNA ( HeLa-pLKO ) . Transduced cells were cloned by limited dilution and selected in the presence of puromycin . We confirmed by FACS analysis that cell surface expression of BST-2 was greatly reduced in HeLa-TKD cells compared to control HeLa-pLKO cells ( Fig 1A ) . We then infected HeLa-pLKO and HeLa-TKD cells with rLCMV/Z-FLAG [35] , and at the indicated times post-infection ( p . i . ) we determined titers of cell-free virus infectious progeny ( Fig 1B ) . At 48 hrs p . i . production of infectious LCMV was slightly higher ( four-fold ) in HeLa-TKD cells than HeLa-pLKO control cells . Consistent with this finding , we observed a slight increase in the numbers of viral antigen positive cells at 48 h p . i . in HeLa-TKD compared to HeLa-pLKO cells ( Fig 1C ) . Expression levels of LCMV Z protein at 16 h p . i . were similar between HeLa-pLKO and HeLa-TKD cells ( Fig 1D ) , suggesting that constitutive expression of BST-2 in HeLa cells did not significantly affect the translation efficiency of viral mRNAs or the early steps of the virus life cycle leading to the release of the vRNP into the cell cytoplasm where it directs virus RNA replication and gene transcription . We next examined whether the modest increase in production of infectious LCMV progeny in HeLa-TKD compared to HeLa-pLKO cells correlated with similar differences in production of total virion particles . For this , we infected HeLa-pLKO and HeLa-TKD cells with rLCMV/Z-FLAG , and at 24 hrs p . i . virion particles present in tissue culture supernatant were collected by ultracentrifugation and cell lysates prepared for protein analysis . Levels of Z protein in cell lysates and virions were determined by Western blot using an anti-FLAG antibody ( Ab ) ( Fig 1D and 1E ) . Levels of Z protein in virion particle preparations were slightly higher ( 1 . 7-fold ) in HeLa-TKD compared to control HeLa-pLKO cells , whereas lysates from infected HeLa-TKD and HeLa-pLKO cells had similar Z protein levels . Consistent with previous findings [39] , we observed that VSV multiplication was enhanced by knock-down of BST-2 ( Fig 1G ) . To investigate whether BST-2 interfered with cell release of LCMV particles , we examined the effect of BST-2 on the efficiency of Z-mediated virus-like particle ( VLP ) production using a well-established assay [2 , 3] . For this we transfected 293T cells with a plasmid ( 0 . 2 μg ) expressing a FLAG-tagged Z and increasing amounts of a plasmid expressing a FLAG-tagged BST-2 , and at 48 h post transfection , we collected VLPs from tissue culture supernatants by ultracentrifugation and prepared cell lysates . Levels of Z protein present in VLPs and cell lysates , and BST-2 levels in cell lysates , were detected by Western blot using an antibody to FLAG . BST-2 exerted a dose-dependent inhibitory effect on VLP production ( Fig 2A ) . It should be noted that we observed this inhibitory effect of BST-2 on Z-mediated VLP production only when we used low ( ≤ 0 . 1 μg ) amounts of BST-2 expressing plasmid to transfect cells . The use of higher amounts of BST-2 expression plasmid resulted in a very dramatic reduction in intracellular expression levels of Z that prevented an accurate assessment of Z-mediated VLP production . We also examined whether BST-2 could directly affect the activity of the LCMV polymerase complex . For this we used an LCMV minigenome ( MG ) rescue assay [5] . We transfected 293T cells with optimized amounts of plasmids required for the intracellular reconstitution of an active LCMV vRNP that directs expression of the chloramphenicol acetyl transferase ( CAT ) reporter gene , together with increasing amounts of a BST-2 expressing plasmid . BST-2 exhibited a dose-dependent inhibitory effect on the expression levels of MG-directed chloramphenicol acetyltransferase ( CAT ) reporter gene expression ( Fig 2B ) . The interpretation of this finding was complicated by our observation that BST-2 appears to generally inhibit pol-II mediated expression in transient transfection assays . This was evidenced by the reduced expression of the Firefly and Renilla luciferase genes , using psiCHECK2 plasmid , which possesses the Firefly luciferase gene driven by HSV-TK promoter and Renilla luciferase gene driven by SV40 promotor , respectively ( Fig 2C and 2D ) . Increasing amount of exogenous BST-2 did not , however , affect cell viability ( Fig 2E ) . To further investigate the effect of BST-2 over-expression on LCMV multiplication , we compared the propagation and production of LCMV infectious progeny between Huh7 . 5 . 1 , which have non-detectable levels of BST-2 , and Huh7 . 5 . 1 cells that constitutively express high levels of BST-2 ( Huh7 . 5 . 1/BST2 ) . To facilitate the assessment of cell-to-cell propagation of LCMV , we used a tri-segmented recombinant LCMV that expressed GFP ( r3LCMV/GFP ) [34] . Numbers of GFP positive cells were similar in both Huh7 . 5 . 1 and Huh7 . 5 . 1/BST2 cells at 24 hrs p . i . following challenge with different multiplicities of r3LCMV/GFP infection ( moi = 0 . 001 , 0 . 01 , and 0 . 1 ) ( S1A Fig ) . In addition , both Huh7 . 5 . 1 and Huh7 . 5 . 1/BST2 cells exhibited similar kinetics ( S1B Fig ) and levels of viral RNA synthesis ( both replication and transcription ) following infection with LCMV as determined by Northern blot ( S1C Fig ) . We also examined the effect of BST-2 over-expression on production of LCMV infectious progeny in 293T cells . For this we transfected 293T cells with either BST-2 or control plasmid , followed by infection with LCMV ( moi = 0 . 01 ) ( S2A Fig ) . At 48 hrs p . i . we collected tissue culture supernatant to determine LCMV infectious titers . We observed a very modest ( two-fold ) , statistically significant , decrease in production of infectious LCMV progeny by 293T cells that were transfected with BST-2 ( S2A Fig ) . This experiment could have underestimated the inhibitory effect of BST-2 on LCMV release due to the contribution of viral titer from cells that did not get transfected . However , consistent with previous findings [39] , we observed about five-fold reduction in VSV infectious titers produced by BST-2-transfected compared to control 293T cells transfected with pGFP ( S2B Fig ) . Intriguingly , we observed that 293T cells over-expressing BST-2 , but not GFP , became highly resistant to de novo LCMV infection ( S2C Fig ) , which likely accounted for the modest reduction in production of infectious virus particles . Several enveloped viruses have been shown to encode gene products that can counteract the antiviral activity of BST-2 via different mechanisms [13–16 , 40] , including down-regulation of BST-2 on the cell surface , degradation of intracellular BST-2 , and sequestration of BST-2 in intracellular compartments different from the site where virion release takes place . We therefore examined whether LCMV infection affected the overall and cell surface BST-2 expression levels , or subcellular localization of BST-2 . For this we infected HeLa cells with LCMV ( moi = 0 . 01 ) , and at 48 hrs p . i . we determined expression levels of BST-2 by Western blot using an anti-BST-2 Ab [31] . We did not observe differences in BST-2 expression levels between LCMV- and mock-infected control HeLa cells ( Fig 3A ) . Likewise , we did not observe significant differences in cell surface expression of BST-2 between LCMV- and mock-infected control HeLa cells as determined by FACS analysis ( Fig 3B ) . To assess whether LCMV infection altered the subcellular distribution of BST-2 , we infected HeLa cells with rLCMV/Z-FLAG ( moi = 0 . 01 ) , and at 24 and 48 hrs p . i . we fixed the cells with 4% PFA and stained them with antibodies to both BST-2 and FLAG . Both rLCMV/Z-FLAG and mock-infected control cells exhibited the same subcellular distribution of BST-2 , which localized predominantly to the peri-nuclear region in juxtaposition to the LCMV Z protein ( Fig 3C ) . To further confirm that LCMV infection does not antagonize BST-2-mediated restriction of virion release , we examined the effect of LCMV infection on production of Ebola virus ( EBOV ) matrix protein ( VP40 ) -mediated VLP production , a process known to be inhibited by BST-2 expression [13 , 41 , 42] . 293T cells were transfected with plasmids expressing EBOV VP40 alone or together with BST-2 , followed by infection with rLCMV/Z-FLAG ( moi = 5 ) . At 16 hrs p . i . we collected VLP-containing tissue culture supernatant and cells . After clarification from cell debris , we collected VLPs by ultracentrifugation and then analyzed cell lysates and VLPs by Western blot for VP40 expression . Ectopic over-expression of BST-2 significantly reduced EBOV VP40-mediated VLP production , a finding consistent with published data [41] . LCMV infection did not affect levels of VP40-mediated VLP production and did not alleviate BST-2 induced inhibition of VP40-mediated VLP production ( S3 Fig ) . We next examined whether the modest effect of BST-2 on LCMV multiplication in cultured cells had any implications in the establishment of a persistent viral infection in vivo . To this end , we infected wild type ( WT ) and BST-2 knockout ( KO ) mice intravenously ( i . v . ) with LCMV clone 13 ( Cl-13 ) , a strain known to establish a chronic viral infection in immune competent adult mice [32 , 33] . Immunohistochemical analysis of WT spleens at day 3 p . i . revealed increased expression of BST-2 in the marginal zone and red pulp relative to uninfected control mice and infected BST-2 KO mice ( Fig 4A ) . Following i . v . inoculation , LCMV localizes to the splenic marginal zone [43] , the area where we observed elevated BST-2 expression levels ( Fig 4A ) . Further delineation of the anatomy was achieved using CD169 staining . CD169 is a cell adhesion molecule found on the surface of splenic marginal zone macrophages . CD169+ macrophages demarcated the ring of splenic BST-2 staining in WT mice at day 3 p . i . ( Fig 4B ) . In addition to marginal zone macrophages , BST-2 was also expressed in B cells , CD4+ T cells , CD11b+ myeloid cells , and plasmacytoid DCs ( pDCS ) at this time point ( S4A Fig ) . Based on the anatomical expression pattern of BST-2 at early times after Cl-13 infection , we postulated that BST-2 might contribute to the early confinement of LCMV within the splenic marginal zone . To examine this hypothesis , we compared the anatomical distribution of Cl-13 in the spleens of WT vs . BST-2 KO mice at day 3 p . i . ( Fig 4C ) . Consistent with our hypothesis , Cl-13 was no longer confined to the splenic marginal zone in BST-2 KO mice . Elevated expression of viral NP was observed in the splenic white and red pulp ( Fig 4C and 4D ) , indicating that the virus had escaped from the marginal zone in the absence of BST-2 . Because early viral control is a critical component in the development of successful adaptive immune responses , we next examined the impact of BST-2 deficiency on the development of LCMV-specific CD8+ and CD4+ T cell responses . As surrogates for the anti-viral T cell response , we selected two well-described immunodominant peptides for ex vivo stimulation assays: GP33-41 ( CD8 ) and GP61-80 ( CD4 ) [37 , 38] . Prior to infection , most splenic immune cell populations were comparable between WT and BST-2 KO mice ( S4B Fig ) . In BST-2 KO mice at day 8 p . i . , we observed a significant reduction in the absolute number of splenocytes ( Fig 5A ) , already suggesting a reduction in T cell clonal expansion . This was confirmed by our ex vivo peptide stimulation assays , which revealed a significant reduction in the percentage of endogenous IFNγ+ TNFα+ GP61-80-specific CD4+ T cells and GP33-41-specific CD8+ T cells in BST-2 KO mice relative to WT controls ( Fig 5B ) . To ensure that this defect was not linked to a T cell intrinsic function of BST-2 , we adoptively transferred a combination of 2 , 000 WT DbGP33-41 Thy1 . 1+ CD8+ T cells and 2 , 000 I-AbGP61-80 CD45 . 1+ CD4+ T cells i . v . into WT and BST-2 KO mice . These T cell receptor transgenic cell populations are commonly referred to as P14 [18] and SMARTA cells [20] , respectively . One day following injection of naïve P14 and SMARTA cells , mice were infected with Cl-13 . As with the endogenous anti-viral T cell response , the absolute number ( Fig 5C ) and function ( Fig 5D ) of both P14 and SMARTA cells were significantly reduced in day 8 BST-2 KO mice relative to WT controls . These data indicate that the effect of BST-2 deficiency on the adaptive response is T cell extrinsic . The impact of BST-2 deficiency on LCMV-specific T cell function at day 8 post-infection was quite significant . We therefore assessed whether this dysfunction was set into motion at earlier time points post-infection . The splenic red pulp and marginal zone are known to express an abundance of immunoregulatory molecules ( e . g . IL-10 and PD-L1 ) following Cl-13 infection [44 , 45] , and we postulated that early viral movement into these regions could draw anti-viral T cells away from the splenic white pulp where they are normally primed . To address this possibility , we adoptively transferred naïve mOrange+ P14 cells into WT and BST-2 KO , and quantified their anatomical distribution in the spleen at day 4 p . i . Interestingly , we observed a higher proportion of P14 CD8+ T cells in the splenic red pulp , and a corresponding decrease in the white pulp , of BST-2 KO mice relative to WT controls ( Fig 6A and 6B ) . We also observed a decreased percentage of carboxyfluorescein succinimidyl ester ( CFSE ) diluted P14 cells in spleens of BST-2 KO mice at this same time point , indicative of reduced proliferation ( Fig 6C and 6D ) . These data demonstrate that virus-specific CD8+ T cells mis-localize in the spleens of BST-2 KO mice , which likely causes them to proliferate less during the critical early stages of T cell priming . The dysfunctional anti-viral T cell response observed in BST-2 KO mice led us to ultimately assess the impact this has on viral control in vivo . Cl-13 is known to persist in most peripheral tissues , including the blood , until ~day 60 p . i . [46] . In BST-2 KO mice , we observed elevated LCMV clone 13 titers in the blood beginning at day 5 , and at all time points measured thereafter , until the virus was ultimately controlled at day 90 ( Fig 7A ) . Control of Cl-13 in BST-2 KO mice required 30 additional days compared to WT mice ( Fig 7A ) . The brain can serve as a reservoir for long term viral persistence in Cl-13-infected WT mice [47] . Interestingly , viral titers remained significantly elevated in the brains of BST-2 KO at nearly 8 months p . i . ( Fig 7B ) . Lastly , we tested whether BST-2 also plays a role in controlling an acute LCMV infection . LCMV Armstrong ( Arm ) differs from Cl-13 by 3 amino acids , and ARM is usually cleared within a week of i . v . inoculation , and is difficult to detect in the blood at any time point p . i . [32 , 33 , 46] . However , in BST-2 KO mice , we noted significantly elevated titers of LCMV Arm in the blood at day 5 p . i . ( Fig 7C ) . These results indicate that BST-2 plays a role in controlling LCMV in vivo .
Innate immunity is the first line of host defense against viral infections . Accordingly , to facilitate the completion of a successful infection many viruses encode molecules that counteract different components of the host innate immune response . Arenavirus NP has been shown to interfere with both induction of IFN-I [49–52] and activation of NF-kB , a critical player in the host inflammatory response to infection [53 , 54] . Likewise , we have documented that the IFN-I inducible BST-2 interferes with LASV Z/GPC-induced VLP production [22] . Similar to that described for HIV-1 [22 , 25] , BST-2 appears to retain LASV VLP on the cell surface . Consistent with this observation , BST-2 was reported to inhibit production of LASV infectious progeny [55] . However , whether these findings are unique to LASV or generally applicable to arenaviruses , remains unknown . In this study , we investigated the contribution of BST-2 to the antiviral defense mounted against the prototypic arenavirus , LCMV . Knock-down of BST-2 in HeLa cells ( Fig 1A ) , known to express a constitutively high level of BST-2 , caused a reduction , although very modest , on production of infectious progeny ( Fig 1B ) that correlated only partially with a similar minimal reduction in production of virus particles ( Fig 1C ) . BST-2 did not affect LCMV entry , virus RNA replication , transcription , or viral mRNA translation efficiency ( Fig 1D ) . These findings suggested , as documented for other viruses [11 , 22 , 55] , that BST-2 might also affect LCMV cell release , a process driven by the LCMV Z protein [2] . To evaluate this hypothesis , we examined the effect of BST-2 on Z-mediated VLP release and found that BST-2 exhibited a dose-dependent inhibitory effect on Z-mediated VLP release ( Fig 2A ) , a finding consistent with those documented for LASV [22] and other enveloped viruses [11 , 12 , 41 , 55] . Unexpectedly , BST-2 inhibited in a dose-dependent manner replication and expression of an LCMV mini genome ( MG ) ( Fig 2B ) . However , the interpretation and significance of this finding is confounded by our finding that in cell transfection assays , BST-2 exhibited a dose-dependent inhibitory effect on expression of co-transfected Renilla or Firefly luciferase , suggesting that under our experimental conditions , BST-2 expression might interfere with RNA Pol-II polymerase mediated gene expression ( Fig 2C and 2D ) , which is consistent with other published results [56] . Likewise , 293T cells over-expressing plasmid supplied BST-2 exhibited a rather modest restriction to LCMV multiplication ( S2A Fig ) . Unexpectedly , we observed that 293T cells expressing very high levels of BST-2 were rather refractory to LCMV infection ( S2C Fig ) , a finding that might have accounted for the minor differences in production of LCMV infectious progeny between BST-2 expressing and control 293T cells . Constitutive expression of BST-2 in Huh7 . 5 . 1 did not have any significant effect on LCMV propagation and production of infectious LCMV progeny ( S1A Fig and S1B Fig ) . Accordingly , LCMV RNA synthesis , both genome replication and gene transcription , were not affected in Huh7 . 5 . 1/BST2 compared to Huh7 . 5 . 1 cells ( S1C Fig ) . The apparent discrepancy between observations presented in S1 Fig and S2 Fig is best explained by the significantly higher expression levels of BST-2 in transfected 293T cells compared to Huh7 . 5 . 1/BST-2 cells . Several enveloped viruses encode molecules that counteract the antiviral activity of BST-2 via a variety of mechanisms . For example , HIV-1 Vpu induces degradation of BST-2 [11 , 12] , whereas Ebola virus GP sequesters BST-2 from the virus budding membrane region [13 , 42 , 57–61] . Cell surface expression , overall protein expression , and subcellular localization of BST-2 were not affected upon LCMV infection ( Fig 3 ) . This in addition to the inability of LCMV infection to counteract the BST-2-mediated inhibitory effect on the budding activity of VP40 ( S3 Fig ) further supported our conclusion that LCMV does not encode a BST-2 antagonist . Despite the relatively modest effect of BST-2 on LCMV in vitro , the absence of this antiviral protein played a critical role in shaping the early viral distribution and subsequent immune defense in vivo . Following intravenous inoculation , LCMV localizes initially to the splenic marginal zone [43] . This area is inhabited by CD169+ marginal zone macrophages that can capture materials from the blood and bear resemblance to subcapsular sinus macrophages in draining lymph nodes [62] . The speed and efficiency with which LCMV moves from marginal zone macrophages into the surrounding white and red pulp likely influences its ability to establish persistence . Interestingly , we observed elevated BST-2 expression in the marginal zone region where LCMV Cl-13 localized early after infection ( Fig 4 ) . This is consistent with local induction of BST-2 synthesis by IFN-I [63] . The functional importance of early expression of this antiviral protein during LCMV infection was established in BST-2 KO mice , where we observed a failure to strictly confine LCMV Cl-13 to the splenic marginal zone at day three post-infection . BST-2 is known to have properties that promote immune function [17 , 64–66] , but this failure to confine LCMV Cl-13 is most consistent with the ability of BST-2 to sequester or “tether” viruses [63] . BST-2 is known to affect replication of some viruses [17 , 63 , 67–71] , but not others [72 , 73] in vivo . This is likely due to the ability of BST-2 to bind specific viral proteins , the degree of IFN-I induced BST-2 synthesis , and the role of BST-2 in directly promoting antiviral immune responses . The contribution of BST-2 to the immune defense against LCMV Cl-13 is intriguing , because its absence induced profound antiviral T cell dysfunction in a cell extrinsic manner ( Fig 5 ) . LCMV Cl-13 is known to induce immune exhaustion in LCMV-specific CD8+ and CD4+ T cells [32 , 74 , 75] , but this functional exhaustion was significantly elevated in BST-2 KO mice . In addition , BST-2 deficiency reduced the ability of wild type LCMV-specific CD8+ T cells to proliferate , which was likely due to elevated splenic viral titers and the early migration of CD8+ T cells into the splenic red pulp ( Fig 6 ) and / or faulty priming by BST-2 KO dendritic cells [65 , 66] . The splenic red pulp is known to express high levels of immunoregulators ( e . g . IL-10 and PD-L1 ) that suppress antiviral T cell cytokine production and proliferation [44 , 45] . We postulate that BST-2 plays an important role in filter organs like the spleen by confining circulating viruses to the splenic marginal zone . This gives the adaptive immune system time to respond and eventually eradicate the pathogen . However , in the absence of BST-2 , a persistence prone virus like LCMV Cl-13 gained an advantage by escaping from the marginal zone more quickly . In fact , serum viral titers were significantly elevated in BST-2 KO mice , even though Cl-13 was eventually controlled in the serum by day 90 ( Fig 7 ) . Titers remained elevated in the brain , which is considered a site of long term persistence following LCMV Cl-13 infection [47] . We also observed that a non-persistent strain of LCMV ( Armstrong ) had an advantage in BST-2 KO mice . Collectively , these data demonstrate BST-2 is an important component of the IFN-I inducible defense against LCMV . By interfering with the budding process mediated by the LCMV Z protein , BST-2 appears to slow down the spread of this virus both in vitro and in vivo . While the inhibition observed in vitro was modest , the antiviral impact of this protein was amplified into a much larger effect in vivo . These findings highlight the importance of sequestering viruses to specific anatomical compartments in vivo . Viral confinement to regions like the splenic marginal zone can shape the ensuing adaptive immune response and influence states of persistence . It will be important in future studies to determine whether the spread of other more pathogenic arenaviruses ( e . g . LASV and JUNV ) are similarly influenced by BST-2 . | BST-2 ( or , tetherin ) is an antiviral protein that inhibits the release and spread of many enveloped viruses and is upregulated as part of the innate immune defense against infections . By studying the lymphocytic choriomeningitis virus ( LCMV ) model of infection , we sought insights into how BST-2 shapes the early viral distribution and immunological defense against a virus during the establishment of persistence . We also studied how the antiviral activity of BST-2 in cell culture compared to the immunological defense mounted by this protein in LCMV-infected mice ( in vivo ) . We observed that BST-2 only modestly affected the production of LCMV in cultured cells by interfering with the viral budding process , but played a crucial role in controlling LCMV in vivo . In the absence of BST-2 , a persistent strain of LCMV was ineffectively contained in the spleen at early time points post-infection , resulting in impaired antiviral T cell responses and viral control systemically . These data demonstrate that BST-2 is important in shaping the anatomical distribution and adaptive immune response against a persistent viral infection in vivo . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"hela",
"cells",
"293t",
"cells",
"spleen",
"immunology",
"biological",
"cultures",
"microbiology",
"viral",
"structure",
"cell",
"cultures",
"cytotoxic",
"t",
"cel... | 2018 | BST-2 controls T cell proliferation and exhaustion by shaping the early distribution of a persistent viral infection |
Herpesviruses infect most humans . Their infections can be associated with pathological conditions and significant changes in T cell repertoire but evidences of symbiotic effects of herpesvirus latency have never been demonstrated . We tested the hypothesis that HCMV and EBV-specific CD8 T cells contribute to the heterologous anti-viral immune response . Volume of activated/proliferating virus-specific and total CD8 T cells was evaluated in 50 patients with acute viral infections: 20 with HBV , 12 with Dengue , 12 with Influenza , 3 with Adenovirus infection and 3 with fevers of unknown etiology . Virus-specific ( EBV , HCMV , Influenza ) pentamer+ and total CD8 T cells were analyzed for activation ( CD38/HLA-DR ) , proliferation ( Ki-67/Bcl-2low ) and cytokine production . We observed that all acute viral infections trigger an expansion of activated/proliferating CD8 T cells , which differs in size depending on the infection but is invariably inflated by CD8 T cells specific for persistent herpesviruses ( HCMV/EBV ) . CD8 T cells specific for other non-related non persistent viral infection ( i . e . Influenza ) were not activated . IL-15 , which is produced during acute viral infections , is the likely contributing mechanism driving the selective activation of herpesvirus specific CD8 T cells . In addition we were able to show that herpesvirus specific CD8 T cells displayed an increased ability to produce the anti-viral cytokine interferon-γ during the acute phase of heterologous viral infection . Taken together , these data demonstrated that activated herpesvirus specific CD8 T cells inflate the activated/proliferating CD8 T cells population present during acute viral infections in human and can contribute to the heterologous anti-viral T cell response .
Over the course of the human lifetime , we are exposed to and infected by many different organisms which may be eliminated or may persist . The co-existence of microorganisms in humans is mainly perceived to have negative consequences for health and wellbeing , but examples of potential symbiotic relationship between the host and microbes start to be recognized [1] , [2] . Classic examples of microorganisms establishing persistent infections in humans are Epstein Barr virus ( EBV ) and human cytomegalovirus ( HCMV ) which are both from ubiquitous herpesviridae family of viruses which infect more than 90% of the human populations . These viruses are associated with the development of specific tumors ( i . e . Burkitt's Lymphoma ) and they can reactivate with significant pathological consequences in immunocompromised hosts [3] , [4] . Nevertheless , in most of the cases , herpesvirus infections are subclinical and well tolerated , even though they cause a robust distortion of T cell repertoire [5] , [6] with HCMV and EBV-specific CD8 T cell known to represent up to 20% of total CD8 T cell population [7] , [8] , [9] . Our inherent effort to maintain such a large population of virus-specific T cells , is seen as a necessity to suppress CMV and EBV reactivation in humans [9] , [10] , [11] . This would imply that the sole function of herpesvirus specific memory effector CD8 T cells is to act against CMV and EBV infected cells . However , evidence in animal models have shown that effector or memory CD8 T cells can provide immune protection against infection with unrelated intracellular pathogens through production of Interferon γ ( IFN-γ ) [12] . Such data open the possibility that the large population of HCMV and EBV-specific CD8 T cells present in humans might contribute to the immunological response against other pathogens . Thus , we set out to evaluate whether CD8 T cells specific for herpesviruses can contribute to the anti-viral T cell response triggered by heterologous acute viral infection in humans . CD8 T cell responses to acute viral infections were analyzed sequentially ( from onset to recovery ) by measuring the population of activated/proliferating CD8 T cells in patients with acute Hepatitis B Virus ( HBV ) , influenza , dengue and adenovirus infections . The combination of activation and proliferation markers ( CD38 , HLA-DR , Ki-67 and Bcl-2 ) expressed by CD8 T cells have been recently proposed to identify the whole population of virus-specific effector CD8 T cells induced by viral infection [13] . These results were obtained in subjects receiving attenuated virus vaccines ( Smallpox and Yellow Fever ) , and activation ( CD38/HLA-DR ) and proliferation markers ( Ki-67/Bcl-2 low ) were only expressed by CD8 T cells specific for the vaccine but not by CD8 T cells of different specificities . In contrast , we demonstrate here that acute symptomatic viral infections trigger an expansion of activated/proliferating CD8 T cell populations of variable sizes , comprising CD8 T cells specific for the infecting virus but these populations are also invariably inflated by CD8 T cells specific for persistent herpesvirus infections . The increased sensitivity of HCMV and EBV-specific CD8 T cells to IL-15 is the likely explanation of this in vivo observation . In addition , HCMV and EBV specific CD8 T cells demonstrate , at the peak of acute infection , an increased ability to secrete IFN-γ suggesting that they might functionally contribute to the heterologous acute anti-viral immunity .
We initially evaluated the size and the expansion kinetics of CD8 T cell population during acute hepatitis B infection . The frequency and quantity of CD8 T cells expressing CD38/HLA-DR and Ki-67/Bcl-2 phenotypic markers was analyzed in 20 patients with acute hepatitis B . Samples were collected at multiple time points from onset of disease ( HBsAg+ , ALT>1000 U/L ) to full recovery ( HBsAg- at least 1 month after onset ) . A remarkably large expansion of activated CD8 T cell pool was detected . CD38/HLA-DR markers were expressed by approximately a quarter of total CD8 T cells ( mean 23% , range 12–68% ) at the onset of clinical hepatitis . The frequency of CD38/HLA-DR+ CD8 T cells decreased consistently at the second time point ( 8–10 days later , mean 12%; range 4–22% ) and at the time of recovery it returned to the normal level ( mean 3% , range 0 . 9–10% ) , detectable in healthy controls ( Figure 1 , left panel ) . CD8 T cells co-expressing Ki-67 and low Bcl-2 followed identical kinetics . The peak of Ki-67/Bcl-2 low CD8 T cells was detected at the onset of disease ( mean 14% , range 4 . 5–27% of total CD8 T ) and contracted abruptly after 10 days ( mean 5% , range 0 . 8–11% ) and at the resolution minimal proliferation was detected ( mean 0 . 8% , range 0 . 4–1 . 6% ) ( Figure 1 , right panel ) . To analyze whether the population of activated/proliferating CD8 T cells included HBV-specific CD8 T cells , HBV-specific pentamers were used to directly visualize these cells in 5 HLA-A201+ patients . Figure 2 A and B shows that the expression of activation markers of HBV-specific CD8 T cells followed the kinetics of expression of the total CD8 T cell population . HBV-pentamer+ CD8 T cells expressed activation markers and proliferated at the onset of disease but not at the recovery phase ( Figure 2 A and B ) . These results demonstrate that HBV-specific CD8 T cells are represented within the total population of activated/proliferating total CD8 T cells . Then we tested whether CD8 T cell specific for other common viruses ( CMV , EBV , influenza ) quantitatively contribute to the total pool of activated/proliferating CD8 T cells . A set of HCMV , EBV or Influenza pentamers ( Supplementary Table S1 ) was used to detect CD8 T cells specific for these common infections . We visualized a sizeable ex vivo frequency of HCMV , Influenza and EBV-specific CD8 T cells in 13 acute hepatitis B patients and their expression of CD38/HLA-DR and Ki-67 was tested at the onset of acute hepatitis and after recovery . A remarkably different profile of CD8 T cell activation was detected in relation to the CD8 T cell specificity . While influenza-specific CD8 T cells were neither activated ( 8 out of 8 patients ) nor proliferating ( 5 out of 5 tested patients ) at all time points ( Figure 3 A and Supplementary Figure S1 ) , HCMV and EBV-specific CD8 T cells were activated ( HCMV mean 12 . 5%; EBV mean 30% ) and proliferating ( HCMV mean 4 . 9%; EBV mean 8% ) ( Tables 1–2 and Figure 3 A ) in all the acute HBV patients where such cells were detectable . The expression of CD38/HLA-DR and Ki-67 markers in HCMV and EBV-specific CD8 T cells followed the same expression kinetics of total and HBV-specific CD8 T cells and contracted after recovery as shown on Figure 3 B ( patient 12 ) . The differential phenotype of CD8 T cells specific for different viruses during acute hepatitis B was well represented in a patient ( patient 10 , Supplementary Figure S1 ) where the different CD8 T cells specificities co-exist in different activation states . As already shown in Figure 2 , at the peak of acute hepatitis , HBV-specific CD8 T cells are mostly activated ( 77% ) and proliferating ( 65% ) . At the same time point , a proportion of HCMV-specific CD8 T cells are also expressing activation ( 20% ) and proliferation ( 6% ) markers , while influenza-specific are in a complete resting phenotype . Taken together , these data demonstrated that , at least in acute hepatitis B infection , a sizable proportion of CD8 T cells specific for persistent viruses are activated during acute heterologous infection . Evidence of activation of unrelated virus specific CD8 T cells has been also reported in HIV infection [14] , but our results clearly differ from the ones obtained in attenuated virus vaccine recipients [13] , where activation of Influenza , HCMV and EBV was not reported . Thus we tested whether our observation was peculiar to acute HBV infection or whether it represents a common feature in other acute viral infections in human . Samples from patients with acute Dengue ( n 12 ) , Influenza A ( n 12 ) , Adenovirus ( n 3 ) infections were collected at the onset of disease ( represented in these patients by fever >38° C ) , after 5–7 days and after recovery ( ∼21 days ) and frequency of CD8 T cell population expressing activation ( CD38/HLA-DR ) and proliferation ( Ki-67 ) markers was measured ( Figure 4 A ) . Differences in the magnitude and kinetics among diseases with different etiology were found . Adenoviral infection elicited a minimal activation of CD8 T cell population ( mean 3 . 5% ) , which is only slightly higher than that of healthy individuals ( mean of 5 healthy controls 2 . 4% ) . In addition , the peak frequency of activated total CD8 T cells in dengue and influenza infections is detected 5–7 days after onset of fever unlike that of HBV , where the peak frequency is seen at the onset of disease ( Figure 4 A ) . These different profiles are compatible with the fact that the onset of disease in acute hepatitis ( jaundice ) is represented by liver injury and coincides with the peak of adaptive immune response [15] , [16] while dengue and influenza infections trigger a strong innate immune reaction ( febrile status being a clinical manifestation ) and thus , in these infections , full maturation of virus-specific adaptive immunity is expected to peak ∼5–7days after infection . Nevertheless , despite the lower quantity of total activated CD8 T cells in dengue , influenza , adenovirus patients as well as in 3 subjects with fever of unidentified etiology , the CD38/HLA-DR expression profile on HCMV or EBV specific CD8 T cells was similar to that in acute HBV infection . Figure 4 B summarizes the results obtained in the patients with the indicated pathologies , where a sizeable ex vivo frequency of HCMV or EBV specific CD8 T cells was detected . HCMV and EBV specific CD8 T cells ( lower panel ) express activation markers to a level even higher to what is detected in the global CD8 T cells populations ( upper panel ) . Unfortunately , the paucity of the cells obtained in these patients didn't allow us to analyze also Ki-67 expression on HCMV and EBV specific CD8 T cells , but overall these results demonstrate that activation of CD8 T cells specific for persistent viral infection ( HCMV-EBV ) is a constitutive feature of acute anti-viral immunity in human . Interestingly , we were able to study a patient with acute influenza infection in whom , influenza-specific CD8 T cell expansion didn't coincide with the profile of total activated CD8 T cells ( Figure 4 C , P14 ) . In this subject , the influenza-specific CD8 T cells ( specific for matrix protein 58–66 epitope ) could be visualized only 5 days after onset of symptoms . In contrast , HCMV-specific ( pp65 123–131 ) CD8 T cell frequency was comparatively constant at different time points ( 1 . 7% onset; 1 . 8% +5 days; 1 . 4% +14 days ) , and already co-expressed activation markers ( 30% ) at the onset of disease ( Figure 4 C ) . Thus , in this patient , the whole CD38/HLA-DR+ population before the expansion of the CD8 T cells specific for the acutely infected virus seems to be composed of herpesvirus-specific activated CD8 T cells . We investigated the possible mechanisms of the selected activation of HCMV and EBV- specific CD8 T cells in patients with heterologous acute viral infections . CD8 T cell cross-reactivity , reactivation of the HCMV or EBV infection and/or activation mediated by cytokines can be implicated in this phenomenon . Cross-reactivity between HCMV or EBV-specific CD8 T cells with epitopes present in the acute heterologous virus infection seems unlikely . The cross-reactive potential of HCMV-specific CD8 T cells is very uncommon [7] and our data do not support cross-reactive mechanisms either . We could detect activation of CD8 T cells specific for two distinct immediate early and latent EBV epitopes ( HLA-B8 RAKFKQLL , BZLF-1 190–197 , and HLA-B8 FLRGRAYGL , EBNA-3A 193–201 ) in an acute HBV patient ( Supplementary Figure S2 ) . If cross-reactivity was responsible of this activation , it would require that both epitopes share sequence or structural similarity with HBV virus , an unlikely scenario , based on a sequence similarity search ( NCBI PubMed BLAST ) , which demonstrated no sequence overlap ( lowest E value obtained = 11 ) between these EBV epitopes and HBV proteome . Reactivation of HCMV and EBV could be a plausible cause , and it might explain why CD8 T cells specific for Influenza are not activated in acute HBV infections . To investigate this possibility , HCMV and EBV DNA levels were tested longitudinally in the serum . However , we did not find any evidence of HCMV or EBV reactivations . HCMV-DNA and EBV-DNA titers were below the level of detection in all patients ( HBV , Dengue , Influenza , Adenovirus and fever of unidentified etiology ) from the onset of acute heterologous viral infections to recovery ( data not shown ) . Importantly , although HCMV and EBV reactivations are usually associated with the expansion of HCMV/EBV- specific CD8 T cells [10] , [17] , [18] significant changes in the EBV or HCMV specific T cells quantity were not observed through the course of acute infections ( Figure 3 B and Supplementary Figure S2 ) . We therefore analyzed whether cytokines produced during acute viral infections [19] can be responsible for the differential expression of activation markers by EBV- , HCMV- and influenza-specific CD8 T cells . PBMC or purified CD8 T cells of healthy subjects containing resting EBV , HCMV and Influenza specific CD8 T cells were incubated with different concentrations of IL-15 , IL-2 , IL-7 , IFN-γ , IFN-α and TNF-α and the expression of HLA-DR and CD38 on EBV , HCMV and Influenza specific CD8 T was analyzed at different intervals ( Figure 5 A ) . We detected that after 24 and 48 hours of incubation , IL-15 ( at 1 and 10 ng/ml ) induced CD38/HLA-DR expression in HCMV and EBV specific CD8 T cells while the other inflammatory cytokines did not activate EBV and HCMV specific CD8 T cells . Similar to the in vivo findings , influenza-specific CD8 T cells were not or only weakly activated by addition of any of the tested cytokines ( Figure 5 A and B ) . Prolonged incubation times ( 3 to 5 days ) did not alter the activation profile . Figure 5 B shows the results obtained in one healthy subject where HCMV , EBV and Influenza specific CD8 T cells were simultaneously detected . Incubation of total PBMC with IL-15 induces expression of CD38/HLA-DR molecules in EBV and HCMV specific CD8 T cells ( 44% and 37% respectively ) but only in few influenza-specific CD8 T cells ( 7% ) . The specific effect of IL-15 on HCMV and EBV-specific CD8 T cells was confirmed in other healthy subjects where individual specificities were detected ( HCMV = n4; EBV = n4 ) . Similar results were obtained incubating total PBMC or CD3+ CD8+ purified cells ( not shown ) . Thus , IL-15 , a cytokine that has been shown to induce T cell activation in mice [20] and human [21] , [22] and is known to be produced during acute viral infections ( [19] and personal data ) induces preferential CD38 , HLA-DR up-regulation of HCMV and EBV-specific CD8 T cells rather than influenza-specific ones . Having observed that a proportion of HCMV and EBV-specific CD8 T cells are activated during heterologous acute viral infection , we sought to analyze their functional profile . The limited quantity of cells available in patients with acute viral infections precludes an extensive evaluation of the functional profile directly in our patient sample . Thus , since IL-15 mimics the differential activation state of HCMV , EBV and influenza-specific CD8 detected in patients with acute viral infections , we performed a series of functional experiments using PBMC of healthy individuals activated with IL-15 . We first tested whether IL-15 can differentially trigger T cell activation in HCMV , EBV and Influenza specific CD8 T cells in vitro . PBMC of healthy individuals were incubated with or without IL-15 for 48 hours and HCMV , EBV and influenza-specific CD8 T cells were tested for their ability to produce anti-viral cytokines ( IFN-gamma , IL-2 and TNF-alpha ) using intracellular cytokine staining . Note that the cytokines measurement on CD8 T cells specific for the different viruses requires their visualization with the specific HLA-class I/peptides pentameric complex ( pentamers ) . The pentamer staining can potentially trigger T cell stimulation through direct interaction of the TCR with the synthetic MHC-class I peptide complexes of the pentamers[23] , [24] , [25] . Thus , to distinguish whether IL-15 can directly trigger T cell activation ( TCR-independent stimulation ) or enhance the T cell activation triggered by MHC/peptide pentamer ( TCR-dependent stimulation ) , the intracellular production of IFN-gamma on CD8 T cells was analyzed adding the MHC/peptide pentamers either before ( TCR-dependent stimulation ) or after ( TCR-independent stimulation ) the incubation time of intracellular cytokine staining . A schematic representation of the experimental design is presented in Figure 6 A . In accordance with previous studies [21] , [22] , IL-15 elicited a spontaneous production of IFN-gamma on T cells . However , the level of IFN-gamma production was modest and present in CD8 cells irrespective of their specificity . Dot plots displayed in Figure 6 A illustrate these results obtained in one representative subject . Increased production of other cytokines ( IL-2 , TNF-alfa ) was less striking ( not shown ) . In contrast , we observed that HCMV and EBV-specific CD8 T cells incubated with IL-15 and stained with MHC/peptide pentamer at the beginning of the intracellular cytokine assay showed an increased ability to produce IFN-gamma . More than 70% of IL-15 pulsed HCMV and EBV-specific CD8 T cells produced high quantity of IFN-gamma while in the absence of IL-15 , MHC-pentamer staining stimulate only a minority of HCMV and EBV-specific CD8 cells ( Figure 6 A–B ) . Importantly , IL-15 incubation has a modest effect on Influenza specific CD8 T cells ( Figure 6 A–B ) . Thus , our in vitro experiments showed that IL-15 is not only able to preferentially activate HCMV and EBV-specific CD8 T cells , but can also modulate their functional responsiveness to the TCR-dependent stimulation mediated by MHC-pentamer staining . Having defined a different functional profile on in vitro activated HCMV and EBV-specific CD8 T cells , we tested whether such features could be detected in vivo . In line with the experiments in vitro , MHC-peptide pentamer stimulation was detected preferentially on HCMV and EBV-specific CD8 cells present during the acute phase of HBV infection ( Figure 6 C and D ) . Figure 6 C shows the results obtained in a representative patient ( P24 ) with acute hepatitis and with a sizeable population of activated HCMV-specific CD8 T cells ( 20% ) . While the spontaneous production of IFN-gamma was identical in HCMV-specific CD8 cells present at the onset and at recovery of acute hepatitis B ( Figure 6 C- unstimulated ) , 12% of HCMV specific CD8 cells present at the onset of acute hepatitis B against only 4% of the ones present at recovery produced IFN-gamma after MHC-pentamer stimulation ( Figure 6 C ) . In addition to the higher frequency of IFN-gamma producing cells , the amount of the cytokine produced during the onset was higher than that during the resolution , as visualized by the difference in mean fluorescence intensity ( MFI ) ( 1152 at onset and 507 resolution ) . Figure 6 B shows the cumulative results obtained in 6 subjects with detectable HCMV ( P9 , P10 , P7 , P24 ) , EBV- ( P13 ) and influenza-specific CD8 ( P24 , P11 ) at the onset and recovery of acute hepatitis B . Bars indicate the % increase of IFN-gamma producing CD8 T cells at onset of acute hepatitis in comparison with recovery . Thus , persistent virus specific CD8 T cells produce more anti-viral cytokines after TCR-mediated activation during acute phase of heterologous viral infection .
We demonstrate here that activation of CD8 T cells specific for persistent viral infection ( HCMV-EBV ) is a constitutive feature of acute anti-viral immunity in human . Our conclusions differ from the ones obtained in attenuated virus recipients [13] , which have suggested that activated ( CD38/HLA-DR+ ) and proliferating ( Ki-67+ ) CD8 T cells are exclusively constituted of CD8 T cells specific for the acutely infecting virus . However , our patients with activated/proliferating HCMV and EBV responses had a symptomatic viral infection with a high level of inflammation , whereas those subjects vaccinated with attenuated viruses , by definition , should not exhibit any pathology of acute infection . Of note , the presence of activated HCMV and EBV specific CD8 was also detected during other pathological human viral infections [14] , [26] , [27] further supporting our conclusion that activation of CD8 T cells specific for persistent infection is a consistent phenomenon during symptomatic viral infections . In contrast to HCMV and EBV-specific CD8 cells , we observed that CD8 T cells specific for influenza were not activated during the acute phase of heterologous acute viral infection . Thus , our data show that memory CD8 cells specific for persistent and non-persistent viruses not only differs in term of phenotypic profile in healthy individuals [28] , but respond differently to the pathological condition triggered by an heterologous acute viral infection . We can only speculate about the causes of the variable behavior of CD8 T cells specific for the different pathogens . A plausible explanation is that , while influenza-specific CD8 T cells are true memory CD8 cells without any recent encounter to their specific ligand , EBV and HCMV specific CD8 cells might experience a continuous or repetitive exposure to the specific antigens . The accumulation over time of herpesvirus-specific CD8 T cells in healthy subjects [29] , [30] and work in animal model of HCMV infection [31] , have suggested that EBV and HCMV antigens are constantly available for T cell stimulation . The Ag-exposure might modulate the functional state of HCMV and EBV-specific CD8 cells and program them to respond to cytokines produced during acute viral infections . The differential functional state of herpesvirus specific CD8 T cells when compared with influenza-specific was confirmed by our in vitro data . We clearly demonstrate that IL-15 triggers in vitro the activation/proliferation of HCMV , EBV specific rather than influenza-specific CD8 T cells . Based on these in vitro data , we favor the idea that the detection of activated/proliferating HCMV and EBV specific CD8 T cell is mediated principally by the presence of IL-15 during acute phase of viral infections . This makes HCMV/EBV reactivation or indeed cross-reactivity a less likely explanation for this phenomenon . However , it is important to stress that this causative link is hypothetical since the level of IL-15 required to activate HCMV/EBV in vitro ( 1–10 ng/ml ) is higher than what we detected in the serum of the patients in this study ( always lower then 50 pg/ml in any viral infection , data not shown ) . Such inconsistency should be taken into account , even though the serum cytokine levels cannot define their actual concentrations in the target organ or lymph node . We cannot exclude that a reactivation of EBV/HCMV infection is occurring in our patient population and thus directly driving the HCMV or EBV-specific CD8 T cell activation . We couldn't demonstrate any virological evidence of HCMV and EBV reactivation , but the negative virological tests do not exclude a HCMV and/or EBV viral reactivation is present elsewhere outside the blood compartment and is immediately curtailed by activated HCMV and EBV specific CD8 T cells . A similar scenario was suggested to occur in patients with acute Hantavirus infection where an increased EBV-DNA titers were found only in subjects without measurable EBV-specific T cell response [26] . However , it has been reported that HCMV and EBV reactivation is associated with the expansion of HCMV/EBV specific CD8 T cells [10] , [17] , [18] , which was not observed in any of our patients ( Figure 3 B and Supplementary Figure S2 ) . What appears clear from our data is that the contribution of the activated/proliferating HCMV/EBV specific CD8 T to the size of activated total CD8 T cells is not negligible , but at the contrary can alter the quantitative measurement of anti-viral CD8 T response during acute viral infections . A mean of , respectively , 30% and 12 . 5% of EBV and HCMV-specific CD8 T cells express activation markers during the acute phase of different viral infections and since the combined population of both HCMV-EBV specific CD8 T cells might exceed 20% of total CD8 T cells [7] , [9] it is plausible to conclude that EBV/HCMV-specific CD8 T cells can inflate the number of total activated CD8 T cells . The presence of activated/proliferating CD8 T cells specific for HCMV and EBV during the early phases of different acute viral infection raises several questions . First , it will be interesting to evaluate whether CD8 T cells specific for other persistent viruses ( i . e . HSV1 and 2 ) can actually behave like HCMV or EBV specific CD8 T cells and thus further contribute to the anti-viral CD8 T cell acute response . A further question might address the biological significance of the herpesvirus specific CD8 T cell activation during heterologous acute viral infections . There is a possibility that the activation/proliferation state of HCMV/EBV specific CD8 T cells counteracts the potential attrition exerted by the expansion of CD8 T cells specific for the acutely infecting virus [32] and therefore might be important for preventing the reactivation of HCMV/EBV infection . In this regard , our data differ from previous reports in acute HBV infected patients [33] , since we did not observe any significant loss of HCMV or EBV specific CD8 T cells . On the contrary , HCMV and EBV-specific CD8 T cell frequency was remarkably constant during the different phases of acute heterologous viral infections and the observed mild proliferation of HCMV and EBV specific CD8 T cells ( Figure 3 A and Supplementary Figure S2 ) might represent a compensatory mechanism counteracting the attrition exerted by the expansion of CD8 T cell specific for the acutely infected virus [30] , [34] . In addition , the observation that activation of HCMV and EBV specific CD8 T cells present during the acute phase of heterologous viral infections is associated with a functional increase in the MHC-pentamer mediated CD8 T cell activation further supports the idea that such events might have a broader biological significance . We can only speculate about the physiological significance of the increased MHC-pentamer mediated CD8 T cell activation . However , a plausible interpretation is that the HCMV and EBV-specific CD8 cells during acute heterologous viral infection are less dependent to possible co-stimulatory effect mediated by additional molecules provided by their target during T cell recognition . Alternatively , the increased response to pentamer-mediated staining might indicate a lower requirement of MHC-class I complexes necessary for T cell activation [24] . These various possibilities will need further investigation , but what our data clearly demonstrate is that functional differences in the ability to produce IFN-gamma are present in different phases of heterologous acute viral infection . The increased likelihood of activated HCMV , EBV specific CD8 T cells to produce antiviral cytokines after recognition of HCMV and EBV antigens might be beneficial not only in the control of HCMV/EBV reactivation but can actively contribute to the global anti-viral immune response . Evidences in animal model have already shown that T cell activation of non-antigen specific T cells can contribute to the early response against pathogens [12] , [35] . On the other hand , the detected hyper-responsiveness of HCMV and EBV- specific CD8 T cells can have an impact on immunopathogenesis of the viral infections [36] . Heterologous immunity have been observed to alter pathogenesis of different viral diseases [37] , [38] . In conclusion , we show that the CD8 T cell population activated during acute viral infection is not constituted exclusively by CD8 T cells specific for the newly infected virus . On the contrary , this population is inflated by the presence of activated T cells specific for herpesvirus , directly demonstrating the ability of persistent virus infections to leave a functional imprint on the acute anti-viral T cell response in humans with functional consequences that will require further elucidation .
Samples were taken from patients or healthy volunteers attending clinics in Singapore ( Dengue , Influenza , Adenovirus infections , fevers of unidentified etiology and healthy volunteers ) and Italy ( HBV infection ) . Local Review board and Ethical Committees approved the study . Total number of patients is 50: HBV 20 , Influenza 12 , Dengue 12 , Adenovirus 3 , patients with fevers of unidentified etiology 3 . Number of healthy volunteers enrolled is 5 . Age of the subjects ranged from 20 to 54 years old . Patients were selected on the basis of fever >38°C ( Dengue , Influenza , Adenovirus ) or jaundice ( HBV ) . Diagnosis of dengue ( detection of dengue virus by PCR ) , influenza A ( + isolation of influenza A from nasal swab ) , adenovirus ( isolation of the virus from nasal swab ) , and HBV ( HBsAg + , anti-HBc IgM+ and HBV-DNA+ ) was performed within 5 days from selection . Acute hepatitis B patients were all HBsAg+ and had ALT>1000 U/L , at the disease onset . All healthy volunteers were asymptomatic . Peripheral blood mononuclear cell isolation from whole heparinized or EDTA blood with Ficoll-Hypaque was performed within 4 hours of drawing . PBMC were analyzed immediately or frozen for subsequent analysis . HBV patients: HBsAg , HBeAg , anti-HBs , anti-HBc IgG and IgM , anti-HBe , anti-HDV , anti-HCV , anti-HIV-1 and -2 were determined by commercial enzyme immunoassay kits ( Abbott Labs , IL , USA; Ortho Clinical Diagnostic , Johnson & Johnson , DiaSorin , Vercelli , Italy ) . HBV-DNA was quantified by PCR ( Cobas Amplicor test; Roche Diagnostic , Basel , CH ) and CMV-DNA was quantified with artus-CMV-LC PCR ( Qiagen , Qiagen Gmbh , Hilden ) , EBV-DNA was tested with EBV R-gene DNA extraction and quantification kit ( Argene , Varilhes , France ) . Dengue detection was performed by RNA isolation from serum samples using RNA extraction kit followed by reverse-transcription into cDNA ( Superscript III First Strand kit , Invitrogen , California , USA ) . The cDNA was PCR amplified for detection of the virus , for determination of serotype and for quantification of viral load as previously published [39] . The serum and nasal swab samples were tested for the presence of Influenza A and Adenovirus using RT-PCR ( Superscript III First Strand kit , Invitrogen , California , USA ) and direct immunofluorescence assay on nasal swabs ( Light Diagnostic Influenza A antibody FITC reagent , Millipore , Billerica , MA ) . Amino acid sequence alignment was done using BLAST from NCBI PubMed ( http://blast . ncbi . nlm . nih . gov/Blast . cgi ) . HLA-peptide pentameric complexes ( pentamers ) were purchased from Proimmune ( Oxford , UK ) . Anti-CD8 ( PE-Cy7 and APC-Cy7 ) , anti-CD3 ( perCP and perCP-Cy5 . 5 ) anti-CD38 ( APC ) , anti-HLA-DR ( Pe-Cy7 ) , anti-KI-67 ( FITC and PE ) , anti-Bcl-2 ( FITC ) , anti-IFN-γ ( FITC and APC ) , anti-IL2 ( FITC , PE and APC ) , and isotype control antibodies were purchased from BD Biosciences , San Jose , CA . Titrated pentamers ( PE ) were added to 50 µl of purified PBMC ( 2×106 cells total ) for 15 min at 25°C in the dark , washed and then panel of titrated antibodies for surface markers were added to pentamer stained or total PBMC . The cells were then fixed and permeabilized using Cytofix/Cytoperm solution ( BD Biosciences , San Jose , CA ) . After washing , intracellular staining was performed for intracellular markers ( Ki-67 , Bcl-2 ) . Cells were then washed 3 times , and fixed with 1% formaldehyde before acquisition on a FACS Canto flow cytometer . Compensation was checked regularly using FASC Diva software . Compensation controls were individually determined for each experimental setup . PBMC were stained with the relevant pentamers ( MHC/peptide pentamer stimulated ) , or left unstained ( unstimulated ) , washed and then incubated for 5 h with 10 µg/ml brefeldin A ( Sigma-Aldrich , St . Louis , MO ) . Following incubation , the unstimualted cells were stained with relevant pentamers and MHC/peptide pentamer-stimulated were left in PBS . Then cells were stained with anti-CD8 and anti-CD3 mAbs for 20 min at 4°C then fixed and permeabilized using Cytofix/Cytoperm solution . Finally , cells were stained with anti-IFN-γ and anti-IL-2 for 30 min on ice , washed , and fixed with 1% formaldehyde before acquisition on a FACS Canto flow cytometer . For analysis of anti-virus-specific CD8 T activation in vitro , freshly isolated PBMC or purified CD8+ T cells were incubated in vitro at 2×106/ml with or without cytokines ( IL-7 , IL-2 , IL-15 , IFN-γ , IFN- α , TNF- α , purchased from RnD Systems , Minneapolis , MN ) . The cells were collected at indicated time points , and the intracellular cytokine staining was performed as described above . This study was conducted according to the principles expressed in the Declaration of Helsinki . The study was approved by the Institutional Review Board of Singapore National Healthcare Group Ethical Domain and Azienda Ospedaliera Universitaria di Parma Ethical Committee hospitals . All patients provided written informed consent for the collection of samples and subsequent analysis . | The majority of humans are infected by herpesviruses , such as Epstein-Barr virus and Human Cytomegalovirus , which rarely cause severe pathology but heavily distort the human T cell repertoire . Up to 20% of cytotoxic T cells can be specific to Epstein-Barr and Cytomegalovirus . It is believed that all these herpesvirus specific T cells are needed to control the persistent infection . However , it has not been explored whether these T cells can contribute to the immune response to a new viral infection . To investigate this possibility , we analyzed the volume of activated virus-specific and total T cells in patients with acute hepatitis B , dengue , influenza and adenovirus infections . We observed that all acute viral infections trigger an expansion of activated T cell population , part of which is specific to infecting agent , and the other part to herpesviruses . Our study provides evidence that persistent herpesvirus infections alter the composition of the T cell population which is activated during new acute viral infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"infectious",
"diseases",
"immunology/immune",
"response",
"virology",
"infectious",
"diseases/viral",
"infections",
"immunology",
"immunology/immunity",
"to",
"infections",
"infectious",
"diseases/gastrointestinal",
"infections"
] | 2010 | Contribution of Herpesvirus Specific CD8 T Cells to Anti-Viral T Cell Response in Humans |
Fatty acid synthesis in plants occurs in plastids , and thus , export for subsequent acyl editing and lipid assembly in the cytosol and endoplasmatic reticulum is required . Yet , the transport mechanism for plastid fatty acids still remains enigmatic . We isolated FAX1 ( fatty acid export 1 ) , a novel protein , which inserts into the chloroplast inner envelope by α-helical membrane-spanning domains . Detailed phenotypic and ultrastructural analyses of FAX1 mutants in Arabidopsis thaliana showed that FAX1 function is crucial for biomass production , male fertility and synthesis of fatty acid-derived compounds such as lipids , ketone waxes , or pollen cell wall material . Determination of lipid , fatty acid , and wax contents by mass spectrometry revealed that endoplasmatic reticulum ( ER ) -derived lipids decreased when FAX1 was missing , but levels of several plastid-produced species increased . FAX1 over-expressing lines showed the opposite behavior , including a pronounced increase of triacyglycerol oils in flowers and leaves . Furthermore , the cuticular layer of stems from fax1 knockout lines was specifically reduced in C29 ketone wax compounds . Differential gene expression in FAX1 mutants as determined by DNA microarray analysis confirmed phenotypes and metabolic imbalances . Since in yeast FAX1 could complement for fatty acid transport , we concluded that FAX1 mediates fatty acid export from plastids . In vertebrates , FAX1 relatives are structurally related , mitochondrial membrane proteins of so-far unknown function . Therefore , this protein family might represent a powerful tool not only to increase lipid/biofuel production in plants but also to explore novel transport systems involved in vertebrate fatty acid and lipid metabolism .
Fatty acids ( FAs ) are building blocks for the majority of cellular lipids , which are essential throughout life of organisms . Besides their role as constituents of biological membranes , plant acyl-lipids are used for diverse functions at different destinations and tissues ( reviewed in [1] ) . For example , triacylglycerols ( TAGs ) in seeds of oilseed plants represent the major form of carbon and energy storage . Cuticular waxes at the surface of plants restrict loss of water and provide protection against pathogen attack . Furthermore , the formation of pollen cell walls is strictly dependent on delivery of modified FAs from tapetum cells in anthers ( reviewed in [2] ) . De novo FA synthesis in plants occurs in plastids ( for overview , see [1 , 3] ) . Growing alkyl chains in the plastid stroma are attached as acyl moieties to acyl carrier protein ( ACP ) , and in seed plants become available for lipid assembly mainly in the form of palmitoyl ( 16:0 ) - and oleoyl ( 18:1 ) -ACP . Part of these long-chain FAs will be integrated into lipids inside plastids ( prokaryotic pathway ) ; the majority , however , is exported to the endoplasmic reticulum ( ER ) for further elongation , acyl editing , and lipid synthesis ( eukaryotic pathway ) . Although it is generally agreed that free FAs are shuttled across plastid envelope membranes , the mode of export still remains enigmatic [4] since until now , no membrane-intrinsic transporter protein could be associated with a direct function in plastid FA export ( for overview , see [1 , 3] ) . On the one hand , a facilitated diffusion of free FAs through the lipid environment of membranes is suggested , which is supported by the recent finding that an acyl-ACP synthase in the cyanobacterium Synechocystis sp . PCC6803 is necessary and sufficient for FA transfer across membranes [5] . On the other hand , several ATP-binding cassette ( ABC ) transporter proteins for lipids , FAs , or acyl-coenzyme A ( CoA ) , and for import of FAs into peroxisomes [6] , as well as FA-transport systems from Escherichia coli , yeast , or mammals , provide evidence for an active mode of transport in plastids . Nevertheless , before transport , acyl-ACP thioesterases at the inner plastid envelope membrane catalyze the hydrolysation of fatty acyl-ACP to free FAs . After crossing both inner and outer plastid envelope membranes ( IE , OE ) , free FAs are re-activated to acyl-CoAs by long-chain acyl-CoA synthetases ( LACSs ) . As demonstrated for the protein LACS9 , these enzymes can attach to the cytosolic face of the plastid OE [7–9] . At the ER membrane , the ABC transporter ABCA9 has recently been described to be involved in FA-uptake , most likely in the form of acyl-CoA , thereby being important for TAG synthesis during seed filling [10] . Once arrived in the ER lumen , plastid-derived FAs are utilized for synthesis of specific lipid classes via the so-called eukaryotic pathway , where phosphatidic acid ( PA ) represents an important intermediate , phosphatidyl-choline ( PC ) is a major membrane phospholipid , and TAGs are the energy storage lipids produced . Subsequently , these eukaryotic lipids are distributed to various subcellular locations . For re-import of eukaryotic lipids into plastids , most likely in the form of ER-derived PA , an ABC transporter system ( TGD1 , 2 , 3 ) at the IE [3] and the PA-binding ß-barrel lipid transfer protein TGD4 in the OE [11] are required . In plastids , the diacylglycerol backbone from these eukaryotic precursors is used for synthesis of the galactolipids MGDG , DGDG ( monogalactosyl- , digalactosyl-diacylglycerol ) , and the sulfolipid SQDG ( sulfoquinovosyl-diacylglycerol ) . In addition , however , a prokaryotic-type pathway also produces MGDG , DGDG , SQDG , and the phospholipid phosphatidyl-glycerol ( PG ) directly from newly synthesized FAs and thus does not require previous FA-export from plastids ( for overview , see [1] ) . Here we describe FAX1 , a novel protein in the IE of plastids that belongs to the Tmemb_14 superfamily of membrane proteins with so-far unknown function . Functional studies in yeast as well as FAX1 mutant analysis in Arabidopsis thaliana clearly demonstrate that FAX1 mediates FA-export from plastids and thus , to our knowledge , represents the first membrane-intrinsic protein described to be involved in this process . In mammals , FAX1 relatives are structurally related mitochondrial membrane proteins , for which the biological task is not yet clear [12–14] . Thus , FAX1 not only is a missing link to explain the mode of plastid FA-export and to improve plant lipid/biofuel production but might also propel the understanding of Tmemb_14 protein performance in general .
The Arabidopsis protein At-FAX1 ( At3g57280 , for fatty acid export 1 ) was previously annotated as potential plastid-targeted and plant-specific solute transporter by proteomic and phylogenetic analysis [15 , 16] . Furthermore , we identified transcripts of At-FAX1 to be up-regulated upon induction of early leaf senescence [17] . To analyze protein function , we isolated the cDNA of FAX1 genes from Arabidopsis and pea ( Pisum sativum ) . For both proteins , chloroplast targeting peptides and four hydrophobic α-helices are predicted ( Fig . 1A ) . By the latter , plant FAX1 clearly groups to the so-called Tmemb_14 superfamily of proteins with so-far unknown function . The Tmemb_14 family is ubiquitous , with members in nearly all eukaryotes and some bacteria ( InterPro|UPF0136 ) . In Arabidopsis , four proteins ( FAX1–FAX4 ) are predicted to be targeted to plastids , while three ( FAX5–FAX7 ) most likely are directed to other , non-plastid membranes via the secretory pathway ( Fig . 1B ) . The plastid-intrinsic FAX1 is restricted to the chlorophyll-containing plant kingdom , with representatives in mono- and dicotyledons as well as in mosses and green algae ( compare InterPro|UPF0136 , [15] ) . Relatives of non-plastid predicted At-FAX proteins , however , can be found in eukaryotes such as mammals , insects , or yeast , and in some bacteria and cyanobacteria ( e . g . , Chlamydiae or Nostocales ) . For all Tmemb_14 proteins , four hydrophobic α-helical domains are predicted ( Fig . 1 ) . However , nuclear magnetic resonance ( NMR ) structure determination of the human Tmemb_14 proteins TMEM14A and TMEM14C [14] showed that only three of these helices are membrane-spanning . TMEM14A contains an amphiphilic N-terminal helix , presumably located at the lipid micelle-water interface , while for TMEM14C an amphiphilic helix that orients perpendicular to the lipid bilayer , is placed between the second and third membrane domain . Amino acid sequences of the plastid FAX1 and the non-plastid At-FAX6 nicely align to both TMEM14A and 14C ( Fig . 1C ) , but structural modeling revealed that the mature At-FAX1 and Ps-FAX1 are more similar to TMEM14C ( Fig . 2A ) . Here , three membrane-spanning and one amphiphilic helix are likely , while the additional N-terminal amino acids of FAX1 proteins might form another α-helical domain not present in TMEM14C . In contrast , the structure of At-FAX6 resembles that of TMEM14A with an N-terminal amphiphilic helix followed by three transmembrane domains ( Fig . 2B ) . With its membrane-spanning domains , FAX1 inserts into the inner envelope membrane ( IE ) of plastids as could be shown by in vivo GFP-targeting and immunoblot analysis . At-FAX1-GFP signals in Arabidopsis protoplasts , which can be detected as rings around chloroplasts ( Fig . 2C ) , point to an envelope localization . This could be confirmed and specified to IE by immunoblot analysis using sub-fractionated pea chloroplasts . In pea IE membranes as well as in Arabidopsis chloroplast envelopes , FAX1 signals appear as a band of about 25kDa ( Fig . 2D ) . In agreement , FAX1 peptides in proteomic analyses of plastid membranes were exclusively detected in IE preparations ( see [16] and references therein ) . To exclude ER localization , we further probed against ER-enriched Arabidopsis microsomal membranes ( see [20] ) , where no FAX1 signals could be detected ( Fig . 2D ) . To study the in vivo function of FAX1 , we analyzed loss-of-function and over-expressing mutant lines in Arabidopsis . We selected fax1–1 and fax1–2 with T-DNA insertions in the first intron and first exon of the FAX1 gene , respectively ( S1A Fig . ) . Reverse transcriptase-polymerase chain reaction ( RT-PCR ) analysis showed that both homozygous alleles represent knockouts for FAX1 ( S1B Fig . ) . To complement this loss-of-function , At-FAX1 cDNA under control of the 35S promoter was introduced into heterozygous fax1–2 plants . Subsequently , two lines homozygous for the fax1–2 allele ( Co#7 and Co#54 ) were selected for further analysis . To stable over-express FAX1 in wild-type plants , the 35S::FAX1 construct was introduced into Col-0 , and two independent lines named ox#2 and ox#4 were identified as FAX1 over-expressors . Quantitative real time RT-PCR revealed that FAX1 transcripts in line Co#7 are at wild-type levels , whereas line Co#54 contains about 12 times more FAX1 mRNA . Over-expression in ox#2 seedlings was mild ( about 2-fold ) , but strong in line ox#4 ( about 200 times more mRNA than in wild type; S1C Fig . ) . Immunoblot analysis confirmed the strength of FAX1 expression in these lines and the knockout in fax1–2 on the protein level ( S1D Fig . ) . Homozygous fax1–1 and fax1–2 knockout mutants both were characterized by reduced biomass at mature rosette stages ( Fig . 3A , Table 1 ) . Full flowering fax1 knockouts were significantly smaller than wild type , had thinner inflorescence stalks , and flowers producing short siliques that contained almost no seeds ( Fig . 3B , C ) . Detailed analysis of different tissues and organs revealed that the decrease in biomass of fax1 lines was detectable throughout the entire plant body , including root , leaf , and stem tissues ( Table 1 ) . Because differences in stem dry weight were slightly more pronounced than in fresh-weight ( FW ) samples , most likely cell wall synthesis was affected . This could be confirmed by ultrastructural analysis of stem tissue ( S2 Fig . ) . Here fax1 knockouts showed small vascular bundles with reduced secondary cell walls ( S2A , D , G Fig . ) . Since the same phenotype was detected in both independent T-DNA insertion lines fax1–1 and fax1–2 , and could be reverted by complementation in lines Co#7 and Co#54 ( Fig . 3 , Table 1 ) , we conclude that the reduced biomass is caused by the loss of FAX1 function . Remarkably , FAX1 over-expressing lines ox#2 and ox#4 were slightly bigger and produced more biomass as well as thicker inflorescence stalks than wild type ( Fig . 3 , Table 1 ) , thus behaving opposite to fax1 knockouts . In stems , this led to about one more hypodermal cell layer and to extended vascular strands , including an increased amount of xylem and phloem vessels , as well as a multi-layered procambuim ( S2C , F , I Fig . ) . Because fresh weight of ox#2 and ox#4 stems was significantly higher than in wild type , but—in contrast to fax1 knockouts—stem dry weight of FAX1ox lines was similar to wild type ( Table 1 ) , the increased biomass of FAX1 over-expressors is most likely mainly due to enhanced production of cells . However , since tracheid walls of ox#2 appeared to be slightly thicker than in Col-0 ( S2H , I Fig . ) , we can’t fully exclude an additional effect on the size of secondary cell walls . Interestingly , the rate of FAX1 overproduction—i . e . , 2-fold for ox#2 , 200-fold for ox#4—did not quantitatively affect the strength of biomass phenotypes , indicating a rather non-linear effect of protein function . To understand the peculiar loss-of-function phenotype of homozygous fax1 knockouts during flower and silique development , segregation analysis of mutant alleles was performed . Self-fertilization of heterozygous fax1–1 and fax1–2 revealed that the ratio of homozygous progeny was 7% and 4% , respectively , pointing to defect male and/or female gametophytes ( Table 2 ) . However , when stigmata from homozygous fax1–2 flowers were fertilized with wild-type fax1–2 pollen , normal seeds with 100% heterozygous fax1–2 mutant alleles were produced , indicating fertile fax1 knockout female gametophytes and sporophyte organs . In contrast , pollination of wild-type stigmata with homozygous fax1–2 anthers , produced short siliques , as observed during selfing of homozygous fax1–2 mutants ( see Fig . 3C ) , and led to an estimated seed yield <0 . 1% of wild type . Furthermore , during manual crossing it became evident that pollen grains of homozygous fax1–2 flowers were improperly released from anthers . To minimize potential anther defects from the paternal sporophyte , we thus pollinated homozygous fax1–2 stigmata with heterozygous fax1–2 anthers , thereby producing 12% progeny homozygous for fax1–2 ( Table 2 ) . In summary , these results point to impaired transmission of male gametophytes ( pollen ) and defects of the male sporophyte ( anther ) in fax1 knockouts , finally leading to the observed male sterility . To further analyze flower development , in particular that of male parts , we examined the morphology of flower tissue from FAX1 mutant lines ( Fig . 4; S3 Fig . ) . In FAX1 wild-type and over-expressors ( ox#2 , ox#4 ) , pollen release by anther dehiscence , transfer to the stigma , and fertilization as indicated by high yield of viable seeds was normal . However , flowers of fax1 knockout mutants showed stigmata with non-pollinated papillae . In addition , fax1 anthers released only very few pollen grains ( Fig . 4A , B; S3A , B , G Fig . ) . In flowers of complemented fax1–2 lines ( Co#7 , Co#54 ) in comparison , more free pollen grains than in fax1 knockouts but less than in wild type were visible , indicating incomplete recovery of pollen release ( S3D , G Fig . ) . In contrast to the rest of the plant organs , where regeneration of fax1 knockout defects in Co#7 , Co#54 lines was 100% ( see Fig . 3; Table 1 ) , complementation of fax1–2 pollen phenotypes was incomplete . This effect was best visualized by the colorless pollen of fax1 knockout and complementation lines ( S3B , D , G Fig . ) , due to the absence ( fax1–2 ) or incomplete ( Co#54 ) assembly of a pollen coat ( compare Fig . 4D–F; S3I–K Fig . ) that normally includes yellow flavonoid and carotenoid deposits ( for overview , see [2] ) . The incomplete complementation was restricted to pollen grains and most likely is due to the 35S promoter system , which in Arabidopsis shows no or reduced activity in pollen grains and anther tissue , respectively [21] . Subsequently , the detailed structure of anther tissue and pollen grains of fax1–2 knockout , Col-0 wild type , and the complementation line Co#54 was visualized by light- and transmission electron microscopy ( TEM ) at the mature , tricellular pollen stage ( Fig . 4C–F; S3H–K Fig . ) . In general , anthers of fax1–2 were smaller than in wild type and the surface of pollen grains appeared to be wrinkled ( Fig . 4C ) . Cross sections revealed an impaired release of fax1–2 pollen , although pollen sacks were wide open , indicating full dehiscence of anthers . Tapetum cells seemed to be degraded as expected for the developmental stage analyzed , however , the locule of fax1–2 anthers was covered by an electron-dense material , which stuck to pollen grains and thus most likely was responsible for improper pollen delivery ( Fig . 4C–E ) . Ultrastructural resolution demonstrated that the well-defined structures of the outer pollen cell wall—i . e . , the exine layer and the pollen coat , which covers the exine surface and its cavities—were absent in fax1–2 knockouts ( Fig . 4E , F ) . The intine , representing the innermost layer of the pollen cell wall and composed of cellulose , pectin , and various proteins , secreted by the microspore ( gametophytic origin , see [22] ) , however , looked intact . In contrast , mature wild-type pollen showed a complete exine , consisting of a flat nexine layer and the sculpted sexine parts tectum and bacula . Furthermore , the latter were filled and covered with the tryphine pollen coat ( Fig . 4E , F ) . Pollen cell walls of Co#54 displayed an intermediate state of biogenesis with visible nexine layers , but incomplete arrangement of tectum and bacula structures as well as pollen coat material ( S3J , K Fig . ) . As described above , these findings point to an incomplete complementation of fax1–2 knockouts in pollen . In conclusion , structural analysis of anthers and mature pollen grains showed that FAX1 is essential for biogenesis of the outer pollen cell wall , in particular for the assembly of exine and pollen coat . Both layers consist of complex biopolymers , such as sporopollenin ( exine ) and tryphine ( pollen coat ) , that are mainly made of FA-derivatives and lipids originating from the tapetum tissue of anthers ( sporophytic origin , see [2] ) . Thus , FAX1 might play a role in delivery of these compounds by mediating FA-export from tapetum cell plastids . Most likely , the electron-dense , sticky material in fax1 knockout anthers that prevents release of pollen grains represents cellular debris of degenerated tapetum cells and/or not-incorporated sporopollenin or tryphine material . Because during analysis of FAX1 mutants an altered surface of epidermal cells was apparent , we investigated structure as well as wax and cutin coverage of epidermis cells from primary inflorescence stalks of FAX1 mutants ( Fig . 5 ) . Microscopic analysis revealed that the width of epidermal cell walls in fax1–1 was strongly reduced when compared to wild type ( Fig . 5A , B ) . As for cell walls in xylem vessels ( S2 Fig . ) , a strong effect was only found for knockout and not for FAX1 over-expressing lines . However , an electron-dense cover at the extracellular side of the cell walls , most likely representing the cutin matrix of the cuticular layer , appeared to be more intense in ox#2 , but reduced in fax1–1 ( Fig . 5B , C ) . To examine the constitution of the cuticular layer , we therefore determined wax and cutin coverage from stem epidermis cells . Surprisingly , the total loads of different wax and cutin species were not altered for all lines analyzed ( fax1–1 , fax1–2 , Col-0 , WT2 , Co#7 , Co#54 , ox#2 , ox#4 , see S1A Data ) . Furthermore , no change in composition regarding aliphatic chain length or functional groups ( e . g . ketones , acids or aldehydes ) could be detected . The only significant difference we found was for C29-ketone wax components , which were reduced in both fax1 knockout lines by more than 50% when compared to stems from all other plants ( Fig . 5D ) . Since cutin contents were unchanged , the different strengths of the outer layer of epidermal cell walls observed by TEM most likely are due to stronger ( ox#2 ) and weaker ( fax1–1 ) crosslinking of the cutin matrix with cell walls . The wax composition of cuticular layers , however , is dependent on plastid FA-synthesis as well as excretion of modified FAs via the plasma membrane of epidermis cells ( for overview , see [1] ) . In parallel to the assembly of sporopollenin and tryphine material in pollen cell walls ( see above ) , FAX1 might thus be necessary for plastid FA-export , previous to synthesis and release of ketone wax components . Because fax1 knockouts revealed a lack of FA- and lipid-derived compounds in pollen as well as stem epidermis cells ( see above ) , we measured free FAs and polar lipid species such as phospho- , sulfo- , galacto-lipids , and triacylglycerols in leaves and flowers of mature FAX1 mutant plants ( S1 Table ) . To spotlight changes in FAX1 mutants compared to wild type , we determined relative values and summarized representatives of significantly different levels , as well as abundant species from each molecule class in the next two figures . For comparison to the overall FA/lipid composition of each tissue , we listed contents in mol% of all significantly different species in S2–S4 Tables , and further estimated the impact of changes in mol% of each molecule class in the next table . In leaves of fax1 knockout plants , levels of 30 FA and polar lipid species ( irrespective of TAGs ) were significantly different from wild type ( S2 Table ) . For free FAs , we observed an increase of plastid-produced FA 18:2 ( Fig . 6A , S2A Table ) and a decrease for FAs 20:0 , 24:0 , 26:0 ( Fig . 6C , S2C Table ) , which are elongated at the ER . Whereas aggregate levels of 34:x glycolipids ( MGDG , DGDG , SQDG ) were only slightly elevated ( Table 3 , S2A Table ) , the highly abundant MGDG 36:6 ( 11 . 7 mol% in wt ) with an ER-made diacylglycerol backbone was considerably less ( 2 . 7 mol% ) than in wild type ( Fig . 6C , S2C Table ) . The eukaryotic-type DGDG 36:6 , however , increased contributing about 0 . 7 mol% more to the overall lipid content ( Table 3 , S2C Table ) . Strong upward changes were observed for phosphatidyl-glycerol ( PG ) species ( 2 . 7- to 5-fold; Fig . 6A , S2A Table ) , leading to an entire gain of up to 3 . 2 mol% ( Table 3 ) of these mainly plastid-derived phospholipids . In contrast , the ER-produced phospholipids phosphatidyl-choline ( PC ) and -ethanolamine ( PE ) mostly decreased in fax1 knockout leaves ( Fig . 6C , S2C Table ) . Here the effect , in particular of highly abundant PC 34:3 , PC 36:6 ( 9 . 3 , 7 . 4 mol% in wt ) , was especially strong and is estimated to primarily contribute to a total reduction of PCs by 8 . 8 mol% ( Table 3 ) . Whereas the overall decrease of PE was about 0 . 5 mol% , phosphatidyl-inositol ( PI ) contents showed a pronounced upward fold change , which , however , only very slightly contributed to the overall lipid composition , and thus leaf PI might rather be involved in signaling ( Table 3 , S2C Table ) . In leaves of FAX1 over-expressing lines ( Fig . 6B , D; Table 3 ) , we found an opposite distribution of free FAs and lipids as in fax1 knockouts . Here , without counting TAGs , 28 molecule species were significantly different from wild type ( S2B , D Table ) . Contents of all differentially regulated and mainly plastid-derived FAs , 34:x glycolipids ( MGDG , DGDG , SQDG ) as well as PG 34:2 dropped ( Fig . 6B , S2B Table ) . A considerable impact on the total lipid content came from reduction of highly abundant molecules such as MGDG 34:5 , 34:6; DGDG 34:2 , 34:3; or SQDG 34:3 , all with levels higher than 2 mol% in wild type . In consequence , the estimated overall reduction was about 2 . 8 , 1 . 9 , and 0 . 9 mol% for 34:x MGDG , DGDG , and SQDG , respectively ( Table 3 ) . For lipids produced by the eukaryotic pathway at the ER , we found a mild decrease of MGDG 36:5 ( 0 . 4 mol% ) and only very minor changes ( 0 . 04–0 . 08 mol% ) for SQDG 36:4 , 36:5 , and PE 34:3 ( Table 3 , S2D Table ) . The effect on PC contents , however , again was quite strong ( total increase of about 3 . 0 mol% , see Table 3 ) , including elevated levels of the abundant PC 34:1 , 34:3 , 36:2 , and 36:3 ( Fig . 6D , S2D Table ) . When compared to leaves , flower tissue of fax1 ko and FAX1ox lines showed a similar differential pattern of free FAs and lipids , which are presumably mainly produced via the prokaryotic pathway ( Table 3 , S3A , B Table ) . Whereas in fax1 , FAs that after synthesis have to be exported from chloroplasts ( i . e . 16:0 , 18:0 , 18:1 ) increased when compared to wild type ( largest change for 16:0 = 0 . 24 mol% ) , the plastid internal FA 18:3 and the plastid external FA 24:0 decreased by about 0 . 1 mol% each ( S3A , C Table ) . In FAX1 ox flowers only a minor increase of FA 18:0 was detected ( S3B Table ) . As found in leaves , overall levels of 34:x MGDG , DGDG , and SQDG increased in knockouts but decreased in over-expressors ( Table 3 ) . The most prominent changes were for MGDG 34:6 ( increase of 0 . 6 mol% in fax1 , S3A Table ) and for MGDG 34:5 ( decrease of 0 . 3 mol% in FAX1ox , S3B Table ) . For several lipid species , which are assembled at the ER , however , patterns in flowers were different and more diverse than in leaves . These included a rise in 36:x MGDG levels ( dominated by +0 . 8 mol % of MGDG 36:6 ) in fax1 knockouts ( S3C Table ) ; an increase and a decrease of about 0 . 45 mol% PE in fax1 and FAX1ox , respectively ( Table 3 ) ; as well as a strong reduction of PC in FAX1ox flowers ( up to 5 . 6 . mol% , Table 3 ) . In fax1 knockout flowers in contrast , PC ( mostly PC 36:6 by-1 . 0 mol%; S3C Table ) and also PI species ( -0 . 26 mol% ) significantly dropped ( Table 3 ) . The most robust differential distribution in both mutant lines and tissues , however , was found for triacylglycerol oils ( Fig . 7; Table 3 ) . Here we determined significant changes for more than half of the molecules measured ( S4 Table ) . More important , however , was a massive decrease of high and low abundant TAGs in fax1 knockout leaves ( Fig . 7A , S4A Table ) and flowers ( Fig . 7C , S4C Table ) as well as a strong increase in FAX1ox leaves ( Fig . 7B , S4B Table ) and flowers ( Fig . 7D , S4D Table ) . Fold changes were highest for low abundant TAGs ( e . g . , 8 . 3-fold decrease for TAG 56:7 in fax1 leaves , S4A Table ) . As expected , the biggest impact on overall TAG content was by significant changes in high abundant species , resulting in a drop of-4 . 3 and-7 . 2 mol% in leaves and flowers of fax1 knockouts and an accumulation of +3 . 2 and +6 . 6 mol% for leaf and flower tissue from FAX1ox lines , respectively ( Table 3 ) . In summary , determination of free FAs and lipids in FAX1 mutants clearly shows that the function of FAX1 in the IE membrane of chloroplasts impacts cellular FA and lipid homeostasis . Overall we found significant differences compared to wild type for more than 50% of all species determined ( Table 3 ) . Together with the observed lack of FA- or lipid-derived compounds in fax1 knockout pollen cell walls and cuticular waxes ( see above ) , these findings point to a function of FAX1 in FA-export from plastids ( for details , see Discussion ) . In baker’s yeast ( Saccharomyces cerevisiae ) import of exogenous long-chain FAs by the so-called vectorial acylation process requires a multiprotein complex , which consists of Fat1p ( the membrane-spanning transport protein ) and Faa1p or Faa4p , acyl-CoA synthetases for intracellular FA activation [25] . In order to test a potential FA-transport function of FAX1 , we thus analyzed growth complementation of the yeast fat1 and faa1/faa4 mutants , which represent knockouts for Fat1p and Faa1p/Faa4p , respectively [26] . Therefore , we transformed the coding sequence of the mature At-FAX1 protein into fat1 and faa1/faa4 cells . Since previous studies revealed that the uptake of the polyunsaturated FA α-linolenic acid ( C18:3 ) into yeast cells was toxic for wild-type but not for fat1 cells [5] , we challenged growth of FAX1-containing yeast mutants by addition of high α-linolenic acid concentrations ( 3 . 6 mM ) to the media ( Fig . 8A–C ) . In drop tests on agar plates , all cells showed normal growth under control conditions ( Fig . 8A , left ) . In addition , yeast mutant cells , transformed with the empty vector only , were resistant to excess α-linolenic acid ( Fig . 8A , right ) . However , fat1 cells expressing the mature At-FAX1 protein died in the presence of α-linolenic acid overload ( Fig . 8A , right ) , indicating that FAX1 is able to restore FA-uptake in fat1 mutants . In contrast , α-linolenic acid induced cell death was not observed in faa1/faa4 cells , neither with nor without FAX1 , pointing to a FAX1 function in FA-transport and not in FA-activation . Furthermore , we monitored a very similar behavior for growth kinetics of the respective yeast cells in liquid media ( Fig . 8B , C ) . Here , a FAX1-mediated toxicity of α-linolenic acid was significant after 18 h when compared to empty vector cells . While this effect was highly significant and strong in fat1 mutants as indicated by a reduction of cell density to about 54% after 29 h ( Fig . 8B ) , only a mild growth inhibition was detected in Δfaa1/faa4 ( density of FAX1 cells was about 78% of control cells after 29 h; Fig . 8C ) . In addition , when compared to vector-only cells grown without α-linolenic acid , we observed a slight growth reduction by addition of α-linolenic acid itself as well as for FAX1-transformed cells in absence of α-linolenic acid , independent of yeast mutant strains ( Fig . 8B , C ) . Whereas the former observation can be explained by unspecific , background uptake of α-linolenic acid provided at excess external concentrations , the latter effect might be due to a general , but minor , toxic effect of FAX1 expression in yeast . To assess specificity of FAX1 for FAs , which have to be exported from chloroplasts in vivo , i . e . , palmitic ( C16:0 ) , stearic ( C18:0 ) , and oleic acid ( C18:1 ) , we performed additional yeast growth complementation assays in the presence of the FA-biosynthesis inhibitor cerulenin and supply of moderate external FA concentrations ( i . e . , 100 μM; Fig . 8D , S4 Fig . ) . Results with rapidly ( S4A–C Fig . ) and non-exponentially growing cells ( S4D Fig . ) allowed definition of a potential substrate specificity of FAX1 , preferring C16:0 over C18:1 and C18:0 FAs ( for details see S4 Fig . and Discussion ) . When we tested α-linolenic acid ( C18:3 ) , which in planta is not exported from plastids ( see [1] ) , as a control in this assay , FAX1 specificity was in the range as for stearic/oleic acid , but significantly lower than for palmitic acid ( Fig . 8D , S4B Fig . ) .
Complementation of the yeast fat1 , but not of the faa1/faa4 mutant , indicates that FAX1 is acting only in membrane transfer of FAs and not in FA-activation . This is in contrast to , for example , yeast and human FA-transporters such as Fat1p and FATPs , which in addition have acyl-activating functional domains [28 , 29] . FAX proteins group into the Tmemb_14 family and thus most likely contain three hydrophobic , membrane-spanning α-helical domains and one amphiphilic helix at the lipid bilayer/water interface . Thus , it is tempting to speculate that the latter might be responsible for binding and transfer of FA-chains across the IE membrane . Once loaded with a FA produced in the plastid stroma , this α-helix might become lipophilic enough to fold into the lipid bilayer and flip FAs over the IE . Furthermore , FAX1 and also FAX2 ( see Fig . 1B ) contain an extended N-terminal region ( gray helix in Fig . 2A ) . Structural modeling indicates that these stretches fold into additional , most likely non-membrane associated α-helices: one for FAX1 , two for FAX2 , respectively . Interestingly , the two anti-parallel helices of the FAX2 N-terminus fit to sequence and structure of a ‘four-helical up-and-down bundle’ of the human apolipoprotein apoE3 , which is involved in lipid transport and binding during formation of lipoprotein particles . Amphiphilic α-helices in the C-terminus of apoE3 are described to bind to lipids and thereby induce a conformational change in the N-terminal helix bundle that allows detergent-like solubilization of lipids and formation of lipoprotein particles ( for overview , see [30 , 31] ) . Therefore , the N-terminal helices of plastid FAX proteins might be involved in similar functions during FA-transport . The different apparent molecular weights observed for FAX1 ( S1D Fig . ) , most likely resulting from discrete conformations and/or packing of membrane domains , support these hypotheses for a transport mode . Once at the intermembrane space , FA-handover from FAX1 to substrate binding proteins , and transport across the OE membrane via a ß-barrel protein might be possible . For plastid re-import of eukaryotic lipids for example , the latter two proteins are represented by TGD2 ( substrate binding ) and TGD4 ( OE ß-barrel , [3 , 11] ) . Furthermore , in E . coli , a similar system has been described for export of lipopolysaccharides , including an ABC transporter that flips the lipid moiety across the inner membrane , transfer proteins in the periplasm , and a ß-barrel protein in the outer membrane [32] . For plastids , subsequently an acyl-CoA synthetase ( ACS ) at the cytosolic face of the OE might finally drive FA-transfer in a passive , carrier-like process . Co-expression of At-FAX1 with LACS4 ( ATTED-II coexpression networks ) , and regulation of LACS1 , 3 , and 5 expression in FAX1 mutants ( S5 Table , S6 Table ) underline a possible cooperation with ACS . The close structural similarity of FAX proteins to the human TMEM14A and TMEM14C , which both localize to mitochondrial membranes , in the future might enable explanation of TMEM14 protein function in vertebrates . Whereas TMEM14C was identified to coexpress with the core machinery of heme biosynthesis and its knockdown causes anemia in zebrafish [12] , TMEM14A was described to stabilize mitochondrial membrane potential and thereby inhibit apoptosis in a yeast system [13] . However , their exact biological function is still unknown . Since animal mitochondria are the site for FA-degradation via ß-oxidation , a role for TMEM14 proteins in FA/lipid homeostasis , energy metabolism or disease ( e . g . , apoptosis ) in vertebrates might be possible . Levels and subcellular distribution of free FAs and polar lipids in Arabidopsis FAX1 mutants mainly correlated with a FA-export function , by which FAX1 influences cellular FA and acyl lipid homeostasis ( for overview , see [1] ) . Most likely because of their toxicity and high metabolic fluxes for primary metabolites , changes in free FAs were not very pronounced . However , very-long–chain FAs ( C20 ) , which are elongated outside plastids and thus require previous export of C16–18 FAs , were significantly reduced in fax1 knockouts ( Table 3 ) . According to acyl-ACP synthesis rates and specificity of thioesterases in Arabidopsis chloroplasts , oleic acid ( C18:1 ) is the major free FA exported from chloroplasts , followed by palmitic acid ( C16:0 ) and only very little amounts of stearic acid ( C18:0; compare [1 , 33 , 34] ) . FAX1 in yeast assays performed best for FA 16:0 ( determined specificity range: 16:0 > 18:1 ~ 18:0 ~ 18:3 ) and thus , most likely , mainly is involved in the plastid export of free palmitic acid but also can transport oleic acid , which at the stromal side of the plastid IE is provided at highest substrate concentrations . The fact that in yeast , FAX1 was also able to transport α-linolenic acid ( C18:3 ) , which in planta is retained inside plastids , indicates that the protein does not discriminate between different degrees of unsaturation , but in general prefers C16 over C18 FAs . In chloroplast IE membranes , FAX1 most likely functions in a passive , carrier-like mode , driven by concentration gradients of free FA substrates ( see above ) . Interestingly , accumulation of export-directed C16–18 FAs in flower tissue of fax1 knockouts ( +0 . 24 > +0 . 04 > +0 . 01 mol% for 16:0 > 18:0 > 18:1 ) , reflect the substrate specificity range of FAX1 in yeast ( compare S3A Table and S4 Fig . ) . Furthermore , non-exported FA 18:3 significantly decreased ( 0 . 14 mol%; S3A Table ) in flowers of fax1 knockouts , thereby maybe pointing to stronger fluxes of FAs into plastid-intrinsic pathways ( e . g . , synthesis of oxilipin hormones ) , due to a block in FA export via FAX1 . Besides changes in free FAs , 65% of the differential lipid patterns depicted in Table 3 underline the hypothesis of plastid FA-export via FAX1 , best documented by the strong reciprocal changes in TAG oils . Here for almost 90% of all significantly distributed TAGs ( compare S4 Table ) , the pattern in both FAX1 mutants and tissues perfectly matched to a FA-export function of FAX1 . The distribution of 34:x glycolipids ( MGDG , DGDG , SQDG ) , which increased in fax1 knockouts but decreased in FAX1 over-expressors , also corresponded to our theory . In this case , we can , however , not exclude a contribution of ER-made species , since the diacylglycerol backbone for the “34”-glycolipids can originate both , from prokaryotic ( from plastids ) and eukaryotic ( from the ER ) phospholipid precursors , respectively . Yet , Arabidopsis is a so-called 16:3 plant , which for galactolipids prefers the prokaryotic pathway with high levels of 16:3 acyl chains . In contrast , ER-derived “34” DAG-backbones contain 16:0 saturated acyl moieties ( compare [1 , 34] ) . Thus , we can assume that MGDG 34:x and DGDG 34:x with more than four desaturated C-bonds are completely synthesized in plastids . For the strong MGDG 34:x reductions in FAX1ox leaves and flowers ( 2 . 8 and 0 . 5 mol% , Table 3 ) and the increase in fax1 flowers ( +0 . 7 mol% ) indeed 34:4 , 34:5 and 34:6 are the major contributing species ( see S2 Table , S3 Table ) and therefore support our hypothesis . Most likely at least the abundant forms of phosphatidyl-glycerol ( PG 34:3 , 34:4 ) are exclusively made inside plastids as well ( see [1] ) , and thus the pronounced overall increase of PG in fax1 leaf tissue ( +3 . 2 mol% ) also mainly is due to a block of FA-export via FAX1 . Our assumption that FAX1 mediates plastid FA-export is further confirmed by a large decrease of PC-levels in fax1 knockout tissues ( up to 8 . 8 mol% ) and a strong increase of PC in FAX1ox leaves ( +3 . 0 mol% ) . However , also , considerable contrasting evidence is found for three ER-made lipids in flower tissue ( i . e . , +0 . 9–1 . 0 mol% MGDG 36:x , DGDG 36:x in fax1; -5 . 6 mol% PC in FAX1ox ) . The latter findings that only apply to lipid species synthesized in the cytosol/ER of flowers might be explained by the inhomogeneity of mature flowers , consisting of leaf , stalk , pollen , ovary , and seed/silique tissues , and/or by a preferential flow of FA-building blocks for lipids into TAG oils during seed development . Furthermore , for the bilayer-forming DGDG 36:x , a plastid export is described to act as surrogate lipid for the lack of PC at , e . g . , phosphate-limited growth conditions ( see [1] and references therein ) . Thus , the observed increase of DGDG 36:x species in fax1 knockout mutants might compensate for the strong decrease of PC in the same tissues ( compare Table 3 ) . In summary , levels and subcellular distribution of free FAs and polar lipids in Arabidopsis FAX1 mutants mainly support a plastid FA-export function of FAX1 . In addition , we can , however , not exclude plastid FA-export via different mechanisms or a bypass by other plastid FAX proteins ( see below ) . Due to this potential functional redundancy of plastid FAX proteins , mutation of FAX1 alone does not affect all lipid species present in plants . Effects in leaf tissue , in particular of fax1 knockouts are somewhat more straightforward and stronger than in FAX1 over-expressing lines . The latter is not unexpected for mutation of a protein involved in transport , which is highly expressed in leaf tissues ( see below and S5 Fig . ) . However , the impact of FAX1 mutation on TAG-oil levels might be of future biotechnological importance . Interestingly , already the enhanced FA-transport by FAX1 was able to significantly increase TAG contents in leaves and flower tissues . Furthermore , our finding is in line with higher TAG when FA-loading to the ER in seeds is improved by over-expression of the ABC transporter ABCA9 [10] . Thus , coupling of the bottlenecks FA-transport ( e . g . , FAX1 , ABCA9 ) with those of FA-synthesis and acyl-editing processes and a seed-specific expression might boost plant oil production in future approaches . Transcripts of At-FAX1 are present in all developmental stages and peak in leaf tissues ( cotyledons , rosette , caulinary , senescent leaves , and flower sepals ) as well as in early pollen development ( S5A–C Fig . ) . Consequentially , the strongest phenotype of fax1 knockout mutants can be observed during growth ( e . g . , reduced rosette leaf size and biomass ) and in particular in pollen grains , leading to almost complete male sterility due to the absence of pollen cell walls and impaired pollen release by anthers . For FAX1 , we propose a function in FA-export from plastids of tapetum cells in anthers , which in fax1 knockouts leads to the strongly impaired assembly of exine layers and pollen coat , most likely because of the lack of FA-precursors for sporopollenin and/or tryphine synthesis ( for a detailed description , see S2 Text ) . Since FAX1 in Arabidopsis belongs to a family of seven proteins , the plastid-predicted FAX2 , 3 , and 4 , whose expression is regulated throughout plant development as well ( S5 Fig . ) , most likely can bypass the loss of FAX1 function in all tissues and organs , leading to the rather mild overall phenotype of fax1 knockouts . Especially in seed tissue ( S5D , E Fig . and S6 Fig . ) , FAX2 , 3 , and 4 most likely play a more prominent role than FAX1 . Indeed transcripts for FAX2 and FAX3—i . e . , the highest plastid FAX genes in seed development and germination—showed to be significantly up-regulated ( 1 . 13- and 1 . 24-fold ) in fax1 knockout flowers . Please note that with a relative signal of 1585 in wild type , FAX2 and FAX3 are strongly expressed in flower tissue we used for microarray analysis ( among highest 9% of all genes on the chip; compare microarray dataset E-MTAB-3090 at www . ebi . ac . uk/arrayexpress ) . The function of FAX1 in FA-delivery for pollen cell wall as well as for cuticular wax assembly is further underlined by differential gene expression in FAX1 mutants ( see S1 Text , S5 Table , S6 Table ) , and by the occurrence of phenotypes similar to fax1 when biosynthesis pathways for FA/lipid-derived precursor material are mutated in Arabidopsis . These include a plastid-intrinsic fatty-acyl-ACP reductase ( AlcFAR2/MS2 ) , involved in primary fatty alcohol synthesis for anther cuticle and pollen sporopollenin formation [35]; as well as cytochrome P450 enzymes ( CYP703A2 , CYP704B1; [36 , 37] ) that hydroxylate FAs , and the ACS ACOS5 [38] that activates FAs for sporopollenin synthesis in the cytosol of anther cells . Furthermore , several long-chain ACS ( LACS1 , 2 , 4 ) are necessary to activate long-chain and very-long–chain FAs for building of cutin and wax as well as pollen exine layers [39 , 40] . In addition , several ABC transporters in the plasma membrane are required for deposition of surface lipids , displaying fax1-like phenotypes upon mutation: ABCG26 for pollen exine formation from tapetum cells , as well as ABCG11 , G12 , and G13 in lipid export from epidermis cells for formation of cuticular wax layers ( for overview , see [41] ) . As for FAX1 , pathways for synthesis of precursors of pollen cell wall and cutin/wax components often overlap . In stems of fax1 knockouts , we further identified strong regulation of two genes involved in wax biosynthesis: AlcFAR3/CER4 and CYP96A15/MAH1 ( S6 Table ) . Because the latter enzyme is catalyzing synthesis of wax ketone components , its differential expression is in line with the observed lack of C29 ketones in fax1 knockout stems . Besides deranged acyl lipid homeostasis , obviously also carbohydrate and cellular/cell wall metabolism was affected in FAX1 mutants as reflected by the impact on plant biomass production and differential regulation of gene expression ( see S1 Text , S7–S10 Figs . ) . In general , we assume that these effects are rather secondary and most likely might result from still unknown signaling events , due to changed FA/lipid homeostasis . Since fax1 knockout plants are short of energy-rich lipids , they most likely turn down anabolic carbohydrate metabolism required for polysaccharide synthesis , resulting in , e . g . , reduced biomass and secondary cell walls . The opposite effect is observed in FAX1 over-expressors , in which an excess of lipids most likely leads to more biomass and the production of additional cell layers in stems . These observations clearly demonstrate regulation of energy metabolism and a close correlation between the availability of FAs/lipids and the utilization of carbohydrates in growth processes . This link is further underlined by the finding that the flow of carbon into oil can be promoted by activating plastid FA synthesis and repressing starch synthesis [42] . In this context FAX1 , to our knowledge , is not only the first membrane protein identified that mediates FA-export from plastids , but FAX1 and its relatives represent key transport proteins and thus—together with enzymes of FA/lipid-synthesis and modification—might provide powerful future tools to modulate plant lipid and bioenergy production [43] .
Experiments were performed on Arabidopsis thaliana ecotype Columbia ( Col-0 , Lehle Seeds; Round Rock , United States ) . The T-DNA insertion lines SAIL_66_B09 ( fax1–1 ) and GABI_599E01 ( fax1–2 ) were purchased from NASC ( Nottingham Arabidopsis Stock Center , Nottingham , United Kingdom ) and GABI-Kat ( MPI for Plant Breeding Research , Köln , Germany ) , respectively . To generate complementation lines of fax1–2 and over-expressing At-FAX1 under the control of the 35S promoter , the coding sequence of At-FAX1 was subcloned into pH2GW7 [44] . At-FAX1/pH2GW7 was transformed into Agrobacterium tumefaciens GV3101 , which was used to transfect heterozygous fax1–2 and Col-0 plants as described [45] . Arabidopsis seeds were sown on soil , vernalized at 4°C in the dark for two days , and grown in a 16 h light ( 22°C; 100 μmol photons ⋅ m–2 ⋅ s–1 ) and 8 h dark ( 18°C ) cycle . At-FAX1 cDNA was purchased as SSP pUNI51 clone U12755 [46] . The corresponding mRNA ( NCBI reference sequence NM_15588 ) is predicted to be 1 , 030 bp long , including 180 bp 5´- and 169 bp 3´-untranslated regions ( UTRs; S1A Fig . ) . For amplification of FAX1 from pea , RT-PCR was performed using pea seedling cDNA as template and oligonucleotide primers designed according to a pea EST contig sequence [47] . The corresponding mRNA molecule was 1 , 115 bp long , with 143 bp 5´UTR , 699 bp coding region , and 273 bp 3´UTR ( GenBank accession no . KF981436 ) . For primer sequences , see S7 Table; for amino acid sequences , see Fig . 1A . To generate a fusion of GFP to the preprotein At-FAX1 , the coding sequence was subcloned into the p2GWF7 plasmid [44] . p2GWF7 provides a fusion of GFP to the C-terminal end of the respective proteins , which are expressed under the control of the constitutive 35S promoter . Transformation and analysis of Arabidopsis mesophyll protoplasts was performed as described [45] . GFP fluorescence was detected at 672 to 750 nm and chlorophyll autofluorescence was monitored at 503 to 542 nm by confocal laser scanning microscopy ( Leica TCS SP5/DM 6000B , argon laser , excitation wavelength of 488 nm ) . Pea chloroplasts isolated from leaf tissue of 10-day-old pea plantlets were sub-fractioned into OE and IE membranes , stroma and thylakoids as described [48] . Chloroplast envelopes , total protein extracts , and microsomal membranes from Arabidopsis plants were prepared as specified in [45] and [20] , respectively . FAX1 antisera were raised in rabbit ( Pineda Antibody Service , Berlin , Germany ) against N-terminal peptide sequences of At-FAX1 ( 17 aa ) and Ps-FAX1 ( 18 aa ) , respectively ( see Fig . 1A ) . Antisera for marker proteins were produced as described previously [45 , 49] . Appropriate amounts of organellar or total cellular proteins were separated by SDS-PAGE , transferred to PVDF membranes and subjected to immunoblot analysis using primary antisera in 1:250 to 1:5000 dilutions in TTBS buffer ( 100 mM Tris-HCl pH 7 . 5 , 150 mM NaCl; 0 . 2% Tween-20; 0 . 1% BSA ) . Non-specific signals were blocked by 3% skim milk powder and 0 . 1% BSA . Secondary anti-rabbit IgG alkaline phosphatase antibodies ( Sigma-Aldrich ) were diluted 1:30 , 000 . Blots were stained using the alkaline phosphatase reaction with 0 . 3 mg/ml nitroblue tetrazolium ( NBT ) and 0 . 16 mg/ml bromochloroindolyl phosphate ( BCIP ) in 100 mM Tris pH 9 . 5 , 100 mM NaCl , 5 mM MgCl2 . Genomic DNA of the T-DNA insertion lines fax1–1 and fax1–2 was screened by PCR genotyping . To identify plants with T-DNA insertion in both At-FAX1 alleles ( homozygous ) , combinations of gene-specific primers that flank the predicted insertion sites with each other and with T-DNA-specific left border ( LB ) primers ( S7 Table ) were used . Positions and orientations of T-DNA inserts and oligonucleotide primers in fax1–1 and fax1–2 are shown in S1A Fig . To verify insertion sites , PCR-genotyping products were sequenced . T1 generations of generated FAX1 over-expression and complementaion lines were selected by hygromycin ( 30 μg/ml ) . Stable insertion of 35S::FAX1 was controlled by PCR-genotyping in all subsequent generations . Therefore , a vector-specific primer in combination with a FAX1 cDNA specific primer was used ( S7 Table ) . In the T2 generation , complementation lines were selected for homozygous alleles of the original T-DNA insertion in fax1–2 ( see above ) , resulting in lines Co#7 and Co#54 . For FAX1 over-expression in Col-0 background , we selected the lines ox#2 and ox#4 in the T2 generation . For microscopic analysis we used 5-week-old plants and dissected anthers from mature flowers or cut 1–2 mm2 stem segments 1 cm above the bottom of the second internode of the primary inflorescence stalk . We analyzed four individual fax1–2 knockouts , two of each Co#7 , Co#54 complementation lines , and five Col-0 wild-type plants for anthers/pollen grains , and pictured stem tissue of independent fax1–1 , fax1–2 knockouts , three ox#2 , four ox#4 over-expressors , as well as seven individual Col-0 wild-type plants , respectively . Tissue was fixed immediately after harvest with 2 . 5% ( w/v ) glutaraldehyde ( 4°C , at least 24 h ) in 75 mM cacodylate buffer ( 2 mM MgCl2 , pH 7 . 0 ) , rinsed several times with fixative buffer , and subsequently post-fixed with 1% ( w/v ) osmium tetroxide for at least 2 . 5 h in fixative buffer at 20°C . After two washing steps in distilled water , samples were stained with 1% ( w/v ) uranyl acetate in 20% acetone , dehydrated with a graded acetone series and embedded in Spurr’s low viscosity epoxy resin [50] . For light microscopy , semithin-sections ( 1–2 μm ) were cut with a glass knife ( Pyramitome 11800 , LKB ) . Ultrathin-sections ( 50–70 nm ) for transmission electron microscopy were prepared with an ultramicrotome ( EM UC6 , Leica ) and post-stained with aqueous lead citrate ( 100 mM , pH 13 . 0 ) . Micrographs were taken at 80 kV with a 268 electron microscope ( Fei Morgagni ) . The second to fourth internode region of primary inflorescence stalks from 7-week-old plants was used for wax and cutin analyses . For each replicate , stem segments from three to four individual plants were pooled , and samples were provided from two independent harvests of each FAX1 mutant line and respective wild-type controls . Determination of wax and cutin coverage of stems was essentially carried out as described previously [51 , 52] . Wax was extracted in chloroform and C24 alkane was added as internal standard . For cutin analysis , exhaustively extracted stems ( 1:1; methanol:chloroform ) were transesterified using methanolic HCl , and cutin monomers were extracted in hexane containing C32 alkane as internal standard . Gas chromatographic and mass spectrometric analysis was carried out after derivatization of extracted wax and cutin with pyridine and BSTFA . For each independent harvest ( 2-times for fax1 knockout , 4-times for FAX1 over-expressing lines ) cauline leaves and flowers ( stage 10–15 , according to [53] ) were sampled from at least ten individual , 7-week-old plants and grinded in liquid nitrogen . To be able to work on tissue of identical sample pools ( i . e . , from 7-week-old plants ) for wax/cutin analysis , FA/lipid determination , and transcript profiling , as well as because FAX1 is highly expressed in cauline leaves ( see S5A Fig . ) , we chose the latter instead of old rosette leaves . Tissue powder of each harvest was portioned into three aliquots of 50 mg , which were used to determine polar lipid and free FA contents . For details on data analysis , see S1 Table . Lipids/FAs were extracted from six ( fax1 k . o ) to 12 ( FAX1 over-expressors ) biological replicates using 1 ml of a pre-cooled ( −20°C ) methanol:methyl-tert-butyl-ether ( 1:3 ) mixture , spiked with 0 . 1 μg/ml PC 34:0 ( 17:0 , 17:0 ) as internal standard . The samples were incubated for 10 min at 4˚C , followed by another 10 min incubation in an ice-cooled ultrasonication bath . After adding 650 μl of UPLC grade water:methanol ( 3:1 ) , the homogenate was vortexed and centrifuged for 5 min in a table top centrifuge . The addition of water:methanol leads to a phase separation producing an upper organic phase , containing the lipids , and a lower phase containing the polar and semi-polar metabolites . The upper organic phase was removed , dried in a speed-vac concentrator , and re-suspended in 500 μl buffer B ( see below ) and transferred to a glass vial . 2 μl of this sample were injected onto a C8 reversed phase column ( 100 mm × 2 . 1 mm × 1 . 7 μm particles BEH C8 , Waters ) , using a Waters Acquity UPLC system . The two mobile phases were water ( UPLC MS grade , BioSolve ) with 1% 1 M NH4Ac , 0 . 1% acetic acid ( buffer A ) , and acetonitrile:isopropanol ( 7:3 , UPLC grade BioSolve ) containing 1% 1 M NH4Ac , 0 . 1% acetic acid ( buffer B ) . The gradient separation , which was performed at a flow rate of 400 μl/min , was 1 min 45% A , 3 min linear gradient from 45% A to 35% A , 8 min linear gradient from 25 to 11% A , 3 min linear gradient from 11% A to 1% A . After washing the column for 3 min with 1% A the buffer was set back to 45% A and the column was re-equilibrated for 4 min ( 22 min total run time ) . Mass spectra were acquired as described [23 , 24 , 54] using either an Orbitrap Exactive mass spectrometer ( Thermo-Fisher ) for fax1 knockout lines or a Waters Synapt G1 ( Waters ) for FAX1 over-expressors , and corresponding wild types , respectively . The spectra were recorded using altering full scan and all-ion fragmentation scan mode , covering a mass range from 100–1 , 500 m/z . The resolution was set to 10 , 000 with 10 scans per second . Spectra were recorded from min 0 to min 20 of the UPLC gradients . The analysis of the spectra ( alignment , peal picking , normalization and peak integration ) was performed with the software package CoMet 2 . 0 ( Nonlinear Dynamics ) according to the instructions of the vendor . For growth assays in yeast , the coding sequence of the mature At-FAX1 protein was subcloned into the yeast expression plasmid pDR195 ( XhoI/BamHI ) . Therefore , we fused the open reading frame of the predicted mature At-FAX1 , starting with aa 34 of the preprotein , behind an “ATG” base triplet by PCR amplification . The yeast mutant strains fat1 ( LS2020-YB332 ) and faa1/faa4 ( LS1849-YB525 ) are specified in [26] . Both strains were transformed with mature At-FAX1/pDR195 and the vector control pDR195 as described [45] . If not denoted elsewhere , liquid cultures of the respective yeast cells were grown to exponential phase in synthetic defined medium ( SD-ura ) , containing 0 . 1% ( w/v ) glucose , 0 . 7% ( w/v ) yeast nitrogen base without amino acids , and necessary auxotrophic amino acids without uracil . Subsequently , 2 μl drops of the cultures were spotted in different dilutions onto SD-ura plates ( 2% agar ) , supplemented with 3 . 6 mM α-linolenic acid ( 0 . 1% , w/v in ethanol ) , and 1% tergitol ( to increase α-linolenic acid solubility ) . For control plates , an equal amount of the solvent ethanol was added instead of α-linolenic acid . Assays in the presence of cerulenin were performed according to [26 , 27] in SD-ura media supplemented with 2% ( w/v ) glucose , 0 . 5% Brij 58 , 0 . 7% KH2PO4 , 10μM cerulenin and either 100μM of palmitic , stearic , oleic or α-linolenic acid . Growth of yeast cells on solid media was documented between 2 to 6 days at 30°C . OD600 measurements were performed in identical liquid media , inoculated to a starting OD600 of 0 . 05 or 0 . 06/0 . 03 for cerulenin experiments with the respective yeast cells . Cultures were continuously shaken at 30°C and the OD at 600nm was determined at indicated time points . Tissue from flowers ( stage 10–15 according to [53] , compare S3 Fig . ) and from the second to fourth internode of primary inflorescent stalks for each harvest was pooled from more than ten individual , 7-week-old plants ( identical sample pool for lipid analysis ) and used for preparation of RNA samples by the Plant RNeasy Extraction kit ( Qiagen ) . RNA ( 200 ng ) of four or five independently harvested samples ( n = 4–5 ) from wild type ( Col-0 and WT2 , segregated from heterozygous fax1–2 ) , fax1 knockout ( fax1–1 and fax1–2 lines ) as well as FAX1 over-expressors ( ox#2 and ox#4 lines ) was processed and hybridized to Affymetrix GeneChip Arabidopsis ATH1 Genome Arrays using the Affymetrix 3´-IVT Express and Hybridisation Wash and Stain kits ( Affymetrix , High Wycombe , UK ) according to the manufacturer’s instructions . Raw signal intensity values ( CEL files ) were computed from the scanned array images using the Affymetrix GeneChip Command Console 3 . 0 . For quality check and normalization , the raw intensity values were processed with Robin software [55] default settings as described [19] . Specifically , for background correction , the robust multiarray average normalization method [56] was performed across all arrays ( between-array method ) . Statistical analysis of differential gene expression of mutant versus wild-type samples was carried out using the linear model-based approach developed by [57] . In total , we analyzed the following comparisons ( see S7 Fig . ) : ( A ) flowers: fax1 knockout ( n = 5 ) versus wild type ( n = 5 ) ; ( B ) flowers: FAX1 over-expressors ( n = 8 , four times each ox#2 , ox#4 ) versus wild type ( n = 5 ) ; ( C ) stems: fax1 knockout ( n = 4 ) versus wild type ( n = 4 ) . The obtained p values were corrected for multiple testing using the nestedF procedure , applying a significance threshold of 0 . 05 in combination with the Benjamini and Hochberg false-discovery rate control [58] . All microarray data are available in the ArrayExpress database ( www . ebi . ac . uk/arrayexpress ) under accession number E-MTAB-3090 . Structural models of At-FAX1 and At-FAX6 were generated by Phyre2 [59] , based on alignments with the PDB entries for human TMEM14C ( c2losA ) and TMEM14A ( c2lopA ) , respectively . Identity of At-FAX1 with its template TMEM14C was 21% and for At-FAX6 with TMEM14A 36% , while confidence of both models was 99 . 9% , thereby indicating a high confidence and accuracy of the core models . Structural alignments were created with PyMOL [60] . | Fatty acid synthesis in plants occurs in chloroplasts—the organelle more commonly known for conducting photosynthesis . For subsequent lipid assembly to be possible in the endoplasmatic reticulum ( ER ) , export of these fatty acids across the chloroplast envelope membranes is required . The mechanism of this transport until now has not been known . We isolated FAX1 ( fatty acid export 1 ) , a novel membrane protein in chloroplast inner envelopes . FAX1 function is crucial for biomass production , male fertility , and the synthesis of fatty acid-derived compounds like lipids , waxes , or cell wall material of pollen grains . Whereas ER-derived lipids decreased when FAX1 was missing , levels of plastid-produced lipids increased . FAX1 over-expressing mutants showed the opposite behavior , including an increase of triacyglycerol oils . Because FAX1 could complement for fatty acid transport in yeast , we concluded that FAX1 mediates the export of free fatty acids from chloroplasts . In vertebrates , FAX1 relatives are structurally related proteins of so-far unknown function in mitochondria . This protein family may thus represent a powerful tool not only to increase lipid oil and biofuel production in plants but also to explore novel transport systems in animals . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | FAX1, a Novel Membrane Protein Mediating Plastid Fatty Acid Export |
Stress granules ( SGs ) are non-membranous cytoplasmic aggregates of mRNAs and related proteins , assembled in response to environmental stresses such as heat shock , hypoxia , endoplasmic reticulum ( ER ) stress , chemicals ( e . g . arsenite ) , and viral infections . SGs are hypothesized as a loci of mRNA triage and/or maintenance of proper translation capacity ratio to the pool of mRNAs . In brain ischemia , hippocampal CA3 neurons , which are resilient to ischemia , assemble SGs . In contrast , CA1 neurons , which are vulnerable to ischemia , do not assemble SGs . These results suggest a critical role SG plays in regards to cell fate decisions . Thus SG assembly along with its dynamics should determine the cell fate . However , the process that exactly determines the SG assembly dynamics is largely unknown . In this paper , analyses of experimental data and computer simulations were used to approach this problem . SGs were assembled as a result of applying arsenite to HeLa cells . The number of SGs increased after a short latent period , reached a maximum , then decreased during the application of arsenite . At the same time , the size of SGs grew larger and became localized at the perinuclear region . A minimal mathematical model was constructed , and stochastic simulations were run to test the modeling . Since SGs are discrete entities as there are only several tens of them in a cell , commonly used deterministic simulations could not be employed . The stochastic simulations replicated observed dynamics of SG assembly . In addition , these stochastic simulations predicted a gamma distribution relative to the size of SGs . This same distribution was also found in our experimental data suggesting the existence of multiple fusion steps in the SG assembly . Furthermore , we found that the initial steps in the SG assembly process and microtubules were critical to the dynamics . Thus our experiments and stochastic simulations presented a possible mechanism regulating SG assembly .
Cells suffer from various environmental stresses including heat shock , chemicals , hypoxia , starvation , osmotic shock , ultraviolet irradiation , and viral infections . Cells respond to these stresses resulting in either survival or apoptosis . Assembly of stress granules ( SGs ) , which are non-membranous cytoplasmic aggregates of mRNAs and related proteins with a size in the order of 0 . 1–2 μm , is one form of cellular response to a stress [1–5] . SGs are reported to contain RNA-binding proteins ( e . g . HuR ) , translation initiation factors ( e . g . eIF4E , eIF4G , eIF3 , and PABP-1 ) , 40S ribosomal subunit , self-oligomerizing proteins ( e . g . TIA-1 and G3BP ) , nuclear transport proteins ( e . g . importin α1 and importin 8 ) , and signaling proteins ( e . g . TRAF2 , RACK1 , and Raptor ) in addition to mRNAs [2 , 5–7] . The 60S ribosomal subunit , HSP90 and ARE-binding proteins hnRNPA1 and hnRNPD are excluded from SGs . Inclusion of the translation initiation factor eIF2α and heat shock protein HSP70 are reported to be cell-type and stress-type specific [8] . S1 Fig shows the translation initiation steps ( thin lines and narrow characters ) together with pathways related to SG assembly ( thick lines and bold characters ) . It has been reported that the SG assembly is usually initiated by the phosphorylation of eIF2α on Ser51 [1 , 8–10] . This phosphorylation inhibits translation initiation by reducing the level of eIF2 ∙ GTP ∙ tRNAMet ternary complex [1 , 11] . The observations led to the hypothesis that SGs act as sites for storing and/or sorting of untranslated mRNAs [1 , 4 , 12 , 13] . It has also been hypothesized that SGs maintain the proper ratio of translation capacity to the pool of mRNAs in response to environmental stress [14 , 15] . In fact , global translation repression is not required for the assembly of SGs [16 , 17] . In addition , exclusion of mRNAs encoding HSP70 and HSP90 from SGs [1 , 2 , 18 , 19] is consistent with these hypotheses , because this enables HSP70 and HSP90 proteins translated under a stress condition to act as chaperones regulating misfolded proteins outside SGs . Thus , the assembly of SGs offers a chance for a cell to decide its own fate . The roles of SGs are typically found in brain ischemia . The ischemic treatment of neurons located in the hippocampal CA3 region led to the phosphorylation of eIF2α resulting in the inhibition of protein synthesis [20 , 21] . Reperfusion with a normal oxygen content solution recovered the protein synthesis in these neurons . However , the recovery did not occur in pyramidal neurons within the hippocampal CA1 region [21–23] . In parallel , the SG assembly was observed in hippocampal CA3 neurons but not in CA1 neurons [20 , 21] . It is known that the CA1 region is more vulnerable to ischemia than the CA3 region [24 , 25] . The fact that no SG is assembled in hippocampal CA1 pyramidal neurons by ischemic treatment is a possible explanation for their vulnerability highlighting the role of SGs described above . In addition , SGs are emerging to play an important role in neurodegenerative disorders including ALS ( amyotrophic lateral sclerosis ) and FTLD ( frontotemporal lobar degeneration ) [20] . These observations support a view that SGs play an important role in the survival of neurons . TIA-1 plays a critical role in the induction of SG assembly ( S1 Fig ) . TIA-1 contains one prion-related domain ( PRD ) and three RNA recognition motifs ( RRM1 , RRM2 , and RRM3 ) [6 , 26] . PRD enables TIA-1 to self-aggregate , while RRM2 and RRM3 bind mRNAs with a different nucleic acid sequence specificity [1 , 6 , 12 , 18 , 27–30] . TIA-1 is required for SG assembly because TIA-1 dominant-negative mutants block SG assembly in response to stress . On the other hand , overexpression of TIA-1 was sufficient to induce SG assembly in the absence of a stress [8 , 29] . TIAR and G3BP are also known to promote aggregate formation in SG assembly [9 , 17 , 31] . Phosphorylation of eIF2α is catalyzed by kinases PKR , PERK , GCN2 , and HRI by the following stress indicators: PKR by heat shock , UV irradiation , and viral infection; PERK by ER stress; GCN2 by starvation; HRI by hypoxia [32–38] . In addition to the initiation of SG assembly by the phosphorylation of eIF2α , it also promotes polysomal disassembly resulting in the accumulation of untranslated mRNPs [33] . From these observations , it is postulated that self-aggregating proteins such as TIA-1 play a major role in aggregate formation , while eIF2α phosphorylation provides constituents of SGs , thus both acting synergistically in the SG assembly . In addition , it is also postulated that TIA-1 actively escorts untranslated mRNA to SGs [6] . SGs were assembled after a short latent period ( <10 min ) by the application of arsenite . The number of SGs reached a peak after approximately 20–30 min ( ~30 SGs ) , and then subsequently decreased [39] . The physical size of SGs was small early in the assembly , and they maturated into larger size later [37 , 40 , 41] . SGs distributed within the cytoplasm without preference to location at the beginning of the assembly , and were gradually confined to the perinuclear region later as a response to arsenite [39] . Microtubules were reported to be critical for SG assembly , because inhibition of their function abrogated SG assembly [41 , 42] . In addition , SGs were reported to move on microtubules by the dynein motor [37 , 42] . It is reasonable to assume that the dynamics of SG assembly should be closely related to its biological roles . The time course over which SG assembly occurs is important in determining how quickly the cell can respond to a stressful environment . The number and the size of SGs can be considered as measures of responsiveness of a cell to environmental stress . The distribution of SGs in a cell can indicate the location in a cell where the translational silencing occurs , and where the most effective loci of translational silencing and recovery take place at specific mRNAs . Although there are reports on the dynamics of SGs as shown above , mechanisms that regulate these dynamics are still largely unknown . We approached these problems by devising experiments and conducting computer simulations . SG simulation is not an extension of conventional deterministic simulation ( DS ) , which is based on differential equations including continuous variables , and concentrations . Since SGs are discrete countable structures , they should be expressed in terms of numbers instead of a concentration . As a result , a stochastic simulation ( SS ) was employed [43] . In our method of SS , molecules undergo a random walk ( RW ) changing their location , making chance collisions leading to a reaction described by a probability function Pr calculated from the classical binding reaction rate constant k . Thus our SS program controls coordinates and states of every single molecule . Simulation results not only showed good agreements with our observations of SG assembly , but also predicted a gamma distribution relative to SG size . The experimental data also yielded a gamma distribution . In addition , we clarified critical parameters determining the dynamics of SG assembly .
First we investigated the assembly of SGs in HeLa cells . To monitor the process of SG assembly in living cells , HeLa cells were transiently transfected with an expression plasmid for green fluorescent protein ( GFP ) -tagged TIA-1 , an SG-nucleating protein . 40 h after transfection , cells were stimulated with arsenite , and the time-lapse fluorescence images were acquired every 5 min for 55 min . SGs were clearly observed at 10 min ( green dots in Fig 1A ) . The size of SGs was small at this stage , but grew larger at a later time ( 50 min ) . In contrast , the number of SGs reached a maximum at 30 min , then decreased as a function of time . These dynamics to exemplify SG assembly are qualitatively shown in Fig 1B and 1C . A small number of SG assembly was observed at 5 min reaching a maximum ( 31 . 7 ± 2 . 7 , N = 9 ) at 30 min , and then decreased to 22 . 0 ± 2 . 8 at 55 min ( Fig 1B ) . During these observable changes in the number of SGs , the average size increased monotonically with time ( Fig 1C ) . In this quantification , the size of SGs were measured by the number of pixels ( Materials and Methods ) . S1 Movie demonstrates the overall dynamics clearly and was similar to previous reports [37 , 39–41] . To validate our SG size measurement , we quantified SGs differently by integrated fluorescent intensity aimed at SG size measurement by volume instead of diameter ( Materials and Methods ) . The time course of SG assembly and its evolution of average size were not changed significantly ( left and middle panels in S2A Fig ) . Next , the TIA-1 requirement for the assembly of SG was tested . TIA-1 possesses three RRMs at the N-terminus , along with a glutamine-rich PRD at the C-terminal region , both of which are essential for TIA-1-mediated SG assembly [9 , 29] . Previous studies reported that the expression of the C-terminal fragment of TIA-1 , which contains the PRD alone , dominantly suppresses SG assembly [9 , 29] . Therefore , we constructed an expression vector encoding GFP-tagged PRD ( GFP-PRD in S2B Fig ) , and examined if the expression of the PRD would affect the SG assembly . COS-7 cells were transiently transfected with either GFP alone ( as control ) or GFP-PRD , and incubated for 48 h . The cells were then treated with 0 . 5 mM arsenite for 50 min and the assembly of SGs was assessed by immunofluorescence microscopy ( Fig 2A and 2B ) . In control cells expressing GFP alone , approximately 90 ± 76% of the cells formed SGs in response to arsenite . In contrast , cells expressing GFP-RPD scarcely showed SG assembly ( 10 . 4 ± 2 . 4% ) ( Fig 2B ) . These results clearly indicate that TIA-1 was required for the assembly of SG upon application of arsenite in COS-7 cells as was reported previously [8 , 29] . We focused on the spatio-temporal dynamics of TIA-1-dependent SG assembly ( i . e . time courses of the number and the size , and the spatial distribution ) in a whole cell . We intended to construct a minimal model instead of a consolidative one that included all pathways shown in S1 Fig . Thus , in the model , SG assembly was initiated by and depended on the self-aggregation of TIA-1 ( Fig 3A ) because our experiments clearly showed the requirement of TIA-1 ( Fig 2 ) . Neither translocation of TIA-1 from the nucleus to the cytoplasm nor de novo synthesis of TIA-1 upon stress application was included in our model , because our experimental data excluded these possibilities ( S3 Fig ) . In addition , TIA-1 translocation was not seen in data from other laboratories [44 , 45] . The O-GlcNAc modification of proteins in the translational machinery [46] was also excluded for simplicity . The latency before the assembly of SGs was observed both in our ( Fig 1B ) and other laboratory’s experiments [39] . Latencies are often observed in biological phenomena . In signal transductions , latency emerges after multiple steps of activation cascade , which acts as rate limiting steps . Among them , the assembly of filamentous actin ( F-actin ) is one typical example [47–52] . The oligomerizations of globular actin ( G-actin ) , which are nucleation steps , are the rate-limiting steps in F-actin assembly . In light of this , we employed a model in which a single TIA-1 ( TIA1 ) formed a dimer ( TIA2 ) , and a trimer ( TIA3 ) ( 3-step model shown in Fig 3A ) . These were the rate-limiting steps in SG assembly . Then TIA3 bound with TIA1 assembling TIA* containing four TIA1 molecules . TIA* further bound with TIA1 , TIA2 , TIA3 , and TIA* resulting in the formation of a larger TIA*aggregate . These steps of SG enlargement are analogous to the elongation step of F-actin . SGs thus formed were assumed to be transported on microtubules . In the present model , TIA1 was a complex of TIA-1 , mRNA and its binding proteins for simplicity . In our SS , molecules undergo random walk ( RW ) by changing their location upon stochastic jumps in 3D space . During a jump , two TIA1 undergo a chance collision ( top panel of Fig 3B ) . This can lead to a binding reaction between two TIA1s . The probability of the binding reaction Pr is calculated by Eq 1 ( Materials and Methods ) , which is a function of the classical rate constant k , time between a jump τ , which is calculated from diffusion coefficient D , collision radius Rc , and calculation time step Δt . The important aspect of Pr is that its usage guarantees the convergence of SS to DS at an infinite number of molecules for any selection of τ , Rc , or Δt as long as Pr<1 [43] . Our SS guarantees this , and we used the same theory and algorithm in the present SSs . The mathematical bases of our SS method are found elsewhere [43 , 53] . In the SS of SG assembly , TIA1 , TIA2 , and TIA3 underwent RW in a 3D model cell with a diameter and thickness of 12 and 1 . 5 μm , respectively ( bottom panel of Fig 3B ) . A nucleus at the center with a diameter and thickness of 4 and 1 . 5 μm was located . SG movement along microtubules was reported [37 , 54] . We assumed that TIA* with 12 or larger number of TIA1s , which was tentatively defined as SG , underwent 1D radial RW simulating transportation on microtubules . Since hindered diffusion of large particles such as SGs was reported [2 , 55] , SG transportation on microtubules was expected to lead to an effective collision between large SGs resulting in a replication of experimentally observed SG dynamics , which are discussed in the following sections . Coordinates and state of every molecules including SG were stored in tables within the SS program , and were updated upon the occurrence of an event ( jump , collision , and reaction ) . The number of TIA1 molecules in a single TIA* ( and also SG ) was stored in the table of the SS program . The size of SG was calculated using this table ( Materials and Methods ) , and the total number of TIA1 in a 3D model cell , which is the summation of the number of TIA1s in all complexes including monomers , was monitored and kept constant . To test the validity of our SS in the 3D model cell shown above , we ran simulations of simple reactions . SS results for the binding reaction ( blue and red open circles in the left panel of Fig 3C ) agreed almost perfectly with those of DS ( blue and red lines ) . The initial number of molecules were 2724 for A and B , and 0 for C . SS results for dimerization reaction , which appeared in the present SG simulation , also agreed almost perfectly with those of DS ( right panel in Fig 3C ) . Insets are snapshots at the beginning and at the end of SSs . These tests of SSs showed clearly that our SS could be applied to the simulation of SG assembly . All SS parameter values in the SG simulation are summarized in S1 Table . Prs for each reaction were calculated by the SS program using Eq 1 with given parameter values of k , D , Rc , τ , and Δt before the start of SS ( Cf . Material and Methods ) . We ran SSs with an initial number of TIA1 equivalent to 9081 ( blue dots at 0 min in Fig 4 ) , which corresponded to a concentration of 100 nM in our 3D model cell . Each dot in Fig 4 indicates a single molecule with different colors for different molecular species . At 6 . 7 min , a small number of TIA2 ( yellow dots , one of which is indicated by a yellow arrowhead ) and TIA3 ( green dots , one of which is indicated by a green arrowhead ) were formed , and a small number of small SGs were also found ( red circles , one of which is indicated by a red arrow surrounded by white line ) . TIA* containing TIA1 smaller than 12 are indicated by small red dots ( one of which is indicated by a red arrowhead ) . At 16 . 7 min , both the number and the size of SGs increased . SGs gradually moved to perinuclear regions at later points in time ( 26 . 7 , 40 , and 60 min ) . These dynamics are clearly seen in the S2 Movie . Note that the number of TIA1 decreased with time as seen by the clearing up of blue dots from the background in the cytoplasmic space . If we plotted the number of SG as a function of time , SS results ( black line in Fig 5A ) agreed well with our observations ( white circles , which were replotted from Fig 1B ) . Gray areas indicate standard deviation ( SD ) at each time point for multiple SSs ( N = 10 ) . SS with smaller runs ( N = 5 ) gave almost the identical result ( S4 Fig ) . This suggests that experimental data with N = 9 is sufficient for analyses . The size of SGs increased monotonically as discovered in the experiment ( black line in Fig 5B ) . White circles were replotted from Fig 1C . Note that the time courses were sublinear both in the experiment and the SS . Next we compared the simulated special distribution of SG with that of experiments . SGs were localized around the nucleus at 50 min during the experiments ( left panel in Fig 5C ) . We found the same distribution in SSs at 50 min ( right panel in Fig 5C ) . There were negligible SGs seen at a distance from the nucleus . If this was compared to an earlier time , distributions from both experiments ( 35 min ) and simulations ( 35 min ) agreed qualitatively ( S5A and S5B Fig ) . Thus we replicated the dynamics of SG assembly qualitatively in our SSs . We tested a simpler model than that shown in Fig 3A , where there were two molecular species , TIA1 and TIA* ( 1-step model shown in S6A Fig ) . TIA* that contained 12 or larger number of TIA1 was defined as SG as in the model shown in Fig 3A . SS results showed that neither latency nor the peak in the time course of the number of SG was found . If we plot the time to 99% of the saturated number of assembled SG by varying kf1 and kf2 in the model shown in S6A Fig , it was far shorter than the experimental observation on the time to peak ( S6B Fig ) . Thus , this simple model could not replicate our experimental observations . We also tested a complex model by employing TIA4 in addition to TIA1-3 ( 4-step model shown in S6C Fig ) . This model also replicated the observed time course of number of SGs ( S6D Fig ) . Next we compared latencies from three SS models and our experiment . The latency was defined as the time to 10% of the maximum number of SGs , which was 5 . 6 min in our experiment ( S7A Fig ) . Latencies for 1- , 3- , and 4-step models were 8 . 3x10-3 , 3 . 8 , and 4 . 2 min , respectively ( S7B Fig ) . These results clearly showed the importance of multiple steps before the assembly of SG . We also tested a DS model , which was based on ‘winners-share-all’ dynamics ( S8A Fig ) . This mechanism was consistent with the observation that almost all SG resources ( TIA-1 ) were collected and shared into a small number of SGs during the course of their assembly [11] . Snapshots of DS results seemed to be consistent with observations ( upper panels of S8B Fig ) . However , there was no latency in the time course of the number of SG ( lower panel of S8B Fig ) . Furthermore , this type of simulation could not involve SG transportation on microtubules , and therefore , neither movement on microtubules of assembled SGs nor their fusion could occur . In fact , large SGs , shown by reddish dots , did not move at all ( S3 Movie ) , which was inconsistent with our experimental observation ( Cf . S1 Movie ) . Thus a DS model shown here was not representative of experimental observations of SG assembly . Next we investigated the dynamics of each molecular species in the model . First we analyzed the time course of complexes ( Fig 6A ) . As the increase in the number of SG , TIA1 decreased monotonically . TIA2 and TIA3 increased rapidly just after the start of the simulation , and then decreased monotonically . The peak number of TIA2 and TIA3 was 66 and only 9 , respectively in this SS . The small number of TIA3 was expected from Fig 4 . This indicates that TIA3 lifetime was relatively short , and quickly made a transition back to TIA2 or forward to TIA* . The number of SG after a peak decreased ( Fig 5A ) . In this decreasing phase , stepwise decrements in the number of SG in a single SS were observed in SSs ( top panel of Fig 6B ) . These decrements should be correlated to some event . We hypothesized that the fusion of SG occurred at time points of these decrements , and found that each decrement coincided exactly with the occurrence of single fusion events . In fact , two SGs at 3200 sec ( white arrowheads in the bottom left panel in Fig 6B ) fused into one larger SG at 3210 sec ( white arrowhead in the bottom right panel ) . Thus , in our simulation , the fusion of SGs was a major reason for the decrease in the number of SGs after attaining its peak . Next we tested the spatial distribution of SGs . In our simulations , the probability of antero- and retrograde transport were 0 . 4 ( pm ) and 0 . 6 ( pn ) , respectively , resulting in perinuclear localization of SGs at 60 min ( left panels of Fig 6C ) . It was expected that if these probabilities were changed , the distribution would also be changed . In fact , SG distributed dispersedly within the cytoplasm ( middle panels ) or near plasma membranes ( right panels ) for pm/pn of 0 . 5/0 . 5 and 0 . 6/0 . 4 , respectively . Thus the probability of antero- and retrograde transport of SGs on microtubules mainly determined the SG distribution in our simulation . We were interested in the SG size distribution , because it might provide us with a better perspective on the dynamics of SG assembly . Bars in the top panel of Fig 7A indicate SG size distribution at 55 min as a result of the simulation , which was fitted by a gamma distribution with a shape and scale parameters of 4 . 50 and 0 . 073 , respectively . The distribution in experiments at 55 min was also fitted by a gamma distribution with a shape and scale parameters of 2 . 54 and 0 . 093 , respectively ( lower panel of Fig 7A ) . There were significant differences between the simulation and the experiment . To explore the reason , we hypothesized that small SGs were not detected in our fluorescence measurement , and estimated that the number of GFP-TIA1 molecules in the smallest SG was 139 ( Materials and Methods ) . If we eliminated SGs smaller than this in our simulated data , the distribution was much similar to that by the experiment with a shape and scale parameters of 1 . 70 and 0 . 14 , respectively ( middle panel of Fig 7A ) . In addition , the shape and scale parameters were almost unchanged by the change in N from 5 to 100 ( S9 Fig ) , suggesting that the SG size distribution was fitted reasonably well with a gamma distribution in our simulation . These results suggest that small SGs were not detected in our fluorescence measurement . To further validate our SG size measurement , we draw SG size distribution using data by integrated fluorescent intensity ( Materials and Methods ) . The SG size was again approximated by a gamma distribution with a shape and scale parameters of 3 . 58 and 0 . 059 , respectively ( right panel in S2A Fig ) . The fact that the SG size was fitted by a gamma distribution both in our experiment and SS suggests that the SG assembly is an accumulation of successive random events , rather than ensembles of single random events . Next we analyzed the change in the dynamics of SG assembly by investigating the change in kinetic parameters using data shown in S10 and S11 Figs . First we investigated the sensitivity of number of SGs at the time of the peak and at 60 min . The peak number of SGs was sensitive to k1-1 , k1-2 , kb2 , and kb3 , but insensitive or only weakly sensitive to other parameters ( black bars in the top panel of Fig 7B ) . In contrast , the number at 60 min was only weakly sensitive to parameters tested ( gray bars in the top panel of Fig 7B ) . The time to peak was sensitive to k1-1 , k1-2 , k2-2 , kb2 , kb3 , and kb4 . But the latency was only weakly sensitive to parameters tested ( bottom panel of Fig 7B ) . Sensitive parameters for the number and the time are shown in solid red and blue boxes in the middle panel of Fig 7B , respectively . Overall , the dynamics of SG assembly was mainly sensitive to rate-limiting steps , and thus they determined the dynamics of SG assembly . Next we investigated the sensitive parameters for determining sublinearity seen in the evolution of the size of SG ( Fig 5B ) . We found that the sublinearity was sensitive to k1-4 and pm/pn ( S12 Fig ) . Larger k1-4 or pm/pn ratio increased the sublinearity . We investigated how microtubules contributed to the multiple aspects of SG assembly . To this purpose we deleted microtubules from our SS by removing 1D radial RW during the SG assembly . Thus SGs underwent 3D RW as other species ( TIA1-3 , and TIA* ) . The D was the same as that for 1D radial RW of SGs . The simulated deletion of microtubules dramatically changed the dynamics . SGs were distributed throughout the cytoplasmic space ( bottom panel of S13A Fig . The peak of the spatial distribution shifted from perinuclear to distant position ( S13B Fig ) , which was similar to that in the presence of SG transportation on microtubules with pm/pn of 0 . 5/0 . 5 ( middle panel of Fig 6C ) . There was no decrease in the number of SGs in the time course of their assembly ( S13C Fig ) . SG was small and distributed at bins of smaller sizes ( S13A and S13D Fig ) . In addition , the normalized SG size distribution was largely different from that in the presence of microtubules ( S13E Fig ) . While the distribution skewed positively in the presence of microtubules , it skewed negatively in their absence . These results strongly suggest that three experimental observations , the decrease in the number of SGs after the peak , the spatial distribution , and the size distribution were regulated by a single common mechanism of 1D radial RW of SGs on microtubules .
We intended to elucidate possible mechanisms for the dynamics of SG assembly in a whole cell both by experiments and mathematical models . To the best of our knowledge , this is the first occurrence to demonstrate SSs of SG assembly in a whole cell model . The SSs result replicated our experimental observations . In addition , the SSs result predicted a gamma distribution describing the SG size , which was also found in our experiments . One of important roles of computer simulations is to show a common mechanism for multiple phenomena . Our SSs replicated multiple aspects of our experimental observations including the decrease in the number of SGs after the peak , perinuclear spatial distribution , and the gamma distribution of SG size with a common mechanism of 1D RW of SG on microtubules . We employed 1D RW of SG in the radial direction . The assumption of radial structure of microtubules does not hold at perinuclear region . However , overall orientation can be assumed to be radial [56 , 57] . Thus we employed this simple assumption in our minimal model . Although the involvement of dynein in the retrograde transportation of SGs was suggested [41 , 42] , there is no firm evidence for the involvement of kinesin in the anterograde transportation of SGs . However , we observed anterograde movement of SGs in our experiments ( Cf . S1 Movie ) . In addition , mRNP was reported to move anterogradely on microtubules driven by kinesin motor protein [37] . Therefore , we hypothesized that SGs was transported anterogradely on microtubules in addition to retrograde movement . We have shown that the SG distribution was modified by probabilities of antero- and retrograde transportation ( pm and pn ) of SG ( Fig 6C ) . If motor proteins associated with a SG are modified posttranslationally by constituents of a SG in cell type- and stress-specific fashion , the probability of the movement direction would be modified too , and the SG distribution will be different by cell type- and stress-specific fashion . This might have a direct relationship to the biological roles of SG . The distribution of SG size was fitted with a gamma distribution both in SSs and experiments ( Fig 7A and S9 Fig ) . It is important to suggest that if the shape and/or scale parameter is different for different type of cells or stresses , existence of different kinetics and/or modifications in SG assembly mechanisms might exist . Thus analysis of SG size distribution will give us additional information for clarifying the mechanisms of SG assembly . A limitation to detect small SGs by fluorescence microscopy could result in a gamma distribution of SG size with different shape and scale parameters ( top and middle panels of Fig 7A ) . However , different quantification method of SG size by integrated fluorescent intensity also gave a gamma distribution ( right panel of S2A Fig ) , suggesting that SG size was intrinsically gamma distributed . Sublinearity in the SG size evolution in experiments was more pronounced than in SSs . ( Fig 5B ) . We have shown that the sublinearity was regulated by pm/pn ( S12 Fig ) . This suggests that the distribution of SG and the sublinearity in the SG size evolution might have a relation . This result was beyond our expectation . Further experiments and simulations are required to further reveal its mechanism and biological role . The difference in the latency between 3- and 4-step models was small , and the latency in a 4-step model was still shorter than the experiment . This may suggest more number of rate-limiting steps are required for replicating experimental observations . However , there is an another possibility of the existence of preprocess before TIA1 . In any case , the present study strongly suggested the existence of multiple rate-limiting steps before the assembly of SGs . In some of the simulation cases , the latency was so long that there was no observable decrease in the number of SG during 60 min of simulations , nor could the peak time be identified ( Cf . S14 Fig ) . If we carefully looked at the evolution of SG size , we found that there was a catastrophic decrease in the SG size before SG assembly in many long latency cases ( blue arrow in the bottom right panel of S15 Fig ) . This should be caused by a stochastic disassembly of SGs , and in extreme cases , no SG would be assembled . For example , such cases were observed for large values of kb2 and kb3 , where latency and the time to peak were not measured because of no SG assembly ( Cf . S11 Fig ) . In the present study , we could not compare the size of SGs between experiments and simulations in an absolute fashion , because we could not know experimentally how many TIA-1 existed in a SG . If this could be measured , models and kinetics would be greatly improved enabling a better replication of the experiments and proposing additional predictions to describe the dynamics of SG assembly . It is also important to know the minimum size of observable SGs and the number of TIA-1 molecules within it .
Full-length TIA-1 or a C-terminal fragment of TIA-1 ( TIA-1-PRD ) was subcloned into pEGFP vector by PCR . HeLa and COS-7 cells were maintained in Dulbecco’s modified Eagle’s medium ( DMEM ) supplemented with 10% fetal bovine serum ( FCS ) , L-glutamate , penicillin and streptomycin . For transient transfection assays , cells grown on a 35-mm-diameter glass bottom dish were transfected with the appropriate expression plasmids using Effectene transfection reagent ( QIAGEN ) according to the manufacturer’s protocol . 36 h after transfection , culture medium was exchanged with fresh DMEM/10%FCS without phenol red . The cells were incubated for another 4 h and then stimulated with arsenite ( 0 . 5 mM ) . Fluorescence images of the living cells were captured using a TE-2000E inverted microscope ( Nikon , Japan ) , equipped with a Planfluor 40× objective ( numerical aperture , 0 . 6 ) , a CoolSNAP HQ charge coupled device ( CCD ) camera ( Photometrics ) , and a xenon lamp . For GFP imaging , a B-2A filter set [a dichroic mirror ( 500 nm , a long-pass ) , an excitation filter ( 450–490 nm ) , an emission filter ( 515 nm cut-on ) ; Nikon , Japan] was used . Image binning was set to 2×2 . Fluorescence images were acquired every 5min from the same field . Fluorescence microscopic images of fixed cells were captured using an inverted Olympus IX81 microscope equipped with a QImaging Retiga EXi digital camera ( IEEE1394 ) and the Universal Imaging Metamorph software ( Molecular Devices ) . The assembly of SGs was determined from the fluorescence images using the Metamorph software package . First , the region of the entire cell area and the cytoplasmic area were manually located and the position of the region of the interest ( ROI ) was set in each fluorescence image in a time-series stack . Next , a median filter ( 25×25 ) was applied to the raw images and the resulting images were used as the background images . Images of SGs were generated by background image subtraction , by which the background images were appropriately subtracted pixel-by-pixel from the raw images . The number and the size of the SGs were quantified using transfluor assay module of the Metamorph software package and the pit-detection algorithm with the parameter describing approximate minimum width of 4 pixels ( ~1 . 3 μm ) was applied to the SG images . We then measured the area size of individual SG , the number of SGs in the cytoplasmic region of the cell , and the x and y coordinates of the each SG dot relative to the nuclear centroid of the cell . In addition , we applied different quantification method , in which a fluorescence spot was surrounded by a circle ( ROI ) using Find Spots command of Metamorph , and ROI higher than the threshold level of fluorescent intensity was defined as SG . Then , the spatial integration of fluorescent intensity within a ROI was calculated aimed at measuring the volume of a SG instead of its diameter . The false-positive dots were carefully checked by looking and manually eliminated from the measurement . Cells grown on glass coverslips were transfected with appropriate plasmids as indicated . Cellular SG assembly was then stimulated with 0 . 5 mM arsenite for 50 min . The cells were then fixed with 1% paraformaldehyde in PBS for 10 min . After washing with PBS , the cells were permeabilized with 0 . 1% Triton X-100 for 5 min , and incubated in the blocking solution BlockAce ( Snow Brand Milk Products ) for 1 h . Cells were then incubated with anti-eIF4E ( Santa Cruz ) antibody for 50 min in PBS containing 2% BSA , washed three times with PBS , and incubated with an Alexa Fluor 546-conjugated rabbit anti-mouse antibody for 30 min . The stained cells on coverslips were washed three times with PBS and were mounted in FluorSave Reagent ( Calbiochem ) . SG assembly was assessed by determination of the number of cells expressing at least two SGs per cell . We assumed that a SG began to assemble by aggregating TIA1 upon stress application . TIA1 was assumed to contain TIA-1 , mRNA , and related proteins for simplification . TIA1 forms dimer ( TIA2 ) , trimer ( TIA3 ) , and larger complex ( TIA* ) . TIA* grew further by binding TIA1 , TIA2 , TIA3 , and TIA* . Reaction probability Pr in SSs was calculated by the following equation [43]: Pr=k/{NA ( 4/3 ) πRc3 ( 2−e−Δt/τ1−e−Δt/τ2 ) /Δt} . ( 1 ) k , NA , Rc , and Δt are binding reaction rate constant between molecule 1 and 2 , Avogadro’s number , collision radius , and calculation time step , respectively . τ1 and τ2 are waiting times before the jump for molecule 1 and 2 , which were calculated by using τ = λ2/6/D . λ and D were jump lengths and diffusion coefficient . The convergence of our SS to DS is guaranteed as long as Pr<1 as shown in the main text , and a theory and analyses for this convergence are found in our previous paper [43] . Rc was used for testing an occurrence of a collision , and Pr was used for decision of actual occurrence of a reaction by the collision . Thus , Rc was not the direct parameter for reactions . In this sense , Rc was not an important parameter in our SS . However , it should be noted that the spatial accuracy of the occurrence of a reaction worsens by the increase in Rc . In contrast , small Rc gives a better spatial accuracy , but Pr would be larger than 1 . This leads to a strategy for the selection of Rc , in which small Rc is selected as long as Pr<1 . Rcs between TIA1 , TIA2 , and TIA3 were 20 nm according to this consideration . The same Rc was employed for the reaction between TIA* and other species ( TIA1 , TIA2 , and TIA3 ) . Rc between two TIA*s was calculated by the summed radii of two TIA*s . Each radius of a TIA* was calculated by the following equation: RTIA*=R0⋅nTIA11/3 , ( 2 ) where R0 is 10 nm and nTIA1 is the number of TIA1 in a single TIA* . The Pr between two TIA*s was assumed to be 1 . TIA* contained 12 or a higher number of TIA1 moved on microtubule by 1D RW . Other species underwent diffusion in a cytoplasmic 3D space . Simulations were run on computers consisting of an Intel Core i7-4770 processor ( 3 . 4GHz ) with Windows 8 OS ( 64 bit ) . SS program was written in C language . Parallelization by Open MP or MPI was not applied in the present SSs . The computational time was about 24 h for a 60 min simulation of SG assembly with an initial number of about 15 , 000 molecules . SS program running on Windows PC can be found in “doi . org/10 . 6084/m9 . figshare . 1295250” on the web . DS model was constructed by A-Cell software [58 , 59] . The model description files can be downloaded from http://www . ims . u-tokyo . ac . jp/mathcancer/A-Cell/A-CellModels/index . html . SS yields a different result for every run because of the use of random variables in the simulations . So we averaged SS results from 5 or 10 runs and calculated SD using Origin Pro 8 . 5 . 0 J SR1 from Origin Lab Corporation . Fitting of SG size distribution with gamma distribution was performed by using R with fitdsitr ( ) function of MASS package . Sensitivity Se was calculated using the following equation: Se=∂PCPC∂PSPS , ( 3 ) where PC and PS are characterizing and simulation parameters , respectively . Sensitivities of a single parameter were calculated at each value in the given range of a parameter shown in S10 and S11 Figs , and they were averaged giving an average sensitivity in the range , which was used in Fig 7B . We defined a characterizing parameter as sensitive , when Se was larger than 25% of the absolute maximum value of sensitivity for the number or the time in SG assembly . We estimated the minimum size of a SG to be detected in our fluorescence measurements . It was reported that ~200 nM EGFP fluorescence signal could be detected above typical cellular autofluorescence [60] . We employed a 1 . 3 μm square window for detecting SGs as shown in the previous section . If spherical shape of a SG was assumed within the window , the number of GFP-TIA1 molecules in the smallest SG was estimated using the following equation: nTIA1=Cf⋅43πr3⋅NA , ( 4 ) where Cf , r , and NA are 200 nM , 0 . 65 μm , and Avogadro’s number , respectively . These yielded nTIA1 of 139 . | Cells suffer from various environmental stresses such as heat shock and viral infection . In response to a stress , small non-membranous cytoplasmic aggregates , stress granules ( SGs ) , are assembled . SGs contain mRNAs and related proteins . Hippocampal CA1 neurons located in the brain , which are vulnerable to ischemia , do not assemble SGs , while CA3 neurons , which are resilient to ischemia , assemble SGs . The dysfunction of SGs has been reported in human diseases including pathogenic viral infection . These observations led to a hypothesis that SGs play an important role in cell fate decisions , and the dynamics of SG assembly would regulate cell fate . However , the conditions that determine the number and distribution of SGs in a cell in response to a stress are largely unknown . We approached this problem by experiments and simulations . Our stochastic simulations replicated the observations . Furthermore , we found that initial steps in the SG assembly process were important to the dynamics of SG assembly , and that SG size resembled the gamma distribution both in simulations and experiments , suggesting the existence of multiple steps in the SG assembly process . To the best of our knowledge , this work was the first to show SG assembly in a whole cell by stochastic simulations . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Spatio-temporal Dynamics and Mechanisms of Stress Granule Assembly |
Multiple syndromes share congenital heart and craniofacial muscle defects , indicating there is an intimate relationship between the adjacent cardiac and pharyngeal muscle ( PM ) progenitor fields . However , mechanisms that direct antagonistic lineage decisions of the cardiac and PM progenitors within the anterior mesoderm of vertebrates are not understood . Here , we identify that retinoic acid ( RA ) signaling directly promotes the expression of the transcription factor Nr2f1a within the anterior lateral plate mesoderm . Using zebrafish nr2f1a and nr2f2 mutants , we find that Nr2f1a and Nr2f2 have redundant requirements restricting ventricular cardiomyocyte ( CM ) number and promoting development of the posterior PMs . Cre-mediated genetic lineage tracing in nr2f1a; nr2f2 double mutants reveals that tcf21+ progenitor cells , which can give rise to ventricular CMs and PM , more frequently become ventricular CMs potentially at the expense of posterior PMs in nr2f1a; nr2f2 mutants . Our studies reveal insights into the molecular etiology that may underlie developmental syndromes that share heart , neck and facial defects as well as the phenotypic variability of congenital heart defects associated with NR2F mutations in humans .
During organogenesis , the initial specification of organ fields generates overlapping populations of progenitor cells that harbor the potential to contribute to multiple organs [1 , 2] . In vertebrates , the anterior lateral plate mesoderm ( ALPM ) , which generates the cardiac progenitor field , develops adjacent to the cranial paraxial mesoderm , which generates the pharyngeal muscle ( PM ) progenitor field , the source of facial and neck muscles [3–5] . In mice , detailed retrospective clonal lineage-tracing has revealed there are rare bi-potent cardio-PM progenitors , which potentially lie at the interface of these progenitor fields and give rise to the heart , pharyngeal , and neck muscles [6–8] . Specifically , craniofacial muscles of the 1st and 2nd pharyngeal arches share progenitors with the right ventricle and outflow tract , respectively [6 , 7] , which are derivatives of the later differentiating second heart field ( SHF ) [9 , 10] . However , muscles of the neck share progenitors from a distinct later-differentiating SHF population that contributes to the pulmonary arterial pole and atria [8] . Thus , these studies have emphasized the integration of developmental potential that generates multiple cardiac and PM progenitor populations during vertebrate development . Given the proximity of the cardiac and PM progenitor fields within the anterior mesoderm of vertebrates , there is significant overlap in the expression of conserved regulators of these lineages . The transcription factors Tbx1 and Tcf21 , in particular , share expression in cardiac and PM progenitors and are required to promote their development [11–14] . In humans , heterozygosity of TBX1 underlies DiGeorge Syndrome , which is characterized by congenital outflow tract and craniofacial defects [15] . Furthermore , studies using knockout ( KO ) mice have demonstrated that Tbx1 is at the top of a complex genetic hierarchy that directs the development of the outflow tract and all PMs [11 , 12] . Within this genetic hierarchy , Tcf21 appears to act downstream of Tbx1 . Compared to Tbx1 , loss of Tcf21 in mice results in less severe outflow tract and PM defects [12] , which is likely due to redundancy with Musculin/MyoR [16] . As in mammals , zebrafish tbx1 mutants have outflow tract and craniofacial defects [14 , 17 , 18] . Furthermore , in zebrafish , tcf21+ progenitors contribute to both ventricular cardiomyocytes ( CMs ) and PMs [5] . However , in contrast to mice , tcf21 in zebrafish is required for the development of almost all PMs [5] . Thus , a conserved network of core transcription factors promotes the development of both cardiac outflow tract and PMs in vertebrates . There is evidence that the origin of bi-potent SHF cardiac and PM progenitors is conserved in chordates [19] . Work in the tunicate Ciona has shed some light on transcriptional signals that drive cardiac and PM fate decisions within distinct precursors of the SHF [20] . Despite the conservation of core factors , including Tbx1 and Nkx homologs , there is currently limited understanding of signals that allocate the cardiac and PM lineages through driving differential fate decisions of progenitors from these adjacent organ fields in vertebrates . Retinoic acid ( RA ) signaling is currently the only known signaling pathway that overtly restricts cardiac specification and promotes craniofacial development in vertebrates [21–26] . However , the mechanisms by which RA signaling may coordinate cardiomyocyte ( CM ) and PM fate decisions from these progenitor fields within the anterior mesoderm are not understood . NR2F proteins ( formerly called COUP-TFs ) are highly conserved orphan nuclear receptor transcription factors whose expression is RA-responsive in many tissues of all vertebrates [27–30] . In mammals , the expression of two NR2F genes , NR2F1 and NR2F2 , overlaps during early embryonic development as well as later in atrial CMs of the heart [29–32] . Despite some overlap in limited cell types , expression of these two genes in mice largely diverges after early stages of embryogenesis , with Nr2f1 and Nr2f2 becoming predominantly expressed in neural and mesendodermal tissues , respectively [27 , 29] . Analysis of individual KO mice has revealed requirements in organs that are consistent with their tissue-specific expression patterns [33–36] . With respect to the heart , global Nr2f2 knockout ( KO ) mice have morphologically smaller atria and sinus venosus [35] . Conditional cardiac-specific Nr2f2 KO mice studies using a Myh6:Cre suggest a later role for Nr2f2 in maintaining atrial CM identity [36] . While zebrafish nr2f2 mutants are not early embryonic lethal and do not have overt cardiovascular defects through at least two weeks of development [37 , 38] , our recent analysis of zebrafish nr2f1a mutants indicates that it is the functional homolog of Nr2f2 in mammals with respect to early heart development [39] . Specifically , zebrafish nr2f1a mutants have smaller atria due to a requirement within atrial CMs to concomitantly promote atrial differentiation and limit the size of the atrioventricular canal ( AVC ) [39] . NR2F1 and NR2F2 are redundantly required for atrial differentiation in human iPSC-derived atrial cells [32] , although NR2F2 seems to have a primary role . Consistent with conserved requirements in vertebrate atrial development , lesions affecting NR2F2 have been associated with variable types of human congenital heart defects ( CHDs ) , in particular atrial septal defects ( ASDs ) and atrioventricular septal defects ( AVSDs ) , but surprisingly also left ventricular outflow tract obstruction ( LVOTO ) [40 , 41] . Therefore , while analysis of vertebrate Nr2f mutant models has provided insight into the molecular etiology of CHDs affecting the atria and AVC , the mechanisms underlying the observed phenotypic variability of CHDs , in particular the origins of ventricular malformations , in humans with NR2F2 mutations are not understood . Here , we identify that RA signaling directly regulates nr2f1a expression within the ALPM of zebrafish embryos and that retinoic acid receptors ( RARs ) can bind an absolutely conserved , yet unconventionally localized , response element . Using zebrafish mutants for both nr2f1a and nr2f2 , we find redundant functions at earlier developmental stages in restricting ventricular CM and promoting PM specification , independent of the later requirement for nr2f1a in promoting atrial differentiation . Cre-mediated genetic lineage tracing shows that tcf21+ progenitors more frequently become ventricular CMs and less frequently contribute to skeletal muscle within the posterior PM in nr2f1a; nr2f2 mutant embryos . Our results support a novel antagonistic mechanism that controls allocation of ventricular CM and PM progenitors within the anterior mesoderm of vertebrates and may help explain the correlation of craniofacial and heart defects as well as the variability found in CHDs associated with NR2F2 mutations in humans .
RA responsiveness of NR2F genes is conserved in chordates [28 , 42–44] . We identified nr2f1a as an RA-responsive gene within the ALPM of zebrafish embryos ( Fig 1A–1C ) , consistent with what other groups have described [28 , 44] . However , the nature of this regulation has not been assessed . Furthermore , although RA signaling affects epigenetic modifiers that control the expression of Nr2f1 in mammalian cells , a direct role for RA signaling has not been shown [30] . We found that RA treatment positively regulates nr2f1a expression after cycloheximide ( CHX ) treatment ( Fig 1D–1H ) , implicating a direct transcriptional regulatory mechanism . To determine if there are putative RA response elements ( RAREs ) for RAR binding sites in the nr2f1a promoter region , we first performed a mVISTA alignment of zebrafish , mouse , and human NR2F1 and NR2F2 genomic sequences . We found a highly conserved region within the 5’-untranslated region ( UTR ) of nr2f1a ( Fig 2A ) . Using the nuclear hormone receptor binding site prediction tool NHRscan in this region , we found a completely conserved direct repeat 1 ( DR1 ) site [45–49] within the 5’-UTR of these genes ( Fig 2B ) . While the location of this DR1 site is atypical , regulatory elements of other genes have been found to overlap with the 5’-UTR [50 , 51] . Despite the conservation of these sites across phyla , the site was not present in the zebrafish paralog nr2f1b , which is not RA responsive [28] . Electrophoretic mobility shift assays ( EMSAs ) and chromatin immunoprecipitation-quantitative PCR ( ChIP-qPCR ) indicated that RARs can bind the nr2f1a DR1 in vitro and in vivo ( Fig 2C and 2D ) . However , this site was not sufficient to respond to RA alone in luciferase assays ( S1 Fig ) . Therefore , our results suggest RA directly regulates nr2f1a expression and may involve interactions with a conserved DR1 RARE , although this atypical site may not be responsive to RA through a canonical activation mechanism . Within the ALPM , zebrafish nr2f1a is expressed immediately posterior to cardiac progenitors during somitogenesis ( Fig 3A ) . However , our recent study of nr2f1a mutants did not reveal requirements for Nr2f1a at these early developmental stages when the cardiac progenitor field is established [39] . Instead , we found that Nr2f1a is required to promote atrial CM differentiation at both the arterial and venous poles of the atrial chamber at subsequent stages of cardiogenesis , consistent with its expression specifically in atrial CMs within the developing cardiac tube [39] . Although zebrafish nr2f2 mutants do not have overt cardiovascular defects through at least two weeks of development [37 , 38] , zebrafish nr2f2 has low levels of expression within the ALPM during somitogenesis and is responsive to RA signaling ( S2 Fig ) , albeit significantly less so than nr2f1a as has been previously shown [28] . Therefore , we wondered if Nr2f2 functions redundantly with Nr2f1a at earlier stages of development within the ALPM . Using established engineered zebrafish nr2f2 deletion mutants [38] , we found that loss of either one or both wild-type ( WT ) nr2f2 alleles in nr2f1a mutant embryos resulted in overall progressively worse pericardial and yolk edemas coupled with blood pooling on the yolk compared to nr2f1a mutants alone ( Fig 3B–3E ) . Similarly , we found that loss of nr2f2 alleles in nr2f1a mutants produced hearts that were more dysmorphic and linear than nr2f1a mutant hearts alone ( Fig 4A–4D ) . Despite the exacerbation of the cardiac dysmorphology in the compound nr2f1a; nr2f2 mutants , we did not observe enhanced reduction of atrial chamber size or expression of AMHC , a marker of differentiated atrial CMs ( Fig 4A–4D ) . Valve markers were also not further expanded with the loss of nr2f2 alleles in nr2f1a mutants ( S3 Fig ) , consistent with a unique role of Nr2f1a in limiting valve development [39] . Surprisingly , in contrast to nr2f1a mutants , which display a specific reduction in atrial CMs ( Fig 4E; [39] ) , counting CMs with the myl7:h2afva-mCherry transgene [52] revealed that loss of one or both nr2f2 alleles in nr2f1a mutants produced an equivalent increase in ventricular CMs without producing any deficit in atrial CMs ( Fig 4E ) . Although we have found that the loss of atrial CMs is not due to early specification defects within the ALPM of nr2f1a mutants [39] , we posited that the specific surplus of ventricular CMs in the nr2f1a; nr2f2 mutants is due to an increase in ventricular CM specification at earlier stages of cardiogenesis because both nr2f1a and nr2f2 are expressed within in the ALPM [28] . Consistent with this idea , in the double mutants we observed a modest expansion of the cardiac progenitor marker Nkx2 . 5 at the 16 somite ( s ) stage ( S4 Fig ) and the amount of differentiating ventricular CMs , indicated by ventricular myosin heavy chain ( vmhc ) , was increased at the 20s stage ( Fig 5A–5E ) . Furthermore , loss of both nr2f1a and nr2f2 appeared to partially repress the ability of RA to inhibit vmhc expression ( S5 Fig ) . Together , these data suggest that Nr2f1a and Nr2f2 function redundantly to restrict the number of differentiating ventricular CMs . Previous analysis suggested that loss of RA signaling does not promote an increase in cardiac progenitor proliferation within the ALPM [53] . Consistent with this data , we did not find an increase in the number of proliferating Nkx2 . 5+ cells at the 16s stage in nr2f1a; nr2f2 mutant embryos ( S4 Fig ) . Thus , we postulated that the surplus ventricular CM progenitors in nr2f1a; nr2f2 mutant embryos , which refers to nr2f1amut with either nr2f2het or nr2f2mut alleles , may be at the expense of an adjacent cell lineage . We reasoned that candidates were the pharyngeal arch arteries ( PAAs ) and PMs , since their progenitors intermingle with the cardiac progenitor population within the anterior mesoderm of zebrafish [5 , 54] . We examined the posterior PAAs and PMs in nr2f1a; nr2f2 mutants at 48 hpf and 96 hpf , developmental time points when these cells have respectively differentiated [55] . Interestingly , we did not detect defects in PAA number and morphology in nr2f1a; nr2f2 mutant embryos carrying the kdrl:EGFP transgene ( S6 Fig ) . However , in contrast to the PAAs , we found the posterior protractor pectoralis ( pp ) , which is proposed to be a homolog of vertebrate neck muscles derived from the occipital LPM [56–60] , was often lost or reduced in nr2f1a; nr2f2 mutant embryos ( Fig 6A–6E ) . Although not as dramatic , the anterior dorsal mandibular ( 1st ) and hyoid ( 2nd ) arch derived muscles were also often smaller and disorganized compared to WT and nr2f1a mutant siblings ( Fig 6A–6D ) . A similar trend with respect to increased pp loss was observed at 75 hpf ( S7 Fig ) . However , for the analysis of the compound mutants we focused on 96 hpf to ensure that any defects were not due to developmental delay . Together , these data suggest that Nr2f1a and Nr2f2 together are required to promote posterior PM development . Due to the inverse effects on ventricular CM and posterior PM development in the nr2f1a; nr2f2 mutants , we sought to understand the relationship of these progenitors . Using two-color ISH to examine the expression of nr2f1a relative to tbx1 and tcf21 , we found that nr2f1a expression does not significantly overlap with tbx1 ( S8 Fig ) . However , nr2f1a and tcf21 expression domains overlap in a caudal region of the ALPM ( Fig 7A ) , interestingly , where lineage tracing has shown tcf21+ progeny give rise to CMs and posterior PM [5] . Despite the overlap in expression , tcf21 expression was not affected in nr2f1a; nr2f2 mutant embryos ( S8 Fig ) . Since the tcf21+ progenitors are overtly specified properly in nr2f1a; nr2f2 mutant embryos , we hypothesized that Nr2f proteins may affect a fate decision of progenitors within the posterior ALPM that can become ventricular and/or PM progenitors . To test this , we first used the inducible tcf21:CreERT2 transgene with the Cre-mediated color-switch line ubi:LOXP-AmCyan-STOP-LOXP-ZsYellow ( CsY ) to permanently label cells that have expressed tcf21+ ( Fig 7B ) . For lineage tracing experiments , nr2f1a homozygous mutants ( nr2f1amut ) coupled with nr2f2 heterozygosity ( nr2f2het ) or nr2f2 mutant homozygosity ( nr2f2mut ) were analyzed together ( referred to as nr2f1a-2mut ) , because our data suggest loss of a single WT nr2f2 allele in nr2f1a mutants produces similar ventricular CM and PM defects as loss of both WT alleles in nr2f1a mutants . Consistent with what has been reported [5] , we found that tamoxifen treatment of embryos containing both transgenes produced labeling of skeletal muscle within the PMs ( Fig 7C–7E ) . Although we did not find a decrease in the frequency of labeled anterior PMs within the 1st and 2nd arches , we found a decrease in the frequency of contribution to the pp in the nr2f1a-2mut embryos ( Fig 7E ) , supporting that Nr2f proteins promote the differentiation of skeletal muscle within the pp . We then reasoned that if Nr2f proteins are influencing a fate decision of ventricular and PM progenitors , tcf21+ progenitors should become ventricular CMs at an increased frequency in nr2f1a; nr2f2 mutant embryos . While we found that using tcf21:CreERT2; ubi:CsY labeled a few CMs , the expression was not as robust as for the PM . Therefore , we used the myl7:CsY transgene in combination with the tcf21:CreERT2 transgene to specifically and permanently label CMs derived from tcf21+ progenitors ( Fig 8A–8E ) . Examining labeled ventricular CMs , we found a trend where nr2f1a-2mut embryos have an increase in the number of embryos with >1 tcf21+-derived ventricular CM labeled compared to control embryos ( S9 Fig ) . Importantly , overall , nr2f1a-2mut embryos on average have about twice as many tcf21+-derived ventricular CMs compared to WT sibling embryos ( Fig 8F ) . Furthermore , there were increased number of labeled ventricular CMs found in nr2f1a-2mut embryos when just examining the pool of embryos that had >1 ventricular CM labeled ( Fig 8G ) , further supporting an increase in the frequency and number of tcf21+-derived ventricular CMs contributing to the ventricles in nr2f1a-2mut embryos . While atrial CMs were also labeled , their labeling was infrequent compared to labeling of ventricular CMs ( S9 Fig ) . We did not find a statistical difference in the frequency or average number of atrial CMs labeled within the populations ( S9 Fig ) . Together , our lineage tracing of tcf21+ progenitors demonstrates that a greater number of their progeny give rise to ventricular CMs in nr2f1a-2mut embryos , while fewer give rise to the pp .
Previous studies have demonstrated that RA signaling is necessary to limit cardiac specification and promote PM development [22 , 61] . With respect to heart development , early RA signaling restricts the posterior border of atrial and ventricular progenitors within the ALPM [25] . Despite similar effects on both cardiac cell types , mechanisms restricting atrial CMs and ventricular CMs downstream of RA signaling appear to be temporally distinct [25] . The present study suggests that Nr2f1a and Nr2f2 function redundantly downstream of RA signaling within the ALPM to regulate these converse effects on ventricular CM and PM specification . While there are numerous similarities between our observations in comparison to RA signaling-deficient embryos [22 , 25] , it is worth recognizing that the heart and PM defects found in nr2f1a; nr2f2 mutant embryos are less severe than what is typically found with RA signaling-deficient embryos . Therefore , we hypothesize that these Nr2f transcription factors likely are part of a larger RA-responsive gene network , including Hox genes and Fgf signaling , that contributes to this allocation of progenitors within the ALPM . Recent work examining Nr2f proteins in cranial neural crest that generate the anterior jaw has suggested significant redundancy with Nr2f1b and Nr2f5 in that developmental context [38] . However , we have not found any evidence of redundancy or genetic interactions with Nr2f1b and/or Nr2f5 in regulating heart development . For example , unlike what is observed with nr2f1a and nr2f2 , nr2f1a+/-;nr2f5+/- intercrosses produce ~25% ( 16/56 ) mutant embryos that are indistinguishable from nr2f1a mutant embryos with respect to the heart and blood pooling . We also have not found evidence for compensatory expression of any nr2f genes in the nr2f1a mutants ( S10 Fig ) . Additionally , we have not found defects in neural crest markers in the nr2f1a; nr2f2 mutants ( S8 Fig ) , suggesting loss of the pp is not secondary to neural crest defects . Our recent work suggests that Nr2f1a alone functions to promote atrial CM differentiation as the heart elongates and atrial CMs mature [39] , which is after it first appears in the ALPM . Here , we demonstrate that the change in the number of atrial CMs is not exacerbated in the nr2f1a; nr2f2 mutant embryos compared to nr2f1a single mutants . Instead , the number of atrial CMs is increased relative to nr2f1a single mutants , despite a similar overt reduction in atrial chamber size and lack of AVC , and equivalent to the number found in control embryos . We posit that these differential effects on the production of atrial CMs are because Nr2f proteins restrict the posterior extent of both atrial and ventricular progenitor fields within the ALPM and that a deficit in differentiating atrial CMs is not observed because the earlier requirements limiting the cardiac progenitor field offset the later requirements promoting atrial CM differentiation . NR2Fs are conserved regulators of atrial chamber development in vertebrates . Zebrafish nr2f1a mutants and mouse global Nr2f2 KOs present smaller atria [35 , 39] , while conditional Nr2f2 KO in the heart at later stages suggests a role in maintenance of atrial CM identity [36] . NR2F2 is required for atrial CM differentiation in human iPSCs [32] . Given these conserved requirements , it is interesting to compare the phenotypes of the nr2f1a; nr2f2 double mutants to the variability and severity of CHDs associated with NR2F2 lesions in humans . It has been proposed there is a direct correlation between the severity of CHDs and types of lesions impacting NR2F2 function [41] . Specifically , nonsense mutations proposed to be more damaging and resulting in significant loss of NR2F2 predominantly are associated with LVOTO , while missense mutations proposed to be less damaging are associated with ASDs and AVSDs [41] . The variable CHDs affecting the arterial pole of the ventricle and the atrial chamber are highly reminiscent of the chamber-specific defects we observe in nr2f1a; nr2f2 double mutants compared to single nr2f1a mutants , which overtly affect the production of ventricular CMs and atrial CMs , respectively . Together , these data support the hypothesis that levels of total Nr2f dosage differentially affect chamber-specific cardiogenic processes within the vertebrate heart . Moreover , we propose that greater loss of NR2F transcription factors , through more damaging alleles or genetic loss , produces ventricular chamber defects due to earlier developmental requirements within the ALPM , while ASDs or AVSDs may occur due to a more modest loss of total NR2F signaling that is required at later stages of atrial CM differentiation . Thus , our studies offer a working model to explain the molecular etiology of congenital LVOTO and ASDs/AVSDs associated with NR2F2 mutations in humans . While NR2F proteins have been studied in numerous development contexts , significant analysis of the requirements for Nr2fs in skeletal muscle have not been reported . Virtually all the Nr2f proteins are expressed in the somites of zebrafish [28] . Interestingly , in mice Nr2f2 is broadly expressed in skeletal muscle , including the somites and the cranial muscles [34 , 41] . Limb-specific Nr2f2 KOs indicate it is required for limb muscle development [34] and mechanistically there is evidence that Nr2f2 can compete with myoD in muscle differentiation [62] . Therefore , there is precedence for Nr2f2 functions in somite-derived skeletal muscle , but requirements in PM development have not been reported . It is interesting to note that cranio-facial defects have been associated with genetic deficiencies that affect both NR2F1 and NR2F2 in humans [40 , 63 , 64] . In two independent cases , similar-sized deletions that eliminate NR2F2 were associated with cranial abnormalities as well as ASDs [40 , 64] . However , overt cranio-facial defects similar to those found in the deficiencies were not reported in patients found to have specific mutations that affect NR2F2 and are associated with CHDs [41] . Therefore , although specific defects in craniofacial muscle were not reported , there is precedence for an association between NR2F gene loss and both craniofacial and CHDs in humans . Recent clonal analysis in mice has suggested there are common cardio-pharyngeal progenitors that contribute progeny to the neck muscles , the arterial pole and atria that are distinct from other cardio-pharyngeal populations of the SHF [8] . Our data are also consistent with a close association of ventricular outflow tract and pharyngeal neck muscle progenitors and a distinction from other SHF progenitors , which arise more anteriorly [65 , 66] . Specifically , while anterior dorsal 1st and 2nd arch muscles are reduced , we predominantly find that nr2f1a; nr2f2 embryos lose the posterior pp muscle ( cucullaris ) , which has been proposed to be homologous to the ALPM-derived trapezius neck muscles in mammals [56–60] . Therefore , reminiscent of the recent retrospective clonal analysis in mice [8] , these results hint at the existence of a distinct posterior progenitor population with cardiac and PM potential that does not correspond to the anterior SHF . Given the existence of bi-potent cardio-pharyngeal progenitors in mice and Ciona [6–8 , 19 , 20] , one interpretation of our results is that RA signaling and consequently Nr2f proteins , at least in part , act on bi-potent cardio-pharyngeal progenitors . Although it is clear from the retrospective clonal lineage analysis in mice that there are multiple populations of bi-potent cardio-pharyngeal progenitors , these populations are rare and only found from examination of large sample sizes [6–8] . While zebrafish tcf21+ , as well as nkx2 . 5+ progenitors , can give rise to ventricular CMs and PMs [5 , 67] , it is not yet clear whether there are bi-potent progenitors with cardiac and PM potential . Although the defects we observe in ventricular CM and PM development of nr2f1a; nr2f2 mutant embryos are less dramatic than with loss of RA signaling , in neither case are the defects subtle enough that a very rare population of bi-potent progenitors is likely being affected . Instead , we favor a model where there is a larger population of progenitors within the ALPM that have the potential to become either ventricular CMs and PM , with signals such as Nr2f proteins functioning downstream of RA signaling to influence their allocation into one of these populations . Overall , our study provides valuable insight into the requirements of Nr2f genes in vertebrate cardiac and cranial muscle development . These studies may help us to further understand molecular and genetic etiology controlling phenotypic variability of CHDs as well as developmental syndromes that have congenital malformations concomitantly affecting the heart , head , and neck muscles in humans .
All zebrafish husbandry and experiments were performed in accordance with protocols approved by the Institutional Animal Care and Use Committee ( IACUC ) of Cincinnati Children's Hospital Medical Center . Adult zebrafish were raised and maintained under standard laboratory conditions . Transgenic lines used were: Tg ( kdrl:nlsEGFP ) ubs1 [68] , Tg ( kdrl:EGFP ) la116 , TgBAC ( −36nkx2 . 5:ZsYellow ) fb7 [69] , Tg ( actc1b:GFP ) zf13 [70] , Tg ( tcf21:nucEGFP ) pd41 [71] , Tg ( tcf21:CreERT2 ) pd42 [72] , Tg ( ubi:LOXP-AmCyan-STOP-LOXP-ZsYellow ) fb5 [69] , Tg ( myl7: LOXP-AmCyan-STOP-LOXP-ZsYellow ) fb2 [69] , Tg ( myl7:h2afva-mCherry ) sd12 [52] and Tg ( hsp70l:EGFP-VP16-RAR- ) c1004 [73] . Mutant alleles used were: nr2f1ael512 and nr2f2el506 [38 , 39] . Single and two-color whole mount ISH were performed using NBT/BCIP ( Roche ) and INT/BCIP ( Roche ) , as previously reported [74] . Digoxygenin- and fluorescein-labeled anti-sense RNA probes for zsyellow ( ZDB-EFG-110824-1 ) , egfp ( ZDB-EFG-070117-1 ) , nr2f1a ( ZDB-GENE-980526-115 ) , nr2f2 ( ZDB-GENE-990415-252 ) , vmhc ( ZDB-GENE-991123-5 ) , nkx2 . 5 ( ZDB-GENE-980526-321 ) , dlx2a ( ZDB-GENE-980526-212 ) , tbx1 ( ZDB-GENE-030805-5 ) , tcf21 ( ZDB-GENE-051113-88 ) , and klf2a ( ZDB-GENE-011109-1 ) were used . Area measurements were performed using ImageJ . Sequences for zebrafish , mouse , and human Nr2f genes plus a 10kb region 5’ and 3’ to the genes were taken from Ensembl ( ensembl . org ) and aligned using mVista ( http://genome . lbl . gov/vista/mvista/submit . shtml ) . Locations , excluding exons , in which there was over 50% conservation between any of the sequences were analyzed for the presence of RARs . Conserved sequences were input into NHRscan ( http://www . cisreg . ca/cgi-bin/NHR-scan/nhr_scan . cgi ) to identify potential RAR binding sites . Total RNA isolation and RT-qPCR was performed using previously reported methods [75] . Briefly , whole embryo RNA was obtained from groups of 30 embryos using Trizol ( Ambion ) and Purelink RNA Microkit ( Invitrogen ) . cDNA was synthesized using 1μg total RNA and the ThermoScript Reverse Transcriptase kit ( Invitrogen ) . RT-qPCR was performed using Power SYBR Green PCR Master Mix ( Applied Biosystems ) in a BioRad CFX-96 PCR machine . Expression levels were standardized to β-actin expression and data were analyzed using the 2-ΔΔCT Livak Method . All experiments were performed in triplicate . Primer sequences for β-actin were reported previously [73 , 75] . All primer sequences used for RT-qPCR are in the S1 Table . All drug treatments were administered to embryos in 2 mL of blue water with drug at specified concentrations in a glass vial with 25–30 embryos/vial at 28 . 5°C . For analysis of nr2f1a and nr2f2 expression , embryos were treated with CHX ( 10 μM , Sigma 48591 ) , RA ( 1 μM , Sigma R2625 ) , and DEAB ( 10 μM , Sigma D86256 ) at tailbud stage for 1 hour . For analysis of vmhc expression in nr2f1a; nr2f2 mutants , embryos were treated with 0 . 05 μM RA at the 3s stage until the 20s stage . Drugs were washed out 3X with embryo water then the embryos were fixed in 4% formaldehyde for analysis . Vmhc-stained embryos were genotyped following imaging . Tamoxifen ( 10 μM , Sigma H7904 ) was administered in 30 mL of blue water with 0 . 003% PTU in petri dishes to embryos at 30%-50% epiboly until embryos were analyzed or through 2 days of development . ChIP-qPCR was performed essentially as previously reported [75] . Hemizygous Tg ( hsp70l:VP16-RAR:EGFP ) c1004 adults were crossed to WT adult zebrafish . The resulting embryos were collected at tailbud stage and heat-shocked at 37°C for 30 minutes . Transgenic embryos were sorted from their non-transgenic control siblings by the presence of GFP . Embryos ( n = 100 ) were dechorionated and fixed in 1% formaldehyde 2 hours after heat-shock . Cells were lysed by gentle pipetting in cell lysis buffer . Nuclei were lysed and DNA was sheered by sonication with glass beads to 200-600bp fragments . Dynabeads ( Invitrogen ) were used to pull down GFP tagged proteins with ChIP-grade polyclonal anti-GFP antibody ( Abcam ab290 ) per manufactures instructions . Samples were de-crosslinked and qPCR was used to quantify the fold difference in enrichment of the DR1 RARE in the nr2f1a promoter and the known DR5 RARE in the Cyp26a1 promoter as compared to a nr2f1a promoter region not containing a RARE . Expression levels were standardized to the no antibody control signal and data were analyzed using the 2-ΔΔCT Livak Method . Primer sequences for cyp26a1 ChIP-PCR were reported previously [76] . Primer sequences for nr2f1a DR1 ChIP-PCR and control are indicated in S1 Table . EMSA was performed essentially as previously reported [77] . Oligonucleotides were designed containing the nr2f1a DR1 site ( GTGTCAAAGTTCA ) , the nr2f1a DR1 site with a targeted mutation in the second half site of the DR1 abolishing the direct repeat ( GTGTCAAAGTCAT ) , and a previously reported Cyp26a1 DR5 site [76] . A complementary oligonucleotide was designed with a 5’ LI-COR IRDye 700 ( IDT ) . The oligonucleotides were annealed and the ends filled with Klenow ( New England Biolabs ) . Zebrafish myc-rarab was in the pCS2+MT . Zebrafish RXRba was cloned into pCS2p+ . Proteins for EMSA were made using the TnT SP6 Quick Coupled Transcription/Translation System ( Promega ) . Protein samples were gently mixed with LI-COR tagged probes and incubated at room temperature for 20 minutes . 4% polyacrylamide gels were run for 2 hours at 150 V . Gels were imaged using an Odyssey CLx LI-COR imager . Embryos were fixed for 1 hour at room temperature in 1% formaldehyde in PBS in 3 ml glass vials . Embryos were washed 1X in PBS and then 2X in 0 . 2% saponin/1X PBS , followed by blocking in 0 . 2% saponin/0 . 5%sheep serum/1X PBS ( Saponin blocking solution ) for one hour . AMHC ( S46 ) and MHC ( sarcomeric myosin; MF20 ) primary antibodies ( Developmental Studies Hybridoma Bank ) were incubated at 1:10 in Saponin blocking solution . Rabbit polyclonal DsRed antibody ( Clontech ) , to detect mCherry , and Living colors anti-RCFP ( Clontech ) , to detect ZsYellow , were used at a 1:1000 dilution . Rabbit anti-GFP ( Abcam ) was used at 1:500 . Rabbit anti-Nkx2 . 5 ( Gene Tex ) was used at 1:250 . Mouse anti-pHH3 ( Abcam ) was used at 1:1000 . All secondary antibodies were used at dilutions of 1:100 . Antibody information is also listed in S2 Table . Cell counts were performed by gently flattening embryos under a coverslip and counting the fluorescent nuclei in each chamber . For all imaging except Nkx2 . 5/pHH3 , embryos were imaged using a Zeiss M2BioV12 Stereo microscope . For Nkx2 . 5/pHH3 , embryos were post-fixed in 2% formaldehyde/1X PBS for two hours and mounted in 1% low-melt agar on 2% agar plates . Images of one side of the embryo were taken using a Nikon A1R Multiphoton Upright Confocal Microscope with a 16X water immersion objective . 200 μm optical sections were taken with the resonance scanner . The promoter fragments for both reporters used were cloned into the Kpn and HindIII sites of the pGL3 ( Promega ) multiple cloning site . The DR1-ef1a construct contains 165 base pairs ( bp ) of the nr2f1a promoter and 5’UTR adjacent to 193 bp of a minimal elongation factor 1a ( ef1a ) promoter ( green ) . The nr2f1a-DR1 construct contains 371 bp that include the promoter and 5’UTR containing the conserved DR1 site . The pGL3-12XRARE-tk vector and dual luciferase assays were reported and performed in HEK293 cells as described previously [78] . To compare two groups , we performed a Student’s t-test or Mann-Whitney test . To compare 3 or more conditions are different , we performed ANOVA analysis . To determine if two proportions were statistically distinct we performed a Chi-squared test or Fisher’s exact test . Statistical analysis was performed using GraphPad Prism . A p value < 0 . 05 was considered statistically significant for all analysis . | Many developmental syndromes include both congenital heart and craniofacial defects , necessitating a better understanding of the mechanisms underlying the correlation of these defects . During early vertebrate development , cardiac and pharyngeal muscle cells originate from adjacent , partially overlapping progenitor fields within the anterior mesoderm . However , signals that allocate the cells from the adjacent cardiac and pharyngeal muscle progenitor fields are not understood . Mutations in the gene NR2F2 are associated with variable types of congenital heart defects in humans . Our recent work demonstrates that zebrafish Nr2f1a is the functional equivalent to Nr2f2 in mammals and promotes atrial development . Here , we identify that zebrafish nr2f1a and nr2f2 have redundant requirements at earlier stages of development than nr2f1a alone to restrict the number of ventricular CMs in the heart and promote posterior pharyngeal muscle development . Therefore , we have identified an antagonistic mechanism that is necessary to generate the proper number of cardiac and pharyngeal muscle progenitors in vertebrates . These studies provide evidence to help explain the variability of congenital heart defects from NR2F2 mutations in humans and a novel molecular framework for understanding developmental syndromes with heart and craniofacial defects . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"fish",
"medicine",
"and",
"health",
"sciences",
"cardiovascular",
"anatomy",
"skeletal",
"muscles",
"vertebrates",
"pharyngeal",
"muscles",
"animals",
"animal",
"models",
"osteichthyes",
"developmental",
"biology",
"model",
"organisms",
"heart",
"experimental",
"organism... | 2019 | Nr2f-dependent allocation of ventricular cardiomyocyte and pharyngeal muscle progenitors |
In Caenorhabditis elegans ( C . elegans ) , the promotion of longevity by the transcription factor DAF-16 requires reduced insulin/IGF receptor ( IIR ) signaling or the ablation of the germline , although the reason for the negative impact of germ cells is unknown . FOXO/DAF-16 activity inhibits germline proliferation in both daf-2 mutants and gld-1 tumors . In contrast to its function as a germline tumor suppressor , we now provide evidence that somatic DAF-16 in the presence of IIR signaling can also result in tumorigenic activity , which counteracts robust lifespan extension . In contrast to the cell-autonomous IIR signaling , which is required for larval germline proliferation , activation of DAF-16 in the hypodermis results in hyperplasia of the germline and disruption of the surrounding basement membrane . SHC-1 adaptor protein and AKT-1 kinase antagonize , whereas AKT-2 and SGK-1 kinases promote , this cell-nonautonomous DAF-16 function . Our data suggest that a functional balance of DAF-16 activities in different tissues determines longevity and reveals a novel , cell-nonautonomous role of FOXO/DAF-16 to affect stem cells .
The forkhead box O ( FOXO ) subfamily of Forkhead transcription factors is conserved from Caenorhabditis elegans ( C . elegans ) to mammals [1] . Mammalian FOXO transcription factors consist of four members: FOXO1 , FOXO3 , FOXO4 and FOXO6 , whereas only one homologue , DAF-16 , is encoded in the C . elegans genome . DAF-16/FOXO proteins are inactivated by the insulin/IGF-1 signaling ( IIS ) through PI3K and the AGC kinases Akt/SGK which promote its cytosolic localization [2]–[7] . Starvation reduces IIS , resulting in nuclear localization and activation of DAF-16 . Stress stimuli also result in nuclear translocation and activate FOXO via JNK and MST1 even in the presence of Akt [8] , [9] . The Akt/FOXO signaling network acts as a critical control mechanism at the intersection between cancer and stem cell biology . FOXO proteins have been considered as tumor suppressors , because of their ability to induce DNA damage repair , cell cycle arrest and apoptosis [10] , [11] . Consistently , loss of functional FOXO is associated with tumorigenesis in various organs [11]–[13] . On the other hand , FOXO proteins are required for the long-term maintenance of both normal and cancer stem cells . Mice with FOXO1 , FOXO3 and FOXO4 triple knockout display a marked reduction of hematopoietic stem cells due to increased physiological oxidative stress [14] . In the cancer stem cells of chronic myeloid leukemia , FOXO3 is enriched in the nucleus and essential for maintaining these cancer stem cells [15] . In C . elegans , DAF-16 is a key regulator of development , longevity and stress response . Active DAF-16 resulting from reduced IIS leads to dauer formation , enhanced stress response and lifespan extension [16]–[18] . Recent studies reveal that the cross-talk between the reproductive system and somatic tissues plays an important role in development and aging of C . elegans . During larval development DAF-16 inhibits robust proliferation of the germline through both cell-autonomous ( the germline ) and -nonautonomous ( the musculature ) mechanism [19] . In adult animals intestinal and muscular DAF-16 slow down reproductive aging via oocyte and germline quality maintenance [20] . Activation of DAF-16 upon reduced IIS also inhibits tumor growth in the germline of gld-1 mutants [21] , indicating that the role of FOXO as tumor suppressor is evolutionarily conserved . On the other side , signals from the reproductive system regulate DAF-16 activity in the soma: Elimination of mitotic germ cells results in nuclear entry of intestinal DAF-16 and extends lifespan [22] , [23] . Ablation of the somatic gonad precursors , however , abrogates the lifespan extension of the germline-ablated animals [23] . Even though several factors , such as KRI-1 , TCER-1 and DAF-9 , have been found to be involved in transduction of such signaling [24] , [25] , the details of the signaling mechanism are still not well known . In a previous study we have shown that the C . elegans p52Shc homolog SHC-1 modulates DAF-16 activity through promoting its nuclear entry ( Neumann-Haefelin et al . , 2008 ) . SHC-1 negatively regulates IIS by inhibition of the insulin/IGF receptor DAF-2 . SHC-1 also associates with MEK-1 , the mitogen-activated protein kinase kinase 7 ( MAPKK7 ) , to activate a C . elegans JNK homolog JNK-1 , thus affecting stress response and longevity [9] , [26] . SHC-1 and MEK-1 also mediate activation of an alternative C . elegans JNK homolog , KGB-1 , upon heavy metal stress [27] . Shc-like proteins have been found in metazoan animals from nematodes to humans , suggesting their roles might also be conserved in evolution [28] . Here , we report a novel role of DAF-16 activity in epidermal cells affecting the reproductive system in a cell-nonautonomous manner , resulting in germline hyperplasia and disruption of the surrounding extracellular matrix of C . elegans . The adapter protein SHC-1 modulates both IIS and JNK pathways to antagonize this hypodermal function of DAF-16 . Our finding reveals a new aspect of DAF-16 activity besides its well known roles in the regulation of longevity , stress response and dauer formation . Our data show that these two aspects affect longevity differently and indicate that the capability of DAF-16 to extend lifespan is dependent on the balance of these two opposing effects .
SHC-1 positively regulates the activity of DAF-16 by both inhibiting DAF-2 and activating JNK signaling pathways [26] . shc-1 mutant animals are generally healthy , grow at a normal rate , and produce normal numbers of offspring [26] , [27] . However , they live about 25% shorter than wild type animals and this reduced lifespan is accompanied by cytoplasmic retention of DAF-16 [26] . Since , according to this model , daf-16 acts downstream of shc-1 , access of wild type DAF-16 , in particular the proportion that escapes phosphorylation by active AKT-1/AKT-2/SGK-1 , should compensate for the loss of shc-1 during the control of lifespan . We tested this hypothesis relying on the frequently used strain TJ356 which expresses the full length daf-16 isoform a fused to GFP in a wild type background [29] . TJ356 animals have increased DAF-16 activity , however , displayed lifespan comparable to wild type ( Figure 1A and Table 1 ) , consistent with a previous report [29] . To our surprise , the daf-16 transgene did not extend , but further reduced the already short lifespan of shc-1 ( Figure 1A and Table 1 ) . Remarkably , about 50% of the shc-1 ( ok198 ) ;Is[daf-16::gfp] adult animals died within the first five days of adulthood . In addition , about half of the animals were sterile and the remaining fertile animals showed a strongly reduced brood size ( 49 vs . 348 , Table S2 ) . In order to exclude that this phenotype is allele-specific or caused by a background mutation linked to the shc-1 locus , we crossed another allele of shc-1 , tm1729 , into the Is[daf-16::gfp] background and observed the same phenotype ( Table 1 and Table S2 ) . We conclude that both daf-16 expression and loss of shc-1 contribute to the observed phenotypes . This suggests that the combination of daf-16 transgene and shc-1 mutant results in a synergistic effect not previously reported in either TJ356 or shc-1 ( ok198 ) . Unexpectedly , raising the adult animals in the presence of 5-fluoro-2′-deoxyuridine ( FUDR ) , an inhibitor of DNA synthesis which blocks cell proliferation , suppressed the early lethality and extended the lifespan of shc-1 ( ok198 ) ;Is[daf-16::gfp] animals up to two fold ( Figure 1B and Table 1 ) . This was surprising , since no lifespan extension was observed upon FUDR treatment of wild type . FUDR is one of the most commonly used drugs in the treatment of colorectal cancer and frequently used in C . elegans lifespan assays to facilitate strain handling due to its ability to block cell proliferation and generation of progeny . Based on this result , we suggest that the substantial lifespan reduction and lethality of shc-1 ( ok198 ) ;Is[daf-16::gfp] may be linked to proliferation and , possibly , abnormal mitosis , and that this detrimental effect is blocked by FUDR . The germline is the only tissue that undergoes mitosis in adult animals , so we asked whether mitosis of the germline leads to the early lethality of shc-1 ( ok198 ) ;Is[daf-16::gfp] animals . The GLP-1 mediated Notch signaling represses meiosis of the germ cells and keeps them in mitotic proliferation from the L3 larval stage . In glp-1 mutants the germ cells stop mitotic cell division and enter meiosis [30] . If the lethality observed in shc-1 ( ok198 ) ;Is[daf-16::gfp] animals is indeed caused by proliferation of the germ cells , inactivation of glp-1 should result in a similar lifespan extension as seen upon FUDR treatment . To test this assumption , we used the temperature sensitive allele glp-1 ( q231 ) , because glp-1 ( q231 ) animals at the restrictive temperature 25°C do not have a pronounced difference of mean lifespan compared to wild type probably due to a dysfunction in the atrophy intestine , in contrast to other long lived glp-1 ( lf ) alleles , such as e2141ts or q158 ( Figure S1 ) [22] , [25] . Therefore , alterations of lifespan caused by detrimental germline proliferation should be easily detectable in this stain . shc-1 ( ok198 ) ;glp-1 ( q231 ) ;Is[daf-16::gfp] animals lived 240% longer than shc-1 ( ok198 ) ;Is[daf-16::gfp] when shifted at the L2 larval stage to the restrictive temperature ( Figure 1C and Table 1 ) . glp-1 mutation also extended lifespan of shc-1;Is[daf-16::gfp] animals when animals are re-shifted to 20°C after larval germline proliferation has ceased ( Figure S2 ) . In addition , the early lethality observed in shc-1;Is[daf-16::gfp] animals was completely abolished . Ablation of the mitotic germline or a glp-1 ( e2141 ) background extends lifespan of wild type animals , which require intestinal DAF-16 and the adaptor protein KRI-1 . [23] . However , FUDR treatment did not extend lifespan of wild type animals ( Figure 1B and Table 1 ) , suggesting that FUDR uses a different mechanism as ablation of the germline to affect longevity of shc-1;Is[daf-16::gfp] animals . To validate this assumption , we further tested whether kri-1 knock-down could suppress the FUDR dependent lifespan extension . kri-1 RNAi treatment did not shorten the lifespan of shc-1;Is[daf-16::gfp] animals fed with FUDR ( Figure 1D and Table 1 ) . Based on these observations , we propose that early lethality of shc-1 ( ok198 ) ;Is[daf-16::gfp] animals is due to a negative input of germline proliferation rather than a general sickness of the strain . We noticed in the strain TJ356 that transgenic daf-16 extended lifespan of wild type animals significantly only in the presence of FUDR ( Figure 1B ) . To further exclude any effect of the transgene used , we generated three independent lines carrying extra-chromasomal daf-16::gfp transgenes in wild type background . In the absence of FUDR , these strains also had lifespan like wild type animals ( Figure 1E ) , whereas they lived 19 . 3% , 20 . 6% , and 26 . 7% , respectively , longer than wild type animals in the presence of FUDR ( Figure 1F ) . Thus , transgenic DAF-16 indeed extends lifespan if germline proliferation is inhibited . Recently , FUDR has been shown to extend lifespan of tub-1 mutant animals [31] . Together , this suggests that the use of FUDR in lifespan assays may be inconspicuous in some genetic backgrounds [32] , but may strongly affect others . Analysis of shc-1 ( ok198 ) ;Is[daf-16::gfp] animals using DIC microscopy revealed that the reproductive system exhibited multiple defects ( Figure 2 ) . The most prominent phenotype was the accumulation of cells in the pseudocoelom in almost all of the one day old adult animals ( Figure 2B , 2J and Figure S3 ) . These cells superficially looked like germ cells , yet were clearly localized outside the gonad . In the background of either glp-1 ( q231 ) or glp-1 ( e2141 ) these extra cells were absent , indicating that they may be germ cells ( Figure S3 ) . They showed substantial variations of shape and size . In some animals the boundary of the gonad was totally disrupted . It was , therefore , not possible to extract the gonad of these animals in an intact form ( data not shown ) . In gonads with still recognizable morphology , a defect in the distal tip cell ( DTC ) migration was observed ( Figure 2C and Figure S4 ) . Analysis of shc-1 ( ok198 ) ;Is[daf-16::gfp] at different larval stages revealed that the gonad ruptured at the L3 larval stage at the proximal side next to the developing somatic gonad primordium . In animals in which germ cells leaked out into the pseudocoelom , a sharp gonad boundary indicative of an intact gonadal basement membrane , as seen in the wild type animals ( Figure 2D ) , was absent in shc-1 ( ok198 ) ;Is[daf-16::gfp] animals ( Figure 2E and Figure S5 ) . The gonadal basement membrane can be visualized by staining with MitoTracker [33] , and MitoTracker staining confirmed its partial disruption and leakage ( Figure 2G and Figure S5 ) . In order to determine the identity of the extra-gonadal cells , we stained shc-1 ( ok198 ) ;Is[daf-16::gfp] animals with an antibody against the germ cell specific protein PGL-1 , a P-granule component . The released cells were PGL-1 positive , corroborating their identity as germ cells ( Figure 2I , Movie S1 ) . We noticed that at the mid-L3 stage the disrupted gonad arms of shc-1 ( ok198 ) ;Is[daf-16::gfp] animals contained significantly more germ cells than in wild type ( 46±14 vs . 35±4 ) ( Figure 2K ) . In shc-1 ( ok198 ) ;Is[daf-16::gfp] animals in which one gonad arm was still intact , we counted more germ cells in the disrupted gonad arm than in the intact one . In adult animals the volume of some of these released cells increased so that they looked like oocytes . However , they lacked the condensed diakinetic chromosomes that are characteristic for oocytes ( Figure 2L ) . Instead , some of them displayed endomitotic chromosomes ( Figure 2M ) . Taking together , these data suggest that shc-1 ( ok198 ) ;Is[daf-16::gfp] animals show abnormal larval germline proliferation and disruption of the gonadal basement membrane . In the following results we focus mainly on the disruption of gonad phenotype . Knock-down of daf-16 expression by RNAi significantly suppressed all phenotypic aspects ( low brood size , sterility , gonad disruption and early adult lethality ) of shc-1 ( ok198 ) ;Is[daf-16::gfp] animals ( Figure 3A , Table 1 and Table S2 ) . The suppression was only partial since knock-down of daf-16 by RNAi feeding was incomplete , indicated by a weak but persistent expression of DAF-16::GFP ( Figure S6 ) . In order to rule out an involvement of background mutations , we generated new transgenic animals carrying extra-chromosomal daf-16::gfp in shc-1 mutants . These strains displayed similar but less severe germline and gonad defects ( Figure 2J and Table 3 ) , which is probably due to the weaker DAF-16::GFP expression in extrachromosomal vs . integrated line ( Figure S7 ) or due to the mosaic inheritance of extra-chromosomal transgenes . The phenotype we observed in shc-1;Is[daf-16::gfp] animals was stronger than in Is[daf-16::gfp] animals , suggesting that reduced SHC-1 activity enhances DAF-16 . This is in apparent contrast to our previous experiments in which we showed that DAF-16 activation is decreased in shc-1 mutant [26] . To understand this obvious discrepancy , we tested whether disruption of gonad phenotype could also be caused by inactivating DAF-16 in either wild type or shc-1 ( ok198 ) mutant background . No gonad disruption was seen in either daf-16 ( mu86 ) or daf-16 ( mgDF50 ) animals ( Figure 3B and Table 2 ) , suggesting that reduction of daf-16 activity was not the cause of this defect . However , about 7% of the shc-1 ( ok198 ) animals showed gonad disruption , yet with a delayed onset . Deletion of daf-16 almost fully suppressed this defect in shc-1 ( ok198 ) animals ( Figure 3B and Table 2 ) . This strongly suggests that the germline phenotype is not a synthetic or artificial phenotype seen only in shc-1;Is[daf-16::gfp] animals , and that DAF-16 and SHC-1 affect the reproductive system of C . elegans in opposite way . Since we have previously shown that SHC-1 promotes nuclear localization of DAF-16 by inhibiting DAF-2 and activating JNK-1 , we expected an increase of cytosolic vs . nuclear DAF-16::GFP in shc-1 ( ok198 ) ;Is[daf-16::gfp] compared to Is[daf-16::gfp] animals [26] . If increased cytosolic vs . nuclear DAF-16::GFP would be the cause of the observed phenotype , then expression of a constitutively nuclear daf-16 mutant ( daf-16 ( 4A ) ::gfp ) should not result in gonad disruption . Expressing daf-16 ( 4A ) ::gfp in wild type animals caused only weak disrupted gonad phenotype ( Figure 3C and Table 3 ) . However , transgenic expression of daf-16 ( 4A ) ::gfp in a shc-1 ( − ) background caused severe defects . We observed that DAF-16 ( 4A ) ::GFP was nuclearly localized and this nuclear localization was not affected by either presence or absence of SHC-1 ( Figure 3D ) . In addition , shc-1 mutation did not affect the expression level of daf-16 ( 4A ) ::gfp ( Figure 3E ) . We also compared GFP intensities in the daf-16::gfp strains and correlated them to the severity of the phenotype . We found that , to result in a comparable phenotype , wild type daf-16::gfp required a higher expression level than daf-16 ( 4A ) ::gfp ( Figure S7 ) . Given that wild type DAF-16 is mostly retained in the cytoplasm , this observation also indicates that nuclear instead of cytosolic DAF-16::GFP is the cause for the phenotype and SHC-1 antagonizes DAF-16 to ensure the gonadal integrity not simply via affecting its subcellular localization . AKT-1 , AKT-2 and SGK-1 are known to phosphorylate DAF-16 directly and inhibit its nuclear entry [4] , [7] . To test whether mutations in akt-1 , akt-2 or sgk-1 that activate DAF-16 also influence the integrity of the gonadal basement membrane , shc-1 ( ok198 ) ;akt-1 ( ok525 ) , shc-1 ( ok198 ) ;akt-2 ( ok393 ) and shc-1 ( ok198 ) ;sgk-1 ( RNAi ) were analyzed ( Figure 4A and Table 2 ) . Neither akt-1 ( ok525 ) , akt-2 ( ok393 ) single mutants nor sgk-1 ( RNAi ) displayed gonad disruption . Inactivation of akt-2 or sgk-1 did not enhance the penetrance of the gonad defect of shc-1 ( ok198 ) animals , either . In contrast , 58 . 5±21 . 0% of first-day adult shc-1 ( ok198 ) ;akt-1 ( ok525 ) animals showed gonad disruption , providing further evidence that this phenotype is not an artificial effect of transgenic daf-16 since this strain does not harbor a daf-16 transgene . In addition , a loss-of-function allele of daf-16 suppressed this phenotype in shc-1;akt-1 animals , verifying its dependence on DAF-16 ( Figure 4A ) . Taken together , these results strongly suggest that AKT-1 ensures gonadal integrity via inhibiting DAF-16 . Since akt-1 is downstream of daf-2 , shc-1;daf-2 double mutant should therefore exhibit the same phenotype as akt-1;shc-1 . To test this , we crossed the temperature sensitive mutant daf-2 ( e1370 ) into shc-1 ( ok198 ) background . At 25°C both daf-2 ( e1370 ) and shc-1 ( ok198 ) ;daf-2 ( e1370 ) mutants formed constitutive dauer larvae with a developmentally arrested germline that prevented the analysis of adult animals . At 20°C , at which daf-2 is partially inactivated , daf-2 ( e1370 ) did not alter the defect in the gonadal integrity in shc-1 mutant ( Figure 4B and Table 2 ) . In order to explore whether DAF-2 inhibits a DAF-16 dependent effect on the germline , we crossed daf-2 ( e1370 ) into shc-1 ( ok198 ) ;Is[daf-16::gfp] animals and quantified the percentage of animals displaying disrupted gonad at 15°C , since at 20°C shc-1 ( ok198 ) ;daf-2 ( e1370 ) ;Is[daf-16::gfp] showed development arrest due to excessive daf-16 transgene expression ( Figure S8 ) . At 15°C , activity of DAF-2 in the daf-2 ( e1370 ) mutant is already reduced , indicated by an extended lifespan [34] . Surprisingly , daf-2 ( e1370 ) significantly reduced instead of increased the penetrance of animals with gonad disruption ( Figure 4B and Table 2 ) , indicating that this phenotype develops only in animals with intact IIS . In addition , the sterile phenotype of shc-1 ( ok198 ) ;Is[daf-16::gfp] animals was completely suppressed by daf-2 mutation and the brood size of the fertile animals was increased ( Table S2 ) . Consistent with the suppression of the defects in the reproductive system , the early lethality in adulthood was abolished and lifespan was extended from 11 . 6 days to 31 . 5 days ( Figure 4C ) . These data suggest that the phenotypes in shc-1 ( ok198 ) ;Is[daf-16::gfp] animals require active DAF-2 . AKT-1 is known to act downstream of DAF-2 to inhibit DAF-16 . Here , however , we observed that loss of akt-1 and loss of daf-2 have divergent consequences for DAF-16 . To address whether the daf-2 mutation could still suppress the defect in the basement membrane in akt-1;shc-1 animals , we examined shc-1 ( ok198 ) ;daf-2 ( e1370 ) ;akt-1 ( ok525 ) triple mutant at 15°C , since only at this temperature the germline of the animals could proliferate after L3 stage . daf-2 mutation suppressed gonad disruption in shc-1 ( ok198 ) ;akt-1 ( ok525 ) animals completely ( Figure 4D , Figure S9 ) , indicating that active IIR DAF-2 contributes to gonad disruption caused by DAF-16 . As we observed antagonistic role of DAF-2 and AKT-1 affecting gonadal integrity , we further explored the downstream effectors of DAF-2 to antagonize AKT-1 . In parallel to AKT-1 , AKT-2 and SGK-1 also transmit input from DAF-2 to inhibit DAF-16 . We asked whether AKT-2 or SGK-1 counteract AKT-1 to control gonadal integrity . Both akt-2 and sgk-1 RNAi clone strongly suppressed disruption of the gonad in shc-1 ( ok198 ) ;akt-1 ( ok525 ) mutant ( Figure 4E and Table 2 ) , indicating that AKT-2 and SGK-1 act downstream of DAF-2 to antagonize AKT-1 . Next we asked whether DAF-2 , AKT-1/2 or SGK-1 also promote tumorous germline proliferation in shc-1;Is[daf-16::gfp] animals . We quantified number of the germ cells of shc-1;Is[daf-16::gfp] L3 larvae in daf-2 , akt-1/akt-2 or sgk-1 mutant background . Knock-down of sgk-1 significantly reduced number of the germ cells ( Figure 4F and Figure S10 ) , suggesting SGK-1 contributes to proliferation of the germline tumor in shc-1;Is[daf-16::gfp] animals . Due to somatic missexpression of PGL-1 in daf-2 and akt-1/akt-2 animals [35] , we were not able to quantify germ cells in these mutants ( data not shown ) . SHC-1 has been shown to inhibit IIS to activate the downstream PI3K signaling [26] . Therefore , one may speculate that active DAF-16 in the presence of hyperactive PI3K signaling is the cause for gonad disruption . The PTEN homolog DAF-18 antagonizes PI3 kinase AGE-1 . Therefore daf-18 mutant should have enhanced PI3K signaling . To test whether active PI3K signaling contributes to the DAF-16 dependent phenotype , we analyzed daf-18 ( e1375 ) ;akt-1 ( ok525 ) and shc-1 ( ok198 ) ;daf-18 ( e1375 ) ;akt-1 ( ok525 ) animals . daf-18 mutation did not enhance gonad disruption of akt-1 ( ok525 ) one day adult animals ( Figure 5A , Table 2 ) , suggesting that active PI3K is not sufficient for DAF-16 to degenerate the basement membrane . However , daf-18 enhanced the defects in shc-1 ( ok198 ) ;akt-1 ( ok525 ) animals up to almost 100% and daf-16 mutation completely suppressed the defect in shc-1 ( ok198 ) ;daf-18 ( e1375 ) ;akt-1 ( ok525 ) animals ( Figure 5A and Table 2 ) , indicating that active IIR/PI3K signaling assists DAF-16 to degenerate the gonadal basement membrane . Taking together , DAF-2 mediated PI3K signaling is necessary but not sufficient for DAF-16 to cause disruption of the gonad . Besides negatively modulating DAF-2 , SHC-1 also activates MEK-1/JNK-1 to affect DAF-16 [26] . Upon heavy metal stress response , SHC-1 also mediates the activation of MEK-1 , which in turn phosphorylates and activates another JNK homolog , KGB-1 [27] . Therefore , in the shc-1 mutant , hyperactive IIS is accomplished by inactive JNK signaling . Since hyperactive IIS is necessary but not sufficient for DAF-16 to derogate the gonadal basement membrane , we asked whether inactivated JNK signaling also plays a role . We generated mek-1 ( ks54 ) ;Is[daf-16::gfp] , jnk-1 ( gk7 ) Is[daf-16::gfp] and kgb-1 ( um3 ) Is[daf-16::gfp] animals . Inactivation of JNK-1 did not further enhance disruption of the gonadal basement membrane in Is[daf-16::gfp] animals ( Figure 5B , Table 2 ) . 85% of mek-1 ( ks54 ) ;Is[daf-16::gfp] animals died at early larval stages . The remaining animals could develop to adulthood . However , 95 . 6% of these animals displayed germ cells outside of the gonad ( Figure 5B and Table 2 ) , phenocopying the shc-1;Is[daf-16::gfp] phenotype . Even though mek-1;Is[daf-16::gfp] animals showed more severe defects in early larval development than shc-1 ( ok198 ) ;Is[[daf-16::gfp] animals , fewer animals were sterile ( 29% vs . 43%; Table S2 ) and the fertile adult animals displayed a higher number of progeny ( 120 vs . 49; Table S2 ) , suggesting that the early larval lethality is not due to a more severe defect in the basement membrane . All kgb-1 ( um3 ) Is[daf-16::gfp] animals died at larval stage . However , most of Is[daf-16::gfp] animals fed with kgb-1 RNAi could bypass the larval lethality and in 75% of these one day old adult animals germ cells in the body cavity were observed ( Figure 5B and Table 2 ) . These results suggest that the antagonistic function of SHC-1 to DAF-16 is mediated via MEK-1 and KGB-1 . The promoter used for the expression of daf-16 in TJ356 Is[daf-16::gfp] is a 6 kb genomic region upstream of the exon 1 of daf-16a isoform , which triggers expression in the hypodermis , intestine , neurons and body wall muscle cells [29] , [36] . Due to a general germline silencing of transgenic promoters in C . elegans , this daf-16::gfp transgene is probably not expressed in the germline . Endogenous DAF-16 in the germline or somatic gonad is not required for this phenotype , since no phenotypic difference of shc-1;Is[daf-16::gfp] animals was observed in daf-16 ( + ) and daf-16 ( mu86 ) null mutant backgrounds ( Table 2 ) . We conclude that DAF-16 outside of the reproductive system is the most likely cause of the phenotype we describe here . In order to test how DAF-16 in different tissues promotes the phenotype and rule out an involvement of weak expression of the daf-16 transgene in the germline , we expressed the transgene in the specific somatic tissues . Since expressing the constitutively nuclear daf-16 ( 4A ) ::gfp resulted in a stronger phenotype compared to wild type daf-16::gfp , we used daf-16 ( 4A ) for this analysis . Transgenic expression of daf-16 ( 4A ) ::gfp in the neurons , intestine , or musculature did not increase the weak shc-1 ( − ) phenotype ( Figure 5C and Table 3 ) . In contrast , expression of daf-16 ( 4A ) ::gfp in the hypodermis was sufficient to cause disruption of the gonad in 80% of one day adult shc-1 ( − ) animals , suggesting that hypodermal DAF-16 causes disruption of the gonadal basement membrane cell-nonautonomously .
In this manuscript we for the first time describe two qualities of DAF-16 function that affect longevity in opposite way . In contrast to the known lifespan extending effect , we discover that DAF-16 can shorten lifespan by inducing a tumor-like germline phenotype . It has probably affected the outcome of previously described experiments , however , to our knowledge this has not been reported before [17]–[36] . The starting point of our experiments was that we , as others before , failed to increase the lifespan of wild type ( daf-2 +/+ ) animals by expressing transgenic copies of daf-16 , unless the lifespan assays were performed in the presence of FUDR [37] ( Figure 1B ) . In this respect daf-16 differs from other key regulators of stress response and lifespan such as skn-1 , which typically shorten lifespan when inactivated , and increase lifespan when being overexpressed as a transgene . Although daf-16 transgene rescued the short life-span of daf-2;daf-16 double mutants , it did not or only modestly increases lifespan in wild type background [17]–[36] . It has been suggested that in a daf-2 ( + ) background , AKT phosphorylation results in cytoplasmic retention of these extra copies of DAF-16 , rendering them inactive [36] . However , phosphorylation was not sufficient to prevent the ability of transgenic SKN-1 to extend lifespan [38] . We found here that the lifespan extending effect of DAF-16 was balanced by its lifespan shortening effect which induced tumor-like growth in the germline . Consistently , blocking germline proliferation by using the cytostatic FUDR or a mutation inhibiting Notch signaling in the germline was sufficient to prevent negative aspects of DAF-16 signaling and increased lifespan of wild type animals containing a daf-16 transgene ( Figure 1B and 1C ) . We found that not only a daf-16 transgene , but also a combination of mutants in the IIS pathway affecting DAF-16 activity caused a pleiotropic phenotype in the reproductive system . This phenotype occurred at low penetrance in wild type animals carrying the daf-16 transgene , which may be the reason why it had escaped detection in previous studies . The penetrance of this phenotype was strongly enhanced in shc-1 mutant background . Its most prominent phenotypic aspect was the disruption of the gonadal basement membrane at the proximal gonad adjacent to the developing somatic gonad primordium , so that eventually germline cells leaked into the body . We counted about 30% more germ cells in disrupted gonad arms of shc-1;Is[daf-16::gfp] L3 larvae compared to wild type animals ( Figure 2K ) . A co-segregation of both phenotypic aspects was particularly obvious in animals in which only one gonad arm was disrupted , and typically contained significantly more germ cells than the other intact gonad arm . This phenotype is distinct from that of gld-1 mutants , in which the basement membrane ruptures in adult animals due to a massive excess of germ cells . gld-1 mutants at the equivalent L3 stage still have intact gonad arms despite their increased number of germ cells [39] . This indicates that in gld-1 mutants , but not in animals described in this study , the basement membrane of L3 animals is mechanically strong enough to withstand an increase in germline nuclei . It is therefore possible that the basement membrane of the gonad in shc-1;Is[daf-16] has a defect on its own , similar as e . g . seen in ten-1 , dgn-1 , ina-1 and epi-1 mutants , which also show disruption of gonadal basement membrane , but without an increase in germline proliferation [40] . An interesting hypothesis is that that the abnormally proliferating germ cells invade into the surrounding extracellular matrix . Although we observed germ cells that penetrate the basement membrane , and dissection and extraction of an intact gonad in those animals failed due to their leakage , we cannot prove at this point that this is an active process . Nevertheless , we discover for the first time that mutations in IIS result in both germline hyperplasia and disruption of the gonadal basement membrane in C . elegans , a phenotype that opposes the beneficial effects of DAF-16 on increasing healthspan . The phenotype in the reproductive system was also observed in shc-1 mutants carrying extrachromosomal arrays with daf-16 . Such arrays in C . elegans are typically not expressed in the germline [41] . In agreement with this generally accepted notion , we never observed germline expression of the GFP tagged daf-16 transgenes . One possibility is that overexpression of DAF-16 in the somatic tissues leads to activation of germline DAF-16 , since it has been shown that active DAF-16 in one tissue elevated its activity in other tissues [42] . However , we observed no difference between shc-1;Is[daf-16::gfp] and shc-1 daf-16;Is[daf-16::gfp] animals , the latter lacking endogenous daf-16 in the germline [19] ( Table 2 ) . Therefore , we consider it more likely that daf-16 expression outside the reproductive system is the cause of the observed phenotype . The fact that expressing daf-16 in the hypodermis was sufficient to provoke this phenotype ( Figure 5C ) further excluded an involvement of very low levels of transgenic DAF-16 in the germline . These data suggest that DAF-16 mediates a cell-nonautonomous signaling to affect germline proliferation and basement membrane integrity . Mutation in shc-1 strongly enhanced the germline and gonad phenotype of Is[daf-16::gfp] animals . In addition , we observed that DAF-16 ( 4A ) was nuclearly localized in both shc-1 ( + ) and shc-1 ( − ) animals ( Figure 3D ) , while disruption of the gonad was only detectable in shc-1 ( − ) background ( Figure 3C ) . These data favor a model that SHC-1 , in addition to its role in affecting the subcellular localization of DAF-16 upon stress response , can also antagonize nuclear DAF-16 . Notably , SHC-1 is localized in both cytoplasm and nucleus [26] . Our data indicate that both active IIS and inactive JNK signaling contribute to the DAF-16 dependent phenotype in the reproductive system ( Figure 4B , 4D , 5A and 5B ) . This may also explain why the mek-1;Is[daf-16::gfp] strain displayed a less severe basement membrane phenotype than shc-1;Is[daf-16::gfp] animals , as only SHC-1 affects both IIS and JNK signaling [26] . Our data also explain previous , contradictory data concerning the lifespan of mek-1 mutants . We reported previously that mek-1 mutants were short lived , while Oh and colleagues found mek-1 mutants to have a lifespan indistinguishable from wild type [9] , [26] . We found that both mek-1 and shc-1 mutant animals lived as long as wild type worms in the presence of the drug FUDR that blocked germline proliferation ( Table 2 ) . However , in the absence of FUDR , mek-1 animals suffered from a pronounced reduction of lifespan that we can now attribute to defects in the reproductive system , rather than to the sensitivity for reactive oxygen species stress as proposed before [26] . We found that in a daf-2 loss-of-function mutant the DAF-16 mediated gonad dysintegrity phenotype of both shc-1;Is[daf-16::gfp] and shc-1;akt-1 animals was reduced ( Figure 4B and 4D ) , thus abrogating lethality in early adulthood ( Figure 4C ) . This was a surprising result , since our experiments so far had indicated that this gonad phenotype was a consequence of too much DAF-16 activity , and the daf-2 mutation should further increase DAF-16 activity . The PTEN homolog DAF-18 antagonizes the DAF-2 and PI3 kinase/AGE-1 pathway , thus a loss-of-function mutant is supposed to result in reduced DAF-16 activity via increased IIS . However , the daf-18 loss-of-function mutation further enhanced the phenotype of a shc-1;akt-1 mutant ( Figure 5A ) . In summary , mutations in both IIS pathway genes daf-2 and daf-18 behave opposite to their known roles in the canonical IIS pathway . Similarly , while the increase of the phenotype in an akt-1 loss-of-function mutant correlated well with the increase of DAF-16 activity , akt-2 and sgk-1 mutations opposed akt-1 , and behaved like daf-2 described above . We conclude that , while daf-2 , akt-1 , akt-2 , and sgk-1 loss-of-function mutants all result in a longevity and stress resistance phenotype consistent with upregulation of DAF-16 nuclear activity , they differ in the way they regulate DAF-16 activity to cause the gonad phenotype . One possible explanation for the antagonistic effect of distinct IIs pathway mutants is that daf-2 , akt-2 , and sgk-1 may provoke a distinct phosphorylation pattern of DAF-16 compared to akt-1 , resulting in different outputs of DAF-16 transcriptional targets that affect germline proliferation/gonadal integrity compared to longevity and stress resistance . Whereas the akt-1 ( ok525 ) mutation might only prevent AKT-1 mediated DAF-16 phosphorylation , daf-2 ( e1370 ) supposedly reduces both AKT-1 , AKT-2 , and SGK-1 phosphorylation , and this may result in controlling distinct downstream genes for long lifespan and gonad integrity , respectively . We consider this model less likely , since there is currently no evidence for C . elegans AKT-1 phosphorylating distinct sites in DAF-16 compared to SGK-1 or AKT-2 [7] . However , we cannot exclude this possibility based on the existing data . Another possible scenario is that mutation in daf-2 , akt-2 , and sgk-1 mediated activation of DAF-16 affects different tissues than akt-1 knock-down mediated activation of DAF-16 , suggesting that the germline reads out and responds to a balance of DAF-16 activities in somatic tissues . This way , DAF-2 could prevent DAF-16 from counteracting hypodermal DAF-16 in some tissues , e . g . the intestine . It has been previously shown that DAF-2 promotes larval germline proliferation via inactivating germline DAF-16 [19] . Together with our results this suggests that DAF-16 activities in the hypodermis and the germline may antagonize each other . Moreover , previous studies suggested that AKT-1 , AKT-2 and SGK-1 have different tissue specificity , expression levels , patterns , or activities to control DAF-16 [4] , [7] . Notably , AKT-1 is expressed in the hypodermis , whereas AKT-2 and SGK-1 are probably not . Consistently , neither down-regulation of akt-2 nor sgk-1 caused disruption of the gonad in shc-1 mutant background ( Figure 4A ) , but instead were both able to suppress the phenotype of akt-1;shc-1 mutants ( Figure 4E ) . Based on our results presented here , we suggest that AKT-1 is active in the hypodermis to prevent DAF-16 mediated signal to the reproductive system , whereas AKT-2 , SGK-1 and even AKT-1 in other tissues may inhibit DAF-16 activity that counteracts hypodermal DAF-16 . Such antagonistic roles of SGK-1 and AKT-1 have been reported in regulation of lifespan [43] . Additional tissue specific studies of these three AGC family kinases will enable a further understanding of such interactions across tissue boundaries . It has been shown in different organisms that FOXO/DAF-16 functions as a tumor suppressor in a variety of cancers by promoting apoptosis or cell cycle arrest [44] . However , several recent data indicate that the function of FOXO/DAF-16 may be more complex than previously thought . An essential role of FOXO3a was proposed in the maintenance of cancer stem cells that are responsible for the reoccurrence of chronic myeloid leukemia [15] . In acute myeloid leukemia FOXO1/3/4 promote leukemic growth and maintenance by inhibiting myeloid maturation and apoptosis [45] . In C elegans , some DAF-16 target genes were identified which stimulate gld-1 germline tumor [46] . All these studies indicate that the function of FOXO/DAF-16 proteins is highly dependent on the cellular context . In mouse , FOXO3a promotes tumor cell invasion through the induction of matrix metalloproteases [47] . Our data show that FOXO/DAF-16 can also promote germline hyperplasia and disruption of the surrounding extracellular matrix through cell-nonautonomous signaling . It will be of great interest to further identify the signaling molecules to the germline activated by DAF-16 , and to investigate whether the gonad dysintegrity phenotype we described involves active invasions of the germ cells into the extracellular matrix and , therefore , resembles the behavior of metastatic tumor cells .
To generate transgenic animals carrying wild type daf-16::gfp or daf-16 ( 4A ) ::gfp , the corresponding constructs were injected into wild type or shc-1 ( ok198 ) animals ( for byEx numbers , see Table 3 ) . 20 ng/µl pRF4 rol-6 ( su1006 ) was used as co-injection marker . GFP expression of the transgenic animals was confirmed using fluorescent microscopy . A formaldehyde fixation procedure was used ( Text S1 ) for the whole worm staining with 1∶200 anti-PGL-1 antibody ( a gift from Dr . Susan Strome ) . Staining of dissected gonads was not possible because of the disruption of the gonad . Basement membranes were stained with MitoTracker Red CMXRos ( invitrogen ) by placing worms in a solution of 10 µM MitoTracker Red at 25°C for 2 hours . The worms were then allowed to recover for 30 minutes on an NGM agar plate and analyzed . Lifespan assays were initiated at the L4 larval stage . Synchronized animals were raised at 15°C prior to lifespan analysis . Then , L4 animals were transferred to the respective temperatures for the assays and examined every day . Animals that showed no response to touch were scored as dead . In assays without FUDR treatment animals were transferred every day onto new plates during the reproductive period . Animals died because of bagging of larvae were censored . In the lifespan assay with FUDR treatment , animals were transferred onto the agar plates containing 0 . 1 mg/ml FUDR and E . coli 24 hours after the L4 larval stage . These were raised on the FUDR containing plates for four days and then transferred onto new plates without FUDR . FUDR was administrated to adult instead of L4 animals in order to exclude a possible interference from seam cell development to the reproductive system , since the epidermal seam cells undergo a final mitotic division at the L4 to adult molt and DAF-16 mediated signaling is sent from the hypodermis . All the lifespan assays were performed at 20°C with exception of those with glp-1 ( q231 ) and daf-2 ( e1370 ) mutants , as glp-1 ( q231 ) required 25°C to inactivate GLP-1 and daf-2 ( e1370 ) ;shc-1 ( ok198 ) ;Is[daf-16::gfp] could develop to adulthood only at 15°C . All of the lifespan assays were performed at least twice . The short lifespan of shc-1 ( ok198 ) ;Is[daf-16::gfp] can only be observed , when no censoring for adult animals dying due to the development defect is done . As reported previously by us ( Neumann-Haefelin et al . , 2008 ) , Is[daf-16::gfp] partially extends lifespan of shc-1 ( ok198 ) mutants , when adult animals dying due to the development defect are censored . All animals in the tests except those with daf-2 ( e1370 ) ;shc-1 ( ok198 ) ;Is[daf-16::gfp] and daf-2 ( e1370 ) ;shc-1 ( ok198 ) ;akt-1 ( ok525 ) , were raised at 20°C and analyzed via DIC microscopy 24 hours after L4 stage . Animals in which at least one germ cell was detected outside of the gonad were scored as positive . The percentage of positive one day old animals was calculated . Per test thirty animals were examined and each test was performed at least three times . shc-1 ( ok198 ) ;daf-2 ( e1370 ) ;Is[daf-16::gfp] and shc-1 ( ok198 ) ;daf-2 ( e1370 ) ;akt-1 ( ok525 ) animals were raised at 15°C and analyzed 24 hours after L4 larval stage . The numbers of anti-PGL-1 positive cells were quantified at the time point that vulva induction takes place . GraphPad Prism 4 . 0 software ( GraphPad Software Inc . , San Diego , USA ) was used to calculate mean value and to perform statistical analysis . | Previous studies have shown that DAF–16/FOXO transcription factor promotes longevity and stress resistance and inhibits tumor progression in the absence of insulin signaling . Here we show that active DAF-16 in the epidermis can shorten lifespan by promoting a tumorous germline phenotype . In contrast to the known inhibitory effect of insulin signaling upon DAF-16 , an active insulin and PI3K signaling are required for DAF-16–mediated signaling to the germline . In addition , AKT-1– and SHC-1–mediated JNK signaling antagonize AKT-2 and SGK-1 to affect the reproductive system . This is to our knowledge the first report about a detrimental effect of DAF-16 on lifespan . Furthermore it emphasizes that DAF-16 activity is highly dependent on the cellular context and communication between different tissues . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"systems",
"biology",
"model",
"organisms",
"genetics",
"biology",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] | 2012 | Cell-Nonautonomous Signaling of FOXO/DAF-16 to the Stem Cells of Caenorhabditis elegans |
Male circumcision reduces acquisition of HIV-1 by 60% . Hence , the foreskin is an HIV-1 entry portal during sexual transmission . We recently reported that efficient HIV-1 transmission occurs following 1 h of polarized exposure of the inner , but not outer , foreskin to HIV-1-infected cells , but not to cell-free virus . At this early time point , Langerhans cells ( LCs ) and T-cells within the inner foreskin epidermis are the first cells targeted by the virus . To gain in-depth insight into the molecular mechanisms governing inner foreskin HIV-1 entry , foreskin explants were inoculated with HIV-1-infeceted cells for 4 h . The chemokine/cytokine milieu secreted by the foreskin tissue , and resulting modifications in density and spatial distribution of T-cells and LCs , were then investigated . Our studies show that in the inner foreskin , inoculation with HIV-1-infected cells induces increased CCL5/RANTES ( 1 . 63-fold ) and decreased CCL20/MIP-3-alpha ( 0 . 62-fold ) secretion . Elevated CCL5/RANTES mediates recruitment of T-cells from the dermis into the epidermis , which is blocked by a neutralizing CCL5/RANTES Ab . In parallel , HIV-1-infected cells mediate a bi-phasic modification in the spatial distribution of epidermal LCs: attraction to the apical surface at 1 h , followed by migration back towards the basement membrane later on at 4 h , in correlation with reduced CCL20/MIP-3-alpha at this time point . T-cell recruitment fuels the continuous formation of LC-T-cell conjugates , permitting the transfer of HIV-1 captured by LCs . Together , these results reveal that HIV-1 induces a dynamic process of immune cells relocation in the inner foreskin that is associated with specific chemokines secretion , which favors efficient HIV-1 entry at this site .
According to an updated report on the global AIDS epidemic ( see www . unaids . org ) , 15 million men are currently infected with HIV-1 worldwide . HIV-1 infection in men has recently gained extensive scientific and public attention following reports of three clinical trials , which clearly demonstrated that male circumcision provides 60% protection from HIV-1 acquisition [1]–[3] . These reports confirmed a multitude of previous similar epidemiological observations [4] , and suggest altogether an important role of the male foreskin as an entry portal for HIV-1 . The male foreskin is a stratified epithelium , made of multiple layers of keratin-forming epithelial cells ( i . e . epidermis ) positioned on top of a connective tissue made of collagen-producing fibroblasts ( i . e . dermis ) [5] . The foreskin consists of two different aspects , outer and inner , which are easily distinguished by the relative decrease in melanocytes in the inner foreskin [6] . While some studies , including ours , have reported that the degree of keratinization of the outer foreskin is higher than that of the inner [7]–[10] , other studies reached opposite conclusions [11] or reported no difference in the degree of foreskin keratinization [12] . Hence , a standardized method to evaluate keratin thickness is required , in order to determine morphologically the difference in keratinization between the outer and inner foreskins , which may provide a protective barrier against HIV-1 entry [7] , [9] , [10] , [13] . Both foreskin epidermis and dermis also contain various immune cells , such as Langerhans cells ( LCs ) , T-cells , dendritic cells ( DCs ) and macrophages [6]–[8] , [10] , [14]–[16] . These immune cells may serve as potential targets for HIV-1 due to their expression of CD4/CCR5 , the principal receptors for HIV-1 [6]–[8] , [10] , [14] , [16] , [17] , as well as alternative HIV-1 attachment receptors , such as the C-type lectins langerin on LCs and Dendritic Cell-Specific Intercellular adhesion molecule-3-Grabbing Non-integrin ( DC-SIGN ) on DCs [10] , [15] , [18]–[22] . Two previous studies showed that the foreskin is susceptible to entry of high doses of cell-free HIV-1 at time points of >24 h [7] , [16] . However , until recently , the exact mechanisms responsible for HIV-1 entry into the foreskin , especially during the very first hours following exposure of the foreskin to HIV-1 , remained to a large extent unknown . To describe the initial events of HIV-1 foreskin entry , we recently developed two novel and complementary models of the adult human foreskin epithelium , namely ex-vivo polarized inner and outer foreskin explants , and immuno-competent stratified foreskin epithelia reconstructed in-vitro from isolated primary inner or outer foreskin cells [10] . In these models , efficient HIV-1 transmission occurs following 1 h of polarized exposure of the inner , but not outer , foreskin to mononuclear cells highly infected with HIV-1 , but not to cell-free virus [10] . Hence , within the inner foreskin , in a first step , HIV-1-infected cells form viral synapses with apical foreskin epithelial cells , leading to polarized budding of HIV-1; In a second step , HIV-1 is rapidly internalized by epidermal LCs and remains intact within their cytoplasm; In a third step , LCs transfer HIV-1 to T-cells within the foreskin epidermis , across conjugates formed between these two cell types [10] . Together , these studies identified the foreskin region permissive to HIV-1 ( i . e . the inner foreskin ) , the conditions allowing for efficient viral entry ( i . e . virus originating from infected cells ) , and the initial cells targeted by HIV-1 ( i . e . LCs and T-cells ) . In the present study , we now explore the molecular mechanisms responsible for potent inner foreskin HIV-1 entry . Thus , our recently developed polarized foreskin explants were inoculated with HIV-1-infeceted or non-infected cells for 4 h . Following viral exposure , the chemokine/cytokine milieu secreted by the foreskin tissue was examined , as well as any related changes in the density and spatial distribution of T-cells and LCs . The involvement of selective chemokines in these processes was confirmed using relevant blocking Abs . The results presented herein demonstrate specific regulation of chemokines secretion within the inner foreskin , correlating with a dynamic process of HIV-1 target cell migration , which may govern the efficient entry of HIV-1 at this site .
T-cells and LCs are the first cells targeted by HIV-1 during early viral exposure of the inner foreskin [10] . The motility of these migratory cells dependents on environmental cues , such as chemokines and cytokines secreted by the tissue . To first screen for potential chemokines and cytokines secreted by the foreskin , inner foreskin explants were inoculated in a polarized manner for 4 h at 37°c with either non-infected or HIV-1-infected peripheral blood mononuclear cells ( PBMCs ) . Explants were then washed and further incubated in culture medium in a non-polarized manner overnight at 37°c , to permit the secretion of high/detectable chemokines/cytokines levels . Quantitative multiplex bead immunoassay and flow cytometry or specific ELISA was then used to measure the levels of twelve chemokines/cytokines . Exposure of inner foreskin explants to non-infected PBMCs ( serving as negative control ) resulted in different levels of chemokines/cytokines secretion: no secretion of IL-23 , IL-17A , IL-10 and IL-1 beta; low levels of IFN gamma ( <20 pg/ml ) ; moderate levels of CCL3/MIP-1 alpha , CCL20/MIP-3 alpha , and CCL5/RANTES ( 100–1000 pg/ml ) ; and high levels of CCL2/MCP-1 , CXCL8/IL-8 , IL-6 and CXCL10/IP-10 ( 2500–18500 pg/ml ) . Next , to identify chemokines/cytokines whose secretion was specifically affected by the interaction of HIV-1 with the tissue , the measured levels following inoculation with HIV-1-infected PBMCs were normalized to that following inoculation with non-infected PBMCs . This enables the calculation of folds increase/decrease secretion . Among the chemokines/cytokines secreted by the inner foreskin , the levels of the highly secreted ones ( i . e . CCL2/MCP-1 , CXCL8/IL-8 , IL-6 and CXCL10/IP-10 ) were not changed by the presence of HIV-1 ( Fig . 1A ) . In contrast , following inoculation with HIV-1-infected PBMCs , the chemokines/cytokines secreted at low/moderate levels were either decreased , including IFN gamma , CCL3/MIP-1 alpha and CCL20/MIP-3 alpha ( Fig . 1B , grey bars; mean actual values were 13 . 8 , 590 . 5 and 443 . 0 pg/ml following inoculation with non-infected PBMCs and 4 . 0 , 310 . 1 and 276 . 9 pg/ml following inoculation with HIV-1-infected PBMCs , respectively; n = 2 ) , or increased , including CCL5/RANTES ( Fig . 1B , black bar; mean actual values were 580 . 3 pg/ml following inoculation with non-infected PBMCs and 953 . 6 pg/ml following inoculation with HIV-1-infected PBMCs , respectively; n = 2 ) . Of note , in outer foreskin explants ( from the same individuals and tested in parallel experiments ) , CCL5/RANTES was also secreted at moderate levels , but remained unchanged following exposure to HIV-1-infected PBMCs ( mean actual values were 291 . 5 pg/ml following inoculation with non-infected PBMCs and 354 . 3 pg/ml following inoculation with HIV-1-infected PBMCs , respectively; n = 2 , p = 0 . 3451 , Student's t-test ) . To further examine whether these alterations in chemokines secretion were already detectable at the end of the infection period , inner foreskin explants were inoculated in a polarized manner for 4 h at 37°c with either non-infected or HIV-1-infected PBMCs , and the apical and basal supernatant fractions were collected immediately . A second array consisting of glass slides spotted with capture Abs to either CCL5/RANTES or CCL20/MIP-3 alpha , as well as specific CCL20/MIP-3 alpha ELISA , were then used to evaluate the early-secreted levels of both chemokines . The obtained values ( i . e . arbitrary relative fluorescent intensities using slides or pg/ml using ELISA ) were translated to folds increase/decrease secretion . Although the secretion levels of both chemokines at this early time point were lower and variable among foreskin tissues from different individuals , inoculation with HIV-1-infected PBMCs modified the secretion of CCL5/RANTES and CCL20/MIP-3 alpha in inner foreskin explants in a similar fashion to that observed at later time points . Hence , compared to non-infected PBMCs , exposure to HIV-1-infected PBMCs resulted in an increase in early CCL5/RANTES that was secreted only into the apical supernatant fraction ( mean fold increase±SEM of 1 . 397±0 . 045 from n = 3 explants; p = 0 . 0419 , Student's t-test ) . In parallel , exposure to HIV-1-infected PBMCs resulted in a decrease in early CCL20/MIP-3 alpha that was secreted both apically and basally ( mean folds increase±SEM of 0 . 521±0 . 183 apical and 0 . 558±0 . 163 basal from n = 3 explants; p = 0 . 0345 and 0 . 0411 , respectively , Student's t-test ) . Finally , immunohistochemical staining of inner foreskin explants for CCL5/RANTES expression showed that this chemokine was expressed by foreskin keratinocytes . Unfortunately , despite using several CCL5/RANTES Abs , the detected signals were low and similar in all conditions , thus not permitting for quantitative evaluation of the differences in CCL5/RANTES levels following exposure to either non-infected or HIV-1-infected PBMCs . Immunohistochemistry appears therefore not sensitive enough to evaluate these differences , which were in contrast observed by the more sensitive and quantitative methods we used ( i . e . ELISA and array slide ) at this early time point . These results show that HIV-1-infected PBMCs affect rapidly the secretion of specific chemokines by the inner foreskin that persists at later time points following viral exposure . CCL5/RANTES exerts chemotactic activity on T-cells [23] , [24] . In the foreskin , the majority of T-cells are located within the dermis , and a minority of T-cells are integrated within the epidermis [6]–[8] , [10] , [14]–[16] . As secretion of CCL5/RANTES increased in the inner foreskin epidermis upon early exposure to HIV-1-infected PBMCs ( see above ) , we next investigated possible changes in T-cell density in inner foreskin following viral exposure . Inner and outer foreskin explants were inoculated comparatively with HIV-1-infected or non-infected PBMCs , and 4 h later the explants were fixed , processed for immunohistochemistry , and stained for CD3 expression . The number of the positively stained CD3+ T-cells was then quantified in both epidermis and dermis . Following 4 h inoculation with HIV-1-infected PBMCs , the density of epidermal T-cells in the inner foreskin significantly doubled ( from 177±31 to 370±58; mean±SEM CD3+ cells/mm2 epidermis from n = 3 experiments; p = 0 . 0423 , Student's t-test; Fig . 2A ) . In contrast , inoculation with non-infected PBMCs did not significantly change the density of epidermal T-cells ( from 151±23 to 199±34; mean±SEM CD3+ cells/mm2 epidermis from n = 3 experiments; p = 0 . 3050 , Student's t-test; Fig . 2A ) . The difference between the densities of epidermal T-cells following exposure to HIV-1-infected compared to non-infected PBMCs at 4 h was significant ( p = 0 . 0316 , Student's t-test ) . The increase in T-cell density in the inner foreskin epidermis following exposure to HIV-1-infected PBMCs was correlated with a significant decrease in their density within the inner foreskin dermis ( from 1201±140 to 582±58; mean±SEM CD3+ cells/mm2 dermis from n = 3 experiments; p = 0 . 0175 , Student's t-test; Fig . 2A ) . Although the density of dermal T-cells decreased following inoculation with non-infected PBMCs , the changes were not statistically significant ( p = 0 . 2030 , Student's t-test; Fig . 2A ) . The difference between the densities of dermal T-cells following exposure to HIV-1-infected compared to non-infected PBMCs at 4 h was significant ( p = 0 . 0269 , Student's t-test ) . T-cell recruitment described herein is not a result of tissue entry of the input PBMCs . Indeed , fluorescently loaded HIV-1-infected input cells were never detected within the epithelium or the basal compartment after their inoculation at the apical surface ( data not shown ) . To gain further insight as to the identity of T-cells that entered the epidermis , we performed additional double immunohistochemical staining experiments of inner foreskin explants and examined CD4 and CD8 expression on CD3+ T-cells . At 4 h , the ratio of epidermal CD8 over CD4 T-cells was 8 . 9±1 . 3 following inoculation with non-infected PBMCs . Following exposure to HIV-1-infected PBMCs , the epidermal densities of both CD4+ and CD8+ cells increased , but the CD8 over CD4 ratio decreased to 3 . 7±0 . 6 ( p = 0 . 0114 , Student's t-test , n = 3 ) , suggesting that HIV-1-infected PBMCs induced preferential recruitment of CD4+ cells into the epidermis . In agreement with these findings , the above-mentioned changes corresponded with migration of T-cells from the dermal into the epidermal compartments across a disrupted basement membrane , as observed at the ultra-structural level in inner foreskin explants following 4 h exposed to HIV-1-infected PBMCs ( Fig . 2C ) . In contrast , no statistically significant changes were observed in the density of epidermal/dermal T-cells in parallel outer foreskin explants exposed to either HIV-1-infected or non-infected PBMCs ( p = 0 . 3873 , 0 . 1333 , 0 . 2355 and 0 . 2142 following exposure to HIV-1-infected or non-infected PBMCs for either epidermis or dermis , respectively; Student's t-test; Fig . 2B ) . T-cells translocation from the dermis into the epidermis was not observed at the ultra-structural level in outer foreskin explants following 4 h exposure to HIV-1-infected PBMCs . These results suggest that HIV-1 originating from infected cells recruits T-cells from the dermis into the epidermis in inner , but not outer , foreskin . To investigate whether the increase in CCL5/RANTES secretion by the inner foreskin epidermis , detected after 4 h viral exposure , is responsible for the observed T-cell recruitment ( Fig . 2 ) , a neutralizing Ab to CCL5/RANTES was used . Hence , inner foreskin explants from additional individuals were exposed for 4 h at 37°c to HIV-1-infected PBMCs either: 1 ) alone; 2 ) in the presence of a control isotype Ab; 3 ) or in the presence of a CCL5/RANTES neutralizing Ab . Exposure to medium or non-infected PBMCs served as controls . Following infection , epidermal single-cell suspensions were prepared , stained for CD3 expression and examined by flow cytometry . In agreement with our CD3+ cell counting by immunohistochemistry ( Fig . 2A ) , inoculation with HIV-1-infected PBMCs resulted in a higher percentage of CD3+ cells in cell suspensions of inner foreskin epidermis , compared to inoculation with either medium of non-infected PBMCs ( Fig . 3A , B ) . Pre-treatment of inner foreskin explants with a control isotype Ab did not affect the observed increase in CD3+ cells upon exposure to HIV-1-infected cells ( Fig . 3B , dark grey bar ) . In sharp contrast , a neutralizing Ab to CCR5/RANTES completely abrogated T-cell recruitment into the epidermis ( Fig . 3B , light grey bar , 80 µg/ml Ab ) . This set of results demonstrates that the recruitment of T-cells into the inner foreskin epidermis induced by HIV-1-infected cells depends on CCL5/RANTES . LCs play a major role during the initial hours of HIV-1 exposure in both the male and female genital tracts , by sampling the mucosal surface and rapidly internalizing HIV-1 that remains intact within their cytoplasm [10] , [25] . In the foreskin , LCs are located within the epidermis but not dermis [6]–[8] , [10] , [14] , [15] . To study possible changes in LC density mediated by HIV-1 , inner and outer foreskin explants were inoculated for 4 h with either HIV-1-infected or non-infected PBMCs , stained for langerin , and the numbers of langerin+ LCs were counted . In the inner foreskin epidermis , LC density did not change upon exposure to HIV-1-infected cells ( from 581±124 to 554±75; mean±SEM langerin+ cells/mm2 epidermis from n = 3 experiments; p = 0 . 4305 , Student's t-test; Fig . 4A ) . In contrast , inoculation with non-infected PBMCs resulted in a significant decrease in LC density ( from 597±87 to 327±64; mean±SEM langerin+ cells/mm2 epidermis from n = 3 experiments; p = 0 . 0322 , Student's t-test; Fig . 4A ) , in correlation with detection of langerin+ cells in the dermis at this time point ( see Fig . 5 below ) . The difference between the densities of epidermal LCs following exposure to HIV-1-infected compared to non-infected PBMCs at 4 h was significant ( p = 0 . 0415 , Student's t-test ) . No statistically significant changes were observed in the density of langerin+ cells in parallel outer foreskin explants exposed to either HIV-1-infected or non-infected PBMCs ( p = 0 . 1578 and 0 . 1274 , respectively , Student's t-test; Fig . 4B ) . To investigate the co-localization of HIV-1 virions with LCs , inner and outer foreskin explants were exposed for 4 h to HIV-1-infected PBMCs , double stained for langerin and HIV-1 , and examined by confocal microscopy . The specificity of the signals recorded was based on parametric settings using specific isotype controls for either langerin or HIV-1 [10] . In inner foreskin , langerin was detected around cell bodies , as well as in LC dendrites reaching the apical surface ( Fig . 4C , red arrowheads ) . In these explants , HIV-1 was detected as free virions or associated with epithelial cells within the epidermis ( Fig . 4C , green arrowheads ) . Notably , HIV-1 was also found co-localized with LCs ( Fig . 4C , green arrows ) including virions that were internalized into LCs ( Fig . 4C , higher magnification insert ) , and was further detected within the dermis . In contrast , in outer foreskin , only few virions were detected within the epidermis ( Fig . 4D ) , in agreement with our previous study [10] . These results suggest that HIV-1 originating from infected cells enters the inner foreskin , and is captured by LCs retained in the epidermis . In contrast , in outer foreskin , HIV-1 fails to enter the epidermis or affect LC density . To follow the localization of LCs following HIV-1 exposure , inner foreskin explants were exposed for 1 h or 4 h to either HIV-1-infected or non-infected PBMCs , stained for langerin , and the distances of each langerin+ cell from the apical surface ( irrespective of their HIV-1 content ) were measured . Similar distances were also measured in normal inner foreskin , serving as the set point for the spatial distribution of LCs . Inoculation with non-infected PBMCs resulted in a time-dependent increase in the mean distance of LCs from the apical surface compared to normal inner foreskin ( Fig . 5A ) . Thus , at 1 h LCs were localized closer to the basement membrane and at 4 h LCs crossed the epidermis/dermis interface and reached the dermis , localizing further away from the apical surface . In contrast , exposure to HIV-1-infected PBMCs resulted in a bi-phasic modification in LC spatial distribution , compared to normal foreskin ( Fig . 5A ) : after 1 h of viral exposure , the mean distance of LCs from the apical surface shortened; at 4 hr these cells were localized near the basement membrane , and fewer LCs reached the dermis compared to exposure with non-infected PBMCs . Based on the above ‘snap-shot’ images and to calculate the total distance covered by LCs during the 4 h infection period , the following mean distances from the apical surface were designated: [a] for normal tissue ( set point distance ) ; [b] and [c] after 1 h and 4 h inoculation with non-infected PBMCs , respectively; [d] and [e] after 1 h and 4 h inoculation with HIV-1-infected PBMCs , respectively ( Fig . 5A ) . The mean distance covered by LCs following 1 h exposure to non-infected PBMCs is therefore represented by [b] minus [a] , as LCs were localized away from the apical surface . In contrast , the mean distance covered by LCs following 1 h exposure to HIV-1-infected PBMCs is represented by [a]-[d] , as LCs were localized closer to the apical surface . At 4 h , the additional distances covered by LCs are represented as [c]-[b] and [e]-[d] , as LCs were localized away from the apical surface following exposure to either non-infected or HIV-1-infected PBMCs . This analysis ( Fig . 5B , C ) demonstrated that at 1 h post-infection , LCs covered a similar distance , but in opposite directions ( 7 . 259±0 . 606 µm basally and 5 . 642±1 . 337 µm apically after exposure to either non-infected or HIV-1-infected PBMCs , respectively ) . Yet , LCs covered a shorter additional distance at 4 h after exposure to non-infected , compared to HIV-1-infected PBMCs ( 10 . 677±1 . 218 µm and 21 . 651±0 . 749 µm for non-infected or HIV-1-infected PBMCs , respectively; p<0 . 0001 , Student's t-test; Fig . 5C ) . These findings suggest that LCs may directly exit the epidermis following inoculation with non-infected cells . However , HIV-1 originating from infected cells induces first the attraction of LCs to the apical surface at 1 h , followed by their entry deeper into the tissue towards the epidermis-dermis interface at 4 h . Following the rapid internalization of HIV-1 by LCs in the inner foreskin at 1 h , intact virions that are sequestered within LCs are transferred to epidermal T-cells across conjugates forming between LCs and T-cells [10] . The number of such LC-T-cell conjugates was further evaluated at 4 h , in inner foreskin explants inoculated with either HIV-1-infected or non-infected PBMCs and double stained for langerin and CD3 . The percentage of langerin+/CD3+ conjugates out of the total number of langerin+ cells was then calculated . Of note , conjugates were defined as those in which LCs and T-cells were in direct contact with a visible interface and common segment of plasma membrane , while excluding those in which LCs and T-cells were only in proximity with a visible space between the two stained cells . At 1 h , 4 . 5±1 . 1% LCs were conjugated with T-cells ( Fig . 6A ) . The proportion of conjugates continued to increase and almost doubled at 4 h , with 7 . 6±2 . 1% of LCs forming conjugates with T-cells ( p = 0 . 0431 , Student's t-test , Fig . 6A ) . These LC-T-cell conjugates were located within the epidermis just above the basement membrane ( Fig . 4C ) , while conjugates were only occasionally detected following exposure to non-infected PBMCs ( Fig . 6B ) . Higher magnification images revealed LCs to be tightly associated with T-cells and HIV-1 virions ( Fig . 6D , E ) . In agreement with our conjugate counting by immunohistochemistry ( see above ) , inoculation with HIV-1-infected PBMCs resulted in a higher percentage of langerin+/CD3+/high forward scatter conjugates in cell suspensions of inner foreskin epidermis , compared to inoculation with non-infected PBMCs ( Fig . 6F ) . Pre-treatment of inner foreskin explants with a control isotype Ab did not affect the observed increase in conjugate formation upon exposure to HIV-1-infected cells . In sharp contrast , a neutralizing Ab to CCR5/RANTES inhibited the formation of LC-T-cell conjugates induced by HIV-1-infected PBMCs ( Fig . 6F ) . These results suggest that LC-T-cell conjugate formation within the inner foreskin tissue following exposure to HIV-1-infected cells is a sustained process during the early hours following exposure to HIV-1 .
HIV-1 is a viral menace that gains access into the body mainly during sexual intercourse , by crossing epithelial barriers that cover the mucosal surfaces of the gastrointestinal , female and male genital tracts . In stratified epithelia , HIV-1 ‘highjacks’ the physiological process of pathogen recognition by LCs [26] in order to invade the body . In the inner foreskin [9] , [10] and vagina [25] , early HIV-1 transmission involves capture of HIV-1 by epidermal LCs . These cells are able to internalize intact HIV-1 virions , due to their close proximity to the mucosal surface and their ability to bind the HIV-1 envelope glycoprotein subunit gp120 via their unique C-type lectin langerin [19] , [21] , [22] . The internalized intact virus is then transferred to T-cells locally within the epithelium [10] , [25] . Hence , LCs and T-cells are the first cells targeted by HIV-1 in both the male and female genital tracts early upon exposure to HIV-1 . To gain further insight as to the molecular/cellular dynamics of HIV-1 entry in the foreskin , we used in the current study our recently developed foreskin explants [10] , which permit for polarized exposure to the virus only via the apical side , as takes place in-vivo . Foreskin explants were inoculated comparatively for 4 h with HIV-1-infected or non-infected cells , as our previous study clearly demonstrated that foreskin transmission of cell-associated HIV-1 is much more efficient compared to that of cell-free virus [10] . In line with these observations , a limited number of studies aimed at ascertaining the infectivity of cell-associated HIV-1 and its potential importance in mucosal transmission , reported that cell-associated HIV-1 was transmitted more efficiently compared to cell-free virus ( e . g . [27] , and recently reviewed in [28] ) . To explore the possible molecular mechanisms that might affect the migratory behavior of T-cells and LCs in the foreskin , we measured the levels of several chemokines and cytokines that may be secreted by the inner foreskin tissue . Our results show that the inner foreskin produces high levels of CCL2/MCP-1 , CXCL8/IL-8 , IL-6 and CXCL10/IP-10 , in line with a recent study that documented a similar patter of secretion in inner foreskin explants [16] . However , HIV-1-infected cells do no affect the secretion of these chemokines/cytokines ( Fig . 1 ) , suggesting that these molecules probably do not contribute to the observed modifications in the spatial distribution of T-cells/LCs reported herein that are mediated by HIV-1-infected cells . In contrast , HIV-1-infected cells up-regulate CCL5/RANTES and down-regulate CCL20/MIP-3 alpha secretion , during both the first and late hours following viral inoculation . We speculate that at the early time points , chemokine secretion represents the levels of pre-synthesized/stored chemokines , rather then de-novo synthesis . Hence , the exact mechanisms by which HIV-1-infected cells modify early/late chemokine secretion may differ . Of note , the PHA/IL-2 activated PBMCs used herein may differ from naturally occurring HIV-1-infected cells in genital secretions and may secrete chemokines due to their experimental activation . Although both non-infected/HIV-1-infected cells secreted CCL5/RANTES and CCL20/MIP-3 alpha when incubated alone , without contact with the foreskin tissue , the secreted levels did not differ between non-infected and HIV-1-infected cells and were low ( data not shown ) . In contrast , chemokines secretion was always higher in the supernatant fractions upon inoculation with the tissue explants , suggesting that the contact between HIV-1-infected cells and the inner foreskin tissue results in the observed modifications in chemokines secretion , which may be contributed by the cells themselves , the tissue , or both . By measuring the density of T-cells in both the foreskin epidermis and dermis , we reveal herein that HIV-1-infected cell induce T-cell recruitment from the dermis into the epidermis , as the density of epidermal T-cell increases , in correlation with a decrease in their density in the dermis ( Fig . 2 ) . This process is restricted to the inner foreskin , as T-cell density remains unchanged in the outer foreskin , in line with previous studies including our own , showing that the inner , but not outer , foreskin is permissive to HIV-1 entry [7] , [10] . Importantly , our studies clearly show that CCL5/RANTES , a known T-cell chemokine [23] , [24] , which is increased following inoculation with HIV-1-infected cells , is responsible for such T-cell recruitment , as an activity neutralizing Ab to CCL5/RANTES completely abrogates this process ( Fig . 3 ) . This principal finding is in line with a recent study showing that CCL5/RANTES mediates specific recruitment of T-cells in human skin xenografts in-vivo [29] . Another study reported that treatment of inner foreskin explants for several days with the chemokine CCL3/MIP-1 alpha , but not with several others , induces T-cell infiltration into the epidermis [30] . However , as we show herein , CCL3/MIP-1 alpha secretion is decreased in inner foreskin explants following exposure to HIV-1 at such late time points ( Fig . 1 ) , suggesting that virus-mediated recruitment of T-cell into the epidermis does not involve this chemokine . Interestingly , the CCL5/RANTES neutralizing Ab used herein not only prevents epidermal T-cell recruitment , but apparently also facilitates exit of T-cell from the epidermis ( i . e . the fold value in Fig . 3B is 0 . 72±0 . 10 for 80 µg/ml neutralizing Ab ) . This observation suggests that endogenous tissue CCL5/RANTES has a physiological role in maintaining T-cells within the epidermis . In addition , we show that HIV-1-infected cells do not affect the density of epidermal LCs ( Fig . 4 ) , but rather modify their spatial distribution ( Fig . 5 ) . Hence , HIV-1-infected cells first induce attraction of LCs to the apical surface , and only later LCs travel towards the basement membrane . In fact , during the first hour of viral exposure , LCs cover a similar distance , but in opposite directions: apically upon exposure to HIV-1-infected cells and basally upon exposure to non-infected cells . As measuring the levels of secreted chemokines after 1 h is technically challenging , the identity of the specific chemokines mediating these opposing effects is currently an open question . Interestingly , the additional migration distance covered by LCs later on after inoculation with HIV-1-infected cells is higher compared to non-infected cells ( Fig . 5 ) , suggesting that LCs are able to migrate more rapidly ( i . e . longer distance during the same time period ) . This observation correlates with the decreased secretion levels of CCL20/MIP-3 alpha , the most potent chemokine for LCs [31] , at this time point . In contrast , inoculation with non-infected cells results in a gradual increase in LC distance from the apical surface along with a decrease in their epidermal density . This process probably reflects the natural emigration of LCs out of the epidermis , a known feature of these migratory cells that is routinely used experimentally for their isolation form mucosal epithelia . We further speculate that HIV-1 might have evolved the ability to modulate chemokines secretion to mediate the initial attraction of LCs apically , in order to facilitate its capture by LCs . This notion is supported by our microscopy studies showing for the first time co-localization of HIV-1 with LCs in-situ at the early hours following viral exposure ( Fig . 4 ) . Such process could permit early local spread of the virus to epidermal T-cells ( either already present within the epidermis and/or newly recruited from the dermis ) and establishment of an HIV-1-infected founder cell population that is crucial for systemic dissemination of the virus later [32] , [33] . Comparatively , in the female genital tract , HIV-1 enters lamina propria macrophages in explanted vaginal mucosa as early as 30 min after inoculation of virus onto the epithelium , and purified vaginal macrophages support substantial levels of HIV-1 replication [34] . In contrast , within the foreskin , dermal macrophages/DCs do not seem to participate in the early events of HIV-1 entry . Their densities and spatial distributions in the dermis remain completely unaffected by the presence of either HIV-1-infected or non-infected cells ( data not shown ) . In addition , intact HIV-1 virions may enter vaginal epithelial cells [35] , and also inner foreskin keratinocytes as shown here ( Fig . 4 ) . Whether these sequestered virions affect foreskin keratinocytes to promote viral entry is still unclear . Productive infection of LCs with HIV-1 following viral capture is limited , and may rely on langerin-mediated degradation of low concentrations of HIV-1 [22] and expression of several host factors that block HIV-1 replication [36] , [37] . In sharp contrast , previous studies have clearly demonstrated that HIV-1 captured by LCs can be efficiently transferred to T-cells across LC-T-cell conjugates , to induce extensive replication of the virus in T-cells [20] , [22] , [38]–[45] . In line with these studies , we recently reported that within 1 h , LCs in the inner foreskin internalize intact HIV-1 virions , which are transferred to T-cells across LC-T-cell conjugates forming rapidly within the epidermis [10] . In the female genital tract , endocytosed virions persist in LCs and remain accessible for transfer to conjugated vaginal T-cells [25] . Hence , LCs serve as storage for HIV-1 , facilitating its later spread to T-cells . Herein , we further followed the process of LC-T-cell conjugate formation in the inner foreskin at later time points . Such conjugated formation is initiated 1 h post HIV-1 inoculation and appears to be a continuous process with increase in conjugate numbers up to 4 h . Importantly , as LC density remains constant in the inner foreskin epidermis ( Fig . 4 ) , while T-cell density increases ( Fig . 2 ) , the recruitment of T-cell into the epidermis fuels this continuous and sustained formation of LC-T-cell conjugates . Thus , inhibiting the recruitment of T-cells into the epidermis may turn clinically useful to limit the local spread of the virus . Clinically approved CCR5 inhibitors [46]–[48] prevent HIV-1 infection of CCR5-expressing cells , and are also efficient microbicides preventing vaginal transmission in female macaques [49] , [50] . As demonstrated herein , CCL5/RANTES recruits T-cells into the inner foreskin epidermis , which drives the continuous process of LC-T-cell conjugate formation . Hence , CCR5 inhibitors may have a dual mechanism of action: directly inhibiting HIV-1 infection , as well as decreasing mucosal T-cell recruitment , to subsequently limit the local spread of the virus from LCs to T-cells . Future studies will hopefully challenge this notion , and examine the efficacy of CCR5 inhibitors as microbicides aimed at preventing inner foreskin HIV-1 entry . Based on our results , Fig . 7 summarizes the ‘chain-of-events’ of early HIV-1 entry in the inner foreskin . At 1 h , the interaction of HIV-1-infected cells with the inner foreskin leads to viral synapse formation , HIV-1 budding and attraction of LCs to the apical surface , which facilitates viral capture by LCs . At 4 h , LCs harboring intact HIV-1 virions in their cytoplasm rapidly travel towards the basement membrane , and T-cells are recruited from the dermis into the epidermis , to fuel the continuous formation of LC-T-cell conjugates within the inner foreskin epidermis . The re-localization of LCs towards the basement membrane may result from decreased secretion of CCL20/MIP-3 , while recruitment of T-cells results from increased CCL5/RANTES secretion , induced by HIV-1-infected cells . Blocking the responsiveness to these chemokines , for instance by CCR5 inhibitors already approved for clinical use , may turn useful as a novel mechanism to limit the local spread of HIV-1 within the inner foreskin epithelium , and perhaps also other mucosal epithelia .
The study was performed according to local ethical regulations following approval by the local ethical committee ( Comité de Protection des Personnes ( GHU Cochin-St vincent de Paul , Paris , France ) . Written informed consent was provided by all study participants and/or their legal guardians . Normal foreskin tissues were obtained from the Urology Service at the Cochin Hospital , Paris , France , from healthy adults ( mean age 32 years old , range 18–57 years ) undergoing elective circumcision due to personal reasons or phimosis , and according to local ethical regulations . Foreskin tissues removed due to cancerous or infectious pathologies were discarded and not used . Informed consents were obtained from all individuals . Tissues were placed in phosphate-buffered saline ( PBS ) supplemented with 20 µg/ml gentamicin ( Gibco Invitrogen , Carlsbad , CA ) , transported to the laboratory immediately following circumcision , and processed within the next 2 h . Foreskin tissues were separated mechanically into inner and outer parts , distinguished by their different colors and morphology , and any remaining fat and muscle tissue was removed from the dermal side . Round pieces from either inner or outer foreskin were cut using a 8 mm Harris Uni-Core , and placed with their epidermal side facing up on top of a polycarbonate membrane ( 12 µm pore size , 12 mm diameter ) of a two-chamber Costar Transwell permeable insert ( Corning Inc , Corning , NY ) . Hollow plastic cloning ring cylinders of 6 mm inner diameter ( VMR , Strasbourg , France ) were glued to the apical surface of each tissue piece , using the two-component biological fibrin sealant Tissucol kit ( Baxter Int . , Vienna , Austria ) , as previously described [10] . Sealing efficiency of the polarized apical chambers was monitored as previously described [10] . PBMCs from healthy donors were separated from whole blood by a standard Ficoll gradient . The HIV-1 primary isolate 93BR029 ( clade B , R5 tropic; NIH AIDS reagent program ) was amplified on phytohaemagglutinin ( PHA ) -stimulated PBMCs as described [51] . HIV-1 p24 antigen was quantified by the p24 Innotest HIV-1 ELISA ( Innogenetics , Gent , Belgium ) , and viral stocks were aliquoted for single use and stored at −80°c . To obtain HIV-1-infected cells , PHA/IL-2 stimulated PBMCs ( 5×106 cells in 10 ml RPMI 1640 ( Gibco ) supplemented with 10% fetal bovine serum ( FBS; PAN Biotech GmbH , Aidenbach , Germany ) , 2 mM L-glutamine , 100 U/ml penicillin , 100 µg/ml streptomycin ( Gibco ) , and 10 U/ml IL-2 ( Roche Diagnostics GmbH , Mannheim , Germany ) ) were inoculated with 200 ng p24 of the HIV-1 primary isolate 93BR029 . Two days later , further non-infected PHA/IL-2 stimulated cells were added ( 20×106 cells in 10 ml of the same medium containing 20 U/ml IL-2 ) . Infection was monitored in culture supernatants by p24 ELISA and by p24 intracellular staining and flow cytometry analysis . Infected cells were used between days 7–14 following addition of virus . Under these infection conditions , 1×106 infected cells released 100–500 pg p24 after 1 hr incubation at 37°c , and 5–7% of the infected cells were positive for p24 . Foreskin explants were exposed apically and in a polarized manner to 1×106 of either HIV-1-infected or non-infected PBMCs ( as negative control ) , in 100 µl RPMI 1640 added into the inner space of the cloning ring cylinders adhered to the tissue explants . Lower chambers were supplemented with 0 . 5 ml RPMI 1640 . Following 4 h incubation at 37°c , explants were either fixed in 4% paraformaldehyde ( PFA ) for further morphological evaluation , or processed for preparation of epidermal single-cell suspensions ( see below ) . Alternatively , following the 4 h polarized viral exposure , explants were washed and further incubated overnight at 37°c in a non-polarized manner in 0 . 5 ml RPMI 1640 . For measurement of secreted chemokines and cytokines , the supernatants fractions were removed , centrifuged , inactivated at 56°c for 45 min , and stored at −80°c . The levels of eleven chemokines and cytokines , secreted by triplicate foreskin explants exposed to either HIV-1-infected/non-infected PBMCs ( 4 h polarized exposure followed by overnight non-polarized incubation ) , were measured by a custom multiplex bead immunoassay kit ( Bender MedSystems , Vienna , Austria ) and quantified by flow cytometry , according to the manufacturer's instructions . Recorded profiles from flow cytometry were exported and analyzed using the FlowCytomix Pro 2 . 3 Software ( Bender ) enabling calculation of the concentration ( pg/ml ) of each analyte in the tested samples . The levels of CCL20/MIP-3 alpha in supernatants collected after similar exposure to the virus , or following short-term polarized infection for 4 h , were quantified with the Quantikine human MIP-3 alpha ELISA kit ( R&D systems , Minneapolis , MN ) , according to the manufacturer's instructions , and translated to actual concentrations ( pg/ml ) with the standard curve obtained using the kit standards . When indicated , CCL20/MIP-3 alpha and CCL5/RANTES levels following short-term polarized infection for 4 h were also evaluated by a custom semi-quantitative Human Cytokine Antibody Array and fluorescent detection ( RayBiotech , Inc . , Norcross , GA ) , according to the manufacturer's instructions . Triplicate foreskin explants were pre-incubated for 30 min at 37°c with either 20 or 80 µg/ml of a goat-anti-human CCL5/RANTES neutralizing Ab ( R&D ) , added to both apical and basal compartments . Normal goat IgG ( Santa Cruz Biotechnology , Santa Cruz , CA ) added at 40 µg/ml served as negative control . Explants were then inoculated with HIV-1-infected or non-infected PBMCs for 4 h as described above , and following removal of the cloning ring cylinders , pooled tissue pieces were incubated with their epidermal side facing up in 1 ml RPMI 1640 medium supplemented with 2 . 4 U/ml Dispase II ( Roche Diagnostics GmbH , Mannheim , Germany ) in a 12-wells plate overnight at 4°C . The epidermis and dermis were then mechanically separated using forceps . Epidermal single-cell suspensions were prepared by incubating pooled epidermal sheets from each triplicate in 1 ml 0 . 05% Trypsin/EDTA ( Gibco ) for 10 min at 37°C , followed by inactivation of trypsin with 1 ml FBS , mechanical disruption using a 10 ml pipette , filtration of released cells through a 40 µm nylon cell strainer and centrifugation . For surface staining of CD3 and langerin , epidermal cell suspensions were resuspended in PBS and transferred to a 96 round-bottom wells plate ( 0 . 1–0 . 5×106 cells/well ) . Cells were then incubated for 30 min on ice with 10 µg/ml of PE-conjugated mouse-anti-human CD3 ( BD Pharmingen , San Jose , CA ) and APC-conjugated mouse-anti-human langerin ( R&D ) mAbs diluted in PBS to a final volume of 50 µl/well . Cells were then washed , centrifuged , and fixed for 15 min at room temperature with 4% PFA . Cells stained with matched isotype control Abs served as negative control . Fluorescence profiles were recorded using a FACSCalibure and results were analyzed using the CellQuest Pro software . Following infection and fixation , foreskin explants were embedded in paraffin . Serial 4 µm sections were cut , deparaffinized in xylene and graded alcohol solutions , microwave heated in 10 mM citrate buffer pH = 6 . 0 for antigen retrieval , cooled down in the same buffer , and washed in PBS . For confocal microscopy , sections were quenched with PBS/50 mM glycine/75 mM ammonium chloride ( Sigma , St . Louis , MO ) for 10 min at room temperature and blocked in PBS containing 20% horse serum ( HS , Vector Laboratories , Burlingame , CA ) and 0 . 1% Tween-20 ( Sigma ) for 1 h at room temperature . Following washes in PBS , sections were incubated overnight at 4°c with primary Abs diluted in PBS/2% HS/0 . 1% tween ( 50 µl/section ) , including goat-IgG-anti-human langerin at 20 µg/ml ( R&D ) ; mouse-IgG-anti-human CD3 at 10 µg/ml ( DakoCytomation , Glostrup , Denmark ) ; a cocktail of several mouse and human monoclonal Abs ( 5–10 µg/ml ) recognizing HIV-1 gp41 ( 41A ( Hybridolab , Pasteur Institute , Paris , France ) 2F5 , D50 , 4E10 ( NIH ) ) , gp120 ( 2G12; NIH ) and p24 ( Kal-1; Dako ) . Sections were then washed with PBS and incubated for 1 h at room temperature with appropriate secondary Abs ( 50 µl/section , diluted 1∶50–1∶100 in PBS ) , including TRITC-conjugated anti-goat IgG ( Beckman Coulter , Villepinte , France ) and FITC-conjugated anti-human/mouse IgG ( Jackson Immunoresearch , West Grove , PA ) , or Alexa488-conjugated anti-goat IgG ( Invitrogen , Cergy Pontoise , France ) and Cy5-conjugated anti-mouse IgG ( Jackson ) . Cell nuclei were then stained with DAPI ( Sigma; 50 µl/section , 10 min at room temperature ) , and the sections washed again in PBS and mounted with MOWIOL medium ( Calbiochem Merck , Darmstadt , Germany ) , containing DABCO antifade ( Sigma; 100 mg/ml ) . Staining was visualized with Leica TCS SP or DMI6000 microscopes ( Leica Microsystems , Wetzlar , Germany ) . Signals were processed based on established parametric settings using specific isotype controls [10] . Acquired image stacks were analyzed with Imaris Software ( Bitplane AG , Zurich , Switzerland ) . Image deconvolution processing was performed by the Maximal Likelihood Estimate procedure of the Huygens software ( Scientific Volume Imaging BV , Hilversum , The Netherlands ) . For immunohistochemistry , the VECTASTAIN Elite-ABC kit ( Vector ) was used according to the manufacturer's instructions . Sections were quenched with 3% hydrogen peroxide for 10 min at room temperature and blocked with PBS/3% HS for 20 min at room temperature . All Abs , including mouse-IgG2b-anti-human langerin ( Abcam , Cambrige , UK ) and mouse-IgG1-anti-human CD3 ( Dako ) , as well as appropriate matching isotype controls , were diluted to 5–10 µg/ml in Ab diluent ( Dako ) and incubated with tissue sections for 30 min at room temperature ( 50 µl/section ) . After washing in PBS , sections were incubated with appropriate biotinylated secondary Abs followed by HRP-conjugated streptavidin ( Vector; 30 µl/section , 20 min at room temperature ) , and visualized with di-amino benzidine ( DAB; Dako ) , 3-amino-9-ethyl carbazole ( AEC; Dako ) or HistoGreen ( AbCys , Paris , France ) peroxidase substrates . Sections were counterstained for 30 sec with hematoxyline solution ( Dako ) and mounted with aqueous MOWIOL or non-aqueous Vectamount ( Vector ) media , according to the recommendations for each substrate . For double antigen staining a similar protocol was used with the addition of a quenching step using LinBlock ( Abcys ) between the first and second antigen staining . Images were taken using the ×40 objective of a Nikon E800 microscope ( Nikon , Tokyo , Japan ) equipped with a CCD QICAM camera ( QImaging , Surrey , BC , Canada ) . For cell quantification , the ImageJ software ( NIH ) was used to count the number of positively stained cells in a minimum of 10 separate fields/stained section , and to calculate the surface of either the epidermis or dermis in each field . Measurement of the distance between 100–200 LCs and the mucosal surface was evaluated using the Arcgis V9 . 3 program ( ESRI , Redlands , CA ) , as described earlier [10] . Foreskin explants were fixed for 1 h with 3% glutaraldehyde . Samples were then postfixed in 1% osmium tetroxide in 0 . 1 M PBS and dehydrated in increasing graded alcochol solutions . After 10 min incubation in a mixture of 1∶2 epoxy propane and epoxy resin , the samples were embed in gelatine capsules with freshly prepared epoxy resin and polymerized at 60°c for 24 h . Sections ( 80 nm ) were then cut with an ultramicrotome ( Reichert Ultracut S , Reichert Technologies , Depew , NY ) , stained with uranyl acetate and Reynold's lead citrate , and observed with a transmission electron microscope ( JEOL 1011 , JEOL , Tokyo , Japan ) , equipped with a GATAN numerical camera ( Gatan , Munich , Germany ) . Pictures were taken and digitalized with Digital Micrograph software ( Gatan , Munich , Germany ) . Statistical significance was analyzed by the Student's t-test . | Circumcision reduces HIV-1 acquisition in men , suggesting that the foreskin is an HIV-1 entry site . We previously showed that two types of immune cells in the foreskin epidermis are the first ones targeted by HIV-1 . Hence , Langerhans cells ( LCs ) positioned in proximity to the external surface rapidly capture incoming HIV-1 . The internalized virus is then transferred to T-cells positioned deeper within the epidermis . Herein , we studied the molecular mechanisms affecting the movement of these motile cells by testing whether HIV-1 alters the secretion of foreskin chemokines ( i . e . molecules that influence cell migration ) . Our results show that HIV-1 increased the secretion of CCL5/RANTES , a potent T-cell chemokine , which mediated T-cell recruitment into the epidermis . In parallel , HIV-1 decreased the secretion of CCL20/MIP-3 alpha , a potent LC chemokine , enabling LCs to travel deeper into the tissue . The two cell types then ‘meet’ to form close contacts that permit the transfer of the virus from LCs to T-cells within the epidermis . Together , these results reveal that HIV-1 modifies foreskin chemokines secretion and the subsequent relocation of its initial immune target cells . Therefore , blocking the responsiveness to these chemokines clinically may limit the local spread of HIV-1 within the foreskin . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"infectious",
"diseases",
"sexually",
"transmitted",
"diseases",
"aids",
"biology"
] | 2011 | HIV-1 Efficient Entry in Inner Foreskin Is Mediated by Elevated CCL5/RANTES that Recruits T Cells and Fuels Conjugate Formation with Langerhans Cells |
Most individuals exposed to hepatitis C virus ( HCV ) become persistently infected while a minority spontaneously eliminate the virus . Although early immune events influence infection outcome , the cellular composition , molecular effectors , and timeframe of the host response active shortly after viral exposure remain incompletely understood . Employing specimens collected from people who inject drugs ( PWID ) with high risk of HCV exposure , we utilized RNA-Seq and blood transcriptome module ( BTM ) analysis to characterize immune function in peripheral blood mononuclear cells ( PBMC ) before , during , and after acute HCV infection resulting in spontaneous resolution . Our results provide a detailed description of innate immune programs active in peripheral blood during acute HCV infection , which include prominent type I interferon and inflammatory signatures . Innate immune gene expression rapidly returns to pre-infection levels upon viral clearance . Comparative analyses using peripheral blood gene expression profiles from other viral and vaccine studies demonstrate similarities in the immune responses to acute HCV and flaviviruses . Of note , both acute dengue virus ( DENV ) infection and acute HCV infection elicit similar innate antiviral signatures . However , while transient in DENV infection , this signature was sustained for many weeks in the response to HCV . These results represent the first longitudinal transcriptomic characterization of human immune function in PBMC during acute HCV infection and identify several dynamically regulated features of the complex response to natural HCV exposure .
Despite the recent breakthrough of highly effective direct acting antiviral therapies , hepatitis C virus ( HCV ) remains a significant public health threat . New infections , especially among people who inject drugs ( PWID ) , are likely to increase in the absence of a prophylactic vaccine [1] . Effective vaccine development is hampered by our limited understanding of how protective immunity is established in the acute stages of natural infections . Acute HCV infection has two dichotomous outcomes , spontaneous resolution ( ~25% of infections ) or chronic infection ( ~75% of infections ) [2] . Immune functions following viral exposure remain incompletely understood due in part to the limited availability of paired pre-infection and longitudinal acute infection research samples from recently exposed , largely asymptomatic individuals . Previous work has established roles for innate and adaptive immunity in the host response to acute HCV . Genetic polymorphisms at the IFNL3/4 locus , NK cell activity , and dendritic cell function influence infection outcomes [3 , 4] . Effective adaptive immunity is also essential for HCV clearance . HCV-specific CD4 and CD8 T cell responses are induced in most acutely infected individuals irrespective of outcome . However , failure to sustain CD4 T cell responses is associated with viral persistence , which in turn leads to CD8 T cell dysregulation and exhaustion [3 , 4] . The role of B cells in acute HCV is less clear . Although not consistent across all studies , anti-HCV neutralizing antibodies have been associated with spontaneous clearance during primary and secondary infections , suggesting that they may contribute to long-term protective immunity [5] . Kinetics and crosstalk between these innate and adaptive responses remain incompletely defined . Systems-level transcriptomic methods have emerged as powerful tools for profiling human immune responses [6] . Examining peripheral blood transcriptome data has provided integrated maps of host response dynamics following vaccination or infection , and the associated interplay of innate and adaptive immune components [7–10] . Studies of the responses to yellow fever and influenza vaccines have identified shared gene expression signatures associated with strong antibody responses [7 , 9–12] . Related studies of influenza [13] and hepatitis B virus ( HBV ) vaccines [14] have underscored the role of baseline inflammation and host factors in determining the outcome of vaccination . Although challenging due to logistical demands and interindividual variation , analogous methods have also been successful in characterizing the human immune response to “real world” acute infections by pathogens such as dengue virus ( DENV ) [15] . We reasoned that similar transcriptomic approaches , with the potential to extract large amounts of data from relatively limited sample material would be useful in characterizing the response to acute HCV . The first microarray studies of acute HCV infection performed on serial liver biopsies from a limited number of chimpanzees showed that innate immune responses are rapidly induced in the liver irrespective of infection outcome [16 , 17] . Spontaneous HCV clearance was associated with upregulation of genes linked to CD4 T cells and lymphocyte migration to the liver [16 , 17] . More recent transcriptomic studies of HCV-specific T cells indicate that metabolic dysregulation during acute infection may influence the outcome of the antiviral T cell response [18] . Transcriptome analysis in human livers demonstrated elevated IFNγ–stimulated gene expression in acute infection , but elevated IFNα–stimulated gene expression during chronic infection [19] . Additional microarray studies described elevated interferon stimulated gene ( ISG ) expression in peripheral immune cells during chronic HCV [20] . Despite our understanding of certain aspects of the host response to HCV , the composition and dynamics of the early antiviral response in acute infection have not been fully defined . Here , we characterize the host response to acute HCV through transcriptional profiling of peripheral blood mononuclear cells ( PBMC ) . We performed RNA-Seq on longitudinal samples collected before , during and after acute HCV infections that resulted in spontaneous resolution or chronic infection . Our analysis provides a detailed characterization of the inflammatory and ISG signatures active in the spontaneous resolution of acute HCV , and identifies similarities with responses to flavivirus vaccines and infections .
Identifying and recruiting research subjects shortly after HCV exposure is challenging due to the typically asymptomatic nature of the infection [2] . The Montreal Hepatitis C Cohort ( HEPCO ) recruits and follows PWID at high risk of HCV exposure and infection [21] . Longitudinal samples from this cohort provide a rare opportunity to explore the dynamics of the immune response to acute HCV . Here , we examined PBMC from 14 individuals ( Tables 1 and 2 ) who became infected with HCV , of whom 6 spontaneously cleared the virus ( Resolution group ) and 8 progressed to chronic infection ( Chronic group ) . We performed RNA-Seq on PBMC samples collected at several time points relative to HCV exposure: Pre-infection , Early acute , Late acute , and Follow up ( Fig 1A ) as described in Materials and methods . We first performed principal component analysis ( PCA ) to assess the degree to which gene expression patterns corresponded with sample group assignments . Within the Resolution group , we observed clear separation ( first principal component , 32 . 6% explained variance , patient-specific variation removed as detailed in Materials and Methods ) of the Early acute samples from Pre-infection and Follow up samples ( Fig 1B , and S1A Fig ) . Some of the Late acute samples also separated from Pre-infection and Follow-up samples but grouped with the Early acute samples; this grouping corresponded with detectable HCV viremia . Within the Chronic group , although some general trends were apparent , PCA did not clearly separate samples according to experimental group ( S1A and S1B Fig ) . The cause of this intersample heterogeneity in the chronic group is not clear . The Resolution and Chronic groups are similar with regard to age and ethnicity ( Tables 1 and 2 ) , and are drawn from the same populations . At this limited sample size , we suspect that confounding self-reported factors ( e . g . other minor non-HCV infections , inflammatory conditions , etc . ) may contribute to variation in gene expression patterns . We reasoned that , given the considerable interindividual variation apparent in human gene expression data generally , and the additional variation potentially introduced by approximate time point sampling of “real world” acute HCV infection , a statistical analysis on appropriately grouped samples would be most likely to provide high quality , generalizable results regarding the immune response to HCV . Therefore , we chose to focus the remainder of our analysis on the Resolution group , which we further partitioned based on viremia status: Early acute , Late acute ( positive HCV viremia ) , Late acute ( negative HCV viremia ) and Follow up . We began our analysis by examining changes in individual gene expression patterns . Applying differential gene expression testing across all time point groups , we detected numerous genes whose expression values were significantly altered during acute HCV infection and resolution ( F-test q-value < 0 . 1 , S1 Table ) . As gene expression dynamics and PCA ( Fig 1B ) suggested that the most pronounced changes occurred in samples with detectable viremia , we conducted pairwise differential gene expression testing for each post-infection time point group versus pre-infection baseline . We detected 853 individual genes differentially expressed at the Early acute time point , with few genes meeting significance thresholds in other groups ( q-value < 0 . 05 , Fig 1C , S2 Table ) . Based on PCA ( and subsequent analyses described below ) , the low number of significant genes detected in the Late acute , positive HCV viremia group was likely due to the small number of samples ( n = 3 ) available for analysis . Overall , these results indicate that during acute infection , the immune response to HCV includes substantial changes to peripheral blood transcriptional signatures . These differences are most pronounced during the Early acute stage of infection , but appear to persist ( at least in part ) during periods of detectable viremia . With a goal of translating gene expression patterns to specific immune functions modulated in acute HCV , we next focused our analysis on the differential regulation of blood transcriptome modules ( BTMs ) [11 , 22] rather than individual genes . Each BTM contains a set of genes with correlated expression patterns , annotated with associated biological functions . Using the MROAST gene set enrichment tool [23] , we found 60 BTMs to be differentially enriched ( q-value < 0 . 05 , details in Materials and methods ) during Early acute infection as compared to Pre-infection baseline ( Fig 2B , S3 Table ) . Of note , as in the above differential gene expression analysis , changes in BTM activity were apparent in the Late acute , positive HCV viremia group but did not clear significance thresholds . At the Follow up time point , BTM activity was indistinguishable from Pre-infection levels . The activities of all differentially enriched BTMs in individual patient samples are presented in Fig 2A . BTMs differentially enriched at the Early acute time point correspond to diverse immune functions and cell types . We further classified enriched BTMs at high level biological annotations based on categories defined by Kazmin et al [24] . Gene expression dynamics and directionality were generally similar for BTMs within the same category . Categories containing BTMs upregulated in the response to Early acute HCV include interferon/antiviral sensing , inflammatory/TLR/chemokines , monocytes , DC activation , and antigen presentation . Of note , upregulation of the “T cell surface , activation ( M36 ) ” BTM corresponded with an increased frequency of HCV-specific CD8+ T cells , as measured by peptide-MHC tetramer analysis ( A2/NS3-1073 ) for HLA-A*0201+ patient series ( n = 3 , S2 Fig ) . Multiple downregulated BTMs were annotated in the B cell category . Taken together , these results indicate that the response to acute HCV infection involves a robust innate antiviral gene expression program in peripheral immune cells , inflammatory signals , and changes associated with innate and adaptive cell types . During acute infection , HCV triggers a potent IFN-mediated antiviral response in the liver [16 , 17 , 19 , 25] . In our PBMC analysis , BTMs associated with the innate antiviral response were strongly upregulated during acute HCV ( Fig 2 , S3 Fig ) . At Early and Late acute time points , this upregulation corresponded with detectable HCV viremia; samples from individuals who achieved viral clearance by the Late acute time point displayed innate antiviral BTM activity similar to Pre-infection levels . Based on these results , we sought to further characterize our observations from BTM analysis using complementary reference data to describe the innate antiviral response . We tested for enrichment of a 277 gene “PBMC ISG set” empirically derived from RNA-Seq analysis of PBMC stimulated ex vivo with Type I IFN [26] . In agreement with BTM results , this ISG collection was significantly upregulated in the Early acute and the Late acute , positive HCV viremia time point groups , but not in the Late acute , negative HCV viremia or Follow up time point groups ( Fig 3A ) . We extended this analysis to define which ISGs best define the PBMC innate antiviral response to acute HCV by intersecting the PBMC ISG set with the list of differentially expressed genes at the Early acute time point ( as compared to Pre-infection ) . The resulting list , comprised of 105 PBMC ISGs , includes ISGs previously implicated in the peripheral blood response to HCV ( CXCL10 ) [27–29] , “classical” ISGs ( Mx1 , OAS1 , ISG15 ) , ISG transcription factors ( STAT1 , STAT2 , IRF7 ) , as well as ISGs associated with immunomodulation ( IL15 , CD38 ) . When intersecting this “acute HCV PBMC ISG signature” with ISG lists derived from IFN stimulated PBMC [30] or HCV liver microarray datasets ( Fig 3B ) , we noted overlap ( 44/105 acute HCV PBMC ISGs ) with ISG induction in acutely infected chimpanzee liver ( [25] , 5–11 weeks post-infection ) , and less overlap ( 16/105 acute HCV PBMC ISGs ) with acutely infected human liver ( [19] , <6 months post-infection ) . This discrepancy may be due to temporal differences in liver biopsy acquisition; the human study sampled from a broader time window and described a predominantly Type II IFN signature ( perhaps reflecting infiltrating adaptive immune cells later in infection ) . Overall , these results indicate that acute HCV infection , despite its hepatotropism , initiates a robust type I interferon response in peripheral immune cells . To evaluate a component of this signature at the protein level , we measured levels of CXCL10 ( IP-10 ) , an important factor in the innate response to HCV [27 , 31 , 32] , in corresponding plasma samples ( available from most of the same patient-timepoint conditions examined by RNA-Seq ) . Similar to the patterns observed in RNA-Seq data , plasma CXCL10 levels were elevated in all samples measured at the Early acute timepoint , and later returned to baseline levels with viral clearance ( Fig 3C ) . Furthermore , plasma CXCL10 values correlated with PBMC CXCL10 expression levels measured by RNA-Seq ( Pearson’s r = 0 . 69 , p = 0 . 00071 ) ( Fig 3D ) . These data indicate that RNA-Seq gene expression measures in PBMC reflect the systemic protein levels of an interferon-induced chemokine during acute HCV infection . Several BTMs related to B cells ( M47 . 0 , M47 . 1 , M69 , S9 ) were found to be downregulated during Early acute HCV infection ( Fig 2B ) . The M54 ( “BCR signaling” ) BTM was a notable exception and was significantly upregulated in the Early acute time point group . Most individual genes within downregulated B cell-associated BTMs demonstrated reduced expression values ( Fig 4A ) . Rather than a decrease in gene expression output by B cells , we suspected that such a general reduction in B cell transcriptional signatures might correspond to a decrease in B cell frequency within PBMC . Therefore , we used flow cytometry to quantify the relative frequency ( Fig 4B ) and fold-change ( Fig 4C ) of CD19+ B cells in samples for which sufficient experimental material remained available . Although the limited sample numbers were insufficient to achieve statistical significance in an analysis incorporating both time point and HCV viremia status , we observed an intriguing trend of diminished B cell frequency as a fraction of total PBMC ( relative to Pre-infection values , per patient ) that appeared to correspond to detectable viremia at Early acute and Late acute time points ( Fig 4B and 4C ) . Peripheral blood transcriptome studies have been effective in providing useful “reference” profiles of protective immune responses to different vaccines [7 , 9–11 , 13] . In an effort to contextualize the Acute HCV response with additional well-characterized responses to different immune challenges , we compared acute HCV PBMC transcriptional profiles to analogous , microarray studies of the PBMC response to live attenuated vaccines ( yellow fever 17D , YFV; influenza , LAIV ) , an inactivated viral vaccine ( trivalent influenza vaccine , TIV ) , a polysaccharide vaccine ( meningococcal polysaccharide vaccine , MPSV4 ) , and a conjugated polysaccharide vaccine ( meningococcal conjugate vaccine , MCV4 ) . After measuring BTM enrichment with GSEA ( pre-ranked ) [33] for each dataset ( peak response vs . pre-vaccination , acute HCV infection vs . Pre-Infection ) , we compared responses by overlap of differentially regulated BTMs ( GSEA FDR 0 . 01 , workflow in S4 Fig ) . Similar to the analysis strategy described by Li et al . [11] , this approach enables qualitative comparisons across methods ( RNA-Seq , different microarray platforms ) , and modulates somewhat the statistical effects of varied sample sizes . The acute HCV response shared more significant BTMs with YFV than with any other vaccine ( Fig 5 ) . Both the acute HCV and YFV responses included interferon/antiviral sensing BTMs , inflammation-associated BTMs , and modules related to T cell proliferation . Although the above vaccine comparisons provide informative functional context regarding the nature of the acute HCV response as compared to defined immune challenges , vaccines , by definition , are not full potency infections . Therefore , in order to evaluate our HCV results in relation to a bona fide viral infection , we compared the acute HCV response ( RNA-Seq data described here ) to the response mounted against acute DENV infection ( publicly available whole blood microarray data ) [15] . First , we partitioned the DENV dataset into three groups based on distinct gene expression profiles ( PCA analysis , S5 Fig ) ; similar groupings were observed by Kwissa et al in their original analysis . These groups correspond to viral load and time post-symptom onset: High viral load ( 2–3 days ) , Moderate viral load ( 4–6 days ) , and Low viral load ( 5–9 days ) . Next , we measured BTM enrichment ( acute infection groups versus matched convalescence “baselines” ) and assessed which BTMs were similarly regulated in acute resolving HCV and acute resolving DENV infection . We observed considerable overlap in the BTM response to acute HCV and acute DENV , which included many of the modules also identified in the YFV comparison ( Fig 6 ) . Concordant enrichment of BTMs was most apparent in the DENV High viral load ( 2–3 days ) condition , with increased activity of innate antiviral BTMs , inflammation BTMs and T cell proliferation BTMs , and decreased activity of B cell BTMs . Although upregulation of BTMs related to T cell proliferation seems to persist through lower viral load/later DENV time points , many BTMs associated with the innate response ( inflammatory/TLR/chemokines , monocytes , a subset of interferon/antiviral sensing BTMs ) do not appear to be differentially enriched beyond the High viral load ( 2–3 days ) group . This analysis suggests that during acute DENV infection , the early host response is characterized by inflammatory signals , B cell changes , and an innate antiviral signature , which is diminished as the response incorporates adaptive immune functions ( i . e . T cell proliferation ) . Although many of the same BTMs are involved , this temporal pattern is in sharp contrast to the timeline of the acute HCV response . At the Early acute stage ( approximately 6 weeks post-infection ) , the HCV response appears similar to the apparently short-lived ( days ) initial response to acute DENV infection .
Here we present a detailed transcriptomic characterization of early events in the human immune response to acute HCV infection in individuals that progress to spontaneous resolution . We detected wide-ranging changes in PBMC gene expression patterns , including those consistent with pronounced innate antiviral programs and inflammatory mediators . The patterns observed in the PBMC response to acute HCV shared many features with effective immune responses to YFV and acute DENV infection . To our knowledge , this study represents the first longitudinal transcriptomic investigation of the PBMC response to acute HCV in humans . Studying the early immune response to HCV in humans is challenging , as infected individuals are not usually recognized until they progress to chronic infection . Even if research subjects are identified during acute stages , obtaining pre-infection samples for comparison is often impossible . We addressed these difficulties in a rare longitudinal study of HCV-naïve PWID [21 , 34] , through which we obtained acute infection and corresponding pre-infection baseline samples . Given this “real world” setting , we observed considerable variability in datasets from different patients . Differences in gene expression patterns could be due to complex factors associated with the PWID population enrolled in this study , including drug use , unstable socioeconomic conditions and exposures to additional ( non-HCV ) infections . Furthermore , the time of HCV infection is necessarily estimated and the samples grouped for each time point are not perfectly synchronized . Problems in assigning datasets to distinct analysis groups were particularly pronounced in samples from individuals who eventually progressed to chronic infection . Although reasons for greater consistency of expression patterns in the Resolution group remain unclear , much of the variation in the Chronic group samples appears unrelated to HCV infection timepoint ( S1 Fig ) . We do not believe this heterogeneity is related to infection outcome . These issues might be overcome with a larger sample size , which was unfortunately not possible in the present study . This study’s original goal was to identify disparities in the initial response to infection that contribute to the differential outcomes of spontaneous infection versus chronicity . Indeed , recently reported analyses of HCV-specific T cells suggest that differences in metabolic networks engaged early in infection contribute to viral clearance [18] . In our datasets , the heterogeneous gene expression patterns measured in the Chronic group precluded a direct comparison to spontaneous resolution at this sample size . However , this longitudinal study did enable a detailed characterization of the peripheral immune response to acute HCV infection that results in spontaneous resolution . Within the Resolution group , we structured our analysis to evaluate changes relative to patient-specific Pre-infection samples and we observed several consistent and concordant transcriptomic patterns across most subjects analyzed . With the range of estimated infection dates , these results suggest that certain immune functions are active to some extent for at least several weeks during acute HCV . Our results define a pronounced innate antiviral gene expression program active in PBMC during the Early acute stage of HCV infection . This response persisted through the Late acute time point in individuals with detectable HCV viremia . These observations are consistent with transcriptomic data from the chimpanzee model demonstrating rapid induction of ISGs in the liver during early acute HCV [16 , 17] . Furthermore , we observed overlap in the PBMC ISG signature identified here and ISG expression patterns in acute HCV liver biopsy samples from infected patients and chimpanzees [19 , 25] . This suggests that the innate antiviral response against acute HCV is not restricted to the liver and that the peripheral blood response corresponds to some extent with that at the site of infection . This response returned to baseline levels following viral clearance , suggesting a dependence on viral RNA . We also observed a decrease in gene signatures associated with B cells . Such changes could reflect a decrease in B cell transcriptional activity , a decrease in the relative frequency of B cells , or both . Based in part on trends observed in flow cytometry analysis of a limited number of samples , we speculate that these patterns result from a diminished fraction of B cells within PBMC during acute HCV viremia . Although it is possible that this pattern could simply be a consequence of a corresponding relative increase in another cell type ( e . g . monocytes , as suggested by BTM enrichment ) , we did not detect significant decreases in BTMs associated with other cell types ( e . g . T cells , NK cells ) . As both RNA-Seq ( as applied here to bulk PBMC samples ) and flow cytometry are relative quantification methods , we cannot ascertain if these changes correspond to a decrease in the absolute frequency of B cells in peripheral blood . However , several recent studies have described mechanisms of virus- and IFN- mediated B cell dysregulation [35–37] . Future studies including sample collection focused on absolute immune cell quantification will be required to further explore this observation in the context of acute HCV infection . Our comparative analyses revealed notable similarities between host responses to acute HCV and to effective vaccines . More specifically , we observed many shared features with the response to YFV , a live attenuated flavivirus that may be thought of as approximating an acute viral infection . As yellow fever virus is a related hepatotropic flavivirus , these similarities are not entirely surprising . In addition , the immune response during Early acute HCV infection shared many features with the immune response to acute DENV infection . Although challenging to make a formal comparison due to discrepancies in study design and timing , these qualitative results suggest that at least at some point during acute HCV , a response similar to that elicited by DENV is induced , likely reflecting the core innate antiviral programs activated in response to RNA virus infection . However , in acute HCV , ISG expression remains elevated long after initial infection and is maintained for many weeks of detectable viremia . This pattern is consistent with models in which adaptive immunity to HCV is delayed as compared to flavivirus infections , despite apparently similar robust IFN responses [16 , 38 , 39] . Even in the case of eventual spontaneous resolution , HCV outpaces these innate antiviral effectors , resulting in prolonged high-level viremia [38 , 39] . Eventually , HCV-specific CD4 and CD8 T cells arise to eliminate the virus , after which innate antiviral signatures rapidly normalize . Failure to prime effective adaptive immunity results in chronic infection that is associated with persistent ISG expression ( reviewed in [3 , 4] ) . This longitudinal study provides an initial assessment of the peripheral immune response to acute HCV at a systems level . Follow-up studies with a larger cohort and sufficient sample availability for complementary experimental methods ( e . g . comprehensive flow cytometry analysis ) will be required to validate these findings beyond the patients described here and determine the impact of these dynamically regulated immune processes on the differential outcome of acute HCV infection .
This study was approved by the ethics committee of the Centre Hospitalier de l'Université de Montréal ( CHUM ) ( Protocol SL05 . 014 ) . All research was conducted according to the principles expressed in the Declaration of Helsinki . All subjects provided written informed consent . Subjects with acute HCV were recruited among high-risk PWID participating in the HEPCO cohort as previously described [21 , 34] . Estimated date of infection ( EDI ) was calculated as the median date between the last HCV negative and the first HCV positive test . Spontaneous viral resolution ( n = 6 ) or chronic infection ( n = 8 ) was defined as the absence or presence of HCV RNA , respectively , at 6 months post EDI ( Cobas Ampliprep/Cobas TaqMan HCV Qualitative Test , version 2 . 0; Limit of detection: 15 IU/ml ) ) . PBMC samples from four time points were examined: i ) Pre-infection baseline ( Variable ) ; ii ) Early acute ( 2–9 weeks , mean 6 weeks ) ; iii ) Late acute ( 15–20 weeks , mean 17 weeks ) ; and iv ) Follow up ( 25–71 weeks , mean 52 weeks ) . Subjects’ clinical characteristics , demographics , testing intervals and actual times post EDI for each sample are presented in Tables 1 and 2 . Cryopreserved PBMC samples were thawed , viability was assessed using trypan blue ( >90% for all samples ) and immediately processed for RNA extraction ( approximately 7x106 PBMC per sample , range: 4 x 106–11 x 106 ) with the Qiagen RNEasy Mini kit . RNA quality was assessed by Agilent Bioanalyzer 2100; all samples exhibited RNA integrity numbers ( RIN ) greater than 8 . RNA-Seq libraries were prepared with the Illumina TruSeq RNA Library Preparation Kit v2 . Libraries were sequenced in multiplex on the Illumina HiSeq 2500 platform in 100 nucleotide , single-end read configuration ( range 3 x 107–7 x 107 total reads per library ) . Samples from the same patient series were always sequenced in the same multiplex pool to minimize batch effects . Reads were mapped to the human genome reference ( hg19 ) using the Tophat ( v2 . 0 . 8b ) alignment tool [40] . Read counts per gene were quantified against Ensembl ( v66 ) transcript reference annotations ( appended with gene annotation for IFNL4 ) using HTSeq-count ( v0 . 5 . 4p3 ) [41] . Analysis was conducted within the R statistical framework . For principal component analysis , read counts were normalized and variance stabilized by regularized log transformation ( rlog ( ) function , DESeq2 package v1 . 18 . 1 ) . Patient-specific gene expression variation was corrected using the removeBatchEffect ( ) function in the limma package ( v3 . 26 . 9 ) , specifying time point and viremia status for preservation . For independent Spontaneous Resolution ( Fig 1B ) and Chronic infection ( S1B Fig ) PCAs , analysis included the top 500 most variable genes across samples from the indicated infection group . For joint Chronic Infection and Spontaneous Resolution PCA ( S1A Fig ) , PCA was performed on a single gene list ( top 1000 most variable genes across all samples ) , and Chronic and Resolution groups were plotted separately to facilitate visualization . In order to incorporate the multiple covariates apparent in the experimental design , differential expression and BTM enrichment analyses were conducted with the voom-limma analysis workflow ( v3 . 26 . 9 ) [42] . Prior to differential expression analysis , a filter was applied to remove genes with low expression values ( genes with greater than one read count per million ( cpm ) in at least four samples were designated as “expressed” ) ; 14 , 138 genes passed filter and were included in subsequent analyses . RNA-Seq read counts were scaled and normalized by the trimmed mean of M values ( TMM ) method ( implemented in the edgeR package [43 , 44] ) and log2 transformed using voom [45] . All differential gene and gene set analyses were based on a linear model specifying covariates for patient , and a categorical joint factor incorporating time point and viremia status ( Pre-infection , Early acute [positive viremia] , Late acute [positive viremia] , Late acute [negative viremia] , Follow up ) . Differential gene and gene set analyses were adjusted for multiple testing by the method of Benjamini and Hochberg [46] . BTM gene memberships and annotations were obtained from Li et al [11] , and enrichment tests were performed with MROAST [23] and the above described linear model . The MROAST tool was selected due to its capacity for complex experimental designs and voom-generated gene weights . Group level BTM activity scores were derived from the proportion of genes in a given BTM contributing to significance ( i . e . genes in BTM with |z| > √2 ) as reported by MROAST , with sign indicating direction of enrichment relative to Pre-infection baseline . For visualizing BTM changes per individual patients ( Fig 2A ) , sample level fold-change activity scores were calculated as the median log2 fold-changes ( sample time point vs . Pre-infection , per patient ) of BTM member genes for each module . PBMC ISG set [26] enrichment testing was performed with ROAST [23] and CAMERA [47] , using the above described linear model . Genes within the PBMC ISG set found to be differentially expressed ( q value < 0 . 05 , S4 Table ) at the Early acute time point relative to Pre-infection baseline were selected as an “Acute HCV PBMC ISG signature . ” The Acute HCV PBMC ISG signature gene list was intersected with lists of differentially expressed genes in different biological contexts as reported by corresponding publications: Acute HCV , Human Liver [19]; Acute HCV , Chimpanzee liver [25]; Type I IFN and Type II IFN , Human PBMC ex vivo [30] . Microarray data for PBMC responses to the following vaccines were obtained from the GEO database: YFV ( GSE13485 ) [9] , LAIV ( GSE29615 ) and TIV ( GSE29617 ) [48] , MPSV4 ( GSE52245 ) and MCV4 ( GSE52245 ) [11] . In an effort to minimize the impact of different study designs , transcriptome platforms ( RNA-Seq , different microarrays ) and sample sizes , the GSEA ( pre-ranked ) approach was used to evaluate BTM activity in each individual dataset as follows ( S4 Fig ) . For each vaccine dataset , differential gene expression analysis for peak response ( day 7 ) versus pre-vaccination baseline ( day 0 ) was performed with limma in parallel analysis workflows . Only those genes represented in all datasets were maintained for subsequent analysis . For each dataset , genes were ranked by moderated t-statistics , and input to the GSEA ( pre-ranked ) module on GenePattern ( http://genepattern . broadinstitute . org ) . Comparative profiles of BTMs enriched in both Acute HCV and vaccine datasets ( CIRCOS plots ) were defined as the intersect of modules enriched at GSEA q value < 0 . 01 ( concordant direction of up/down regulation ) for each dataset ( for Acute HCV , time point vs . Pre-Infection; for vaccines , peak response versus baseline ) . Microarray data on the peripheral whole blood response to acute DENV infection [15] were obtained from GEO ( GSE51808 ) . Samples without paired convalescent controls were not included in analysis . Principal component analysis was used to identify distinct infection groups based on gene expression patterns: High viral load ( 2–3 days ) , Moderate viral load ( 4–6 days ) , and Low viral load ( 5–9 days ) . Differential gene expression analysis ( acute DENV infection versus paired convalescent “baseline” controls ) for each infection group was performed with limma . BTM profiles for each Acute HCV and DENV group were generated using the GSEA ( pre-ranked ) approach as described for vaccine comparisons . Cryopreserved PBMCs were thawed and analyzed using different panels for T cells and B cells phenotyping against the following markers: CD3 ( clone UCHT1 ) , CD4 ( clone RPA-T4 ) , CD8 ( clone SK1 ) , CD10 ( clone HI10a ) , CD19 ( clone AJ25C1 ) , CD20 ( clone 2H7 ) , CD21 ( clone B-ly4 ) , CD38 ( clone HB7 ) , CD56 ( clone NCAM 16 . 2 ) , IgM ( clone G20-127 ) , HLA-DR ( G46-6 ) , all from BD Bioscience ( San Diego , CA ) ; CD1 ( clone L161 ) , CD27 ( Clone O323 ) both from Thermo-Fisher ( Waltham , MA ) ; CD22 ( Clone HIB22 ) , IgG ( clone M1310G05 ) both from BioLegend ( San Diego , CA ) . HLA-A2/NS3-1073 tetramers ( HLA-A2 restricted HCV-NS3 peptide aa 1073–1081 ( CINGVCWTV ) ) were obtained from the NIH Tetramer Core facility ( Emory University , Atlanta , GA . Multiparameter flow cytometry was performed at the flow cytometry core of the CRCHUM using a BD LSRII instrument equipped with violet ( 405 nm ) , blue ( 488 nm ) , yellow-green ( 561 nm ) and red ( 633 nm ) lasers and FACSDiva version 8 . 0 . 1 ( BD Biosciences ) . FCS data files were analyzed using FlowJo version 10 . 0 . 8 for Mac ( Tree Star , Ashland , OR ) . Fluorescence minus one controls were used to set the analysis gates . CXCL10 levels in plasma were quantified using the human CXCL10/IP-10 Quantikine ELISA Kit ( R&D Systems Inc , Minneapolis , MN ) according to the manufacturer’s protocol . | Hepatitis C virus ( HCV ) is a leading cause of liver disease , with an estimated 71 million people infected worldwide . Following exposure , a subset of individuals spontaneously clears the virus while a majority progress to chronic infection . The immune functions active during the initial , acute infection by HCV are not well understood , due in part to difficulties associated with studying the early stages of this disease in humans . Most individuals acutely infected with HCV are asymptomatic and do not initially seek medical care , which presents challenges in conducting prospective longitudinal studies . Here , by employing specimens from a rare research cohort of individuals sampled before infection , during infection , and after spontaneous viral clearance , we use RNA-Seq to characterize the early immune processes active in acute HCV . Our analysis identified a robust innate antiviral gene signature that corresponds with HCV viremia . In addition , comparisons to other immune transcriptomics datasets demonstrated that the immune response to acute HCV shares many features with responses to flaviviruses . These results offer a detailed longitudinal description of immune function active during the spontaneous resolution of acute HCV infection and provide insight into the early events that may contribute to viral clearance . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"hepacivirus",
"pathogens",
"immunology",
"computational",
"biology",
"microbiology",
"viruses",
"multivariate",
"analysis",
"vaccines",
"mathematic... | 2018 | Longitudinal transcriptomic characterization of the immune response to acute hepatitis C virus infection in patients with spontaneous viral clearance |
The pro-apoptotic proteins Bax and Bak are essential for executing programmed cell death ( apoptosis ) , yet the mechanism of their activation is not properly understood at the structural level . For the first time in cell death research , we calculated intra-protein charge transfer in order to study the structural alterations and their functional consequences during Bax activation . Using an electronegativity equalization model , we investigated the changes in the Bax charge profile upon activation by a functional peptide of its natural activator protein , Bim . We found that charge reorganizations upon activator binding mediate the exposure of the functional sites of Bax , rendering Bax active . The affinity of the Bax C-domain for its binding groove is decreased due to the Arg94-mediated abrogation of the Ser184-Asp98 interaction . We further identified a network of charge reorganizations that confirms previous speculations of allosteric sensing , whereby the activation information is conveyed from the activation site , through the hydrophobic core of Bax , to the well-distanced functional sites of Bax . The network was mediated by a hub of three residues on helix 5 of the hydrophobic core of Bax . Sequence and structural alignment revealed that this hub was conserved in the Bak amino acid sequence , and in the 3D structure of folded Bak . Our results suggest that allostery mediated by charge transfer is responsible for the activation of both Bax and Bak , and that this might be a prototypical mechanism for a fast activation of proteins during signal transduction . Our method can be applied to any protein or protein complex in order to map the progress of allosteric changes through the proteins' structure .
Mitochondrial outer membrane permeabilization ( MOMP ) is a hallmark of programmed cell death ( apoptosis ) . Following MOMP , apoptotic proteins from the mitochondrial inter-membrane space are released , causing the activation of cell death proteases which cleave the cell's cytoskeleton and genetic material . MOMP is executed by the Bcl-2 family proteins Bak and Bax that , upon activation during apoptosis , oligomerize and form pores in the mitochondrial membrane [1]–[4] . Bak and Bax oligomerisation is controlled by the interplay of further Bcl-2 proteins [5]–[8] . While pro-survival Bcl-2 proteins bind to and deactivate Bak and Bax [9] , other apoptotic Bcl-2 proteins de-repress this inhibition , leaving Bak and Bax free to oligomerize [10] . Nevertheless , a separate step , whereby a subclass of apoptotic Bcl-2 proteins such as Bim and Bid directly activate Bak and Bax , was proposed to be required for oligomerization [11]–[13] . The activation steps required for Bax oligomerization were extensively investigated [14]–[18] . These steps were found to comprise Bax translocation from the cytosol to the mitochondrial membrane , and changes of Bax conformation . Conformational changes of Bax include exposure of its C-domain , insertion of this C-domain into the membrane , and exposure of the Bax BH3 domain , one of four homology domains of Bcl-2 proteins ( Figure 1 ) . In inactive Bax , the C-domain is tightly bound inside a hydrophobic pocket which we henceforth denote as the ‘BH groove’ . This tight binding was suggested to increase the solubility of Bax and to keep Bax in the cytosol in the absence of stress [16] . Gavathiotis et al . [18] synthesized a helix mimicking the BH3 domain of the activator Bim ( Bim-stabilized α-helix of Bcl-2 domains , Bim-SAHB ) . They subsequently performed NMR spectroscopy to study the interaction of Bax with the Bim-SAHB activator . They found that , in the absence of Bim-SAHB , the Bax activation site was blocked by a largely unstructured loop ( loop 1–2 ) , which opens upon incubation with Bim-SAHB . Using Bax mutants with reduced loop 1–2 mobility , Gavathiotis et al . later demonstrated that the opening of this loop was a prerequisite for Bax activation [19] . Interestingly , the suggested Bax activation site and the Bax C-domain are separated by over 25 Å . Since the binding of Bim-SAHB to Bax is weak and transient , and neither significant disturbances in the helical packing , nor covalent modifications have been observed in Bax upon activation , the mechanism of how C-domain exposure occurs following this activation remains elusive [20] , [21] . Charge transfer was found to be significant in many biomolecular interactions [22]–[24] , and functionally linked to protein structural dynamics [25] . In this paper , we therefore investigate the role of charge transfer during Bax activation by employing an electronegativity equalization model for the calculation of atomic charges . Following our investigation , we propose that a charge transfer network is intimately connected to the way that the activation information travels across Bax , and that a similar network is plausible in Bak .
The Electronegativity Equalization Method ( EEM ) [26] is a fast technique for estimating partial atomic charges , and has been successfully applied to zeolites , small organic molecules and polypeptides [27]–[31] . To use EEM for studying charge transfer during Bax activation , EEM model parameters need to be calibrated to charges of reference molecules . For this purpose we followed our previously published EEM model calibration procedure [32] , with a few modifications that address the complex nature of proteins . The reference data consisted of atomic charges for molecules of two disjoint reference sets RS1 and RS2 , and these charges were calculated using the quantum mechanics ( QM ) scheme detailed in the Methods section . Since calculation of QM charges for large molecules such as proteins would require too high computational costs , previous EEM models available in literature were calibrated to reference charges from small molecules [32]–[38] . Moreover , to make EEM as generally applicable as possible , these calibrations used mostly inorganic or drug-like compounds , which do not reflect the complex nature of proteins as long , non-neutral molecular chains with complex 3D assembly . Therefore , to retain properties that are characteristic for proteins and allow a fast calculation of reference QM charges at the same time , large fragments of experimentally determined protein structures were used as reference sets in the present study ( see the Methods section for details ) . We next determined values for the EEM model parameters by fitting them to the reference QM charges using a least squares algorithm . Prior to fitting , we classified atoms according to two schemes . One scheme was based on chemical elements only ( denoted ‘E’ ) , and the other on chemical elements and maximum bond order for each atom ( denoted ‘EX’ , so that , for example , ‘O1’ indicates simple bonded , and ‘O2’ double bonded oxygen ) . Fitting the model parameters for each of the two atom classification schemes and each of the two reference sets of atomic charges , we obtained four parameter sets , denoted RS1-E , RS1-EX , RS2-E , RS2-EX . Finally , we validated our EEM model by assessing the accuracy of the model in reproducing the original QM charges from reference sets RS1 and RS2 , and from five additional test molecules T1-T5 . Results were evaluated by the average correlation coefficient Ravg ( squared Pearson's correlation coefficient ) , the root mean square deviation RMSDavg , and the average absolute difference Davg . An overview of the EEM model calibration procedure is given in Figure 2 , and the complete details can be found in the Methods section . Overall , the results in Figure 3 suggested that the finer grained atom classification scheme ‘EX’ only modestly improved the accuracy compared to the scheme ‘E’ based on chemical elements alone . The good agreement between QM and EEM charges for all data sets suggested that both atom classification schemes can provide satisfactory calibration results . Moreover , our model was able to compute EEM atomic charges in less than 1 second for any of the reference or test molecules using our previously published implementation [39] . Having developed an EEM based method for rapid calculation of atomic partial charges , we investigated whether atomic charge distribution prior and subsequent to Bax activation would reveal any clues about the mechanisms of the activation . To this end , we obtained the 3D structure of inactive Bax ( Figure 1B ) , and of active Bax in complex with the activator peptide Bim-SAHB ( Figure 1C ) from the Protein Data Bank ( PDB IDs 1F16 [16] and 2K7W [18] respectively ) . We then computed EEM atomic charges using parameter set RS2-E ( Figure 2B ) for both structures , and assessed the absolute charge transfer per residue ( total difference in charge per amino acid residue , ΔQres ) , and the intra-residue charge density reorganization ( root mean square deviation in charge per residue , RMSDres ) . The mathematical derivation of these descriptors can be found in the Methods section , and their values for all Bax residues are available in Table S1 . Experimental evidence suggests that , in inactive Bax , the C-terminal helix is bound tightly to its hydrophobic pocket ( ‘BH-groove’ ) . During activation , this binding gets destabilized , causing the C-domain to subsequently vacate the BH-groove and insert into the mitochondrial outer membrane . Early mutagenesis studies revealed a critical interaction between residues Ser184 and Asp98 at the C-domain-BH-groove interface , whose abrogation is sufficient to immediately activate Bax [16] , [40] . We therefore focused on the changes in charge density distribution in the vicinity of this interaction . While our calculations did not show any change in the charge profile of Ser184 , they indicated that any interaction that might have taken place between Asp98 and Ser184 in the inactive structure has been replaced by an Asp98-Arg94 salt bridge in the active structure ( Figure 4 ) . Upon activation , Arg94 becomes more positive ( see also Table S1 ) , which is suggested to lead to the recruitment of Asp98 , the abrogation of the Asp98-Ser184 interaction , and ultimately the destabilization of the C-domain . This demonstrates that the binding of Bim-SAHB to Bax can activate Bax by destabilizing the interaction between the Bax C-domain and its binding groove . It remains puzzling how the BH groove is influenced by the binding of Bim-SAHB to Bax , given that this interaction occurs on the opposite side of the Bax molecule , at a distance of 25 Å from the BH groove . Interestingly , the residues which showed a transfer of charge one standard deviation higher than average ( Table S1 ) provided a clue as to how the activation information proceeds through the protein . Foremost , significant changes in the net residue charges were found at the Bax activation site , the BH3-domain ( required for oligomerization ) and the C-domain ( required for membrane insertion ) . Since these are all functional sites of Bax , these changes were not unexpected . For example , George et al [41] found that a triple alanine mutant at residues 63–65 ( on the BH3 domain of Bax ) ablated Bax oligomerisation and apoptotic activity , which correlates perfectly with the high charge transfer we found on residues 64 and 65 upon Bax activation ( Table S1 ) . However , in addition to the expected changes , our method surprisingly identified significant charge transfer also on the central helix , inside the hydrophobic core of Bax ( residues Trp107 , Arg109 and Lys119 on helix 5 ) . The presence of significant charges and charge transfer in a predominantly hydrophobic environment suggests that helix 5 acts as a hub which collects and distributes charge density ( Figure 5 ) . We further calculated the intra-residue redistributions of charge density upon activator ( Bim-SAHB ) binding . Significant such redistributions were observed at the Bax activation site , BH groove and loop 1–2 . Since these are the functional regions of Bax , these calculations provide further support for the notion of a charge transfer network that conveys information across the entire Bax molecule ( Figure S1 , Table S1 ) . Hints of such an interaction transfer phenomenon were found by Gavathiotis et al . [19] . They titrated Bax with increasing amounts of Bim-SAHB and observed small , but reproducible dose-responsive changes in NMR resonance behavior for the backbone N atoms of residues on the Bax C-domain , as well as on helix 5 inside the hydrophobic core of Bax . They concluded that the binding of the activator induces reverberations in the core of the Bax protein , which serve to mobilize the C-domain ( allosteric sensing ) . Our charge analysis explains these reverberations by a network of charge transfer through the entire Bax molecule . Unlike Bax , Bak is present at the outer mitochondrial membrane in absence of apoptotic stimuli . Evidence suggests that the inactive form of Bak gets recruited to the mitochondrial outer membrane and forms complexes with the membrane protein VDAC2 . Upon apoptotic stimuli , pro-apoptotic Bcl-2 proteins such as Bid transiently bind to Bak . This binding breaks down the VDAC2/Bak complex and exposes the BH3 domain of Bak , which is essential for Bak oligomerization [42]–[46] . As the activation information may be conveyed by a similar charge transfer network to induce abrogation of VDAC2/Bak binding , we wondered whether a comparable transfer hub may exist also in Bak . Since residues that are essential for functionality are most often conserved in proteins with similar functions , we therefore first performed the sequence alignment of Bax and Bak . While the sequence identity between the two proteins was rather low ( ClustalW2 score 19% , see Figure S2 ) , we found that the residues involved in the charge transfer network in Bax were conserved in Bak . These homologous residues were Trp125 , Arg127 and Arg137 ( Figure 6A ) . We subsequently compared the 3D structures of Bax and Bak ( PDB ID 2IMT [47] ) , and found that above Bak residues were organized in a very similar manner to their Bax homologues ( Figures 5B and 6B ) . These findings suggest that the mechanism of charge transfer via the hydrophobic core of Bax is also plausible for Bak , and that similar residues may also play an essential role during Bak activation .
Allosteric proteins are characterized by a regulatory site that is distinct and often well distanced from the protein's active site . Regulation of the protein's activity which occurs via this distinct site is termed allosteric regulation . Recent reports indicate that allosteric regulation is particularly important during cell signaling processes , where it has been shown to stabilize receptor proteins , or to be responsible for the rapid , stress induced release of dormant signaling proteins bound to the cytoskeleton [48] , [49] . An interesting structure-function analysis of Bax performed by George et al . [41] concluded that monomeric Bax may be held in an inactive conformation by multiple helices in the absence of stress , and that Bax may be activated through perturbation at multiple sites . Nevertheless , later Gavathiotis et al . identified a unique and well defined activation site on Bax [18] , and subsequently demonstrated that binding of an activator BH3 peptide induces reverberations in the core of the Bax protein , a phenomenon they named allosteric sensing [19] . The present study found that this allosteric regulation is mediated by a charge transfer network , which conveys the activation information from the Bax activation site to the functional regions of Bax without compromising the structure of the BH groove ( essential for pro-apoptotic activity ) . As charge transfer is significantly faster than domain rearrangements , the charge transfer mediated alosteric regulation in Bax also allows for a swift control of the apoptotic fate [50] . In addition to suggesting that charge transfer mediated allosteric regulation is responsible for Bax activation by pro-apoptotic Bcl-2 proteins , our charge profile analysis also indicated several residues that actively mediate this charge interaction , providing an opportunity for further in-depth mutagenesis studies or even pharmacological intervention . We first confirmed that the abrogation of the Asp98-Ser184 interaction , which has been reported to be responsible for the mobilization of the C-domain from the BH groove [16] , [40] , indeed occurs during Bax activation . We propose that Arg94 plays an essential role in this abrogation , as it can sequester Asp98 and prevents the formation of the stabilizing Asp98-Ser184 interaction in active Bax . Indeed , previous mutational studies [41] showed that a triple alanine Bax mutant at residues 92 to 94 is biologically inactive , supporting our finding that residue Arg94 plays a role in Bax activation . Furthermore , we found that helix 5 acts as a central hub for the charge transfer network in Bax . We identified three residues , Trp107 , Arg109 and Lys119 , that may act as the main mediators of this charge transfer . Helix 5 has been found to react to Bim-SAHB binding in NMR experiments [19] . It was then found that the Bim-SAHB-induced opening of the Bax loop 1–2 is essential for Bax activity , and that this opening induces reverberations in the protein's hydrophobic core . A deeper look at the NMR data from their supplement ( Figure S1D from [19] ) reveals that activator binding induces pronounced chemical shifts in the Bax backbone N atoms in the area of residue Trp107 even when the mobility of loop 1–2 is restricted by chemical tethering . In Figure S1B from the same publication [19] , we observe that opening this loop similarly affects the backbone N atoms in the vicinity of Lys119 . While the authors [19] did not explicitly focus on these residues , our charge calculations make it reasonable to assume that they are indeed important for allosteric Bax activation . Moreover , another study [41] found that triple alanine Bax mutants at residues 109–111 or 118–120 showed decreased biological activity in the presence of the activator tBid . Therefore , influencing the activity of Trp107 , Arg109 or Lys119 may readily influence the biological activity of Bax . Because of their positioning , residues Trp107 and Arg109 are easily accessible and therefore excellent drug targets . The results of our investigation agree well with the mutational study of George et al [41] , in that helix 5 is a central mediator of Bax activity . Both studies further agree that Arg94 is essential for Bax oligomerisation , and that residues Arg109 or Lys119 may influence the biological activity of Bax . In addition , George et al . suggested that the block of four central residues ( 113–116 ) is mandatory for Bax activity . Comparatively , the amount of charge transferred by these 4 residues upon Bax activation was only slightly above average ( Table S1 ) . Nevertheless , we note that this block of residues resides at the centre of the Bax molecule and is very bulky , and therefore these residues are very likely essential for maintaining the helical fold of Bax . Therefore , the observation of George et al . that a quadruple mutation in the centre of the Bax molecule impairs biological activity can be easily explained by change of stability of the helical fold , rather than by a disruption of the Bax activation mechanism . Finally , by performing sequence and structural alignment of Bax and Bak , we identified that Bak residues Trp125 , Arg127 , and Arg137 potentially constitute a similar hub of charge transfer inside the Bak protein . The membrane protein VDAC2 was reported to recruit Bak to the mitochondrial membrane in the absence of apoptotic stimuli . Upon apoptotic stimuli , this VDAC2/Bak binding can be abrogated by Bcl-2 proteins that transiently bind to Bak [42]–[46] . Cheng et al . [42] showed that Bak mutations at residues Leu78 ( within Bak's BH3 domain ) , Trp125 , Gly126 and Arg127 ( on Bak's central helix ) impair complex formation with VDAC2 , and thus concluded that VDAC2 binding to Bak must occur in the proximity of the above mentioned residues . Having identified those residues that may constitute a transfer charge hub during Bak activation , we propose a charge transfer mediated allosteric activation mechanism of Bak that is similar to that of Bax . We propose that transient activator binding in the vicinity of Arg137 is transmitted through allosteric sensing to the VDAC2 binding site , which is in the neighborhood of Trp125 and Arg127 . Subsequently , Bak disengages VDAC2 , exposes its BH3 domain and oligomerizes . Since the relevance of residue Arg137 has not been assessed to date , its investigation may further our understanding of the details of Bak activation and its de-repression of VDAC2 . Understanding the structural changes of Bax and Bak during apoptosis provides important insights into the mechanisms of Bax and Bak activation , which steps and key players are involved , how aberrant protein folding or mutations may influence the protein's function , and how their activation may be influenced by inhibitors or synthetic drugs . While X-ray crystallography and NMR spectroscopy provide excellent experimental techniques to obtain 3D-structures , further theoretical data analysis tools are needed to obtain better mechanistic and functional insights into the structural aspects of protein activation . We report here the first successful application of the Electronegativity Equalization Method to study protein activation during programmed cell death , which enabled us to detect and track the allosteric effects responsible for Bax activation by BH3-only proteins . We have thus shown how knowledge of atomic charges can yield insight into biological phenomena even without further simulations or intricate computations . Moreover , the methodology we developed is directly applicable to other molecular systems , and thus of interest in biomedical and pharmacological research .
The Electronegativity Equalization Method ( EEM ) [26] was employed here to estimate the charge transfer upon Bax activation using structural data obtained from the Protein Data Bank ( PDB ) . EEM enables the determination of connectivity- and geometry- dependent atomic charges . Various formalisms are available in literature [32]–[38] , [51]–[53] . In the present implementation , we focused on the original work by Mortier et al . [26] with a minor modification as suggested by Yang and Wang [54] and described below . Besides enabling a fast and versatile calibration , this formalism estimates atomic charges via a set of coupled linear equations which can be efficiently solved by a Gaussian elimination procedure ( see below ) . EEM is based on the Electronegativity Equalization Principle [55] , which was proven within the Density Functional Theory [56] and which states that electronegativity is equalized throughout a molecule ( ) . The electronegativity of each atom i in this molecule can be approximated as a linear function of several terms: The first term is the effective electronegativity ( i . e . , the electronegativity of the neutral atom , corrected for the presence of the molecular environment ) . The second term is the charge of the atom qi multiplied by its effective hardness ( i . e . , the hardness of the neutral atom corrected for the presence of the molecular environment ) . Hardness was defined by Parr and Pearson [57] as the second derivative of the energy with respect to the charge , and can be thought of as the resistance to change in charge . The last term accounts for the electrostatic interaction with every other charged atom j in the molecule . k is an adjusting factor first introduced by Yang and Wang [54] . Setting and , the molecular electronegativity can be written as: Considering the total molecular charge Q to be the sum of all partial atomic charges qi ( ) , a system of equations results , from which the partial atomic charges qi and the molecular electronegativity can be calculated: The corrections for electronegativity and hardness cannot be measured [26] . Therefore , the effective electronegativity and hardness contributions given by and respectively were calibrated in this study . The additional parameter k was also determined , as it allows for a computationally cheap and straightforward sampling of the ( A , B ) parameter space , as previously demonstrated by Svobodová Vařeková et al . [32] . The EEM equation of the molecular electronegativity can be rearranged as a linear equation in A and B for each atom in the system: The above linear equations can be grouped together according to the type of atom they refer to , as each parameter will be valid only for a particular type of atom . The classification of atoms into types can be done according to various criteria . As schemes in literature use different levels of granularity [e . g . ] , [ 32] , [33 , 35] , two schemes of atom classification were tested in the present study ( see also the Results section ) . For each atom type , the parameters A and B can be determined by least squares minimization , provided that the values of all the other variables in the equation are known . Here the interatomic distances were calculated from the 3D atomic coordinates . The reference atomic charges were calculated by the QM scheme described below . For each molecule , the value of the global electronegativity was approximated as the harmonic average of the electronegativities of its constituent atoms [58]:where n is the number of atoms in the molecule , and the values of correspond to Pauling electronegativities [59] , [60] . The extra parameter k present in this particular formalism was sampled on several intervals . For each discrete value of k , the least squares minimization was performed in order to obtain the ( A , B ) x parameters , where x goes over all atom types considered . Upon internal validation , the result was the set of parameters [k , ( A , B ) x] which gives the best Ravg ( see below ) between the reference QM values and the predicted EEM values for atomic charges . A scheme of the calibration step is given in Figure 2B , while a detailed description of this procedure can be found in the work of Svobodová Vařeková et al . [32] . Obtaining appropriate reference data is essential for the accuracy and applicability of a predictive model . The reference data used in this study consisted of atomic charges for molecules of two disjoint reference sets RS1 and RS2 , and these charges were calculated using quantum mechanics ( QM ) ( see below ) . These reference molecules were fragments extracted from calcium containing proteins which were obtained from PDB and whose structures had been determined by X-ray crystallography or solution state NMR experiments . Each of the fragments consisted of amino acid chains , calcium ions and water molecules , and was obtained from the 3D structure of its parent protein using the program Triton [61] . The fragments were curated manually to ensure that , while they are sufficiently small for QM calculations , they remained biochemically meaningful . For each fragment , reference QM atomic charges were obtained from a Mulliken population analysis performed at the HF/6-31G* theory level using the program Gaussian 03 [62] . An overview of the composition of all fragments used for EEM model calibration is given in Table 1 , while the 3D structures of these fragments are available online in PDB format at www . ncbr . muni . cz/~ionescu/Supporting_Data_Sets . zip . Reference sets RS1 and RS2 were used for model calibration , and internal and external validation ( see below ) . Five additional test molecules T1-T5 were used for external validation . A brief summary regarding the reference data is given in Figure 2A . The accuracy of the EEM models in reproducing the original QM charges from reference sets RS1 and RS2 , and from five additional test molecules T1–T5 was evaluated by internal and external validation . In the internal validation step , the charges predicted by the EEM model with parameter sets RS1-E and RS1-EX ( RS2-E and RS2-EX respectively ) were compared against QM charges from the associated reference set RS1 ( RS2 respectively ) . In the external validation step , EEM and QM charges were compared for five test molecules T1–T5 which were not contained in the original reference sets RS1 and RS2 , but were obtained in a similar manner . Since the reference sets RS1 and RS2 were disjoint , two further external validations were performed . Therefore , EEM charges obtained by using parameter sets RS1-E and RS1-EX ( RS2-E and RS2-EX respectively ) were compared against QM charges from the non associated reference set RS2 ( RS1 respectively ) . A schematic representation of the EEM model validation step is given in Figure 2C . The correlation between the sets of QM and EEM charges was assessed by three indicators . The first indicator was the average correlation coefficient ( squared Pearson's correlation coefficient ) , computed for each molecule , and averaged over all molecules in a set:where the index i = 1 …N described all atoms in molecule I , and represented average atomic charges in molecule I , and were standard deviations of the atomic charges in molecule I , nI was the number of atoms in molecule I , and N was the number of molecules in a given set . The second indicator was the root mean square deviation , computed for each molecule , and averaged over all molecules in a set: The third indicator was the average absolute difference , computed for each molecule and averaged over all molecules in a set: The EEM charge calculations for both Bax structures were done using the program EEM_SOLVER [39] which implemented the above mentioned EEM formalism and employed the parameter set RS2-E developed in the present study . Two indicators were employed in order to quantify the changes in the charge profile of Bax upon activation . The first indicator was the total difference in charge per amino acid residue:where denoted atomic charges in the active Bax , denoted atomic charges in the inactive Bax , and nres was the number of atoms in the residue . ΔQres assessed the amount of charge that had been transferred to or from each residue . The second indicator was the root mean square deviation in atomic charge per residue:RMSDres assessed the intra-residue charge density redistributions . The Bax/Bak sequence alignment was done for the UniProtKB/Swiss-Prot entries Q07812 ( BAX_HUMAN ) and Q16611 ( BAK_HUMAN ) , and was performed using ClustalW2 with default parameters on the EBI server [63] . The structural models were visualized using VMD [64] . | Apoptosis is a physiological form of cell death that is fundamental for development , growth and homeostasis in multi-cellular organisms . Deviations in the apoptosis machinery are known to be involved in cancer , neurodegenerative disorders , and autoimmune diseases . The proteins Bax and Bak are essential for executing apoptosis , yet the mechanism of their activation is not properly understood at the structural level . To understand this mechanism , we investigated how the electronic density is reorganized ( i . e . , how charge is transferred ) inside the Bax molecule when Bax binds a functional peptide of its natural activator protein . We identified the specific interactions responsible for the exposure of the functional sites of Bax , rendering Bax active . Furthermore , we found a network of charge transfer that conveys activation information from the Bax activation site , through the hydrophobic core of Bax , to the well-distanced functional sites of Bax . This network consists of three residues inside the hydrophobic core of Bax , which are present also in the hydrophobic core of Bak , suggesting that these residues are functionally important and thus potential drug targets . We provide a straightforward and accessible methodology to identify the key residues involved in the fast activation of proteins during signal transduction . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"empirical",
"methods",
"biochemistry",
"protein",
"chemistry",
"computational",
"chemistry",
"protein",
"interactions",
"proteins",
"ab",
"initio",
"methods",
"protein",
"structure",
"regulatory",
"proteins",
"macromolecular",
"assemblies",
"biology",
"chemistry",
"drug",
... | 2012 | Charge Profile Analysis Reveals That Activation of Pro-apoptotic Regulators Bax and Bak Relies on Charge Transfer Mediated Allosteric Regulation |
Aneuploidy is known to be deleterious and underlies several common human diseases , including cancer and genetic disorders such as trisomy 21 in Down's syndrome . In contrast , aneuploidy can also be advantageous and in fungi confers antifungal drug resistance and enables rapid adaptive evolution . We report here that sexual reproduction generates phenotypic and genotypic diversity in the human pathogenic yeast Cryptococcus neoformans , which is globally distributed and commonly infects individuals with compromised immunity , such as HIV/AIDS patients , causing life-threatening meningoencephalitis . C . neoformans has a defined a-α opposite sexual cycle; however , >99% of isolates are of the α mating type . Interestingly , α cells can undergo α-α unisexual reproduction , even involving genotypically identical cells . A central question is: Why would cells mate with themselves given that sex is costly and typically serves to admix preexisting genetic diversity from genetically divergent parents ? In this study , we demonstrate that α-α unisexual reproduction frequently generates phenotypic diversity , and the majority of these variant progeny are aneuploid . Aneuploidy is responsible for the observed phenotypic changes , as chromosome loss restoring euploidy results in a wild-type phenotype . Other genetic changes , including diploidization , chromosome length polymorphisms , SNPs , and indels , were also generated . Phenotypic/genotypic changes were not observed following asexual mitotic reproduction . Aneuploidy was also detected in progeny from a-α opposite-sex congenic mating; thus , both homothallic and heterothallic sexual reproduction can generate phenotypic diversity de novo . Our study suggests that the ability to undergo unisexual reproduction may be an evolutionary strategy for eukaryotic microbial pathogens , enabling de novo genotypic and phenotypic plasticity and facilitating rapid adaptation to novel environments .
Aneuploidy , a condition in which cells have an abnormal number of chromosomes , can cause deleterious effects in organisms throughout the eukaryotic tree of life . Aneuploidy underlies several common human genetic diseases , including trisomy 21 in Down syndrome and trisomy 13 in Patau syndrome [1] , [2] , and is detected in more than 90% of solid tumors [3] , [4] . The presence of even a single extra chromosome in primary mouse embryonic fibroblasts ( MEFs ) results in proliferative defects and metabolic aberrations [5] . In Caenorhabditis elegans and Drosophila melanogaster , aneuploidy is often lethal [6] , [7] , and in budding and fission yeasts , aneuploidy can inhibit cellular proliferation [8] , [9] . However , aneuploidy can be advantageous in fungi by conferring antifungal drug resistance and enabling rapid adaptive evolution . Aneuploidy evokes transcriptomic and proteomic changes in the model yeast Saccharomyces cerevisiae [8] , [10] . Mutations in a deubiquitinating enzyme of S . cerevisiae , which arose during the evolution of an aneuploid isolate , leads to improved proliferation of many aneuploid strains , likely by promoting degradation of aberrant proteins produced in perturbed stoichiometric ratios [11] . Rancanti et al . found that aneuploidy facilitated adaptive evolution in yeast cells lacking the conserved motor protein Myo1 involved in cytokinesis [12] . Moreover , it has been found that chromosomal duplication may confer a selective advantage to S . cerevisiae under stress conditions by promoting genomic instability and mutation , and although it is a transient solution , the short-lived aneuploid intermediate may also serve as a capacitator of evolution [13] , [14] . In the human pathogenic fungi Candida albicans and Cryptococcus neoformans , aneuploidy can confer resistance to commonly used antifungal drugs such as fluconazole [15]–[18] . In C . albicans , haploids can even arise from concerted chromosome loss from diploid progenitors [19] . In C . albicans , an isochromosome 5 can arise in response to fluconazole treatment and confers drug resistance because the left arm of Chr 5 encodes Erg11 ( lanosterol 14α demethylase ) , the target of fluconazole , and Tac1 , a transcription factor that activates expression of drug export pumps [15] , [17] . Similarly in C . neoformans , Chr 1 disomy confers azole resistance because this chromosome harbors ERG11 and AFR1 ( which encodes the major azole efflux pump ) and aneuploidy can also influence the virulence of this pathogen [16] , [20] , [21] . Interestingly , a similar drug-resistance phenotype has been observed in the protozoan parasite Leishmania , in which resistance to front line antimonial-based drugs similarly emerges via aneuploidy [22] , [23] . C . neoformans is a globally distributed human fungal pathogen that causes life-threatening meningoencephalitis [24] . Cryptococcus predominantly infects individuals with compromised immunity , such as HIV/AIDS patients . The United States Centers for Disease Control ( CDC ) has reported that C . neoformans causes more than one million cases of cryptococcosis annually with more than 620 , 000 attributable mortalities , resulting in approximately one-third of all AIDS-associated deaths [25] . This fungal pathogen has now surpassed tuberculosis as a common cause of death in Africa . C . neoformans is a basidiomycetous fungus that usually grows in the environment as a haploid , budding yeast . This species has a bipolar mating system with two mating types: a and α [26]–[28] . In response to a variety of environmental conditions , a and α cells secrete lipid-modified pheromones that induce cell–cell fusion , and the resulting dikaryon undergoes a dimorphic transition to hyphal growth [28]–[31] . Ultimately , the hyphal tips form basidia fruiting bodies wherein nuclear fusion and meiosis occur , and multiple rounds of mitosis and budding result in the production of four long chains of infectious spores decorating each basidium . These spores are readily aerosolized and cause infections in humans and animals when inhaled [32] . While C . neoformans has a defined a-α opposite sexual cycle , >99% of natural isolates are of the α mating type [33] . C . neoformans var . neoformans ( serotype D ) α cells can undergo α-α unisexual reproduction to generate spores under laboratory conditions [34] . Recent population genetic studies provide evidence that multiple pathogenic lineages of C . neoformans var . grubii ( serotype A ) and the sibling species responsible for the Vancouver outbreak Cryptococcus gattii ( serotype B and C ) undergo α-α sexual reproduction in nature , but this remains to be documented under laboratory culture conditions [30] , [35]–[41] . Under nutrient-limiting conditions α haploid cells form a diploid or a haploid monokaryotic hyphae . The haploid hyphae grow to form basidia where a late diploidization event occurs . In the diploid hyphae diploidization of the nuclear content may be induced early and results in a diploid monokaryon that initiates hyphal growth and the formation of apical basidia . Early diploidization may occur through endoreplication or cell–cell fusion . Cell fusion , followed by nuclear fusion , may be induced in ménage à trois matings where a cells serve as pheromone donors to stimulate α-α fusion between genetically different or clonal cells [34] . Endoreplication may occur in cells that undergo DNA replication without cell division or they may undergo nuclear division and then fusion . In all cases , the resulting diploid nucleus undergoes meiosis and multiple rounds of mitosis generate the meiotic progeny , the basidiospores . In previous studies , we found that the key meiotic regulators , the endonuclease Spo11 and the meiotic recombinase Dmc1 , are required for spore production and germination during unisexual reproduction [34] , [42] . Deletion of SPO11 or DMC1 severely impairs sporulation and the fewer meiotic progeny generated are largely inviable , providing further evidence that unisexual reproduction is a meiotic process [34] , [42] . Remarkably , unisexual reproduction has been recently discovered to occur in another common systemic human fungal pathogen , C . albicans [43] . C . albicans is known to undergo a parasexual cycle involving heterothallic fusion of α and a mating type cells followed by stochastic , concerted loss of chromosomes [44] . Similarly , fusion of α-α and a-a cell unions via homothallic parasexual reproduction occurs when a pheromone-degrading protease ( Bar1 ) is inactivated or in ménage à trois matings , in which a third mating partner serves only as the pheromone donor [43] . Sex is costly , and thus the question arises as to why C . neoformans or C . albicans would undergo inbreeding/selfing unisexual reproduction , which would limit the amount of genetic diversity inherited from the parents , as opposed to a-α sexual reproduction , which promotes outcrossing and genetic exchange . We hypothesize that unisexual reproduction is a hypermutagenic process that generates genetic diversity de novo and that the resulting progeny can thereby more rapidly adapt to changing environments than cells produced asexually by mitosis . To test this hypothesis , we isolated progeny generated via selfing α-α unisexual reproduction and subjected them to phenotypic and genotypic analyses . Remarkably , we found that α-α unisexual reproduction generates frequent phenotypic diversity , including temperature sensitivity , fluconazole resistance or sensitivity , and increased melanin production among other novel traits . Further comparative genomic hybridization ( CGH ) analysis revealed that the majority of phenotypically diverse progeny are aneuploid . Phenotypic diversity is caused by this aneuploidy , as loss of the aneuploid chromosomes restored both euploidy and the wild-type parental phenotype . Aneuploidy and phenotypic variation was also found to occur at a similar rate following a-α bisexual reproduction but not as a result of mitotic asexual vegetative growth . Our findings show that sex can generate phenotypic and genotypic diversity de novo in the pathogenic yeast C . neoformans with implications for other eukaryotic microbes and pathogens , including other fungi and parasites that are common pathogens of humans .
To investigate whether phenotypic and genotypic changes are generated by selfing unisexual reproduction , we analyzed the meiotic progeny produced via this process . To this end , we solo cultured the promiscuous , hypersexual haploid strain XL280α under conditions that support robust α-α unisexual reproduction ( V8 media or FA ( filament agar ) for ∼2 weeks ) and isolated spores by microdissection . The hyperfilamentous haploid strain XL280α [34] , [45] generates abundant hyphae , homozygous diploid intermediates , and haploid meiotic spores when grown on mating-inducing media all by itself . Xl280 is a haploid F1 progeny descended from a cross between two exceptionally well-validated haploid sibling parental strains , whose genomes have been both sequenced , B3501α and JEC20a ( which is isogenic to the sequenced strain JEC21α , with the exception of the mating-type locus ) ( Figure 1A ) [34] , [45]–[47] . Thus , while XL280α shares markers with both parents , it only has one copy of each chromosome ( it is haploid ) and is not heterozygous anywhere in its genome . When it undergoes unisexual reproduction with itself , it forms a transient homozygous diploid ( generated through either α-α cell fusion between mother and daughter cell or via endoreplication ) that is identical throughout the diploid , euploid genome , in which each gene is present in two identical copies . Hyphae produced by unisexual reproduction grow to form terminal basidia wherein the diploid nucleus undergoes meiosis to generate abundant recombinant progeny . That this is a meiotic process has been established in previous studies [34] , [42] that showed both key meiotic genes SPO11 and DMC1 are required for spore production and viability . The parental strains B3501α and JEC20a share ∼50% genetic identity [46]; thus , their F1 progeny ( XL280 ) would be expected to share ∼75% genetic identity with both B3501α and JEC21α , whose genomes have been sequenced [46] . Therefore , we used the available JEC21 genomic sequence and genome microarray slides as a foundation to study the genome of XL280 and its α-α unisexual reproduction [48] , [49] . We performed next-generation sequencing ( NGS ) using high-throughput Illumina sequencing to compare the genomes of XL280α and JEC21α [50] . In total , 47 , 891 , 302 sequences ( paired-end reads ) of 75 bp in length were generated , providing 160-fold coverage of the XL280 genomic sequence , which was then analyzed using the known JEC21 genome as the reference for the sequence assembly ( Figure S1 and Table S1 ) . All but 4 , 535 , 426 sequence reads ( ∼10% ) were used in the genome assembly , and of these remaining reads , the majority were of low quality ( 90% ) . We assembled the Illumina reads using a combination of de novo assembly ( Velvet , [51] ) and reference genome assembly ( BWA , [52] ) ( see Figure S1 and Materials and Methods for detailed procedure ) . XL280 shares 100% identity with JEC21 over 81% of the overall genome and is 99 . 88% identical at the sequence level overall ( Figure 1C ) . Among the SNPs and small indels identified , 9 , 021 were located in exonic regions , 10 , 254 in intergenic regions , nine in tRNA genes , and 3 , 849 in intronic regions ( 31 of which differed in the 5′ splice-site and 23 in the 3′ splice-site ) ( Table S2 ) . Among the exonic SNPs and indels , 107 SNPs resulted in the introduction or deletion of stop codons , 68 indels resulted in a frame shift , 14 indels resulted in the loss of amino acids without a frame shift , and 3 , 943 SNPs resulted in nonsynonymous amino acid substitutions in 960 ORFs ( Table S2 ) . Further analysis of the modified ORFs will enable the elucidation of the phenotypic differences between XL280 and JEC21 , for example with respect to the hypersexual phenotype [45] . CGH was performed using a 70-mer oligonucleotide microarray covering all predicted ORFs of the JEC21 genome [49] to detect chromosome copy number alterations . Based on CGH , the genomes of XL280 and JEC21 ( Figure 2A ) are quite similar and span 14 conserved , linear chromosomes . This enabled use of JEC21-based microarray slides for CGH analysis of XL280 unisexual progeny . Because this microarray covers only ORFs , no differences could be detected in centromeric or telomeric regions . There is an ∼28 kb region absent in XL280; in JEC21 this sequence lies near the right end of Chr 5 and is conserved in many C . neoformans strains [53] . In addition , XL280 is missing one copy of the left end of either Chr 8 or 12 , which is a 63-kb duplicated region located on both Chr 8 and 12 of JEC21 [54] . NGS data further confirmed these findings ( Figure 2B ) . The chromosome content of strain XL280 was compared to JEC21 by clamped homogenous electric field ( CHEF ) gel electrophoresis to detect any chromosomal translocations . Chr 5 and Chr 8/9 of XL280 are smaller than those of JEC21 , while Chr 13 of XL280 is larger ( Figure 2C ) . To ascertain where the sequence similarity of these chromosomes lies , we excised each chromosomal band and performed band array analysis . The configurations of Chr 9 and 12 differed between XL280 and JEC21 ( Figure S2 ) . Chr 9 of XL280 hybridized to Chr 8 and Chr 12 of JEC21; Chr 12 of XL280 hybridized to Chr 8 of JEC21 . These findings were confirmed by Southern hybridization ( Figure 2D ) . Previous studies reported that this chromosomal translocation and duplication occurred during generation of the JEC21α/JEC20a congenic strain pair [54] . During the cross of strains B3501α×B3502a , Chr 9 and Chr 12 formed a dicentric chromosome via telomere–telomere fusion and then broke to form new versions of Chr 12 and Chr 8 in strains JEC21/JEC20 [54] . Our results indicate that Chr 9 and Chr 12 of XL280 are more similar to those of the B3501α parent , whereas other chromosomes are more similar to the JEC20a parent ( Figure 2E ) . This comprehensive analysis of the XL280 genome reveals how its genome is derived from two well-validated parental reference strains and allowed detailed molecular analyses of unisexual progeny . Solo culture of strain XL280 on V8 sexual reproduction-inducing media resulted in robust production of F1 meiotic spore progeny via selfing α-α unisexual reproduction . This process involves ploidy changes ( 1n→2n→1n ) via self cell–cell fusion or endoreplication , meiosis , and sporulation to produce haploid meiotic progeny . In total , 90 α-α unisexual reproduction meiotic progeny were isolated by spore micromanipulation and germination and were examined in a battery of phenotypic analyses . We assessed major virulence factors of C . neoformans ( growth at 37°C and melanin production on niger seed ( NS ) or L-DOPA media ) , sensitivity to the antifungal drugs fluconazole ( FLC ) or FK506 , and unisexual reproduction ( self-filamentous growth ) ( Figure S3 ) . In these analyses , six of 90 F1 progeny ( ∼7% ) showed a distinct phenotype compared to the haploid parental strain XL280 and the remaining 84 progeny showed no phenotypic differences from the wild-type ( Figure 3A and Figure S3 ) . The six isolated variant F1 progeny strains all exhibited temperature-sensitive ( TS ) growth ( Figure 3A ) . Two progeny ( MN77 and MN89 ) were sensitive to the calcineurin inhibitor FK506 and produced more melanin , indicating they may have similar genetic changes . MN35 was FLC-sensitive , while the other four progeny ( MN27 , MN55 , MN77 , and MN89 ) were relatively FLC-resistant ( Figure 3A ) . On V8 mating-inducing media ( V8 agar ) , MN27 underwent more robust unisexual development , while MN35 , MN55 , MN77 , and MN89 generated fewer hyphae ( Figure 3A ) . Although the two genomes of cells undergoing α-α unisexual reproduction are identical , phenotypic changes to these and other conditions ( Figure S4 ) are clearly evident in even a relatively small sample of progeny produced by this process , which indicates that α-α unisexual reproduction generates phenotypic plasticity de novo since there is no preexisting genetic diversity to admix . We next tested if standard mitotic growth might also lead to phenotypic changes like meiotic unisexual reproduction . A similar set of 96 isolates derived from XL280 by mitotic asexual growth exhibited no phenotypic variation from wild-type in the same battery of phenotypic tests ( Figure S5 ) in which phenotypic variation was readily detected in meiotic unisexual reproduction . Thus , we conclude that phenotypic variation occurs following meiotic unisex but not standard mitotic asexual growth . Phenotypic variation is often attributable to genetic changes . To establish the molecular basis of phenotypic variation of the unisexual progeny , we examined changes in ploidy , whole-chromosome aneuploidy , chromosomal translocations , and single nucleotide polymorphisms ( SNPs ) by FACS , CGH , CHEF , and Illumina NGS , respectively . FACS analysis revealed that the hypersexual progeny MN27 was diploid , in accord with previous findings that diploid strains are ( 1 ) intermediates or products of unisexual reproduction and ( 2 ) hyperfilamentous ( Figure 3B ) [34] . None of the other 90 F1 α-α unisexually produced progeny were diploid ( Figure 3B and unpublished data ) . Three progeny ( MN7 , MN55 , and MN89 ) were analyzed by Illumina NGS to examine the presence of SNPs . In total , only three SNPs were identified in the ∼20 Mb genomes of the three isolates based on 48 , 644 , 802 sequences of 75 bp generated for MN7 , 46 , 089 , 962 for MN55 , and 44 , 650 , 780 for MN89 ( Figure 3C ) . One SNP located at position 878 , 429 on Chr 7 of MN55 resulted in an L318P coding substitution in the hypothetical protein CNJ03170 . A second SNP was located at position 491 , 632 on Chr 8 in both MN7 and MN55 , resulting in a cytosine to thymine change in the 5′ UTR of the CNH02880 gene . The third SNP was located at position 704 , 393 on Chr 9 of MN7 and resulted in an H66A amino-acid substitution in the co-chaperone Hsc20 ( CNI02610 ) , a heat shock protein that functions in iron–sulfur cluster assembly [55] . To examine whether this SNP is responsible for the TS phenotype of isolate MN7 , we performed a complementation test by introducing the WT allele of HSC20 into the MN7 TS strain . We first documented that the hsc20-1 mutant allele is recessive based on analysis of an hsc20-1/HSC20 diploid fusion product of MN7 and XL280α . While the MN7 strain is TS , the MN7×XL280α diploid fusion product was temperature resistant , similar to the XL280α parent . The WT allele of HSC20 was cloned in a plasmid under control of its native promoter and terminator and introduced into strain MN7 by biolistic transformation . In multiple independent transformants , we found that a single copy of the WT HSC20 allele complemented the TS phenotype of MN7 and restored WT growth similar to the XL280α parent at higher temperature ( Figure 3D ) . Based on these findings we conclude that the TS phenotype of progeny MN7 is attributable to the H66A amino-acid substitution in the co-chaperone Hsc20 . Based on CHEF analysis of the chromosomal karyotype , chromosome size differences were detected in the α-α unisexual F1 progeny MN27 , MN55 , and MN89 ( Figure 4A ) . For MN27 and MN55 , an extra chromosomal band was present and migrated more rapidly than Chr 8/9 . For MN89 , an extra chromosomal band was observed to migrate more rapidly than Chr 6/7 . To further analyze these anomalous chromosomes , we excised the novel chromosomal bands from the CHEF gel and extracted and fluorescently labeled the DNA to generate probes . Band CGH analysis revealed that the excised chromosomal bands covered all ORF regions ( the JEC21 expression array covers ORF and not centromeric or telomeric regions ) of Chr 9 , Chr 9 , or Chr 7 in isolates MN27 , MN55 , or MN89 , respectively ( Figure 4B ) ( XL280 Chr 9 hybridizes to Chr 8 and 12 on the JEC21 array ) . Because all ORFs were recognized despite a shortened chromosomal length , we hypothesized that centromeric or telomeric regions ( not represented on the array ) may have suffered significant deletions in the respective strains . To test this hypothesis , we chose MN89 as an example and designed specific probes that hybridize to the centromeric and telomeric regions of Chr 7 ( Figure 4C ) . Restriction enzyme digestion of whole chromosomes and CHEF electrophoresis followed by Southern hybridization revealed that Chr 7 of MN89 has the same telomere lengths as those of XL280 , but a shortened centromeric region ( Figure 4C ) . Further whole genome sequencing determined that the size of the deleted centromeric region is 10 , 257 bp ( Figure 4D ) . These findings indicate that α-α unisexual reproduction induces phenotypic and genotypic plasticity . However , a common genetic change responsible for the phenotypic changes remained to be identified . CGH analysis of the phenotypically variant progeny revealed that four out of six isolates ( 66 . 7% ) were aneuploid . Isolates MN35 , MN55 , MN77 , and MN89 contained an extra copy of chromosome 13 , 9 , 10 , and 10 , respectively , indicating that a high rate ( >4% ) of aneuploidy is generated as a consequence of α-α unisexual reproduction ( Figure 5A ) . Whole genome sequencing of isolates MN55 and MN89 confirmed the presence of an extra copy of Chr 9 or Chr 10 , respectively , in a majority of cells in the population , as ∼2-fold higher levels of sequence reads were obtained for these chromosomes compared to other genomic regions ( Figure 5B ) . Interestingly , Sionov et al . previously associated Chr 10 disomy with FLC resistance [16] . As a complementary approach to CGH , we developed a facile multiplex PCR-based method to detect aneuploidy . In this approach , 14 primer pairs ( Table S3 ) were combined into a single PCR reaction such that each primer pair represents a separate chromosome and amplifies a specific region of different sizes to generate a ladder of 100 to 1400 bp ( Figure 5C ) . PCR products of greater intensity reflect the presence of extra chromosomes ( red arrows in Figure 5C ) . Bioanalyzer analysis further confirmed a ∼2-fold increase in PCR product yield from the additional aneuploid chromosomes present in 1n+1 isolates ( Figure S6 ) . Using multiplex PCR we screened all of the XL280 meiotic ( 90 ) and mitotic ( 96 ) progeny and we did not detect any additional aneuploid strains from the meiotic progeny and none from the mitotic progeny ( Figure S7 and Figure S8 ) . In S . cerevisiae , strains aneuploid for different chromosomes all share several common phenotypes , including temperature sensitivity [8] . Therefore , we determined whether aneuploid strains of Cryptococcus exhibit a similar phenotype . Consistent with the observed consequences of aneuploidy in S . cerevisiae , all aneuploid C . neoformans strains in our study exhibited a TS phenotype ( Figure 3A ) . Aneuploidy can compromise growth unless it provides a fitness benefit under stress or other distinct conditions . We found that aneuploid strains exhibit a variety of phenotypes potentially associated with virulence , including antifungal drug resistance . To test if the progeny have a competitive advantage relative to the parental strain , we co-incubated an aneuploid progeny ( MN55 , MN77 , or MN89 ) with the wild-type parent XL280α ( marked with the NAT drug resistance marker ) in the presence of fluconazole . As expected , the fluconazole-resistant aneuploid progeny exhibited an increased competitive fitness and outgrew the parental strain in competitive growth assays by a ratio of ∼4∶1 ( Figure 6 ) . Moreover , to investigate the fitness of these aneuploid unisexual progeny , we tested their ability to infect and cause disease in a murine inhalation model . Mice were infected with the aneuploid strains via intranasal instillation and monitored for signs of pulmonary cryptococcosis or meningoencephalitis . In contrast to the view that aneuploidy is deleterious , the aneuploid unisexually generated isolates MN35 ( n+113 ) and MN55 ( n+19 ) were as virulent as the euploid wild-type parent , despite their modest TS growth at 37°C on YPD rich media in vitro ( Figure 7 ) . Previous studies have shown that in the clinically important sibling species C . neoformans var . grubii ( serotype A ) strains that are disomic for chromosome 13 exhibit a melanin defect and impaired virulence in mice [21] . However , we found that disomy 13 in C . neoformans var . neoformans ( serotype D ) did not affect melanin production and the MN35 strain was equally virulent as the wild-type . Thus , the specific consequences of aneuploidy may differ between even closely related lineages . Two other aneuploid strains ( MN77 and MN89 ) were both attenuated compared to wild-type and , interestingly , both are aneuploid for chromosome 10 ( Figure 7 ) . Cells isolated from lungs of infected animals exhibited relatively stable aneuploid phenotypes and were found to retain their aneuploid chromosome in 70–80% of isolates based on multiplex PCR [20 isolates were analyzed from each strain , 15 were aneuploid for MN35 ( 75% ) , 16 for MN55 ( 80% ) , 14 for MN77 ( 70% ) , and 15 for MN89 ( 75% ) ( Figure S9 ) ] . Thus , under the stressful gauntlet of the host , several aneuploid progeny appeared as fit as the euploid parent whereas others were attenuated . To examine whether the observed phenotypic changes are a consequence of aneuploidy , we compared the phenotype of aneuploid strains and euploid derivatives obtained following loss of the extra chromosome . Aneuploid strains are relatively unstable and frequently lose the 1n+1 extra chromosomal copy to return to the 1n haploid euploid state . Here , we analyzed isolate MN77 , which contains an extra copy of Chr 10 and produces increased amounts of melanin , a readily scorable phenotype . We grew MN77 at 37°C to promote loss of the extra chromosome and isolated 22 randomly selected colonies for phenotypic and genotypic testing ( Figure 8 ) . Of these 22 isolates , 14 lost the extra copy of Chr 10 based on multiplex PCR analysis ( Figure 8B ) , and in all cases concomitantly lost the enhanced melanin production phenotype ( Figure 8A ) , therefore suggesting that Chr 10 disomy is responsible for the higher level of melanin produced . The remaining eight isolates that retained the Chr 10 aneuploidy all continued to produce higher levels of melanin ( Figure 8A ) . To test whether Chr 10 disomy is also correlated with other phenotypes , we performed growth dilution assays with fluconazole ( at 30°C ) , FK506 ( at 30°C ) , and for growth at 37°C . Similar to the melanin phenotype , all isolates that had lost the extra copy of Chr 10 exhibited WT phenotypes similar to the XL280 parent , and all isolates that retained the extra copy of Chr 10 exhibited the variant phenotype ( Figure 8C ) . Similar experiments were performed for isolates MN35 ( n+113 ) and MN55 ( n+19 ) , and we found that loss of the extra chromosome was linked with loss of the variant phenotypes ( fluconazole sensitivity and fluconazole resistance , respectively ) ( Figure S10 ) . These results indicate that aneuploidy causes the phenotypic changes observed in disomic strains . To test whether the generation of aneuploidy is specific to α-α unisexual reproduction , we isolated progeny from asexual mitotic reproduction and a-α opposite-sexual reproduction . We isolated 96 single colonies from yeast extract-peptone-dextrose ( YPD ) media following asexual mitotic reproduction and from spores produced on V8 agar following a-α sexual reproduction . No phenotypic or genotypic changes were detected among asexual mitotic progeny ( 0/96 ) ( Figure S5 and Figure S8 ) . To test a-α sexual reproduction , we first investigated 88 F1 progeny from the cross between strains XL280α and JEC20a . To reduce the chance that α-α unisexual progeny from XL280 were mixed with the a-α sexual reproduction progeny from the cross , three times more yeast cells from the a parent JEC20 were mixed with the XL280α cells in the cross . As these strains are ∼81% congenic , but not isogenic , they exhibit different phenotypes under various conditions , and their progeny ( 88 were examined ) showed a range of phenotypes ( Figure S11 ) . Among 11 strains that had the most overt phenotypic changes , four were aneuploid and no other aneuploids were detected in the larger progeny set ( Figure 9A ) . Further PCR and RFLP tests confirmed that these F1 progeny strains were products of meiosis ( Figure S12 ) . In addition , as a previous study showed that the SXI2a homeodomain factor gene is sufficient to drive sexual development of haploid α cells [56] , we introduced the SXI2a gene into strain XL280α to mimic a-α sexual reproduction , generating strain MN140 . 23 ( Figure S13 ) . Among 90 isolates generated from the selfing of this α+SXI2a self-fertile strain , seven strains exhibited phenotypic changes compared to the parental strain ( Figure S14 ) . Further CGH analysis showed that two of these strains were aneuploid ( Figure 9B ) . These results provide evidence that aneuploidy is also generated by a-α sexual reproduction . To further examine this hypothesis , we analyzed a-α progeny from the cross between the isogenic C . neoformans var . grubii strains KN99a and KN99α . Among 88 progeny , 14 strains exhibited phenotypic changes ( Figure S15 ) , and eight were aneuploid ( Figure 9C ) , indicating that a-α sexual reproduction between isogenic strains in the serotype A lineage can also generate phenotypic and genotypic diversity frequently involving aneuploid progeny . In summary , aneuploidy is generated during both α-α unisexual and a-α congenic sexual reproduction in both C . neoformans var . grubii ( serotype A ) and C . neoformans var . neoformans ( serotype D ) .
While sexual reproduction serves to admix genetic diversity from two distinct parents to produce progeny with a diverse genetic repertoire , sex also comes with a series of costs . Such costs include: ( 1 ) metabolic energy that must be devoted to mating and meiosis; ( 2 ) energy and time expended locating a mating partner; ( 3 ) that only 50% of parental genes are transmitted to any given progeny or that two individuals are required to produce one progeny ( resulting in the so-called 2-fold cost of sex ) ; and ( 4 ) the fact that two genomes that have run the gauntlet of adaptive selection are shuffled during the process , breaking apart well-adapted genomic configurations [57] . A central question then is: Why would an organism engage in unisexual reproduction ? In some examples , unisex involves two genetically distinct α mating partners , and this leads to genetic exchange and the production of recombinant progeny , similar to a-α sexual reproduction . Also , α-α mating lowers the barrier to finding a rare mate in a predominantly α population , enabling outcrossing even if no a partners are available . But in other cases , an α isolate undergoes unisexual reproduction all by itself ( via cell–cell fusion or endoreplication ) , and in these cases there is no pre-existing genetic diversity to admix . If there is no heterozygosity of the diploid undergoing meiosis , why undergo sex if the genome is homozygous everywhere ? As shown here , our studies provide direct experimental support for the hypothesis that unisexual reproduction can generate phenotypic and genotypic diversity de novo . Why might this be of selective and adaptive benefit ? In the context of considering the costs of sexual reproduction , unisexual reproduction is one strategy by which several of the costs normally associated with sex can be lessened or mitigated entirely . First , the cost of finding a mating partner is considerably decreased in the case of mother–daughter cell mating in which the two cells are physically juxtaposed or eliminated entirely , such as in endoreplication , in which a single cell transitions from haploid to diploid as a prelude to meiosis . Second , during selfing unisexual reproduction , ∼100% of the parental genes are directly transmitted to the F1 progeny , thus reducing the 2-fold cost of sex . Further , if one considers aneuploidy , >100% of the parental genes are transmitted to progeny . Third , unisexual reproduction eliminates the cost of sex associated with breaking apart well-adapted genomic configurations . Instead , unisexual reproduction preserves well-adapted genotypes by allowing mating between genetically identical cells ( i . e . , mother and daughter cells ) and adding a limited amount of genetic diversity ( including aneuploidy , chromosomal size polymorphisms/deletions , and SNPs ) to a well-adapted genotype . In essence , unisexual reproduction provides a mechanism by which a well-adapted genotype can be changed much more subtly than standard sexual outcrossing , and in the case of well-adapted genotypes has the capacity to provide a more parsimonious route to progeny that are enhanced in competitive fitness in response to subtle changes in environmental conditions . Organisms that reproduce sexually have an advantage over organisms that reproduce asexually , as preexisting genetic diversity will generate novel combinations in the population . Meiosis , through homologous recombination , will generate distinct genetic compositions that may have a selective advantage over the parents in a new hostile environment . Unisexual reproduction is a sexual cycle that can occur between genetically identical cells and , as we observed here , allows the introduction of limited de novo genetic diversity through meiosis . As in most species , meiosis in Cryptococcus is a highly regulated process . Comparative genomics reveals that Cryptococcus species contain the meiotic genes involved in homologous recombination and synaptonemal complex formation [46] , [58] . The genes that support the presence of a meiotic pathway ( termed the meiotic toolkit genes: DMC1 , MND1 , MSH4 , MSH5 , SPO11 , HOP1 , HOP2 , and REC8 ) are highly conserved in Cryptococcus , and two of these ( SPO11 and DMC1 ) have been found to be critical for sporulation and spore germination during unisexual reproduction [34] , [42] . Sex-induced phenotypic and genotypic variation in a clonal population is not restricted to the Cryptococcus genus . Aspergillus nidulans is a filamentous ascomycete that , along with its well-established sexual cycle , also has a parasexual cycle wherein haploid mycelia fuse and then their nuclei fuse to form diploid nuclei [59] . In the parasexual cycle , the diploid mycelium undergoes a transient aneuploid state by repeated loss of whole chromosomes to ultimately regenerate haploid progeny . A recent study has shown that the A . nidulans parasexual cycle can drive adaptive evolution . Hoekstra and colleagues generated homozygous diploid strains that were isogenic with their haploid progenitor [60] . In a laboratory setting , faster growing variants frequently arose from the homozygous diploid strains but not from their haploid parents . Remarkably , all of the faster growing variants derived from a diploid parent were found to be haploids that arose through the parasexual cycle . As few as 3 , 000 generations were sufficient for the emergence of more rapid growth . Through genetic analysis , this study supported a model wherein the diploid serves as a capacitator for evolution . This enables recessive mutations to arise sequentially in the sheltered state of the diploid where they are complemented , and these mutations do not survive in the haploid because they are individually deleterious and swept from the population before a second mutation can arise [60] . The parasexual state assorts and releases these mutations into the haploid state , where they exhibit reverse epistasis and only when in combination confer a benefit ( i . e . , faster growth ) , whereas individually each recessive mutation was deleterious . If parasexual cycles can function as capacitators to generate , store , and then release genotypic and phenotypic diversity de novo , we propose that sexual cycles , including unisexual reproduction , might also serve such a role in which meiosis rather than parasexual chromosome loss is involved . A similar unisexual selfing mechanism has been observed in the human fungal pathogen C . albicans . Alby et al . found that a/a cells of C . albicans lacking the Bar1 protease that destroys the α-factor mating pheromone undergo a/a-a/a unisexual mating , yielding tetraploids that can undergo concerted chromosome loss to complete a parasexual cycle [43] . In addition , the parasexual cycle is induced between two a/a mating partners when mixed in ménage a trois matings with a limited number of α/α cells that serve as pheromone donors to trigger unisexual a-a reproduction [43] . However , the parasexual progeny of C . albicans are characterized by high rates of aneuploidy [44] . Although this chromosomal assortment process is imprecise , it may be retained due to a selective pressure that favors aneuploid strains under certain environmental conditions , such as in patients receiving fluconazole , a situation in which an isochromosome 5 derivative that confers drug resistance often arises [15] . Aneuploidy has been linked to phenotypic changes , including resistance to antifungal drugs in both C . albicans and C . neoformans , which could provide a selective advantage during infection [15]–[17] . The positive impact of aneuploidy on evolution and genetic diversification extends beyond the fungal kingdom to other unicellular eukaryotes . The parasitic protozoan Leishmania is the etiological agent behind one of the most common neglected diseases , leishmaniasis , a global cause of morbidity and mortality . No vaccine is available , and treatment relies heavily on pentavalent antimonial compounds with diminishing efficacy due to emerging drug resistance [61] . Recent studies have shown that aneuploidy is widespread in natural and clinical populations , with variation in the ploidy state ( monosomic , disomic , or trisomic ) for different chromosomes even within the same isolate [62] , [63] . Mosaic aneuploidy in Leishmania generates dynamic genome plasticity , is well tolerated , and can confer drug resistance [22] , [62] , analogous to fungal azole resistance conferred by aneuploidy . Given the finding that aneuploidy can underlie and drive adaptive evolution in S . cerevisiae , C . albicans , C . neoformans , and Leishmania , it seems likely that beneficial impacts of aneuploidy may be even more ubiquitous and remain to be discovered in other saprobic and pathogenic eukaryotic microbes .
All of the animal studies were conducted in the Division of Laboratory Animal Resources ( DLAR ) facilities at Duke University Medical Center ( DUMC ) . All of the animal work was performed according to the guidelines of NIH and Duke University Institutional Animal Care and Use Committee ( IACUC ) . The animal experiments were reviewed and approved by the DUMC IACUC under protocol number A266-08-10 . Strains and plasmids used in this study are listed in Table S4 . Yeast cells were grown on YPD media . Mating of C . neoformans was conducted on 5% V8 juice agar medium ( pH = 7 ) . DNA sequencing was performed on a Genome Analyzer IIx using Illumina Paired Ends technology . The genomic DNA samples were fragmented with a Bioruptor sonicator ( Diagenode ) . Fragmented DNA was used for size selection by extraction ( Qiagen ) of 300 to 400 bp DNA fragments from 2% agarose gels after electrophoresis . Size-selected DNA was used for standard Illumina library preparation protocols . Prepared libraries were sequenced using Paired Ends 2×76 cycles . This approach provided the most suitable sequencing data for SNP and small indel detection and deep sequencing coverage . The Illumina-Solexa data were assembled using a combination of de novo assembly ( Velvet ) and reference genome assembly ( BWA ) to assemble the Cryptococcus genomic sequence . We used BWA and SAMtools to determine the depth of sequence coverage and identify polymorphisms , particularly SNPs and indels . In addition to the use of BWA for reference genome assembly , we checked the completeness of the assemblies by extracting all sequence reads that BWA was unable to align to the reference genome and assembled these reads using Velvet . The resulting contigs were compared back to the genome sequence and also used to search GenBank using Blast . This method allows identification of regions with multiple polymorphisms that cannot be identified by BWA due to the inability of the short reads to be aligned to the reference sequence . In addition , these contigs typically originate from highly repetitive sequence , such as the rDNA and telomeres . With this Cryptococcus dataset , more than 90% of the unaligned reads were low-quality sequence reads . Overall , the assembly was an iterative process . After polymorphisms were identified in an initial round of sequence assembly with BWA and Velvet , a new consensus sequence was generated , and the assembly was repeated . In the second and subsequent rounds of assembly , polymorphisms were identified in regions where there were multiple polymorphisms . After several rounds , no additional polymorphisms were identified , and the assembly was considered complete ( Figure S1 ) . The ploidy of progeny was determined by flow cytometry as described previously [64] . Briefly , cells were collected from overnight YPD liquid medium , washed once in 1× PBS buffer , and fixed in 1 mL of 70% ethanol overnight at 4°C . Fixed cells were washed once with 1 mL of NS buffer ( 10 mM Tris-HCl [pH = 7 . 6]; 250 mM sucrose; 1 mM EDTA [pH = 8]; 1 mM MgCl2; 0 . 1 mM CaCl2; and 0 . 1 mM ZnCl2 ) and then stained with propidium iodide ( 0 . 3 mg/mL ) in 200 µL of NS buffer containing RNaseA ( 1 mg/mL ) at room temperature for 4 h . Then , 50 µL of the stained cell preparation was diluted into 2 mL of 50 mM Tris-HCl ( pH = 7 . 5 ) and sonicated for 1 min . Flow cytometry was performed on 10 , 000 cells and analyzed on the FL1 channel of a Becton-Dickinson FACScan . The strain MN7 was transformed with the dominant selectable drug marker NEO , amplified from plasmid pJAF1 using the universal primers M13F and M13R . The strain MN7 NEO was mixed with strain XL280α NAT and incubated on mating media for 48 h . The mating culture was scraped off the plate , washed in water , serially diluted , and plated on YPD+NAT+NEO medium . The ploidy of the fusion products was determined by FACS . Multiple diploid isolates were phenotypically analyzed at 30°C and 37°C to determine the recessive nature of the TS mutation in strain MN7 . The wild-type HSC20 gene with its native promoter ( 269 bp ) and terminator ( 246 bp ) was amplified from XL280α genomic DNA using the primer pair JOHE38840/JOHE38841 and cloned in the pJAF12 plasmid using KpnI restriction enzyme ( NEB ) [65] . Two independently derived plasmids , pMF86 and pMF89 , were obtained and confirmed by sequencing . The plasmids were introduced into strain MN7 via biolistic transformation , and multiple isolates were obtained and analyzed phenotypically . A single colony was inoculated into 5 mL of liquid YPD medium and grown overnight at room temperature ( RT ) while shaking . Then , 1 mL of the culture was added to 50 mL of yeast nitrogen base minimal medium ( YNB ) with 1 M NaCl ( to suppress capsule formation ) , which was then grown overnight at RT on a shaking platform incubator . Yeast cells were then washed three times in 0 . 5 M NaCl/50 mM EDTA ( pH = 8 . 0 ) . Then , 50 mg of cells were mixed with 450 µL of low-melting point agarose ( Bio-Rad 0 . 5% in 0 . 1 M EDTA [pH = 8 . 0] ) and 20 µL of cell wall lysing buffer ( 25 mg/mL Zymolase in 10 mM KPO4 [pH = 7 . 5] ) and poured into molds . The plugs were solidified for 15 min at RT , transferred into 700 µL of lysing solution ( 0 . 5 M EDTA , 10 mM Tris-HCl [pH = 8] ) , and lysed overnight at 37°C . The next morning , 400 µL of proteinase K solution ( 5% sarcosyl , 5 mg/ml proteinase K , 0 . 5 M EDTA [pH = 8 . 0] ) were added , and the plugs were further incubated at 50°C for 5 h . The plugs were then washed three times for 1 h each with wash solution ( 50 mM EDTA , 10 mM Tris-HCl [pH = 8 . 0] ) . Then , 1% gels made with pulse field certified agarose in 0 . 5× TBE buffer were run with a BioRad CHEF Mapper XA system . Running conditions were according to the CHEF Mapper calculation , with appropriate modifications . Genomic DNA was ultrasonicated to generate ∼500-bp fragments and purified with a DNA Clean and Concentrator kit ( Zymo Research , CA ) . Then , 2 . 5 µg of DNA was used for Cy3 dUTP or Cy5 dUTP labeling reactions using the Random Primer/Reaction Buffer mix ( BioPrime Array CGH Genomic Labeling System , Invitrogen ) . Labeled DNA was then hybridized to microarray slides of 70-mer oligonucleotides for the C . neoformans JEC21 [49] or the C . neoformans JEC21 and H99 genomes ( Washington University , St . Louis , MO ) . After hybridization , arrays were scanned with a GenePix 4000B scanner ( Axon Instruments ) . Data analysis was performed with Genespring GX v7 . 3 ( Agilent Technologies ) and CGH-miner . Band array was performed as described previously [66] . Briefly , chromosomes were separated by PFGE and the bands of interest were excised and treated to extract the DNA . Band DNA was labeled with Cy5 and hybridized to microarrays with Cy3-labeled whole genome DNA . Genomic DNA was extracted using the CTAB protocol as described previously [67] . We utilized the Qiagen multiplex PCR kit for multiplex PCR . PCR reaction mixtures ( 25 µL ) contained 12 . 5 µL of Qiagen multiplex PCR master mix , 2 µM equimolar primer mixture ( Table S3 ) , and 100 to 300 ng of genomic DNA . Thermal cycling conditions included an initial heat activation of 15 min at 95°C followed by 30 cycles of denaturation for 30 s at 94°C , annealing for 90 s at 58 . 7°C , and extension for 10 min at 72°C . For bioanalyzer analysis , 1 µL of PCR reaction was examined with the Experion DNA 12K Analysis Kit ( Bio-Rad ) . Multiplex PCR reactions were analyzed by gel electrophoresis using 1 . 8% agarose gels in 1×TBE buffer ( Tris/Borate/EDTA buffer ) . 2 µL of PCR reaction were loaded in the gels that were run overnight at 30 volts or for 5 h at 100 volts . The intensity of the desired bands , which represent the aneuploid chromosomes , was quantified relative to the control band using the Gel Doc XR+ system of BIO-RAD and Image Lab version 4 software . RFLP analysis was performed as described previously [68] . Based on the SNPs in XL280 and JEC20 , we designed two RFLP markers , RFLP3 and RFLP7 ( primers are shown in Table S5 ) , and digested with NdeI and XbaI , respectively . Other markers included the STE20a and STE20α mating type locus genes to assign the mating type as a or α . Virulence assays were conducted using a murine inhalation model of cryptococcosis . Cohorts of 4- to 8-wk-old female DBA mice were anesthetized through intraperitoneal injection of Nembutal ( 37 . 5 mg/kg ) and infected intranasally with 5×106 cells diluted in sterile PBS . The cells of the inoculum were diluted and plated onto YPD medium to determine CFU and viability . Mice were monitored twice daily , and moribund individuals were euthanized with CO2 . The survival rates were plotted against time using Kaplan-Meier survival curves , generated with Prism 4 . 0 ( GraphPad software , La Jolla , CA , USA ) . The p values were evaluated by a Log-rank ( Mantel-Cox ) test . A p value of <0 . 05 was considered significant . The lungs and brains of the euthanized animals were removed , weighed , and homogenized in 2 ml sterile PBS . The samples were serial diluted and plated on YPD media to count CFUs . Twenty random isolates were colony purified and subjected to multiplex PCR to detect aneuploidy . XL280 , MN55 , MN77 , MN89 ( all NAT sensitive ) , and XL280 marked with the NAT resistance marker integrated at the SPO11 genetic locus ( XL280 NAT ) were used to measure the competitive fitness of the aneuploid strains in the presence of fluconazole . Strains were grown overnight in liquid cultures in YPD , washed with sterile water , and diluted to a density of 1×107 cells/ml . The following strains were mixed in equal ratios: XL280 with XL280 NAT , MN55 with XL280 NAT , MN77 with XL280 NAT , and MN89 with XL280 NAT . The mixtures were spotted on YPD and YPD plus 8 µg/mL fluconazole ( FLC ) and incubated for 3 d at 30°C . Then , the cells were recovered from the plates , washed with sterile water , and plated on YPD to count CFUs . Following 2 d incubation at 30°C , 300 isolates were replica plated onto NAT selective media and survival rates were calculated . The experiments were conducted in triplicate . For the growth curve assays , the strains were grown overnight in liquid cultures , washed with sterile water , and diluted into 4×104 cells in 200 µl of YPD and YPD plus 2 µg/mL fluconazole ( fluconazole concentration was determined experimentally ) . The cells were incubated at room temperature in 96-well plates , and OD600 was measured every 2 h for 48 h using a Tecan-Sunrise microplate reader . All of the experiments were performed in triplicate . | Aneuploidy refers to increases or decreases in the copy number of individual chromosomes ( rather than of the entire haploid or diploid genome ) . In humans , aneuploidy is well known to be deleterious , causing genetic disorders such as Down syndrome ( trisomy 21 ) , and frequently occurring during mitosis in the genesis of cancer . By contrast , aneuploidy in fungi can be advantageous , conferring antifungal drug resistance and enabling rapid adaptive evolution . Cryptococcus neoformans is a globally distributed human pathogen that often infects patients with compromised immunity . It accounts for significant morbidity and mortality associated with HIV/AIDS and is linked to more than one million infections and >600 , 000 deaths per year world-wide . Although C . neoformans has a defined heterosexual cycle involving a and α cells , more than 99% of clinical and environmental isolates are α . Interestingly , C . neoformans α cells undergo α-α unisexual reproduction to generate diploid intermediates and infectious haploid spores . Sex is costly , though , and the question therefore arises as to why C . neoformans would undergo selfing unisexual , meiotic reproduction as opposed to more efficient asexual , mitotic reproduction . We show here that unisexual , meiotic reproduction in C . neoformans results in aneuploidy , creating advantageous genetic diversity de novo . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"mycology",
"genomics",
"emerging",
"infectious",
"diseases",
"medical",
"microbiology",
"genetics",
"microbiology",
"biology"
] | 2013 | Unisexual and Heterosexual Meiotic Reproduction Generate Aneuploidy and Phenotypic Diversity De Novo in the Yeast Cryptococcus neoformans |
Sporadic Creutzfeldt-Jakob disease ( sCJD ) is the most prevalent of the human prion diseases , which are fatal and transmissible neurodegenerative diseases caused by the infectious prion protein ( PrPSc ) . The origin of sCJD is unknown , although the initiating event is thought to be the stochastic misfolding of endogenous prion protein ( PrPC ) into infectious PrPSc . By contrast , human growth hormone-associated cases of iatrogenic CJD ( iCJD ) in the United Kingdom ( UK ) are associated with exposure to an exogenous source of PrPSc . In both forms of CJD , heterozygosity at residue 129 for methionine ( M ) or valine ( V ) in the prion protein gene may affect disease phenotype , onset and progression . However , the relative contribution of each PrPC allotype to PrPSc in heterozygous cases of CJD is unknown . Using mass spectrometry , we determined that the relative abundance of PrPSc with M or V at residue 129 in brain specimens from MV cases of sCJD was highly variable . This result is consistent with PrPC containing an M or V at residue 129 having a similar propensity to misfold into PrPSc thus causing sCJD . By contrast , PrPSc with V at residue 129 predominated in the majority of the UK human growth hormone associated iCJD cases , consistent with exposure to infectious PrPSc containing V at residue 129 . In both types of CJD , the PrPSc allotype ratio had no correlation with CJD type , age at clinical onset , or disease duration . Therefore , factors other than PrPSc allotype abundance must influence the clinical progression and phenotype of heterozygous cases of CJD .
Prion diseases are fatal neurodegenerative disorders affecting humans and various other mammals . They are associated with the misfolding of monomeric prion protein ( PrPC ) into a pathological isoform termed PrPSc that is partially protease resistant , aggregated , and infectious . Human prion diseases include Creutzfeldt-Jakob disease ( CJD ) , Gerstmann-Straussler-Scheinker syndrome ( GSS ) , kuru , fatal familial insomnia ( FFI ) ( for review , see [1] ) , PrP cerebral amyloid angiopathy [2] and variably protease-sensitive prionopathy [1 , 3] . Although CJD occurs in sporadic , genetic , and iatrogenic forms , the most common form is sporadic CJD ( sCJD ) , which occurs at approximately 1–2 cases per million people per year in any given population . Acquired forms of CJD such as kuru , variant CJD , and iatrogenic CJD ( iCJD ) , represent a smaller percentage of all CJD cases . CJD is associated with a wide diversity of clinicopathological features [1 , 4 , 5] , but the factors that determine these different CJD phenotypes are still being elucidated . Sporadic CJD is classified by a neuropathological profile that appears to correlate with the biochemical properties of PrPSc [4–7] as well as the sequence of the patient prion protein gene ( PRNP ) at codon 129 . All three genotypes of a naturally occurring amino acid polymorphism at PRNP codon 129 are found in CJD: homozygous methionine ( MM ) or valine ( VV ) , and heterozygous ( MV ) . Biochemically , the two major PrPSc types associated with CJD , termed Type 1 and Type 2 , can be distinguished by the molecular mass of PrPSc following protease digestion . Type 1 PrPSc has a protease-resistant molecular mass of approximately 21 kDa while a molecular mass of approximately 19 kDa is characteristic of Type 2 PrPSc [1] . Thus , sCJD can occur with six genotype/PrPSc type combinations: MM1 , MM2 , MV1 , MV2 , VV1 , and VV2 . These correspond well to the differing clinicopathological presentation found in patients in the six well-recognized sCJD phenotypic subtypes: MM1/MV1 , MM2 cortical , MM2 thalamic , MV2 , VV1 and VV2 [1 , 8 , 9] . Based on these criteria , as well as transmission studies in non-human primates [10] and transgenic mice expressing different human PRNP genotypes [11 , 12] , 5 major strains of sCJD have been proposed: MM1/MV1 , MV2/VV2 , MM2c , MM2t and VV1 [5] . PrPSc type and PRNP genotype are therefore believed to be two major factors influencing CJD phenotype and pathogenesis . Although not causing disease itself , the presence of M or V at residue 129 in PrPC does affect many aspects of prion disease susceptibility and phenotype [1 , 4–9 , 13] . For example , in a familial form of human prion disease associated with an aspartic acid to asparagine mutation at residue 178 ( D178N ) , a methionine in cis at position 129 manifests as FFI , while a valine in cis at position 129 manifests as CJD [14] . The codon 129 genotype also affects susceptibility to CJD . Methionine and valine homozygosity at codon 129 are overrepresented in sCJD patients when compared to the general population , whereas MV heterozygosity at codon 129 is underrepresented [15–18] . A similar allelic distribution has been observed in human growth hormone associated cases of iCJD in France [19] . The predisposition towards homozygosity in CJD is even more pronounced in variant CJD , in which all probable and definite clinical cases reported to date have been MM homozygous [16] . By contrast , MM homozygosity is less common in human growth hormone associated cases of iCJD in the United Kingdom ( UK ) in which heterozygosity and VV homozygosity appear to predominate [19 , 20] . Thus , the epidemiological data suggest that the effect of homozygosity or heterozygosity at codon 129 on CJD pathogenesis varies depending upon the type of CJD . There can be considerable variation within heterozygous MV1 and MV2 sCJD patients with regard to disease duration and onset [6 , 9] . This variability may be due to multiple factors including the relative abundance of different populations of PrPSc . For example , in the brain the amount of protease-sensitive PrPSc , a conformationally distinct population of PrPSc that is aggregated but susceptible to digestion with proteinase K [21] , has been correlated with disease progression rate . Multiple forms of PrPSc , including not only Type 1 and Type 2 but also PrPSc with atypical N-termini , are often found within the same sCJD brain [22 , 23] suggesting that PrPSc type variation may be involved . PrPSc with either M or V at residue 129 ( PrPSc-M129 and PrPSc-V129 , respectively ) has also been identified in brain material from two different heterozygous sCJD patients [24 , 25] . This latter observation is particularly intriguing since the presence of two different PrP allotypes in the same brain can often lead , in a dose-dependent manner , to inefficient PrPSc formation and increased disease incubation times [26 , 27] . Thus , a ready explanation for the variability in disease onset and duration in heterozygous cases of CJD would be quantitative differences between PrPSc allotypes containing M or V at residue 129 . However , the relative abundance of each PrPSc allotype in MV heterozygous cases of sCJD is unknown . In order to determine whether differences in CJD phenotype correlated with differences in the relative amounts of PrPSc-M129 and PrPSc-V129 , we extracted protease-resistant PrPSc from brain tissue of multiple sCJD and iCJD patients heterozygous for M and V at codon 129 in PRNP . Tandem mass spectrometry was then used to differentiate peptides containing M at residue 129 from those containing V . Our results show that the relative abundance of PrPSc-M129 and PrPSc-V129 was highly variable between individual sCJD cases . Furthermore , the PrPSc allotype ratio differed between the sCJD and iCJD patient groups and did not correlate with CJD type , age at clinical onset , or disease duration .
In previous MS-based analysis of purified hamster and mouse PrPSc , we found that one of the most commonly identified PrP peptides spanned residues 111–136 ( PrP111-136 ) [28 , 29] . This peptide has also been identified by MS in PrPSc purified from sheep scrapie [30] . In order to determine if MS analysis could also detect PrP111-136 in human PrPSc , PK treatment followed by PTA precipitation was used to isolate PrPSc from a sample of brain homogenate from a heterozygous case of sCJD . The same procedure was used on brain samples from several non-CJD neurological controls and the samples were compared . Only 5 PrP-specific peptides were found in one of the 3 non-CJD controls , indicating that the PrPSc enrichment protocol used yielded little or no PrPC . By contrast , 525 total PrP peptides were found in the sCJD sample the most common of which are shown in S1 Table . Many of these peptides spanned the region of PrP from residues 111–136 and showed variable levels of methionine oxidation ( S1 Table ) . Importantly , the MV polymorphism at residue 129 was distinguishable based upon the unique fragmentation patterns of peptides containing either M or V ( Fig 1a and 1b ) . The same PrP111-136 peptides were also found , albeit with a much lower rate of methionine sulfoxidation , using rHuPrP-M129 and rHuPrP-V129 purified from E . coli ( S1a and S1b Fig ) . Analysis of synthetic PrP111-136 peptides containing either M or V at residue 129 confirmed that each peptide yielded distinctive spectra with precise differences between MS/MS fragments corresponding to the presence of either M or V at residue 129 ( S1c and S1d Fig ) . Finally , MS analysis using purified rHuPrP demonstrated that PrP111-136 peptides containing M or V at residue 129 could be differentiated with 100% specificity ( S2 Fig ) . The detection of peptides with either M or V at residue 129 in PrPSc from a heterozygous case of sCJD ( Fig 1 ) confirmed earlier studies demonstrating the presence of both PrPSc allotypes in two cases of heterozygous sCJD [24 , 25] . Therefore , both PrPSc-M129 and PrPSc-V129 might contribute to sCJD pathogenesis . However , the relative abundance and thus potential extent of the contribution of each allotype to disease pathogenesis is unknown . Spectral counting , which is defined as the number of MS spectra identified for a protein , is a semi-quantitative technique that is often used as a practical method for estimating protein abundance [31] . It provides a robust label-free estimate for comparing the relative abundance of proteins in or between sample groups [32] . We first determined the quantitative relationship between PrP concentration and spectral counts using MS analysis of a solution of purified rHuPrP stoichiometrically adjusted to contain different molar ratios of rHuPrP-M129 and V129 ( S2 Fig ) . There was a good correlation between spectral counts and the concentration of rHuPrP-M129 ( R2 = 0 . 89 ) . At a molar ratio of 50:50 , 56 ± 2% of the PrP111-136 peptides were identified as M129 while approximately 44 ± 2% contained V129 . Thus , spectral counting allowed us to estimate the relative abundance of M or V at residue 129 in a heterozygous mixture of human PrP , albeit with a slight bias towards detection of the M versus the V PrP111-136 peptide . Therefore , we used a spectral counting approach to semi-quantitatively determine the relative amounts of PrPSc-M129 and PrPSc-V129 in brain tissue derived from heterozygous cases of sCJD . Patient brain samples from 14 cases of heterozygous sCJD were analyzed and their clinical and molecular data are presented in Table 1 . The prominent neuropathological features of each case are described in S2 Table and exemplified by images of PrP immunohistochemistry performed on cerebral cortex ( CC ) and , where relevant to this study , cerebellar cortex ( CbC ) samples ( Fig 2a ) . No PrPSc is detected in non-CJD brain using this technique ( for an example see S3a Fig ) . The pathological and molecular features correspond broadly to the histopathological types proposed by Parchi et al [5] . However , in certain cases , such as case 3 , there was a discrepancy between molecular and histopathological findings ( see S2 Table ) suggesting an atypical case of MV1 sCJD . Age of onset for the sCJD cases analyzed ranged from 53–77 years ( mean ± SD = 64 . 7 ± 7 . 3 ) while disease duration ranged from 4–21 months ( mean ± SD = 10 . 9 ± 5 . 7 ) . Molecular analysis had previously identified 5 of the cases as being of the MV1 sCJD subtype ( cases 1–5 ) and 9 as being of the MV2 sCJD subtype ( cases 6–14 ) ( Table 1 ) . Western blot analysis comparing a sample from the specimen used for MS ( Fig 3a , middle lane of each blot ) to reference standards for Type 1 PrPSc from sCJD MM1 ( Fig 3a , left lane of each blot ) and Type 2 PrPSc from sCJD VV2 ( Fig 3a , right lane of each blot ) confirmed the expected presence of Type 1 or Type 2 protease-resistant PrP in these samples . Analysis of samples from one sCJD MV1 case specimen consistently showed a mobility slightly ahead of the Type 1 reference standard ( Fig 3a , case 4 ) . The blots further resolved the presence of a Type 2 doublet ( 2d ) in 6 out of the 9 MV2 subtype cases ( Table 1 and Fig 3a ) . The relative amounts of the two doublets varied between samples from equivalence ( case 10 CC ) to barely detectable amounts of the least abundant band ( case 7 CC ) . None of the sCJD specimens examined contained both Type 1 and 2 PrPSc ( Fig 3a ) . No PrPSc is detectable in PK-treated non-CJD brain using this technique ( for an example see S3b Fig ) . PrPSc enriched from the cerebral cortex or cerebellar cortex of the 14 cases of heterozygous sCJD as well as 3 non-CJD neurological controls was analyzed by MS . For each patient sample , the number of PrP spectra spanning residues PrP111-136 was determined . Using the data analysis criteria described in the materials and methods , an average of >5 PrP111-136 spectra were identified in every sCJD sample examined but not in the non-CJD controls . The abundance of PrPSc-M129 relative to PrPSc-V129 varied between MV1 samples ( Fig 3b and Table 1 ) , ranging from 100% to 34% of the total PrPSc detected . Similar results were obtained for the MV2 samples where the percentage of PrPSc-M129 ranged from 88–22% ( Fig 3b and Table 1 ) . For each sCJD molecular type approximately 20–30% of the cases significantly favoured PrPSc-M129 , 20–30% favoured PrPSc-V129 , and all other cases had an equal distribution of PrPSc-M129 and PrPSc-V129 ( Fig 3b ) . We next determined if the relative abundance of PrPSc-M129 and PrPSc-V129 could also vary between brain regions within an individual sCJD patient . The relative abundance of PrPSc-M129 and PrPSc-V129 in the CbC was compared to that of the CC in one MV1 and three MV2 sCJD cases ( Table 1 ) . There were no major differences in the amount of PrPSc-M129 ( Fig 4 ) or the type of PrPSc deposition ( Fig 2a ) in the cerebral and cerebellar samples from cases 3 and 7 . However , the amount of PrPSc-M129 ( Fig 4 ) as well as the type of PrPSc deposition ( Fig 2a ) differed significantly between brain regions in the two other cases , 9 and 14 . While only 4 sCJD cases were analyzed , the data suggest that within the same patient the amount of PrPSc-M129 and PrPSc-V129 can differ markedly between different regions of the brain with distinctive neuropathologies . In order to determine if differences in the amount of each allotype were associated with phenotypic variables , the relative abundance of PrPSc-M129 was compared to the sCJD molecular subtype , age of onset , and duration of clinical disease . Statistical analysis revealed no significant correlation between the amount of PrPSc-M129 and sCJD molecular subtype , duration , and age of onset ( p > 0 . 1 , Table 1 ) . For example , sCJD MV1 case 3 , where 100% of PrPSc was M129 , had a similar age of onset and disease duration to sCJD MV1 case 4 where PrPSc-M129 represented only 34% of the total PrPSc identified ( Table 1 ) . Thus , despite significant differences between cases , the MV ratio of the PrPSc allotypes did not appear to overtly influence disease onset or progression . There was evidence that PrPSc allotype might influence disease phenotype in some heterozygous sCJD patients . In two cases , 9 and 14 , differences in the PrPSc allotype ratio between brain regions ( Fig 4 ) were associated with different neuropathologies ( Fig 2a ) . However , when all of the MV sCJD cases were compared , there was no definitive correlation between the amount of PrPSc-M129 and various neuropathological features such as the presence of amyloid plaques , pattern of PrPSc deposition or type of spongiform change ( S2 Table ) . Next we determined the relative abundance of PrPSc-M129 in brain samples from 6 cases of iCJD , all of whom were heterozygous at codon 129 in PRNP ( see Table 2 for clinical and molecular data ) . The prominent neuropathological features of each case are described in S2 Table and exemplified by images of PrP immunohistochemistry performed on cerebral cortex samples in Fig 2b . The mean age of onset for the iCJD cases was younger than that of the sCJD cases , ranging from 27–33 years ( mean ± SD of 30 . 3 ± 2 . 4 ) while disease duration was longer than that of the sCJD cases , ranging from 8–32 months ( mean ± SD of 17 . 8 ± 8 . 5 ) . Western blot analysis of a tissue sample from the specimen used for MS using the method of Parchi et al . [33] showed that Type 2 PrPSc predominated in all six cases . In 4 of the 6 samples , this took the form of a Type 2 doublet ( 2d ) and in the other two it was found to be a mixture of Type 2 and a smaller amount of Type 1 ( 2+1 ) ( Table 2 and Fig 5a ) . As with the heterozygous cases of sCJD , there was no significant correlation between the amount of PrPSc-M129 and disease duration or age of onset ( p > 0 . 3 , Table 2 ) . However , as shown in Fig 5b , 4 of the 6 iCJD samples had significantly less PrPSc-M129 than PrPSc-V129 . When the average relative abundance of each PrPSc allotype for all cases was compared between the sCJD and iCJD groups , PrPSc-M129 was significantly less abundant in the iCJD cases than in the sCJD cases ( p = 0 . 01 using the unpaired Student’s t-test ) . Overall , our results suggest that the composition of PrPSc among MV heterozygous cases of iCJD is less variable and biased towards a higher proportion of PrPSc-V129 when compared to heterozygous cases of sCJD . Epidemiological , genetic and neuropathological data all suggest that human growth hormone associated cases of iCJD in the UK may be the result of contamination with the MV2 or VV2 subtype of sCJD [19 , 20] . We therefore compared the PrPSc allotype ratio to the major neuropathological phenotypes represented in both the sCJD and iCJD cases . As with disease onset and duration , there was no correlation between sCJD PrPSc allotype and Type 1 and Type 2 sCJD cases with similar neuropathological phenotypes ( Fig 6 ) . Both the sCJD MV1 + 2C and MV2K + 2C cases segregated into three distinct groups based on PrPSc allotype ratio with no significant difference in the mean percentage of PrPSc-M129 between the two types ( Fig 6a and 6b ) . Interestingly , sCJD MV2K and iCJD MV2K had a somewhat similar PrPSc allotype distribution with no statistically significant difference in the mean percentage of PrPSc-M129 ( Fig 6c and 6d ) . With the caveat that only 2 cases of sCJD MV2K were available for analysis , these data are nonetheless consistent with the hypothesis that UK cases of human growth hormone associated iCJD are the result of exposure to the V2 strain of human prion .
Efficient conversion of PrPC to PrPSc can be strongly dependent upon homology at a single amino acid residue in PrPC [35] , including residue 129 [36 , 37] , and differences in key amino acid residues can effect disease onset and progression [27 , 38 , 39] . However , we found no evidence that M or V at residue 129 in PrPSc correlated with disease onset or progression . This is consistent with some studies of familial human prion disease which also suggest that disease phenotype does not necessarily correlate with the presence of M or V at residue 129 in mutant PrPC [24 , 40] . It is possible that , in heterozygous cases of sCJD , the polymorphism at residue 129 does not necessarily lead to protease-resistant PrPSc conformations that are significantly different . Since PrPSc conformation is thought to encode prion strain phenotypes [1] , PrPSc-M129 and PrPSc-V129 might therefore contribute minimally to conformational differences affecting disease phenotype . Unknown host factors or other forms of abnormal PrP which may or may not be equivalently infectious , such as those with atypical protease-digestion profiles [23] or increased protease-sensitivity [21] , may have a more significant influence on age of onset or duration of clinical disease than the presence of an M or V at residue 129 in Type 1 or Type 2 PrPSc . The epidemiological associations of Type 1 PrPSc with methionine homozygosity and Type 2 PrPSc with valine homozygosity within sCJD patient cohorts could be used to argue that the presence of methionine at position 129 of the prion protein predisposes it to misfold to a Type 1 conformation , whereas the presence of valine at the same position predisposes it to misfold to a Type 2 conformation . If this were so , then one might predict that Type 1 PrPSc in heterozygous sCJD cases would be largely composed of PrPSc-M129 whereas Type 2 PrPSc in heterozygous sCJD cases would be largely composed of PrPSc-V129 [37] . Data from transmission studies in transgenic mice [11 , 12 , 41] and in vitro assays of PrPSc formation [37] are consistent with this prediction . It was therefore surprising that in our study we saw no correlation between PrPSc type and PrPSc allotype ( Fig 3b ) . Out of the 5 MV1 sCJD cases analyzed , PrPSc-M129 was the dominant allotype in only one case while PrPSc-V129 was the most prevalent allotype in 3 of the 9 MV2 cases analyzed ( Fig 3b ) . It is therefore unlikely that the amino acid residue at codon 129 in PrPC predisposes to accumulation of either the Type 1 or Type 2 PrPSc conformations associated with the MM1/MV1 and VV2/MV2 strains of sCJD . Despite the distinctive transmission properties of MV1 and MV2 sCJD when inoculated into transgenic mice expressing human PrPC [11 , 12 , 41] , the sCJD neuropathological phenotypes MV1 + 2C and MV2K + 2C could not be distinguished based upon their PrPSc allotype ratios ( Fig 6a and 6b ) . These data are based primarily on the relative abundance of PrPSc-M129 and PrPSc-V129 in the cerebral cortex . It is possible , however , that there may be some correlation with phenotypic differences between brain regions in the same patient . In cases 9 and 14 , where distinctive pathologies were observed in the cerebral cortex and cerebellar cortex ( Fig 2 ) , the PrPSc allotype ratios were also significantly different ( Fig 4 ) . By contrast , in two cases where neuropathology was similar between these two brain regions ( cases 3 and 7 , Figs 2 and 4 ) , the PrPSc allotype ratios were also similar . While these data are suggestive that PrPSc allotype may correlate with regional differences in neuropathology , further analysis of multiple brain regions from a greater number of patients will be necessary to prove this correlation . Different strains of sCJD could account for the variability in PrPSc allotype ratio between patients , although the consistent transmission properties of MV1 and MV2 sCJD into transgenic mice expressing human PrPC argues against this [11 , 12 , 41] . An alternative explanation for the heterogeneity in the MV PrPSc allotype ratio may lie in its proposed aetiology . In sCJD the initiating event is thought to be the spontaneous , but stochastic , misfolding of PrPC into PrPSc . Since either PrPC-M129 or PrPC-V129 could potentially misfold leading to PrPSc accumulation and disease , the PrPSc allotype might be expected to vary from patient to patient . Such variation is what we observed in our analysis where the relative abundance of PrPSc-M129 and PrPSc-V129 was independent of sCJD type , differing not only between sCJD cases , but also between brain regions from a single patient . Thus , our data are consistent with either PrPC-M129 and/or PrPC-V129 refolding into PrPSc by chance . Our data may also provide some insight into the relative tendency of PrPC-M129 and PrPC-V129 to spontaneously adopt the PrPSc conformation . In approximately one-third of all of the heterozygous sCJD cases analysed , PrPSc-M129 was the most abundant allotype , PrPSc-V129 was the most abundant allotype , or both allotypes were equally represented ( Fig 3b ) . The relatively equal representation of the different possible PrPSc allotypes does not appear to be consistent with one PrPC molecule being more predisposed to misfold into PrPSc than the other , as has been suggested by several in vitro studies [42–44] . Rather , our data are consistent with what would be expected if PrPC-M129 and PrPC-V129 were equally likely to spontaneously misfold into PrPSc . In this context , the differences in relative abundance between the two allotypes could be interpreted as indicating which PrPC molecule initially misfolded into PrPSc . Thus , in cases of heterozygous sCJD where PrPSc-M129 predominates , PrPC-M129 could have misfolded into PrPSc prior to PrPC-V129 . However , it is important to note that we have only analyzed a small portion of the brain from the end stage of a long and complex disease process . It is more likely that there are multiple variables , both host and prion specific , that contribute to the relative abundance of each PrPSc allotype at the end stage of disease . For example , the presence of M or V at residue 129 in PrPC could affect a non-rate-limiting step in PrPSc formation and accumulation which occurs after the initial misfolding of PrPC . A similar hypothesis was proposed by Hosszu et al . who reported that methionine or valine at residue 129 had no measurable effect upon the folding , dynamics , or stability of PrPC [45] . By contrast , there is evidence that there are differences in the stability of the downstream products of PrPC misfolding . PrPSc from homozygous cases of sCJD demonstrates a broad spectrum of stabilities [21] while short peptide ‘steric zipper’ structures derived from homozygous recombinant PrP molecules form more stable crystalline structures than the hypothetical structures modeled from heterozygous molecules [46] . Therefore , while similar free energy barriers for both PrPC-M129 and PrPC-V129 would suggest that they may be equally likely to misfold into PrPSc , variability in the thermodynamic stability of the PrPSc end products may be more important in determining the final relative abundance of each PrPSc allotype . In contrast with sCJD , the heterozygous cases of iCJD analyzed were more homogeneous , with a higher proportion of PrPSc-V129 in most of the cases ( Fig 5b ) . As with the sCJD cases , aetiology may provide an explanation for the greater PrPSc allotype homogeneity of the iCJD samples . In cases of iCJD prior to 2003 which were linked to human growth hormone therapy in the UK , 96% had the VV or MV PRNP genotype at codon 129 [19] . In patients diagnosed after 2008 , this percentage had decreased to 86% [20] . Based on the earlier cases , Brandel et al . [19] first proposed that the UK cases were caused by either sCJD VV2 or sCJD MV2 , a conclusion supported by the more recent study [20] . Our data showing that 1 ) PrPSc-V129 is the dominant PrPSc allotype in the majority of the UK human growth hormone-associated iCJD cases analyzed and 2 ) that the PrPSc allotype distribution between sCJD MV2K and these cases ( Fig 6c and 6d ) is very similar , are also entirely consistent with this conclusion . It is thus possible that the PrPSc-M129 to PrPSc-V129 ratio in heterozygous iCJD patient cohorts might be a means of determining the prion strain that initiated infection . Unlike sCJD , iCJD is a secondary infection of CJD in humans . Selective pressures in the original host with sCJD as well as in the secondary host with iCJD may influence the final PrPSc allotype ratio . If PrPSc-V129 in the infectious material was of the VV2 type , when transmitted to a second host it would likely interact most efficiently with PrPC-V129 to form PrPSc . As a result , PrPSc-V129 would more than likely be the predominant PrPSc allotype in the infected iCJD patient . Used in this way , mass spectrometry analysis of heterozygous PrPSc could be the molecular equivalent of in vivo traceback studies that have used the transmission properties of iCJD into transgenic mice [47 , 48] or non-human primates [10] to determine the prion strain originally responsible for infection in groups of patients .
Trypsin was purchased from Promega . Burdick & Jackson water and acetonitrile ( ACN ) were purchased from VWR . Imperial Coomassie blue stain and iodoacetamide were purchased from Thermo-Fisher Scientific . Formic acid ( FA ) , dithiothreitol ( DTT ) , trifluoroethanol ( TFE ) , phosphotungstic acid ( PTA ) , benzonase , and protein extraction reagent ( 7M urea , 2M thiourea , 1% C7BzO , 40 mM Tris , pH 10 . 4 ) were purchased from Sigma-Aldrich . Human brain samples were obtained from the National CJD Research & Surveillance Unit Brain and Tissue Bank in Edinburgh , UK , which is part of Edinburgh Brain Banks . For the purposes of this study , samples were pseudoanonymized using a Brain Bank reference number . Cerebellar and cerebral cortex tissue from two additional sCJD patients were obtained from the University of Verona , Italy . These tissues were obtained at autopsy and sent to the Neuropathology Unit at the University of Verona for statutory definite diagnosis of CJD . All UK cases had consent for research and their supply and use in this study was covered by LREC 2000/4/157 ( National Creutzfeldt-Jakob disease tissue bank: acquisition and use of autopsy material for research on human transmissible spongiform encephalopathies , Professor James Ironside , amended date: 9th October 2007 ) . Ethical approval for the acquisition and use of human brain material was obtained from the National Institutes of Health ( NIH ) Office of Human Subject Research ( Exempt #11763 and #12725 ) and no patient identifiable data was transferred to the NIH . Human recombinant prion proteins spanning residues 23–231 and containing either methionine or valine at residue 129 ( rHuPrP-M129 and rHuPrP-V129 , respectively ) were cloned into a pET41a vector using NdeI-XhoI restriction enzymes . The proteins were expressed as inclusion bodies using Escherichia coli Rosetta cells and then purified as described for recombinant hamster PrP [49] . Briefly , guanidine denatured PrP from bacterial inclusion bodies was clarified by centrifugation , bound to NiNTA resin ( Qiagen ) , and then subjected to on-column refolding without reducing agents using a non-denaturing refolding buffer ( 10mM Tris , 100mM sodium phosphate , pH 8 ) . After elution using refolding buffer adjusted to pH 5 . 8 and 500mM imidazole , PrP was immediately sterile-filtered , dialyzed into 10mM ammonium formate pH 4 . 5 , and stored at -80°C at a concentration of ~0 . 3mg/mL until needed . Consistent with our previous results [49] , all PrP preparations were approximately 99% pure . In addition , the molecular weight of purified recombinant PrP was verified by intact mass analysis using a Sciex 4000 QTrap system . Synthetic PrP111-136 peptides containing either M or V at residue 129 were diluted in water with 3% ACN/0 . 1% formic acid and used without further modification after having been subjected to HPLC and MALDI in order to establish purity and the correct molecular weight . Concentrations of purified rHuPrP-M129 and rHuPrP-V129 proteins were determined individually by absorbance at 280nm and then adjusted to contain molar ratios of 0:100 , 25:75 , 50:50 , 75:25 , or 100:0 rHuPrP-M129 to rHuPrP-V129 , respectively . Final mixtures containing 0 . 3mg/mL total rPrP were mixed 1:1 with 2X sample buffer ( Life Technologies ) , boiled for 5 min and subjected to SDS-PAGE . Coomassie blue staining displayed a single strong band which was excised and subjected to in-gel trypsin digestion as described below . At least 8 lanes ( n = 8–9 ) were used per individual rPrP mixture and each was subjected to mass spectrometry analysis as described below . Approximately 2g samples of cerebral cortex ( CC ) and ( in some cases ) cerebellar cortex ( CbC ) from sCJD , iCJD and non-CJD neurological controls were analysed . All cerebral cortex samples were of grey matter enriched frontal cortex with the exception of cases 3 and 14 which were from the occipital cortex . Additionally , grey matter enriched cerebellar cortex samples were analyzes from cases 3 , 7 , 9 and 14 . The choice of frontal cortex reflects both its involvement in MV1 sCJD , MV2 sCJD , and UK human growth hormone-related MV iCJD as well as tissue availability . The selection of cerebellum as a second region to analyze was based on it providing a contrast to the cerebral cortex as well as the known pronounced involvement of the cerebellum in MV2 sCJD . The diagnosis of definite sCJD or definite iCJD had previously been reached using internationally recognised criteria ( http://www . cjd . ed . ac . uk/documents/criteria . pdf ) and the cases had been characterised in terms of their neuropathology , PRNP codon 129 genotype , and predominant PrPSc type . The 20 CJD cases examined comprised 14 cases of sCJD in heterozygous patients ( 5 of the MV1 and 9 of the MV2 molecular subtypes ) and 6 cases of iatrogenic CJD in heterozygous patients ( all of which were of the MV2 molecular subtype ) . The three non-CJD neurological cases analyzed were initially suspected of having CJD , but post mortem examination resulted in an alternative diagnosis . These three patients were a 78 year old male with a pathological diagnosis of Lewy body dementia , a 66 year old female with a pathological diagnosis of arteriosclerosis and a seventy year old female with a pathological diagnosis of arteriosclerosis . Western blot analysis had previously failed to detect PrPSc in the brain in these cases and the PRNP codon 129 genotypes of the three patients were VV , MV and MM respectively . A detailed neuropathological review of the sCJD and iCJD cases analysed in this study was conducted and cases were sub-classified with reference to the sCJD histotypes reported by Parchi et al [5] . To confirm the presence of PrPSc and to provide a definitive type for the PrPSc present in the brain specimen being used for mass spectrometry , a 100mg tissue sample was removed from each brain specimen and analysed by western blotting . The stringent sample preparation and proteinase K digestion conditions of Parchi et al 2009 were used [33] that allow distinction to be made between Type 1 ( 1 ) , Type 2 ( 2 ) , a mixture of Type 2 and Type 1 ( 2+1 ) and doublets of Type 2 ( 2d ) . Brain tissue samples for mass spectrometry were homogenized in phosphate buffered saline ( PBS , pH 7 . 4 ) to a final concentration of 20% ( w/v ) using a Mini-BeadBeater-8 ( Biospec ) and stored in aliquots at -80°C until needed . Enrichment of PrPSc was performed using PTA precipitation as described previously [50] with some modifications [51] . Briefly , 250 μL of a 20% brain homogenate was diluted to 10% with PBS and then mixed with 500μL of 4% sarkosyl/PBS . Following incubation at 37°C for 30 min , samples were treated with benzonase at 50U/mL for 1 hour and then clarified by a 5 min centrifugation at 5 , 000 g . The supernatants were then digested with PK ( 50 μg/mL ) for 1 hour at 37°C . PTA was added to a final concentration of 0 . 3% ( w/v ) and the sample incubated at 37°C for 2 hours . Samples were centrifuged for 30 min at 16 , 100 x g , the pellets suspended in 200μL of PBS with 200mM EDTA , incubated at 37°C for 30 min , and centrifuged as before . The pellets were resuspended in 200μL PBS , incubated at 37°C for 30 min , and the sample again collected by centrifugation . The final pellet was used for analysis of PrPSc by mass spectrometry as detailed below . Samples of PrPSc pelleted by PTA precipitation were solubilized in 25μL of protein extraction reagent ( 7M urea , 2M thiourea , 1% C7BzO ) and DTT added to a final concentration of 14mM . Following a 30 min incubation at 37°C , iodoacetamide was added to a final concentration of 75mM and the sample incubated in the dark at room temperature for 30 min . The reaction was quenched by the addition of 1M DTT to a final concentration of 200mM followed by the addition of 12μL of 4X NuPAGE sample buffer ( Life Technologies ) . Denatured , reduced and alkylated samples of enriched PrPSc were briefly heated to 100°C using a heating block prior to loading onto a 10% Bis-Tris 1 . 5mm gel for SDS-PAGE . Gels were stained with Coomassie blue ( Imperial stain , Thermo-Fisher Scientific ) but the bands corresponding to PrPSc were often too faint to be detected by eye . Therefore , the area of the gel lane encompassing molecular mass markers from ~10kDa to 80kDa , including the area of the gel containing PrPSc , was divided into 8 gel slices . In-gel trypsin digestion was then performed as described previously [28] . Each patient sample was analyzed at least 3 times ( i . e . 3 technical replicates ) as detailed above except for the CC from cases 4 , 7 and 9 where only two technical replicates were done . Thus , each technical replicate was composed of 8 individual LC-MS/MS runs corresponding to 8 gel slices excised from a single lane of the Commassie blue stained gel . Trypsin-digested peptides were identified by LC-MS/MS ( abbreviated as MS throughout the manuscript ) using an Agilent 1200 HPLC system interfaced with a 6330 XCT Ultra Ion Trap via a chip-cube nanospray source . The mass spectrometer was externally calibrated using a tuning mix provided by Agilent . Data-dependent MS acquisition was performed with dry gas ( purified air ) set to 4L/min at 350°C , MS capillary voltage 1800V , and a maximum accumulation time of 150ms . The MS scan range was set to 300–1400 m/z in the Ultrascan mode . Four parent ions were selected for each MS/MS cycle with a fragmentation amplitude of 1 . 0V . Protein digests were loaded onto the HPLC chip ( Agilent #G4240-62006 , Zorbax 300SB-C18 , 5μm , 75μm x 150mm ) with an autosampler and washed with Buffer A ( 3% ACN/H2O and 0 . 1% FA ) prior to elution at 300nL/min by reverse-phase chromatography . The gradient was 3–30% Buffer B ( 90% ACN and 0 . 1% FA ) over 30 min , to 50% B by 40 min , 80% B by 50 min , 90% B by 60 min , 97% B by 70 min , and back to 3% B by 78 min . This run was followed by a 4 min post-run re-equilibration at 3% B with a total run time of approximately 80 min . Raw data were processed into MGF peak lists using MASCOT Distiller v2 . 4 . 3 . 0 . The MGF files were searched using MASCOT Daemon against a target database ( www . uniprot . org ) filtered for human taxonomy consisting of 20 , 235 entries which included an entry for PrP111-136 containing V at residue 129 in addition to the default Swiss-Prot entry containing the full PrP sequence with M at residue 129 . The trypsin/P search parameters for MASCOT protein identification consisted of one missed tryptic cleavage allowed with a fixed carbamidomethylation ( +57 , Cys ) and a variable oxidation ( +16 , Met ) . Mass tolerances of 2 . 0 and 1 . 0 Daltons were used for parent and monoisotopic fragment ions , respectively . The resulting DAT files generated by MASCOT were used as input files for spectral counting using the ProteoIQ bioinformatics platform ( Premier Biosoft Inc ) [52] with the constraints that only MASCOT ion scores of ≥ 30 and only peptides of ≥ 7 amino acids in length were considered in these calculations . The number of PrP spectra from each sample were then compared to the other samples that had been processed using the same protocols and the same instrument parameters . In PTA precipitates from the 3 non-CJD brains examined , only 5 PrP peptides were detected in a single sample following a total of 72 MS runs . This indicated that little or no PrPC was present in the final non-CJD PTA pellet . A background threshold of ≥5 total PrP peptides was therefore used to exclude any potential contribution of PrPC to the final PTA pellet and all 3 non-CJD samples were considered negative for PrPSc . By contrast , all PTA precipitated sCJD and iCJD samples had total PrP spectral counts at least 26-fold above the background threshold of 5 spectral counts . Only western blot positive samples yielding greater than 5 PrP111-136 peptides were utilized for semi-quantitation as spectral counts of less than 5 spectra are considered to be unreliable for spectral counting [32 , 53] . Mean and standard deviation ( SD ) were derived using the individual technical replicates described above . The percentage of PrPSc-M129 and PrPSc-V129 was calculated by dividing the number of peptides containing M129 or V129 by the total number of peptides containing residue 129 and multiplying by 100 . The statistical significance of multiple data sets was determined using a 1-Way ANOVA with Dunnett’s post-test . A 50:50 mix of rHuPrP-M129 rHuPrP-V129 was set as the control group . The potential relationship between disease onset or duration and the percentage of PrPSc containing M at residue 129 was analyzed by determining the Pearson correlation coefficient r using correlation analysis . The unpaired Student’s t-test was used to analyze the percentage of PrPSc-M129 between different brain regions and the percentage of PrPSc-M129 in sCJD versus iCJD . All calculations were done using the GraphPad Prism software package , version 6 . 04 . | In Creutzfeldt-Jakob disease ( CJD ) , heterozygosity at residue 129 for methionine or valine in normal prion protein may affect disease phenotype , onset and progression . However , the relative contribution of each prion protein allotype to the infectious , disease associated form of prion protein ( PrPSc ) is unknown . Here we report the novel observation that in heterozygous cases of sporadic CJD the PrPSc allotype ratio is highly variable . This case-by-case variability is consistent with the origin of sporadic CJD being the spontaneous , but random , misfolding of either host prion protein allotype into infectious PrPSc . By contrast , in heterozygous cases of iatrogenic CJD in the United Kingdom resulting from exposure to contaminated human growth hormone , the PrPSc allotype ratio is much more homogeneous and consistent with exposure to infectious PrPSc containing valine at residue 129 . Surprisingly , the PrPSc allotype ratio did not correlate with disease onset or duration in either disease type . Thus , factors other than PrPSc allotype ratio likely influence the clinical progression of heterozygous cases of CJD . Moreover , our results suggest that the ratio of methionine to valine in PrPSc may be a means of determining the origin of prion infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"valine",
"chemical",
"compounds",
"aliphatic",
"amino",
"acids",
"neurodegenerative",
"diseases",
"cerebellar",
"cortex",
"brain",
"organic",
"compounds",
"amino",
"acids",
"infectious",
"diseases",
"zoonoses",
"proteins",
"homo... | 2016 | The Distribution of Prion Protein Allotypes Differs Between Sporadic and Iatrogenic Creutzfeldt-Jakob Disease Patients |
In natural ecosystems , hundreds of species typically share the same environment and are connected by a dense network of interactions such as predation or competition for resources . Much is known about how fixed ecological niches can determine species abundances in such systems , but far less attention has been paid to patterns of abundances in randomly varying environments . Here , we study this question in a simple model of competition between many species in a patchy ecosystem with randomly fluctuating environmental conditions . Paradoxically , we find that introducing noise can actually induce ordered patterns of abundance-fluctuations , leading to a distinct periodic variation in the correlations between species as a function of the phenotypic distance between them; here , difference in growth rate . This is further accompanied by the formation of discrete , dynamic clusters of abundant species along this otherwise continuous phenotypic axis . These ordered patterns depend on the collective behavior of many species; they disappear when only individual or pairs of species are considered in isolation . We show that they arise from a balance between the tendency of shared environmental noise to synchronize species abundances and the tendency for competition among species to make them fluctuate out of step . Our results demonstrate that in highly interconnected ecosystems , noise can act as an ordering force , dynamically generating ecological patterns even in environments lacking explicit niches .
Species abundances and their variation over time are quantities of fundamental importance in any ecosystem: understanding the forces that shape them is a key part of central problems in ecology , ranging from conceptual questions about the role of neutral processes [1] , [2] to practical issues in biodiversity conservation [3] . One major driver of changes in species abundances is environmental influences which vary across time and space , such as the weather [4]–[6] . A classic example of an ecological phenomenon caused by such environmental noise is the Moran effect , the tendency for a shared fluctuating environment to synchronize the variations in abundance among species and across space [7]–[10] . This effect has now been studied in systems with colored noise [11]–[13] and species dispersal [14] , and in small food webs [15]–[19] . The synchronizing effect of noise , however , is opposed by negative interactions between species ( e . g . through resource competition or predation ) which cause compensatory dynamics: when the abundance of one species increases , the abundance of others tend to decrease , creating out-of-step variations [20] . Although significant progress has been made towards quantifying the total impact of each of these factors [21]–[23] , it remains unknown how the tension between them influences the dynamics in natural ecosystems . In such systems , many phenotypically distinct species are embedded in a tangled web of direct and indirect interactions that make it hard to predict the effect of even simple disturbances [24]–[26] , and non-trivial collective effects could play a significant role . For instance , even in the absence of noise species interactions can lead to static , clumped patterns across phenotype space [27] , providing a possible explanation for the widely observed tendency for species in a given ecosystem to cluster around a few preferred body sizes [28] , [29] . Such phenotypic patterns could be ubiquitous but have received relatively little attention [30] . The idea that the interplay between environmental noise and inter-species interactions could lead to non-trivial effects is supported by both theoretical and empirical studies of ecosystem dynamics . Even single- or few-species ecological models exhibit a range of complex behaviors , including bifurcations and chaos [31] , strong amplification of environmental noise [32]–[34] , noise-induced oscillations [35] , [36] , and pattern formation driven by demographic fluctuations [37] . Empirical observations in nature and laboratory experiments have similarly revealed complex dynamics , including chaotic behavior [38] , [39] , environmental noise and density-dependence intermingling in determining single species abundances [9] , and cases where synchrony in the abundance of a single species across landscapes propagates down a food-web [40] . In this article , we show that environmental noise can indeed lead to robust , dynamic patterns in phenotype space . We introduce a simple model of the combined effect of noise and competition in an ecosystem with many species differing in their reliance on growth rate and efficiency , respectively , for survival . To focus on dynamically emerging patterns rather than on pre-imposed niche differences , we use a minimalist patch-model framework in which all species compete for a single resource and undergo periodic , global dispersal between the patches . Each species is entirely defined simply by its rate of growth and its efficiency in turning resources into offspring . We start by considering the model behavior in a fixed environment , showing that it allows many species to coexist stably . We then introduce external environmental noise and show that it gives rise to systematic and robust alternating patterns of species-species correlations which are accompanied by the formation of dynamic clusters of abundant species in phenotype space . Finally , we show that these patterns directly reflect a balance between the tendency of noise to synchronize different species and the tendency of competitive interactions to create abundance-differences .
Our patch model is similar to both the theoretical model proposed by Wilson [41] and to ( the metapopulation version of ) the experimental yeast system of MacLean and Gudelj [42] . The specific formulation was inspired by the rich microbial communities found in soil ( which exhibit many of the same broad ecological patterns as macroscopic species [43] ) , but its basic features – patchiness , repeated environmental disturbances , and the presence of a range of different phenotypic strategies – are shared by many ecosystems . In this sense , for instance , our model is similar to a model of competition between grasses analyzed by Tilman [44] , [45] . Hence , we believe that our conclusions will also be relevant to many macroscopic ecosystems . A key feature of the soil environment , as experienced by microbes , is its granular nature , with dividing cells typically found in separated pockets in the soil matrix [46] . These communities are not static: cells are constantly dispersed by weather and fresh resources are added and washed away continuously . Our model describes an ecosystem of N species competing for a single resource on multiple patches containing a fixed amount of the resource ( Figure 1 ) . The dynamics consists of repeated , two-phase cycles of local reproduction of individuals on their patches until the resource is depleted , followed by global dispersal to fresh patches ( representing periodic environmental influence due to e . g . rainwater ) . The appearance of full nutrient patches can represent either the dispersal to existing but hitherto unoccupied locations or the addition of new resource by the environmental disturbance ( e . g . deposited by water flow ) . Each species is described by two basic metabolic parameters , growth rate and efficiency [47] , allowing us to consider the behavior of many species spread along continuous phenotype axes . Since efficiency would not confer an advantage unless resource availability is what limits growth , the model assumes that dispersal happens only after all resources have been exhausted . This assumption applies whenever disturbances are rare compared to the typical rates of growth , either because the dispersal events are intrinsically spaced out or because the resources are so finely divided that they only support short bursts of growth . An example of the first case is ecosystems where dispersal represents a yearly occurrence ( e . g . for seeding plants ) , while the second case is likely to apply to e . g . microbes feeding off scattered organic matter in soil or the ocean ( ‘marine snow’ [48] ) . For simplicity , we assumed that all nutrient patches are identical and always contain the same amount of resource at the beginning of a cycle . We also worked in the limit of infinitely many patches and hence infinitely large populations , allowing us to consider the impact of environmental noise on species abundance without complications due to demographic stochasticity . Growth cycle number t starts with a global seeding pool in which the abundance per patch of each species is given by the vector n ( t ) = ( n1 ( t ) , n2 ( t ) , … , nN ( t ) ) . From this pool , a fraction α of individuals randomly gets seeded onto a new collection of patches , while the remaining fraction , ( 1−α ) , of the cells is washed out of the system . We assumed α is very small so that the probability that a patch receives a total of m1 individuals of species 1 , m2 of species 2 etc . is a product of Poisson probabilities: ( 1 ) where m = ( m1 , m2 , … , mN ) . The two traits characterizing each species are: ( 1 ) growth rate , μ – the rate of exponential reproduction on a nutrient patch while resources are available , and ( 2 ) efficiency in turning nutrients into offspring , Y – the number of offspring that can be produced by a single individual if it consumes all the resource on a patch . After seeding , each individual of species k starts replicating at rate μk while consuming the shared resource on its patch at a rate of 1/Yk units per offspring . Growth on a given patch stops when the resource on that patch is depleted . The time at which this happens ( T ) is a function of the initial abundance of each species on the patch , as well as of their growth rates and efficiencies , i . e . T = T ( m;μ , Y ) , where the vectors μ and Y represent the growth and efficiency parameters for all species , respectively ( see Methods ) . The final abundance of species k , averaged across all patches with this seeding , is then simply ( 2 ) Since the interval between dispersal events is assumed to be longer than all growth-times , only the final abundances matter . The new average per-patch abundances , n ( t+1 ) , after all growth has stopped is found by averaging these final abundance over all possible seeding configurations: ( 3 ) where f ( m ) = ( f1 ( m ) , f2 ( m ) , … , fN ( m ) ) . Equation 3 is the fundamental dynamical equation for the per-patch abundances at the end of growth phase . It expresses the fact that final species abundances in one cycle determine the abundances in the next by setting the probabilities of the various possible initial seedings . Details of the model and simulations are given in the Methods section . We note that dispersal and the availability of new resources are assumed to be linked . Such linkage is natural if both are driven by the same external factor ( e . g . rainfall dispersing bacterial cells and depositing new resources ) or if one of them is driving the other . For instance , dispersal can effectively generate new resources if empty patches with new resources are always available and are simply being invaded by dispersal . While models of competition for a single resource typically lead to competitive exclusion – a single species comes to dominate and drives all others extinct [49] , [50] – division into patches can allow many species to coexist [45] , [51] . Indeed , numerical simulations of our model for fixed α showed that many species can be stably maintained ( Figure 2 ) , and it can be argued explicitly that arbitrarily many species can coexist if the amount of resource on each patch is very large ( see Methods ) . The stabilizing mechanism that makes coexistence possible can be understood as a frequency-dependent selection during the growth-phase . When the total population density fluctuates up , patches are more likely to be seeded with more species , which intensifies competition and promotes selection for fast growth . If fast-growing species are also less efficient , their increased frequency drives the total population density back down . Conversely , when the population density is decreased , species have a higher probability of growing on patches with few or no competitors . This allows high-efficiency species to grow to high densities even if they are growing slowly , leading to an increase in the overall population . These growth-phase selection pressures – favoring speed ( μ ) and yield ( Y ) , respectively– are examples of R- and K-selection [52] , and can also be interpreted in terms of different levels of selection introduced by the division of the population into isolated groups [53] . The frequency-dependent fitness can lead to stable , steady-state solutions ( fixpoints ) , n* , of Equation 3 such that n ( t+1 ) = n ( t ) = n*: species abundances relax back to their steady state values following small perturbations ( Figure 2 ) . For such stabilization to work , however , constraints must prevent species from optimizing both growth and efficiency simultaneously and hence form a ‘super-species’ that will drive all other species extinct [50] . Cost-benefit reasoning suggests that such trade-offs will indeed generically be present , e . g . high efficiency will typically require more extensive metabolic machinery and hence divert energy away from cellular reproduction [54] , and plants must divide their resources between e . g . root and seeds [55] . Such trade-offs have indeed been found empirically in a number of contexts [55]–[58] , and trade-offs between the rate and efficiency of resource utilization has been shown to allow two distinct strains of yeast to coexist [42] . As our focus is on the dynamics of the ecosystem rather than its assembly through evolution , we will assume the existence of appropriate μ-Y trade-offs which allow community coexistence . Because of the stabilizing mechanism , trade-offs do not uniquely fix μ and Y for each species; instead , a range of different values are possible ( each leading to different steady state abundances ) , albeit the range of parameters choices narrows as two species become very similar ( Supplementary Figures S1 and S2 ) . To have an unbiased baseline , we chose sets of parameters that lead to equal species abundance at steady state , i . e . nk* = n0 for all species k . Given n0 , μ , and α , we can numerically solve the fixpoint equation n ( t+1 ) = n ( t ) for the species efficiencies Y using Equations 1 and 3 – see Figure 2A . We introduced shared environmental noise through fluctuations in the dispersal dilution factor α which represents the strength of the environmental disturbance and affects all species in each step . Specifically , we drew an independent , random α-value in each cycle ( white noise ) from a fixed log-normal distribution . This choice is convenient for keeping the expectation value of the long-term dilution factor fixed as we changed the noise intensity , but our conclusions do not depends on the exact distribution ( see Methods ) . The environmental noise was strongly amplified: a 15% variation in α around the mean causes both the total abundance and that of individual species to fluctuate over several orders of magnitude ( Figure 3A ) . Individual species exhibited short ‘bursts’ of high abundance and occasionally maintained a relatively high abundance over long periods . No single species permanently gained the upper hand – instead , there was a constant , slow turnover of species , reminiscent of that observed in plankton communities [59] . But while the fluctuations in the abundance of any single species are erratic , the competitive interactions acted to create a striking coherent pattern in the relative fluctuations of different species . At any typical time , the most abundant species formed clusters in phenotype space , separated by ‘valleys’ of low-abundance species ( Figure 3B and Supplementary Figures S3 , S4 , and S5 ) . Due to the turnover of dominant species , the number , and height of clusters changed over time , but the peak-and-valley pattern itself was robust . Furthermore , peaks tended to have approximately the same width in phenotype space . This clustered pattern remained when averaging over many cycles , albeit with a smaller amplitude ( Figure 3B , bottom panel ) , and also appeared across replica systems started at different random configurations . Increasing the noise intensity has little impact on the typical size of the clusters , but naturally leads to larger abundance differences . At very high noise levels , non-linear effects – presumably related to the stabilizing mechanism discussed above – stabilizes rare species at low densities , leading to clusters separated by very distinct valleys ( Supplementary Figure S6 ) . Extinction of species can occur at very high noise levels , but was never observed at the noise strengths discussed in this paper . To understand how the phenotypic clusters are formed , we looked at the pair-wise correlation between species abundances in simulations of the complete model and constrained versions of it ( data series of 105 cycles ) . When plotted as a function of the phenotypic difference between them , the correlation between two species in the complete model alternates between positive and negative values ( Figure 4A , purple ) , reflecting the clustering we observed in Figure 3B ( since ‘peak-species’ move in synchrony with one another , but out of step with ‘valley-species’ ) . To separate the contribution of noise and species interaction to this oscillatory pattern , we repeated the simulation with the exact same noise ( same series of α-values ) while artificially fixing the abundance of either all but one , or all but two species , to their steady state values . These two types of simulations maintain the properties of the steady state while singling out the contribution of the noise itself and the pair-wise interactions combined with noise , respectively . For the single-species version , we simulated each species separately ( N simulation runs ) and computed pair-wise correlations between the different simulations; for the pairs , we simulated all pairs ( N2 simulations ) and computed the correlation of every pair of species within the corresponding simulation . We found that when each single species fluctuates independently , the full dynamics is determined by the noise and all species remain strongly positively correlated with each other regardless of how different they are ( Figure 4A , black; no interactions – see also Supplementary Figure S5 ) . Allowing pairs of species to fluctuate keeps similar species positively correlated , but causes species which are sufficiently phenotypically different become anti-correlated ( Figure 4A , green; pair-wise interactions ) . Hence , one- or two-species dynamics lead to the standard behaviors – Moran effect and compensatory dynamics , respectively . The latter effect is also visible in the response to an instantaneous increase in the abundance of a single species: the abundances of the other species drop ( Supplementary Figure S7 ) . The combination of noise and pair-wise interactions account correctly for the positive correlation between close species and for the negative correlation with some distant species , as seen in the complete model . However , pair-wise interactions alone are not sufficient for explaining the alternating patterns of multiple peaks of positive and negative correlations: this is a collective phenomenon requiring the interaction of many species . It only appears as we increase the number species allowed to fluctuate ( Supplementary Figure S8 ) . The mechanism behind the species clustering in phenotype space can be understood as a dynamic balance between the smoothing ( synchronizing ) effect of noise and the roughening effects of interactions . When the system is perturbed by a change in the dilution parameter α , all the species change their abundances by similar amounts and in the same direction , generating a relatively smooth ( uniform ) change in the abundance profile across phenotype space . As shown above , if the species do not interact with each other they will move up and down in almost perfect lockstep and hence maintain a flat uniform profile ( equal abundances ) . But if the species do in fact all compete , moving in lockstep means that every species experiences either increased or decreased competition from all the others after a perturbation and hence quickly gets pushed back to the fixpoint . If , for instance , all species simultaneously become more abundant , the resulting shortage of food will quickly decimate each one of them . Now suppose instead that the system is in a state where some species are above their fixpoint abundances and others below it – i . e . have an abundance profile that oscillates up and down . In that case , each species experiences a combination of less competition from species that are below their normal abundance and more competition from over-abundant species . These competitive differences partially cancel each other out , leading to a decreased pull on the abundance of each species and hence a slower relaxation back to the steady state . The more rugged the profile , the slower the relaxation: if similar species can have very different abundances , they can better cancel out each other's effects . We conclude that noise tends to generate smooth abundance profiles across phenotype space but , conversely , that the most stable profiles are the very jagged ones . We therefore expect that the typical abundance profile we observe is one that is neither completely flat nor maximally jagged , but instead changes smoothly between high and low abundances i . e . exhibits clusters of abundant species . This heuristic argument can be tested rigorously by considering a simplified version of our model ( Figure 4B ) . By expanding Equation 3 around the fixpoint n* and keeping only the leading ( linear ) terms , we obtain a good approximation for weak noise ( see Methods ) . The interactions between species are now described by a single N×N matrix J , and the eigenvectors of this matrix describe N independent deformations of the abundance profile around the steady state . These basic deformations can be sorted by their smoothness in phenotype space and are ordered accordingly on the x-axis in Figure 4B – three example profiles are illustrated in the bottom panels . The presence of both positive and negative elements in all but the first deformation is a direct reflection of compensatory dynamics: they involve some species growing more abundant while others become rarer . For each deformation , we calculated its propensity to be generated by noise ( Figure 4B , squares ) , and the time it takes for it to decay back to the flat steady state ( Figure 4B , triangles ) – see Methods for details . The results confirm the argument above: the two properties change in opposite directions as the profiles become more jagged . The environmental noise tends to generate smooth deformations , but the jagged deformations are much more long-lived . Statistically , the typical profile will therefore be one showing smooth peaks a few species wide ( Figure 4B , red line peaking at middle smoothness ) . Changing the noise intensity multiplies the amplitude of each deformation with the same constant and so does not affect the typical cluster size ( see Methods ) . This analysis agrees excellently with what we observe in our simulations: persistent clustering , with clusters having the same typical size even though the exact abundance profile is constantly changing due to the stochastic noise ( compare Figure 3C and the middle of the bottom panels in Figure 4B ) . The amplitude distribution ( red line in Figure 4B ) also agrees well with simulations ( Supplementary Figure S9 ) . The linear analysis also reveals the origin of the strong noise amplification: Although the parameters were not chosen to bring this about , the system is very close to instability , with the most jagged abundance deformation taking τ∼107 cycles to decay back to the fixpoint ( for the parameters used in Figures 3 and 4 ) . By the same token , a permanent shift in α ( a press perturbation ) will lead to significant shift in the stead-state abundances; the stabilizing mechanism discussed above acts only on changes in the abundances themselves ( see also Supplementary Figure S10 ) .
Our results show that the interplay between environmental noise and species interactions can induce robust patterns of alternating correlations between species abundances , leading to dynamic clustering of abundance in phenotype space . We demonstrated that the fundamental basis for this pattern is the dynamic balance between synchrony caused by noise ( Moran effect ) and the compensatory dynamics caused by the species interactions . Environmental noise is thus not merely a randomizing or synchronizing force , but can actively create ecological patterns that do not directly reflect fixed external factors like niches . These are collective phenomena requiring the presence of many species , suggesting that few-species ecological models may miss entire classes of dynamic behavior that could be important in natural ecosystems . By pointing to environmental noise as an important structuring factor in ecosystems , these results could cast new light on a number of empirical observations . For instance , metabolic theory suggests that body mass M is linked to maximal growth rate through the scaling relation [60] , so the clusters we observe across different growth rates could be directly reflected in cluster in the space of body mass . And indeed , body size cluster have been found to be dynamic in several cases , with the location of the clusters and their number changing over time [61]–[63] . Our model provides a simple mechanism for such itinerant clusters and at the same time offers a way to reconcile metabolic theory , which suggest the existence of single optimal body size , with the empirical observation that species rarely cluster at a single optimum [29] . Dynamic phenotypic clustering also implies that even species which are all direct competitors can arrange themselves into distinct sub-groups whose abundances fluctuate in synchrony for long periods of time ( Figure 4A ) . This lends support to the suggestion that the apparent lack of strong negative correlations between species found in large-scale empirical studies [64]–[66] could be due to obscuring effects rather than the actual absence of negative interactions [67] . The formation of phenotypic clusters bears some resemblance to the classical concept of limiting similarity: the idea that competition puts a limit on how similar the phenotypes of coexisting species can be , and hence implying that two neighboring species must have a finite stretch of unoccupied phenotype space between them [68] . The sensitivity to environmental fluctuation in our model means that at a permanent shift in α could drive some species extinct and thus effectively lead to a new , larger phenotypic separation of neighboring species . Conversely , for Lotka-Volterra models it has been shown that a very small perturbation in the parameters can shift the system from allowing the coexistence of arbitrarily similar species to requiring a finite phenotypic difference [69] . If environmental fluctuations drive such an ecosystem back and forth between these two regimes fast enough to keep many species from going extinct , the result could be bands coexisting species similar to the clusters we observe . As with all ecological modeling , we have made a number of simplifying assumptions . Firstly , we have ignored spatial structure beyond that provided by the division into patches . Secondly , we have worked in the limit of an infinite population size and hence neglected demographic noise ( neutral ecological drift ) . Finally , we have assumed a pre-existing trade-off between efficiency and growth rate . The question of how such tradeoffs can evolve and how they affect ecosystem stability is complicated [70]–[73] , and it would be interesting to understand it in the framework of our model . Indeed , the noise-induced clusters describe here could themselves play a role in speciation and the maintenance of genetic diversity [74]–[76] . Our model assumes that all patches contain the same amount of resource and deviations from this assumption are beyond the scope of this mode . However , we expect that if the resource amount on each patch was drawn independently from a fixed distribution in each round , the noise would simply average out and the model would converge to a steady state of coexistence similarly to that observed in our model . A slightly different natural variation would be to consider noise that affects the average amount of resources available on each patch rather than the dilution factor . A change in the amount of resource per patch is equivalent to a uniform rescaling of all efficiencies ( see Methods ) and therefore , like a change in dilution , will generically shift the balance between fast and slow species . We would therefore expect such fluctuations to cause qualitatively the same effects as we observe . Another possible variation of our model is to allow dispersal to occur before growth has finished on all patches . This would lower the advantage conferred by higher efficiency , so coexistence would require a steeper trade-off between growth-rate and efficiency . Indeed , in the limit of dispersal time much shorter than growth time , the model simply converges to exponential growth in a well- mixed environment; the efficiency becomes irrelevant and the fastest species takes over the population . The appearance of dynamic phenotypic clusters in such a minimal simplified model suggest that species clustering in phenotype space could be a generic property of ecologies with many interacting species subject to noise . Indeed , the underlying mechanism is quite general and other noisy systems involving many interacting parts , e . g . neuronal or molecular networks , might exhibit similar effects . This mechanism could also work independently along several axes to create clusters in multi-dimensional phenotype spaces which could be seen as temporary ecological guilds [77] . Indeed , general metabolic theory suggests that body mass linked to many other ecological quantities by similar simple scaling relations [78] so if the clustering in the space of growth-rates transfer to body masses , as we argued above , it should also be reflected in patterns along still other phenotypic axes . It will be interesting to see whether such noise-induced abundance patterns can be directly observed in natural or laboratory-based experimental ecosystems , particularly microbial ones [79] .
The full model is defined by Equations 1–3 . To compute the final abundances for a given initial seeding , we first find the growth-time ( T ) given the available amount of resource , ( R ) . Since all species grow freely , the number of offspring ( not counting the original ancestor ) of a single individual of species k at a time t is exp ( μkt ) −1 , and each new offspring removes 1/Yk units of resources . Starting from mk individuals , the total amount of resources consumed by the population of species k on a given patch is thus mk ( exp ( μkt ) −1 ) /Yk . Hence , T is the solution to the equation . ( 4 ) This equation defines a growth time T for every initial configuration m , given a set of growth rates μ and efficiencies Y . Changing the value of R is equivalent to scaling all the Y-values by a common factor , so we set R = 1 for convenience ( this is the choice used in this paper ) . In that case , Y is simply the per-patch number of offspring produced by a single seeded individual in the absence of competitors . We assumed that the environmental disturbances arrive at intervals longer than the time needed for even the slowest species to grow to saturation , i . e . the time between disturbances is longer than the largest T-value . Hence , the resources will always be completely exhausted on every patch and the time it took for this to happen ( which varies depending on the seeding of the given patch ) plays no further role . The final abundances for a given seeding averaged over all patches with this seeding , f ( m ) , are now given by Equation 2 . Using the average is consistent since we work with an infinite population; however , for a finite population , the stochastic growth differences between individual patches starting with the same seeding could change the results . With the exception of the rather trivial case N = 1 , we cannot analytically solve Equation 4 , so we used numerical solutions for the simulations . Similarly , for N>1 we cannot analytically do the sum in Equation 3 since it depends on quantities than can only be found numerically . We therefore approximated it by summing over a finite number of seedings , imposing the condition that the combined probability of all neglected configurations was less than 10−7 ( evaluated at the fixpoint ) . The resulting finite sum was over all seedings that involved at most M seeded individuals in total , where M was picked to satisfy the probability-condition . All simulations program were written in MATLAB and run on the Harvard Medical School supercomputing cluster ( Orchestra ) . Because our model involves the solution of the transcendental Equation 4 , a rigorous general proof of coexistence is difficult to provide . However , we can get close by drawing on similarities with the patch model of Tilman [45] , in which simplified competitive dynamics makes it possible to prove that an arbitrarily large number of species can coexist . Consider making the amount of resources on each patch very large or , equivalently , rescaling all efficiencies by a common large factor , s≫1: In this limit , the growth-time T clearly also goes to infinity . Expanding Equation 4 , we see that on a given patch , T becomes dominated by the contribution from the highest-μ species present , with corrections due to other species falling off exponentially in T . Neglecting all but the fastest species and choosing an α such that for all seedings that contribute significantly , we thus arrive at a ‘complete dominance approximation’ ( for R = 1 ) : Plugging this into the dynamical equation , we can now do the sum and get a set of explicit fixpoint equations ( we order the species so that μ1<μ2<…<μN ) : As in Tilman's model [45] , the fixpoint abundance of a species now depends only on its own parameters and those of the species that are stronger competitors ( have a higher μ ) . We can thus solve this hierarchy of equations for the efficiencies by working from the top and plugging the solution of each equation into those below . This allows us to find arbitrarily large sets of coexisting species . We introduce environmental noise by drawing the dilution factor α from a log-normal distribution with probability density ( 5 ) where θ and ω are the mean and standard deviation of the logarithm of α , respectively . This gives a smooth , peaked distribution of tunable width that automatically implements the constraint that α>0 . We made this choice since the long-term dilution rate – the expectation value of the product of many consecutive αs – is set by the expectation value of log ( α ) ( cf . [80] ) which we can control directly through θ . Had we instead kept the expectation value of α itself constant , we would have introduced changes in the expectation value of log ( α ) when changing the noise strength and hence biased the competition towards species that are either very efficient or very fast . To avoid this trivial bias , we kept θ constant as we increased the noise intensity ( ω ) in all simulations . Comparison with the linearized model ( see below ) shows that the exact choice of distribution for α is unimportant for the crucial features of the model . To test the stability of a fixpoint n* , we write n ( t ) = n*+Δn ( t ) and α ( t ) = α0+Δα ( t ) , and expand the dynamic equation ( Equation 3 ) in powers of Δn and Δα ( α0 is the dilution factor at the fixpoint ) . In the limit of low noise ( Δα/α0→0 ) , the fluctuations will be small and we need only keep the leading terms . We thus arrive at the linear approximation: ( 6 ) where the matrix J and the vector r have elements ( 7 ) ( all derivatives evaluated at the fixpoint ) . The formulas for the derivatives can be derived from Equation 3 , but must again be evaluated numerically for N>1 . The fixpoint is stable if all eigenvalues λk of the matrix J satisfy |λk|<1 ( complex modulus less than unity ) ; we explicitly checked that this conditions was fulfilled this for the parameter sets used in the article . The dependence of the elements of J on the phenotypic distance between species is illustrated in panel ( A ) of Supplementary Figure S7 . We now introduce white , Gaussian noise defined by ( 8 ) where <…> indicate averages over the noise distribution . These are the only properties of the noise we will make use of , so the exact noise distribution will not play a role . We split the system into N independent eigenmodes by diagonalizing J: ( 9 ) where λk is the kth eigenvalue of J ( all real and positive for the parameters used ) , p = S−1r , and q = S−1Δn ( the matrix S is the diagonalizing matrix whose columns are the N distinct eigenvectors of J ) . Using the noise properties ( 8 ) and the fact that |λk|<1 ( stable system ) , the average squared amplitude as t→∞ is given by ( 10 ) If we set pk = 0 ( no noise ) , we find ( 11 ) where the relaxation time τk is given by ( 12 ) In our system , all the eigenvalues are real and close to 1 , and can hence be written as λk = 1−εk with 0<εk≪1 . Hence , we find Therefore , we can write the equilibrium squared amplitude as a product of the coupling to the noise ( γk ) and the relaxation time ( τk ) : ( 13 ) The values of γk , ( 1−λk ) −1≈τk , and are plotted in Figure 4B ( squares , triangles and red dots , respectively ) – to facilitate visualization , the first two quantities have been rescaled so that their maximum value is 1 . Notice that the noise strength σα2 appears as an overall factor and hence does not affect the shape of the amplitude spectrum . The squared mode amplitudes for a simulated time-series of abundances , n ( t ) , can be found simply by normalizing to the fixpoint and transforming into the eigenbasis: ( 14 ) The average is performed over the simulated cycles . Comparisons of simulated data and the exact linear results from different number of species are shown in Supplementary Figure S9 . To each eigenvalue λk , there corresponds an eigenvectors v ( k ) of J , the elements of which specifies a deformations of the abundances away from the fixpoint . For these deformations , the influence of each species is balanced so that they all return to the fixpoint at the same rate . Since the fast species are superior competitors , the components in each v ( k ) corresponding to fast species must therefore be correspondingly smaller . To make the oscillations in the profiles more visible , we have therefore plotted a weighted version of the profiles in Figure 4B . In the weighted eigenvectors , each component is multiplied by the average interaction the corresponding species has with other species , compensating for the trivial decrease in component values with competitive ability . The interaction between species in the linearized model is given by the matrix ΔJ = J−I , where I is the unit matrix . The weighted eigenvectors thus have elements ( 16 ) where I is the unit matrix . The three plots below the main panel in ure 4B are plots of the components of the weighted vectors for k = 1 , 8 , and 15 . For comparison , both the weighted and unweighted forms of these three vectors are plotted in Supplementary Figure S11 . | In natural ecosystems , hundreds of species with different characteristics typically live side by side , some competing for the same foods and some preying on others . A central question in ecology is how the abundance of a given species in such an ecosystem depends on its particular characteristics ( its phenotype ) . Clearly , fixed environments can favor certain phenotypes ( thick fur in a cold climate ) , but what happens when environmental conditions fluctuate randomly as e . g . the weather does ? We investigated this question using a simple mathematical model of an ecosystem with many competing species . We found that , paradoxically , randomness in the environment can lead to the appearance of ordered clusters of abundant species with similar phenotypes , with the species adopting intermediate phenotypes being much less abundant ( a mountains-and-valleys pattern ) . The clusters move around so that different phenotypes are favored at different times . We found that these effects arise from the tension between the tendency of noise to level out difference in abundances and the tendency of competition to create larger abundance differences . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"ecology",
"ecosystems",
"population",
"modeling",
"population",
"dynamics",
"theoretical",
"ecology",
"biology",
"population",
"ecology",
"population",
"biology",
"microbial",
"ecology"
] | 2011 | Dynamic Phenotypic Clustering in Noisy Ecosystems |
HIV-1 does not persistently infect macaques due in part to restriction by several macaque host factors . This has been partially circumvented by generating chimeric SIV/HIV-1 viruses ( SHIVs ) that encode SIV antagonist of known restriction factors . However , most SHIVs replicate poorly in macaques unless they are further adapted in culture and/or macaques ( adapted SHIVs ) . Therefore , development of SHIVs encoding HIV-1 sequences derived directly from infected humans without adaptation ( unadapted SHIVs ) has been challenging . In contrast to the adapted SHIVs , the unadapted SHIVs have lower replication kinetics in macaque lymphocytes and are sensitive to type-1 interferon ( IFN ) . The HIV-1 Envelope ( Env ) in the chimeric virus determines both the reduced replication and the IFN-sensitivity differences . There is limited information on macaque restriction factors that specifically limit replication of the more biologically relevant , unadapted SHIV variants . In order to identify the IFN-induced host factor ( s ) that could contribute to the inhibition of SHIVs in macaque lymphocytes , we measured IFN-induced gene expression in immortalized pig-tailed macaque ( Ptm ) lymphocytes using RNA-Seq . We found 147 genes that were significantly upregulated upon IFN treatment in Ptm lymphocytes and 31/147 were identified as genes that encode transmembrane helices and thus are likely present in membranes where interaction with viral Env is plausible . Within this group of upregulated genes with putative membrane-localized proteins , we identified several interferon-induced transmembrane protein ( IFITM ) genes , including several previously uncharacterized Ptm IFITM3-related genes . An evolutionary genomic analysis of these genes suggests the genes are IFITM3 duplications not found in humans that are both within the IFITM locus and also dispersed elsewhere in the Ptm genome . We observed that Ptm IFITMs are generally packaged at higher levels in unadapted SHIVs when compared to adapted SHIVs . CRISPR/Cas9-mediated knockout of Ptm IFITMs showed that depletion of IFITMs partially rescues the IFN sensitivity of unadapted SHIV . Moreover , we found that the depletion of IFITMs also increased replication of unadapted SHIV in the absence of IFN treatment , suggesting that Ptm IFITMs are likely important host factors that limit replication of unadapted SHIVs . In conclusion , this study shows that Ptm IFITMs selectively restrict replication of unadapted SHIVs . These findings suggest that restriction factors including IFITMs vary in their potency against different SHIV variants and may play a role in selecting for viruses that adapt to species-specific restriction factors .
The macaque models of HIV-1 infection have been critical to the understanding of retroviral pathogenesis as well as for testing antiretroviral therapies and candidate vaccines for HIV-1 . However , multiple species-specific host factors restrict HIV-1 replication in macaque cells [1] . To overcome these restrictions chimeric SIV/HIV-1 viruses ( SHIVs ) , which encode the SIV antagonists of known restriction factors are used to infect macaques to model HIV-1 infection . Existing SHIV/macaque models typically employ SHIVs that encode HIV-1 sequences from viruses that were adapted by viral passage in cell culture , and often these viruses are from chronic stages of infection . However , there is evidence that the chronic stage HIV-1 variants are distinct from HIV-1 variants that are selected for transmission in humans [2 , 3] . In addition , SHIVs encoding HIV-1 sequences derived directly from humans typically require further adaptation in vitro in macaque cells and/or in vivo by serial macaque-passage [1] in order to obtain pathogenic viruses that establish persistent infection in macaques . These variant chimeric viruses used in the SHIV/macaque models are thus “adapted” SHIVs . We have previously determined that most circulating , transmitted HIV-1 Envelope ( Env ) variants , including the transmitted/founder variants , do not use macaque CD4 entry receptor efficiently [4] and thus SHIVs generated using these Envs replicate poorly in macaque cells . The adaptation of SHIV Env sequences in macaques increases replication and pathogenicity of SHIVs [5–11] but also leads to antigenic changes in Env that can limit their utility for vaccine and therapeutic approaches [12] . SHIVs encoding circulating HIV-1 variants obtained directly from the newly infected patients without adaptation ( termed unadapted SHIVs ) that maintain the antigenic properties of the transmitted variants are desired as challenge viruses for vaccine and therapeutic studies . However , most attempts at generating these SHIVs have failed as unadapted SHIVs replicate poorly , if at all , in macaque cells and do not establish persistent infection [13] . The virus-host dynamics that contribute to the differences in replication of adapted and unadapted SHIVs in macaques are not well defined . One of the main host determinants that exerts immune pressure on viruses in vivo is the type-1 interferon ( IFN ) response . Upon infection , detection of viral pathogen-associated molecular patterns ( PAMPs ) by host pattern recognition receptors ( PRRs ) activates a signaling cascade that results in production of IFNα/β and other inflammatory cytokines [14] . IFNα/β in turn activates autocrine or paracrine IFN-signaling pathways that results in expression of hundreds of IFN-stimulated genes ( ISGs ) –thus inducing an “antiviral state” . Proteins encoded by certain ISGs such as APOBEC3s , TRIM5 , Tetherin/BST-2 , and MxB/Mx2 potently block lentiviral replication [15 , 16] and are referred to as “restriction factors” . Lentiviruses have in turn evolved evasion/escape mechanisms and encode viral antagonists to counteract the restriction factors . These antagonistic virus-host interactions have resulted in an evolutionary “arms race” that drives continuous rounds of selection for advantageous mutations in the restriction factor genes [17] . Due to viral-host coevolution and because viruses can evolve faster than their hosts , lentiviral antagonists in turn adapt to the restriction factors encoded by their natural hosts . Thus , restriction factors are less active against wild-type viruses replicating in their natural host but act as potent barriers against cross-species transmission [15 , 16] . HIV-1 infection in humans and SIV infection in macaques induces a robust IFN response during acute infection in vivo [16 , 18–20] . Despite the induction of an IFN response , HIV-1 and SIV replication persists in their respective natural hosts . Recent studies suggest that transmitted strains of HIV-1 are more IFN-resistant than strains obtained from chronic stages of infection [21 , 22] , although this has not been seen in all studies [23 , 24] . However , in the case of unadapted SHIV infection in macaques , the selection pressure may be distinct as the HIV-1 portion of the genome has been adapted in humans and SHIVs therefore may be under a different selection pressure that is more similar to a cross-species transmission event . In some cases , adapted rapidly replicating , pathogenic SHIVs have been derived by rapid serial macaque-passage , performed within the first two weeks of infection , a time during which macaques mount a robust IFN response to infection [18 , 20] , potentially implicating escape from the IFN response as contributing to the adaptation process . Recently , we examined the sensitivity of a panel of adapted and unadapted SHIVs to IFNα in macaque lymphocytes and found that the unadapted SHIVs were potently inhibited by IFNα despite encoding the SIV antagonists of known restriction factors [25] . In contrast , SHIVs encoding adapted HIV-1 variants were largely resistant to IFNα . This difference mapped to HIV-1 Env , which was also a major determinant of replication differences , with adapted SHIVs demonstrating more rapid replication kinetics than unadapted SHIVs in macaque lymphocytes . These findings suggest that there are macaque-specific ISGs that restrict replication of unadapted SHIVs , and the adaptation of Env potentially plays a role in evading or antagonizing the macaque IFN response . The goals of this study were to characterize the host IFN response in macaque lymphocytes and to identify IFN-induced host factor ( s ) that specifically inhibit replication of unadapted SHIVs . To this end , we measure the IFN-induced gene expression in immortalized pig-tailed macaque ( Ptm ) CD4+ lymphocytes and identify previously uncharacterized interferon-induced transmembrane protein ( IFITM ) genes as candidate host factors for the observed IFN inhibition of unadapted SHIVs . The IFITM family of proteins are small ( 125–135 amino acids ) transmembrane proteins with a type II transmembrane protein topology that are expressed at a basal level in multiple cell types [26 , 27] . Of the five IFITM genes encoded by humans , IFITM1 , IFITM2 , and IFITM3 are induced by IFN and display broad antiviral activity against a number of enveloped viruses [28 , 29] . Recent studies have shown that human IFITMs ( IFITM1 , 2 and 3 ) are incorporated into HIV-1 virions and impair viral fusion and cell-to-cell spread [30–33] . Here , we demonstrate that Ptm IFITMs are packaged at higher levels in unadapted SHIVs when compared to adapted SHIVs , that depletion of IFITMs partially rescues the IFN sensitivity of unadapted SHIVs , and that depletion of basal levels of IFITMs increases replication of unadapted SHIVs in the absence of IFN . These findings identify IFITMs as IFN-induced host factors that limit replication of SHIVs in macaque lymphocytes in an Env-dependent manner .
In order to characterize the IFN response in macaque lymphocytes and to identify IFNα-induced host factor ( s ) that might inhibit replication of unadapted SHIVs , we measured IFNα-induced gene expression in immortalized Ptm CD4+ lymphocytes . Triplicate cultures of Ptm lymphocytes were left untreated or treated with IFNα at a concentration similar to that observed in natural SIV and HIV-1 infection ( 1000 U/ml ) [18 , 20 , 34] . Twenty-four hours later , RNA was isolated , and RNA-seq libraries were prepared for sequencing . Comparable numbers of reads were obtained for each sample ( average ~21 . 5 million reads for untreated and ~21 . 2 million reads for IFNα-treated , respectively ) and similar percentages of reads were unambiguously mapped ( average ~65 . 3% for untreated and ~64 . 9% for IFNα-treated , respectively ) to the pig-tailed macaque genome ( M . nemestrina Mnem 1 . 0 ) . A total of 198 genes were found to be significantly differentially expressed ( |logFC| ≥ 0 . 585 & FDR 5% ) upon IFNα treatment ( Fig 1 ) . 147/198 genes were found to be significantly upregulated and 51/198 genes were found to be significantly downregulated upon IFNα treatment ( Fig 1 , S1 Table ) . In order to identify candidate IFNα-induced host factor ( s ) that inhibit replication of unadapted SHIVs , we prioritized the upregulated genes and applied the following criteria . Because the differences in IFNα-induced inhibition between SHIVs mapped to Env we prioritized upregulated genes with predicted subcellular localization at sites of viral entry or assembly , such as intracellular and plasma membranes where interaction with Env is plausible . For this , we determined the amino acid sequences of the open reading frames ( ORFs ) of the 147 significantly upregulated genes and used TMHMM v2 . 0 [35] to predict which of these encode transmembrane helices . We found that 31/147 genes were predicted to encode at least one transmembrane helix ( Fig 1 , S2 Table ) . Based on these criteria of differential expression upon IFNα treatment and predicted membrane localization , we selected two previously uncharacterized Ptm genes ( LOC105494124 and LOC105494127: both predicted to be IFITM3-like genes ) and IFITM1 ( XM_011762253 ) for subsequent analysis ( Fig 1 ) . Notably , IFITM1 and IFITM3 from other species have been reported as broad-spectrum , anti-viral factors that impair viral fusion [28 , 29] . Previous work to determine the number and location of IFITM sequences within animal genomes focused broadly on vertebrates [36] or primates [37] . As a result , variation in IFITM genes amongst closely related species remains largely undescribed . Since our RNA-seq analysis showed increased expression of multiple Ptm IFITM-related genes , including some uncharacterized sequences , we wanted to understand the location , sequence , and evolutionary relationship of previously described human IFITMs and the IFNα-upregulated Ptm IFITMs ( Fig 1 ) . We mapped the relative position of all IFITM genes within the canonical IFITM locus , focused on comparing the genome of three macaques ( pig-tailed , rhesus , and crab-eating ) with human . The macaque IFITM loci were mapped using the human IFITM1 , IFITM3 , and IFITM5 sequences as queries of each macaque genome assembly . IFITM1 and IFITM5 had a single clear hit in each genome as compared to the human locus , suggesting one-to-one orthology of these genes ( Fig 2A; top panel; purple and red ) . In contrast , IFITM3 mapped to three unique locations within the IFITM locus of each macaque genome . An analysis of the synteny of these sequences across macaques and humans shows that one of these sequences is syntenic with human IFITM3 ( Fig 2A; top panel; orange ) , while the other two sequences are not found in humans . Of these macaque IFITM3 sequences , two encode apparently intact IFITM3 genes ( Fig 2A; top panel; green and orange ) , while the third copy has been pseudogenized with mutations that introduce numerous stop codons ( Fig 2A; top panel; gray ) . Humans have a duplication of IFITM3 , named IFITM2 [36] , and a phylogenetic tree shows that all the intact and pseudogenized macaque IFITM3 genes group together with the human IFITM2 and IFITM3 sequences ( Fig 2B and S1 File ) . While IFITM2 has only been described in humans , chimpanzees , and gorillas [37] , we find that macaques also have a duplication of IFITM3 within the IFITM locus . This copy LOC105494124 , here called IFITM3A , retains an intron and an ORF of the same length ( ~402nt ) as the parental IFITM3 ( LOC105494127 ) . In addition to the IFITM3 sequences within the IFITM locus of the macaques , we found numerous shorter sequences ( 100-400bp ) outside of the IFITM locus with sequence identity to IFITM3 greater than 80% . These hits represent partial and full-length copies of IFITM genes . Two of these sequences were identified in the RNA-seq analysis , denoted as IFITM3L sequences on chromosomes 16 and 9 ( Fig 2A; bottom panel; brown , pink ) . These copies are both found in the intronic region of genes and have lost the intron present in IFITM sequences within the IFITM locus . Based upon their phylogenetic grouping with IFITM3 , their dispersion in the genome , and their lack of introns , these sequences are likely retrocopies of IFITM3 , often referred to as pseudogenes of IFITM3 . However , both of these retrocopies preserve a putative ORF similar in length to IFITM3 and we find these copies to be expressed . We compared the magnitude of IFNα-dependent induction of these IFITM-genes by mapping the reads from our RNA-seq data to the Ptm genomic loci . We found that the bulk of the RNA-seq read coverage for IFITM3A and IFITM1 was from uniquely mapping reads ( S1 Fig ) . For IFITM3 , there were a considerable number of ambiguous reads that did not map uniquely , particularly in the first exon ( S1 Fig ) . However , there was still enough coverage from the uniquely mapping reads to suggest that this locus was expressed at the RNA level . In addition , we measured the levels of IFITM protein expression upon IFNα treatment in Ptm lymphocytes . Out of five commercially available IFITM antibodies tested , we found two that react with Ptm IFITMs and selected the one with the highest reactivity for subsequent analysis . Importantly , the selected anti-human IFITM3 antibody used in this study does not distinguish between Ptm IFITM3 and Ptm IFITM3A ( S2 Fig ) ; therefore , the immunoblots using this antibody are indicated as “IFITM3/3A” . Our immunoblotting results indicate that IFITM3/3A is induced by IFNα in a dose dependent manner in Ptm lymphocytes ( Fig 2C ) . Lastly , we performed immunofluorescence imaging of the Ptm IFITMs in mouse embryonic fibroblasts ( MEFs ) lacking the IFITM locus ( IFITMdel ) [38] and HEK293T cells , which also lack baseline IFITM expression [39] . In both cell types , the IFITMs were observed in intracellular clusters primarily in the perinuclear region , and IFITM1 staining was also observed at the plasma membrane ( S3 Fig ) . The cellular localization of Ptm IFITMs is similar to what is reported for human IFITMs [30 , 40] . Incorporation of human IFITMs ( IFITM1 , 2 and 3 ) into HIV-1 virions impairs viral fusion and cell-to-cell spread [30–33] . In addition , human IFITMs affect HIV-1 Env processing and virion Env incorporation [41] . Thus , we explored whether Ptm IFITMs contribute to the sensitivity of SHIVs to IFNα treatment in macaque lymphocytes and whether they explain Env-dependent differences in IFNα sensitivity between adapted and unadapted SHIVs . We employed a prototypical , macaque-adapted SHIV ( SHIV AD8-EO ) and a prototypical , unadapted SHIV ( SHIV Q23AE ) . SHIV AD8-EO was derived by five serial-macaque passages followed by further adaption in macaque PBMCs [8] while SHIV Q23AE was generated from a HIV-1 Env from early in infection and then modified with a single amino acid substitution ( A204E ) to allow the Env to use macaque CD4 for entry [42] . In order to determine whether or not the Ptm IFITMs are differentially incorporated into unadapted versus adapted SHIV virions , Ptm lymphocytes were infected with SHIV AD8-EO and SHIV Q23AE . Virions were harvested nine days post-infection and an amount of virus equivalent to 10 ng of SIV p27 were immunoblotted . We found that Ptm IFITM3/3A was readily detected in the unadapted SHIV Q23AE virions but not in the adapted SHIV AD8-EO ( Fig 3A ) . When four times more virions ( 40 ng of SIV p27 ) were immunoblotted , IFITM3/3A could be detected , but the level of IFITM3/3A in the adapted SHIV AD8-EO virions was lower than the unadapted SHIV Q23AE ( Fig 3A ) . Virions were also purified by analytical sucrose density gradient fractionation to remove secreted , soluble cellular proteins and budding cellular microvesicles . Immunoblot analyses of the gradient fractions with IFITM3/3A and SIV p27 antibodies revealed co-fractionation of IFITM3/3A with SIV p27 suggesting that IFITM3/3A are incorporated in the SHIV virion ( Fig 3B ) . Consistent with the results in Fig 3A , we observed that IFITM3/3A is packaged at higher levels in the unadapted SHIV Q23AE virions in comparison to the adapted SHIV AD8-EO virions . Importantly , the basal and IFNα-induced levels of IFITM3/3A were similar in uninfected cells or cells infected with adapted SHIV A8-EO or the unadapted SHIV Q23AE ( Fig 3C ) . These findings suggest that infection with SHIV does not affect steady state or IFNα-induced IFITM levels in Ptm lymphocytes , but rather the Ptm IFITMs are differentially incorporated into an unadapted SHIV relative to an adapted SHIV . Given the differences in IFITM packaging between the two prototype SHIVs , we determined whether the differences observed in IFITM incorporation are specific to the two SHIVs used in this study or also characteristic of other adapted and unadapted SHIVs . We employed a panel of eight SHIVs that we previously used to examine IFNα sensitivity in Ptm lymphocytes [25] . The panel includes four SHIVs encoding HIV-1 Env variants adapted through passage in cell culture and/or macaques ( AD8-EO , AD8-OG , SF162P3 , and 1157ipd3N4 ) and four SHIVs encoding HIV-1 Env variants isolated directly from infected individuals without culture or macaque adaption , three of which represent variants from early in infection ( QF495AE , Q23AE , MG505GV , and BG505AE ) and all of which encode a single amino acid substitution ( A204E or G312V ) to allow macaque CD4 use . The four unadapted SHIVs in this panel were potently inhibited by IFNα ( IC50 range 2 to 164 U/ml ) whereas the four adapted SHIVs were resistant to IFNα ( IC50 range 2454 to >5000 U/ml ) [25] . We utilized these SHIVs to examine virion incorporation of IFITMs . When virions equivalent to 5 ng of SIV p27 were immunoblotted , Ptm IFTIM3/3A was readily detected in the unadapted SHIV virions but not in the adapted SHIVs ( Fig 4A ) . Thus , we observed that IFITM3/3A virion incorporation was linked to sensitivity to IFNα for this virus panel . We did detect some IFITM3/3A incorporation in adapted SHIV virions when 3 . 6- to 9 . 6-times more virions were loaded ( Fig 4B ) suggesting that IFITM3/3A is not totally excluded from these viruses . Taken together , our results suggest that Ptm IFITMs are packaged at higher levels in the unadapted SHIV virions in comparison to the adapted SHIV virions . In order to determine the contribution of Ptm IFITMs in restricting SHIV replication , we generated IFITM-knockout ( KO ) cell pools . We designed CRISPR guide RNA ( crRNA ) predicted to target Ptm IFITM1 , IFITM3A , and IFITM3 ( Table 1 ) . crRNA that did not target any macaque genes was used as a Non-Targeting Control ( NTC ) . Because of the high sequence similarity between the Ptm IFITMs , one of our crRNA targeted both IFITM1 and IFITM3/3A ( this double KO is referred to as “M1+M3” ) . We generated four independent batches of IFITM-KO cell pools to examine whether IFITMs affect IFNα sensitivity of unadapted SHIV . IFNα-induced IFITM protein expression was confirmed through immunoblotting of IFNα-treated KO cells . The levels of IFITM3/3A in the IFNα-treated , KO cells were 5 . 9- to 12 . 5-fold lower compared to the IFNα-treated , NTC cells ( Fig 5A; left panel; 8 . 3-fold for M1+M3 KO , 12 . 5-fold for IFITM3A KO , and 5 . 9-fold for IFITM3 KO ) , and comparable to control cells that were not treated with IFNα , suggesting a partial KO and/or presence of unedited cells in the KO pool . The levels of IFITM1 in the IFNα-treated , M1+M3 KO cells were lower compared to the IFNα-treated ( 100 fold ) or IFNα-untreated ( 17 fold ) control cells ( Fig 5A; right panel ) . The KO cells were infected with the unadapted SHIV Q23AE or adapted SHIV AD8-EO at a multiplicity of infection ( MOI ) of 0 . 02 to allow spreading viral infection . The infected cells were cultured in the presence of 1000 U/ml of IFNα over a nine-day time course . Viral replication was also measured in the IFNα-untreated NTC cells to determine baseline replication . In order to obtain a quantifiable measure of IFNα sensitivity , we measured the ratio of the area under the curve ( AUC ) of the replication curve in the IFNα-treated cells to the AUC of the replication curve in the untreated , NTC cells . Consistent with our prior results , the prototype unadapted SHIV Q23AE exhibited a pronounced IFNα-induced inhibition of viral replication ( 6 . 7-fold at day 9 post infection ) whereas the adapted SHIV AD8-EO was largely resistant to IFNα ( AUC ratio 0 . 86 vs 0 . 96 , p = 0 . 0091 two-tailed student’s t-test ) ( Fig 5B and 5C ) . Next , we compared the IFNα sensitivity of adapted and unadapted SHIV in IFITM-KO cells . We observed that in comparison to the NTC cells ( AUC ratio 0 . 86 ) , KO of M1+M3 ( AUC ratio 0 . 93 , p = 0 . 0054 ) or IFITM3 ( AUC ratio 0 . 92 , p = 0 . 0156 ) , but not IFITM3A ( AUC ratio 0 . 85 ) , resulted in modest but statistically significant rescue of IFNα-induced inhibition of SHIV Q23AE replication ( Fig 5B and 5D ) . In contrast , IFITM KOs had no significant effect on IFNα sensitivity of adapted SHIV AD8-EO ( Fig 5C and 5E ) . Thus , the differential activity of IFITMs on the unadapted SHIVs accounts for some , but not all , of the IFNα sensitivity of these viruses . We have previously observed that the IFNα sensitivity of the SHIVs positively correlates with the replication capacity of the virus [25] . Thus , we hypothesized that basal levels of Ptm IFITMs have the potential to limit SHIV replication . We employed the IFITM-KO cell pools described above and assessed the ability of unadapted SHIV Q23AE and adapted SHIV AD8-EO to replicate in the absence of IFNα over a nine-day time course . Reduction of IFITM protein expression was confirmed through immunoblotting of KO cells . The levels of IFITM3/3A in the KO cells were lower than or comparable to the NTC cells ( Fig 6A ) . Since the IFITM3/3A antibody does not distinguish between IFITM3 and IFITM3A , we are unable to rule out weather the observed IFITM levels in the IFITM3 KO cells is due to the unedited cells in the KO pool or cross-detection of IFITM3A . Consistent with our previous results , we observed that the unadapted SHIV Q23AE replicates slowly with peak virus levels of ~104 pg/ml of SIV p27 at six days post-infection ( Fig 6B ) . In contrast , the adapted SHIV AD8-EO replicates rapidly reaching peak virus levels of >106 pg/ml of SIV p27 by six days post-infection ( Fig 6C ) . In order to compare the replication kinetics of the adapted and unadapted SHIVs across the IFITM-KO cells , we determined AUC for the replication curves from Fig 6B and 6C . When compared to the NTC cells ( AUC 30 . 1 ± 0 . 6 ) , a modest but statistically significant increase in SHIV Q23AE replication was observed in the M1+M3 KO cells ( AUC 32 . 1 ± 0 . 5 , p = 0 . 002 ) and the IFITM3 KO cells ( 31 . 6 ± 0 . 7 , p = 0 . 011 ) , despite comparable IFITM3/3A protein levels , but not in the IFITM3A KO cells ( AUC 29 . 9 ± 0 . 6 , Fig 6D ) . In contrast , IFITM KOs had no effect on the replication capacity of adapted SHIV AD8EO ( Fig 6E ) . Thus , basal levels of IFITMs selectively limit replication of unadapted SHIV in macaque lymphocytes . We have observed that the unadapted SHIVs have 2 . 7- to 14 . 3-fold lower virion Env levels than adapted SHIVs and that virion Env content correlates with replication capacity and sensitivity to IFNα [25] . A recent study reported that the human IFITMs impair HIV-1 infectivity by decreasing Env processing in the host cell; increasing Env shedding; and reducing virion Env incorporation [41] . In order to determine the impact of Ptm IFITMs on Env processing , virion incorporation , and viral infectivity , we co-expressed SHIV AD8-EO or SHIV Q23AE and Ptm IFITM in HEK293T cells . The cell and virion lysate were immunoblotted and infectivity of cell-free virions were measured on TZM-bl reporter cell line . Consistent with other studies [37 , 40 , 43] , we did not observe any differential effects of Ptm IFITM expression on Env processing in the producer cells and Env incorporation into viral particles ( S4A and S4B Fig ) . In addition , Ptm IFITM expression did not affect infectivity of SHIV virions produced during 48 hours in these experiments ( S4C Fig ) . This is perhaps not surprising given that the effect of IFN on replication is only apparent several days after infection in a spreading infection . Consistent with our previous study [25] , we did observe that the unadapted SHIV Q23AE has lower virion Env levels than adapted SHIV AD8-EO and there are differences in the migration pattern of gp160/gp120 , likely attributable to the glycosylation levels of the Env . Thus , our findings that the Ptm IFITMs limit replication of unadapted SHIV Q23AE during spreading infection lend support to the previously reported observations that IFITM-mediated restriction is most potent when IFITMs are present in both the donor and the target cells [30 , 40] .
Here we identify a novel example of cross-species restriction in which macaque-specific restriction factors selectively restrict replication of SHIV based on circulating , transmitted HIV-1 Env variant . In this study , we employed transcriptional profiling to define the repertoire of ISGs induced by IFNα in Ptm CD4+ lymphocytes . On the basis of this profiling , IFITMs , which have been implicated as anti-viral factors for a number of different viruses , including lentiviruses , were identified as a candidate restriction factors in these cells . We documented IFITM gene duplications in the macaque genomes and evaluated the ability of previously uncharacterized Ptm IFITMs to restrict SHIVs . We found that a prototypical , macaque-adapted SHIV is resistant to IFITM-mediated restriction , whereas a prototypical , unadapted SHIV , which encodes a circulating HIV-1 Env variant , is inhibited by Ptm IFITMs . Further , we demonstrate that IFITM virion incorporation tracks with the IFNα sensitivity of the SHIVs . For example , IFITMs are packaged at higher levels in IFNα-sensitive SHIVs when compared to IFNα-resistant SHIVs . Overall , our results suggest that the increased replication that results from adaptation of SHIVs in macaques may in part reflect an adaptation to an IFITM-mediated restriction . The evolutionary “arms race” that results from antagonistic virus-host interactions has led to many evolutionary innovations that accelerate host adaptation to a virus [17] . The duplication of restriction factor genes is one such strategy through which the host is able to explore multiple evolutionary trajectories to select for advantageous mutations in restriction factor genes . For example , increase in the copy number of the restriction factor could allow the host to rapidly evolve against a number of different viruses and/or collectively target a given virus through different mechanisms to constrain viral evolution . Thus , it is not surprising that many restriction factor gene families , such as APOBEC3 [44 , 45] , Mx1 [46] , and TRIM5 [47 , 48] , have undergone gene duplications . In our RNA-seq dataset , we found sequence reads that uniquely map to multiple locations in the IFITM gene locus in the Ptm genome that are not found in the human genome , suggesting IFITM gene duplications . The duplication of IFITM genes has been previously reported for vertebrates [36] and a recent study reported recurrent duplication and divergence of IFITM3 in the primate genomes [37] . An in-depth analysis of IFITM gene duplications specific to the macaque genomes was lacking , though several of the macaque IFITM3 copies we describe were included in previous analyses ( S5 Fig ) . Here , we demonstrate that macaques have three copies of IFITM3 genes that map to three separate loci within the IFITM locus . We found that a duplicated IFITM3 gene ( LOC105494124 , here called IFITM3A ) encodes an ORF of the same length as the parental IFITM3 gene ( LOC105494127 ) that differs by 10 . 5% at the amino acid level . Moreover , IFITM3A is distinct from the human IFITM3 duplication , named IFITM2 , which is found in chimpanzees and gorillas in addition to humans [37] . Interestingly , we find that all three macaque species ( pig-tailed , rhesus , and crab-eating ) contain an IFITM3 pseudogene within the IFITM locus , which is absent in humans . These pseudogene sequences group with high confidence in a clade that is separate from both the IFITM1 and intact macaque IFITM3s ( Fig 2B ) . Phylogenetically , we are unable to resolve whether these pseudogenes were born before or after the hominoid/Old World monkey split . Though , their presence in all three macaques and absence in humans suggests they were likely born in at least the last common ancestor of these macaques , but likely after the branching of Old World monkeys and hominoids . These pseudogene sequences are preceded by a long branch that presumably reflects a period of rapid but neutral evolution after birth or pseudogenization of this sequence . Together , these species restricted duplications of IFITM3 ( IFITM2 in humans , chimps , gorillas , and IFITM3A , IFITM3 pseudogene , and IFITM3Ls in macaques ) suggest this gene may be recurrently duplicated during primate evolution . In addition to genes within the IFITM locus , our RNA-seq analysis identified expression of two IFITM3-like retrogenes ( here called IFITM3L ) , which are present in the macaque genome on chromosome 16 and 9 . These IFITM3L copies are both found in the intronic region of genes and have lost the intron present in IFITM sequences within the IFITM locus . These putative IFITM retrogenes await a described function . In this study , we demonstrated that Ptm IFITMs are differentially incorporated in SHIV virions with higher levels in unadapted SHIVs in comparison to adapted SHIVs . It has been suggested that human IFITMs directly interact with HIV-1 Env [41] , and their restriction activity maps to Env- and Gag-dependent virion incorporation [30] . Thus , it is plausible that the Envs encoded by unadapted SHIVs display greater colocalization and/or interaction with Ptm IFITMs . In contrast , the Envs encoded by adapted SHIVs evade and/or antagonize this interaction . Moreover , our data rule out the alternative possibility that the lower IFITM incorporation in the adapted SHIVs is due to reduced induction and/or degradation of IFITMs as similar steady-state and IFNα-induced levels of IFITM3 was observed in cells infected with adapted or unadapted SHIVs . By generating IFITM-KO Ptm lymphocytes and infecting them with a prototype , macaque-adapted SHIV or a prototype , unadapted SHIV encoding HIV-1 Env isolated directly from an infected individual , we demonstrated that IFITMs partly contribute to the IFNα-induced inhibition of unadapted SHIV . We found that IFITM1 and parental IFITM3 , but not the duplicated IFITM3A , determines IFNα sensitivity of unadapted SHIV . Because of the high sequence similarity between the Ptm IFITMs , the CRISPR guide RNA designed to target IFITM1 also targeted IFITM3/3A . However , we could not discern the identity of IFITM3s ( IFITM3 and/or IFITM3A ) that were targeted in these double “M1+M3” KO cells , as the only available anti-IFITM3 antibody does not distinguish between the two Ptm IFITM3s . Thus , due to the targeting of more than one IFITM by CRISPR guide RNA and because of the cross-reactivity of IFITM antibody , we cannot determine which specific IFITM contributes to anti-viral activity or whether multiple IFITMs are playing a role in inhibiting unadapted SHIV . Because there is extensive duplication of IFITMs in the macaque genome , including presence of IFITM3L genes on different chromosomes , we did not attempt combinatorial knockouts so as to avoid spurious effects due to gene-editing and recombination across multiple chromosomes . Our findings that IFITM KOs only result in partial recovery of viral replication of unadapted SHIVs to levels observed in the IFNα-untreated Ptm lymphocytes , suggest that either there is functional redundancy between Ptm IFITMs or ISGs other than IFITMs also contribute to IFNα-induced inhibition of unadapted SHIVs . In a previous study , we found a significant positive correlation between viral replication kinetics and the sensitivity of SHIVs to IFNα in Ptm lymphocytes [25] . Here , we demonstrate that in contrast to adapted SHIV , depletion of basal levels of Ptm IFITMs increases the replication of unadapted SHIV in the absence of IFNα treatment . This suggests that the constitutively expressed IFITMs could contribute to the observed differences in the replication fitness of adapted vs unadapted SHIV , which could in turn affect their sensitivity to IFNα . Consistent with other reports [37 , 40 , 43]; we did not observe any differential effects of Ptm IFITM on Env processing and incorporation in SHIV virions . In contrast to the adapted SHIV , we did observe lower virion Env levels in the unadapted SHIV . Thus , one possibility is that lower virion Env levels in combination with IFITM virion packaging , decreases the replication kinetics of unadapted SHIVs rendering them sensitive to IFNα . In contrast , higher Env content and lower IFITM packaging in adapted SHIVs promotes higher replication , resulting in saturation of other IFNα-induced restriction factors . An alternative possibility is that high Env expression in cells infected with adapted SHIVs leads to increased cell-to-cell viral transmission thereby evading IFITM-mediated restriction . In support of this hypothesis , two recent studies demonstrated that in contrast to cell-free HIV-1 infection , cell-to-cell HIV-1 transmission renders it less sensitive to IFITM-mediated restriction [30 , 49] . Thus , differences in the sensitivity of unadapted vs adapted SHIVs to Ptm IFITMs could also be explained by cell-to-cell transmission , raising the interesting possibility that the increased replication fitness of SHIVs adapted in macaques is a reflection of better cell-to-cell virus spread . Adaptation of SHIVs to macaques typically involves serial macaque-passage to increase the replication capacity , transmissibility and pathogenicity [5–11] . As macaque IFNα-induced restriction factors can antagonize HIV-1 gene products encoded by SHIVs , successful adaptation of SHIVs in macaques likely involves overcoming these restrictions . This can potentially explain why adapted SHIVs cause persistent infection in macaques and the more biologically relevant unadapted SHIVs do not . If IFITMs constitute a selective force in vivo to control virus replication , then we hypothesize that the serial macaque-passage increases the replication capacity of SHIVs by overcoming IFITM-mediated restriction . Interestingly , such passaged viruses do become more resistant to IFNα inhibition with passage [25] . IFITMs may also play a role in restricting replication of simian tropic HIV-1 ( stHIV-1 ) in macaque lymphocytes . In support of this hypothesis , two studies have suggested the presence of unidentified , IFNα-inducible host factor ( s ) that target stHIV-1 at an early stage of viral life cycle in Ptm lymphocytes [50 , 51] . Moreover , these studies determined that the antiretroviral restriction factors—TRIM5 , APOBEC3 , BST-2/tetherin , and SAMHD1 are not responsible for this IFNα-induced restriction . Thus , it will be of interest to evaluate whether IFITMs restrict stHIV-1 in macaque lymphocytes . Lastly , it is well established that Envs from most transmitted HIV-1 strains demonstrate poor affinity for macaque CD4 [4] thus , SHIVs derived using such Envs replicate poorly , if at all , in macaque lymphocytes . Because it is difficult to study infection in unadapted SHIVs that lack mutations that have been identified to enhance macaque CD4 use [52 , 53] , we cannot draw conclusions on whether efficient engagement of macaque CD4 also affects IFITM-mediated restriction in macaques . Collectively , these studies may shed light on new approaches to further improve the SHIV/macaque models by rationally designing SHIVs to avoid key macaque restriction factors while maintaining as much as possible of the HIV-1 character of the virus .
HEK293T ( ATCC CRL-3216 ) , IFITMdel Mouse Embryonic Fibroblasts ( MEFs ) [38] , and HeLa TZM-bl cells ( NIH AIDS Reagent program catalog no . 8129 ) were cultured in Dulbecco's modified eagle medium ( DMEM , Gibco ) supplemented with 10% fetal bovine serum ( FBS , Gibco ) , 2 mM L-glutamine ( Gibco ) , and 1x Anti-anti ( anti-microbial/anti-mycotic , Gibco ) . Immortalized pig-tailed macaque ( Ptm ) CD4+ lymphocytes [54] were cultured in Iscove’s modified Dulbecco’s medium ( IMDM ) supplemented with 10% FBS , 2 mM L-glutamine , 1x Anti-anti , and 100 U/ml of interleukin-2 ( Roche ) ( complete IMDM ) . The following full-length proviral plasmids were used to generate viruses used in this study: SHIV AD8-EO , SHIV AD8-OG , SHIV 1157ipd3N4 , SHIV QF495AE , SHIV Q23AE , SHIV MG505GV and SHIV BG505AE ( S3 Table ) . Replication-competent SHIVs were generated by transfecting 2x106 HEK293T cells with 4 μg of proviral plasmid DNA and 12 μl of Fugene 6 transfection reagent ( Roche ) following manufacturer's protocol . Forty-eight hours post-transfection , virus-containing supernatant was harvested , passed through a 0 . 2 μm sterile filter and concentrated ~10-fold using Amicon Ultracel 100 kDa filters ( Millipore ) . Replication-competent stock of SHIV SF162P3 [6] was generated by expanding the virus in immortalized Ptm lymphocytes as described previously [25] . Aliquots of replication-competent SHIV stocks were stored at -80°C . The viral titer of each SHIV stock was determined by infecting TZM-bl cells and staining for β-galactosidase activity 48 hours post-infection [55] . For transient co-transfection experiments , HEK293T cells ( 2 . 5x106 cells/well in a 6-well plate ) were seeded 24 hours prior to transfection . Cells were co-transfected with plasmids encoding SHIV proviral DNA ( 1 μg ) and Ptm FLAG-IFITM variant or an empty vector control ( 0 . 5 μg ) using Fugene 6 transfection reagent ( Roche ) following manufacturer's protocol . Six replicate wells were transfected per sample . Forty-eight hours post-transfection , cells and virus-containing supernatant were harvested for subsequent analysis . RNA was extracted and purified from the Ptm lymphocytes using the RNeasy Mini kit ( Qiagen ) following manufacturer's protocol . Total RNA integrity was checked using an Agilent 2200 TapeStation ( Agilent Technologies ) and quantified using a Trinean DropSense96 spectrophotometer ( Caliper Life Sciences ) . RNA-seq libraries were prepared from total RNA using the TruSeq RNA Sample Prep Kit v2 ( Illumina ) and a Sciclone NGSx Workstation ( PerkinElmer ) . Library size distributions were validated using an Agilent 2200 TapeStation ( Agilent Technologies ) . Additional library QC , blending of pooled indexed libraries , and cluster optimization were performed using Life Technologies’ Invitrogen Qubit 2 . 0 Fluorometer ( Life Technologies-Invitrogen ) . RNA-seq libraries were pooled ( 6-plex ) onto a flow cell lane . Sequencing was performed using an Illumina HiSeq 2500 in rapid mode employing a paired-end , 50 base read length ( PE50 ) sequencing strategy . Image analysis and base calling were performed using Illumina's Real Time Analysis v1 . 18 software , followed by 'demultiplexing' of indexed reads and generation of FASTQ files , using Illumina's bcl2fastq Conversion Software v1 . 8 . 4 . Reads of low quality were filtered prior to alignment to Mnem 1 . 0 using TopHat v2 . 1 . 0 [56] . Counts were generated from TopHat alignments for each gene using the Python package HTSeq v0 . 6 . 1p1 [57] . Genes with low counts across all samples were removed , prior to identification of differentially expressed genes using the Bioconductor package edgeR v3 . 12 . 0 [58] . A false discovery rate ( FDR ) method was employed to correct for multiple testing [59] , with differential expression defined as |log2 ( ratio ) | ≥ 0 . 585 ( ± 1 . 5-fold ) with the FDR set to 5% . To predict which proteins may contain transmembrane helices , TMHMM v2 . 0 [35] was used with amino acid sequences from genes identified as being significantly differentially expressed . All RNA-sequencing FTP data files are available from the NCBI GEO database ( accession number GSE126594 ) . Human IFITM1 , IFITM2 , IFITM3 , IFITM5 genes were collected from NCBI ( NM_003641 , NM_006435 , NM_021034 , NM_001025295 ) . The IFITM locus for each macaque species was mapped using BLAT on UCSC genome browser ( for M . mulatta/BCM Mmul_8 . 0 . 1/rheMac8 and M . fascicularis/Macaca_fascicularis_5 . 0/macFas5 ) or BLASTN ( for M . nemestrina/Mnem_1 . 0 ) and extracting the aligned sequences . An alignment of IFITM nucleotide sequences from the three macaque species and humans was created using MAFFT with auto algorithm parameters [60] within Geneious version 11 . 1 . 4 [61] . The IFITM phylogenetic tree was created using PHYML with NNIs topology search , BioNJ initial tree , HKY85 nucleotide substitution model , and 100 bootstraps [62] . Synteny maps were built using pairwise dotplots ( in Geneious ) of the IFITM locus amongst species in addition to the pattern of sequence grouping in the phylogenetic tree . Whole cell extracts were prepared by lysing the cells in radioimmunoprecipitation assay ( RIPA ) cell lysis buffer ( 50 mM Tris pH 8 . 0 , 0 . 1% SDS , 1% Triton-X , 150 mM NaCl , 1% deoxycholic acid , 2 mM PMSF ) . For virion incorporation , virus containing supernatants from the infected or transfected cell cultures were centrifuged at 650 x g for five minutes at room temperature . Cell-free supernatant was filtered through 0 . 2 μm filter and then pelleted through a 25% sucrose cushion by ultracentrifugation for at 28 , 000 rpm for 90 minutes at 4°C . Virus pellets were lysed in 70 μl of RIPA buffer for 10 minutes at room temperature . The concentration of SIV p27 in the viral lysates was determined by SIV p27 antigen ELISA ( Advanced BioScience Laboratories ) , and normalized amounts of lysate were subjected to SDS-PAGE and immunoblotted . Standard Western blotting procedures were used with the following antibodies: SIV p27 ( ABL catalog no . 4323 ) , IFITM3 ( Proteintech catalog no . 11714-1-AP ) , IFITM1 ( Proteintech catalog no . 11727-3-AP ) , GAPDH ( BioRad catalog no . MCA4739P ) , FLAG ( OriGene catalog no . TA100023 ) , and HIV-1 gp120 ( NIH AIDS Reagent program catalog no . 288 ) . The IFITM3 antibody ( Proteintech catalog no . 11714-1-AP ) used in this study reacts with both Ptm IFITM3 and Ptm IFITM3A ( S2 Fig ) ; therefore , the immunoblots using this antibody are labeled as “IFITM3/3A” . Protein expression was quantified by measuring the band intensities using ImageJ software . Sucrose density gradient fractionation was performed as described previously [63] with some modifications . Briefly , 8x106 Ptm lymphocytes were infected at an MOI of 0 . 02 by spinoculation at 1200 x g for 90 minutes at room temperature . After spinoculation , cells were washed 4x with 1 ml of complete IMDM , re-suspended in 9 . 2 ml of complete IMDM and plated in two wells of a 6-well plate . Five hours after the initial infection , IFNα-2a ( PBL Assay Science ) was added to the culture at a final concentration of 1 , 000 U/ml . Every three days , two-third of the culture was replaced with fresh , complete IMDM containing IFNα-2a . On day 9 post-infection , the virus containing supernatant from infected cultures were collected , filtered through 0 . 2 μm filter and pelleted through a 25% sucrose cushion by ultracentrifugation at 28 , 000 rpm for 2 hours at 4°C . Pelleted virions were resuspended in 200 μl phosphate-buffered saline ( PBS ) , loaded on 20–70% ( w/v ) linear sucrose density gradients in PBS , and ultracentrifuged at 35 , 000 rpm for 16 hours at 4°C in SW41 rotor ( Beckman ) . Ten 1 ml fractions were collected from the top of the gradient , equilibrated in 10% trichloroacetic acid , and subjected to SDS-PAGE and immunoblotted with SIV p27 and IFITM3 antibodies . IFITM knockout cell pools were generated by electroporation of CRISPR/Cas9 ribonucleoproteins ( crRNPs ) as described previously [64] with some modification . Briefly , IFITM knockout cell pools were generated by electroporating Ptm lymphocytes with custom IFITM targeting crRNAs ( Table 1 ) . crRNA ( IDT ) and tracrRNA ( IDT catalog no . 1072534 ) were resuspended at 160 μM in 10 mM Tris pH 7 . 4 . 1 μl crRNA was complexed at an equimolar ratio with 1 μl tracrRNA and incubated for 30 minutes at 37°C followed by addition of 2 μl of 40 μM Cas9-NLS ( UC Berkeley MacroLab ) and further incubation at 37°C for 15 minutes to create the IFITM-targeting crRNP complexes . 3 . 5 μl crRNP was added to 5x105 Ptm lymphocytes resuspended in Amaxa SG Cell Line 96-well Nucleofector Kit ( Lonza catalog no V4SC-3096 ) and electroporated using a Lonza 4D Nucleofector according to the manufacturer’s protocol . After electroporation , 80 μl of prewarmed complete IMDM was added , followed by 30-minute recovery by incubating at 37°C . Eight replicate electroporations were carried out for each crRNA . Following recovery , cells from the eight replicate electroporations were pooled and resuspended at a density of 1 . 6x106 cells/mL in 2 . 5 ml complete IMDM in a 12-well plate . IFITM knockout was analyzed by Western blotting at 3 , 7 and 14 days post-electroporation . Replication of SHIVs was accessed as described previously [25] . Briefly , 1x106 Ptm lymphocytes were infected at an MOI of 0 . 02 by spinoculation at 1200 x g for 90 minutes at room temperature . After spinoculation , cells were washed 4x with 1 ml of complete IMDM , re-suspended in 1 . 2 ml of complete IMDM and plated in two wells of a 48-well plate . Five hours after the initial infection , IFNα-2a ( PBL Assay Science ) was added to one well at a final concentration of 1 , 000 U/ml . Every three days , two-third of the cultures were harvested and cell-free supernatant were separated by pelleting at 650 x g for five minutes at room temperature . Cultures were replenished with fresh , complete IMDM , including with IFNα-2a if appropriate . SIV p27 concentrations were determined using a SIV p27 antigen ELISA ( Advanced BioScience Laboratories ) . The data and statistical analyses were performed using Prism version 6 . 0c ( GraphPad Software ) . For imaging experiments , 105 HEK293T cells or IFITMdel MEFs were grown on sterilized glass coverslips in 12-well plates and transfected with 500 ng of plasmid encoding Ptm HA-IFITM variant or an empty vector control using Fugene 6 transfection reagent ( Roche ) following manufacturer's protocol . 24 hours post-transfection , cells were fixed with 4% paraformaldehyde in phosphate buffered saline ( PBS ) for 20 minutes , permeabilized with 0 . 1% Triton X100 in PBS for 10 minutes , and blocked with 2% FBS in PBS for 10 minutes . Cells were stained with the anti-HA primary antibody ( 1:1000 , HA . 11 Covance ) and Alexa Fluor 488-conjugated goat anti-mouse secondary antibody ( 1:1000 , Life Technologies ) . Coverslips were mounted with Prolong Gold Antifade reagent containing DAPI ( Life Technologies ) and images were taken using an Olympus Fluoview FV10i confocal microscope . | Macaque model systems are critical gatekeepers for testing HIV-1 prevention methods and for studies of HIV-1 transmission and pathogenesis . HIV-1 does not persistently infect macaques due to inhibition of the virus by several macaque-specific restriction factors necessitating the use of chimeric SIV/HIV-1 viruses ( SHIVs ) . Existing SHIV/macaque models typically employ SHIVs that encode HIV-1 sequences from viruses amplified in culture and further adapted in macaques ( adapted SHIVs ) . Development of SHIVs encoding circulating HIV-1 variants derived directly from infected humans ( unadapted SHIVs ) , which best model HIV-1 infection in humans , has been challenging as these SHIVs replicate poorly in macaque cells . While some host restrictions to HIV-1 replication in macaques have been defined , there is limited information on macaque-specific restriction factors that limit replication of circulating HIV-1 variants . Here , we demonstrate that the macaque interferon-induced transmembrane proteins ( IFITMs ) selectively restrict replication of unadapted SHIVs . These findings may help develop new approaches to improve the SHIV/macaque model of HIV-1 infection by rationally designing clinically-relevant SHIVs that overcome restriction by macaque-specific restriction factors . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"vertebrates",
"viral",
"structure",
"animals",
"mammals",
"retroviruses",
"primates",
"immunodeficiency"... | 2019 | Macaque interferon-induced transmembrane proteins limit replication of SHIV strains in an Envelope-dependent manner |
The cores of globular proteins are densely packed , resulting in complicated networks of structural interactions . These interactions in turn give rise to dynamic structural correlations over a wide range of time scales . Accurate analysis of these complex correlations is crucial for understanding biomolecular mechanisms and for relating structure to function . Here we report a highly accurate technique for inferring the major modes of structural correlation in macromolecules using likelihood-based statistical analysis of sets of structures . This method is generally applicable to any ensemble of related molecules , including families of nuclear magnetic resonance ( NMR ) models , different crystal forms of a protein , and structural alignments of homologous proteins , as well as molecular dynamics trajectories . Dominant modes of structural correlation are determined using principal components analysis ( PCA ) of the maximum likelihood estimate of the correlation matrix . The correlations we identify are inherently independent of the statistical uncertainty and dynamic heterogeneity associated with the structural coordinates . We additionally present an easily interpretable method ( “PCA plots” ) for displaying these positional correlations by color-coding them onto a macromolecular structure . Maximum likelihood PCA of structural superpositions , and the structural PCA plots that illustrate the results , will facilitate the accurate determination of dynamic structural correlations analyzed in diverse fields of structural biology .
Biological macromolecules , like proteins and catalytic RNAs , are dynamic structures . Each of the atoms in a macromolecule is coupled with other atoms via covalent bonds and various non-covalent interactions . This large and complex network of interconnections produces correlated structural dynamics , in which a perturbation or movement of one structural element covaries with the positional displacement of other elements . Thus , over a given time frame , macromolecules exist as an ensemble of correlated substates which span a large configurational space . Relevant time scales for dynamic structural change can range from picoseconds in molecular dynamics studies , to milliseconds for large structural movements , to millennia in evolutionary analyses of conformational perturbations due to amino acid substitutions . Understanding the correlated dynamics of such systems is essential for mapping structure to function . However , structural biologists currently have few tools for analyzing the correlations found in an ensemble of structures . Previous work characterizing the structural correlations in macromolecules has been limited to analysis of molecular dynamics ( MD ) simulations . Two general methods have been used to extract major modes of functionally relevant motions: normal mode analysis [1] and principal components analysis ( PCA ) of atomic covariance matrices [2 , 3] . Studies using these methods have largely shown that protein motions are dominated by only a few major distinct modes of correlated movement . Normal mode analysis assumes that dynamics are harmonic . In contrast , PCA does not make this assumption , and it has been found to be useful for finding major modes when the dynamics are highly anharmonic , which is more biologically realistic since proteins have multiple energetic minima [1] . In standard practice , PCA of an MD trajectory first involves removal of arbitrary rotational and translational effects by conventional least-squares superpositioning [4–8] . From this least-squares superposition one then calculates a covariance matrix , which is subsequently used as input for eigendecomposition in PCA ( also see [9] ) . However , the use of least squares is problematic in both theory and practice . As a statistical technique , least squares relies on two strong physical assumptions: that all atoms have the same variability , and that each atom is uncorrelated with the others . When these assumptions do not hold , least squares can give very misleading results [10] . In biomolecular applications , individual atoms in a superposition do not have equal variances , as some regions superposition closely while others show more conformational heterogeneity . Similarly , the atoms in macromolecular structures are strongly correlated by physical coupling via chemical bonds . Thus , both of the assumptions of least squares are violated in real biological data . In fact , performing PCA of a least-squares superposition is logically contradictory; the least-squares method assumes that no correlations exist , yet PCA is then performed on the least-squares derived covariance matrix to analyze those “nonexistent” correlations . We use a maximum likelihood ( ML ) method that overcomes the drawbacks of conventional least-squares superpositioning methods [11–13] . Unlike least squares , ML superpositioning is valid in the presence of heterogeneous variances and correlations , thereby providing more accurate superpositions [12 , 13] and corresponding covariance ( and correlation ) matrices . Rather than performing separate superpositioning and covariance matrix calculation steps , our ML superpositioning method simultaneously determines the optimal superposition and the optimal covariance matrix . We show that , as expected , PCA of our ML superposition provides markedly more accurate structural correlations than those extracted from least-squares superpositions . Furthermore , we show that use of the correlation matrix , rather than the covariance matrix , automatically corrects for biases that may be introduced due to experimental uncertainty in atomic positions or due to large differences in the magnitude of dynamic motion . We provide examples of the generality of the method by applying it to alternate crystal forms of the same protein , nuclear magnetic resonance ( NMR ) ensembles , and distant homologs with differing amino acid sequences .
We performed two simulation analyses to confirm the ability of our ML method to accurately determine the structural correlations found in sets of conformationally similar molecules . Two sets of conformationally perturbed protein structures were generated randomly by assuming a Gaussian distribution with known mean and known covariance matrices ( and , hence , based on known correlation matrices; see Figure 1A and 1E ) . In this case , the covariance matrix is a mathematical description of the positional variation and correlations among the atoms in an ensemble of molecular structures ( for more background regarding covariance and correlation matrices , see Methods ) . Two different covariance matrices were used: one with a range of variances , yet no correlations , and another with the same range of variances plus strong correlations ( the corresponding “true” correlation matrices are plotted in Figure 1A and 1E ) . The correlation structure and the range of variances are typical of NMR solution structures found in the PDB database ( see Methods ) . We then randomly translated and rotated each of the perturbed structures . Both least-squares and ML superpositions were performed independently on these two sets of simulated protein structures to obtain estimates of the true covariance/correlation matrix that was used to generate the structures ( Figure 1B–1D and 1F–1H ) . We found that , when calculated from an ML superposition , both the covariance matrix and the corresponding correlation matrix are considerably more accurate than those calculated from least-squares superpositions ( Figure 1 ) . When compared to the true ( known ) correlation matrix , the least-squares correlation matrix is highly biased , showing an artifactual pattern of correlation ( Figure 1B and 1F ) . As shown in Figure 1C and 1G , the least-squares correlation matrix remains artifactually biased even when the majority of highly variable atoms are excluded from the analysis , as often done in common practice ( “truncated least squares , ” where disordered regions are subjectively removed from the analysis with intent to obtain lower RMSDs ) . Interestingly , the least-squares procedure imparts a highly similar , artifactual correlation structure regardless of the true correlations ( compare Figure 1B and 1C , with no true correlations , to Figure 1F and 1G , in which the structures had true strong correlations ) . In contrast , the ML-based correlation matrix reliably recapitulates the true complex patterns of correlation ( Figure 1D and 1H ) . To extract major modes of structural correlation from a superposition , we use the statistical method of principal components analysis ( PCA; see Methods ) . PCA produces multiple principal components , each of which represents the predominant modes of structural correlation within the superposition . Generally , only the first few principal components ( that is , those with the largest eigenvalues ) are of practical interest , since they usually account for the majority of correlations in the data . As shown below , when significant covariation exists in a family of structures , PCA based on a least-squares superposition will yield erroneous principal components , resulting in artifactual modes of correlation . As with the correlation matrices , we found that the principal components determined from an ML superposition are likewise more accurate than principal components from a least-squares superposition ( Figure 2 ) . In these images , the largest ( or first ) principal component has been plotted in color on a single representative structure from the superposition . We refer to these types of graphs as “PCA plots . ” Red regions are correlated with each other , meaning that these regions tend to “move together” on average within the set of structures . Similarly , blue regions are also correlated with each other . However , the red regions are anti-correlated with the blue regions , meaning that red and blue regions tend to “move” in opposition to each other . White regions represent atoms whose positions are completely uncorrelated . In the first analysis , the PCA plots shown in Figure 2A–2D were calculated from simulated structures that had no bona fide correlations among their atoms ( using the correlation structure plotted in Figure 1A ) . Nevertheless , the largest principal components from the least-squares superpositions indicate a substantial , yet completely artifactual , mode of correlation , even when only the well-ordered residues are included in the superposition ( compare the true first principal component in Figure 2A with Figure 2B and 2C ) . In contrast , the first principal component from the ML superposition faithfully shows very little correlation , as indicated by the lack of colored patterns ( Figure 2D ) . PCA of the ML superposition also avoids the need for a subjective judgment on which residues to remove from the analysis . In the second , complementary analysis , protein structures were simulated which had strong correlations , using the correlation matrix plotted in Figure 1E . As before , the first principal component from the least-squares superposition indicates a large , artifactual mode of correlation , which is still present even when the highly variable residues are excluded ( Figure 2F and 2G ) . PCA of the ML superposition , however , accurately estimates the true correlation ( Figure 2H ) . Results from our ML method differ most from the conventional least-squares method when there is a wide range of variances among the atoms ( that is , when some regions of the structures are well-superpositioned and other regions are highly disordered ) and when correlations are strong . As the variances for the atoms become more uniform , and as the correlations approach zero , our method converges on the conventional least-squares method . Even so , the poor performance of the least-squares PCA method persists despite the removal of the majority of the most highly variable residues ( residues 1–5 at the N-terminus; see Figure 2C and 2G ) . Thus , with the improved accuracy of ML superpositions , PCA can be used reliably to find the major modes of positional variation and dynamical correlation within a family of structures . The method presented here for identifying major modes of structural correlation is general , and in principle it can be used to analyze any structural superposition , including independent solutions of the same protein , different homologous proteins , or a series of MD conformations . As one example , Figure 3 shows the second principal component from an ML superposition of a series of 10 crystal structures structures of the 70S ribosomal subunit from Haloarcula marismortui , including nine structures of the subunit bound to different antibiotics [14–16] . Remarkably , the majority of the correlation is localized to the active site of ribosome , the subunit interface , and the active site cleft , which binds the actively transcribed mRNA , tRNAs , translation factors , and the nascent polypeptide . The regions of strong correlated positional displacement also roughly correspond to regions of high RNA sequence conservation ( see , for example , Figure 5 of [14] ) . Thus , this PCA plot suggests that conformational perturbations of the ribosome during binding by various antibiotics are accompanied by correlated changes in distant yet functionally important regions . Our method can also be used to analyze the correlated conformational changes that have occurred during the evolution of protein homologs . The ML superposition and first principal component for a set of homologous telomere end-binding protein OB-fold domains are shown in Figure 4 . The PCA plot indicates a clear correlation between the two upper loops in blue and also within the red β-barrel , a fact that is otherwise difficult to ascertain from inspection of the structural alignment alone . The two blue loops are known to be critical for recognition of the proteins' single-stranded DNA ligand [17 , 18] . Thus , this PCA analysis implies that these loops ( and also the β-barrel ) have co-evolved in terms of conformation during the divergence of these domains from a common ancestor [19–21] . The correlations found in PCA plots are also useful for analyzing ensembles of solution structures of macromolecules solved by NMR spectroscopy . For instance , Figure 5A and 5C shows the largest principal mode of correlation from solution structures of ubiquitin solved by dynamic ensemble refinement , which takes into account the dynamic heterogeneity of a protein as measured by NMR relaxation experiments in addition to NOE distance constraint data [22] . Two independent NMR refinements of the ubiquitin structure are shown to give a sense of the reproducibility of our ML PCA method [22 , 23] . Two key residues in the core of the protein , Val5 and Ile30 , pack against each other and are highly anti-correlated , indicating that during the “fluid-like” dynamic motion of the protein's interior these residues move in opposition to each other . Val5 and Ile30 are both members of a small set of core residues that have been implicated in forming a folding nucleus in ubiquitin [24] . Furthermore , these residues are notable for being some of the most highly conserved among ubiquitin homologs [25] , for exhibiting the slowest rates of hydrogen exchange in the protein [26] , and for decreasing the thermodynamic stability of the protein when mutated [27] . Together with these experimental results , ML PCA suggests that strongly correlated residues in ubiquitin are important for its folding and stability . Our method is reminiscent of previous work that has used PCA of covariance matrices to extract major modes of functionally relevant motions from MD trajectories [2–8] . However , the interpretation of PCA of a covariance matrix is problematic , as that method results in modes of covariation that are a convolution of both the correlation and the variance of the atoms ( see Equation 3 in Methods ) . In structural superpositions , two very different factors contribute to the conformational variance: ( 1 ) random experimental uncertainties and ( 2 ) dynamic motion or conformational heterogeneity . Because we use the correlation matrix , rather than the covariance matrix , our method cleanly separates pure correlations from the variance , and thus the resulting principal components can be interpreted as bona fide modes of correlation . For instance , often the variances in a covariance matrix are composed of stochastic contributions that can be physically irrelevant or uninteresting . In NMR ensembles , the variance of each atom reflects not only the dynamics of that atom but also the number of experimental constraints for the position of that atom . Highly uncertain regions of a structure can therefore dominate the largest principal component from a covariance matrix , thereby artifactually inflating the importance of these imprecise regions . An example is shown in Figure 5B , where the disordered C-terminal tail of ubiquitin has a large variance largely due to experimental imprecision ( from a paucity of NOE distance constraints ) , resulting in its unilateral contribution to the largest principal component of the covariance matrix . PCA of a correlation matrix , on the other hand , circumvents this problem by down-weighting uncertain regions in proportion to their variances ( see Equation 2 and compare Figure 5A and 5C with Figure 5B and 5D ) . Furthermore , in an MD trajectory , a highly mobile loop with little correlated movement with other parts of the structure can nevertheless dominate the first mode of covariation . As a result , the largest principal components from the covariance matrix will primarily represent large magnitude motions with little or no real correlated movement . Covariance matrix PCA is useful , then , for analyzing major modes of motion when coordinate precision is high . However , covariance PCA is generally uninformative about true conformational correlation . In sum , correlation matrix PCA produces modes of pure correlation that are independent of the uncertainties in atomic positions , since the variance components have been normalized away ( Equation 2 ) . Our ML method thus provides correlations that are unlikely to be artifacts of experimental imprecision or of the magnitude of dynamic motions in localized regions of the structure . Our maximum likelihood method provides principal components that accurately describe the modes of coordinated motions and correlations found in an ensemble of structures . By using correlation matrices rather than covariance matrices , the modes of correlation that are found are largely free of artifacts that can result from experimental imprecision and the magnitude of dynamic motion . Taken together , various experimental results suggest that highly correlated residues from PCA plots are likely to be functionally significant . Thus , maximum likelihood PCA of structural superpositions , and the structural PCA plots that illustrate the results , should prove to be of wide utility in analyzing and comparing macromolecules in diverse fields of structural biology .
A covariance matrix is a mathematical description of the variation and covariation among members of a dataset . In the case of macromolecular structures , the covariance matrix describes the positional variation and correlations among the atoms observed in properly superpositioned family of structures . For example , given a protein K amino acids in length , here we consider the K × K covariance matrix representing the covariation of each of the K α-carbons with each of the others . If the orientations of the structures are known with certainty , then the diagonal elements σi , i of the covariance matrix Σ are simply the variances for each of the atoms . Each off-diagonal element σi , j≠i is the covariance of the ith atom with the jth atom . The elements σi , j of the covariance matrix Σ can be defined as: where denotes the arithmetic average of yi over all i , and the xi here are 3-vectors representing the 3-D x , y , and z coordinates of each atom . The correlation matrix C , on the other hand , is a simple function of the covariance matrix that has been normalized by the variances , leaving only pure correlations . Each element ci , j of the correlation matrix C is given by: Unlike a covariance matrix , the diagonal elements of a correlation matrix all equal 1 , and the non-diagonal elements range from −1 to 1 ( corresponding to perfect negative correlation and positive correlation , respectively ) . Clearly , the accuracy of both the covariance and correlation matrices directly depends on the accuracy of the superposition . Note that , if the covariance matrix is known , then the correlation matrix is also necessarily known . However , the transform is not symmetric , as the correlation matrix does not contain all the information needed to reconstruct the covariance matrix; the variances are also required: Major modes of structural correlation within a given structural dataset were found using the statistical method of principal components analysis ( PCA ) . To perform PCA , the correlation ( or covariance ) matrix is diagonalized by spectral decomposition . The resulting eigenvectors are ranked according to their corresponding eigenvalues , largest to smallest . The eigenvector with the largest eigenvalue corresponds to the first principal component , which summarizes the major mode of correlation ( or covariance ) in the data . The second principal component corresponds to the second largest mode of correlation , and so on . Unless otherwise indicated , all examples reported here used PCA of the correlation matrix , although our program THESEUS will also perform PCA on the covariance matrix if desired ( see Implementation ) . A statistical likelihood model for superpositioning structures . A detailed treatment of the following likelihood analysis can be found elsewhere [12 , 13] . We present here a simplified account of the ML method and its rationale , focusing on simultaneous estimation of the covariance matrix in the macromolecular structural superpositioning problem . In the following , we specifically consider the superpositioning problem per se , as opposed to the structural alignment problem . We assume that the one-to-one correspondence between atoms or residues ( i . e . , the alignment ) is known . Consider superpositioning N different structures ( Xi , i = 1…N ) , each with K corresponding atoms . Each structure is mathematically represented as a K × 3 matrix of K rows of atoms . We assume a statistical perturbation model in which each macromolecular structure Xi is drawn from a matrix normal ( Gaussian ) probability distribution [28 , 29] . Each structure Xi to be superpositioned is considered as an arbitrarily rotated and translated Gaussian perturbation of a mean structure M: where ti is a 1 × 3 translational row vector , 1K denotes the K × 1 column vector of ones , and Ri is an orthogonal 3 × 3 rotation matrix . The entries of the K × 3 matrix Ei are filled with normal random errors , each with mean zero , i . e . , Ei ∝ NK , 3 ( 0 , Σ , I3 ) . The K × K covariance matrix Σ describes the ( spherical ) variance of each atom and the covariances among the atoms . The likelihood equation for matrix Gaussian superpositioning . In the superposition problem with arbitrary translations , the covariance matrix Σ is poorly identified and singular unless it is parametrically constrained . Thus , to render the covariance matrix estimable , we assume that its eigenvalues are hierarchically distributed according to an inverse gamma probability density . An inverse gamma distribution is physically reasonable , as extremely small or large variances are relatively unlikely . The full joint log-likelihood for a structural superposition is then the sum of the log-likelihood for the eigenvalues of the atomic covariance matrix and the log-likelihood for the multivariate matrix normal density [30 , 31] corresponding to the statistical model given by Equation 4 . The full superposition log-likelihood is thus where |U| denotes the determinant of a matrix U , denotes a squared Frobenius Mahalanobis matrix norm , and α and γ are the scale and shape parameters , respectively , of an inverse gamma distribution for the K eigenvalues ( λj ) of the atomic covariance matrix Σ: ML superposition solutions . In the following , we briefly give the ML solutions for each of the unknown parameters of the superposition log-likelihood equation from above . Each observed structure must be translated to its row-weighted centroid: where t̂" i is the ML estimate of the translation: The optimal rotations are calculated using a singular value decomposition ( SVD ) . Let the SVD of an arbitrary matrix D be UΛV' . Then , the ML rotations R̂" i are estimated by The mean structure is estimated as the arithmetic average of the optimally translated and rotated structures: Finally , the ML estimate of the atomic covariance matrix is given by: where the unconstrained ML estimate of the covariance matrix is: Because the estimate of the covariance matrix Σ is a function of the other unknown parameters , the ML solutions given above must be solved simultaneously by numerical methods [12 , 13] . We use an iterative algorithm based on the Expectation-Maximization ( EM ) method [32 , 33] . The algorithm assumes that the alignment ( the one-to-one correspondence among atoms/residues in the structures ) is known a priori , and it aims to determine the ML superposition given that alignment . In brief: 1 . Initialize: Set = I . Randomly choose one of the observed structures for use as the mean structure M̂" . 2 . Translate: Translate ( i . e . , center ) each according to Equation 7 . 3 . Rotate: Calculate each rotation R̂" i ( Equation 8 ) , and rotate each translated structure by setting . 4 . Estimate the mean: Recalculate the average structure M̂" ( Equation 9 ) . Return to step 3 and loop until convergence . 5 . Estimate the inverse gamma distributed eigenvalues: Estimate ( Equation 11 ) and find its sample eigenvalues . Estimate the inverse gamma parameters by iteratively fitting them to the eigenvalues of the ML estimate of the covariance matrix , treating the zero eigenvalues ( or the smallest variance ) as missing data in an expectation-maximization algorithm . 6 . Estimate the atomic covariance matrix: Modify according to Equation 10 . Return to step 2 and loop until convergence . 7 . PCA: Perform a principal components analysis on the correlation matrix ( or corresponding covariance matrix ) . If all variances are assumed to be equal and all covariances are assumed to be zero ( i . e . , Σ ∝ I ) , then this algorithm corresponds to the classical least-squares algorithm for the simultaneous superpositioning of multiple structures [34–37] . The algorithm presented above ( like that of Theobald and Wuttke [13] ) is similar to that given in Theobald and Wuttke [12] , with three exceptions . First , the algorithm of [12] is much more general , e . g . , it is applicable to data in an arbitrary number of dimensions . Here we assume D = 3 for 3-D , spatial data . Second , here no scaling factors are necessary ( i . e . , βi = 1 for all structures ) , since molecules are inherently in the same scale , as bond lengths are fixed by the laws of physics . Third , we further assume that the variance about each atom is spherical ( i . e . , Ξ = I ) , an assumption that greatly simplifies the calculations . The algorithm described above for calculating ML superpositions and performing PCA of the estimated covariance matrix is implemented in the command-line UNIX program THESEUS [12 , 13] . THESEUS operates in two different modes: ( 1 ) a mode for superpositioning structures with identical sequences and ( 2 ) an “alignment mode , ” which superpositions homologous structures with different residues given a known alignment ( for instance , as determined from a sequence alignment program or from a structure-based alignment program ) . THESEUS does not perform structure-based sequence alignments , which is a distinct bioinformatic problem [38] . As with all superposition methods , THESEUS requires an a priori one-to-one mapping among the atoms/residues ( i . e . , it requires a known alignment ) . With NMR models or different crystal structures of identical proteins , the one-to-one mapping is trivial . When superpositioning different molecules with different sequences , however , a sequence alignment must be provided as a guide . THESEUS accepts sequence alignments in standard CLUSTAL and A2M ( FASTA ) formats . In addition to the ML superposition for a set of structures , THESEUS will calculate the principal components of either the covariance or correlation matrix . For input , THESEUS takes a set of standard PDB formatted structure coordinate files ( http://www . wwpdb . org/docs . html [39 , 40] ) . PCA analysis is requested with the “-Pn” command line option , where “n” is substituted with the number of principal components desired ( usually three are sufficient ) . PCA of the correlation matrix is performed by default; the “-C” option specifies that the covariance matrix should be used . Each principal component is written into the temperature factor field of two output files: ( 1 ) a PDB formatted file of the optimal ML superposition ( each structure is represented as a different MODEL ) and ( 2 ) a PDB formatted file of the estimate of the mean structure . Principal components can then be visualized as PCA plots ( described in Results/Discussion ) with any visualization software , such as PyMOL [41] , RasMol [42] , or MolScript [43] , that can color the structures by values in the temperature factor field . Two artificial datasets of protein coordinates were prepared as described previously [12] . Briefly , for each set , 300 protein structures were generated randomly , assuming a matrix Gaussian error distribution with a known mean protein structure and known atomic covariance matrix . The α-carbon atoms from model 1 of PDB:ID 2sdf ( the human cytokine stromal cell-derived factor-1 protein [44] ) were used as the mean protein structure ( 67 atoms/landmarks , squared radius of gyration = 152 Å2 ) . The 67 × 67 atomic covariance matrices were based on values calculated from the superposition given in 2sdf , with variances ranging from 0 . 0452 to 79 . 2 Å and correlations from 0 to 0 . 99 . Thus , in this simulation , the variances range over 3 . 2 orders of magnitude , a value that is typical for NMR solution structure ensembles ( of 3 , 150 single-domain NMR families in the PDB database , the average range for the variance is 2 . 9 ± 1 . 1 ( SD ) orders of magnitude ) . The first simulated set of structures used a diagonal covariance matrix in which all covariances were set to zero . The second simulated set of structures used the full covariance matrix . Hence , both sets were generated with the same variances , differing only in their correlation structure . After generating the perturbed protein structures , each was then randomly translated and rotated . Our ML superposition procedure was then performed on these simulated data sets , providing estimates of the atomic covariance matrix , along with estimates of the coordinates of the mean structure and of the original “true” superposition before translations and rotations had been applied . Default THESEUS parameters were used ( version 1 . 2 . 6 ) , except that the full covariance and correlation matrices were estimated with the “-c” command line option . For comparison , conventional least-squares superpositions were also calculated for the same dataset . The corresponding sample covariance and correlation matrices were calculated based on these least-squares superpositions . In order to show the effect of discarding a subset of highly variable ( “disordered” ) regions , separate least-squares analyses were performed using all atoms and also excluding residues 1–5 from the N-terminus , the atoms with the highest variance ( referred to as “truncated least squares” ) . Images of rendered macromolecules in Figures 2 , 4 , and 5 were made with POVScript+ [43 , 45] and Raster3D [46] . Figure 3 was made with PyMOL [41] . | Biological macromolecules comprise extensive networks of interconnected atoms . These complex coupled networks result in correlated structural dynamics , where atoms and residues move and evolve together as concerted conformational changes . The availability of a wealth of macromolecular structures necessitates the use of robust strategies for analyzing the correlated modes of motion found in molecular ensembles . Current strategies use a combination of least-squares superpositions and statistical analysis of the structural covariance matrix . However , the least-squares treatment implicitly requires that atoms are uncorrelated and that each atom has the same positional uncertainty , two assumptions which are violated in structural ensembles . For example , the atoms in the proteins are connected by chemical bonds , covalent and non-covalent , resulting in strong correlations . Furthermore , different atoms have different variances , because some atoms are known with less precision or have greater mobility . Using maximum likelihood ( ML ) analysis , we have developed a technique that is markedly more accurate than the classical least-squares approach by accounting for both correlations and heterogeneous variances . The improved ability to accurately analyze the major modes of dynamic structural correlations will benefit a diverse range of biological disciplines , including nuclear magnetic resonance ( NMR ) spectroscopy , crystallography , molecular dynamics , and molecular evolution . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"archaea",
"yeast",
"and",
"fungi",
"computational",
"biology",
"eukaryotes",
"paramecium",
"etc)",
"molecular",
"biology",
"ciliates",
"(tetrahymena"
] | 2008 | Accurate Structural Correlations from Maximum Likelihood Superpositions |
Dengue is associated with significant economic expenditure and it is estimated that the Asia Pacific region accounts for >50% of the global cost . Indonesia has one of the world’s highest dengue burdens; Aedes aegypti and Aedes albopictus are the primary and secondary vectors . In the absence of local data on disease cost , this study estimated the annual economic burden during 2015 of both hospitalized and ambulatory dengue cases in Indonesia . Total 2015 dengue costs were calculated using both prospective and retrospective methods using data from public and private hospitals and health centres in three provinces: Yogyakarta , Bali and Jakarta . Direct costs were extracted from billing systems and claims; a patient survey captured indirect and out-of-pocket costs at discharge and 2 weeks later . Adjustments across sites based on similar clinical practices and healthcare landscapes were performed to fill gaps in cost estimates . The national burden of dengue was extrapolated from provincial data using data from the three sites and applying an empirically-derived epidemiological expansion factor . Total direct and indirect costs per dengue case assessed at Yogyakarta , Bali and Jakarta were US$791 , US$1 , 241 and US$1 , 250 , respectively . Total 2015 economic burden of dengue in Indonesia was estimated at US$381 . 15 million which comprised US$355 . 2 million for hospitalized and US$26 . 2 million for ambulatory care cases . Dengue imposes a substantial economic burden for Indonesian public payers and society . Complemented with an appropriate weighting method and by accounting for local specificities and practices , these data may support national level public health decision making for prevention/control of dengue in public health priority lists .
Dengue is an arboviral infection transmitted between humans by Aedes mosquitoes . Globally , dengue is a major public health concern that has rapidly spread across the tropics and subtropics . [1 , 2] Between 1990 and 2013 the estimated number of global dengue cases doubled every decade , [3] and up to 3 . 9 billion people remain at risk in endemic countries . [4] Recent global modelling studies estimate between 55–100 million dengue cases occur annually; and estimate an increasing dengue mortality reaching over 38 , 000 deaths in 2016 . [3 , 5 , 6] Of the global population at risk , more than 70%–or about 1 . 8 billion people–live in the Asia-Pacific region and as such , Asians contribute the most to the overall burden of dengue . [1] In addition , the incidence of the severe forms of disease is higher in Asia-Pacific compared with other endemic regions perhaps for reasons of genetic susceptibility , but more likely because secondary infection is more common , due to the higher levels of endemicity and that all four dengue virus serotypes continually co-circulate . [7–10] In Indonesia , Ae . aegypti and Ae . albopictus are the primary and secondary vectors for transmission , respectively . The average number of annual dengue cases reported to health authorities in Indonesia was more than 129 , 000 for the period between 2004 and 2010 , the second highest incidence rate in the world after Brazil . [1] Reporting of dengue in Indonesia is acknowledged to be incomplete and reporting procedures vary widely among the provinces . [11] A 2013 cartographical modelling study estimated that approximately 7 . 6 million dengue infections may have occurred in in Indonesia in 2010 , the majority of which went unreported . [5] The disease typically is most common in urban areas , however , rural areas are increasingly affected . [7] Furthermore , the traditionally cyclical epidemic outbreaks of dengue appear to have become more erratic in recent decades . [9] The costs associated with dengue illness are substantial , in 2012 the WHO ranked dengue as the most important mosquito-borne viral disease across the globe , noting that outbreaks “exert a huge burden on populations , health systems and economies in most tropical countries of the world” . [1] Recognizing the substantial impacts in endemic regions , several economic burden studies have been conducted in various regions of the Americas , [12–18] and several countries in Asia and South Asia including Thailand , [19] Malaysia , [20 , 21] India , [22] Singapore , [23] Cambodia [24] and the Philippines . [25] These studies confirmed the considerable direct and indirect impact of dengue on individuals , families and communities . In Indonesia , some initial insights could be derived from the study by Shepard and colleagues , who estimated the annual economic burden of dengue in 12 countries of Southeast Asia at US$950 million; for Indonesia the annual cost over the period 2001–2010 was US$323 million . [26] A subsequent estimate based on revised global dengue incidence estimates and extrapolations of costs from scientific literature estimated the costs in Indonesia in 2016 to have been US$2 billion . [27] Another study by Stahl and colleagues estimated the cost of dengue outbreaks by conducting a literature review and case studies in four countries including Indonesia . [28] The estimated costs of an Indonesia dengue outbreak in 2011 were US$6 . 75 million ( adjusted to 2012 US$ ) . However , these studies did not collect local empirical cost data and instead relied on estimates derived from a literature review on unit costs for inpatient and outpatient care and used extrapolations of proportionality of costs from other nearby countries . [26–28] One study conducted in Surabaya , Indonesia in 2007 examined treatment costs at the hospital level and estimated inpatient costs per episode related to dengue were in the range of 1–2 million Indonesian Rupiah ( IDR ) or approximately US$106–212 . However , the scope of this study was limited to that single area and did not provide country-wide estimates for total healthcare costs . [29] We are not aware of a study which has collected comprehensive primary data on the economic burden of dengue in Indonesia . Such studies are needed to inform policy making , provide information to support healthcare resource allocation including prioritizing research and disease prevention and control measures , as well as promote public awareness . [30] Due to the country’s economic , geographic and sociological heterogeneity , the best way to represent national level burden and expenditure would be to use data from multiple sites and treatment facilities , taking a broad economic perspective . The aim of this study was to estimate the economic burden–including direct and indirect costs–associated with hospitalized and ambulatory dengue cases in Indonesia , first by determining costs at the facility level across three provinces , then extrapolating these using local epidemiological data to make the first , empirically-derived national economic burden of dengue estimates for Indonesia .
The ethics Committee of the Faculty of Public Health at Universitas Indonesia approved this study . Ethical approval for data collection at public hospitals and health centres was received from the local authorities ( Dinas Kesehatan or District Health Office ) . Interview participants or their parents/guardians signed informed consent ( signed assent forms were also required for those aged 8–18 years ) before study entry . In Indonesia , tertiary healthcare facilities are divided into type A facilities , which provide the full spectrum of specialist medical services and type B facilities , where specialist services are limited . Both types provide basic and supportive care to both in- and out-patients . Of the 34 provinces in Indonesia , three were selected to represent low- ( Yogyakarta ) , medium- ( Bali ) and high-income ( Jakarta ) ; from these three a total of nine facilities were selected for inclusion in the study . Public and private healthcare facilities were selected according to their research experience; and to provide a range of dengue and cost perspectives , including those treating inpatients and outpatients . Four facilities were selected in Jakarta: RSUPN Cipto Mangunkusumo ( public type A hospital ) , RSUD Pasar Rebo ( public type B hospital ) , RS Pelni ( private hospital ) and Tambora ( Puskesmas [a sub-district level public health centre] ) ; three facilities in Yogyakarta: RSUD Wirosaban ( public type B hospital ) , RS Bethesda ( private hospital ) , Puskesmas kota Yogyakarta ( Puskesmas ) ; two facilities in Bali: Sanglah Hospital ( public hospital ) and Puskesmas VI Denpasar ( Puskesmas ) . Patient records were randomly selected from a list of all age-stratified dengue diagnoses , maintained in facility diagnosis log books , in the 12 months preceding the beginning of the study . We planned to assess 50 inpatient and 50 outpatient records from each hospital ( total sample from six hospitals = 600 ) ; and 50 outpatient records from each Puskesmas ( total sample from three sites = 150 ) . It was expected that the sample would comprise an equal number of children ( ≤18 years old ) and adults ( ≥ 19 years old ) due to the approximately equal distribution of dengue cases occurring in these categories . Additionally , we intended to interview 30 inpatients and 30 outpatients or their respective parents/guardians at each hospital ( total sample from six hospitals = 360 ) and 30 outpatients or their parents/guardians from each Puskesmas ( total sample from three sites = 90 ) . Sample sizes were chosen to be operationally feasible and sufficiently large that analysis methods based on the normal distribution may be used for the analysis . Direct medical costs were retrospectively assessed through review of medical records and billing/charges made to patients who received treatment at selected hospitals or Puskesmas ( sub-district level health centres ) in the 12 months prior to the beginning of the study ( April 1st 2014 until March 31st 2015 ) with a diagnosis of dengue or dengue haemorrhagic fever ( with ICD 10 code A90 and A91 ) . Direct non-medical costs and indirect costs were assessed from data collected during face-to-face interviews with patients or their parents/guardians at selected hospitals or Puskesmas . Patients with clinically diagnosed/laboratory confirmed dengue or those with evidence of fever >38˚C for >1 day , plus symptoms compatible with dengue fever were recruited to participate in two interviews . The first was a face-to-face interview with patients or their parents/guardians using a questionnaire and conducted at the health facility at discharge/during an ambulatory visit . The second interview was conducted by telephone two weeks later to determine subsequent costs of treatment and any absenteeism from work/school . Direct non-medical costs were defined to include all expenses incurred due to the treatment , such as meals , transport , accommodation for care givers , etc . The interviews documented: the use of medical services; missed schooling; lost work productivity; out-of-pocket spending ( e . g . transportation , meals , hotel/house rental , etc ) and income lost due to the episode of illness . In the event that participants chose not to disclose their income and in the absence of reliable data on average wages including in the informal economy , we applied the standard minimum wages as a proxy , which are regulated in Indonesia and differ by province . Lost productivity was not calculated for children; rather , for each affected school child lost productivity was calculated for the caregiver ( as a result of leaving work to care for the child ) . Costs were expressed in US dollars ( as of 2016 with a conversion rate: US$1 = IDR13 , 000 ) . For those regions where a particular type of facility was not included in the study , gaps in the data were filled via weighted adjustment from neighbouring sites . For example , private outpatient costs were captured by recording treatment bills paid by the patient in Jakarta . Because private outpatient facilities were absent in Yogyakarta , these costs were estimated by adjusting Jakarta values weighted according to outpatient public costs for Jakarta and Yogyakarta . In Bali , private outpatient and inpatient costs were estimated based on the ratio observed in Yogyakarta . Passive reporting of dengue in Indonesia is mandatory within 72 hours of diagnosis according to SEARO-WHO dengue diagnosis guidelines 2011 . [31] Notification follows diagnosis by clinical and/or laboratory confirmation ( by detection of NS1 antigen and/or IgM/IgG ) . Cases are reported to provincial health offices and pooled at the provincial and national levels by the Directorate General of CDC . [32] Costs at the national level were estimated by multiplying cost per case ( outpatient/inpatient ) by an estimate of the number of cases occurring in Indonesia in 2015 . National burden estimates were generated using a ) provincial-level surveillance data from each of the 34 provinces; b ) estimates of hospitalization rate derived from an expert consensus technique in Indonesia;[11] and c ) a study which observed a magnitude of dengue under-reporting of 11 . 5-fold in the placebo group of a dengue vaccine clinical trial in Jakarta , Bandung and Denpasar , Bali . [33] The expert panel that gave rise to the estimates of hospitalization rate has been described previously;[11] briefly , it comprised a group of Indonesian dengue experts ( clinicians , hospital managers , epidemiologists and Ministry of Health officials ) who reviewed existing data sources and made iterative estimates of epidemiological parameters by which full burden estimates could be made . These were balanced against published analyses . [3 , 5 , 33–37] The panel concluded that 60% of dengue cases in Indonesia were hospitalized; a figure which , when combined with an estimated reported hospitalization rate and under-reporting factor of 11 . 5 , generated the final expansion factor for hospitalized patients ( EFH; 7 . 65 ) and expansion factor for ambulatory patients ( EFA; 45 . 90 ) used for calculation of the cost-of-illness . The numbers of ambulatory and hospitalized dengue cases for each province in Indonesia during 2015 were estimated by multiplying these expansion factors by the numbers of reported cases in each province . To calculate the economic burden of dengue nationally the three sites in our study: Jakarta , Yogyakarta and Bali , were used as references for other provinces arranged into three groups according to their fiscal capacity index ( FCI ) . Yogyakarta was the reference for low FCI province ( FCI <0 . 5 ) , Bali for middle FCI ( 0 . 5–2 . 0 ) and Jakarta for high FCI ( >2 . 0 ) provinces . Unit outpatient and inpatient costs of each province were proportionally weighted by the consumer price index ( CPI ) or Indeks Harga Konsumen ( IHK ) using Jakarta , Bali and Yogyakarta dengue unit costs as the baseline . By multiplying the number of ambulatory/hospitalized cases by the unit cost estimates for each province , the total economic burden in each province was calculated . [11] To assess the uncertainty surrounding estimated overall dengue burden , [38] deterministic sensitivity analyses were performed to examine the effect of parameters’ variations i . e . costs in each setting ( inpatient , outpatient , by region ) and expansion factor . Each parameter was manually varied by an arbitrary value of ±10% to examine the impact on the total economic burden . Calculations were performed using Microsoft Office Excel 2010 .
A total of 615 patient records were reviewed for the retrospective , direct medical cost calculation ( 262 in Jakarta , 251 in Yogyakarta and 102 in Bali ) during the period from the 15th of June to the 31st of July 2015 . The regional distribution of patients included , by province , inpatient/outpatient and dengue classification is shown in Table 1 . A total of 199 patients were involved in the prospective phase of the study ( 94 , 43 and 62 from each site ) ; data were collected from interviews during the period from the 3rd of August to the 15th of September 2015 . Combining both retrospective and prospective elements , the study sample was 68% of the enrolment target . The total costs ( combined direct and indirect costs ) per patient episode for outpatient cases were , US$103 , US$252 and US$179 for Yogyakarta , Bali and Jakarta , respectively . For inpatients these costs were US$689 , US$989 and US$1071 respectively . With the exception of inpatient costs in Jakarta , direct medical costs were higher from private hospitals compared with public facilities . Table 2 describes cost of illness results per episode for each site . Direct medical cost were the largest proportion of costs for inpatient care , while indirect costs were the largest proportion of costs for outpatient care . Outpatient costs in Jakarta were slightly lower than in Bali . Overall , the mean length of hospital stay was 3 . 9 days . By region , it was 4 . 4 days in Yogyakarta , 3 . 5 for Bali and 3 . 8 days in Jakarta . The results of the extrapolated regional costs ( by CPI and FCI ) are shown in Table 3 . Jakarta was the province with the highest dengue-related cost , followed by Yogyakarta , West Java and West Kalimantan . The annual total cost of dengue-related illness in Indonesia was estimated at US$381 . 5 million , with US$354 , 802 , 570 for hospitalised and US$26 , 249 , 519 for ambulatory cases ( Table 3 ) . Considering the total number of inpatient cases and costs , the average cost per dengue patient was lowest in region 1 ( $346 . 38 ) and highest in region 3 ( US$535 . 91 ) . Similarly , average cost per outpatient was lowest in region 1 ( US$34 . 38 ) and highest in region 2 ( US$84 . 48 ) . Results from the sensitivity analyses are presented in the Tornado diagram ( Fig 1 ) , which represents baseline value ( per US$ million ) . The parameters included in sensitivity analysis were the costs of outpatient and inpatient treatment by province; and EFA and EFH . Variation in any of these parameters resulted in overall economic burden varying from US$166–557 million . The greatest variation in the final estimate followed variation in outpatient costs in Jakarta; followed by costs in outpatient facilities in type A clinics , and in Bali .
We estimated the average annual economic burden of dengue-related illness in Indonesia in 2015 to be US$381 . 5 million with more than 90% of this cost associated with hospitalized care . Jakarta was the province associated with the greatest cost , which is a function of the greater population and the higher average costs of treating hospitalized dengue episodes . In Jakarta , inpatient , direct medical costs were higher from public facilities than in private hospitals . This was thought to result from the fact that the public study sites included Ciptomangunkusumo Hospital which is a type A public hospital , a top referral hospital in Indonesia and therefore responsible for treating the most severe cases requiring intensive , expensive , specialist care . Sensitivity analyses identified uncertainty around outpatient cost in Jakarta as the variable with the largest impact on the overall economic burden , due to the relatively higher cost of episodes in Jakarta , and their frequency . Notably , the overall estimates are directly influenced by the expansion factors used to estimate the number of cases . These numbers were derived from high-quality epidemiological studies in tandem with local expert opinion . But studies have shown reporting completeness can be affected by changes in disease severity , level of epidemic activity and other external factors , which could limit the generalizability of these numbers at different time points . 2015 was a fairly “typical” year in Indonesia , with the number of cases being close to the average from 2010–2016 . [39 , 40] Future analyses will hopefully allow for a more refined understanding of the level of dengue reporting in Indonesia . Our estimate of the cost per episode in type B hospital was~US$150 ( ~IDR 2 million ) , which is consistent with a previous Indonesian estimate from East Java of IDR1–2 million published in 2008 . [29] Our unit cost estimates are also similar to those reported in the regional analyses of Shepard in 2013 and 2016 . [26 , 27] Our study found that dengue is associated with considerable economic burden , which is in agreement with other studies conducted in Asian countries , especially those in Thailand and the Philippines . In Thailand , Philippines and Malaysia , total economic burdens were estimated at US$486 million ( in 2005 costs ) , [19] US$345 million ( in 2012 costs ) , [25] US$102 . 25 million ( in 2009 costs ) , [20 , 21] respectively . However , estimates in the much smaller ( Singapore ) and larger ( India ) countries were considerably higher than our estimate at more than US$1 billion in each country . [23 , 41] With regard to existing national level burden estimates for Indonesia , our results are similar to those published by Shepard and colleagues in 2013 who concluded that the annual economic burden of dengue for Indonesia was US$323 million . [26] This was slightly lower than our 2015 estimate , caused by an increasing disease burden; and slightly higher outpatient unit costs . However , this group refined their estimates in a 2016 [27] publication using a different method of epidemiological burden estimation and concluded that the dengue burden in Indonesia was US$2 billion . [27] Costs in our study were calculated from primary data sources and clinically diagnosed dengue , including medical record review and patient interview . Unit costs were broadly similar to those estimated by Shepard and colleagues and the variation is predominantly driven by different epidemiological estimates: Shepard and colleagues’ estimated >11 million annual dengue cases , while we assumed ~640 , 000 . Such variation in dengue burden estimates are difficult to reconcile; the paper by Shepard and colleagues applied regression methods from the Global Disease Burden group; in contrast we used local surveillance data combined with expert opinion and empirical under-reporting calculation . Much of this variation likely stems from case definitions , particularly those around mild cases of dengue whose clinical and economic significance is very difficult to calculate with confidence , and whose full economic impacts are very difficult to measure . In addition , the Shepard 2016 study included estimates for non-medical cases ( i . e . patients that did not seek professional medical advice but may have had laboratory testing or purchased therapeutic products outside the professional healthcare system ) , which we did not include in this analysis . The strength of this study is that it is based on empirical , patient-specific data for medical care and out-patient costs in Indonesia . Furthermore , it considered both public and private hospitals and included costs derived from different treatment settings and economic backgrounds . To address limitations in the available passive surveillance data , expansion factors were used to fully describe the number of dengue cases and expert opinion employed to desegregate data into outpatient and inpatient cases . We consider this approach , underpinned by gold-standard epidemiological clinical trial data with local expert opinion to stratify cases by severity , is likely a realistic representation of the health-seeking dengue case population in Indonesia . The costs captured from the three reference provinces ( Jakarta , Bali and Yogyakarta ) were extrapolated to other regions based on weighted average costs linked to the consumer price index to ensure relevant estimates from other regions . Other variables such as type of hospitals ( private/public , type A or B ) were also taken into account in the extrapolation to get a mixed representation of healthcare setting throughout the country . We acknowledge several limitations to our study , mostly due to the patients’ clinical pathway i . e . most patients generally received outpatient services at type B hospitals , hence had an impact on type A sample size; also the number of ambulatory patients was generally lower than expected ( potentially due to the local regulation at Jakarta and Yogyakarta whereby laboratory-confirmed dengue patients were referred directly to hospital ) . Furthermore , we did not enroll as many patients as planned and were only able to achieve 68% of the target enrolment . The primary reason for lower-than-expected enrolment was the relatively small number of dengue cases occurring in 2015 , especially in Yogyakarta in which enrolment was especially challenging . Outpatient recruitment was additionally complicated by local clinical practice guidelines which advise that all dengue cases should be hospitalized . There is uncertainty in income loss calculations due to illness because most patients or their parents/guardians did not disclose their actual income during the interviews; so the national minimum wage was used as proxy . Some studies also included ‘outside hospital costs’ , such as vector control activities , in the overall cost estimates , but this was beyond the scope/focus of our study . Lastly , our estimates are based on data from one year ( 2015 ) , corresponding to the period over which primary data were collected . As a result , the estimates are subject to vary with epidemic activity .
The total direct costs of dengue illness in Indonesia were estimated at US$381 . 15 million . Our analysis provides results that are relevant to public health policymakers in Indonesia , helping to strengthen local knowledge and informing decision-making regarding the prevention and control of dengue in public health priority lists . These results can also be used in health economic studies of novel dengue prevention and control technologies or vaccine programs . | Dengue , an infection transmitted by mosquitos , is a public health concern particularly in tropical/subtropical areas and the Asia Pacific region where it is associated with a significant cost to society . Indonesia has one of the world’s highest dengue burdens but Indonesia-specific data on cost are lacking . To estimate the annual economic burden of dengue in Indonesia , this study collected data from public/private hospitals and health centres in three provinces ( Yogyakarta , Bali and Jakarta ) during 2015 . We estimated cost of illness using the societal perspective: calculations of costs included those that were directly paid by the healthcare system , as well as costs incurred by the patients ( or their family/care givers ) and their lost productivity . The costs from the three provinces were then used as the basis for extrapolating cost of illness in Indonesia . The authors confirmed that dengue imposed a substantial economic burden for Indonesian public payers and society . Based on 2015 data , the authors estimated total economic burden of dengue in Indonesia at US$381 . 15 million . Of this , US$355 . 2 million related to patients treated in hospitals and US$26 . 2 million was for patients treated in health centres . Establishing a better understanding of the burden of dengue in Indonesia will help to guide public health decision-making at a national level and support prevention and control initiatives for this disease . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Conclusion"
] | [
"medicine",
"and",
"health",
"sciences",
"geographical",
"locations",
"tropical",
"diseases",
"social",
"sciences",
"indonesia",
"health",
"care",
"inpatients",
"global",
"health",
"neglected",
"tropical",
"diseases",
"patients",
"public",
"and",
"occupational",
"health... | 2019 | Economic burden of dengue in Indonesia |
Although sleep appears to be broadly conserved in animals , the physiological functions of sleep remain unclear . In this study , we sought to identify a physiological defect common to a diverse group of short-sleeping Drosophila mutants , which might provide insight into the function and regulation of sleep . We found that these short-sleeping mutants share a common phenotype of sensitivity to acute oxidative stress , exhibiting shorter survival times than controls . We further showed that increasing sleep in wild-type flies using genetic or pharmacological approaches increases survival after oxidative challenge . Moreover , reducing oxidative stress in the neurons of wild-type flies by overexpression of antioxidant genes reduces the amount of sleep . Together , these results support the hypothesis that a key function of sleep is to defend against oxidative stress and also point to a reciprocal role for reactive oxygen species ( ROS ) in neurons in the regulation of sleep .
A sleeping animal is vulnerable to predators and other dangers in its environment for a large portion of the day . Despite these daily risks , sleep is an evolutionarily conserved behavior throughout the animal kingdom [1–3] , suggesting that sleep serves important functions . In support of this , prolonged episodes of acute sleep deprivation in both rodents and invertebrates cause an increased need to sleep [4–7] , cognitive impairment [8 , 9] , increased metabolic rate [6 , 10] , and death [6 , 10 , 11] . It remains unclear whether these effects are due to loss of sleep or due to the intense stress associated with acute sleep deprivation . Epidemiological studies have revealed that chronic sleep restriction , or shortened sleep duration , in humans is associated with metabolic disorders [12] , cardiovascular disease [13] , inflammation [14 , 15] , psychiatric disorders [16] , and even premature mortality [17 , 18] . Similar to experimental results involving acute sleep deprivation , it is unclear whether these defects are due to the loss of sleep itself , to associated disruptions in circadian rhythm , or from the very factors that cause sleep loss , such as shift work , aging , or psychological stress . Thus , while current research in both humans and model organisms has demonstrated an important role for sleep in learning and memory [19–22] , it has been difficult to identify underlying functions for sleep essential to the organism’s survival or fitness . Sleep is thought to be regulated by two distinct types of mechanisms: those that control the timing of sleep , such as the circadian system , and those that control the duration of sleep , also called sleep homeostasis mechanisms [23 , 24] . While the molecular mechanisms underlying circadian regulation have been well characterized , molecular mechanisms regulating sleep homeostasis are less well defined but are thought to be neuronally based [24–29] and context dependent—that is , sleep deprivation or other stress conditions may induce different homeostasis pathways than baseline sleep . Because acute sleep deprivation increases sleep need and results in extended sleep duration at the animal’s next opportunity to sleep , many models of sleep homeostasis propose a feedback mechanism in which the wake state increases sleep-promoting factors , such as adenosine or overall synaptic strength [24 , 29] . The sleep state then clears or abrogates these factors to allow the wake state . A controversial hypothesis for the function of sleep is the free radical flux theory of sleep , proposed in a theoretical paper by Reimund in 1994 . Reimund proposed that reactive oxygen species ( ROS ) accumulate in neurons during the wake state and that sleep allows for the clearance of ROS in the brain [30] . ROS are chemically reactive by-products of metabolism , which , when not properly neutralized , cause damaging covalent modifications that inhibit the function of proteins , lipids , and DNA and can lead to cell death . Thus , the free radical flux hypothesis proposed that the core function of sleep is to act as an antioxidant for the brain . Despite the appeal of this hypothesis , data to support it are conflicting . While some groups have reported decreased antioxidant capacity and oxidative damage in the brains of sleep-deprived rats and mice [31–34] , other reports have contradicted these findings [35–37] . As a result , the Reimund hypothesis has fallen out of favor as a model for sleep function . Notably , all studies testing the Reimund hypothesis focused on the effects of acute sleep deprivation . In contrast to acute sleep deprivation , the relationship between chronic sleep restriction and oxidative stress has not been thoroughly investigated , despite the physiological relevance of chronic sleep restriction widespread in modern society [38] . In recent years , the fruit fly has become a powerful , genetically tractable model system for the study of sleep [39 , 40] . Forward genetic screens have identified a number of Drosophila mutants that are short sleeping and retain intact circadian rhythms . Loss-of-function mutations in ion channels and ion-channel regulators , including sleepless , which regulates the potassium channel Shaker and nicotinic acetylcholine receptors ( nAChRs ) , have been shown to reduce sleep [20 , 26 , 41 , 42] . Other short sleep–causing mutations include the redeye allele of the nAChRα4 subunit [43] , the fumin allele of the dopamine transporter ( DAT ) [44] , and loss of function of the putative ubiquitin ligase adaptor encoded by insomniac ( inc ) [45 , 46] . It has been hypothesized that these mutations cause short sleep by increasing neuronal excitability [24] . These mutants allow researchers to investigate the effects of chronic short sleep independent of circadian defects . While the specific genes affected vary widely and it is not clear whether these mutants sleep less than controls because of reduced sleep need or an inability to sleep , the common phenotype of these diverse mutants is chronic short sleep . Thus , together these mutants provide a system for identifying a “core” or essential function of sleep; we hypothesized that if chronic short sleep has negative effects on health , these diverse short-sleeping Drosophila mutants might share a common physiological defect independent of the specific mechanism driving their short sleep . In this study , we sought to identify a physiological defect common to short-sleeping flies that might provide insight into the function and regulation of sleep . We found that diverse short-sleeping mutants are sensitive to acute oxidative stress , exhibiting shorter survival times than controls , and that increasing total sleep duration of wild-type flies promotes survival after oxidative challenge . We further showed that neuronal overexpression of antioxidant genes in wild-type flies reduces sleep . Our data demonstrate that one function of sleep is to increase the organism’s resistance to oxidative stress and support the hypothesis that sleep abrogates neuronal oxidative stress; these results also point to a role for neuronal ROS in the homeostatic regulation of sleep .
To identify specific physiological functions of sleep ( Fig 1A ) , we first focused on neuron-specific RNA interference ( RNAi ) of the inc gene , which has been shown to cause short sleep [45 , 46] . inc encodes a putative adaptor protein for Cullin-3 ( Cul3 ) , an E3 ubiquitin ligase expressed in both the brain and the body . Cul3 is involved in a number of crucial biological processes , and inc null mutants have reduced lifespan [45] . In contrast , neuron-specific RNAi of inc was reported to cause short sleep without affecting lifespan [45] , suggesting that reduction of Inc activity in nonneuronal tissues affects lifespan in a sleep-independent manner . For this reason , we used flies expressing neuron-specific inc-RNAi as our initial model of short sleep . We verified that animals expressing an upstream activation sequence ( UAS ) -inc-RNAi construct via the pan-neuronal driver elav-GAL4 , hereafter referred to as neuronal inc-RNAi flies , exhibited a 30% reduction in total sleep time relative to isogenic controls carrying one copy of either the inc-RNAi construct or elav driver alone ( Fig 1B , p < 0 . 0001 relative to either control; S1A Fig ) . We further confirmed that neuronal inc-RNAi flies exhibit normal lifespan compared to controls ( Fig 1C , p > 0 . 5 compared to either control ) , consistent with a previous report [45] and with recent findings on inbred short-sleeping Drosophila lines that have normal lifespan [47] . This result confirms earlier findings that chronic short sleep does not itself shorten lifespan . Changes in sleep are often associated with altered metabolic energy storage . In humans and mice , sleep loss is associated with metabolic dysfunction such as obesity [48 , 49] , and in flies , starvation suppresses sleep behavior [50] and prolonged sleep is associated with increased starvation resistance [51] . We tested whether neuronal inc-RNAi flies have altered starvation resistance , which reflects altered metabolic energy stores . We found that the mortality rate of inc-RNAi flies after starvation was intermediate between normally sleeping control flies containing either the elav driver or the UAS-inc-RNAi construct alone ( Fig 1D , p = 0 . 0592 compared to elav control , p = 0 . 0493 compared to inc-RNAi control ) , suggesting that short sleep does not affect metabolic energy storage in neuronal inc-RNAi animals . Acute sleep deprivation has also been associated with immune dysfunction in humans , rats , and mice [52–55] . Work in flies has shown that acute sleep deprivation can also augment the immune response [56] . To assay for defects or enhancement in immunity because of chronic short sleep , we injected neuronal inc-RNAi flies with different bacterial pathogens , including Streptococcus pneumoniae , a gram-positive pathogen that has been well characterized in Drosophila ( Fig 1E ) , Providencia rettgeri , a gram-negative natural pathogen found in wild-caught Drosophila ( Fig 1F ) , Listeria monocytogenes , and Staphylococcus aureus ( S1B and S1C Fig ) . In each case , neuronal inc-RNAi flies died at the same rate as one or both of their genetic controls . To further test whether chronically reduced sleep causes deficits in immune function , we examined the response of short-sleeping fumin mutants that lack a functional DAT [44] . We confirmed earlier findings that fumin mutants exhibit short sleep ( an approximately 95% reduction in sleep relative to controls ) ( S1D Fig ) . We found that fumin mutants responded variably to these pathogens ( S1E–S1H Fig ) . The lack of a consistent immunity defect across different pathogens in both neuronal inc-RNAi flies and fumin mutants suggests that chronic short sleep does not have a dramatic or common impact on immune function in Drosophila . We next set out to test whether sleep is required to defend against oxidative stress ( Fig 2A ) [30] . We compared the survival of neuronal inc-RNAi flies relative to controls when subjected to two different treatments that induce oxidative stress by increasing ROS levels ( Fig 2B ) . We first injected neuronal inc-RNAi flies with a lethal dose of paraquat , an herbicide that catalyzes the production of superoxide anions [57] . We found that neuronal inc-RNAi flies died at a significantly faster rate after paraquat injection than controls ( Fig 2B , left panel , p < 0 . 0001 relative to either control ) . To determine whether neuronal inc-RNAi flies have a specific sensitivity to superoxide anions or if they are also sensitive to other forms of oxidative stress , neuronal inc-RNAi flies and controls were fed hydrogen peroxide ( H2O2 ) , an oxidant that produces highly reactive hydroxyl radicals and has been shown to alter locomotor activity when fed to flies [58] . Similar to paraquat injection , neuronal inc-RNAi flies were sensitive to H2O2 feeding compared to controls ( Fig 2B , right panel , p < 0 . 0001 relative to either control ) . These results indicate that short-sleeping neuronal inc-RNAi flies are susceptible to oxidative stress . To verify that oxidative stress sensitivity is caused by the reduction in inc expression , rather than an off-target effect of RNAi , we next tested inc null mutants for paraquat sensitivity . We confirmed that inc null mutants exhibit a 50% reduction in sleep ( S2A Fig , p < 0 . 0001 for both inc1 and inc2 mutants , relative to controls ) , as previously reported [45] . Consistent with neuronal inc-RNAi flies , inc null mutants died faster than controls when injected with paraquat ( Fig 2C , p < 0 . 0001 for both inc1 and inc2 mutants , relative to controls ) . Furthermore , because Inc is a putative adaptor for the Cul3 ubiquitin ligase , we predicted that reduction of neuronal Cul3 activity would also cause paraquat sensitivity . As previously reported [45] , neuronal Cul3-RNAi flies exhibit a 60% reduction in sleep ( S2B Fig , p < 0 . 0001 relative to either control ) ; here we found that neuronal Cul3-RNAi flies were also sensitive to paraquat injection ( Fig 2D , p < 0 . 0001 relative to either control ) . Thus , chronic short-sleeping inc null mutants and Cul3-RNAi flies are sensitive to oxidative stress induced by elevated ROS levels , similar to neuronal inc-RNAi flies . To determine whether sensitivity to oxidative stress is caused specifically by the reduction in inc or Cul3 activity or whether it is more broadly associated with loss of sleep , we next tested for sensitivity to oxidative stress in three different short-sleeping mutants , each carrying mutations in different genes with varied functions: sleeplessΔ40 ( sleepless ) , DATfumin ( fumin ) , and nAChRα4rye ( redeye ) ( Fig 3A ) . We first confirmed , as previously reported [42–44] , that each mutant spends significantly less time sleeping than its isogenic control ( Fig 3B–3D , left panels , p < 0 . 0001 for each; S1C , S3A and S3B Fig ) . We next tested these short-sleeping mutants for sensitivity to oxidative stress . Relative to controls , we found that each mutant was sensitive to both paraquat injection ( Fig 3B–3D , middle panels , p < 0 . 0001 for each ) and H2O2 feeding ( Fig 3B–3D , right panels , p < 0 . 0001 for each ) . Thus , our finding that this molecularly diverse set of short-sleeping mutants has a common susceptibility to oxidative challenge raises the possibility that sleep itself is required for proper response to oxidative stress . Because short-sleeping mutants exhibit sensitivity to oxidative stress , we next tested whether extending sleep duration promotes resistance to oxidative stress . We increased sleep by either genetic manipulation or pharmacological treatment and measured the effect on survival after oxidative challenge . For the genetic approach , we used transgenic flies in which sleep-inducing neurons were activated by the expression of a neuron-activating bacterial sodium channel [21] . For the pharmacological approach , we treated wild-type animals with the sleep-inducing drug Gaboxadol [19 , 59] . It was previously shown that total sleep time is increased by constitutively activating neurons in the dorsal Fan-shaped Body ( dFB ) , a sleep-promoting region in the fly brain [21] . We verified this phenotype using a previously established dFB driver ( 23E10-GAL4 ) [60] to drive expression of the neuron-activating bacterial sodium channel construct UAS-NaChBac [61] and observed a 40% increase in sleep duration in dFB>NaChBac flies ( Fig 4A , left panel , p < 0 . 0001 relative to either control; S3C Fig ) . We then subjected dFB>NaChBac flies to oxidative stress by either paraquat injection or H2O2 feeding . In both cases , dFB-activated flies died at a slower rate than controls ( Fig 4A , middle and right panels , p < 0 . 001 for each ) . Thus , genetically activating the dFB to increase sleep promotes resistance to oxidative stress . To further test whether extended sleep duration can increase survival of acute oxidative stress , we used an independent pharmacological method of sleep induction . Wild-type animals were fed the GABAA receptor agonist Gaboxadol , which induces sleep in Drosophila [19 , 59] . We observed a 25% increase in total sleep time in Gaboxadol-treated animals ( Fig 4B , left panel , p < 0 . 001; S3D Fig ) and a corresponding increase in resistance to paraquat injection relative to vehicle-fed controls ( Fig 4B , right panel , p < 0 . 0001 ) . Together , these results demonstrate that two different methods of increasing sleep both promote resistance to oxidative stress , consistent with the idea that oxidative stress resistance is a physiological function of sleep ( Fig 4C ) . If sleep clears ROS from neurons , one would expect short-sleeping flies to exhibit higher baseline levels of ROS in the brain . Quantitation of ROS in live brains is extremely difficult , possibly due to tight feedback control of ROS levels via the induction of antioxidant gene expression . As an indirect measure of ROS , we measured the expression of genes known to be activated by high levels of ROS by performing quantitative reverse transcription polymerase chain reaction ( qRT-PCR ) on the heads of neuronal inc-RNAi flies and controls ( Fig 5A ) . These genes include the antioxidant genes superoxide dismutase 1 ( SOD1 ) , catalase , the glutathione-S-transferases GSTS1 and GSTO1 , and; the mitochondrial stress response genes hsp60 , ClpX , and Pink1; and the endoplasmic reticulum stress response gene BiP , which was previously shown to be induced by sleep deprivation [40 , 62–64] . We found that neuronal inc-RNAi flies exhibited increased expression of all of these genes except catalase and BiP ( Fig 5B–5I ) . While neuronal inc-RNAi flies had modestly elevated BiP expression in the head ( Fig 5I ) , the difference was not significant . Thus , the increased baseline expression of antioxidant genes and mitochondrial stress genes in neuronal inc-RNAi flies is consistent with short sleep causing increased ROS levels in the brain . If one function of sleep is to clear ROS from the brain , then it is plausible that ROS itself may be a factor that triggers sleep , perhaps when it reaches a certain critical threshold . To determine whether neuronal ROS levels play a role in the regulation of sleep , we reduced ROS levels in the brains of otherwise wild-type flies by driving neuronal overexpression of the antioxidant genes catalase , SOD1 , or SOD2 using the elav-Gal4 driver ( Fig 6A ) . SOD1 or SOD2 overexpression resulted in a significant reduction in the total amount of sleep , with an average decrease in total sleep of 10% and 16% , respectively ( Fig 6B , p < 0 . 05 compared to either control; S3F and S3G Fig ) . catalase overexpression resulted in a similar trend but did not reach significance compared to the driver control ( Fig 6B , S3E Fig ) . Our observation that reducing neuronal ROS levels reduces sleep amount suggests that ROS levels reflect sleep need and play a role in the regulation of sleep ( Fig 6C ) .
Although sleep appears to be evolutionarily conserved across all animal species [1–3] , the physiological function of sleep remains unclear . Our data show that chronic sleep restriction sensitizes flies to two types of oxidative stress: paraquat injection and H2O2 feeding ( Figs 2 and 3 ) . Conversely , increasing sleep through either genetic or pharmacological methods promotes resistance to oxidative stress ( Fig 4 ) . Thus , our data suggest that one important function of sleep is defense against oxidative stress . The molecular mechanisms underlying the susceptibility of short-sleeping mutants to acute oxidative stress and whether this susceptibility is due to the effects of oxidative stress on the brain or other , nonneuronal tissues of the body remains unclear . It is possible that increased baseline ROS levels in neurons or other tissues sensitize short sleepers to acute oxidative stress . Other investigators have found that accumulation of cellular ROS was associated with susceptibility to acute oxidative challenge [65 , 66] . Chronic sleep loss may lead to accumulated mitochondrial damage that , in the presence of an acute oxidative stress , triggers cell death pathways . Another possibility is that short sleepers are less able to detect or respond to acute oxidative challenge in specific tissues . Testing these hypotheses will be an important focus for future investigation . Our data also suggest that short-sleeping animals accumulate higher baseline ROS levels in the brain . While ROS levels in the brain are difficult to measure directly , we observed increased expression of antioxidant and mitochondrial stress response genes in the heads of short-sleeping neuronal inc-RNAi flies , consistent with increased ROS levels in the brain . Other studies have similarly observed that sleep-deprived animals display increased expression of genes induced by high ROS levels . Induction of the antioxidant regulator cap’n’collar ( cnc ) was observed in fly heads when flies were exposed to recurrent sleep fragmentation [67] , and its mammalian homolog Nuclear factor ( erythroid-derived 2 ) -like 2 ( nrf2 ) was reported to be induced in the cerebral cortex of mice after 6 hours of sleep deprivation [68] . Sleep deprivation has also been associated with activation of the unfolded protein response in the ER in fly heads and mouse brains [40 , 62–64] . Because both the ER- and mitochondrial unfolded protein responses can be induced by high levels of ROS , we hypothesize that both genetic and environmental sleep loss increase baseline ROS levels that , depending on the specific method of sleep deprivation , genetic background , and tissue tested , are reflected in the activation of different response pathways . Finally , we found that increasing antioxidant gene expression in the brain causes short sleep , suggesting that decreasing neuronal ROS levels will promote the wake state . Emerging evidence demonstrates that ROS can act as crucial signaling molecules in a number of biological processes [69 , 70] , and it has been demonstrated that injecting an oxidant into the rat brain induces sleep [71] . One study showed modest effects of lifelong , low-dose paraquat feeding on sleep in flies [72] . Thus , ROS levels , either directly or indirectly through the activation of oxidative stress responses , appear to induce sleep . Taken together , our results support a model for a bidirectional relationship between sleep and oxidative stress , in which one function of sleep is to act as an antioxidant for both the body and the brain , increasing the organism’s resistance to acute oxidative challenge and reducing ROS levels in the brain; moreover , neuronal ROS play a role in the regulation of sleep and wake states ( Fig 6C ) . Thus , with chronic sleep restriction , the animal accumulates higher ROS levels in the brain and is sensitive to acute oxidative stress . Identifying the physiological functions and key regulators of sleep is critical to understanding the negative effects on health associated with chronic sleep restriction . In the United States , average sleep time is steadily decreasing [73] , and one third of adults sleep less than the recommended 7 hours per night [38] . Sleep restriction is correlated with a variety of diseases [12 , 13] , many of which are also associated with oxidative stress [74–78] . Sleep disturbances have been implicated as a predictor for Alzheimer , Parkinson , and Huntington’s diseases [79–82] , and in all of these diseases , oxidative damage has been reported in the brains of patients postmortem [83–85] . Because oxidative stress can induce protein misfolding and aggregation through protein damage , neuronal accumulation of ROS is a plausible contributing factor in the pathogenesis of neurodegenerative diseases . Thus , understanding the role of sleep in defense against oxidative stress and the role of ROS in regulating sleep could provide much-needed insight into the pathology and treatment of neurodegenerative diseases .
The following flies were used to manipulate inc and Cul3 as described previously [45]: UAS-inc-RNAi ( VDRC stock #18225 ) , elavC155-Gal4 , UAS-Dicer ( dcr ) ( Bloomington stock #24651 ) , inc1 deletion mutant , and inc2 transposon insertion mutant ( CG32810f00285 ) , all in the same genetic background ( w1118 iso31 or Bloomington stock #5905 ) , along with the isogenic iso31 strain used for outcrossing . UAS-Cul3-RNAi ( NIG stock #11861R-2 ) was in the NIG w1118 background and compared to its isogenic control . For neuronal Cul3 knockdown experiments , the UAS-Dicer line ( Bloomington stock #24651 ) was crossed into the elavC155-Gal4 line . Parental controls used for experiments were obtained by crossing expression driver ( e . g . , elav-Gal4 ) and RNAi construct ( e . g . , UAS-inc-RNAi ) lines to the outcrossed wild-type line ( e . g . , iso31 ) for heterozygous controls , accounting for differences in complex phenotypes affected by genetic background . In case the absence of the white gene , which encodes an ABC transporter , has an effect on survival after paraquat or H2O2 exposure , red-eyed controls were used with the red- and orange-eyed inc1 and inc2 mutants; these w+ controls were generated by outcrossing w+ from an Oregon-R background for eight generations with the iso31 stock ( Bloomington stock #5905 ) . redeye , sleeplessΔ40 ( imprecise excision mutants ) , and their corresponding background-matched controls were obtained from Amita Sehgal ( University of Pennsylvania ) . sleeplessΔ40 was used instead of sleeplessP1 because sleeplessP1 flies were sensitive to CO2 , which made paraquat injection experiments difficult to interpret . Male sleeplessΔ40 flies also exhibited some wounding sensitivity , whereas females did not , so female sleeplessΔ40 flies were used in the paraquat injection experiments ( S4 Fig ) . Male sleeplessΔ40 were used in H2O2 feeding experiments . fumin mutants and their background-matched controls were obtained from Rob Jackson ( Tufts University ) . UAS-NaChBac [61] was obtained from Paul Shaw ( Washington University , St . Louis , MO ) and 23E10-Gal4 [60] was obtained from Jeffrey Donlea ( University of Oxford ) ; both were outcrossed for eight generations with the iso31 stock . As described above , parental controls used for experiments were obtained by crossing expression driver ( 23E10-Gal4 ) and transgene construct ( UAS-NaChBac ) lines to the outcrossed wild-type line ( iso31 ) for heterozygous controls . The following stocks were obtained from the Bloomington Stock Center ( BDSC , Bloomington , IN ) and outcrossed 6–8 generations into the iso31 background: UAS-SOD1 ( #24754 ) , UAS-SOD2 ( #24492 ) , and UAS-cat ( #24621 ) . All flies were raised at room temperature on standard molasses food ( 5 . 85% cornmeal , 2 . 675% yeast , 0 . 575% agar , 3% v/v blackstrap molasses , 0 . 14% methylparaben , 0 . 5% v/v propionic acid ) and kept on cornmeal food ( 4% cornmeal , 2 . 15% yeast , 9% dextrose , 0 . 75% agar , 0 . 095% methylparaben ) post-eclosion in a temperature- ( 25°C ) and humidity- ( 55% ) controlled incubator with a 12-hour light–dark cycle . Four- to ten-day-old males were used for all experiments , unless otherwise noted . Individual flies were loaded into plastic tubes containing cornmeal food and allowed to acclimate for 1 day . Sleep was monitored for 4–5 days using Drosophila Activity Monitors ( either DAM2s or DAM5s , an older model of DAM5M with a single beam per tube ) ( Trikinetics , Waltham , MA ) . Activity was recorded as beam-breaks in 1-minute bins and analyzed using PySolo software [86] or Microsoft Excel , with sleep defined as a 5-minute period of inactivity . Graphing and statistical analysis were performed using GraphPad Prism ( survival assays and scatterplots ) and PySolo ( 24-hour sleep profiles ) . When comparing two groups: an unpaired t test was performed when standard deviations were similar , and an unpaired t test with Welch’s correction was performed when standard deviations were not similar ( F test p < 0 . 5 ) . When comparing three groups , a one-way ANOVA was performed and followed by a post hoc Tukey test to compare means when significance was detected . For starvation assays , flies were transferred to tubes containing 1% agar and loaded into Drosophila Activity Monitors . Time of death was determined by complete loss of movement . Flies were collected on the day of eclosion and allowed to mate overnight . Total flies per genotype ranged from 74 to 225 . Numbers were roughly equivalent for each group within different trials . Males were separated into groups of 20 per vial . Flies were transferred to new vials every 2–7 days and scored for death at time of transfer . Lifespan experiments were performed in at least two independent trials . Injections were carried out with a pulled glass capillary needle . A custom-made microinjector ( Tritech Research , Los Angeles , CA ) was used to inject 50 nL of liquid into the abdomen of each fly . Volume was calibrated by measuring the diameter of the expelled drop under oil . Death was assayed visually at least daily , with a typical n = 60 for both bacterial infections and paraquat injections . For each experiment , a smaller set of flies was injected with vehicle alone to ensure that wounding caused minimal death . The following bacterial strains were used for injections: S . pneumoniae ( strain SP1 , a streptomycin-resistant variant of D39 ) obtained from Elizabeth Joyce ( University of California , San Francisco , CA ) was grown standing in Brain Heart Infusion media ( BHI , Teknova , Hollister , CA ) at 37°C with 5% CO2 , frozen into aliquots with 10% glycerol , pelleted and resuspended upon thawing , and injected at an OD600 of 0 . 015–0 . 05; P . rettgeri ( strain Dmel , a natural pathogen isolated from wild-caught D . melanogaster [87] ) obtained from Brian Lazzaro ( Cornell University ) was grown shaking in LB at 37°C and injected at an OD600 of 0 . 003–0 . 005; L . monocytogenes ( strain 10403S ) obtained from Julie Theriot ( Stanford University ) was grown standing in BHI at 37°C and injected at an OD600 of 0 . 075–0 . 2; and S . aureus strain 12600 ( ATCC ) was grown shaking in BHI at 37°C and injected at an OD600 of 0 . 0001–0 . 001 . Postinjection , flies were kept in a 29°C incubator for the remainder of the experiment to allow for optimal infection , with the exception of P . rettgeri injection , in which case optimal infection was achieved at 25°C . All OD600 measurements were made using a Genesys 10S Vis Spectrophotometer ( ThermoScientific , Waltham , MA ) , blanked against the corresponding sterile media for the given culture . Cultures were then diluted in sterile media to the desired OD . For paraquat injections , paraquat ( methyl viologen hydrate , Fisher Scientific , Hampton , NH ) was dissolved in water to a concentration of 3–5 mM . Paraquat solution was either stored at 4°C for up to 1 month or frozen in aliquots and thawed as needed . For every experimental genotype treated with paraquat injection , we conducted mock injections with ddH2O to control for wounding sensitivity ( S4 Fig ) . These assays were performed in two ways . In one method , flies were transferred to vials containing a folded Kimwipe soaked with 1 . 5 mL of a 5% sucrose , 1%–4% H2O2 solution . Thirty percent H2O2 ( Sigma-Aldrich , St . Louis , MO ) was diluted in ddH2O to a concentration of 1%–4% depending on the death rate for the given genotype , titrated to complete death within several days . Flies were flipped onto a freshly soaked Kimwipe every 2 days and death was assayed visually and recorded daily . This method allows very rapid setup ( typical experiment used 40 flies/genotype ) but provides relatively low-resolution survival kinetics . In the second method , flies were transferred to 5 mm tubes containing a piece of a soaked Kimwipe and loaded into Drosophila Activity Monitors , in which case death was determined by a complete loss of movement . Control flies were kept on 5% sucrose alone to ensure that death did not occur by starvation or desiccation . This method provides high-resolution survival kinetics but requires more time-intensive setup ( typical experiment used 30 flies/genotype ) . We found that all our results for short-sleeping mutants were consistent between the two methods . Survival curves for starvation assays , lifespan experiments , bacterial infections , paraquat injections , and H2O2 feeding assays are all plotted as Kaplan-Meier graphs . Log-rank analysis was performed using GraphPad Prism ( GraphPad Software , La Jolla , CA ) . All experiments were performed with a minimum of three independent trials and yielded statistically similar results , except where noted . Graphs and p-values in figures are from representative trials . Age-matched , 6–8-day-old flies were anesthetized on ice and decapitated between ZT2 and ZT5 . Fifteen to twenty heads per sample were homogenized in TRIzol ( Invitrogen ) , and a phenol-chloroform extraction was performed to isolate nucleic acids . Samples were treated with DNAse ( Invitrogen , Carlsbad , CA ) to isolate RNA and then diluted to a concentration of about 60 ng/μL . RevertAid First Strand cDNA synthesis kit ( ThermoFisher , Waltham , MA ) was used to convert RNA to cDNA . Quantitative RT-PCR was performed using a Bio-Rad CFX Connect Real-Time qPCR machine , with Express Sybr GreenER qPCR SuperMix ( Invitrogen , Carlsbad , CA ) and the following primer sets: SOD1: For: GGAGTCGGTGATGTTGACCT Rev: GGAGTCGGTGATGTTGACCT GSTS1: For: CACCAGAGCATTTCGATGGCT Rev: ACGACTGCAATTTTTAGACGGA GSTO1: For: ACGACTGCAATTTTTAGACGGA Rev: CCGATCGCCGGGAGTTCATGTAT catalase: For: TTCTGGTTATCCCGTTGAGC Rev: GGTAATGGCACCAGGAGAAA hsp60: For: TGATGCTGATCTCGTCAAGC Rev: TACTCGGAGGTGGTGTCCTC ClpX: For: AAAATGCTCGAAGGCACAGT Rev: TTGAGACGACGTGCGATAAG Pink1: For: TCGGTGGTCAATGTAGTGC Rev: CCACTCGGAAGATTCCACTGC BiP: For: GCTATTGCCTACGGTCTGGA Rev: CATCACACGCTGATCGAAGT actin: For: TTGTCTGGGCAAGAGGATCAG Rev: ACCACTCGCACTTGCACTTTC Analysis was performed using the Standard Curve method . Total cDNA concentration was normalized to actin expression . Data are represented as mean ± SEM . Five to six biological replicates ( containing 15–20 heads each ) per experiment . Gaboxadol hydrochloride ( Sigma-Aldrich , St . Louis , MO ) was dissolved in water and added to melted cornmeal food to a final concentration of 0 . 1–0 . 2 mg/mL . Flies were flipped onto Gaboxadol-containing food for 3 days prior to paraquat injection and remained on Gaboxadol-containing food postinjection . Control food was made by adding the appropriate amount of vehicle alone ( H2O ) to melted cornmeal food . | Most animals sleep; humans sleep nearly a third of their lives . Yet the fundamental functions of sleep remain unknown . Here , we used short-sleeping Drosophila mutants to uncover a role for sleep in resistance to oxidative stress . Oxidative stress is an imbalance of reactive oxygen species and antioxidant responses . Although these short-sleeping mutants have defects in diverse pathways , they all exhibit sensitivity to oxidative stress . Moreover , increasing sleep in wild-type flies increased resistance to oxidative stress . This suggests that one function of sleep is to defend against oxidative stress . Finally , reducing oxidative stress in neurons of wild-type flies reduces their sleep , suggesting that oxidative stress also regulates sleep . Taken together , our results support an intriguing hypothesis for a bidirectional relationship between sleep and oxidative stress: oxidative stress triggers sleep , which then acts as an antioxidant for both the body and the brain . These results have implications for human patients suffering from chronic sleep restriction and diseases associated with oxidative stress . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"sleep",
"deprivation",
"rna",
"interference",
"oxidative",
"stress",
"sleep",
"enzymes",
"enzymology",
"dismutases",
"animals",
"sleep",
"disorders",
"animal",
"models",
"physiological",
"processes",
"drosophila",... | 2018 | A bidirectional relationship between sleep and oxidative stress in Drosophila |
The Kaposi's sarcoma-associated herpesvirus gene products K3 and K5 are viral ubiquitin E3 ligases which downregulate MHC-I and additional cell surface immunoreceptors . To identify novel cellular genes required for K5 function we performed a forward genetic screen in near-haploid human KBM7 cells . The screen identified proteolipid protein 2 ( PLP2 ) , a MARVEL domain protein of unknown function , as essential for K5 activity . Genetic loss of PLP2 traps the viral ligase in the endoplasmic reticulum , where it is unable to ubiquitinate and degrade its substrates . Subsequent analysis of the plasma membrane proteome of K5-expressing KBM7 cells in the presence and absence of PLP2 revealed a wide range of novel K5 targets , all of which required PLP2 for their K5-mediated downregulation . This work ascribes a critical function to PLP2 for viral ligase activity and underlines the power of non-lethal haploid genetic screens in human cells to identify the genes involved in pathogen manipulation of the host immune system .
Manipulation of the cellular machinery of the host by viruses is essential to ensure their successful replication . This is particularly clear in the interactions between viruses and the immune system , as demonstrated by the large DNA viruses which encode multiple genes that manipulate the cell surface expression of many different immunoreceptors [1] . K3 and K5 are two genes encoded by Kaposi's sarcoma-associated herpesvirus ( KSHV ) which were originally identified through their ability to degrade major histocompatibility complex class I ( MHC-I ) molecules [2] , [3] . These genes encode membrane-bound E3 ubiquitin ligases , which use their N-terminal RING-CH domain to direct the polyubiquitination and subsequent endolysosomal degradation of target immunoreceptors [4] . Although K3 seems primarily focussed on MHC-I , K5 , with which it shares 40% amino acid identity , is more promiscuous and targets a variety of additional cell surface immunoreceptors for degradation . These include the NKT cell ligand CD1d [5] , the MHC-I-related molecule HFE [6] , the co-stimulatory molecule B7-2 [7] , the adhesion molecules ICAM-1 [7] , PECAM [8] and ALCAM [9] , the NK cell ligands MICA , MICB and AICL [10] , and the cellular restriction factor tetherin [11] , [12] . How a single ligase is able to target such a structurally diverse range of molecules for degradation , whilst retaining specificity , is not well understood . Although microscopy localises both K3 and K5 to the endoplasmic reticulum ( ER ) [2] , substrate ubiquitination occurs in the late secretory pathway , including the plasma membrane [13] , [14] . K3 and K5 must therefore traffic through the secretory pathway to the plasma membrane , where recruitment of serial E2 conjugating enzymes by the viral RING-CH domain leads to lysine-63-linked ( in the case of K3 ) [15] , or mixed lysine-11- and lysine-63-linked ( in the case of K5 ) polyubiquitin chain formation and the ESCRT-mediated endolysosomal degradation of target immunoreceptors [14] , [16] To further elucidate cellular genes required for K5 function , we took advantage of the recent development in forward genetic screens in the near-haploid human KBM7 cell line . Forward genetic analysis , the concept of starting with a biological process and proceeding through to gene discovery , has a proven track record of elucidating gene function , particularly in yeast . This approach has been challenging to apply to cultured mammalian cells , owing to the difficulty in generating bi-allelic mutations in diploid cells . This problem has recently been circumvented with the demonstration that the near-haploid KBM7 cell line can be used to perform genetic screens in cultured human cells [17] . KBM7 cells were originally isolated from a patient with chronic myeloid leukaemia [18] and are haploid apart from disomy of chromosome 8 and the sex chromosomes [19] . Insertional mutagenesis of these cells with a gene-trap retrovirus generates a library of knockout cells [20] , which can then be screened for individual mutants defective in the cellular process under investigation . Thus far this technique has been applied principally to lethality-based screens to study the mechanism of action of cytotoxic drugs , bacterial toxins and viruses which kill KBM7 cells [17] , [20]–[26] . In a proof-of-concept experiment , we recently showed that non-lethal haploid genetic screens could be successfully performed in KBM7 cells . The phenotypic enrichment of mutagenized KBM7 haploid cells displaying an altered cell surface phenotype by fluorescence-activated cell sorting ( FACS ) allowed us to identify known components of the MHC-I antigen presentation pathway [27] . Here we extend this approach by showing that non-lethal genetic screens in KBM7 cells can also be performed using integrated transgenes and allowed us to identify a function for the uncharacterised proteolipid protein 2 ( PLP2 ) . We show an absolute requirement for PLP2 to mediate the export of the K3 and K5 viral E3 ligases out of the endoplasmic reticulum , and consequently to allow their ubiquitinating activity . Furthermore , using our recently developed proteomic technique ‘plasma membrane profiling’ [28] , we identify an additional 74 targets of K5 , all of which are likely to be dependent on PLP2 . This ability to selectively enrich mutant cells by cell sorting together with the use of reporter constructs enables non-lethal haploid genetic screens to be used to identify the genetic components of essentially any cellular process .
To identify novel host factors required for the function of the KSHV-encoded E3 ubiquitin ligase K5 , we performed a forward genetic screen in near-haploid KBM7 cells . The rationale for this approach was our prediction that inactivation of gene ( s ) essential for K5 function in K5-KBM7 cells would rescue the surface expression of K5 target proteins to wild-type levels ( Figure 1A ) . Lentiviral expression of K5 in KBM7 cells decreased cell surface expression of known K5 targets , including MHC-I , B7-2 , ICAM-1 , PECAM and tetherin ( Figure 1B ) . We chose to screen on the most downregulated cell surface target B7-2 , which was barely detectable on the surface of K5-expressing KBM7 cells ( Figure 1B ) . Mutagenesis of K5-KBM7 cells with gene-trap retrovirus ( Figure S1 ) generated rare B7-2high cells , which were selectively enriched from the bulk population of B7-2low cells via two rounds of FACS for high surface B7-2 expression ( Figure 1C ) . Following the second sort a relatively pure population of B7-2high cells was established , from which genomic DNA was extracted . Retroviral integration sites were then mapped using splinkerette-PCR followed by 454 pyrosequencing [29] , which revealed 14 independent retroviral insertions on the X chromosome in the gene encoding proteolipid protein 2 ( PLP2 , also known as protein A4 ) ( Figure 1D and Figure S2 ) . No other clusters of retroviral insertion sites were found elsewhere in the genome , and as such PLP2 represents the only bona fide hit from the genetic screen . A major advantage of this technique is that human somatic cell knockouts of the gene of interest are generated during the selection process . Single cell cloning of the enriched B7-2high population identified a PLP2 gene-trap cell ( K5-PLP2GT ) , which expressed no detectable PLP2 protein by immunoblot , but , importantly , did still express K5 protein ( Figure 2A ) . To demonstrate that the inactivation of PLP2 was indeed responsible for the loss of K5 activity , we showed that exogenous expression of PLP2 in K5-PLP2GT cells restored the K5-mediated downregulation of B7-2 ( Figure 2B ) , proving that PLP2 is necessary for K5 function in KBM7 cells . PLP2 was also required for the downregulation of other known K5 targets , as the surface expression of MHC-I , ICAM-1 and PECAM were all at wild-type levels in K5-PLP2GT cells; in each case , re-expression of PLP2 restored the K5-mediated downregulation ( Figure 2C ) . Similar results were obtained using four other independent K5-PLP2GT clones . To demonstrate that PLP2 was required for K5 function in a cell type other than KBM7 , we found that K5 was unable to decrease cell surface MHC-I expression in hepatocellular carcinoma HepG2 cells , which do not express PLP2 [30] ( Figure 2D ) . Co-expression of PLP2 restored K5 function in these cells , resulting in a decrease in cell surface MHC-I expression ( Figure 2E ) . The requirement for PLP2 for K5 function is therefore independent of cell type . Given the essential role of PLP2 in K5 function , we wanted to determine whether it was also required for K3 , and the cellular orthologues of K3 and K5 , the membrane-associated RING-CH ( MARCH ) family of proteins [31] . This is best examined in a PLP2 knockout cell in the absence of K5 , and therefore required excision of the K5 transgene from the genome of K5-PLP2GT cells . To generate this cell line , we repeated the phenotypic screen , but used a K5 expression construct flanked by rox sites ( Figure 3A ) . Analogous to the cre/loxP system , dre recombinase catalyses the excision of DNA flanked with rox sites [32] ( Figure S3 ) . The repeat screen again generated PLP2 gene-trap cells . Following dre treatment we established a clonal line ( PLP2GT ) from which K5 had been excised , as shown by the lack of K5 protein expression by immunoblot ( Figure 3B ) and the lack of any downregulation of the K5 target MHC-I upon re-expression of PLP2 ( Figure 3C ) . K3 was unable to decrease MHC-I levels when expressed in PLP2GT cells , but its activity was restored upon re-expression of PLP2 ( Figure 3C ) . A similar effect was also seen in HepG2 cells ( Figure 3D ) . However , MARCH1 , MARCH8 and MARCH9 all efficiently downregulated their cell surface targets in PLP2GT cells ( Figure S4 ) , suggesting that PLP2 is not an essential factor for the MARCH proteins . K3 and K5 are responsible for the downregulation of MHC-I from the cell surface during KSHV infection [33] . Given that PLP2 is essential for the function of both K3 and K5 when the viral ligases are expressed individually in cultured cells , we would predict that PLP2 expression would also be critical for MHC-I downregulation by K3 and K5 in KSHV-infected cells . To test this hypothesis we used short-hairpin RNA ( shRNA ) lentiviral vectors to knockdown PLP2 expression in BC-3 cells , which are latently infected with KSHV . We designed four independent shRNA vectors against PLP2 and tested their efficacy to knockdown PLP2 expression by immunoblot ( Figure S5A ) , and by their ability to inhibit the K5-mediated downregulation of MHC-I in the monocytic THP-1 cell line ( Figure S5B ) . The two leading hairpins ( sh1-PLP2 and sh4-PLP2 ) , together with a non-targeting hairpin ( shControl ) were used to transduce BC-3 cells , and knockdown of PLP2 was confirmed by immunoblot ( Figure 3E ) . PLP2 knockdown by itself did not significantly affect cell surface MHC-I levels ( Figure 3F ) . Lytic reactivation of KSHV in BC-3 cells was then induced using sodium butyrate , and cell surface levels of MHC-I measured by flow cytometry ( Figure 3G ) . In wild-type and shControl-transduced BC-3 cells , butyrate treatment resulted in a decrease in MHC-I levels . In contrast , MHC-I was not downregulated in either shPLP2-transduced knockdown cell line; indeed MHC-I levels actually increased , probably as a result of an interferon effect that could no longer be overcome by K3 and K5 . Therefore , PLP2 is essential for the K3- and K5-mediated downregulation of MHC-I in KSHV-infected cells . PLP2 is a small 17 kDa integral membrane protein which on bioinformatic analysis contains a MARVEL ( MAL and related proteins for vesicle trafficking and membrane link ) domain , characterised by an M-shaped four-membrane spanning architecture with both N- and C-terminal tails in the cytoplasm [34] ( Figure 4A ) . Although the precise function of the MARVEL domain remains unclear , it is found in the myelin and lymphocyte protein ( MAL ) , physin , gyrin and occludin protein families , some of which are implicated in vesicle trafficking and membrane apposition events [34] . To understand the requirement for PLP2 in K3 and K5 function , we determined whether PLP2 was required for the ubiquitinating activity of these viral ligases . The robust ubiquitination of MHC-I induced by both K3 and K5 in wild-type KBM7 cells , as assessed by either immunoblot or radiolabelling , was totally abrogated in the PLP2 knockout cell ( Figure 4B–C ) . Therefore , PLP2 is required for the ubiquitination of K3 and K5 substrates . K3 and K5 ubiquitinate their substrates in the late secretory pathway [13] , [14] . The loss of ubiquitinating activity in the PLP2-knockout KBM7 cells , together with the described trafficking function of MARVEL domain-containing proteins , suggested a role for PLP2 in trafficking the viral ligases and/or their substrates to the late secretory compartment where ubiquitination can occur . However the vast majority of K3 and K5 localise to the ER [2] , [35] , preventing detection of any trafficking defect by microscopy . We therefore took a biochemical approach and visualised trafficking of the viral ligases through the secretory pathway using Endoglycosidase H ( EndoH ) sensitivity . Although K3 and K5 are not themselves glycosylated ( and artificial insertion of an N-linked glycan abrogates their function , unpublished data ) , the K3-associated MHC-I heavy chains do contain an N-linked glycan which permits this approach . The K3-associated MHC-I which remain EndoH sensitive are localised to the ER , while K3-associated MHC-I which acquire EndoH resistance have exited the ER and trafficked beyond the medial-Golgi . By [35S]-methionine radiolabeling and pulse-chase analysis of K3-KBM7 cells in the presence and absence of PLP2 ( Figure 5A ) we found that K3 associated with EndoH-sensitive MHC-I at the 0 time point , irrespective of PLP2 expression ( Figure 5B ) . After 35 minutes , in the presence of PLP2 , almost all the K3-associated MHC-I had acquired EndoH resistance , implying that the K3•MHC-I complex had trafficked beyond the medial-Golgi ( Figure 5B , lane 4 ) . In contrast , in the PLP2-deficient cells , the K3-associated MHC-I remained exclusively EndoH-sensitive ( Figure 5B , lane 8 ) . The loss in signal of K3-associated MHC-I in the PLP2-deficient cells is probably due to the MHC-I escaping K3 and trafficking to the cell surface . These data show an essential requirement for PLP2 in the export of the K3•MHC-I complex from the ER . Importantly , these results cannot be explained by the lack of PLP2 affecting the normal maturation of MHC-I , as we observed no difference in the egress of MHC-I not bound by K3 in the presence or absence of PLP2 ( Figure 5C ) . Furthermore , the requirement for PLP2 for the normal trafficking of the K3•MHC-I complex suggests that PLP2 likely interacts with K3 in the ER and facilitates its export . In support of this model , we could readily detect PLP2 bound to K3 or K5 following immunoprecipitation of FLAG-tagged K3 or K5 from HeLa cells ( Figure 5D ) . To determine the subcellular localisation of PLP2 , we initially confirmed detection of endogenous PLP2 by immunofluorescence ( Figure S6A ) . In HeLa cells , endogenous PLP2 concentrated primarily in a perinuclear compartment and in tubules emanating from that compartment , and expression of PLP2 tagged with mCherry at its N-terminus ( mCherry-PLP2 ) assumed a similar distribution ( Figure S6B ) . This compartment co-stained with markers of the recycling endosome , as seen with Rab11-GFP , internalised CD59 ( a GPI-anchored protein that enters the cell by clathrin-independent endocytic pathways ) [36] , and , to some extent , the transferrin receptor ( Figure S6C ) . A similar localisation of PLP2 was also observed in KBM7 cells ( Figure S6D ) . Therefore PLP2 is an integral membrane protein which , following its insertion into the ER and traffic through the secretory pathway , concentrates in recycling endosomes . If PLP2 is indeed required for the export of K3 and K5 from the ER , an ER-retained PLP2 mutant should not support K5 function . We therefore inserted a C-terminal di-lysine KKAA ER retention/retrieval motif [37] onto the C-terminus of PLP2 , which resulted in its redistribution from tubular recycling endosomes to the ER in HeLa cells , as confirmed by co-localisation with the ER marker calnexin ( Figure 6A ) . In K5-PLP2GT cells , the PLP2KKAA mutant still bound K5 ( Figure 6B ) , but could not compensate for the lack of PLP2 and was unable to rescue the K5-mediated downregulation of B7-2 ( Figure 6C ) . This was not merely a consequence of adding extra residues onto the C-terminal tail of PLP2 , as both PLP2KDEL ( an ER retention signal for soluble proteins of the ER lumen ) [38] and PLP2KDAS rescued the K5-mediated downregulation of B7-2 in K5-PLP2GT cells ( Figure 6C ) . Taken together , these data support a model whereby PLP2 binds K3 and K5 in the ER and is required for their export into the late secretory pathway where the viral ligases can ubiquitinate their substrates . To assess what proportion of K5 targets were dependent on PLP2 expression for their K5-mediated downregulation , we performed ‘Plasma Membrane Profiling’ [28] , [39] , [40] , a recently-developed proteomic technique which compares the relative abundance of cell surface proteins between different cell types . WT KBM7 , K5-KBM7 and K5-PLP2GT cells were grown in stable isotope labelling by amino acids in cell culture ( SILAC ) media and after selective enrichment for plasma membrane proteins , plasma membrane protein expression was quantified by mass spectrometry ( Figure 7A ) . Comparison of the plasma membrane proteome of WT KBM7 versus K5-KBM7 cells identified cell surface proteins downregulated by K5 , while comparison of K5-KBM7 cells versus K5-PLP2GT cells identified target proteins dependent on PLP2 for their K5-mediated downregulation ( Table S1 , S2 , S3 ) . We quantified 470 plasma membrane proteins , of which 83 were downregulated >3-fold from the plasma membrane upon expression of K5 ( Figure 7B , Figure S7 and Table S1 ) . Of these 83 target proteins , 73 ( 88% ) were clearly dependent on PLP2 for K5-mediated downregulation ( Table S2 ) . These included the known K5 targets , including MHC-I , B7-2 , ICAM-1 , PECAM , ALCAM and the IFN-γ receptor , thereby validating the dataset ( Table S3 ) . However , the majority of the downregulated proteins ( 74/83 , 89% ) had not previously been identified as K5 targets . In particular , multiple members of three families of proteins were found to be downregulated by K5: receptor-type tyrosine phosphatases ( PTPRA , PTPRF , PTPRK , PTPRM and PTPRS ) , ephrin receptors and their ligands ( EPHA2 , EPHB3 , EPHB4; EFNB1 , EFNB2 and EFNB3 ) and plexin receptors and their semaphorin ligands ( PLXNA1 , PLXNB1 , PLXNC1 , PLXND1; SEMA-4G ) . Four of the novel targets ( CD32 , CD33 , CD99 and EPHB4 ) were confirmed by flow cytometry ( Figure 7C ) , and one additional target , Plexin-A1 , was confirmed by immunoblot ( Figure 7D ) . The mass spectrometry dataset additionally suggested a small subset of K5 targets whose degradation might be PLP2 independent ( Table S2 ) , and these were investigated further ( Figure S8 ) . Myelin protein zero-like protein 2 ( MPZL2 , also called epithelial V-like antigen ) was most strongly downregulated by K5 , in a seemingly PLP2-independent manner . Immunoblotting confirmed that MPZL2 was degraded in the presence of K5 and was not detected in K5-PLP2GT cells ( Figure S8A ) . However , MPZL2 protein was readily detected in a different K5-PLP2GT clone , suggesting that its K5-mediated degradation was likely to be PLP2-dependent ( Figure S8B ) . Two other seemingly PLP2-independent K5 targets , the Mast/stem cell growth factor receptor Kit and the interleukin-9 receptor were confirmed as K5 targets , but were also PLP2-dependent in a second , independent K5-PLP2GT clone ( Figure S8C ) . Therefore the identification of a small number of PLP2-independent K5 targets likely represent a peculiarity of the particular K5-PLP2GT clone used , and was not generalizable . This emphasises the importance of checking the phenotypes of several clones when analysing mutants derived from KBM7 screens . We conclude that PLP2 is likely to be required for the downregulation of all K5 targets .
Non-lethal haploid genetic provide a novel approach to identify genes required for the downregulation of plasma membrane proteins by the viral E3 ubiquitin ligase K5 . Our screens revealed an absolute requirement for the poorly-characterised protein PLP2 in the K5-mediated downregulation of B7-2 and other K5 targets . Subsequent biochemical analysis demonstrated that a lack of PLP2 prevented K3 and K5 from ubiquitinating their substrates by impairing the export of the viral ligases from the ER . By using plasma membrane profiling to compare the relative abundance of cell surface proteins in the presence and absence of K5 and PLP2 we identified 74 novel plasma membrane protein targets of K5 . Almost all of these were PLP2 dependent , confirming a critical role for this protein in the downregulation of K5 target proteins . This is the first description of the mutagenesis of near-haploid KBM7 cells coupled with phenotypic selection by FACS to identify a novel gene in a cellular process . Phenotypic enrichment by cell sorting to select rare genetic mutants provides a complementary approach to the previously described live/dead screens using KBM7 cells . We demonstrate here that haploid screens can be successfully performed using stably-integrated transgenes , and , although we used the viral gene K5 , this approach is equally applicable to a fluorescent reporter . The ability to design screens based on altered expression of a genetically-encoded reporter vastly increases the number of cellular processes that can be examined using this approach . However this technique also has its limitations . Additional genes known to be required for K5 activity besides PLP2 were not identified . For example , RNAi-mediated depletion of the E2 ubiquitin-conjugating enzymes UbcH5 and Ubc13 abrogates ubiquitination by K5 [14] , but these genes were not identified in our haploid screens . Given that the retroviral gene-trap vector is able to integrate in essentially all expressed genes [20] , our inability to isolate mutations in the genes encoding UbcH5 and Ubc13 is likely to reflect their requirement for normal cell growth . Indeed , mutant cells need to both survive and maintain normal growth characteristics for around three weeks to emerge from the two-stage cytometry selection . Mutations that result in either cell death or a growth disadvantage will likely be lost from the final pool of selected cells . This work describes a function for PLP2 , a protein which was originally identified as enriched in colonic epithelial cells , where it was suggested to multimerise and assume characteristics of an ion channel [41] . PLP2 binds the ER-resident protein Bap31 [30] and the chemokine receptor CCR1 [42] , and decreased PLP2 expression has been implicated in X-linked mental retardation [43] and melanoma metastasis [44] . However , no function is ascribed to PLP2 . Our demonstration that PLP2 is required for the traffic of the K3•MHC-I complex out of the ER supports the proposed role of MARVEL domain-containing proteins in regulating membrane trafficking events [34] . The few well-characterised proteins of the MARVEL family associate with specialised cholesterol-rich membrane microdomains and regulate membrane apposition events . For example in T cells MAL is required to traffic Lck to the immunological synapse [45] , while synaptophysin is involved in regulating synaptic vesicle endocytosis in neurons [46] . Why might K3 and K5 need PLP2 to escape from the ER ? Such protein-assisted export has been reported for a number of other proteins . The proteolytically-inactive rhomboid iRhom2 mediates the export of TACE from the ER [47] , [48] , and the nucleotide-sensing toll-like receptors TLR7 and TLR9 require UNC93B1 for traffic from the ER to endolysosomes [49] . Examples can also be found in yeast , where the gene product Shr3p , notably also a small four transmembrane-spanning integral membrane protein that shares the same topology as PLP2 , mediates the recruitment of amino acid permeases into transport vesicles for ER export [50] . The reported association between PLP2 and Bap31 is of special interest in this regard . Bap31 is implicated in both the degradation [30] , [51] and export of proteins from the ER [52] , [53] , suggesting a key role as a quality control factor in protein triage in the ER . Furthermore , Bap31 binds K3 and K5 [54] , suggesting that Bap31 may sequester the viral ligases in the ER such that PLP2 is required to extract them from Bap31 to facilitate export . Alternatively PLP2 may serve as a recruitment factor for the packaging of the trimeric Bap31•PLP2•K5 complex into ER exit sites . Given that we identify PLP2 in our screens rather than Bap31 , the former is perhaps more likely . Of note , all three proteins are unusual in containing charged residues within their transmembrane domains , which may mediate protein associations within the hydrophobic environment of the ER membrane . Clearly PLP2 has not evolved to facilitate the trafficking of viral ligases , and its endogenous role remains unclear . Cellular proteins may also require PLP2 to exit the ER , and we are undertaking further studies to identify such PLP2 client proteins . It was surprising that none of the MARCH proteins , the cellular orthologues of K3 and K5 , were affected by the absence of PLP2 ( Figure S4 ) . Our previous work found PLP2 to be upregulated on the cell surface upon overexpression of MARCH9 in Sultan B cells [55] , and more recently many of the MARCH proteins were also shown to bind Bap31 [54] . PLP2 may still regulate MARCH function , but another protein substitutes for PLP2 to allow MARCH-mediated downregulation of cell surface targets . Alternatively the viral ligases may exploit PLP2 to facilitate their trafficking independently of the normal endogenous role of PLP2 . While our experience would suggest that viral gene products are more likely to appropriate endogenous function of cellular genes , rather than invent new ones , the steady-state localisation of PLP2 to recycling endosomes ( Figure S6 ) suggests additional cellular functions for PLP2 in the endosomal system . Indeed , a recent proteomic experiment identified PLP2 as a component of clathrin-coated vesicles [56] . Our proteomic dataset underscores the remarkable ability of K5 to specifically target a wide variety of structurally diverse plasma membrane proteins for degradation . In this regard , KSHV has adopted an alternative strategy to other herpesviruses . Human cytomegalovirus , for example , encodes multiple genes which each downregulate a few cell surface ligands [57] , while in the case of KSHV a single gene appears to perform this function . The selective advantage to the virus of downregulating ligands for cytotoxic T cells ( MHC-I ) , costimulatory molecules ( B7-2 ) or NK cell ligands ( MICA and MICB ) is readily explained . However , this is less clear for some of the novel K5 targets such as the ephrin signalling pathway , implicated in normal thymocyte maturation and T cell modulation [58] , or semaphorin/plexin signalling , involved in immune cell migration [59] . We suggest additional , as yet to be identified , roles for these cell surface proteins and it will be interesting to see which of these new substrates are also targeted by other viral proteins . We envisage a number of potential mechanisms by which K5 targets so many different receptors: K5 could interact directly with each of its targets , K5 could bind a common adaptor protein which provides a platform from which it can ubiquitinate its targets , or K5 could be recruited to a specialised membrane microdomain where it ubiquitinates its targets . The general requirement for PLP2 suggests that K5 uses at least a broadly similar mechanism to downregulate its targets . Given the propensity for the partitioning of MARVEL domain proteins into specialised membrane microdomains , we speculate that the role of PLP2 is not simply to export K5 from the ER , but also to traffic it to specific site ( s ) for target ubiquitination . Identifying common features shared between the large number of K5 targets may shed further light on this important issue . In summary , we have found an essential role for PLP2 in facilitating the ubiquitination and degradation of cell surface immunoreceptors by the viral E3 ubiquitin ligases K3 and K5 . In addition we identified 74 novel plasma membrane targets of K5 , all of which are likely to be PLP2-dependent; PLP2 is therefore a critical host factor for KSHV immune evasion . Further work will be required to elucidate the normal cellular function of PLP2 and to investigate the potential for interfering with PLP2 function to modulate the pathogenesis of KSHV infection .
KBM7 cells , THP-1 cells and BC-3 cells were maintained in IMDM supplemented with 10% fetal calf serum and penicillin/streptomycin . HeLa cells and HepG2 cells were grown in RPMI 1640 plus 10% fetal calf serum and penicillin/streptomycin . For reactivation of latent KSHV , BC-3 cells were treated with 2 mM sodium butyrate for 24 hours . For SILAC analysis , KBM7 cells were grown in SILAC RPMI 1640 ( Thermo Pierce ) , 10% dialysed FBS ( JRH Biosciences ) , and penicillin/streptomycin . SILAC media was supplemented with either light ( Arg 0 , Lys 0 , Sigma ) , medium ( Arg 6 , Lys 4 , Cambridge Isotope Laboratories ) or heavy ( Arg 10 , Lys 8 , Cambridge Isotope Laboratories ) amino acids at 50 mg/l and L-proline at 280 mg/l . Incorporation of heavy label was >98% for both arginine and lysine-containing peptides . The PLP2 expression vector pcDNA3 . 1-PLP2myc/his was a kind gift from Prof . Jiyoung Kim ( Kyung Hee University , Korea ) [42] , the lentiviral expression vector pHRSIN-UCOE-EGFP a kind gift from Prof . Adrian Thrasher ( University College London , UK ) [45] , the dre recombinase expression vector pCAGGS-Dre a kind gift from Prof . Francis Stewart ( TU Dreseden , Germany ) ( Anastassiadis et al . , 2009 ) , and the lentirviral shRNA expression vector pHR-SIREN a kind gift from Prof . Greg Towers ( University College London , UK ) . Primary antibodies used were as follows: mAb W6/32 ( recognises conformational MHC-I ) , mAb HC10 ( anti-MHC-I heavy chain ) , mAb 3B10 . 7 ( anti-MHC-I heavy chain ) , mouse α-B7-2 ( BU63 ) , mouse α-ICAM1-APC ( BD ) , mouse α-PECAM-APC ( BD ) , mouse α-tetherin ( a kind gift from G . Towers , University College London ) , mouse α-FLAG M2 ( Sigma ) , rabbit α-calreticulin ( Thermo ) , mouse α-calnexin ( mAb AF8 , a kind gift from M . Brenner , Harvard Medical School , USA ) , mouse α-SLAM ( Biolegend ) , mouse α-ubiquitin ( VU-1 , Lifesensors ) , mouse α-CD71 ( Santa Cruz ) , mouse α-CD59 ( Santa Cruz ) , mouse α-Rab11 ( Abcam ) , mouse α-CD32 ( BD ) , mouse α-CD33 ( Biolegend ) , mouse α-CD99 ( Abcam ) , rabbit α-EPHB4 ( R&D Systems ) and rabbit α-MPZL2 ( Proteintech ) . The mouse monoclonal antibody against K5 was a gift from Prof . Klaus Früh ( Oregon Health and Science University , USA ) and the rabbit polyclonal antibody against PLP2 was a kind gift from Prof . Gordon Shore ( McGill University , Canada ) [30] . Fluorophore-conjugated secondary antibodies were obtained from Molecular Probes , and HRP-conjugated secondary antibodies were obtained from Jackson ImmunoResearch . K5 was expressed from a modified version of the lentiviral expression vector pHRSIN-UCOE-EGFP [45] . A rox site was inserted into the EcoRI site upstream of the UCOE element , the EGFP replaced with K5 with an N-terminal FLAG tag , and a puromycin resistance cassette and additional rox site inserted downstream . Otherwise the lentiviral expression vectors pHRSIN-PSFFV-GFP-PPGK-Puro and pHRSIN-PSFFV-GFP-PPGK-Hygro were used , with the gene of interest cloned in place of GFP as a BamHI-NotI fragment . Lentivirus was produced by transfecting 293ET cells with the lentiviral vector plus the packaging plasmids pCMVΔR8 . 91 and pMD . G using TransIT-293 ( Mirus ) according to the manufacutrer's instructions . The viral supernatant was collected 48 h and 72 h later and target cells transduced by spin infection at 1800 rpm for 45 min . For shRNA-mediated knockdown of PLP2 expression , hairpin oligonucleotides were annealed , cloned into the pHR-SIREN lentiviral vector cut with BamHI and EcoRI , and sequence verified . Lentivirus was then made as above in 293ET cells and used to transduce BC-3 cells . Hairpins were designed using Clontech's RNAi target sequence selector; the forward oligonucleotide sequences used were: sh1-PLP2: 5′-GAT CCG TGG TGA TCC TGA TCT GCT TTT CAA GAG AAA GCA GAT CAG GAT CAC CAT TTT TTG -3′ sh2-PLP2: 5′-GAT CCG CGG TGA TTG AGA TGA TCC TTT CAA GAG AAG GAT CAT CTC AAT CAC CGT TTT TTG-3′ sh3-PLP2: 5′-GAT CCG TGC ACA CCA AGA TAC CAT TTT CAA GAG AAA TGG TAT CTT GGT GTG CAT TTT TTG-3′ sh4-PLP2: 5′-GAT CCG CGG CAA TCC TCT ACC TGA TTT CAA GAG AAT CAG GTA GAG GAT TGC CGT TTT TTG-3′ shControl: 5′-GAT CCG TTA TAG GCT CGC AAA AGG TTC AAG AGA CCT TTT GCG AGC CTA TAA CTT TTT TG-3′ The haploid genetic screen was carried out as described [27] . Briefly , 5×107 near-haploid K5-KBM7 cells were mutagenised with gene-trap retrovirus , grown for 7 days , and then sorted by FACS for high B7-2 expression . A further sort to purify the B7-2high population was carried out a further 7 days later , before genomic DNA was extracted and the retroviral integration sites mapped by splinkerette-PCR and 454 pyrosequencing [27] , [29] . Cells were washed with PBS and incubated with primary antibody for 30 min at 4°C , washed once with PBS , and then incubated with fluorophore-conjugated secondary antibody for 30 min at 4°C . Following fixation with 4% PFA , samples were analysed on a FACSCalibur ( BD ) . FACS was carried on an Influx ( BD ) cell sorter . HeLa cells were grown overnight on glass coverslips; KBM7 cells were allowed to adhere to coverslips in serum-free media . Cells were fixed with 4% PFA , permeabilised with 0 . 5% Triton X-100 and then blocked for 1 h with 4% BSA dissolved in PBS+0 . 1% Tween-20 ( PBS-T ) . Primary antibody was then applied for 1 h , the coverslips washed in PBS-T , and the fluorophore-conjugated secondary antibody applied for 45 min . Coverslips were mounted in Prolong anti-fade reagent ( Molecular Probes ) and imaged using a Nikon LSM510 laser scanning confocal microscope ( Zeiss ) . Images were processed using Adobe Photoshop ( Adobe , CA ) . Cells were lysed in 1% Triton X-100 or 1% Digitonin in TBS plus 10 mM iodoacetamide ( IAA ) and 0 . 5 mM phenylmethylsulfonyl fluoride ( PMSF ) for 30 minutes on ice . The postnuclear supernatants were heated to 70°C in SDS sample buffer for 10 minutes , separated by SDS-PAGE , and transferred to a PVDF membrane ( Millipore ) . Membranes were blocked in 5% milk in PBS+0 . 2% Tween-20 , probed with the indicated antibodies , and reactive bands visualised using West Pico ( Thermo Fisher Scientific ) . To visualise ubiquitinated MHC-I using the VU-1 antibody , membranes were first cross-linked with 0 . 5% glutaraldehyde and processed according to the manufacturer's instructions . For immunoprecipitation , lysates were pre-cleared with protein A and IgG-Sepharose and incubated with primary antibody and protein A-Sepharose for 2 h at 4°C . Following three washes in lysis buffer , samples were eluted in SDS sample buffer and processed as above . Cells were starved for 1 h in cysteine- and methionine- free media , labelled for the indicated time with [35S]-cysteine and [35S]-methionine ( Amersham ) and chased in media containing excess cold cysteine and methionine . Cells were then lysed in 1% digitonin plus IAA and PMSF and processed for immunoprecipitation as above . For re-precipitation , the washed primary immunoprecipitates were dissociated in 1% SDS for 1 h at 37°C , diluted 20-fold in 0 . 5% Triton X-100 , and then re-precipitated overnight with the appropriate antibody and protein A-Sepharose beads . Samples were then washed in lysis buffer , separated by SDS-PAGE , dried and analysed by autoradiography . For EndoH treatment , the samples were split into two , and EndoH ( NEB ) added to one of the samples following the manufacturer's instructions . Following incubation for 1 h at 37°C , samples were analysed by SDS-PAGE and autoradiography . PMP was performed as described previously [28] , [40] . Briefly , 2 . 5×108 of each SILAC-labelled cell type were pooled in a 1∶1∶1 ratio . Surface sialic acid residues were oxidized with sodium meta-periodate ( Thermo ) then biotinylated with aminooxy-biotin ( Biotium ) . The reaction was quenched , and the biotinylated cells incubated in a 1% Triton X-100 lysis buffer . Biotinylated glycoproteins were enriched with high affinity streptavidin agarose beads ( Pierce ) and washed extensively . Captured protein was denatured with DTT , alkylated with iodoacetamide ( IAA , Sigma ) and digested with trypsin ( Promega ) on-bead overnight . Tryptic peptides were collected and fractionated ( described below ) . Glycopeptides were eluted using PNGase ( New England Biolabs ) . A total of 30 µg of tryptic peptide was subjected to high pH reversed phase HPLC ( HpRP-HPLC ) fractionation into 124 fractions [40] . Fractions were combined into 62 samples prior to analysis with the PNGase fraction using a NanoAcquity uPLC ( Waters ) coupled to an LTQ-OrbiTrap XL ( Thermo ) . Mass spectrometric analysis , database searching and data processing were performed as described previously [40] . Briefly , fractionated tryptic peptides were separated using a gradient of 3 to 25% MeCN over 20 min and to 45% MeCN by 30 min . Unfractionated PNGase peptides were separated with a gradient of 3 to 25% MeCN over 60 min and to 45% MeCN by 80 min . PNGase samples were acquired in triplicate . MS data was acquired between 300 and 2000 m/z at 60 , 000 fwhm . CID spectra were acquired in the LTQ with MSMS switching operating in a top 6 DDA fashion . Raw MS files were processed using MaxQuant version 1 . 3 . 0 . 5 [60] H . sapiens Uniprot database ( downloaded 29/05/13 ) . Significance A values were calculated and Gene Ontology Cellular Compartment ( GOCC ) terms added using Perseus version 1 . 2 . 0 . 16 ( downloaded from http://maxquant . org ) . We assessed the number of PM proteins identified as described previously [40] . | Viruses manipulate the cellular machinery of the host to facilitate their replication and evade the host immune response . Kaposi's sarcoma-associated herpesvirus ( KSHV ) , a gammaherpesvirus linked to the development of Kaposi's sarcoma , encodes two viral E3 ubiquitin ligases K3 and K5 which target multiple cell surface immunoreceptors for destruction . Here we employ a novel genetic screen in the haploid human cell line KBM7 to identify cellular proteins required for K5 function . This revealed an essential role for the poorly characterised protein proteolipid protein 2 ( PLP2 ) ; K3 and K5 hijack PLP2 to facilitate their export out of the endoplasmic reticulum , which is necessary for ubiquitination and subsequent degradation of their substrates . Furthermore we identified many new cell surface receptors targeted by K5 , all of which are likely to be dependent on PLP2 . Therefore , PLP2 is likely to be a key host factor to allow KSHV immune evasion . Overall this work provides further insight into the function of this family of viral E3 ubiquitin ligases and paves the way for further study of the role of PLP2 in normal cellular function . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Haploid Genetic Screens Identify an Essential Role for PLP2 in the Downregulation of Novel Plasma Membrane Targets by Viral E3 Ubiquitin Ligases |
Heterozygous mutations in the PRPF31 gene cause autosomal dominant retinitis pigmentosa ( adRP ) , a hereditary disorder leading to progressive blindness . In some cases , such mutations display incomplete penetrance , implying that certain carriers develop retinal degeneration while others have no symptoms at all . Asymptomatic carriers are protected from the disease by a higher than average expression of the PRPF31 allele that is not mutated , mainly through the action of an unknown modifier gene mapping to chromosome 19q13 . 4 . We investigated a large family with adRP segregating an 11-bp deletion in PRPF31 . The analysis of cell lines derived from asymptomatic and affected individuals revealed that the expression of only one gene among a number of candidates within the 19q13 . 4 interval significantly correlated with that of PRPF31 , both at the mRNA and protein levels , and according to an inverse relationship . This gene was CNOT3 , encoding a subunit of the Ccr4-not transcription complex . In cultured cells , siRNA–mediated silencing of CNOT3 provoked an increase in PRPF31 expression , confirming a repressive nature of CNOT3 on PRPF31 . Furthermore , chromatin immunoprecipitation revealed that CNOT3 directly binds to a specific PRPF31 promoter sequence , while next-generation sequencing of the CNOT3 genomic region indicated that its variable expression is associated with a common intronic SNP . In conclusion , we identify CNOT3 as the main modifier gene determining penetrance of PRPF31 mutations , via a mechanism of transcriptional repression . In asymptomatic carriers CNOT3 is expressed at low levels , allowing higher amounts of wild-type PRPF31 transcripts to be produced and preventing manifestation of retinal degeneration .
The penetrance of a disease-causing mutation corresponds to the proportion of individuals who carry such variant and develop clinical symptoms . In the majority of Mendelian disorders penetrance is 100% , but incomplete penetrance is far from being uncommon [1] . Although in medical genetics penetrance is still largely uncharacterized at the molecular level , it is usually determined by genetic or epigenetic factors , and sometimes even by environmental modifiers [2] . Retinitis pigmentosa ( RP ) is a group of inherited degenerative diseases of the retina that cause the progressive death of photoreceptors , the neurons of the eye that are sensitive to light . Typically , patients affected by RP first suffer from night blindness , most often during adolescence . Rod and cone photoreceptor cells start to degenerate from the mid periphery to the far periphery and the center of the retina , resulting in the so-called tunnel vision . Later in life , central vision is also lost , leading to legal or complete blindness [3] . Clinically , RP is a highly-heterogeneous disease , reflecting not only genetic heterogeneity ( mutations in different genes ) , but also inter-individual diversity ( penetrance and expressivity ) [4] . The PRPF31 gene encodes in humans a pre-mRNA processing factor . In autosomal dominant RP ( adRP ) due to mutations in PRPF31 penetrance of the disease can be incomplete . Specifically , in families with PRPF31 mutations it is not uncommon to observe the presence of asymptomatic individuals who have affected parents , affected children , or both [5]–[8] . Although they carry the same PRPF31 mutation as their affected relatives , asymptomatic subjects show no visual impairment , even at older ages , and normal to slightly reduced electroretinographic recordings [7] . PRPF31 mutations causing adRP are largely null alleles , such as deletions , nonsenses , or DNA changes leading to premature termination codons and to mRNA degradation [9]–[14] . Patients are therefore hemizygotes for PRPF31 , suggesting that the molecular pathophysiology of the disease is due to the functional loss of one allele and to haploinsufficiency [10] , [12] , [15] . The ubiquitous expression of PRPF31 has allowed a number of functional studies to be performed in immortalized lymphoblastoid cell lines ( LCLs ) from patients and asymptomatic carriers of mutations [16]–[18] . In particular , it has been shown that penetrance of mutations is due to the differential expression of the PRPF31 allele that is not inactivated by mutations , in both symptomatic and asymptomatic individuals . Unlike affected persons , asymptomatic carriers naturally express high amounts of functional PRPF31 mRNA , a phenomenon that compensates for the mutation-induced loss of one allele and prevents manifestation of symptoms [16]–[18] . This variable expression of PRPF31 seems to be present within the general population [16] and therefore asymptomatic carriers of mutations would be individuals that by chance are “high expressors” . Furthermore , protection from PRPF31 mutations ( and therefore variable PRPF31 expression ) is itself an inheritable character [16] , [19] . In an elegant meta-analytic study , McGee et al . [19] have shown that protective alleles , named isoalleles , are inherited by carriers of PRPF31 mutations from the parent who does not transmit the mutation ( i . e . they are in trans with respect to the mutation ) . Furthermore , such isoalleles would be responsible for the majority of incomplete penetrance cases , and map to chromosome 19q13 . 4 , in proximity to PRPF31 itself [19] . The same study also indicated that these isoalleles were not the only modulators of PRPF31 penetrance , since some individuals with discordant phenotypes carried an identical wild-type haplotype for the isoalleles on chromosome 19 . Another genetic element potentially capable of influencing the penetrance of PRPF31 mutation was later mapped to chromosome 14q21–23 [16] . In this study , we search for and identify the major modifier gene responsible for penetrance of PRPF31 mutations , through the analysis of LCLs from a very large family with adRP due to a PRPF31 microdeletion [6] , [20] .
The region on chromosome 19q13 . 4 harboring the main modifier gene for PRPF31 penetrance was determined by McGee et al . to lie between microsatellite markers D19S572 and D19S926 [19] . This interval contains 118 genes , including 50 protein-coding genes , 50 miRNAs and 18 pseudogenes . Based on data from lymphoblast studies describing the nature and the possible mechanism of action of the penetrance modifier gene [16]–[18] , we selected protein-coding genes that were consistently expressed in LCLs , as detected by q-PCR ( 18 genes ) . We also excluded some of the genes that in this region belong to the leukocyte receptor cluster ( LRC ) and are implicated exclusively in leukocyte functions . We were left with 10 sequences , namely: NDUFA3 , TFPT , CNOT3 , LENG1 , MBOAT7 , TSEN34 , RPS9 , LILRB3 , ILT7 , and NALP2 . We then measured by q-PCR the mRNA expression levels of these genes in LCLs from 4 asymptomatic and 6 affected individuals from the RP856/AD5 family ( Table S1 and Figure S1 ) . All genes showed consistent expression across the family members . Of these , only CNOT3 showed a statistically significant difference in mRNA expression between the two groups of individuals ( p<0 . 01 ) ( Figure 1 and Figure S1 ) . Unexpectedly , CNOT3 trend of expression was the opposite to that of PRPF31 , as it showed lower expression in asymptomatic than in the affected carriers of PRPF31 mutations ( Figure 1B ) . This phenomenon was particularly clear when expression of CNOT3 and PRPF31 were paired by cell lines and the relevant regression lines calculated ( Figure 1C ) . Assessment of CNOT3 protein by quantitative western blotting confirmed the differential expression detected by q-PCR ( Figure 1D ) . CNOT3 belongs to the Ccr4-Not complex , a conserved multi-protein structure involved in the regulation of gene expression [21] . To investigate if CNOT3 could influence PRPF31 expression , we silenced its expression in ARPE-19 cell lines , by using two different siRNA sequences . Suppression of CNOT3 resulted in significant increase of PRPF31 mRNA and protein ( p<0 . 001 , Figure 2 ) . This effect was very specific , as no influence was observed in negative controls and in TFPT expression , a neighboring gene sharing part of the promoter with PRPF31 ( Figure S2 ) . CNOT3 can negatively regulate transcription by either directly binding to the promoter of target genes or by affecting their mRNA rate of degradation [22] , [23] . To understand which could be the mechanism through which CNOT3 modulates PRPF31 expression , we incubated LCLs from two asymptomatic-affected pairs with Actinomycin D , a drug that inhibits de novo transcription , and then measured the rate of decay of PRPF31 mRNA . No statistically significant difference was observed between the asymptomatic and affected individuals ( Figure S3 ) , suggesting that the modulation of PRPF31 expression happens most probably at the transcriptional level . To test this hypothesis , we performed a Chromatin ImmunoPrecipitation ( ChIP ) assay in LCLs from 3 healthy individuals , using an anti-CNOT3 antibody and serum IgG as a negative control . To confirm that CNOT3 enrichment of a target DNA region was due to a specific immunoprecipitation rather than to a random precipitation of DNA , we designed primers targeting genomic regions that were not supposed to be bound by CNOT3 . Primers targeting CNOT3 promoter were used as a positive control , since it has been previously shown that CNOT3 self-regulates its expression by binding to its own promoter [23] . Both qualitative and quantitative PCR showed a statistically significant enrichment in PRPF31 promoter sequences in DNA that was immunoprecipitated by the CNOT3 antibody , compared to that exposed to serum IgG ( Figure 3A , 3B ) . In order to identify genetic markers that could be associated with variable expression of CNOT3 and therefore with penetrance of PRPF31 mutations , we sequenced the entire CNOT3 genomic region by next-generation sequencing ( NGS ) in one asymptomatic-affected sibling pair . We identified five polymorphic variants ( rs36643 , rs56079424 , rs36661 , rs4806718 , rs1055234 ) that differed between the two subjects . These five variants were subsequently analyzed in a second asymptomatic-affected sibling pair from the same pedigree , showing that only alleles of rs4806718 , lying in intron 17 of CNOT3 , segregated with the trait . This SNP was then sequenced in a total of 38 asymptomatic and affected individuals from the RP856/AD5 family , as well as from an unrelated family for which the modifier gene for PRPF31 penetrance was also found to be linked to chromosome 19q13 . 4 [24] ( Figure 4 ) . Association between the C allele of rs4806718 with the affected status and the T allele with the asymptomatic status was moderately significant ( p = 0 . 04 , by Fisher exact test ) .
Despite penetrance being an old concept in genetics , little is known about its molecular causes , especially in inherited human diseases . Notable positive examples include dominant erythropoietic protoporphyria , caused by mutations in the FECH gene , and dominant elliptocytosis , due to mutations in SPTA1 . In these disorders , an imbalance of expression between the wild-type and the mutated alleles causes the manifestation of the symptoms [25]–[27] . Similar mechanisms determine penetrance of PRPF31 mutations , since asymptomatic carriers are individuals who display increased levels of wild-type mRNA alleles , which in turn compensate for the deficiency caused by the mutation [16]–[18] . However , unlike erythropoietic protoporphyria and elliptocytosis , in PRPF31-linked adRP the molecular causes of such beneficial hyper-expression have remained , up to now , unexplained . Previous mapping studies have shown that the penetrance and expression of PRPF31 is influenced by at least two loci: one , likely having a major effect , lies within the same chromosomal region as PRPF31 ( proximal modifier ) , the other is on chromosome 14 ( distant modifier ) [16] , [19] . Our previous work has also demonstrated that both modifiers would act through diffusible elements ( e . g . transcription factors ) since their effects on PRPF31 mRNA expression concerns equally both copies of the gene [16] . This observation probably explains the failure of previous attempts to identify the proximal modifier as a polymorphic variant of the PRPF31 sequence itself , according to the FECH or SPTA1 models . Based on this previous knowledge , we reasoned that the expression of the proximal modifier of PRPF31 mutations should correlate with that of PRPF31 . Therefore we started assessing mRNA levels of genes that reside within the mapped 19q13 . 4 interval , by using the same cellular model successfully used in previous studies of PRPF molecular genetics , and in particular of PRPF31 penetrance [10] , [15]–[18] , [28] , [29] . Specifically , we studied cells derived from members of one of the largest pedigrees known to segregate a PRPF31 mutation , family RP856/AD5 [6] , [20] , for which incomplete penetrance could also be , at least in part , determined by the proximal modifier [19] . Following a filtering process based on both in silico analyses and on mRNA expression , we were left with only 10 candidates . Of these , only one , CNOT3 , showed a pattern of expression that significantly correlated to that of PRPF31 . Interestingly , its trend of expression was inverse to that of PRPF31 , raising the possibility that CNOT3 may be a negative regulator of PRPF31 expression . CNOT3 encodes a protein that is part of the Ccr4-Not multi-subunit complex , an evolutionary conserved multimeric structure involved in modulation of gene expression [21] , [30]–[34] . Evidences that CNOT3 could be a negative regulator of transcription have been provided in yeast [31] , and then confirmed in human cell lines , by the identification of a conserved motif at its C-terminus , called the Not-Box . This motif was originally identified in another subunit of the complex , CNOT2 , where it was shown to repress reporter gene activity upon promoter targeting [35] . We confirmed the role of CNOT3 as a negative regulator of PRPF31 expression by siRNA-mediated silencing experiments in ARPE-19 cells . Specifically , we observed that 70% depletion of CNOT3 induced approximately a 2-fold increase in PRPF31 expression , but had no effects on TFPT , a gene that is contiguous to PRPF31 and shares with it part of the promoter [36] . CNOT3 can modulate transcription of its targets by the direct binding to their promoters [23] or by promoting the recruitment of deadenylases at the 3′ end of their transcripts [22] . Our data provide evidence showing that regulation of PRPF31 expression should be mainly at the transcriptional level . First , we observed that decay of PRPF31 mRNA was roughly the same in cells from individuals expressing different levels of CNOT3 , disfavoring gene modulation through post-transcriptional mechanisms . Second , we showed by ChIP that CNOT3 could bind directly to the bona fide PRPF31 promoter . In their work , McGee et al . identified the chromosomal interval containing the proximal modifier through linkage analysis , a technique that searches for relationships between phenotypes and physical elements on the DNA sequence [19] . This implies that variable expression of CNOT3 must be determined by a DNA variant that is present in this same region , possibly within CNOT3 itself . Given their supposedly high frequency within the general population , these isoalleles would very likely be polymorphic elements . Our search for CNOT3 DNA changes that would be present in asymptomatic but not in affected carriers of mutations ( or vice versa ) resulted in the identification of particular alleles of rs4806718 . Are these the isoalleles originally mapped by McGee et al . ? Although statistically significant , the association between rs4806718's C allele and disease ( and the T allele with an unaffected status ) was not perfect . This phenomenon can be explained by the presence of additional factors capable of determining PRPF31 penetrance , such as the one mapped on chromosome 14 [16] . These modifiers could interfere with or even mask the effects of rs4806718 alleles , ultimately allowing the “wrong” rs4806718 variant to be associated with either phenotype . Such a hypothesis is in perfect agreement with the original data on PRPF31 isoalleles , as a few discordant phenotype-genotype associations concerning the mapped locus for the proximal modifier were also clearly recognized . Amongst other examples , 2 siblings from the last generation of RP856/AD5 had discordant phenotypes but concordant haplotypes [19] , [37] . These same individuals , genotyped by us at the rs4806718 locus , were found indeed to share the same parental allele . Furthermore , if the modifier allele is truly inherited from the parent who does not transmit the mutation , then the chance that this does not forcibly correspond to an rs4806718 allele is relatively high in RP856/AD5 , given the number of spouses external to the family who are present in this pedigree . Another important element to consider is whether rs4806718 alleles have a direct effect on CNOT3 expression , or whether the two factors are simply in linkage disequilibrium with other elements ( e . g . transcription enhancers ) lying somewhere else in the region . According to in silico prediction tools , the rs4806718 C variant , which has a frequency of 0 . 38 in the European population , could affect CNOT3 splicing by decreasing the binding energy for one acceptor splice site . Therefore , at least potentially , rs4806718 alleles could represent the true PRPF31 isoalleles . Taken together , all our observations suggest that CNOT3 is the modifier gene on chromosome 19q13 . 4 that is responsible for penetrance of PRPF31 mutations . Through direct repression of PRPF31 transcription and in virtue of its own variable expression , CNOT3 would differentially reduce the amount of available PRPF31 mRNA , thus determining incomplete penetrance . Although further studies on the physiological role of CNOT3 in human cells and tissues are definitely needed , our data open the way for a possible treatment of PRPF31-linked RP through the inhibition of this transcriptional regulator .
This study involved 10 individuals from the British family RP856/AD5 , segregating an 11-bp deletion in exon 11 of PRPF31 ( c . 1115_1125del ) [6] , [20] . Our research has been conducted in accordance with the tenets of the Declaration of Helsinki and has been approved by the IRBs of our Institutions . Lymphoblastoid cell lines derived from peripheral blood leukocytes of each individual were either obtained from the Coriell Cell Repositories or through the immortalization of peripheral blood leukocytes . Cells were grown and maintained as previously described [18] . The human retinal pigment epithelial cell line ARPE-19 ( kindly provided by Dr . Yvan Arsenijevic ) was grown and maintained at 37°C with 5% CO2 in N1 medium ( DMEM/F12 complemented with 2 . 5 mM L-glutamine , 56 mM NaHCO3 , and 10% fetal bovine serum ) . Lymphoblasts were harvested during their exponential growth phase ( 500 , 000–1 , 000 , 000 cells/ml ) and RNA was isolated from 107 cells using the QIAGEN RNeasy Mini Kit , following the manufacturer's instructions . The only modification to the protocol concerned the DNase treatment , since we used double the amount of enzyme compared to the suggested quantity . RNA concentration was measured with the Dropsense 96 spectrophotometer ( Trinean ) . cDNA synthesis was carried out as previously described [10] . Most of the primer sequences used in this study were annotated in the qPrimerDepot database ( http://primerdepot . nci . nih . gov/ ) . These sequences are specifically designed to span exon-exon junctions , thus avoiding genomic DNA to be amplified during q-PCR . To design other primer sequences , which were not present in the qPrimerDepot database , we used the Primer Blast tool from NCBI ( http://www . ncbi . nlm . nih . gov/tools/primer-blast/ ) . To validate each primer pair for q-PCR we first optimized the primer amounts ( 50–200 nM ) , and then loaded 10 µl of the q-PCR product obtained on a 1% agarose gel , in order to check the specificity of the amplification product . Finally , a standard curve using a control cDNA template was used to test each primer pair's efficiency . We considered as acceptable ranges of efficiency between 90 and 110% , corresponding to standard curve slopes between −3 . 6 and −3 . 1 . All primer pairs used for this study are listed in Table S2 . For GAPDH and PRPF31 amplification we used primers and probes previously described [16] . All genes but PRPF31 and GAPDH were amplified with the Sybr Green PCR Master Mix ( Applied Biosystems ) . Q-PCR reactions were performed as published [16] . After having assessed that PCR efficiencies for all genes were comparable , mRNA expression of each of them was normalized with respect to GAPDH , using the ΔΔCt method . Total protein was extracted from lymphoblastoid cell lines in RIPA buffer as reported before [10] . ARPE-19 whole cell lysate was obtained by scraping the cells into 150 µl of lysis buffer ( 20 mM Tris HCl , pH 8 . 0 , 150 mM NaCl , 10% glycerol , 2 mM EDTA , 1% TritonX-100 ) complemented with protease and phosphatase inhibitors , and incubated on ice for 15 minutes followed by a centrifugation at 14 . 000 rpm for 30 minutes at 4°C . Proteins concentration was measured with the BCA protein assay kit ( Pierce ) , using BSA to generate a standard curve . Anti-PRPF31 antibody was raised in rabbit as previously described [10] . Rabbit anti-CNOT3 antibody was purchased by Bethyl Laboratories . This targets residues 525 to 575 of the human CNOT3 protein ( NP_055331 . 1 ) , allowing detection of a 117-kDa protein . Mouse anti-β-actin antibody ( Sigma ) was used as a loading control . Equal amounts of proteins were loaded and run on an 8% SDS-PAGE gel . Proteins were transferred to a nitrocellulose membrane and blocked in 5% milk overnight at 4°C or alternatively for 1 hour at room temperature . The incubation of all primary antibodies was performed for 1 hour at room temperature using the following dilutions: anti-PRPF31 ( 1∶500 ) , anti-CNOT3 ( 1∶2 , 000 ) , and anti-β-ACTIN ( 1∶2 , 500 ) . The membrane was washed 3 times with 0 . 05% Tween-20 in TBS . Rabbit and mouse HRP-conjugated secondary antibodies were diluted 1∶1 , 000 in 2% milk and incubated for 1 hour at room temperature . Bands were detected using enhanced chemioluminescence ( Pierce ) . Signal detection via the Odyssey infrared imaging system ( LI-COR ) was performed by using fluorescently-labeled secondary antibodies provided by LI-COR , diluted 1∶5 , 000 in 0 . 5% milk and incubated in the dark , for 1 hour at room temperature . The membrane was then washed twice with 0 . 05% Tween-20 in TBS and once in PBS to remove residual Tween-20 prior to the laser scanning . We used two different siRNA sequences targeting CNOT3 ( QIAGEN , FlexiTube siRNA , Hs_CNOT3_5 and Hs_CNOT3_8 , 1 nmol ) and a negative control siRNA for human genes ( Santa Cruz Biotechnology ) . One day before transfection ARPE-19 cells were seeded at a concentration of 2×105 cells/well in a 6 well-plate , and transfection was achieved by using 5 µl Lipofectamine ( Invitrogen ) and 50 pmol siRNA . RNA was extracted 48 hrs after transfection . Lymphoblasts grown at a concentration of ∼8 million cells in a T75 flask were treated with Actinomycin D ( 5 µg/ml in DMSO ) ( Sigma ) by adding it directly to the medium . Cell pellets were collected at seven different time points ( 0–24 hrs ) and total RNA was extracted and analyzed by q-PCR . Three control lymphoblastoid cells from the Centre d'Etude du Polymorphisme Humain ( CEPH ) were grown to have 107 cells per ChIP experiment . DNA and proteins were cross-linked by adding 1% formaldehyde directly to the medium and by incubating the cells on a rotating hybridization oven at 37°C for 10 minutes . To quench cross-linking , we then added 125 mM glycine and incubated the cells at 37°C for 5 minutes . Cells were pelleted by centrifugation ( 800 g for 5 minutes at 4°C ) and washed twice with cold PBS , supplemented with protease inhibitors . Optimization of the chromatin shearing was performed by using a Covaris sonicator , to obtain on average cross-linked DNA fragments of 150–400 bp . ChIP was performed using buffers provided with the Ep-iT Chromatin Immunoprecipitation kit ( Bio-AAA ) . Immunoprecipitation was performed using three different antibodies: anti-CNOT3 , anti-pol2 ( Bio-AAA ) as a positive control for IP , and serum IgG ( Santa Cruz Biotechnology ) as a negative control for IP . Antibody-protein-DNA complexes were collected on protein A agarose beads ( 2 hrs , 4°C ) , then washed with the low salt buffer , high salt buffer , LiCl buffer , and TE buffer ( pH 8 . 0 ) provided in the kit to remove non-specific binding . Complexes were eluted from the beads by using the elution buffer ( 0 . 1 mM NaHCO3 and 1% SDS ) in an orbital shaker . Cross-links were removed by an overnight incubation at 65°C . Ribonuclease and proteinase K digestion were added to remove specific contaminants , before the eluted DNA was extracted once in 25∶24∶1 phenol-chloroform-isoamyl alcohol and once in 24∶1 chloroform-isoamyl alcohol . DNA was ethanol precipitated , washed in 70% ethanol , and finally eluted in TE . ChIP-PCR was performed using the GoTaq DNA Polymerase ( Promega ) and 0 . 5 µl of the ChIP DNA , by using standard cycling conditions and primers described in Table S3 . GAPDH primer sequences are the ones provided by Millipore for the EZ-ChIP kit , while primers for DHFR have been previously described [38] . Two microliters of ChIP DNA were also amplified by q-PCR using Sybr Green PCR Master Mix ( Applied Biosystems ) and the PRPF31 promoter primer pair ( Table S3 ) . CNOT3 genomic region was amplified by 3 overlapping long-range PCRs ( Table S4 ) , for a total length of 34 Kb . PCR was performed in 20 µl using TaKaRa LA Taq and GC buffer I ( Takara Bio Inc . ) . Final primers concentration was 1 µM , and 200 ng of genomic DNA were used as template . PCR amplification conditions were: an initial step at 94°C for 1 minute , 30 cycles of denaturation at 98°C for 5 seconds and annealing/extension at 68°C for 15 minutes , and a final extension step at 72°C for 10 minutes . Long-range PCR products were sequenced with an Illumina HiSeq 2000 machine , to obtain coverage values in the range of thousands of reads . Mapping of the reads and variant detection was performed by using the CLCbio Genomics Workbench software . Differences of gene expression between asymptomatic and affected individuals were tested by t-test , and likelihood computed by 100 Monte Carlo label-swapping simulations per each gene . One-way ANOVA followed by Bonferroni's multiple comparison tests was used to analyze the effect of CNOT3 silencing on the expression of the target genes . The enrichment of PRPF31 promoter sequence after CNOT3 immunoprecipitation compared to the serum IgG was evaluated by using the Mann Whitney non-parametric statistical hypothesis test . In figures , p<0 . 05 is indicated by one star , p<0 . 01 by 2 stars , and p<0 . 001 by 3 stars . | Retinitis pigmentosa ( RP ) is an inherited disorder of the retina that is caused by mutations in more than 50 genes . Dominant mutations in one of these , PRPF31 , can be non-penetrant . That is , some carriers of mutations suffer from the disease while others do not display any symptoms . In these particular individuals , functional PRPF31 transcripts are expressed at higher levels compared to affected persons , thus compensating for the deleterious effects of the mutated allele . Up to now , the nature of such a stochastic and protective effect was unknown . In this work , we identify CNOT3 as the modifier gene responsible for penetrance of PRPF31 mutations . We show that CNOT3 is a negative regulator of PRPF31 expression and modulates PRPF31 transcription by directly binding to its promoter . In asymptomatic carriers of mutations , CNOT3 expression is lower , allowing higher amounts of PRPF31 to be produced and therefore inhibiting the development of symptoms . Finally , we find that a polymorphism within a CNOT3 intronic region is associated with the clinical manifestation of the disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"biology",
"human",
"genetics",
"genetics",
"and",
"genomics"
] | 2012 | CNOT3 Is a Modifier of PRPF31 Mutations in Retinitis Pigmentosa with Incomplete Penetrance |
The granular layer , which mainly consists of granule and Golgi cells , is the first stage of the cerebellar cortex and processes spatiotemporal information transmitted by mossy fiber inputs with a wide variety of firing patterns . To study its dynamics at multiple time scales in response to inputs approximating real spatiotemporal patterns , we constructed a large-scale 3D network model of the granular layer . Patterned mossy fiber activity induces rhythmic Golgi cell activity that is synchronized by shared parallel fiber input and by gap junctions . This leads to long distance synchrony of Golgi cells along the transverse axis , powerfully regulating granule cell firing by imposing inhibition during a specific time window . The essential network mechanisms , including tunable Golgi cell oscillations , on-beam inhibition and NMDA receptors causing first winner keeps winning of granule cells , illustrate how fundamental properties of the granule layer operate in tandem to produce ( 1 ) well timed and spatially bound output , ( 2 ) a wide dynamic range of granule cell firing and ( 3 ) transient and coherent gating oscillations . These results substantially enrich our understanding of granule cell layer processing , which seems to promote spatial group selection of granule cell activity as a function of timing of mossy fiber input .
The granular layer of the cerebellar cortex consists of populations of granule cells ( GrCs ) , Golgi cells ( GoCs ) , unipolar brush cells , and Lugaro cells [1–3] . GrCs are excitatory [1] and form the largest population of neurons not only in the cerebellum , but also in the entire brain [4] . GoCs are inhibitory and are known as interneurons of the granular layer [1 , 5] . The granular layer of the cerebellar cortex receives its input from different parts of the brain primarily through mossy fibers [6] . The mossy fibers excite both GrCs and GoCs through their typical axonal boutons called ‘rosettes’ [7 , 8] . Within the cerebellar cortex , the GrCs excite GoCs through parallel fibers [9–11] and ascending axons [12] and the GoCs in turn inhibit numerous GrCs through sagittal branching of their axons [5 , 13] . So there exists a feedback loop between the GoCs and GrCs [14] , which has a similar structure to that of the pyramidal-interneuron gamma rhythm generation ( PING ) model of the neocortex [15] . In addition , the GoCs are connected together by gap junctions [16–18] and have also been reported to inhibit each other sparsely [19] . Previous studies have proposed different roles for cerebellar GrCs . Jörntell and colleagues have suggested that GrCs function as signal-to-noise enhancing elements [20 , 21] , based on their observation that GrCs in the C3 zone of decerebrated cats receive identical inputs through mossy fibers that are modality specific , have the same receptive field type and are similarly encoded . Other studies proposed that GrCs provide a bank of various temporal patterns ( tapped delay line model , spectrum models ) that can be used to generate learned temporal responses such as in classical conditioning experiments [22–25] . According to this view , the GrC population is endowed with a variety of time constants so that the different GrCs are active at different moments during conditioned stimuli [26] . One of the earliest proposals for the GrC function was by David Marr , suggesting that they act as low noise sparse encoders [27] , a popular hypothesis supported by some recent electrophysiological and modeling studies [28–30] . In this theory , each GrC represents a combination of a few mossy fibers that provide diverse input , where the number of activated GrCs at any time is small compared to the total number of GrCs , i . e . sparse firing , to facilitate discrimination of binary input patterns by Purkinje cells . In support of this , GrCs receive only a few mossy fiber inputs [31] and exhibit low background firing rates partly due to the presence of tonic GABAergic input [29] . With such a synaptic structure , sparsely firing GrCs could losslessly encode a wide range of spatial input patterns [30] . Sparse encoding assumes relatively uncorrelated GrC activity , however , there are multiple anatomical and physiological mechanisms that promote correlation of GrCs . For instance , it has been demonstrated that bursting of a single mossy fiber afferent can lead to bursting of many GrCs [32] . Furthermore , granule cells in the flocculus respond with high mutual correlation during vestibule-ocular reflex tasks , due to the activation by unipolar brush cells [33] . In addition , spillover mechanisms could conceivably create spatially correlated GrC input [34] . The sagittal arrangement of mossy fiber rosettes are likely to create anisotropic spatial correlations [7 , 35] . Recent in vivo imaging of the granular layer reports a lack of sparse activity in GrCs [36] . Hence , a detailed analysis on how granular layer network mechanisms contribute to spatiotemporal encoding of parallel fiber activity is timely , particularly considering the anatomical interactions between the cell populations characterized by the high density of GrCs [4] ( Fig 1A–1F ) . Early physiologically detailed models of the granular layer , constructed in 1D and 2D , have suggested the presence of robust oscillations in the granular layer of the cerebellar cortex due to the feedback loop between GoC and GrC’s [14 , 37] . The oscillations cease if there is very low mossy fiber activity , or a dominant excitation of GoCs by mossy fibers or high tonic inhibition of GrCs in the network [14 , 37 , 38] . The 2D model suggested that gap junctions between GoCs increase the power of feedback loop driven oscillations [37] . Recently it has been suggested that the emergence of network oscillations can also be linked to NMDA receptors at parallel fiber-GoC synapses [39] . Comparable oscillations have also been experimentally observed in the local field potentials ( LFP ) recorded in the granular layer , in the 10–25 Hz range in the paramedian lobule of primates [40] and in the range of 7–8 Hz in Crus IIa of awake rats [41] . While previous detailed computational models were studied with a limited repertoire of mossy fiber stimuli such as spatially uniform and monotonic ones , etc . [14 , 37 , 39] , the mossy fiber firings in vivo exhibit a variety of temporal and spatial patterns . Vestibular mossy fibers provide slow rate-coded inputs that linearly encode head velocity [33 , 42 , 43] . In response to sensory stimulation , mossy fibers in Crus I and Crus IIa generate high frequency bursts [29 , 32] , and metronome mossy fibers of the lateral reticular nucleus ( LRN ) spike synchronously [44 , 45] . Furthermore , in response to peripheral stimulation , each body part is represented multiple times in the form of patches in the granular layer , where adjacent patches represent non-adjacent body parts , forming a so called fractured somatotopy [46 , 47] . Here we simulate a 3D large-scale network model of the granular layer activated by patches of mossy fibers inputs with realistic firing patterns such as slow rate modulation or rapid bursting , to study how spatiotemporal interactions between the neurons determines holistic network dynamics .
First , we simulated the granular layer network ( Fig 1 ) model with spontaneous background firings of all mossy fibers . The mossy fiber firing rate was 5 Hz , which is comparable to experimental observations: Cuneate mossy fibers fire spontaneously around 9 Hz in vivo [21] . Mossy fibers of the LRN of the brainstem fire regularly ( spontaneous ) in a wide range from 2–23 Hz [21 , 45] . Mossy fiber boutons in vivo from crus I and crus II of cerebellar cortex are spontaneously active around 4 Hz [32] . With the background inputs , the GrCs and GoCs in the model fired with a mean frequency of 1 . 01±0 . 09 Hz and 8 . 18±0 . 61 Hz respectively , and these match values from in vivo recordings , ~1 Hz ( GrCs in Crus I-IIa anaesthetized of rats ) [29] and ~8 Hz ( GoCs in Crus I-II of anaesthetized rats ) [48] . A characteristic feature of the network activated with diffuse mossy fiber input is widely distributed oscillations of GoCs and GrCs ( Fig 2 ) , driven by the feedback loop from GoCs onto GrCs , and vice-versa . The loop consists of AMPA and NMDA receptors of the GrCs activated by the mossy fiber input , AMPAergic receptors in the GoC population activated by the parallel fiber/ascending axon input , and GABAergic receptors in the GrCs activated by the GoCs . Oscillations in baseline could readily be seen in single cell activities of GoCs ( Fig 2A and 2E ) , but were less obvious in single GrC firing for background input , as they are sparsely active ( Fig 2C ) . As observed in a previous 2D network model of the granular layer of cerebellar cortex [37] , gap junctions between GoCs increased the synchrony of GoC and of GrC firing in case of low frequency diffuse mossy fiber input but had less effect when in addition a patch of mossy fibers was activated more strongly ( compare Fig 2B and 2D ) . The mossy fiber firing rate was vital in controlling the firing rates of GrCs and GoCs and of the network oscillation frequency . In addition to the baseline input of 5 Hz , we activated the mossy fibers in patches of 100 or 200 μm in radius , which we will call the ON patch , over a range of input frequencies . This protocol simulated the patch-like mossy fiber activations observed in vivo [49] . In those simulations , GoCs showed highly synchronized oscillations ( Fig 2F ) while GrCs exhibited more loose synchronization ( Fig 2H ) . Oscillation frequency increased with the frequency of the activated mossy fibers ( Fig 3A ) together with the firing frequency of GoCs ( Fig 3B ) and GrCs ( Fig 3C ) , regardless of the patch size . The presence of gap junctions slightly increased the firing frequency of GoCs ( Fig 3B ) and GrCs ( Fig 3C ) for high mossy fiber firing rates . Unlike the previous one-dimensional model of granular layer [14] where oscillations disappeared for mossy fiber firing rate below 15 Hz , oscillations were still observed in our model for 5 Hz mossy fiber background firing rate . Each model GoC needed to receive inhibition on its apical dendrites to obtain experimentally observed firing rates ( see Methods ) . The origin of this inhibition is not conclusively known , but we postulated that it largely originates from nucleocortical neurons [50 , 51] , with a total conductance for each GoC of 2160 pS based on experimental results . This made GoCs fire in the experimentally observed range in vivo . Each GoC also received inhibition on basal dendrites from other GoC cells [19] . This basal inhibition was about ~10 pS/Hz* ( average firing rate of presynaptic GoCs ) , which is much weaker than the inhibition on apical dendrites . However , the effect of GoC-GoC inhibition remained limited even when it was artificially increased . Increasing the synaptic conductance of GoC-GoC inhibition 8 times decreased the GoC firing rate only moderately from ~64 Hz to ~54 Hz with mossy fibers constantly firing at 80 Hz . The same condition changed oscillation frequency barely ( from 52 . 1 Hz to 54 Hz ) , but the oscillation power decreased to 45% . Therefore , increased GoC-GoC inhibition weakened the oscillations , with small effect on their frequency . At the single neuron level , the response to mossy fiber input was quite stochastic ( Figs 3E and S1 . See also S1 Movie ) . This paper , an initial description of our results , will mostly emphasize the average behavior of the network , which is quite complex . But this only summarizes the rich and stochastic network dynamics , shown in the Supplementary Movies . The synchronization index of the GoCs and GrCs increased with the firing frequency of mossy fibers in the activated patch and with the size of the patch ( Fig 3D ) , as a larger number of GoCs became involved in each oscillation . Elimination of gap junctions reduced the synchronization index but did not eliminate synchronization at higher mossy fiber firing rates . The divergence rate from a single mossy fiber to GCs is quite large ( see S1 Table ) . Moreover , mossy fiber rosettes occupy a restricted volume of the granular layer allowing for activation of circumscribed patches ( Fig 1C ) . The sparse coding hypothesis predicts that GrC firing should remain sparse , also for high input conditions . However , in our model the GrC activity within the activated volume climbed quickly with increasing mossy fiber input frequency , from just a few cells to about half the cells in a patch ( in a long integration window of 100 ms , Fig 4A ) . We defined the dynamic range as the ratio of maximal to minimal activation of GrCs ( see Methods ) . The largest dynamic range was found for a physiologically relevant integration window of 1 ms , where increasing the mossy fiber firing rate from 10 to 80 Hz caused 10 . 1 times more GrCs to fire in the ON patch . The sparse baseline firing rate of the GrCs likely represents an almost quiescent network , and this sparse firing quickly transforms into dense firing upon stimulation . Adding tonic inhibition ( see the last section of Results ) slightly increased the dynamic range . Figs 5–7 and S2 Fig demonstrate how the network responded when the input was time dependent , particularly when firing rates were slowly modulated ( Fig 5A ) . For this , we stimulated the mossy fibers in single or double ON patches of 100 μm radius in various configurations that were 1 ) single patch ( Fig 6C , inset ) 2 ) double patches along the transverse axis with 800 μm of the center-to-center distance ( Fig 6G and 6K , inset ) , 3 ) double patches along the sagittal axis with 400 μm distance ( S2H and S2L Fig , inset ) . Volumetric maps of the network activity in response to a double patch input along the transverse axis are shown in S3 Fig and S2 Movie . In both the single and double-transverse patch paradigm , GoCs along the parallel fiber axis showed a high degree of oscillatory synchrony with little effect of firing rate co-modulation ( Fig 5B ) . This was observed not only in the cross-correlation between two ON patches ( Fig 5D ) but also in the cross-correlation of an ON patch with a non-stimulated patch , which we will call an OFF patch ( Fig 5E ) . This demonstrated the effectiveness of the common parallel fiber input to the GoCs along the transverse axis . On the other hand , the effect of firing rate modulation was much more pronounced in the population activity of the GrCs ( Fig 5C ) . In particular , the GrCs in the OFF patches ( along the transverse direction ) showed anti-correlation of their firing rate with the ON patch GrCs , while the synchronized oscillations could still be observed on a shorter time scale ( Fig 5G ) . Therefore , the spatial structure of the average firing rate and the correlations was strikingly different between the GoCs and GrCs . There was only a small spatial dependence of the GoC firing rates along the transverse axis ( Fig 6A , 6E and 6I ) . They exhibited a stable and high cross-correlation along the transverse axis and showed only a slight decay with distance even after discounting the effect of firing rate co-modulation ( Fig 6B , 6F and 6J ) , indicating that the correlation was due to a high degree of synchronization ( Fig 5E ) driven by the parallel fiber inputs . On the other hand , the GrC firing rates displayed an on-beam inhibition [52] featuring an activated ON patch surrounded by laterally inhibited cells along the transverse axis ( Fig 6C , 6G and 6K ) . In a single ON patch paradigm ( Fig 6C ) , the ON patch GrCs fired at 47 . 7±1 . 9 Hz while those in the OFF patch ( separated by 500 μm ) fired at a below-baseline average rate of 0 . 86±0 . 11 Hz , due to increased GoC inhibition in OFF patches . The GrCs along the transverse axis were more strongly correlated within and between two ON patches than between the ON-OFF pairs ( Fig 6D , 6H and 6L ) , with a stronger effect of firing rate modulation ( Fig 5F ) . We compared the two-patch condition for identical ( Fig 6E–6H ) mossy fiber input frequency with simulations where the patches received different input frequencies ( Fig 6I–6L ) , but overall there was little difference suggesting that the spatial input pattern is more important than the input frequencies . Spatial profiles of cross-correlations of long term firing rates ( red curves in Fig 5D–5G ) also clearly exhibited these structures for both GoCs ( Fig 7A ) and GrCs ( Fig 7B ) along the transverse axis . In the single ON patch stimulation paradigm , the on-beam inhibition is clearly seen in the GrC rate correlations along the transverse axis ( Fig 7B ) . This is a little more pronounced without gap junctions between GoCs since the GrC firing rates becomes less variable due to a decrease of spike synchronization , but gap junctions did not alter the spatial structure qualitatively . Therefore , the synaptic connectivity in the granular layer results in a winner-take-all mechanism where OFF patch GrCs get inhibited by GrCs that receive strong mossy fiber input ( ON patch ) through parallel fiber mediated feedback inhibition . In the absence of common parallel fiber input , when activity was measured in patches along the sagittal axis in response to slow rate modulation , GoCs were not activated and GrCs were not inhibited in the OFF patches ( S2A and S2C Fig ) . When two patches were activated in this configuration , GrC correlation reduced as the distance between the patches increased ( S2H and S2L Fig ) and the GoCs exhibited less synchrony ( S2F and S2J Fig ) than for the transverse configuration ( Fig 6F and 6J ) . Patches of GoCs in this configuration were correlated if they were close to each other ( 100 μm ) but the correlation between the patches rapidly decreased with distance . As reported previously [37] , gap junctions had a limited effect on the synchrony between GoCs along the parallel fiber axis , evoked by patch activation . Without gap junctions between GoCs , the synchrony was reduced among GoCs ( for the single patch condition , Fig 6B ) along the parallel fiber axis when compared with the control condition but was maintained all along the parallel fiber axis and didn’t exhibit any decay with distance . This was observed even when we activated two mossy fiber patches along the transverse axis with different firing rates ( Fig 6J ) . The difference in correlation between control and gap junction block conditions decreased for two-patch activation due to increased overall parallel fiber activity ( Fig 6F and 6J ) . Removal of gap junctions between GoCs reduced the synchrony of both GrCs ( S2D , S2H and S2L Fig ) and GoCs ( S2B , S2F and S2J Fig ) along the sagittal axis . To summarize , the application of slow rate coded input to the granular layer model of the cerebellar cortex synchronized the GoCs along the transverse axis , resulting in an on-beam inhibition that makes OFF patch GrCs exhibit anti-correlation on a longer time scale and increases the separation between ON patch and OFF patch GrCs with respect to firing rate . GrC ascending axons , which excite the basolateral dendrites of GoCs [12] , play an important role in shaping the activity of the network , especially the GoC synchronization . In Fig 6 ( see also Fig 8A ) , the cross-correlation of GoCs between two remote ON patches ( separated by 800 μm ) along the transverse axis ( 0 . 64±0 . 02 ) is less than that of the corresponding patches ( 0 . 69±0 . 02 ) for a single patch activation configuration . This is contrary to the intuition that cross-correlation should be larger in the two-patch activation configuration due to an increased activation of shared parallel fiber inputs . We first checked whether this feature is due to differences in the temporal structure of mossy fiber inputs between the activated patches . We eliminated the difference in temporal input structure between the two activated ON patches ( see Methods ) and calculated the cross-correlation . This procedure did not affect the cross-correlation in the two-patch activation condition ( 0 . 64±0 . 01; Fig 8B ) . Next , we eliminated the ascending axon inputs to GoCs ( reduced their peak synaptic conductance to zero ) and calculated the cross-correlation in the same manner as above . Fig 8C shows that when ascending axon inputs are blocked , the cross-correlation between the ON patches ( 0 . 73±0 . 03 ) becomes higher than that of the corresponding patches for single patch activation configuration ( 0 . 70±0 . 02 ) . Moreover , the removal of ascending axon inputs to GoCs , resulted in an overall increase in the cross-correlation for both activation paradigms ( Fig 8C ) . Because ascending axon inputs represent highly 'localized' input sources to GoCs in both ON patches , they result in reduction of the parallel fiber mediated synchronization . In a single patch activation configuration , cross-correlation varied non-monotonically with distance along the transverse axis and this is due to ascending axon inputs to GoCs ( Fig 8A ) . Here the correlation decreases steeply until 400 μm , which is the last OFF-ON pair , and therefore the ascending axons in the ON patch resulted in a reduced correlation ( 0 . 55±0 . 02 ) . Beyond this point , the correlation is a measure between OFF-OFF pairs and recovers up to 0 . 69±0 . 02 at a distance of 800 μm . For the two ON patch configuration , cross-correlation reaches a plateau for distances beyond 400 μm due to the localized ascending axon inputs in the second ON patch ( Fig 8A ) . Increasing the strength of the ascending axon to GoC connections , by increasing their synaptic weights by 20% or 40% of their original values led to a small decrease in cross-correlation of 2–5% along the transverse axis ( not shown ) . We conclude that ascending axon inputs to GoCs reduce the parallel fiber mediated synchronization of GoCs . Besides the slow rate modulation that we have used so far , some mossy fibers can also respond to sensory stimulation in vivo on a much shorter time scale by bursting at an extremely high frequency of a few hundred hertz [32] , sustained for a few tens of milliseconds . This signal is reliably transmitted to the GrCs that show similar bursting [29] . We simulated this by activating the mossy fibers in the selected ON patch ( es ) with burst type inputs . Each mossy fiber in the patch was given nine input bursts , with a duration of 10 ms and a firing rate of 500 Hz ( Fig 9A ) . During the 10 ms burst , all the GoCs that fired emitted only one spike ( S1I and S1J Fig ) and the GrCs spiked 3–4 times ( S1K and S1L Fig ) . As a result , the GoC PSTH showed a sharp peak ( Fig 9B , inset ) , while the GrC PSTH exhibited a prominent broad peak ( Fig 9C inset ) . GrCs were strongly correlated between the activated patches ( Fig 9F ) . GoC firings were highly synchronized for both ON-ON ( Fig 9D ) and ON-OFF ( Fig 9E ) patch configurations . All along the transverse axis , GoC firing was sharply synchronized for both single and double patch conditions ( Fig 10B and 10F ) . The ON patch GrCs along the transverse axis were strongly correlated ( Fig 10H ) . The correlation decreased between the ON and OFF patch GrCs and as a result the correlation vs . distance relationship shows the shape of an inverted bell . Volumetric maps of the network activity in response to a single burst activation of a single patch are shown in S4 Fig and S3 Movie . Along the sagittal axis , the correlation decreased with distance for both GoCs ( S5B and S5F Fig ) and GrCs , ( S5D and S5H Fig ) except when the correlation was measured between the two ON patches ( S5F and S5H Fig ) , which showed strong stimulus driven correlations even with a distance of 400 μm between the two patches . We observed that the on-beam inhibition of GrCs also emerged with the bursting input ( Figs 10C and volumetric representation of cell activities in S4C and S4D ) since OFF patch GrCs were silenced due to inhibition by strongly firing GoCs that were activated by the parallel fibers . This on-beam inhibition predicts that GrC spikes can be gated by synchronous GoC firing along the transverse axis . We examined how strongly synchronized inhibition can regulate GrC firing if multiple sets of the mossy fibers along the transverse axis are bursting , particularly when there are relative time delays between the bursts . This replicates in vivo conditions , where the patches of mossy fibers could get activated along the transverse axis at various latencies in response to peripheral activation . We activated two mossy fiber patches separated by 500 μm along the parallel fiber axis ( Fig 11 , inset ) with the same double patch burst activation paradigm but with different latencies between them . We discovered that the feedback inhibition due to the first patch parallel fibers inhibited the GrC excitation in the second patch when the arrival of the feedback inhibition coincided with their mossy fiber excitation of GrCs . Therefore , when the latency of mossy fibers in the second patch was around 5 ms , the GrC excitation was less effective and GrC firing rate decreased by 20% ( Fig 11 ) compared to synchronous activation of the patches . Because of the slow spillover component of GoC inhibition this effect persisted for intervals up to 15 ms and then slowly declined with a return to baseline firing responses at 30 ms intervals . We conclude that synchronized GoC inhibition can strongly regulate GrC firing , even in the presence of excitatory drive , and this mechanism makes the earliest firing GrCs dominate the network activity . Therefore , it may be more appropriate to describe the competition among inputs in the granular as first-take-all instead of winner-take-all . The activity of ON patch GrCs in response to the burst input was not only characterized by firing during the input , but also showed a long transient even after the offset of the mossy fiber burst ( black line in Fig 12A ) . This was unexpected since there are no other sources to excite GrCs other than mossy fibers in our model . We found that the long transient was due to a long-term gain increase in the firing rate by activation of NMDA receptors . GrCs express NMDA receptors [53] that contain the GluN2C subunit [54] and NMDA mediated currents are known for their non-linear voltage dependence and slow kinetics [55] . The NMDA receptors exhibit voltage dependent block at hyperpolarized membrane potentials due to partial block by magnesium ions , and the membrane needs to be depolarized enough to remove this block . Supralinear synaptic summation [56] can be a mechanism to deliver such depolarization to the NMDA receptors . Therefore , the spiking of GrCs caused by strong mossy fiber input could result in NMDA receptor unblocking and their slow decay kinetics caused effective elevation of the resting membrane potential for a considerable period of time ( Fig 12A , inset ) . In this way , the NMDA receptors can implement a winner-keep-winning mechanism , enabling the GrCs that have already spiked to fire again more easily by improved integration of the subsequent inputs . To further investigate how the effectiveness of this gain change , we delivered a weak probe asynchronous mossy fiber inputs ( rate: 20 Hz , duration: 20 ms ) with different time delays after the burst offset ( Fig 12D ) . Then , we measured GrC firing during the probe inputs and compared it the control conditions where there is no probe input or when the probe inputs were delivered to an OFF patch . If there is no ON patch specific gain change , the firing rate change in the ON patch GrCs with a probe input should be equal to the rate increase in the OFF patch GrCs due to a probe input . Instead , we found a supra-linear increase in GrC firing , even up to 150 ms after a burst offset ( Fig 12E ) . Therefore , the mossy fiber burst input induced a long lasting gain increase in ON patch GrCs . The same simulation was repeated with a reduced conductance of NMDA receptors at the mossy fiber-GrC synapse ( Fig 12A and 12C ) . In this case , while there was little change in the response of the ON patch GrC population during the burst , the rebound activity was significantly reduced ( Fig 12A and 12C ) . In response to the probe mossy fiber input , ON patch GrCs exhibited reduced firing rate ( Fig 12F ) with reduced conductance of NMDA receptors ( decrease in firing rate from 44 . 4 Hz to 10 . 8 Hz in response to an asynchronous input at 2 ms after the burst offset ) . The ON patch response amplitudes to the probe stimulus were now closer to those of the OFF patch GrCs . The winner-keep-winning mechanism was robust to changes in the strength of GoC to GrC synaptic inhibition ( not shown ) . Our results show that the NMDA receptors in GrCs cause a long-term change in their input/output function , which implements a winner-keep-winning mechanism in bursting GrCs . GrCs reliably burst with a single bursting mossy fiber and this has been proposed as a mechanism for reliable signal transmission [32] . Here our results demonstrate that mossy fiber bursts can also be a mechanism for regulating the signal processing property of GrCs and changing how the GrC population filters subsequent mossy fiber inputs . GrCs possess extra-synaptic GABAA receptors ( receptors with δ subunit ) [57 , 58] . In many neurons of the central nervous system , extra-synaptic GABA receptors mediate a form of 'tonic' GABAergic current and play an important role in their baseline excitability [57] . We simulated the effect of tonic inhibition on network oscillations by quantifying the power and frequency of ON patch network oscillations when activated with constant mossy fiber input of different firing frequencies . We modeled the tonic inhibition as a tonic Cl- conductance of 88 pS in the GrCs ( see Methods ) . We found that the tonic inhibition reduced the power of network oscillations ( Fig 13A , 13B and 13D ) . For a mossy fiber frequency of 30 Hz , tonic inhibition reduced the peak power of GoC network oscillations from 336±34 to 154±22 ( Fig 13A ) . We observed a fairly constant reduction in power of network oscillations for all values of input mossy fiber firing rate ( Fig 13D ) . In contrast , the effect of tonic inhibition on oscillation frequency was small ( Fig 13C ) . For a mossy fiber firing frequency of 30 Hz , oscillation frequency for control condition was 40 . 4±0 . 5 Hz while that in the presence of tonic inhibition was 38 . 2±0 . 8 Hz ( Fig 13C ) . Therefore , although tonic inhibition of cerebellar GrCs reduces the power of network oscillations , robust oscillation still arise with sufficiently strong mossy fiber input without any effect on the oscillation frequency .
Our model can be considered a superset that reprises findings of previous 1D [14] , 2D [36 , 38] and 3D [38] network models such as feedback oscillations , while suggesting new dynamical phenomena implied by physiology and anatomy . Previous models were significantly smaller and did not analyze network responses as a function of complex mossy fiber activation . Earlier models in [14] and [37] lacked the fine spatial and temporal structure of mossy fiber activation in the cerebellar cortex and did not include the ascending axon input to GoCs . Similarly , the model in [39] approximated mossy fiber input by current injection . A 3D network model in [38] described network dynamics of the granular layer in response to spontaneous and burst input patterns and replicated the center-surround inhibition observed in experiments in sagittal slices , where the parallel fibers are cut [62] . Because of the absence of any significant parallel fiber contribution , this model could not produce the spatial interactions we described here . Our network model is based on recent conductance based models of individual neuron types [38] and , network topology including both the long folium axis and parasagittal axis in the cerebellar cortex . A potential limitation of our network model is that it extends only for 700 μm along the sagittal axis and therefore does not capture rostro-caudal distribution of a number of structures ( e . g . , rostro-caudal extent of mossy fiber axonal arborization is usually greater than 1000 μm in the granular layer [7] ) . The model does not include NMDA receptors at parallel fiber-GoC synapse because they caused depolarization block in the GoC model and have been reported to be absent in adult animals [11 , 12] . A previous 2D model suggested that NMDA receptors at parallel fiber-GoC synapse can cause activity dependent state transitions in the granular layer [39] . Moreover , the GoC model used in our study does not incorporate the fine branched morphology of GoC dendrites found in the granular layer [11] . Gap junctions between GoCs probably occur more frequently along the sagittal plane of the folia as GoC dendrites follow the zebrin boundaries of Purkinje cells above them [63] . But lack of proper experimental data forced us to model them without any directional dependence . Finally , to achieve physiological low GoC firing rates it was necessary to include dendritic inhibition from sources outside of the granular layer . Inhibition of GoCs has been an unresolved issue since a recent study claimed that MLIs do not inhibit GoCs [19 , 51] and our results emphasize the importance of extracortical inhibition for normal GoC function . Conversely , the effect of GoC to GoC inhibition is modest due to its weak strength [19] . One of the strongest features of the simulated network dynamics is the network wide oscillation driven by a feedback loop between the GrCs and GoCs , and mediated by the parallel fibers , which synchronize the GoCs . The synchronized activity of GoCs , were stronger all along the transverse axis as in experimental studies [10] . Consistent with this result [10] , we found that common parallel fiber input drives the synchronized spiking in GoCs , but the gap junctions also contributed significantly [18 , 64] particularly when mossy fiber firing frequency was low . The GrCs in the network , whose spike timings are controlled by cycles of GoC inhibition , exhibit a less precise synchronization . Afferent mossy fibers that project to the granular layer exhibit a wide variety of firing patterns . Experimental studies have shown that mossy fibers exhibit slow rate modulation during a variety of behaviors [42 , 65] , but can also exhibit burst activity in response to sensory stimulation [20 , 21 , 66] firing at more than 700 Hz [32] . We studied how the network responds to and encodes these different physiological input patterns . With one exception , the observed synchronization patterns differed little . This may seem surprising because the mechanisms are fundamentally different , driven by feedback inhibition for the slow rate modulated input , and caused by locking to the strong stimulus for burst input . In response to both types of input , GoCs exhibited parallel fiber mediated synchronization extensively along the transverse axis and this GoC activity powerfully regulated GrC firing . As a consequence , the GrCs correlations showed much more dependence on their location relative to the stimulus than on stimulus type or frequency . Separate GrC populations along the transverse ( parallel fiber ) axis , fired with significant correlations only when they both received mossy fiber inputs , regardless of whether the inputs were slow rate modulated or bursting . However , along the sagittal axis , the stimulus driven correlations in the GrCs became strong only with simultaneous bursting mossy fiber inputs . On the other hand , input by the ascending axons of GrCs [12] was found decrease synchronization of GoCs along the parallel fiber beam because they are highly local . Additionally , the coherence of granular layer network oscillations was affected by tonic inhibition , which is present only in the GrCs [57 , 58] , without any effect on the oscillation frequency . Extra-synaptic GABAA receptors that mediate tonic inhibition are known to be involved in many neuro-psychiatric disorders and also in memory and cognition [57] , such as hippocampus-dependent learning and memory [67] . In the cerebellum , tonic inhibition improves the representation of sensory information in granule cells [68] , whereas it is unclear how it affects motor learning [69] . In experiments , network oscillations in the granular layer have been probed by the LFP ( reviewed in [40] ) . Our simulation can be augmented by recently developed softwares to compute the LFP directly [70] or via hybrid schemes [71–73] , to predict how the LFP signal depends on physiological factors , which can be verified in extracellular recording experiments [74] . GrC population activity is characterized by two distinctive patterns at two time scales . On long time scales , there is an on-beam inhibition effect due to global inhibition of the unstimulated GrCs along the parallel fiber axis , which implements a first-take-all type mechanism . On shorter time scales , the GrC activities are regulated by the time window imposed by the timing of synchronized GoC spikes , which can regulate precision in timing , particularly regarding different latencies in the onsets of mossy fiber inputs . This is in contrast to a recent modeling study [38] that showed a much smaller inhibitory surround ( < 100 μm in diameter ) around an excited center . However , since that model was limited in space and had few parallel fiber contacts per GoC ( ~100 ) , this was probably due to limitation of the model size . A recent in vitro study [75] also suggested that GoCs provide fast feedback inhibition to GrCs , based on the observation that a GoC receives inputs mostly from nearby GrCs but also some input from distant GrCs . Parallel fiber synapses may deliver much smaller input to a GoC soma compared to ascending axon synapses [12] . However , it has been observed that weak common inputs to individual cells can lead to robust synchronization , not only in the cerebellar network [14] but in many contexts [76] . Furthermore , our model predicts that the earliest GoC inhibition should dominate and this coincides with a recent experimental observation that the majority of GrCs receive early , not late , GoC inhibition [77] . We also observed that NMDA receptors in the GrCs play an important role by inducing a long-term increase in the GrC output gain after ( burst ) spiking , even in the presence of the lower voltage-dependence due to their GluN2C subunits [54] . Therefore , the GrCs that have already fired upon early mossy fiber inputs can integrate subsequent inputs much better than other cells in the network , which we called the winner-keep-winning mechanism . NMDA receptors have been well known for their role in supralinear synaptic integration in many systems including GrCs [56] , which can contribute to information gating ( e . g . , [78] ) . The winner-keep-winning mechanism is a combined effect of two phenomena due to NMDA receptors , sustained depolarization [56] and voltage dependent synaptic integration [79] , that gives an additional advantage ( long-term gain upregulation ) to GrCs that respond to bursting inputs . Furthermore , NMDA receptors in GrCs are known for their roles in synaptic plasticity . It has been proposed that this plasticity can tune the relative latency between the GrC firing and mossy fiber input , which in turn dictates whether the GrC firings can pass the time window imposed by the GoC feedforward inhibition[62 , 80] . All the mechanisms that we have discussed , the network mediated first-take-all and the cell intrinsic property based winner-keep-winning , give a predominant advantage to the GrCs that are activated earlier by the mossy fiber inputs while the others are suppressed . While this leads to a sparse spatial organization of the granular layer output , the activity within activated regions of the GrC can be quite dense due to the high dynamic range of the GrC population . This pattern of activation is compatible with the described fractured somatotopy of tactile inputs in crus II of the cerebellum [46 , 47] . The response to the two-patch configuration can be considered a simulation of patches activated by the same tactile input . Moreover , the larger amplitude of responses observed to the late input from sensory cortex , compared to the preceding trigeminal input [46] , could be explained by the NMDA mediated increase of the GrC gain if the respective mossy fibers synapse onto the same GrCs . Note , however , that our two patch simulation results also apply to co-activated mossy fiber inputs carrying different modalities [81] . In Marr’s pioneering theoretical work [27] and following studies [30] , the GoCs also play the role of regulating how many GrCs activate , but the spatiotemporal aspect of GoC firing has been largely ignored . In the mushroom body in the insect olfactory system , the synchronous and oscillatory firing of inhibitory interneurons maintain sparse firing of excitatory neurons [82] . In our model , the GoCs are governed by a similar principle since they oscillate , discharge synchronously over an extended spatial scale , and impose a narrow time window leading to effectively inhibiting a large number of GrCs , contributing to strongly spatially restricted activation . However , contrary to insect olfaction and to Marr’s theory [25] , our model predicts that GrC activity within the activated patch depends on the strength of the mossy fiber stimulus and is often not sparse . The nature of coding by the granular layer has been actively debated: Jörntell and colleagues in their study in C3 zone of decerebrated cats have found little evidence of sparse coding [20 , 21 , 83] . In C3 zone of cats , the authors reported that GrCs receive similar kind ( unimodal ) of mossy fiber inputs [20] , whereas diverse mossy fiber inputs should converge at a GrC ( multimodal ) for sparse coding to work effectively . Also , GrCs in their study were not silent at rest and fired a barrage of spikes in response to peripheral activation [20] . However , other studies in mouse cerebellar cortex demonstrated that GrCs receive multimodal mossy fiber input . For example , Huang et al reported convergence of proprioceptive ( external cuneate nucleus ) and pontine ( basilar pontine nucleus ) inputs in various regions of the cerebellar cortex [84] . Convergence of multimodal mossy fiber inputs ( vestibular , visual ) is found in the GrCs of the vestibulocerebellum [85] and in the hemispheres ( tactile , auditory and visual ) [81] . Our model is neutral towards the convergence discussion because we did not specify what information is carried by the mossy fiber input . The sparse coding by the GrCs hypothesis has recently also been challenged based on experimental observations of dense coding by GrCs [36] and that the GrCs also rate code the rate modulated MF inputs [83 , 86] ( see also [87] ) . Similarly , our model showed that , despite strong temporal patterning by the GoCs , the GrC population rate follows the rate modulated mossy fiber input quite well , resulting in a large dynamic range . Furthermore , spatially separated GrC populations can co-activate , when each of them are stimulated by a different mossy fiber group . Note that this would be impossible if GoC inhibition is purely based on an asynchronous rate code , since no time window for co-activation would be allowed . Therefore , the rich spatiotemporal dynamics of our model provides a unified viewpoint for the resolution of experimental controversies about coding in the cerebellar granular layer . Our simulations suggest that oscillations characterize the basic network activity of cerebellar granular layer network along with stochastic spiking of GoCs and GrCs and rich spatio-temporal dynamics . A first-take-all mechanism based on the network structure and NMDA receptor mediated winner-keep-winning mechanisms further characterize the spatiotemporal dynamics of granular cell firing . Wide dynamical range indicates a large flexibility in the allocation of granule cells , ranging the encoding from sparse to dense . Based on our results , we suggest that the unique anatomy of the cerebellar granular layer , coupled with cellular and network mechanisms promote spatial group selection of GrC activity as a function of MF input timing and spatial organization .
We used previously published models of GrCs and GoCs [38] except that the dendritic morphology of a GoC was modified: two shorter ( 60 μm long ) baso-lateral dendrites were constrained to the granular layer and the other two , longer ( ~166 μm long ) , apical dendrites extended into the molecular layer as in [1] . We also reduced the diameter of the dendrites to 2 . 4 μm to match the electrical and firing properties to the original model . All the cell and synapse models ( see below ) were simulated at the temperature of 37°C . For simulations with tonic inhibition , we included a tonic conductance of 88 pS with a reversal potential at -73 mV in a GrC model , which resulted in ~260 pS of total tonic conductance [32] , which includes stationary activation of GABAergic synapses in the baseline condition . Our granular layer network model is based on detailed anatomical information previously published [1 , 4 , 5 , 7 , 16 , 18 , 19 , 31 , 35 , 89 , 90] . The 3D network model has dimensions of 1500 μm along the transverse axis , 700 μm along the sagittal axis and 430 μm along the vertical axis ( Fig 1 ) . The granular and molecular layers were 200 μm thick each , with a 30 μm thick Purkinje cell layer between them . The number of neurons in the network was determined in the following way: We first calculated the number of GoCs in the network using the anatomical GoC density ( 9500 cells/mm3 ) [18] . From this we calculated the number of GrCs in the network using the GrC to GoC ratio [4] . The total number of GoCs in the network for the above-mentioned network dimension was 1 , 995 and total number of GrCs amount to 798 , 000 . The somatic centers of all the neurons were uniformly distributed in the granular layer . After this , we determined the connectivity between the neurons based on connectivity rules that we will explain in the following section . The neurons were then connected with experimentally validated synapses and gap junctions with corresponding conduction delays depending on their mutual distances . The conduction velocity of parallel fiber axons was set to 0 . 3 m/s [91 , 92] , while that of mossy fiber and GoC axons was 2 m/s [93] . Mossy fiber rosette distribution was based on that of LRN mossy fiber axons [7]: Rosettes of a single primary collateral of LRN axon distribute widely along the parasagittal axis , but along the transverse axis the spread is limited . As a result , rosettes of a single LRN axon are arranged in sagittal strips parallel to each other along the transverse axis . A similar parasagittal arrangement of mossy fiber rosettes is also reported in other studies [35] . We first constructed mossy fibers with a density of 5000 fibers/mm2 , which is based on the projection density of mossy fibers in C1 zone of Paramedian lobule of the cerebellum [94 , 95] . Due to network size limitations , the distribution of mossy fiber rosettes in the model is based on that of a single primary collateral of a LRN axon [7] . For each mossy fiber in the model , the rosettes were distributed according to the rosette cluster distribution of primary collaterals of LRN axon [7] . We used another experimental data set about the distribution of pontine mossy fiber rosettes ( [96] , private communication with Daria Rylkova ) to optimize their distribution in the model . For each mossy fiber in the model , we adjusted the extent of spread of rosettes both along the long axis and sagittal axis until the amount of overlap between mossy fibers closely matched that of the pontine mossy fiber data . We measured the amount of overlap between different mossy fibers in the model and experimental ( pontine ) data as follows: We divided the entire volume into a number of small cubes and counted the number of distinct mossy fibers represented in each cube . From this data , we calculated the relative number of cubes representing 0 , 1 , 2 , 3 and 4 distinct mossy fibers . This was then repeated for different cube sizes . We computed the final mossy fiber density using anatomical ratio of glomerulus to GrC [31] . In order to eliminate boundary effects , we also instantiated mossy fibers around the network when the rosettes projected into the model . Total number of mossy fibers that project at least one rosette into the model was 2109 and total number of rosettes was 29519 . The connectivity between neurons in the network is based on anatomical connectivity patterns observed in the cerebellar granular layer [5 , 31 , 89 , 97 , 98] . The model has synapses projecting from excitatory mossy fibers to GrCs , mossy fibers to GoCs , inhibition by GoCs of GrCs , excitation by GrCs of GoCs through ascending axons and parallel fibers ( Fig 2 ) . In addition to the synapses listed above , GoCs are connected by gap junctions [16 , 18] and inhibitory synapses [19] . Convergence , divergence and synaptic parameters for each synapse in the model are described in S1 Table . Except for synapses in GrCs ( see below ) , the time course of synaptic conductance Gsyn ( t ) was modeled according to the standard double exponential equation [99] Gsyn ( t ) =gmax×N×[exp ( −tτdecay ) −exp ( −tτrise ) ] ( 1 ) where τrise and τdecay are rise and decay time constant respectively . gmax is peak synaptic conductance , and N is a normalization factor that makes the maximum of Gsyn ( t ) equal to gmax . τrise and τdecay were obtained by fitting Eq 1 to the respective experimental traces . The connectivity between mossy fibers and GrCs is based on the maximum length of the GrC dendrite [31] . For each GrC , we formed a sphere of radius of 30 μm around its center and connected to rosettes within that sphere in a probabilistic manner . On average , each GrC received 4 . 5±1 . 5 ( 2–7 ) distinct mossy fiber connections . We used the mossy fiber-to-GrC synapse model of [38] ( deterministic version ) with the following modifications: First , neurotransmitter diffusion is approximated by a cascade linear process [100] , dPdt=−rfastP , dTdt=−rTT+P−rT1 ( T−I1 ) , dI1dt=r1T ( T−I1 ) −r12 ( I1−I2 ) , dI2dt=r21 ( I1−I2 ) −r23 ( I2−I3 ) , dI3dt=r32 ( I2−I3 ) , ( 2 ) where T is the concentration of diffused neurotransmitter . At each presynaptic spike , P is transformed as P→P+y where y represents a diffusing fraction of released neurotransmitter , controlled by a synaptic facilitation/depression mechanism . The parameters are given as rfast = 4/τD , rT = 6 . 2/τD , rD1 = 20/τD , r1D = 9 . 09/τD , r12 = 4 . 9/τD , r21 = 1 . 71/τD , r23 = 0 . 55/τD , and r32 = 0 . 333/τD . The diffusion time constant τD is given by τD=100Rd2/4D where Rd = 1 . 03 μm and D = 4 μm2/ms [38 , 79] . This scheme provided a good approximation of the AMPA and NMDA activation over a wide range of presynaptic inputs ( S6A Fig ) . Second , we set the desensitization constant of the NMDA receptor to 12×10−4 ms-1 [79] . Finally , the voltage dependence of the NMDA receptors is modeled as f ( V ) =11+[Mg]oKMgexp ( −V/δV ) ( 3 ) where [Mg]o = 1 mM , KMg = 1 . 77 mM , and δV = 22 . 4 mV [54] . Conductance parameters of the receptors were adjusted to have GrC firing ~1 Hz in the baseline condition ( mossy fiber firing at 5 Hz ) and also its ~6 fold increase in the absence of inhibition [68] . For connectivity between mossy fibers and GoCs , we assumed a sphere of radius 100 μm and connected GoCs and rosettes within that sphere probabilistically . Each GoC in the model received an average of 13 . 7±6 . 5 ( 1–36 ) distinct mossy fiber connections . The mossy fiber to GoC synapse is glutamatergic ( only AMPA receptors ) whose synaptic parameters were obtained from the experimental recordings of GoC EPSCs [101] . Inhibitory connections between GoCs and GrCs were based on the extent of axonal arborization of the former . GoC axons exhibit a parasagittal organization ( Fig 2 ) [5] . Distribution of their axonal boutons is about 650 μm along the parasagittal axis and about 180 μm along the medio-lateral axis . We assumed a connection probability that generated 8 . 4±3 . 2 ( 1–22 ) GoC synapses per GrC on average . Synaptic parameters were based on experimental data [13] and included an indirect spillover component ( S6B Fig ) . The IPSC decay consisted of two components: the transient component with a time constant of 5 ms and an indirect spillover component with a time constant of 35 ms ( that contributed to 10% peak amplitude ) . The IPSC rise time constant was 3 ms . Connections from GrCs to GoC via parallel fibers/ascending axons were generated using our custom tool , the Boundary Representation Language ( BREP ) [102] . In this method , the geometric structures associated with the connectivity ( parallel fibers/ascending axon and GoC apical dendrites ) were described as points in space along a straight line in three dimensions . The ascending axon of each GrC was modeled as a straight vertical line of length 200 μm with points separated by 50 μm . The parallel fibers of each cell were modeled as two straight lines of length 1000 μm each ( extending on either side of their bifurcation from the ascending axon in the molecular layer ) with points separated by 7 . 5 μm . Small random perturbations were added to both . GoCs also had random angular displacements of their dendritic points . In each GoC , the dendritic elements were modeled as lines that lie on the surface of an inverted cone of height 332 μm ( apical dendrite ) or 6 μm ( baso-lateral dendrite ) . Each dendritic element was created with a randomly chosen angle from a normal distribution with mean ( 30° , 120° ) for apical and ( -20° , -240° ) for basolateral and standard deviation of 10° . The elements were first rotated on to the circumference of a circle of radius 100 μm and raised ( apical in molecular layer ) or lowered ( basolateral in granular layer ) thereby forming an inverted cone . The GoC axons were represented as uniformly distributed random points in a rectangular area ( boundaries in μm: transverse [-45:45] , sagittal [-160:160] , vertical [-75:75] ) relative to the soma position . Once the points associated with each geometric cell structure were generated , we used a K-d tree data structure [103] to order the points and performed fast nearest neighbor searches . We assumed a connectivity radius of 30 μm and 5 μm for ascending axon and parallel fiber connections , respectively . The connectivity probability was chosen to achieve the target number of connections . On the average , each GoC in the model received about 554±302 ( 55–1245 ) ascending axon connections . The number of parallel fiber synapses ( 4759±1037 ( 2512–6582 ) ) on a single GoC in the model was calculated based on the density of parallel fiber synapses in the molecular layer [98] and also based on the fact that approximately 9% of them are formed on structures other than Purkinje neuron spines [97] . Both ascending axon and parallel fiber synapses on GoC dendrite are AMPAergic with time constants and maximal synaptic conductance described in S1 Table . We connected the GoCs with gap junctions [16 , 18] and inhibitory connections [19] . Inhibitory and gap junctional connectivity between GoCs were also generated by BREP . The probability distribution function ( Boltzmann function ) for gap junction connectivity was based on an experimental published data [16] and the conductance decayed as a function of distance [18] as g = β exp ( -λx ) where β = 1 . 659 nS and λ = 0 . 01259 μm-1 . Each GoC had about 13 . 7±4 . 6 ( 1–31 ) gap junctions on average . For inhibitory connections between the GoCs , we used the experimental measurements ( 20% connectivity probability at 50 μm ) from ref . [19] , coupled with the gap junction connection probability data . Each GoC received inhibitory input from 2 . 2±1 . 6 ( 0–10 ) GoCs on average . As we tuned the model with the background mossy fiber firing of 5 Hz by varying synaptic conductances , we discovered that the firing rate of GrCs and GoCs tended to covary with a ratio of GrC:GoC ≈ 1:30 . This suggested that GoCs needed extra inhibitory inputs to reproduce in vivo observations of GrC:GoC ≈ 1:8–18 . Recent studies also suggested that the inhibitory inputs mostly originate from extracortical neurons [50 , 51] , which are not in our model . To simulate the effect of those inhibitory inputs , we included tonically active GABA receptors in the apical dendrites of the GoCs . We estimated that they should roughly correspond to ~150 synapses with a peak conductance ~180 pS [51] and also a 5 ms decay constant in in vivo-like conditions . A range of activation rates from 15 to 20 Hz robustly resulted in ~1 Hz and ~10 Hz firing of GrCs and GoCs , respectively , with a 5 Hz mossy fiber input , and we chose 16 Hz , which led to the resulting total conductance of 2160 pS . We introduced random variations in the cellular and synaptic parameters as follows: GoC and GrC soma diameters were randomly varied by up to 20% , and their initial resting membrane potential was also varied in the range -60 to -75 mV . For each type of synapse , the peak conductances were varied in a manner so that they had a coefficient of variation of 0 . 25 . We generated firing of each mossy fiber by using a leaky integrate-and-fire ( LIF ) neuron model driven by a noisy current input: the membrane voltage of the model was given by dVdt=−V−Eτ+βμ ( t ) +σ ( t ) ξ ( t ) , where τ = 1 ms , E = -70 mV and ξ ( t ) was a Gaussian white noise updated every 1 ms . The spike threshold was at V = -60 mV . After a spike , a refractory period of 1 . 1 ms was imposed and then V was reset to E . μ ( t ) and σ ( t ) were controlled by a common parameter ν ( t ) as μ ( t ) = gNν ( t ) and σ ( t ) =gNν ( t ) where N = 1000 and g = 5 μV/ms . We chose β = 0 . 01 to ensure that the model would fire mostly due to noisy fluctuations in the input . We first generated a table of constant ν vs . the output firing rate , and used it to calculate ν ( t ) for a certain target firing rate by linear interpolation . The background firing of each unstimulated mossy fiber was generated by the LIF neurons firing at 5 Hz as in the in vivo recordings [32] . The stimulated mossy fibers were activated in patches of 100 or 200 μm radius , with different input patterns such as slow rate modulation or bursting . In the slow rate modulation paradigm , the input was defined by upper and lower bound frequencies , each lasting 300 ms . The lower bound frequency was kept constant at 10 Hz for all simulations . Upper bound frequency was varied between 50–60 Hz . A combination of an upper bound and lower bound is an epoch , which lasts for 600 ms and each of the rate modulated mossy fiber stimuli consists of five epochs ( Fig 5A ) . The firing rate epochs were smoothed with a Gaussian kernel ( σ = 50 ms ) and the LIF spike trains were generated based on the rate . In the burst input paradigm , we activated mossy fibers in patches of 100 μm with bursts of frequency 500 Hz and duration 10 ms . Mossy fibers were activated in single or two patches either along the parallel fiber axis or sagittal axis . For simulations involving ascending axon mediated de-synchronization of GoCs along the transverse axis , we eliminated the difference in temporal structure in mossy fiber input between the two activated patches ( patch 1 and 2 ) in the following way: For each GoC in patch 1 , a corresponding GoC was randomly picked ( without replacement ) from patch 2 . Mossy fiber connectivity to the GoC from patch 2 was made identical to that of the GoC from patch 1 . We recorded spike times of all neurons during the course of simulation . Simulations were repeated 5 times with different global random seeds , which also affected the network structure . Data was then analyzed using MATLAB version R2011b ( Mathworks , MA , USA ) software . Spike times were transformed into spike trains with 1 ms long time bins . We often evaluated the average activity of specific neurons within a certain region by taking the average of the corresponding spike trains . Oscillations were measured by binning the spikes of the population ( GoC or GrC ) in 1 ms long time bin . Power spectral density of the resulting oscillations was calculated and oscillation frequency was taken as the frequency corresponding to peak power in the power spectral density . The synchronization index for GoC oscillations was calculated as the proportion of total number of GoCs involved in each oscillatory cycle , calculated by integrating the area under each oscillatory cycle . The firing rate of ON patch neurons in Figs 9 , 10 and 11 was computed for a time period of 30 ms from the burst onset . For Figs 6 and S2 ( slow rate coded input ) , the firing rate of ON patch neurons was computed for a period of 100 ms ( during the upstroke of the epoch when mossy fiber firing rate is maximum ) . The firing rate of OFF patch neurons was computed during the corresponding time period . All indicated values in the study represent mean ± standard deviation . Dynamic range of GC activation ( Fig 4 ) was quantified for both ON patch and OFF patch GrCs ( both transverse and sagittal axis ) when simulated with different mossy fiber firing rates . This was done by calculating the percentage of active GrCs in different time windows ( 1 , 10 , 100 ms ) over the course of the entire simulation and averaging it across different data sets . Dynamic range was then calculated using the formula , Dynamicrange=maxvalueminvalue . The volumetric reconstructions of cell activity are achieved by binning the cells spiking within any millisecond of simulation in a voxel of 10 μm . This produces a 3D histogram of spike counts per voxel . This volume is then convolved with 3D Gaussian kernels normalized by the maximum value of the kernel . The resulting voxel value is proportional to the maximum number of cells active in the voxel . The result is passed as color and alpha components to the MATLAB function Vol3D ( http://www . mathworks . com/matlabcentral/fileexchange/22940-vol3d-v2 ) . Cross-correlations were computed between the average activities of neurons in two selected patches , regions of 100 or 200 μm radius in the model . We first formed the spike trains of all the neurons with 1 ms time bins . The average activity yA for patch A is given by , yA= ( ∑i=1NAAi ) NA ( 4 ) where NA is the number of the GrCs or GoCs in A . The cross-correlation function ( CCF ) between region A and B is given by CCFAB ( t ) =1L×ZAB∑s=1L−t ( yA ( s+t ) –yA¯ ) × ( yB ( s ) −yB¯ ) ift≥0 , CCFAB ( t ) =1L×ZAB∑s=1−tL ( yA ( s+t ) –yA¯ ) × ( yB ( s ) −yB¯ ) ift<0 , ( 5 ) where ZAB= ( Var[yA]×Var[yB] ) and L is the length of yA , B . CCF was computed for t = ±300 time lags . We defined cross-correlation ‘c’ as oscillatory synchrony after discounting the effect of firing rate modulation . We denote the measured correlation coefficient by ‘a’ ( zero-time lag correlation CCF ( t = 0 ) ) , the expected coefficient from firing modulation only by ‘b’ , and compute c as c = a-b . In all sections of the results we report the value of ‘c’ as cross-correlation . To find the effect of firing rate co-modulation , the average population activities for each patch were low pass filtered below 10 Hz , which was above the frequency range of our input firing rate modulation . We computed CCFs based on them according to Eq 4 , except that the normalization factor ZAB is still based on the unfiltered spike trains . This scheme made it easier to compare cross-correlations at two different time scales ( e . g . , blue and red lines in Fig 5D–5G ) . Statistical significance of CCF ( t = 0 ) is non-parametrically evaluated by counting the number of the outliers nout in {CCFshuffled} whose amplitude exceeded that of CCF ( t = 0 ) . CCFshuffled was computed for t = ±300 time lags in the following way: In each of the patches , we divided each simulation epoch , which can be the period of the entire simulation or each stimulation protocol depending , into a number of small sub-epochs ( of length 10 ms ) and randomly shuffled the GrC and GoC spike trains in the divided sub-epochs . This gave us the average activity zA and zB from the shuffled spike trains of two patches . Nshuffle was chosen to be 50 . CCFshuffled was calculated from zA and zB according to Eq 4 . Then , the p-value of CCF ( t ) was estimated by an empirical type-I error rate , p=noutNtotal , where Ntotal = Nshuffle × ( 2 × nlag + 1 ) . We assumed a confidence interval of 99% and p<0 . 01 was considered to be significant . We also tried to increase the statistical power by combining the results from multiple simulations . In this case , CCFs were appropriately averaged and the p-values were obtained from combined observations . Error bars in cross-correlation were obtained by bootstrap resampling . | The cerebellum is an organ of peculiar geometrical properties , and has been attributed the function of applying spatiotemporal transforms to sensorimotor data since Eccles . In this work we have analyzed the spatiotemporal response properties of the first part of the cerebellar circuit , the granule layer . On the basis of a biophysically plausible and large-scale model of the cerebellum , constrained by a wealth of anatomical data , we study the network dynamics and firing properties of individual cell populations in response to 'realistic' input patterns . We make specific predictions about the spatiotemporal features of granule layer processing regarding the effects of the gap junction coupled network of Golgi cells on a spatially restricted input , in an effect we denominate first-takes-all . Furthermore , we calculate that the granule cell layer has a wide dynamic range , indicating that this is a system that can transmit large variations of input intensities . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"cell",
"physiology",
"medicine",
"and",
"health",
"sciences",
"action",
"potentials",
"nervous",
"system",
"membrane",
"potential",
"junctional",
"complexes",
"electrophysiology",
"neuroscience",
"gap",
"junctions",
"nerve",
"fibers",
"neuronal",
"dendrites",
"animal",
... | 2017 | Spatiotemporal network coding of physiological mossy fiber inputs by the cerebellar granular layer |
Several mechanisms have been proposed to explain how ion channels and transporters distinguish between similar ions , a process crucial for maintaining proper cell function . Of these , three can be broadly classed as mechanisms involving specific positional constraints on the ion coordinating ligands which arise through: a “rigid cavity” , a ‘strained cavity’ and ‘reduced ligand fluctuations’ . Each operates in subtly different ways yet can produce markedly different influences on ion selectivity . Here we expand upon preliminary investigations into the reduced ligand fluctuation mechanism of ion selectivity by simulating how a series of model systems respond to a decrease in ligand thermal fluctuations while simultaneously maintaining optimal ion-ligand binding distances . Simple abstract-ligand models , as well as simple models based upon the ion binding sites in two amino acid transporters , show that limiting ligand fluctuations can create ion selectivity between Li+ , Na+ and K+ even when there is no strain associated with the molecular framework accommodating the different ions . Reducing the fluctuations in the position of the coordinating ligands contributes to selectivity toward the smaller of two ions as a consequence of entropic differences .
The ability of some biological molecules to discriminate between different ions is crucial for their function . This differentiation is important , for example , in the generation ( or regulation ) of the action potential during cellular signalling , and the maintenance of an electrochemical gradient across the cell membrane [1] . Indeed , without this ability to discriminate between ions , a cell would quickly die . Of particular interest is how such molecules are able to distinguish between the monovalent cations Na+ and K+: these ions are both spherical , they have identical charges , and they differ in atomic radius by only 0 . 38 Å . It is incredible that some proteins , such as potassium selective ion channels , can discriminate between these two ions at nearly diffusion limited rates [2]–[5] . Although it is generally agreed that selectivity depends on a difference in free energy relative to bulk water of one ion compared to the other at some position within the transit pathway ( i . e . how well the loss of free energy from dehydration is recouped by coordination with the protein ) , there are several different proposals which attempt to explain how this difference in free energy occurs . These proposals fall into three broad categories related to: In this study we focus on the last category . To date , three different cavity effects have been proposed that can lead to ion selectivity: the ‘rigid cavity’ , the ‘strained cavity’ , and the ‘reduced ligand fluctuation’ ( RLF ) mechanisms . We discuss each in turn below; Table 1 summarises their key similarities and differences . The ‘rigid cavity’ mechanism is perhaps the easiest to understand [24]–[26] . It suggests that ion selectivity is created by the ligand framework maintaining a certain fixed position ( i . e . cavity size ) about an ion regardless of the type of ion that is coordinated . Specific positions will be energetically more favourable for one ion type over another , thus contributing to selectivity for that ion . For example , when the smaller ion is favoured because the binding site is too small to fit the larger ion , this is often termed ‘size selectivity’ . If the favoured ion is larger and sits more favourably in the cavity , this mechanism is commonly called ‘snug fit’ . In reality the positions of the ligands will never be completely fixed , and their thermal motion is often larger than the size difference between Na+ and K+ . Taking these thermal fluctuations into account , it has been demonstrated , in principle , that if the ligands fluctuate about some fixed average configuration for different ions this will create ion selectivity [10]–[12] . The question of which particular ion is selected by a given cavity site depends strongly upon the actual positions to which the ligands are constrained . Even this picture of a rigid cavity is probably too simplistic as the ligands are likely to fluctuate about different average positions when coordinating different sized ions . If the difference in the average positions is less than the difference in ion radii , one may still consider this situation to be a ‘rigid’ cavity . Our studies of many proteins suggest that the difference in average ion-ligand distance when coordinating Na+ and K+ is almost always similar to the difference in ionic radii , suggesting that a true rigid cavity is uncommon in proteins [12] . Unlike a ‘rigid cavity’ , a ‘strained cavity’ allows for the average ion-ligand distances to adjust according to different ion types . However , in this case the adjustment comes at an energetic cost , called ‘strain’ . Strain may be realised as a deformation within the ligand itself , or as a deformation of the ligand/protein scaffold , be it local [8] , [27] or non-local to the ligand site [12] , [19] , [28] . Non-local strain may itself precipitate a conformation change in the protein ( an extreme version of the effects of strain ) thus further influencing ion selectivity [28] . A rigid cavity can be considered as an extreme form of a strained cavity , wherein the coordinating ligands resist any attempt to adjust to a new ion type in the binding site , perhaps due to an even larger cost in energy of deforming the protein scaffold . A continuum exists between the two , characterised by the degree of change in the average position of the ligands upon a change in ion type . As already noted , a rigid cavity is unlikely to exist in proteins , due to the inherent flexibility of these structures . The idea that differentiation between ions could be achieved through a rigid cavity mechanism was first suggested by Mullins [24] , [29] , [30] who was investigating selectivity in ion channels . It was suggested that a rigid pore of an appropriate size could allow favourable interactions with K+ , but be too big for Na+ , leading to unfavourable interactions . This mechanism was supported voltage clamp experiments by Bezanilla and Armstrong [25] which suggested the pore was lined with backbone carbonyl oxygens , the particular arrangement of which mimicked bulk water more closely for K+ than Na+ . More recently , Doyle et al . [26] have purported to suggest that ion selectivity in KcsA resulted from a rigid cavity mechanism . Whether it be a poor choice of words by the authors or misinterpretation by others ( or a combination of the two ) , it seems that this explanation was offered as a caricature of a strained cavity mechanism , a point that is clarified in a later study [28] . The strained cavity mechanism has also been shown to play a role in ion selectivity in some ionophores , such as valinomycin [31]–[33] , where the small amount of scaffolding between ion coordinating groups can leave the molecule sensitive to subtle sizes differences between coordinating ions . We propose that the rigid and strained cavity mechanisms are two domains of the same continuum; the term ‘strained cavity’ will herein be used to encompass this , except for when contrast between the two is required . Both share the common feature of requiring a resistance to changes in the positions of the coordinating ligands in order to generate ion selectivity . Could ion selectivity be generated in an ion binding site through a ‘cavity’ like mechanism without the need for strain ? In a previous investigation using simple abstract-ligand models it was noted that ‘cavity’ based mechanisms could still create selectivity even when both the cavity size is not fixed and when there is no strain associated with adjustment of the ligand positions [12] . In this case , the only ‘cavity’ factor controlling the binding energy of the ions was the degree of thermal motion associated with the ligands as the ligands were free to adopt their optimum positions for each ion type . This is the essence of the ‘reduced ligand fluctuation’ ( RLF ) model for ion selectivity that is the focus of this investigation . Reduced ligand fluctuations ( i . e . small values of root-mean-square ( RMS ) deviations in positions of the atoms forming the ligands from their average positions ) of the ligands in the ion binding site compared to the rest of the protein were noted in atomistic molecular dynamics ( MD ) simulations of LeuT [12] , [34] , a leucine transporter , and GltPh , an aspartate transporter [12] . This situation contrasts with other molecules we studied previously , in which the RMS fluctuations of the ligands in the ion binding site were not notably smaller than the average across the protein [12] . It was also demonstrated , at least in the case of GltPh , that the Na+ binding sites were able to accommodate K+ ( i . e . the sites were able to change the ion-ligand distance to a more favourable one for K+ ) , ruling out a rigid cavity mechanism creating ion selectivity in this molecule . It could be the case that there is an energetic penalty in adjusting to K+ ( strain ) , however , this is difficult to quantify [12] , [18] . In this investigation we try to better understand how reducing ligand fluctuations creates ion selectivity , improving on the simple models used in previous work [12] by using better force field parameterisations , and by investigating a greater variety of model systems . Simplified model systems have been used to study ion selectivity in a range of molecules including potassium channels [8] , [9] , [12] , [13] , [18] , [19] , [22] , [27] , a sodium channel [35] , NaK channels [27] and kainate receptors [36] . In addition , the previous work never addressed exactly how reducing the thermal motion of the ligands leads to selectivity . Using the more detailed model systems , and by analyzing binding-energy components , we are able to propose an answer here . Of course , the RLF mechanism is not mutually exclusive with the other means of obtaining ion selectivity previously mentioned; it may work in concert with these other effects . However , in this investigation , we wish to study it in isolation , so as to clearly discern it from these other possibilities .
One can elucidate the effect of the RLF mechanism on the selectivity of an ion binding site by investigating how a series of ‘abstract’ model systems ( pioneered by Noskov et al . ) [8] respond to a reduction in the RMS fluctuation on the coordinating ligands . These model systems consist of a number , , of abstract ligands ( in this case based on formaldehyde ) , where each oxygen atom is confined to a 3 . 5 Å sphere about an ion , either Li+ , Na+ or K+ . This first constraint is enforced by a one-sided harmonic potential with a very large force constant . This spherical constant is not meant to precisely define the coordination numbers of the ions , but rather to control the number of ligands near to the ion as a model of the composition of a biological ion binding site . The position of the central ion is fixed , while each atom in the coordinating ligands can be further constrained by placing them in an additional harmonic potential , centered at a nominated position . The amount of thermal fluctuation of the ligands can be controlled by altering the force constant , . The choice of physical location at which the harmonic restraint is placed is very important . In order to isolate RLF from strain , the harmonic restraint needs to be placed at an ‘optimal’ ion-ligand distance for each ion type . This position is defined as the maximum of the first peak in the radial distribution function from dynamics simulations of Li+ , Na+ or K+ surrounded by ligands with no harmonic restraint . The systems where the position of the coordinating ligands are controlled are referred to as with ligands , where M+ = Li+ , Na+ or K+ , with geometries determined by the vertices of optimal coordination polyhedra for n = 4 , 6 , 8 and by packing circles on spheres for n = 5 , 7 [37] . is the distance between the center of the ion and the center of the coordinating atom . In order to quantify ion selectivity , the free energy to exchange two ions between bulk water and our model binding sites is calculated . As each model system is allowed to adopt an ‘optimal’ cavity size ( or ion-ligand distance ) for each ion type , a series of free energy perturbation molecular dynamics ( FEP MD ) simulations are required in order to describe the contribution to ion selectivity from this mechanism in a meaningful way . The overall reaction to be investigated is an exchange reaction of the ions and between an optimally sized hypothetical model system for each ion ( at ion-ligand distances and ) with controllable fluctuations and bulk water: ( 1 ) To calculate and other binding energies we begin by effectively ‘morphing out’ the positional constraint on the ligands around one of the ions , . This is achieved by having two sets of ligands in the simulation: one set where each atom is subjected to a harmonic constraint ( ) and one with no harmonic constraint ( ) . ( 2 ) The system with no harmonic constraint , ( but still with a 3 . 5 Å radial constraint ) can now undergo an exchange reaction with another ion bound in bulk water: ( 3 ) This exchange reaction consists itself of two separate FEP MD simulations: ( 4 ) ( 5 ) The values for used here were calculated from the free energies of solvation by Joung and Cheatham [38] . Now the constraints can be morphed into the model system containing : ( 6 ) The change in exchange free energy for the overall reaction is then given by ( 7 ) will be positive if is thermodynamically preferred in the ion binding site and negative if is preferred . This quantity can also be studied as the thermal fluctuations ( i . e . the value of ) is reduced without reference to the energy involved in bringing the ion into the site from the bulk . The contribution to ion selectivity due only to the RLF mechanism alone , , can be defined as: ( 8 ) Abstract model systems consisted of abstract ligands ( based on formaldehyde with partial charges of carbon +0 . 5 , oxygen −0 . 5 and hydrogen 0 . 0 ) where each oxygen atom is confined to a 3 . 5 Å sphere by use of a spherical flat-bottomed , steep harmonic potential constraint . Reducing the ligand fluctuations was achieved by confining each atom to a harmonic potential , varying the force constant , , from to in 0 . 5 increments . This harmonic potential was placed at the vertices of optimal coordination polyhedra for , and at geometries governed by packing circles on spheres for [37] , at a distance from the ion corresponding to the first peak in the radial distribution function of a harmonically unconstrained ( but still with the 3 . 5 Å spherical constraint ) simulation determined for each ion type . A harmonic restraint is used for this purpose as a first order approximation; the restraining potentials exhibited in nature would probably be somewhat anharmonic and anisotropic . Two sets of ligands , i . e . , are required in order to conduct FEP MD between harmonically constrained and harmonically unconstrained ligands with one set annihilated and the other set exnihilated during the simulation . However , both endpoints represent ligands coordinating to an ion . Errors in were minimised by using a large number of windows , with values of , then for then then incrementing by 0 . 05 to , then for then . Forward and reverse morphs were conducted for each ion/model/ combination . The maximum difference in the forward and reverse morphs for was 0 . 94 kcal/mol , with an average difference of 0 . 27 kcal/mol . The maximum error in ( the summation of errors from , and ) is estimated to be 1 . 1 kcal/mol with an average error of 0 . 62 kcal/mol . Energies were averaged over 4 ns for each window . Softcore potentials were utilised using a van der Waals radius shift coefficient of 1 . A cut-off distance of 12 Å and a switching distance of 10 Å is used for electrostatic and van der Waals interactions . FEP MD simulations where the ion is being morphed used only one set of ligands , as the ligands are not bound by a harmonic constraint . was varied from 0 to 1 in 0 . 05 increments . Energies were averaged over 4 ns for each window . Softcore potentials were utilised using a van der Waals radius shift coefficient of 1 . A cut-off distance of 12 Å and a switching distance of 10 Å is used for electrostatic and van der Waals interactions . All simulations were conducted using NAMD2 [39] with the CHARMM27 force field [40] at 310 K with 1 fs timesteps . Force field parameters for Li+ , Na+ and K+ are from Joung and Cheatham [38] . The volumes occupied by the coordinating ligands were calculated using the VolMap tool in VMD [41] , with a resolution of 0 . 1 Å , and an in-house Fortran program with an isosurface value of 1 . 0 . FEP MD simulations were conducted identically to the harmonically constrained abstract models and the harmonically unconstrained models discussed in the previous section . To set up these systems the positions of the atoms coordinating to Na+ in each ion binding site , along with the atoms directly bonded to these and the ions themselves were extracted from the crystal structures of GltPh , PDB accession code 2NWX [42] , and LeuT , PDB accession code 2A65 [43] . These four model sites were then energy minimised with Li+ , Na+ and K+ present as the central ion . These minimised structures provided the initial starting coordinates for further simulations along with the coordinates to which the harmonic constraints were placed ( i . e . ) . was calculated by extracting the average total potential energy from the first and last window of each FEP MD simulation and combining them in a fashion as described for in the theory section . was calculated for each using .
The ion selectivity of the abstract models , including contributions from the RLF mechanism , is plotted versus the force constant , , for binding sites with 4–8 ligands in Fig . 1 and 2 . For comparison we also plot two sets of results for the strained cavity mechanism , that is , when the same location of the restraint is used for both ions rather than using an ‘optimal’ position for each ion type . Fig . 1 demonstrates that each abstract model system displays an inherent selectivity for K+ over Na+ when there is little or no restraint on the fluctuation of the ligands , in line with results from previous studies by ourselves [12] and others [8] . It must be stressed that this inherent selectivity is only for this particular type of ligand . Naturally , as increases the strained cavity models become more selective for the ion to which the positions of the ligands are optimised ( blue , Na+ , and red , K+ , lines in Fig . 1 ) . This effect can be quite large ( tens of kcal/mol ) and tends to plateau for the largest values of tested in this study , where the strained cavity becomes a ‘rigid’ cavity . The selectivity of the abstract models for Na+/Li+ is a little more complicated than for Na+/K+ . Each abstract model displays an inherent selectivity for Na+ when there is little or no positional constraint on the coordinating ligands , as Fig . 2 shows . The effects of introducing the strained cavity begin to show as the strength of the restraint increases; the positions to which the ligands are constrained determine the selectivity ( green , Li+ and blue , Na+ , lines ) . However , there is some anomalous behavior , especially for the six ligand case ( Fig . 2 C ) where strong restraints at both Li+ and Na+ optimised positions increase selectivity toward Na+ . A more subtle situation arises when the position of the restraint is different for each ion , i . e . when we consider the RLF mechanism without any strain . Although the difference looks small on the scale of Fig . 1 , a 2–5 kcal/mol increase in selectivity toward Na+ occurs for the exchange reaction with K+ with ligands when the size of the thermal fluctuations is reduced ( increased ) . The majority of this change occurs for between and , plateauing for ( see Fig . 3 to see this plotted in a different scale ) . This change in ion selectivity alters these models from K+ selective sites to Na+ or non-selective sites . For , the selectivity in the already K+ selective site is further enhanced by 2–3 kcal/mol . A similar situation arises in the exchange reaction between Li+ and Na+ ( Fig . 2 ) . Again , the most drastic changes occur for the strained cavity model as increases . The changes in selectivity due the RLF mechanism are again smaller than those for the strained cavity but the trends are similar to that seen for Na+/K+ in the cases with 4 or 5 ligands , with increasing moving selectivity toward Li+ . The situation with is more confusing with a reduction in fluctuations causing selectivity toward Na+ for . The model also has a much larger change in ( 29 kcal/mol ) than the models ( 3–4 kcal/mol ) . Also , the RLF results in the models do not fall within the bounds of the strained cavity results . Having shown that restricting the fluctuations in the positions of the ligands creates selectivity for one ion over another even in the absence of strain , the question remains , how does this occur ? If we decompose into the enthalpic , , and entropic , , components the driving force behind this change in selectivity becomes apparent . In the exchange between Na+ and K+ ( Fig . 3 ) for and Na+ and Li+ ( Fig . 4 ) for , the contribution follows very closely with indicating this selectivity is largely due to entropy differences . Intuitively one would expect the change in the available number of states ( as you decrease the allowed fluctuations ) to be largest for the larger ions for the following reasons . The number of possible configurations for coordination in the ( only bound by a 3 . 5 Å constraining sphere ) is greater for the larger ion than the smaller ion because of the greater volume available at the larger ion-ligand distance , as depicted in Fig . 5 . As the positional restraint is increased ( ) , the number of states become approximately equal for different ion types . Hence the change in entropy between a non-restrained and restrained system is largest for the larger ion . This can be shown to be the case by considering the difference in volume sampled by the coordinating oxygen atoms as their fluctuations becomes more and more constrained . For instance , this change in volume for the four fold coordination state is 3640 Å3 for Li+ , 5050 Å3 for Na+ and 5820 Å3 for K+ when comparing and . Reducing the thermal fluctuations on the ligands causes a greater decrease in entropy when they coordinate a larger ion compared to a small one . As a consequence , this reduction of thermal fluctuations favours small ions binding in the site . A different situation arises with for K+/Na+ systems and for the Li+/Na+ systems . In the former both the enthalpic and entropic terms play a role , while the latter is dominated by the enthalpic contribution . Analysis of these situations shows that the reason for the different behaviour is due to the difficulty in packing a large number of ligands around the smaller ions . As the 3 . 5 Å constraining sphere does not precisely define the coordination numbers , it is possible for ligands to form a second coordination shell about the small ions when the number of ligands is large . Increasing the force constant , , brings all the ligands to a uniform distance , yielding enthalpic changes in addition to the entropic changes seen for the cases with fewer ligands . The motivation for proceeding with this investigation was the realisation that in at least two amino acid transporters , the aspartate transporter , GltPh , and the leucine transporter , LeuT , the ligands forming the two Na+ binding sites display reduced RMS fluctuation in their positions compared to similar atom types elsewhere in the protein . The RMS fluctuation of the oxygen atoms forming the four sites ranged between 0 . 3 and 0 . 5 Å , whereas the other oxygens in the protein had values larger than 0 . 7 Å [12] , [34] . This is thought to be the result of extensive hydrogen bonding networks in the vicinity of the ion binding sites , as shown to be the case with one of the LeuT sites [44] . Additional constraint may be imparted upon the coordinating ligands if they belong to an amino acid in a more rigid secondary structure , such as backbone carbonyl oxygens of -helices . The ion binding sites in both LeuT and GltPh contain many of these backbone carbonyl oxygen atoms ( table 2 ) . More generally , there may also be sterical effects that limit the motion of the coordinating ligands . Attempts have been made to explain the Na+ selectivity in LeuT . Yamashita et al . [43] suggested that it could be a result of a more snugly fitting site for Na+ than the larger K+ , which would upset hydrogen bonding or packing interactions in the protein . This is in line with the strained cavity mechanism described by Lockless et al . [28] in K+ channels . Other investigations suggest that the first binding site ( Na1 ) achieves Na+ selectivity over both Li+ and K+ due to the strong electrostatic interactions resulting from the coordinating carboxylate ligands , while the second binding site ( Na2 ) achieves this through a strained cavity mechanism [10] , [44] . To the best of our knowledge , similar investigations have not been undertaken for GltPh . Therefore , we investigate the effect on ion selectivity of reducing the fluctuations in the ligands forming each binding site in the transporters by constructing corresponding model systems . The model sites for GltPh were constructed from the outward facing , Na+ and aspartate bound crystal structure [42] . Only two ( Na1 and Na2 ) of the three Na+ binding sites are considered , as the exact nature of the third is still under debate [45]–[47] . Models for the two LeuT Na+ binding sites were constructed from the Na+ and leucine bound crystal structure [43] . For each model , the coordinating atom , and atoms bonding directly to these , were used to construct simple dipolar ligands in order to model the electrostatic environment experienced by the bound ion . The composition of each model is detailed in table 2 . The initial coordinates of these atoms were taken from the crystal structure and then allowed to energy minimise with Li+ , Na+ and K+ independently . This gave us the final optimal coordinates for each ion type at which harmonic constraints were applied . Simulations were conducted to investigate RLF as described earlier for the abstract ligand models . As the amount of allowed fluctuation in the ligands of the amino acid transporter models are reduced ( increased ) , the change in the free energy of the exchange reaction between two ions ( ) behaves in a very similar manner to the abstract models; the decrease in fluctuation contributes selectivity to the smaller of the two competing ions ( Fig . 6 ) . If we recall that the most of the oxygen atoms in GltPh and LeuT displayed RMS fluctuations greater than 0 . 7 Å , we see from Fig . 6 C and D that there is little to no contribution toward ion selectivity in this region . However , this contribution becomes significant for RMS fluctuation values observed for the oxygen atoms at the ion binding sites ( the grey regions in Fig . 6 C and D ) . In this model , the ligands are able to adopt a preferred ion-ligand distance , and at no energy cost ( in contrast to the strained cavity mechanism ) , yet a degree of ion selectivity is still created by the reduction in ligand fluctuation . A decomposition of the free energy change in each of the sites into the enthalpic and entropic contributions clearly demonstrate that this effect is primarily a consequence of entropy differences ( Fig . 7 ) . It is evident from the non-zero values of at large RMS fluctuations ( small ) of the ligands in Fig . 6 that the chemical nature of the ligands and/or coordination numbers play a role in creating ion selectivity in the ion binding sites of LeuT and GltPh . As the RMS fluctuations decrease , the contribution from RLF merely adds to this . Nevertheless , in the absence of a strained cavity , it is crucial for enhancing selectivity for Na+ over K+ in LeuT . While a strained cavity and the chemical nature of the ligands may play a role in creating selectivity in themselves , we hope to show here that the observed selectivity could also involve a contribution from the RLF mechanism . Note that the 8 ligand coordinated Na+/K+ abstract model is very similar to the crude model S2 K+-selective binding site in the selectivity filter of the potassium ion channel KcsA investigated previously by Thomas et al . [13] There is a very slight to no increase ( kcal/mol towards K+ selectivity ) in the region corresponding the RMS fluctuations ( 0 . 75 Å ) [8] of the filter , suggesting that the RLF mechanism does not play a role in KcsA . Of course this result for KcsA depends on Na+ and K+ binding at the same sites in the selectivity filter; a view which has been challenged with the proposal of distinct binding sites for the two ions [48]–[50] . It should be noted that our study of the RLF mechanism and previous work by Yu et al . [10] differ in one very significant way . In the latter , constraints are placed at the crystallographic coordinates for the Na+ model binding sites for LeuT and the K+ model binding site for KcsA . This means that there is only one set of constraint positions for both ion types ( Na+ and K+ ) and thus their analysis includes the influence of a strained cavity , which is to say that a change in enthalpy , as well as entropy , will influence selectivity . While we do not deny that such an effect may play a role , we have isolated the RLF mechanism by allowing the ligands to freely adapt to each ion type . This means that the positions of the constraints on the ligands are optimal for each ion type , eliminating any ‘strained cavity’ effect from this analysis . The simple models of the ion binding sites in LeuT and GltPh were not designed so as to quantitatively reproduce experimental and more detailed simulation results , only to show how the RLF mechanism may influence the overall selectivity . Other factors may become important when considering the selectivity of the protein as a whole , such as the coupling between the ion selective sites [51] . However , these simple models are able to qualitatively reproduce experimental [43] and more detailed simulation [44] results for Na+/K+ selectivity in LeuT . In fact , Na2 changes from a K+ to a Na+ selective site when the reduction in the ligand fluctuations are accounted for . When compared to experimental [42] and more detailed simulations [12] , Na+/K+ selectivity in GltPh is qualitatively reproduced for Na1 , while Na2 is rendered essentially non-selective with the RLF effect . A table comparing results from this study to experimental and more detailed simulations can be found in text S1 . What conclusions can we draw from this ? Given that the amino acid transporter model binding sites are exceedingly simple , any conclusion drawn will be tentative . However , even though these models may be crude , they do demonstrate that reducing the fluctuation of the coordinating ligands , can affect ion selectivity even if there is no strain in the protein . Again it is shown that the RLF mechanism is primarily a consequence of entropy differences . As this mechanism relies heavily on entropic factors , experimental investigations into the temperature dependence of ion selectivity in these amino acid transporters could perhaps shed further light on its role in biological systems . Reducing the thermal fluctuation in the positions of the coordinating ligands affects the binding of Li+ , Na+ and K+ differently and is able to contribute toward ion selectivity , even when there is no strain associated with the protein adapting to different ions . This contribution to ion selectivity is due to entropic differences arising with different ions in the site , resulting from the larger difference in accessible states for the ligands surrounding the larger ions than the small ones when the thermal fluctuations are reduced . Thus , this mechanism of ion selectivity favours of small ions over larger ions . | Differentiating between Na+ and K+ ions is important for many cellular processes , such as nerve conduction and the regulation of membrane potentials . Different biological molecules utilise different methods to discriminate between ions . In this work , the reduced ligand fluctuation mechanism of ion selectivity is described . This entropy-driven mechanism is due to the limited thermal fluctuations of the atoms in some macromolecular ion binding sites . The elucidation of this mechanism offers a more complete picture of the ways in which the fundamental process of ion selectivity can be achieved . | [
"Abstract",
"Introduction",
"Methods",
"Results/Discussion"
] | [
"molecular",
"mechanics",
"molecular",
"dynamics",
"chemistry",
"biology",
"computational",
"chemistry",
"biophysics",
"simulations",
"biophysics"
] | 2013 | An Entropic Mechanism of Generating Selective Ion Binding in Macromolecules |
Semi-conservative segregation of nucleosomes to sister chromatids during DNA replication creates gaps that must be filled by new nucleosome assembly . We analyzed the cell-cycle timing of centromeric chromatin assembly in Drosophila , which contains the H3 variant CID ( CENP-A in humans ) , as well as CENP-C and CAL1 , which are required for CID localization . Pulse-chase experiments show that CID and CENP-C levels decrease by 50% at each cell division , as predicted for semi-conservative segregation and inheritance , whereas CAL1 displays higher turnover . Quench-chase-pulse experiments demonstrate that there is a significant lag between replication and replenishment of centromeric chromatin . Surprisingly , new CID is recruited to centromeres in metaphase , by a mechanism that does not require an intact mitotic spindle , but does require proteasome activity . Interestingly , new CAL1 is recruited to centromeres before CID in prophase . Furthermore , CAL1 , but not CENP-C , is found in complex with pre-nucleosomal CID . Finally , CENP-C displays yet a different pattern of incorporation , during both interphase and mitosis . The unusual timing of CID recruitment and unique dynamics of CAL1 identify a distinct centromere assembly pathway in Drosophila and suggest that CAL1 is a key regulator of centromere propagation .
Centromeres are the chromosomal regions that mediate correct assembly of the kinetochore , a multi-protein structure necessary for attachment to spindle microtubules and faithful chromosome segregation in mitosis and meiosis . Centromeres are composed of DNA associated with nucleosomes that contain the H3 variant CENP-A ( CID in Drosophila ) , and numerous constitutively bound centromeric proteins [1] . Specific underlying DNA sequences are neither necessary nor sufficient for centromere function in many eukaryotes , in contrast to the requirement for conserved , centromere-specific proteins such as CENP-A [1] . Accurate chromosome segregation also requires that the number and positions of centromeres be stably inherited through cell and organismal generations . DNA replication in mid to late S phase generates two copies of centromeric DNA [2] , [3] , but little is known about how passage of the replication fork affects the integrity of centromeric chromatin , how centromeric proteins are redistributed , and how intact centromeres are recreated after replication and accompanying nucleosome dilution . CENP-A assembly does not require DNA replication , in contrast to the replication-dependence of histone H3 deposition [2] , [4] . Surprisingly , the timing of CENP-A replenishment during the cell cycle is not the same in different eukaryotes . In human HeLa cells , newly-synthesized CENP-A protein is recruited to centromeres during late telophase and G1 , and requires mitotic exit [5] . GFP-CID and GFP-CENP-C recruitment in Drosophila syncytial embryos is initiated earlier in mitosis during anaphase . Interestingly , anaphase loading is not observed in later embryonic stages [6] , where the cell cycle timing of loading has not been determined . GFP-CID was also previously reported to be deposited in G2 phase in Drosophila Kc167 cells [4] . What is conserved between Drosophila and human cells is that there is a delay between centromeric DNA replication ( S phase ) and CENP-A replenishment ( mitosis or G1 ) . Interestingly , this means that the main function of centromeres , i . e . kinetochore assembly and chromosome segregation in mitosis , occurs with only half of the maximal amount of CENP-A in these organisms [5] . In contrast , in organisms such as S . pombe [7]– , Arabidopsis [10] , and Dictyostelium [11] the kinetochore is assembled on chromatin containing a full CENP-A complement , which has led to the proposal that mitotic and post-mitotic CENP-A recruitment may have been acquired more recently during evolution [11] . Finally , the signaling event ( s ) responsible for initiating centromere replenishment have not been identified in any of these organisms . Despite differences in the timing of centromere replenishment in the cell cycle , both S . pombe and human cells contain homologous proteins that are essential for CENP-A assembly , specifically the Mis18 complexes and the CENP-A partner Scm3/HJURP [12]–[16] . The timing of CENP-A assembly in human cells approximately coincides with centromere localization of HJURP [12] , [13] . The human Mis18 complex ( which contains hMis18α , hMis18β and M18BP/hKNL2 ) is recruited at centromeres at the end of mitosis [12] , [13] , [17] , [18] , slightly before CENP-A and HJURP [13] , [19] and has been proposed to ‘prime’ centromeres to receive new CENP-A [17] . Studies of the turnover of several constitutive centromere proteins indicated that CENP-C displays dynamic exchange in G1 and G2 [20] . The timing and mechanisms controlling the replenishment of additional constitutive centromeric components ( e . g . the CCAN [21] , [22] ) in human cells are not known . Functional screens and database searches have failed to identify hMis18 , M18BP1 or Scm3/HJURP homologs in Drosophila [16] , [18] , [23] , [24] , so it is unclear whether non-homologous proteins perform analogous functions in this organism . Centromeric CID localization in Drosophila requires CENP-C , CAL1 , Cyclin A and Rca1 [23] , [24] . CID , CAL1 and CENP-C interact physically , and are interdependent for centromere localization . CAL1 is particularly intriguing because it has only been found in Dipteran species , displays unusual localization dynamics in mitosis [23] , and appears to limit how much CID and CENP-C can be incorporated at centromeres [25] . The anaphase recruitment of CENP-A and -C observed in Drosophila syncytial embryos may not be representative for this organism , since these embryos undergo very fast ( 8–10 min ) nuclear divisions that lack G1 and G2 phases . Furthermore , the distribution to daughter cells and turnover of CID , CAL1 and CENP-C have not been determined . Thus , establishing the cell cycle dynamics of nascent centromere protein assembly in cells with a complete cell cycle is crucial for determining the general mechanism of centromere replenishment in flies . Here , we report the dynamics of centromere protein redistribution and replenishment in Drosophila tissue culture cells . The SNAP tag system [5] was used to track the behavior of CID , CAL1 and CENP-C in pulse-chase and quench-chase-pulse experiments . CID and CENP-C are stable proteins , whose centromeric levels decrease by 50% at each cell division , as predicted for semi-conservative segregation and inheritance of centromere components . In contrast , CAL1 is less stably associated with the centromere because its centromeric levels decrease 66% after one cell cycle . Tracking of newly-synthesized protein demonstrates that CID is recruited during metaphase in Drosophila tissue culture cells , prior to what was reported for early fly embryos and HeLa cells , while CENP-C is recruited during both mitosis and interphase . Interestingly , CAL1 is recruited at centromeres during prophase , before CID , and , similarly to HJURP and CENP-A , physically associates with pre-nucleosomal CID [13] . These findings establish the temporal events leading to centromere replenishment in Drosophila cultured cells , and provide evidence that CAL1 plays a crucial role in targeting CID to centromeres .
Faithful transmission of the centromere locus depends on re-assembly of centromeric components that have been either diluted or disrupted by centromeric DNA replication . Semi-conservative inheritance of CENP-A was previously demonstrated in HeLa cells [5] . To establish how CID , CENP-C and CAL1 are inherited during cell division , we generated clonal cell lines expressing SNAP-CID , SNAP-CAL1 or SNAP-CENP-C fusion proteins ( see Materials and Methods ) . The three centromeric proteins were expressed from the Copia promoter at comparable levels to the endogenous proteins [23] , exhibited the expected centromeric localization , and did not perturb normal chromosome segregation ( data not shown ) . The functionality of SNAP-CID was confirmed by the observation that segregation and viability defects associated with RNAi depletion of endogenous CID were rescued by SNAP-CID expression ( Figure S1 ) . SNAP-tagged proteins were pulse-labeled using the cell permeable fluorescent substrate tetramethylrhodamine ( TMR ) . TMR reacts with both centromeric SNAP and any SNAP from a soluble protein pool . Therefore , to determine the turnover of the pre-existing centromere pool , TMR signal intensity at centromeres was quantified 24 h after labeling ( Day 1 ) and again after a 24 hr chase ( i . e . after one doubling , Day 1 , Figure 1A and Materials and Methods ) . This analysis showed that , on average , CID levels decreased to 49% after one cell division ( Figure 1B , 1C ) . The distributions of intensity values for Day 1 and Day 2 demonstrate that the averages reflect the behavior of individual cells ( Figure S2A ) . We conclude that pre-existing CID is distributed equally to daughter cells , as observed previously for CENP-A in HeLa cells [5] . These results also suggest that , on average , CID nucleosomes are segregated semi-conservatively to sister chromatids during DNA replication , but does not exclude conservative segregation of individual CID blocks . The redistribution of pre-existing CAL1 and CENP-C during cell division could be different from that of CID , since neither appear to be nucleosome components [23] , [25] , [26] . Pulse-chase experiments using SNAP-CENP-C showed that the average TMR signal at Day 2 was 48% of the intensity at Day 1 , indicating equal distribution of pre-existing CENP-C ( Figure 1C , Figure S2B ) . However , TMR signal on Day 2 was only 34% for CAL1 ( Figure 1C , Figure S2C ) , which is significantly lower than the 49% observed for CID ( p<0 . 0001 ) . Thus , 66% of pre-existing centromeric CAL1 is exchanged at each round of division . Similarly to CID , the distributions of intensity values for Day 1 and Day 2 for CENP-C and CAL1 are consistent with the averaged values ( Figure S3A , S3B ) . We conclude that pre-existing CID and CENP-C are stably retained at centromeres through cell division , and are diluted to levels consistent with 50∶50 segregation during replication in S phase . In contrast , CAL1 protein displays higher turnover ( see Discussion ) . Faithful centromere propagation requires that new CID , CENP-C and CAL1 are recruited with each cell cycle . To determine the timing of deposition of centromeric proteins , we performed quench-chase-pulse experiments on clonal S2 cell lines expressing SNAP-tagged CID and GFP-tubulin ( Figure 2A ) . Asynchronous cells were treated with the BTP-blocking agent to quench pre-existing SNAP-CID ( Materials and Methods and Figure S4A ) , chased for 1 , 2 , 10 or 24 h to allow synthesis of new SNAP-CID , TMR labeled ( pulse ) and imaged ( Figure 2A ) . All manipulations were carried out in conditioned medium ( see Materials and Methods ) , and FACS analysis showed that cell cycle distributions remained constant throughout the experiment ( Figure S4B ) . S2 cells have a cell cycle that is approximately 24 h long , and newly-synthesized CID is already detected by TMR labeling 1 h after the BTP-block . To establish the approximate cell cycle stage ( s ) when cells first display centromeric TMR signal , cells were scored for the presence or absence of the mitotic-specific marker phosphorylated histone H3 Ser10 ( phospho H3 ) [23] . Only 3–5% of S2 cells are in mitosis [27] , and the length of mitosis is approximately 30 min [23] . Surprisingly , at 1 h after the BTP-block , 53% of cells in mitosis contained newly-synthesized SNAP-CID at centromeres ( p<0 . 0001 compared to interphase ) , and 97 . 4% of all cells with centromeric TMR signal were in mitosis ( Figure S4C ) . At later time-points after the BTP-block ( 2 , 10 , and 24 h ) , the percentage of mitotic cells that were TMR positive increased to 70% , 83% and 89% , respectively ( Figure S4C ) , consistent with cells having more time to synthesize new CID protein before entering mitosis . Although 95–97% of S2 cells are in interphase , only 2% of interphase cells contained TMR labeled CID at the 1 h time-point ( Figure S4C ) . It is possible that some CID loading occurs in G1 , S or G2 phases . However , since approximately 10% of S2 cells are in G1 ( Figure S4A ) , loading exclusively in G1 would result in more than 2% TMR positive interphase cells . Furthermore , the efficiency of BTP-block was 97% in these experiments ( see Materials and Methods ) , which could by itself account for the low frequency of interphase TMR-CID signals . In addition , 6–10% of cells complete mitosis in a 1 h interval after addition of BTP-block , and were scored as interphase in our quantitation . Although the percentage of TMR positive interphase cells increased at later time points , the frequencies were much lower than observed for mitotic cells ( Figure S2C , p<0 . 0001 compared to mitosis for 2 h and 10 h ) , consistent with CID recruitment during the previous mitosis . Thus , although we cannot exclude that some CID incorporation occurs in G1 , as observed in human cells , the vast majority must occur in mitosis . CID loading in mitosis was also observed in a clonal Kc167 cell line expressing SNAP-CID , demonstrating that these results are not specific to S2 cells ( Figure S5 ) . To further assess the contribution of interphase to SNAP-CID loading , we quantified the number of interphase cells displaying TMR-CID in quench-chase-pulse experiments performed in S2 cells arrested with colchicine ( see Materials and Methods ) . Colchicine disrupts microtubules and thus prevents cells from exiting mitosis and re-entering the cell cycle; thus , interphase cells displaying centromeric TMR-CID must have recruited SNAP-CID without going through mitosis . Clonal SNAP-CID cells were treated with colchicine for 2 h , BTP blocked and chased for 4 h in the presence of colchicine after which they were subjected to TMR labeling . We observed that 64% of interphase cells did not contain any centromeric TMR-CID , and the remaining 36% contained minimal TMR-CID signal compared to mitotic cells ( Figure S6A ) . These results confirm that interphase contributes minimally to new CID loading and that the majority of interphase cells that were scored as TMR-CID positive in our time courses ( Figure 2 , Figure S4 ) were either non-BTP blocked cells or cells that re-entered the cell cycle after recruiting SNAP-CID in the previous mitosis . We conclude that CID loading occurs predominantly in mitosis in S2 and Kc167 cells , and that minimal CID recruitment occurs during interphase , which distinguishes Drosophila from human cells , where the majority of CENP-A is recruited to centromeres in G1 [5] . To more precisely determine the specific stage ( s ) of mitosis when new CID is recruited , microtubules ( GFP-tubulin ) , phospho H3 immunofluorescence ( IF ) and DNA morphology ( DAPI ) were used to identify cells in prophase , metaphase , anaphase , telophase and cytokinesis . The earliest mitotic stage where new CID was detectable was metaphase; 67% of metaphase cells and 0% of prophase cells were TMR positive 1 hr after the BTP-block ( p<0 . 0001; Figure S6B , Figure 2B and 2C ) . Cells in anaphase , telophase and cytokinesis also displayed TMR-labeled CID at centromeres ( Figure 2B ) . However , we observed that the total TMR intensity at centromeres did not increase between metaphase and cytokinesis ( Figure 2D ) . Thus , the presence of newly-synthesized SNAP-CID in cells in anaphase/telophase and cytokinesis likely results from CID loading in the previous metaphase , and not from ongoing CID recruitment during later mitotic stages . To determine if endogenous CID is recruited to centromeres in metaphase , as observed for SNAP-tagged CID , we compared the total CID intensity in metaphase and interphase using IF in S2 cells ( Figure 3 ) . S2 cells have approximately 13 chromosomes whose centromeres form 3–6 distinguishable foci throughout interphase . De-clustering occurs in early mitosis , making individual centromeres distinguishable . Because of centromere clustering in interphase , each centromere focus is composed of several centromeres , making quantitative comparison of individual centromere IF signals between interphase and metaphase cells unfeasible . Therefore , we quantified the total nuclear CID intensity in interphase and metaphase cells ( see Materials and Methods ) . If endogenous CID is recruited in metaphase , then G1 , S and G2 cells should have similar total amounts of CID at centromeres , and metaphase cells should have double that amount . Indeed , the mean total CID intensity of metaphase cells was approximately 2-fold higher than in interphase ( p<0 . 0001; Figure 3B ) , confirming that the cell cycle timing of endogenous CID recruitment is similar to that of SNAP-tagged CID . In summary , our data show that newly-synthesized CID is assembled at centromeres during metaphase in S2 cells , and that recruitment occurs in a discrete ‘pulse’ during this stage . We cannot exclude the possibility that the process initiates slightly earlier , e . g . in G2 or prophase , and that centromeric CID needs to accumulate through metaphase to be detectable by SNAP-labeling and CID IF . Our experiments collectively show that CID assembly in Drosophila S2 cells occurs earlier in mitosis than observed in syncytial embryos ( anaphase ) , and in human HeLa cells ( late telophase through G1 ) [5] , [6] . The timing of assembly for CID , CENP-C and CAL1 could differ , despite the fact that they physically interact and are interdependent for centromere localization [23] , [25] . Furthermore , the timing of CENP-C centromeric recruitment has only been analyzed in early syncytial Drosophila embryos , which have unusual cell cycles [6] . The cell cycle timing of new CAL1 and CENP-C recruitment to centromeres was determined by quench-chase-pulse experiments , using clonal S2 cell lines expressing SNAP-CAL1 or SNAP-CENP-C . Similar to what was observed for CID , newly-synthesized CAL1 was visible at centromeres 1 h after the BTP-block , predominantly in mitotic cells; 63% of mitotic cells contained CAL1 TMR signal at centromeres compared to only 1% of interphase cells ( p<0 . 0001 compared to interphase; Figure 4A , 4B , Figure S7 ) . The vast majority of all cells with centromeric TMR signal were in mitosis ( 99% , Figure S7 ) , since new CAL1 was only observed in 1% of interphase cells . This likely represents incomplete BTP blocking or CAL1 loading in the previous mitosis ( see above ) . In contrast to CID , analysis of specific stages of mitosis showed that new CAL1 protein was detected as early as prophase; 33% of prophase cells contain TMR-CAL1 1 h after the BTP-block ( p<0 . 0001 compared to interphase; Figure 4A , 4B ) . Quantitative comparison between cells in prophase , metaphase and cytokinesis showed no significant change in TMR-CAL1 centromeric intensity at these different mitotic stages ( Figure 4C ) . We conclude that new CAL1 is predominantly loaded in a single ‘pulse’ during prophase , prior to assembly of newly-synthesized CID in metaphase . Newly-synthesized CENP-C was not observed until 10 h after the BTP block in quench-chase-pulse experiments , at which point 47% of interphase and 50% of mitotic cells ( all stages of mitosis ) were positive for TMR ( Figure S8A , S8B ) . The longer time required to observe TMR signals for CENP-C , compared to CAL1 and CID , makes it difficult to distinguish CENP-C deposition in interphase from mitotic deposition in the previous cell cycle . In contrast , 10 h after BTP-block of pre-existing SNAP-CID and SNAP-CAL1 , approximately twice as many mitotic cells display new CID and CAL1 relative to interphase cells ( Figure S4C and Figure S7 ) . These results suggest that CENP-C recruitment occurs in both interphase and mitosis , and that new CENP-C recruitment occurs from a pool of soluble CENP-C , which is blocked by BTP-treatment and therefore not visible by TMR labeling after short chases . Thus , the dynamics of CENP-C recruitment in Drosophila cultured cells are similar to those observed in human tissue culture cells [20] , but differ from the dynamics observed in early Drosophila embryos , where CENP-C's recruitment was observed in anaphase [6] . The observation that CID recruitment is restricted to metaphase raises the question of what signals and mechanisms regulate this process . It was previously proposed that robust kinetochore/microtubules interactions could provide a ‘signal’ for centromere replenishment [28] , which would be consistent with CID loading during metaphase . We therefore examined whether new CID recruitment occurs in cells treated with the microtubule destabilizing drug colchicine . S2 cells have an intact spindle checkpoint [29] and respond to colchicine treatment by accumulating cells with condensed , phospho H3 positive chromosomes and no spindle microtubules . SNAP-CID cells expressing GFP-tubulin were incubated in the presence of colchicine , treated with BTP-block , chased and labeled with TMR to detect newly-synthesized CID ( Figure 5A ) . We observed that 84% percent of the colchicine-arrested cells contained TMR CID , compared to 92% of untreated ( control ) cells ( p = 1; Figure 5B ) . Thus , an intact mitotic spindle is not required for new CID recruitment in metaphase in S2 cells , similar to observations in HeLa cells and early Drosophila embryos [5] , [6] . However , in those systems loading occurs after anaphase onset , so the spindle assembly checkpoint ( SAC ) had to be inactivated to assess CENP-A/CID assembly in the presence of microtubule disrupting drugs . Since CID is loaded during metaphase in S2 cells , we can conclude that CID recruitment occurs independently of intact mitotic spindles , mitotic checkpoint satisfaction and chromosome segregation . Furthermore , given that colchicine treatment abolishes inter-kinetochore tension , our results also exclude the contribution of tension and chromatin stretching in promoting new CID deposition [28] . The observation that CID recruitment occurs independently of spindle function , checkpoint silencing and chromosome segregation , leaves open the question of what cell-cycle events regulate CID assembly in metaphase . Regulated ubiquitination of cyclins and other substrates followed by proteasome-mediated protein degradation are crucial to proper cell cycle progression , including the metaphase-anaphase transition [30]–[32] . Therefore , we analyzed whether proteasome inhibition by treatment with MG132 affects CID metaphase recruitment . SNAP-CID S2 cells were incubated for 2 h with 25 µM MG132 , treated with BTP-block , chased for 4 h with or without MG132 , and labeled with TMR . Treatment with MG132 did not affect the percentage of mitotic cells , however the percent of anaphases was significantly lower than in control cells ( data not shown ) . While control cells efficiently recruited new SNAP-CID ( 55% of metaphases displayed TMR-CID; Figure 6A , 6B ) , treatment with MG132 dramatically prevented efficient recruitment ( 5% of metaphases displayed TMR-CID , p<0 . 0001; Figure 6 ) . When the chase was carried out in the absence of MG132 , new CID loading returned to levels comparable to control cells ( Figure 6A , 6B ) . These results demonstrate that proteasome activity is crucial for efficient CID loading during metaphase , and suggest that one or more proteasome-targets must be degraded prior to or during metaphase in order to recruit new CID . One of the key events in metaphase is the Anaphase Promoting Complex ( APC ) -mediated ubiquitination of Cyclin A ( CYCA ) [30] , which is then targeted for destruction by the proteasome [31] , [32] . Cyclin A is centromere-associated and is required for CID localization [23] . Since MG132 inhibits new CID loading , we investigated whether CYCA degradation is required for new CID recruitment in metaphase , using a non-degradable CYCA mutant . S2 cells expressing SNAP-CID were transfected with a plasmid carrying CYCA lacking the destruction signals ( ND-CYCA ) , fused to GFP [23] , which cannot be degraded via the APC and causes a delay in metaphase and defective anaphases [33] . 24 h after transfection with ND-CYCA , cells were BTP-blocked , chased for 4 h and labeled with TMR to detect new SNAP-CID . Cells transfected with ND-CYCA showed a statistically significant decrease in the percent of cells in metaphase with new SNAP-CID compared to controls ( 77% versus 97%; p = 0 . 0011; Figure 7A , 7B ) . Furthermore , in cells transfected with ND-CYCA , the intensity of TMR-labeled SNAP-CID was significantly weaker in most cells ( p = 0 . 0256; Figure 7A and 7C ) . We conclude that degradation of Cyclin A contributes to the process of new CID loading . The observation that MG132 treatment has a more dramatic effect suggests that CYCA is not the only proteasome-mediated degradation target involved in CID loading in metaphase . In summary , these observations demonstrate that proteasome-mediated degradation of CycA and other key targets is essential for centromere assembly in flies . CAL1 displays functional similarities with two different sets of proteins required for CENP-A loading in human cells . CAL1 and components of the hMis18 complex are recruited to centromeres before CID and CENP-A , respectively , and exhibit similar dynamics in time-lapse analysis ( loss from centromeres during mitosis [23] ) . However , hMis18 proteins do not bind CENP-A , whereas CAL1 physically associates with CID on chromatin and in yeast two hybrid assays [23] , [25] . This property is instead shared with HJURP , which binds to human CENP-A in both chromatin and chromatin-free extracts [12] , [13] . These observations led to the proposal that HJURP is a chaperone that facilitates targeting of new CENP-A assembly to centromeres [12] , [13] , [19] , [34] . To determine if CAL1 also associates with pre-nucleosomal CID , we analyzed the distributions of CAL1 , CID and CENP-C in different cellular fractions ( chromatin-free ( S1 ) , chromatin-bound ( S2 ) , histone-containing ( S3 ) and nuclear-matrix bound , insoluble material ( S4 ) ) prepared from S2 cell lines stably expressing FLAG-tagged CID ( Figure 8A ) . The majority of CID and CENP-C protein were present in the S2 , S3 and S4 fractions , whereas CAL1 protein was detectable in all four fractions , including chromatin-free extracts ( S1 ) ( Figure 8B ) . Immunoprecipitation with FLAG beads from chromatin-free extracts clearly identified the presence of CID ( Figure 8C ) . Furthermore FLAG-CID specifically pulled down CAL1 , indicating that FLAG-CID and CAL1 interact in pre-nucleosomal complexes ( Figure 8C ) . In contrast , CENP-C was not present in these FLAG-CID precipitates , indicating that CENP-C association with CID and CAL1 is limited to chromatin-bound complexes ( Figure 8C; [23] , [25] ) . CAL1 is the first protein to be shown to bind to pre-nucleosomal CID in Drosophila . The finding that pre-nucleosomal CID interacts with CAL1 raises the possibility that CAL1 may be acting as a CID chaperone , targeting CID to centromeres at the appropriate cell cycle phase in a manner similar to HJURP ( see Discussion ) .
We describe the cell cycle dynamics of three essential centromere components in Drosophila cells . CID , CAL1 and CENP-C display different turnover and assembly dynamics , despite the fact that these essential centromeric components interact physically , and are interdependent for centromere localization [23]–[25] . SNAP-tagged CID , CAL1 and CENP-C are expressed from the identical Copia promoter [23] , thus it is unlikely that these distinctions are due to different rates of new protein synthesis . Using a pulse-chase strategy , we show that CID levels are reduced by ∼50% after one cell cycle , which could result from semi-conservative distribution of pre-existing CID nucleosomes , or random redistribution of parental CID-H4 tetramers [35] , to replicated sister chromatids . While CID and CENP-C display stable association with centromeres and 50∶50 distribution after each cell cycle , 66% of TMR-CAL1 is replaced by new protein . Thus , CAL1 is either less stably bound , or its replenishment involves partial removal of pre-existing protein . Alternatively , CAL1 could undergo an even higher turnover and our quantification could be an underestimation; CAL1 could be entirely recruited de novo and the measured centromeric TMR-CAL1 could reflect recruitment from an initial soluble pool at the time of labeling . An additional difference is that while SNAP-CID and CAL1 are detectable at centromeres 1 h after quenching the SNAP epitopes , 10 h of chase time are necessary for CENP-C to be visible by TMR labeling . This suggests that at each cell cycle the recruitment of CID and CAL1 relies for the most part on newly-synthesized protein , while CENP-C recruitment also involves a pre-existing non-centromeric or soluble pool . Indeed , the cellular fractionation analysis demonstrated the presence of low levels of CENP-C in chromatin-free extracts ( Figure 8B ) , supporting the possibility that there is a soluble pool of CENP-C available to replenish the centromere-associated CENP-C diluted during the cell cycle . CENP-C is targeted to centromeres during multiple cell cycle stages , consistent with previous findings in human cells [20] . In contrast , newly-synthesized CAL1 and CID are recruited to centromeres during discrete stages of mitosis . Using quench-chase-pulse time-courses in both asynchronous and arrested cultures , we demonstrate that the contribution of interphase to CID loading is minimal , since the percent of interphase cells displaying newly-synthesized SNAP-CID and the signal intensity of TMR-CID differ dramatically from those measured for mitotic cells ( Figure 2 and Figure S6A ) . These observations distinguish Drosophila from human HeLa cells , where CENP-A is recruited during G1 [5] , from fission yeast , where CENP-A assembles at centromeres in both S and G2 phases [7] , [8] as well as from plants ( G2 ) and Dictyostelium ( G2/prophase ) [10] , [11] . Both new CID and CAL1 are assembled at centromeres in mitosis , but each protein is recruited during discrete stages: prophase for CAL1 and metaphase for CID . It is possible that CID and CAL1 loading are initiated simultaneously in prophase , but CAL1 levels accumulate faster than CID at centromeres . Regardless , the observed temporal distinction suggests that CAL1 acts upstream of CID recruitment ( summarized in Figure 9 ) . Incorporation of nascent CAL1 at centromeres during prophase could be mediated by binding to pre-existing centromeric CID and CENP-C . This could in turn promote new incorporation of nascent CID during metaphase , either by gap-filling or exchange of space-holder histone H3 ( Figure 9 ) . Interestingly , a similar temporal distinction has been described for the human centromere proteins hMis18α , β and M18BP , which localize to centromeres in anaphase , before new CENP-A assembly in late telophase/G1 [5] , [13] , [17]–[19] . The lack of any physical interaction between hMis18α , β , M18BP and CENP-A [12] , [13] , [17] , [21] , and the observation that hMis18α can localize to centromeres even if CENP-A is depleted [17] , has led to the proposal that this complex may ‘prime’ centromeres to receive new CENP-A [17] from the HJURP chaperone , whose centromeric targeting coincides temporally with deposition of new CENP-A [12] , [13] . Homologs for hMis18 complex components and HJURP ( or the budding and fission yeast Scm3 homologs ) have not been identified in the Drosophila genome . Collectively our data support a model in which CAL1 performs functions attributable to both HJURP and hMis18 , despite the lack of sequence homology . hMis18 proteins are recruited to centromeres before CENP-A [5] , [13] , [17]–[19] , and CAL1 loading precedes CID assembly . However , the hMis18 complex does not interact with CENP-A , whereas CAL1 and CID are associated in chromatin-free extracts , identifying the first Drosophila protein that binds CID in its pre-nucleosomal form . HJURP also interacts with pre-nucleosomal CENP-A [13] , and both HJURP and CAL1 strongly colocalize with nucleoli [12] , [23] , [25] . Thus , CAL1 could ‘prime’ the centromere in prophase , and also mediate CID recruitment directly in metaphase ( Figure 9 ) . We previously showed that gross-levels of centromeric GFP-CID and GFP-CENP-C did not visibly change through the cell cycle in time-lapse analysis [23] , consistent with the 50∶50 segregation observed here during one division . In contrast , GFP-CAL1 levels were significantly reduced in metaphase , increased again in telophase , and remained stable through interphase [23] . The transient reduction in GFP-CAL1 levels at metaphase is intriguing , given that it coincides with new CID assembly . The observation that newly assembled TMR-CAL1 intensities were constant from prophase to cytokinesis ( Figure 4C ) suggests that most of the GFP-CAL1 reduction at metaphase and increase at telophase involves pre-existing protein . One model to account for these observations is that free CAL1 ( not bound to CID ) is recruited to centromeres in prophase where it performs a yet undefined ‘priming’ function; then , the subset of CAL1 bound to pre-nucleosomal CID escorts it to centromeres in metaphase while ‘old’ CAL1 is displaced ( Figure 9 ) . The interdependency of CAL1 , CID and CENP-C in centromere localization [23] could be explained by the requirement of pre-existing CID and CENP-C for CAL1 assembly in prophase . The loading of CID and CAL1 in specific , early stages of mitosis also raises questions about the nature of the signal ( s ) that initiate assembly of centromeric chromatin . Centromere replenishment signaling by kinetochore-microtubule interactions [28] is inconsistent with our demonstration that CID loading in metaphase is not affected by colchicine treatment , and therefore does not require spindles ( as also observed in human cells [5] ) , SAC inactivation , chromosome segregation , or inter-kinetochore tension . However , we previously showed that premature activation of the Anaphase Promoting Complex , by Cyclin A or RCA1 depletion , interferes with CID localization to centromeres , demonstrating that centromeric chromatin assembly is linked to key regulators of mitotic progression [23] . Interestingly , Cyclin A localizes to centromeres and is degraded in metaphase [23] , [36]; here we demonstrate that metaphase loading depends on proteasome activity , which could include degradation of key mitotic regulators . MG132 treatment prior to BTP block prevented CID loading ( Figure 6 ) while transfecting cells with a non-degradable form of CYCA abrogated new CID recruitment in a subset of cells ( 23% , Figure 7B ) , and TMR-CID levels were significantly reduced in most cells . One possibility to explain the stronger impact of proteasome inhibition is that proteasome targets in addition to CYCA need to be degraded for efficient CID deposition . Alternatively , the presence of centromeric endogenous CYCA , which is probably degraded normally in the presence of excess ND-CYCA , might trigger a sufficient signal to initiate CID incorporation in some cells . Interestingly , Cyclin A is degraded in the presence of microtubule drugs and escapes inhibition of the APC by the SAC [37] , which would explain why new CID recruitment takes place efficiently in the presence of colchicine ( Figure 5 ) . Proteasome and ubiquitin-ligase activities have been implicated in controlling proper CENP-A centromeric incorporation by degradation of euchromatic CENP-A in budding yeast and Drosophila [38]–[41] . Understanding the relationship between the CENP-A degradation pathway and our implication of proteasome activity in the recruitment of nascent CENP-A will require further investigation . It is unclear at this point how degradation of CYCA contributes to CID assembly . One possibility is that high CDK-CYCA activity at the centromere inhibits CID recruitment , and that local inhibition of CDK activity through degradation of CYCA or other substrates triggers CID assembly . Understanding the role of degradation of Cyclin A and other APC and proteasome substrates in CID recruitment will be crucial to elucidating how centromere assembly is coupled to the cell cycle . The dynamics of centromere replenishment in Drosophila cultured cells differs from those observed in S . pombe [7] , [8] and human HeLa cells [5] . Early syncytial fly embryos display slightly later recruitment of new CID in anaphase , but this difference could be due to the unusually short nuclear cycles that lack G1 and G2 phases [6] . Although CENP-A loading in HeLa cells is first observed in telophase , it is possible that the primary signal to initiate CENP-A loading ( e . g . inhibiting local CDK-CYCA activity at the centromere ) is conserved , and occurs during prophase or metaphase in both Drosophila and human cells . It is also puzzling that key proteins required in trans for CENP-A assembly , such as HJURP and CAL1 , are not always conserved , in contrast to the universality of centromeric chromatin components such as CENP-A and CENP-C [1] , [16] , [18] , [23] , [24] , [34] . It is possible that highly diverged proteins , such as CAL1 , perform the same function ( s ) as human regulators such as HJURP and hMis18 . Thus , although our data challenges the universality of centromere propagation dynamics in metazoans , it will be important to determine whether some mechanisms and signals required for CENP-A replenishment are conserved , despite different times of assembly in the cell cycle , and the lack of conservation for key regulatory proteins .
S2 cells were co-transfected with either pCopia-SNAP-CID , SNAP-CAL1 , or SNAP-CENP-C and pCoHygro ( Invitrogen ) and pAC-GFP-tubulin ( gift of G . Goshima ) using the Cellfectin reagent ( Invitrogen ) . Polyclonal stable cell lines were generated by hygromycin selection . Clonal lines were generated as described in Zhang et al . ( submitted ) . In brief , single cells were sorted into individual wells of a sterile 96-well plate using a DAKO-Cytomation MoFlo High Speed Sorter ( UC Berkeley FACS Facility ) containing 1000 untransfected S2 feeder cells in 200 µL of serum containing medium . After 1 week , the media was replaced with serum medium containing medium supplemented with hygromycin to establish monoclonal lines . Individual clonal lines were checked for expression and one line for each SNAP-tagged centromeric protein was used in our experiments . To ensure that the SNAP-CID protein fusion is functional , RNAi was carried out using the soaking method as previously described [23] , with double-stranded RNA with homology to the 3′UTR of the CID mRNA . These regions were amplified from genomic DNA by PCR . The primers contained the T7 promoter and were as follows: 3′UTR Forward TCCAAAAGAGAAGTTTAGG , Reverse CTCAATGACATGTTATTTATTTG . RNA was synthesized and precipitated using the Ambion kit following manufacturer's instruction , it was then denatured for 30 min at 65°C and re-annealed overnight . Cells were processed for IF with anti-CID ( 1∶1000 ) , anti-tubulin ( Sigma , 1∶500 ) and anti-SNAP ( NEB; 1∶50 ) and imaged as described below . In duplicate experiments , exponentially growing clonal S2 cells stably expressing SNAP-tagged centromeric proteins and GFP-tubulin were incubated in 300 µl of serum containing medium ( SM ) containing 4 µM tetramethylrhodamine ( TMR ) for 15 min . After 3 washes with 5 ml of SM , cells were counted , diluted to 1×106 cells/ml and plated in a 12 well plate . Samples were taken , counted , fixed and mounted right after TMR ( Day 0 ) and after 24 h ( Day 1 ) . Cell counting confirmed that cell numbers doubled during the 24 h period . Cells were imaged immediately after mounting and 10 fields of cells ( 200–300 cells ) were acquired on a PersonalDV microscope ( Applied Precision ) using a 60×/1 . 42 Olympus oil immersion objective . Increments in z were set at 0 . 3 µm , sample thickness was 11 µm , and the bin was set to 2×2 . 100× bin 1×1 images were also acquired to make the figures . All images were scaled in Softworks , maintaining the parameters constant between samples , saved as . psd files and figures were assembled in Adobe Illustrator . Clonal S2 cells stably expressing SNAP-tagged CID , CAL1 and CENP-C and GFP-tubulin or clonal Kc167 cells expressing SNAP-CID were diluted at 1×106 cells/ml in serum medium two days prior to experiments . In addition , extra flasks of cells were prepared at the same concentration two days prior to experiments so that conditioned medium ( CM ) could be harvested . 1 ml of cells were plated in 12-well culture plates and allowed to settle . Medium was removed and replaced with CM containing 12 µM bromothenylpteridine ( BTP; BTP- block ) to quench SNAP-tagged protein , then incubated for 30 min with gentle rocking . Cells were washed four times with 1 ml of serum medium and the last wash was incubated for 30 min . One well of cells was harvested prior to the 30 min wash for the 0 h time-point , to ensure adequate quenching of SNAP-tagged protein . All other samples were harvested at 1 , 2 , 10 , and 24 h following the addition of BTP . Once harvested , cells were pelleted at 600 g for 5 min , and then resuspended in CM containing 4 µM TMR to label the newly-synthesized SNAP-tagged protein . Cells were allowed to incubate for 15 min with gentle rocking . Cells were washed four times with 1 . 5 ml of CM and the last wash was incubated for 30 min . Cells were pelleted , resuspended in 1× PBS , settled on a glass slide , and fixed with 3 . 7% formaldehyde in PBS-T ( PBS with 0 . 1% Triton X-100 ) for 10 min . Slides were washed three times for 5 min in PBS-T , rocking , and then were blocked in 5% milk in PBS-T for 20 min . Slides were incubated with 30 µl of PBS-T 5% milk containing a polyclonal anti-phospho H3 antibody ( Millipore; 1∶1000 dilution ) for 2 hours at room temperature in a humid chamber . Slides were washed three times for 5 min in PBS-T , with gentle rocking , and then were incubated with Alexa 647 anti-rabbit antibody ( Molecular Probes; 1∶500 dilution ) for 45 min at room temperature in a humid chamber . Slides were washed three times for 5 min in PBS-T , with gentle rocking , and were then mounted on coverslips with SlowFade Gold Reagent ( Invitrogen ) containing 2 . 9 µM DAPI . Slides were imaged using a 60×/1 . 42 Olympus oil immersion objective on a PersonalDV microscope ( Applied Precision ) keeping exposure constant between all samples . Cells were manually scored for the cell cycle stage and for the presence or absence of centromeric TMR . Any daughter cell pair connected by a midbody was categorized as cytokinesis . More than 3 independent experiments were carried out , which showed similar results . At least 100 mitotic and 100 interphase cells were scored per experiment . Images were scaled in Softworks , maintaining the scaling constant between samples , saved as . psd files and figures were assembled in Adobe Illustrator . P-values were calculated in InStat ( GraphPad ) . 5×105 cells were centrifuged for 5 min at 600 g at room temperature and resuspended in 150 µl of PBS . 350 µl of ice-cold 100% 200 proof ethanol was added drop wise while vortexing cells gently . Cells were incubated at 4°C for 24 h , washed twice with 1 ml of PBS and then resuspended in 1 ml of PBS-T containing 20 µg/ml Propidium Iodide and 0 . 2 mg/ml RNAse A and incubated at 37°C for 15 min . Samples were analyzed on a Beckman-Coulter EPICS XL flow cytometer and the data was analyzed in FlowJo . Approximate percentage of cell in each cell cycle phase was estimated using the Watson-pragmatic model in Flowjo , eliminating doublets resulting from cells in cytokinesis/G1 . S2 cells stably expressing SNAP-tagged centromeric proteins and GFP-tubulin were diluted at 1×106 cells/ml in serum medium two days prior to experiments and conditioned medium was prepared as above . 1 ml of cells was plated in duplicate wells of 12-well culture plates and allowed to settle . CM was prepared by harvesting medium from the additional flasks , filtering through a 0 . 22 µm filter , and diluting 1∶1 with serum medium . Once settled , cells were incubated for 2 h with either 12 . 5 µM colchicine or 25 µM MG132 in CM , washed three times with 1 ml CM , then incubated for 30 min with CM containing 12 µM BTP block followed by 3 washes in CM , the last wash being incubated for 30 min . One well of cells was harvested prior to the 30 min wash and treated with TMR as above for the 0 h time-point , where ∼50 metaphase cells were observed to ensure complete blocking of SNAP proteins ( 91% of interphase cells were efficiently blocked by treatment with BTP in these experiments ) . After BTP-block , samples were incubated in the presence of 12 . 5 µM colchicine or 25 µM MG132 for additional 4 h to allow synthesis of new SNAP-CID protein . Cells were then TMR labeled , fixed , stained with anti-phospho H3 antibody , mounted and imaged as described above . Presence or absence of TMR labeled centromeres was scored manually in two independent experiments ( N = 50 ) . Presence or absence of TMR-CID was also scored in interphase cells ( phospho-H3 negative; N = 230 ) in the colchicine quench-chase-pulse ( Figure S4 ) . P-values were calculated in InStat ( GraphPad ) . Stable S2 cells expressing SNAP-CID were transfected with FUGENE ( Roche ) following the manufacturer's instructions with 2 µg of the pCopia-GFP-Δ55-CYCA plasmid , which was previously described [23] . 24 h post-transfection , cells were BTP-blocked as above , chased for 4 h and then incubated with TMR as described . Efficiency of the BTP-block was determined in both transfected ( n = 66 ) and mock-transfected ( n = 288 ) cells and was found to be 93% and 98% efficient , respectively . Imaging and manual scoring was carried out as described above . To determine the intensity of TMR-CID , the sum of of pixel intensity in the different z sections was averaged between 3 centromeres in each metaphase cell , the values obtained were subdivided in 5 groups ( n = 35 ND-CYCA transfected metaphases; n = 22 mock-transfected metaphases ) . P-values were calculated in InStat ( GraphPad ) and the graph in Figure 5C was made in Prism ( GraphPad ) . In Softworx Suite , images were deconvolved with the method set to enhanced ratio , the number of cycles set to 5 , and noise filtering set to medium . The images were then quick projected with the method set to max intensity . The images were exported as TIFF files , without scaling to min/max/exp values , with the destination computer set as Windows PC/Linux , and the output size set as 16-bit grey . The TIFF files were analyzed using a MATLAB ( R2007a ) script designed to measure total fluorescence intensity of TMR spots within a cell nucleus , the total area of those TMR spots and the median pixel intensity in a region within the nucleus but excluding the TMR spots . These data were exported as a text file and imported into a Microsoft Excel document . The true TMR intensity per cell was calculated by subtracting from the total TMR intensity the product of the TMR spot area and median pixel intensity of the nuclear region outside the spots . Statistical outliers were removed using the 1 . 5*IQR method . The remaining values were averaged across cells and within each day to make the comparison between days . The values for each experiment were normalized to Day 0 , and then the resulting value was averaged between the two experiments . P-values were calculated using Student's t-test . S2 cells stably expressing SNAP-tagged CID or CAL1 and GFP-tubulin were blocked with BTP as described for the quench-chase-pulse experiments above . Samples were taken at 0 h and 4 h , labeled with TMR , washed and fixed as previously described . Cells were then stained with anti-phospho H3 antibody ( Millipore; 1∶1000 dilution ) for 2 hours at room temperature followed by staining with Alexa 647 anti-rabbit antibody ( Molecular Probes; 1∶500 dilution ) for 45 min at room temperature . Slides were mounted on coverslips using 30 µl of SlowFade Gold Reagent ( Invitrogen ) containing 2 . 9 µM DAPI and imaged on a Deltavision microscope as described above . To compare the TMR intensity between cells in different mitotic stages between 9–13 cells per stage from two independent experiments were analyzed . The images were deconvolved with Softworx ( in the “Ratio” mode , with 5 iterations ) and quick projected . Using the 2D Model function , polygons were generated for individual cells in the DAPI channel to contain the entire DAPI area and the polygons were then propagated through the TMR ( TRITC ) or CID channel . The true TMR intensity ( or CID intensity ) per cell was calculated by subtracting the background for the TMR channel from the total TMR intensity within the DAPI mask . For the cells undergoing cytokinesis , the TMR intensity value for that image is the sum of the TMR intensity values for both daughter cells . Fisher's exact test ( TMR-CID intensity in metaphase versus cytokinesis ) , one-way ANOVA ( TMR-CAL1 in prophase , metaphase and cytokinesis ) , and Mann-Whitney Test ( total CID intensity in metaphase versus interphase cells ) were used to determine the p-values using InStat ( GraphPad ) . 5×107 S2 cells stably expressing FLAG-tagged CID ( where CID is expressed as a N-terminal fusion with FLAG under the pCopia promoter [23] ) were washed in PBS before resuspension in CSK/Triton buffer ( 10 mM PIPES pH6 . 8; 100 mM NaCl; 1 mM EGTA; 300 mM Sucrose; 3 mM MgCl2; 1 mM DTT; 0 . 5% Triton-X100; 1× EDTA-free protease inhibitor ( Roche ) ; 1 mM PMSF ) to extract the cytoplasmic and soluble nuclear fraction ( S1 ) . The remaining pellet was washed in CSK/Triton buffer and then resuspended in CSK buffer ( without Triton-X100 ) with the addition of 25 U of RNase-free DNaseI ( Promega ) before incubation at 37°C for 30 minutes . 4M ( NH4 ) 2SO4 was added to a final concentration of 250 mM to disrupt the nuclear membrane and extract the chromatin-bound fraction ( S2 ) . After centrifugation the pellet was washed in CSK buffer before resuspension in CSK/NaCl buffer ( CSK buffer+2M NaCl ) to extract the histone-containing fraction ( S3 ) . The remaining insoluble fraction containing nuclear-matrix bound material , along with any precipitated proteins , was washed twice in CSK/NaCl buffer and then resuspended in 8M urea ( S4 ) . Total protein concentrations in the four fractions ( S1–4 ) were determined using the 660 nm Protein Assay ( Pierce ) and subsequently 40 µg of total protein were used for analysis by Western blot . Western blot with α-Tubulin antibodies ( 1∶1000 , Sigma ) was used to verify the extraction of soluble proteins ( S1 ) , histone H3 antibodies ( H3K4 dimethylated , 1∶1000 , Abcam ) confirmed the extraction of nuclear soluble proteins ( S2 ) and chromatin-associated ( S3 ) fractions , lamin antibodies ( 1∶1000 , Hybridoma bank ) was used to follow the chromatin-insoluble/matrix-associated fractions S3–S4 . FLAG-CID was detected using anti-FLAG antibodies ( Sigma ) ; CENP-C was detected with using affinity purified guinea-pig antibodies [23]; CAL1 was detected using affinity purified rabbit polyclonal antibodies ( gift of Aaron Straight [23] ) . The cytoplasmic and soluble nuclear S1 fraction was added to 10 µl of anti-FLAG M2 agarose ( Sigma ) and incubated for 2 h , 4°C with rotation . The immunoprecipitated proteins were then washed with 100 volumes of CSK buffer before elution of the bound material by addition of 300 µg/µl 3×FLAG peptide ( Sigma ) in 20 µl of CSK buffer . The input ( S1 ) and eluted fraction were analyzed by Western blotting using antibodies as described above except for CID , which was detected with affinity purified rabbit polyclonal antibodies; 50 µg of total input protein and 25% of immunoprecipitated material were used for Western blot analysis . | The centromere is essential for kinetochore formation , chromosome attachment to spindle microtubules , and equal segregation of the genome to daughter cells . Centromeres are epigenetically inherited through a unique type of chromatin which contains centromere-specific proteins . At each round of DNA replication , centromeric proteins become diluted and must be replenished to ensure faithful maintenance of the centromere locus through cell division . Whether divergent eukaryotes share a common strategy for centromere identity and propagation remains an unanswered question . Here , we examine how Drosophila centromere proteins re-distribute after replication , and we determine the cell-cycle dynamics of their replenishment . We show that three chromatin components required for centromere maintenance display distinct dynamics during the cell cycle; surprisingly , two components are assembled at centromeres during mitosis . These results suggest a new model for regulation of centromere assembly in Drosophila , which emphasizes a key role for the Dipteran-specific protein CAL1 . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology"
] | 2011 | Assembly of Drosophila Centromeric Chromatin Proteins during Mitosis |
In both insects and mammals , spermatids eliminate their bulk cytoplasm as they undergo terminal differentiation . In Drosophila , this process of dramatic cellular remodeling requires apoptotic proteins , including caspases . To gain further insight into the regulation of caspases , we screened a large collection of sterile male flies for mutants that block effector caspase activation at the onset of spermatid individualization . Here , we describe the identification and characterization of a testis-specific , Cullin-3–dependent ubiquitin ligase complex that is required for caspase activation in spermatids . Mutations in either a testis-specific isoform of Cullin-3 ( Cul3Testis ) , the small RING protein Roc1b , or a Drosophila orthologue of the mammalian BTB-Kelch protein Klhl10 all reduce or eliminate effector caspase activation in spermatids . Importantly , all three genes encode proteins that can physically interact to form a ubiquitin ligase complex . Roc1b binds to the catalytic core of Cullin-3 , and Klhl10 binds specifically to a unique testis-specific N-terminal Cullin-3 ( TeNC ) domain of Cul3Testis that is required for activation of effector caspase in spermatids . Finally , the BIR domain region of the giant inhibitor of apoptosis–like protein dBruce is sufficient to bind to Klhl10 , which is consistent with the idea that dBruce is a substrate for the Cullin-3-based E3-ligase complex . These findings reveal a novel role of Cullin-based ubiquitin ligases in caspase regulation .
Caspases are a family of cysteine proteases that have received considerable attention because of their critical roles in inflammation and apoptosis [1–4] . Caspases are expressed as weakly active zymogens in virtually all cells of higher metazoans , and their conversion to the active enzyme is tightly controlled by many different signaling pathways . Historically , most efforts to understand the mechanism of caspase regulation have focused on activator proteins , such as Apaf-1 and FADD , which promote the assembly of active initiator caspase protein complexes [5–9] . Once activated , apoptotic initiator caspases cleave and activate effector caspases , which in turn cleave a variety of important cellular targets , thereby promoting the execution of cell death [10 , 11] . Given the widespread expression of pro-caspases that have the potential to auto-activate in a proteolytic cascade , one might expect that efficient mechanisms exist to prevent inappropriate caspase activation in cells that should live . Furthermore , activation of apoptotic effector caspases does not always result in cell death . For example , apoptotic caspases have been shown to play a critical role for cell differentiation , proliferation , NF-κB signaling , and dendritic pruning [1 , 3 , 12–21] . At this time , the mechanisms that prevent unwanted cell killing by restricting caspase activity are poorly understood , but there are strong reasons to explore the role of inhibitory proteins . One important family of caspase inhibitors are the inhibitor of apoptosis proteins ( IAPs ) , which can bind to and inhibit active caspases in both insects and mammals [22 , 23] . The most compelling evidence for a critical role of IAPs in caspase regulation has come from studies in Drosophila . Drosophila IAP1 ( Diap1 ) encodes an E3 ubiquitin ligase that is strictly required to prevent inappropriate caspase activation and apoptosis [24–27] . In live cells , Diap1 promotes the ubiquitination and degradation of the apoptotic initiator caspase Dronc , and mutations in the RING domain of Diap1 that abrogate E3-ligase activity lead to a dramatic increase of Dronc protein , effector caspase activation , and cell death [28 , 29] . On the other hand , in cells that are destined to undergo apoptosis , Diap1 is inactivated by Reaper-family ( RHG ) proteins [24 , 26 , 27] . Reaper stimulates the self-conjugation and degradation of Diap1 , thereby irreversibly removing this critical caspase inhibitor [30] . Likewise , induction of apoptosis in thymocytes induces the auto-ubiquitination and degradation of mammalian IAPs [31] . These and other observations reveal a critical role of the ubiquitin pathway in the regulation of apoptosis [30 , 32–37] . Ubiquitin-mediated protein degradation is a tightly regulated process , in which proteins are tagged with ubiquitin moieties through a series of enzymatic reactions involving an E1-activating enzyme , E2-conjugating enzyme , and E3 ubiquitin ligase , which determines substrate specificity . Tagged proteins are then degraded by the 26S proteasome [38–40] . However , thus far no other E3 ligases besides IAPs have been implicated in the direct regulation of caspases . Here we provide evidence that a Cullin-3–based multiprotein complex plays a critical role in caspase activation in Drosophila . Cullins are major components of another type of E3 ubiquitin ligase that serve as scaffolds for two functional modules: a catalytic module , composed of a small RING domain protein that recruits the ubiquitin-conjugating enzyme , and a substrate recognition module that binds to the substrate and brings it within proximity to the catalytic module [41 , 42] . The human genome encodes seven different Cullins: Cullin-1 , −2 , −3 , −4A , −4B , −5 , and −7 [41 , 42] . The SCF ( Skp1-Cullin-1-F-box ) complexes are , so far , the best-characterized Cullin-dependent E3 ligases . More recently , the molecular composition and function of the Cullin-3–dependent E3 ligase complex has also been described [43–48] . In this complex , Broad-complex , Tramtrack and Bric-a-Brac ( BTB ) domain-containing proteins mediate binding of the Cullin to the substrate , whereas the Skp1/F-box heterodimer fulfill this function in the SCF complex [41 , 42 , 49] . During the past decade Cullins have been implicated in a variety of cellular activities [41] . However , very little is known about their involvement in the regulation of caspase activation and apoptosis . Here , we describe the identification of cullin-3 mutants from a genetic screen for mutants that abrogate effector caspase activation during terminal differentiation of Drosophila spermatids . In this process , also known as spermatid individualization , spermatids eliminate the majority of their cytoplasm and organelles in an apoptosis-like process that requires canonical cell death proteins , including apoptotic caspases [12 , 50] . Although caspase activation in this system does not lead to death of the entire cell , sperm individualization resembles apoptosis in the sense that many cellular structures are removed into the “waste bag , ” which resembles an apoptotic corpse without the nucleus . Another example where apoptotic proteins are used for cellular remodeling is the caspase-dependent pruning of neurites [14 , 51] . Like in spermatid individualization , the apoptotic machinery is used in a spatially restricted way to destroy only parts of a cell [14 , 51–54] . In our screen , we isolated several cullin-3 alleles with mutations in a testis-specific N-terminal Cullin-3 ( TeNC ) domain . We show that the small RING domain protein , Roc1b , interacts with Cullin-3 in spermatids to promote effector caspase activation . We also identified a BTB-domain protein , Klhl10 , that selectively binds to the testis-specific form of Cullin-3 , but not to somatic Cullin-3 . Mutant alleles of klhl10 were isolated that block effector caspase activation and cause male sterility . Finally , the giant IAP-like protein dBruce binds to Klhl10 in S2 cells , suggesting that dBruce may be a substrate for the Cullin-3–dependent ubiquitin ligase complex . Together , these results define a novel Cullin-3–dependent E3 ubiquitin ligase complex that regulates effector caspase activation in Drosophila spermatids . Given the conserved nature of these proteins , our findings may have important implications for caspase regulation in other systems .
During sperm development in Drosophila , a group of 64 post-meiotic spermatids remain initially interconnected by cytoplasmic bridges that result from incomplete cytokinesis [55] . These spermatids are later separated from each other by the caudal movement of an actin-based individualization complex ( IC ) in a process termed “individualization . ” During spermatid individualization , the majority of the cytoplasm and cellular organelles are removed and get deposited into “waste bags” [55–58] . This process shares several features with apoptosis and requires apoptotic effector caspases [12 , 15 , 50 , 59] . To gain insight into the regulation of caspases in this system , we screened for mutants that lack staining for an antibody detecting processed caspase-3 ( CM1 ) [15 , 50 , 60–62] . We screened a collection of more than 1 , 000 male-sterile mutant lines defective in spermatid individualization that were previously identified among 12 , 326 viable mutants [63 , 64] . Testes from each line were stained with CM1 , and 33 CM1-negative alleles representing 22 different complementation groups were identified . However , the vast majority of male-sterile mutants were CM1-positive , even though many displayed severe defects in spermatid individualization . Therefore , consistent with our earlier observations , caspase activation at the onset of spermatid individualization is independent of many other aspects of sperm differentiation [12 , 50] . To distinguish mutants that specifically affect caspase-3 activation from ones that affect general aspects of spermatid differentiation , we used a monoclonal antibody that detects polyglycylated axonemal tubulin ( AXO 49 ) as an advanced differentiation marker [65 , 66] . In wild-type spermatids , the pattern of AXO 49 staining is identical to that of CM1 ( Figure S1 ) . While mutants in eight of our complementation groups abrogated AXO 49 staining , mutations in the remaining 14 groups retained AXO 49 staining , indicating that these genes act downstream of the general signal ( s ) required for the initiation of spermatid individualization . One of these complementation groups was represented by five different alleles that we termed “medusa” ( mds; in Greek mythology , Medusa represents both life and death ) . mds1 is AXO 49–positive but completely negative for CM1 as a homozygote or in trans to deficiencies that cover the corresponding region ( Figure 1B–1D and Figure S2 ) . The remaining four mds alleles retained various levels of CM1 staining but failed to complement the sterility of mds1 , suggesting that they are hypomorphic alleles . All these mutations were later mapped to the Drosophila cullin-3 gene and were thus designed cul3mds1–5 ( Zuker lines Z2–1089 , Z2–4870 , Z2–4061 , Z2–1270 , and Z2–1062 , respectively; Figure 1E–1G ) . cul3mds2 contains an unrelated lethal mutation in the background and therefore was analyzed in trans to the other alleles or deficiencies in the region . Drosophila apoptotic effector caspases , such as drICE and Dcp-1 , can display DEVD cleaving activity [67–69] . We have previously shown that wild-type adult testes also contain DEVDase activity and that this activity is reduced in cyt-c-d mutant testes [50] . To provide independent evidence for a requirement of Cullin-3 in caspase activation , we measured DEVDase activity in cul3mds1 mutant testes . Whereas lysates of wild-type testes displayed significant levels of DEVDase activity , activity in cul3mds1 mutant testes was reduced to background levels , comparable to the reduction achieved with the potent caspase inhibitor Z-VAD . fmk ( Figure 1I and 1J ) . These results confirm that cullin-3 is required for the activation of effector caspases in spermatids . To map the mds alleles , we first searched for genomic deletions that failed to complement the sterility of the mds males . Utilizing the “deficiency kit” from FlyBase , the male-sterility was mapped to genomic segment 35C1-35D1 on the left arm of the second chromosome ( Figure S2 ) . We then performed similar complementation tests with available mutants in this region and found that lethal cullin-3 mutants [70 , 71] failed to complement the sterility of mds mutant males , suggesting that the mds alleles may represent a unique class of mutations in the cullin-3 gene ( see below , Table S1 ) . Because the Drosophila cullin-3 gene was previously termed gft [71] , we will henceforth refer to the lethal cullin-3 alleles as cul3gft . To determine the molecular nature of the mds mutations , we analyzed first the cullin-3 genomic organization . The cullin-3 gene consists of 14 exons , 11 of which contain coding sequences ( Figure 2A; a partial genomic map was provided in [71] ) . Our genetic and molecular analyses identified a new exon , 1D , and suggested that the cullin-3 gene codes for two major isoforms that are somatic and testis specific ( Figure 2A and 2B ) . A lethal P-element insertion in the 5′ untranslated region ( UTR ) of the cullin-3 gene , cul3gft[06430] , which failed to complement the lethality of other gft alleles , complemented the sterility of the mds alleles , suggesting that these noncoding sequences are only required for the somatic function of cullin-3 ( Figure 2A , Table S1 , and unpublished data ) . Additionally , genomic PCR followed by sequencing analysis revealed that the mds1 mutant contains a deletion in the intron that is flanked by exon 2 and exon 3 , suggesting that this intron contains sequences that are only required for the function of cullin-3 in spermatids ( Figure 2A and 2C ) . Finally , reverse-transcriptase ( RT ) -PCR as well as sequence analyses of several independent clones from adult testis and somatic cDNA libraries confirmed the presence of two major cullin-3 mRNA isoforms , cul3Soma and cul3Testis ( Figure 2A and 2B; see Materials and Methods for details about the cDNA clones ) . While both isoforms share extensive similarity ( exons 3–11 ) , cul3Soma contains a unique , 20-amino-acid-long N-terminal polypeptide ( encoded by exon 2 ) , and cul3Testis contains a unique , 181-amino-acid-long TeNC domain encoded by exon 1D ( Figure 2A and 2B; part of exon 1D is incorrectly annotated in FlyBase as an independent gene , CG31829 ) . We also identified three different mRNA isoforms of cul3Soma , but these only differ in their 5′ UTRs ( encoded by exons 1A , 1B , and 1C , Figure 2A ) . Cullins were previously thought to be universally expressed . Therefore , to the best of our knowledge , cul3Testis represents the first tissue-specific Cullin identified in any organism . To determine the molecular nature of the mds alleles , we sequenced PCR-amplified genomic fragments of the cullin-3 locus from these mutants . As expected from the genetic analysis , all mds alleles contained mutations in or near exon 1D and hence affect only the testis-specific isoform: cul3mds1 has a 181-bp deletion that eliminates part of 5′ UTR of cul3Testis ( orange brackets in Figure 2A and 2C ) . The two hypomorphic alleles , cul3mds3 and cul3mds4 , contain a C-to-T transversion at positions 341 and 347 , which convert glutamine to stop codons at amino acids 8 and 10 , respectively . As a result , translation may initiate downstream of the normal translation initiation site ( orange stars in Figure 2A and Figure S3 ) . cul3mds2 contains a G-to-A transversion of a splice donor site in the intron that is flanked by exons 1D and 3; this presumably abrogates splicing between these exons ( Figure 2A ) . cul3mds5 has a G114-to-A transversion within the 5′ UTR of cul3Testis ( Figure 2A ) . In contrast , three of the lethal gft alleles that failed to complement the sterility of mds−/− males—cul3gft[GR18] , cul3gft4 , and cul3gft2—contain mutations in exons 4 , 10 , and 11 , respectively , that are shared by both isoforms of cullin-3 ( purple stars in Figure 1A; [71] ) . Transheterozygous combinations between cul3mds1 and four strong gft alleles—cul3gft2 , cul3gft[GR18] , cul3gft1 , and cul3gft[d577]—were sterile , and their elongated spermatids were CM1-negative but AXO 49–positive ( Figure 2E , 2G , 2H , Table S1 , and unpublished data ) . Other mds/gft combinations with weaker alleles produced reduced levels of cleaved caspase-3 staining , wild-type levels of axonemal tubulin polyglycylation , and decreased fertility ( Figures 1H , 2I , Table S1 , Figure S4 , and unpublished data ) . Collectively , these results suggest that a testis-specific isoform of cullin-3 is required for effector caspase activation and spermatid individualization . Our genetic analyses suggested the existence of two functionally distinct isoforms of cullin-3 , cul3Testis , and cul3Soma . One possible explanation for this is that the two isoforms are differentially expressed . To test this idea , we examined the distribution of cullin-3 transcripts in the testis and the soma . Comparative RT-PCR experiments were performed using specific primers in the unique 5′ UTRs of cul3Soma and cul3Testis and a reverse primer in their common 3′ UTR ( black arrows in Figure 3A ) . cul3Soma was the only isoform detectable in the soma of adult females ( which lack testes ) , and cul3Testis was the major isoform in testes ( Figure 3B ) . Dissected testes contain both germ cells and somatic cells , such as the testicular wall , muscles cells , and cyst cells . To determine whether cul3Testis is germ-cell specific , we also analyzed RNA from a mutant lacking germ cells . Both the somatic and testis forms of cullin-3 were expressed in wild type , but only cul3Soma was detected in the germ-cell–less reproductive tracts of adult males derived from oskar−/− mothers ( Figure 3C ) . Since cul3Testis is not detectable in adult females , this indicates that cul3Testis expression is restricted to male germ cells , and that cul3Soma expression is mainly , if not exclusively , restricted to somatic cells ( Figure 3B and 3C ) . Finally , consistent with the idea that promoter and 5' UTR sequences of cul3Testis are absent in cul3mds1 mutants ( Figure 2A ) , neither cul3Testis transcripts nor protein were detected in cul3mds1−/− testes . Therefore , cul3mds1 has both the genetic and molecular properties of a cul3Testis null allele ( Figure 3B and 3D ) . These results suggest that differential expression of cul3Testis in the male germline and cul3Soma in somatic tissues accounts for the distinct phenotypes ( male sterility versus lethality ) of the different classes of cullin-3 mutations . The N-terminal region of Cullins is thought to mediate binding to a specific substrate recognition module [41 , 42] ( Figure 9 ) . To test whether the unique TeNC domain is required for the function of Cul3Testis , we tested whether expression of Cul3Soma , which lacks a TeNC domain , was able to functionally substitute for the loss of Cul3Testis in developing spermatids . For this purpose , we generated transgenic flies that express the coding regions of either cul3Testis or cul3Soma under the control of the cul3Testis promoter and 5' and 3' UTRs ( Figure 4A; see also Materials and Methods ) . At least three independent transgenic lines for each of these constructs were crossed to cul3mds1 flies , and proper expression of the transgenes was confirmed by RT-PCR analysis ( Figures 4B and 4C ) . We examined the ability of these transgenes to rescue caspase activation , spermatid individualization and male sterility of cul3mds1 flies . As expected , transgenes with either one or two copies of the cul3Testis open reading frame ( ORF ) fully rescued CM1-staining , spermatid individualization , and male fertility ( Figure 4E and 4F; note the reappearance of cystic bulges and waste bags ) . This proves that both the caspase and sterility phenotypes seen in cul3mds1 mutant flies are due to the loss of cullin-3 function . We next tested the ability of cul3Soma to functionally substitute for the loss of cul3Testis . Neither one nor two copies of cul3Soma rescued spermatid individualization or male fertility , although we observed very low levels of CM1-staining ( Figure 4H and 4I ) . Since the cul3Soma and cul3Testis ORF transgenes were expressed under the same promoter and at comparable levels , we conclude that the TeNC domain is necessary for efficient caspase activation and spermatid individualization . Cullins contain a C-terminal cullin homology domain ( CHD ) that can bind small RING domain proteins , which in turn recruit a ubiquitin-conjugating enzyme ( E2 ) to generate the catalytic module [42 , 49 , 72] . The Drosophila genome contains three small RING domain proteins , Roc1a , Roc1b , and Roc2 , all of which are capable of activating ubiquitin conjugation in vitro [73] . Loss of Roc1a function causes lethality , and targeted disruption of roc1b was previously reported to cause male sterility [73 , 74] . Furthermore , Cullin-3 preferentially co-immunoprecipitates with Roc1b , indicating that both proteins form a complex [74] . We therefore examined whether loss of roc1b function affects caspase activation and individualization of spermatids . We found that roc1bdc3−/− spermatids displayed reduced levels of CM1 staining and failed to individualize ( Figure 5A ) . To test whether roc1b genetically interacts with cul3Testis , we generated double mutants between roc1bdc3 and the hypomorphic cul3mds alleles . Homozygous mutants for either cul3mds or roc1bdc3 showed moderate levels of CM1 staining ( Figure 1E–1G and Figure 5B ) . In contrast , CM1 staining was completely abolished in spermatids of the double mutants , demonstrating that cul3Testis genetically interacts with roc1b to promote caspase activation in spermatids ( Figure 5C ) . These results support the idea that Roc1b is a functionally relevant partner of Cullin-3 in vivo . Cullin-3–dependent E3 ligases use BTB domain containing proteins for substrate recognition [44 , 45 , 49] . The number of genes encoding BTB domain containing proteins is very large , with an estimated 140–250 proteins in Drosophila [42 , 49] . We performed a yeast two hybrid ( Y2H ) screen using the coding region of Cul3Testis as bait to identify potential protein partners for Cul3Testis in a library of adult Drosophila cDNAs ( Figure 6; see also Materials and Methods ) . Several cDNA clones , which encode for Drosophila orthologues of three BTB domain–containing proteins , Spop ( CG9924 ) , Ipp ( CG9426 ) , and Klhl10 ( CG12423 ) , were isolated in this screen ( Figure 6 ) . Notably , both mouse Klhl10 , which shares 46% identity with its Drosophila counterpart and mouse Spop as well as Drosophila Spop were previously shown to interact with Cullin-3 [44 , 75–78] . Given our previous results implicating the TeNC domain in Cul3Testis function , we asked whether any of these proteins bind preferentially to this domain . For this , we examined interactions between these BTB domain–containing proteins and Cul3Testis , Cul3Soma , or the TeNC domain alone in two different yeast strains . Whereas Spop and Ipp interacted with either Cul3Testis or Cul3Soma , Klhl10 interacted with Cul3Testis only ( Figure 6A and 6B ) . However , the TeNC domain alone was not able to bind to any of the BTB domain–containing proteins in this assay . We conclude that the TeNC domain is required but not sufficient for BTB protein binding . These results identify Klhl10 as a potential partner of Cul3Testis in spermatids . If Klhl10 is indeed a physiologically relevant Cullin-3 binding partner in vivo , mutations in klhl10 should affect the function of this E3 complex and thus block caspase activation . To test this hypothesis , we searched for loss-of-function mutations in this gene . Genetic analysis of the klhl10 gene is complicated because of its position within a heterochromatic , cytologically unmapped portion on the 2nd chromosome . However , we were able to identify seven klhl10 alleles ( klhl101–7 ) in our collection of CM1-defective mutants . All alleles were defective in spermatid individualization , were recessive male-sterile , failed to complement each other , lacked CM1 staining but were AXO 49–positive ( Figure 7A–7F and unpublished data ) . This phenotype is virtually identical to the loss of Cul3Testis function . By using RT-PCR and sequence analyses , we identified mutations in six of these klhl10 alleles ( Figure 7I and 7J ) . Five of these alleles , klhl102–6 ( Zuker lines Z2–1331 , Z2–0960 , Z2–2739 , Z2–3284 , and Z2–4385 ) have mutations in highly conserved amino acids of the Kelch repeats , a domain that mediates interaction with the substrate ( colored stars and amino acid residues in Figure 7I and 7J , respectively; more details on the molecular nature of these mutations are in the legends for Figure 7 ) . A sixth mutation , klhl107 ( Z2–3353 ) contains a G508-to-A transversion that converts a highly conserved alanine ( A170 ) to threonine in the BTB domain ( gray star in Figure 7I ) . No mutations were identified in the ORF of klhl101 ( Z2–1827 ) , suggesting that this allele carries a mutation in a regulatory region . To prove that these mutations are indeed responsible for the observed phenotypes , we conducted transgenic rescue experiments . Expression of the klhl10 coding region under the control of the cul3Testis promoter ( together with cul3Testis 5′ and 3′ UTRs , see Materials and Methods ) completely restored CM1 staining and rescued all the sterility phenotypes associated with klhl10 mutant alleles ( Figures 7G and 7H ) . Collectively , these results suggest that Cul3Testis interacts functionally and physically with Roc1b and Klhl10 to promote caspase activation and spermatid individualization in Drosophila . Our results suggest that a Cul3-Roc1b-Klhl10 E3 ubiquitin ligase complex functions at the onset of spermatid individualization . To explore this further , we investigated the level and spatiotemporal distribution of ubiquitinated proteins during spermatid individualization , and the consequences of loss of Cul3Testis and Klhl10 function on this pattern . For this purpose , we stained wild-type testes with the FK2 monoclonal antibody , which specifically detects ubiquitin-conjugated proteins but not free ubiquitin . At the onset of individualization , a steep gradient of ubiquitinated protein expression is detected from the nuclear heads of the spermatids to the tips of their tails ( yellow arrows in Figure 8A ) . During the caudal translocation of the IC , ubiquitinated proteins became completely depleted from the newly individualized portion of the spermatids ( the region that is flanked by a white arrowhead and a white arrow in Figure 8A ) . The staining remained abundant , however , in the pre-individualized portion of the spermatids , with the highest levels seen in the cystic bulge ( CB; Figure 8A–8C ) . At the end of individualization , the newly formed waste bag ( WB ) contained high levels of ubiquitinated proteins ( Figure 8D ) . This spatiotemporal pattern of protein ubiquitination is very similar to the distribution of active effector caspase ( compare Figure 8A–8D to Figure S1B and S1E or to Figure 2 in [12] ) . This striking correlation supports the idea that protein ubiquitination facilitates effector caspase activation in individualizing spermatids . Next , to test whether the observed ubiquitination process depends on an intact Cul3-Klhl10 complex , we stained cul3Testis and klhl10 mutant spermatids with the FK2 antibody . The overall level of protein ubiquitination was dramatically decreased in elongated spermatids from both mutants ( Figure 8E and 8F ) . Furthermore , this reduction was specific to late elongated spermatids , because protein ubiquitination during early stages of spermatid maturation was not significantly affected in cul3Testis−/− and klhl10−/− flies ( Figure 8G–8I ) . These results show that protein ubiquitination during spermatid individualization is largely mediated by the Cul3-Klhl10 complex . Therefore , we conclude that the Cul3-Roc1b-Klhl10 complex is functionally active as an E3 ubiquitin ligase to promote protein ubiquitination during spermatid individualization . Our results suggest a simple working model in which the Cul3-Roc1b-Klhl10 complex promotes caspase activation via ubiquitination and degradation of a caspase inhibitor ( Figure 9 ) . The best-characterized family of endogenous caspase inhibitors is the IAP family [23 , 79] . Diap1 is essential for the survival of most , if not all somatic cells [24 , 27 , 28 , 30 , 32 , 80 , 81] . However , it appears that Diap1 is not the major caspase inhibitor in this context . If Diap1 was a substrate for Cullin-3–mediated protein degradation , we would have expected to see an increase of this protein in cul3 mutants . However , no significant differences in Diap1 protein levels between wild-type and cul3Testis−/− and klhl10−/− mutant testes were detected ( Figure 10A ) . Another candidate is the giant , 4852–amino acid-long , IAP-like protein dBruce . dBruce function is necessary to protect sperm against unwanted caspase activity , because loss of dbruce function causes degeneration of spermatid nuclei and male sterility [12 , 24] . To further investigate possible interactions between dBruce and the Cul3-Roc1b-Klhl10 complex , we tested whether the substrate recruitment protein Klhl10 can bind to dBruce . For this purpose , we expressed tagged versions of Klhl10 and portions of dBruce in S2 cells and performed co-immunoprecipitation ( co-IP ) experiments ( Figure 10B and 10C ) . In this system , Klhl10 efficiently immunoprecipitated both a dBruce “mini gene” ( consisting of the first N-terminal 1 , 622 amino acids , including the BIR domain , and the last C-terminal 446 amino acids that contain the UBC domain; Figure 10B ) . Furthermore , a tagged peptide with the first N-terminal 387 amino acids of dBruce that includes the BIR domain ( amino acids 251–321 ) is sufficient to bind to Klhl10 in this assay ( Figure 10C ) . These data are consistent with the idea that dBruce is a substrate for the Cullin-3-based E3-ligase complex .
We identified four different cullin-3 transcripts , three of which are somatic and share the same coding region , and one male germ-cell–specific transcript . The latter is transcribed from a separate promoter and contains a unique first exon that encodes the N-terminal TeNC domain . This result is somewhat surprising , because there have been no previous reports of tissue-specific expression of different Cullin isoforms with distinct biochemical and functional properties . Our results indicate that the testis-specific TeNC domain plays an important role for binding to the substrate recognition protein , Klhl10 . The substrate specificity of Cullin-based E3 complexes is generally achieved through a variety of substrate recognition adaptors that may be differentially available in different cell types [41 , 49 , 72] . Similar to the SCF and ECS complexes , the BTB-containing substrate recognition proteins bind via their BTB domains to the N-terminal region of Cullin-3 [49] . In the Drosophila Cul3Testis isoform reported here , the TeNC domain is required for strong binding to Klhl10 , caspase activation , spermatid individualization , and fertility . Cul3Soma , which lacks the TeNC domain but is otherwise nearly identical to Cul3Testis , bound much more weakly to Klhl10 and failed to rescue spermatid individualization and fertility of cul3Testis−/− flies . We conclude that the TeNC domain is important for proper binding of Cul3Testis to Klhl10 . However , this domain is not sufficient for binding to Klhl10 , indicating that additional sequences shared between Cul3Testis and Cul3Soma also contribute to this interaction . The TeNC domain is highly conserved among eight different Drosophila species with an evolutionary divergence of up to 40 million years ( Figure S3 ) . Spermatozoa of D . melanogaster are 300 times longer than human spermatozoa , and other Drosophilids can produce even longer sperm , with a length up to 6 cm [86] . It is possible that the TeNC domain of Cul3Testis evolved to facilitate coordinated regulation of caspase activation along the entire length of these giant spermatids . However , there is some indication that Cullin-3 can regulate caspases in tissues that do not contain a TeNC domain ( see below ) . Therefore , it is possible that other factor ( s ) can substitute for the TeNC domain to promote the assembly of a similar Cullin-3–based E3 complex in species or tissues that do not express the TeNC domain . Ubiquitin pathway proteins have well-established roles in the regulation of the cell cycle , DNA damage checkpoint , signal transduction , and in the regulation of apoptosis [37 , 87–91] . Cullin-3–based E3 ubiquitin ligase complexes were previously implicated in various biological processes , including cell-cycle control , Hedgehog signaling , and Wnt signaling [78 , 82–85] . Our current study points to a previously unknown link between the ubiquitin-proteasome protein degradation system and caspase activation during late spermatogenesis . It is unlikely that the Cullin-3–based complex regulates caspases at the mRNA level , because transcripts of effector caspase drice and initiator caspase dronc are present in cul3mds1 mutant testes ( Figure S5 ) . A more likely model is that the Cul3-Roc1b-Klhl10 complex promotes degradation of a caspase inhibitor ( Figure 9 ) . According to this model , the ubiquitination and degradation of this hypothetical caspase inhibitor at the onset of spermatid individualization would de-repress effector caspases and promote sperm differentiation . Whereas Diap1 is an essential caspase inhibitor in most somatic cells in Drosophila , it appears that it is not a major substrate for the Cul3-Roc1b-Klhl10 complex , because no significant differences in Diap1 protein levels between wild-type and cul3Testis−/− and klhl10−/− mutant testes were detected ( Figure 10A ) . On the other hand , the BIR domain region of another IAP-like protein , dBruce , can bind to the substrate recruitment protein Klhl10 in S2 cells ( Figure 10B and 10C ) . These results suggest that dBruce is at least one of the substrates for the Cullin-3–based E3-ligase complex . Importantly , it has been previously shown that loss of dbruce function causes degeneration of spermatid nuclei and male sterility , suggesting that dBruce function in spermatids is tightly controlled to prevent unrestrained caspase activity [12 , 24] . Another interesting question raised by our results is how spermatids can survive high levels of apoptotic effector caspase activity . Since transgenic ectopic expression of the effector caspase drICE leads to spermatid death ( EA , MB , HS , unpublished results ) , we propose that caspase activity in spermatids is restricted to specific subcellular compartments . A related phenomenon has been observed during the caspase-dependent pruning of neurites [14 , 51] . This process is similar to spermatid individualization in that it uses the apoptotic machinery for the destruction of parts of a cell [14 , 51–54] . Interestingly , a requirement for the ubiquitin-proteasome system in the process of axon pruning was also reported [53] . These similarities suggest that the processes of axon pruning and spermatid individualization may use similar mechanisms to restrain the activity of apoptotic proteins for cellular remodeling . In neurons , synaptic activity can lead to local remodeling of synaptic proteins by localized proteasome-mediated degradation [92] . Likewise , it is possible that the proposed caspase inhibitor is only locally degraded , which would allow for localized caspase activity in developing spermatids . There is some evidence that the somatic isoforms of Cullin-3 may also regulate caspase activity in other tissues . Loss of cullin-3 function causes an increase in the number of Drosophila sensory organ precursors and external sensory organs [71] . This phenotype is reminiscent of decreased activity of the apoptotic proteins Ark , Dronc , Dcp-1 , or cytochrome C-d [93–97] . Therefore , Cullin-3 may also play a role to regulate caspase activity in other non-apoptotic processes [13 , 81] . However , based on our results , we would expect that substrate recognition in the soma is mediated by proteins other than Klhl10 . A recent report suggest that mammalian KLHL10 and Cullin-3 can interact in vitro and that Cullin-3 is highly expressed during late murine spermatogenesis [77] . In addition , KLHL10 was shown to be exclusively expressed in the cytoplasm of developmentally advanced murine spermatids , and mice carrying a null klhl10 allele are infertile due to defects during late spermatid maturation [98] . These data suggest that a similar E3 complex may function in late mammalian spermatogenesis and that the defects in klhl10 mutant mice may be due to lack of caspase-3 activity . Despite apparent anatomical differences between insect and mammalian spermiogenesis , there are similarities in the removal of bulk spermatid cytoplasm . Like in insects , intracellular bridges between spermatids and the bulk of the cytoplasm are eliminated during mammalian spermatogenesis . In addition , residual bodies , which contain the extruded cytoplasm of the mammalian spermatids show high levels of active caspase-3 expression and may be homologous to the insect waste bag [99 , 100] . Furthermore , targeted deletion of the mouse Sept4 locus , which encodes the pro-apoptotic protein ARTS , causes defects in the elimination of residual cytoplasm during sperm maturation [99] . Finally , a recent study reported a high frequency of mutations in klhl10 from infertile oligozoospermic men [101] . These intriguing anatomical and molecular similarities between spermatid individualization processes in Drosophila and mammals suggest that further studies on the link between the ubiquitin-proteasome system and apoptotic proteins during sperm differentiation in Drosophila may provide new insights into the etiology of some forms of human infertilities .
yw flies were used as wild-type controls . The Zuker mutants Z2–1089 ( cul3mds1 ) , Z2–4870 ( cul3mds2 ) , Z2–4061 ( cul3mds3 ) , Z2–1270 ( cul3mds4 ) , Z2–1062 ( cul3mds5 ) , Z2–1827 ( klhl101 ) , Z2–1331 ( klhl102 ) , Z2–0960 ( klhl103 ) , Z2–2739 ( klhl104 ) , Z2–3284 ( klhl105 ) , Z2–4385 ( klhl106 ) , and Z2–3353 ( klhl107 ) were obtained from C . S . Zuker ( University of California at San Diego , United States ) ; the gft mutants cul3gft1 , cul3gft2 , cul3gft3 , cul3gft4 , cul3gftGR18 , and cul3gftd577 from M . Ashburner ( University of Cambridge , United Kingdom ) ; the osk[301]/TM3 and osk[CE4]/TM3 lines from R . Lehmann ( Skirball Institute , NYU School of Medicine , New York , United States ) ; roc1bdc3 from R . J . Duronio ( University of North Carolina at Chapel Hill , North Carolina , United States ) ; the deficiency lines DF ( 2L ) ED3 from the Bloomington Stock Center; and the deficiency line DF ( 2L ) Exel8034 from Exelixis . The following BDGP's cul3Testis EST clones: AT08710 , AT10339 , AT08501 , AT07783 , AT21182 , AT19493 , and AT03216 and the cul3Soma EST clones SD20020 and RE58323 were either completely or partially sequenced , and some of them were used as templates in PCR reactions for subcloning . The tr-cul3Testis and tr-cul3Soma rescue constructs were generated as follows: a 979-bp fragment of the presumed promoter region and 5′ UTR and a 345-bp fragment from the 3′ UTR of cul3Testis were PCR amplified from genomic DNA ( forward primer CACATTGGAGCATCGTTAAA and reverse primer GAGATTGCTACGCTGGTCCA with added NsiI and StuI restriction sites , respectively ) and the BDGP's EST clone AT07783 ( forward primer GGCCCACAAAAAGTAGCA and reverse primer AGAGAATATCAAGAAATATATTAGAGGG with added NheI and Acc65I restriction sites , respectively ) , and subcloned in a sequential order into the PstI + StuI and SpeI + Acc65I sites , respectively , of the CaSpeR-4 vector ( from V . Pirrotta ) . Subsequently , the complete coding regions of cul3Testis ( a 2 , 817-bp fragment ) and cul3Soma ( a 2 , 336-bp fragment ) were PCR amplified from the BDGP's EST clones AT07783 ( using the forward primer ATGCAAGGCCGCGATCCCCG and reverse primer TTAGGCCAAGTAGTTGTACA with added XhoI and NheI restriction sites , respectively ) and SD20020 ( using forward primer ATGAATCTGCGGGGAAATCC and reverse primer TTAGGCCAAGTAGTTGTACA with added XhoI and NheI restriction sites , respectively ) , and ligated into the XhoI and XbaI restrictions sites between the cul3Testis 5′ and 3′ UTRs within the CaSpeR-4 vector , to generate tr-cul3Testis and tr-cul3Soma , respectively . To generate the tr-klhl10 rescue construct , the ORF of klhl10 ( a 2 , 320-bp fragment ) was PCR amplified from the BDGP's EST clone AT19737 ( using the forward primer ATGAGTCGTAATCAAAACG and reverse primer CTATGTACGACGACGAATTT with added SalI and XbaI restriction sites , respectively ) , and ligated into the XhoI and XbaI restriction sites between the cul3Testis 5′ and 3′ UTRs within the above vector . Standard Drosophila techniques were used to generate transgenic lines from these constructs . The technical details of the screen were described in the supplementary 4 section in [50] . Cleaved effector caspase antibody staining of young ( 0–2 d old ) adult testes was carried out as described in [12] using a rabbit polyclonal anti-cleaved Caspase-3 ( Asp175 ) antibody ( CM1 , Cell Signaling Technology , Cat . # 9661; http://www . cellsignal . com ) diluted 1:75 . The only changes are that the subsequent TRITC-phalloidin ( Sigma; http://www . sigmaaldrich . com ) incubation for staining of the actin filaments was carried out for 5 min in room temperature , and the slides were subsequently rinsed twice for 10 min in PBS . Axonemal tubulin polyglycylation antibody staining was carried out using the mouse monoclonal antibody AXO 49 ( a kind gift from Marie-Helene Bre , University of Paris-Sud , France ) diluted 1:5 , 000 . The mouse anti-multi ubiquitin monoclonal antibody ( FK2 , Stressgen; http://www . assaydesigns . com ) was used at a dilution of 1:100 . Genomic DNA was isolated from 25–50 adult flies using the High Pure PCR Template Preparation Kit ( Roche; http://www . roche . com ) . Genomic DNA ( 2 μg ) was used to amplify overlapping fragments from the cullin-3 or klhl10 loci in wild-type and homozygote mutant lines . PCR reactions were carried out using DyNAzyme EXT DNA polymerase ( Finnzymes; http://www . finnzymes . fi ) , according to the manufacturer instructions . The products were purified using the High Pure PCR Product Purification Kit ( Roche ) , concentrated by evaporation , and sequenced in a GENEWIZ sequencing facility . 180 testes were dissected from newly eclosed wild-type or cul3mds1 homozygote males , collected into 0 . 5 μl standard skirted tubes ( Fisherbrand #05-669-25; http://www . fischersci . com ) , standing on ice and containing 70 μl of testis buffer ( 10 mM Tris-HCl [pH 6 . 8] , 183 mM KCl , 47 mM NaCl , 1mM EDTA , and 1mM PMSF ) , homogenized using a Pellet Pestle Motor ( Kontes; http://www . kimble-kontes . com/ ) , and subsequently transferred into three new tubes ( 30:30:10 μl ) . The tubes with 10 μl of the testes extracts were used for Western blot analysis to control for the protein amount in the samples by probing with anti-β-tubulin antibody ( E7; 1:1000; Hybridoma Bank; http://dshb . biology . uiowa . edu/ ) . Either Z-VAD ( 20 μM final concentration; Enzyme Systems Products; http://www . mpbio . com/landing . php ) or DMSO was added to each of the 30 μl tubes , and the samples were transferred to a 96-well assay white plate ( Costar #3610 , Corning; http://www . corning . com ) , and allowed to incubate for 10 min at RT . Caspase-Glo 3/7 reagent ( Promega; http://www . promega . com ) was added to a final volume of 200 μl and the signal was detected with a multiwell plate reader ( SPECTRA max M2 , Molecular Devices; http://www . moleculardevices . com ) . Luminescence readings were obtained every 2 min; therefore , each time interval in the figure represents an average of five readings . Three experiments were performed that gave similar results . Total RNA was extracted by using the Micro-to-Midi Total RNA Purification System ( Invitrogen; http://www . invitrogen . com ) according to the manufacturer's recommendations . Ten–twenty young adult testes or male reproductive tracts and ten adult females were used to obtain enough RNA for 5–10 RT-PCR reactions . The samples were collected into 1 . 5-ml Eppendorf tubes that were standing on ice and containing 300 μl of the Invitrogen kit's lysis buffer and 3 μl of 2-mercaptoethanol , homogenized using a Pellet Pestle Motor ( Kontes ) , and subsequently purified using the same kit . In cases when the genomic DNA had to be removed , the 30 μl of the RNA was incubated with 4 μl of RQ1 DNase and 3 . 8 μl of appropriate buffer ( Promega; http://www . promega . com ) for 1 . 5 h at 37 °C , and subsequently purified again with the Invitrogen kit . The RNA was stored in −80 °C or immediately used for RT-PCR reactions using the SuperScriptTM III One-Step RT–PCR System with Platinums Taq DNA polymerase ( Invitrogen ) . The Mastercycler Gradient PCR machine ( Eppendorf; http://www . eppendorf . com ) was programmed as follows: 50 °C for 30 min for the RT step followed by 94 °C for 2 min , and the amplification steps of 94 °C for 30 s , 60 °C for 30 s , and 68 °C for 1 min . A master-mix was prepared and aliquoted to five tubes , each of which was amplified for 17 , 20 , 25 , 30 , or 35 cycles . Absence of genomic DNA in RNA preparations was verified by replacing the RT/Taq mix with only Taq DNA polymerase ( Invitrogen ) . The comparative RT–PCR reactions in Figure 3 were performed using two pairs of primers in a same reaction mix: For cul3Testis the forward primer TCTCATGCAAGGCCGCGATC and the reverse primer CGGGTTATTGGCTGGCGGTC amplified a 2 , 997-bp cDNA fragment ( and a 3 , 715-bp genomic fragment ) , whereas for cul3Soma , the forward primer CATTGATTGCCGCCGAGGAA and the reverse primer CGGGTTATTGGCTGGCGGTC amplified a 2 , 642-bp cDNA fragment ( and a 4 , 911-bp genomic fragment ) . For amplification of the 868-bp fragment of the transgenic tr-cul3Testis ( and the 858-bp endogenous fragment ) in Figure 4B , the forward primer GAGACCCGAATCGCGAGTAG and the reverse primer GCATTCTTTAAGCTGGCCCA were used . For amplification of the 335-bp fragment of the transgenic tr-cul3Soma in Figure 4C , the forward primer GAGACCCGAATCGCGAGTAG ( specific for cul3Testis promoter ) and the reverse primer CATTTTGCCCTCCTTCTTGG ( specific for cul3Soma ) were used . To simultaneously amplify a 496-bp fragment of the endogenous cul3Testis , the reverse primer GGAGGCGTTGGGCACATTGA was also used in the same reaction . For amplification of the 538-bp fragment from drice mRNA in Figure S5A , the forward primer GCCCACCTTGAAGTCTCGCG and the reverse primer CAGGATGTCCAGCCGCTTGC were used . For amplification of the 527-bp fragment from dronc mRNA in Figure S5B , the forward primer CCACCGCCTATAACCTGCTG and the reverse primer CTGCACATACGACGAGGAGG were used . The Cul3Testis and Cul3Soma “bait” constructs were generated as follows: a 2 , 817-bp fragment containing the entire Cul3Testis coding region was PCR amplified from the BDGP's EST clone AT19493 ( forward primer CAAGGCCGCGATCCCCG and reverse primer TTAGGCCAAGTAGTTGTACA with added EcoRI and PstI restriction sites , respectively ) , and a 2 , 336-bp fragment containing the entire Cul3Soma coding region was PCR amplified from the BDGP's EST clone SD20020 ( forward primer AATCTGCGGGGAAATCCTC and reverse primer TTAGGCCAAGTAGTTGTACA with added EcoRI and PstI restriction sites , respectively ) . Both were subcloned in frame to the GAL4 DNA-binding domain using the EcoRI and PstI sites of the pGBKT7 vector ( Matchmaker , Clontech; http://www . clontech . com ) . For the TeNC domain “bait” construct , a 600-bp fragment containing the entire TeNC ORF was PCR amplified from wild-type genomic DNA ( forward primer CAAGGCCGCGATCCCCG and reverse primer GGGATATTAAGACTTTCGCT with added EcoRI and BglII restriction sites , respectively ) and subcloned in frame to the GAL4 DNA binding domain using the EcoRI and BamHI sites of the pGBKT7 vector ( Matchmaker , Clontech ) . Two hybrid screens were performed using Saccharomyces cerevisiae strain AH109 and an adult Drosophila cDNA library ( Matchmaker , Clontech ) . Selection was accomplished on synthetic complete medium lacking tryptophan , leucine , and adenine for 3–7 d at 30 °C . To test for LacZ activity , positive “prey” cDNA clones were isolated and transformed into the Y187 yeast strain , which was pre-transformed with the appropriate “bait” constructs . The UAS-dbruce “mini gene” was generated as follows: a 5 , 022-bp fragment encoding the N-terminal 1 , 622 amino acids of dBruce ( including the BIR domain ) was cleaved by EcoRI and XhoI from the BDGP's EST clone LD31268 and subcloned into the EcoRI and XhoI sites of the pUASt vector to generate the UAS-dbruce-5′ vector . Next , a 1 , 990-bp fragment encoding the C-terminal 446 amino acids of dBruce ( including the UBC domain ) was cleaved with SalI and XbaI from clone T1A-ClaI ( originally identified as clone T1A in the T . Hazelrigg testis cDNA library by J . Agapite and was further cleaved with ClaI to remove the first 357 bp and then self ligated ) , and subcloned in frame into the XhoI and XbaI sites of the UAS-dbruce-5′ vector to generate the UAS-dbruce “mini gene” plasmid . Forty to sixty testes from wild-type and mutant adult males were used to prepare extracts in 30 μl of cell lysis buffer ( 20 mM HEPES–KOH [pH 7 . 6] , 150 mM NaCl , 10% glycerol , 1% Triton X-100 , 2 mM EDTA , 1× protease inhibitor cocktail , and 1 mM DTT ) . Total protein was used for Western blot analysis using either mouse anti-CUL-3 ( 1:1 , 000; BD Transduction Laboratories , cat #611848; http://www . bdbiosciences . com ) [102] or rabbit anti-Diap1 antibody [30] . To generate the anti-dBruce antibody , sequence encoding the C-terminal 446 amino acids of dBruce was cleaved from the T1A-ClaI cDNA clone ( see above ) using SalI and XmaI and then cloned into the XhoI and XmaI sites of a derivative plasmid of the pET14b ( Novagen; http://www . novagen . com ) . The expression plasmid was then transformed into BL21/DE3/pLys , followed by 2-h IPTG induction to express His6-dBruce-C-term . This protein was purified by nickel affinity chromatography and used to raise rat polyclonal antibody ( Covance; http://www . covance . com ) . This antibody recognizes the dBruce “mini gene” band on a Western blot ( 1:1 , 000 ) . For immunoprecipitation reactions , S2 cells were co-transfected with Actin-Gal4 , UAS-PrA-klhl10 , and either UAS-dbruce “mini gene” or UAS-HA- ( dbruce ) BIR plasmids . For negative controls , S2 cells were co-transfected as above but with UAS-PrA-GFP instead of UAS-PrA-klhl10 . Cells were lysed 48 h post-transfection , and the extracts were then incubated with Dynabeads which are conjugated to rabbit IgG ( Dynal , Invitrogen ) at 4 °C for 1 . 5 h . Bound proteins were eluted by boiling in 3 × SDS loading buffer and detected with anti-dBruce antibody ( for the presence of dBruce “mini gene” ) or anti-HA antibody ( for the presence of HA- ( dBruce ) BIR ) . | Caspases are a family of proteases that play important roles in programmed cell death ( apoptosis ) . These enzymes also have nonlethal functions , for example , in inflammation , cell differentiation , and cellular morphogenesis . During maturation , sperm cells eliminate the majority of their cytoplasm and organelles as they are transformed into highly specialized DNA delivery vehicles . Although caspase activation does not kill the entire cell in this case , sperm maturation resembles apoptosis in the sense that many cellular structures are degraded . An important unresolved question is how the lethal activity of apoptotic caspases is regulated to prevent the unwanted death of cells . Here , we show that a Cullin-3–based enzyme complex is required for caspase activation during sperm differentiation in Drosophila . Cullins are known to target cellular proteins for degradation , but their role in caspase regulation was not previously recognized . Our results suggest that a specific Cullin-3 enzyme complex activates caspases by degrading potent caspase inhibitors , thereby providing a model for how apoptotic proteins are regulated during cellular remodeling . Importantly , components of this Cullin-3 enzyme complex are also required for fertility in mice and humans , indicating that this mechanism has been conserved in evolution from fruit flies to humans . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"cell",
"biology",
"drosophila"
] | 2007 | A Ubiquitin Ligase Complex Regulates Caspase Activation During Sperm Differentiation in Drosophila |
The honeybee olfactory system is a well-established model for understanding functional mechanisms of learning and memory . Olfactory stimuli are first processed in the antennal lobe , and then transferred to the mushroom body and lateral horn through dual pathways termed medial and lateral antennal lobe tracts ( m-ALT and l-ALT ) . Recent studies reported that honeybees can perform elemental learning by associating an odour with a reward signal even after lesions in m-ALT or blocking the mushroom bodies . To test the hypothesis that the lateral pathway ( l-ALT ) is sufficient for elemental learning , we modelled local computation within glomeruli in antennal lobes with axons of projection neurons connecting to a decision neuron ( LHN ) in the lateral horn . We show that inhibitory spike-timing dependent plasticity ( modelling non-associative plasticity by exposure to different stimuli ) in the synapses from local neurons to projection neurons decorrelates the projection neurons’ outputs . The strength of the decorrelations is regulated by global inhibitory feedback within antennal lobes to the projection neurons . By additionally modelling octopaminergic modification of synaptic plasticity among local neurons in the antennal lobes and projection neurons to LHN connections , the model can discriminate and generalize olfactory stimuli . Although positive patterning can be accounted for by the l-ALT model , negative patterning requires further processing and mushroom body circuits . Thus , our model explains several–but not all–types of associative olfactory learning and generalization by a few neural layers of odour processing in the l-ALT . As an outcome of the combination between non-associative and associative learning , the modelling approach allows us to link changes in structural organization of honeybees' antennal lobes with their behavioural performances over the course of their life .
Olfactory coding and its modification by learning have been extensively studied in the honeybee , Apis mellifera , both at the behavioural and neural levels [1–5] . Honeybees are able to discriminate between odours , or mixtures of odours , and generalise from a trained odour to perceptually similar odours [6–10] . The protocol typically used to study these capacities is the olfactory conditioning of the proboscis extension reflex ( PER ) [3 , 11 , 12] . The protocol relies on pairing an odorant as a conditioned stimulus with sucrose solution as a reward signal , i . e . as an unconditioned stimulus; in this case , the bee learns to associate the conditioned stimulus with the reward and subsequently responds with proboscis extension to the conditioned stimulus [13 , 14] . In certain forms of non-elemental olfactory learning ( configural learning ) , bees are trained to discriminate single odorants from their mixture; reinforcement assigned to the single odours has a different valence compared to that of the odour mixture , so that ambiguity arises at the level of odour components [6 , 15 , 16] . For instance , in negative patterning discrimination , two individual odours A and B are rewarded while the mixture AB is non-rewarded ( i . e . A+ , B+ vs . AB- ) . During training , each odour component is as often rewarded as non-rewarded so that discrimination requires learning for instance that A alone is different from A in the presence of B . Interestingly , honeybees learn to solve negative patterning discriminations [6] while fruit flies Drosophila melanogaster are unable to learn this task [6 , 17 , 18] . However , the key circuitries underlying this cognitive capacity have only recently started to be elucidated [5 , 19] . A honeybee’s antennae contain ~60 , 000 olfactory receptor neurons that transform chemical features of the environment into spatiotemporal patterns of neural activity ( Fig 1 ) [20] . Axons of different types of olfactory receptor neurons extend to a primary olfactory centre , the antennal lobe that contains 165 spherical structures known as glomeruli ( Fig 1A ) [21 , 22] . Glomeruli are sites of synaptic contacts between afferents of olfactory receptors , inhibitory local neurons ( LNs ) connecting glomeruli , and excitatory projection neurons ( PNs ) ( ~800 ) conveying the processed olfactory message to higher-order centres such as mushroom bodies and the lateral horn . Two types of inhibitory local neurons , heterogeneous and homogeneous , are distinguished in the antennal lobe , depending on their arborisation pattern [21–24] . The glomeruli are laterally interconnected via local neurons or indirectly through projection neurons ( Fig 1A and 1B ) [4 , 22 , 24] . Most projection neurons convey odour information to higher brain regions through a dual pathway [25 , 26] . Projection neurons located in the dorsal region of the antennal lobe form the so-called lateral antennal lobe tract ( l-ALT ) which extends to the lateral horn and then further to the mushroom bodies . Projection neurons in the ventral region of the antennal lobe form the medial antennal lobe tract ( m-ALT ) , which first projects to the mushroom bodies and then to the lateral horn ( for more detail see; [4 , 27–29] ) . Interestingly , l-ALT projection neurons can be found in honeybees and other Hymenoptera but not in Drosophila , which only has m-ALT projection neurons in its olfactory system [28 , 30] . Recent studies suggested that different features of odorants might be processed separately by these two parallel tracts of projection neurons [25 , 26] . Sucrose reward representation is mediated by a giant octopaminergic neuron termed the VUMmx1 neuron , whose activity can substitute for real sucrose in PER olfactory conditioning [31] . Importantly , VUMmx1 contacts the olfactory circuit at three main regions , the antennal lobes , the mushroom bodies , and the lateral horn , thus providing multiple , spatially segregated opportunities for odour-sucrose associations . Although much progress has been made in understanding the physiological properties of projection neurons belonging to these pathways , the roles of these parallel pathways and their contribution to elemental and non-elemental olfactory learning is still unknown . Computational models have been developed to understand functions and mechanisms of these pathways . A recent firing-rate model of the antennal lobe demonstrated that inhibitory local neurons with a global gain neuron could replicate different coding characteristics of the l-ALT and m-ALT pathways [32] . It also has been shown that lateral inhibition provided by local interneurons increases linear separability of odour representations in the antennal lobes , and improves the linear classifier in odour discrimination [33] . The m-ALT pathway that feeds into the mushroom bodies is thought to play a central role in olfactory learning and memory . Heisenberg’s model [34] , which is followed by most computational models of associative learning , describes how odour information is encoded in the Kenyon cells ( the mushroom bodies’ constitutive neurons ) and the connections that the m-ALT projection neurons make with them . Wessnitzer et al . [19] modelled the Drosophila olfactory system from the neural coding stage in the antennal lobe to the mushroom body extrinsic neurons . Their spiking neural network can learn both elemental and non-elemental conditioning tasks , similarly to a recent model of the honeybee mushroom body [5] . Earlier studies produced ambiguous results referring to the question of whether bees can learn elemental associations without the higher-order processing provided by the mushroom bodies [35 , 36] . However , a recent study used selective pharmacological blocking of mushroom bodies and of sub-areas of these structures , and showed that in the absence of functional mushroom bodies , bees fail at learning complex ( configural ) discrimination but can still learn simple olfactory discrimination [15] . Although computational models have focused on the mushroom bodies in analyses of associative olfactory learning , it is usually neglected that both the antennal lobe and the lateral horns possess the basic circuitry to support olfactory learning ( i . e . connectivity between odour and sucrose pathways ) [4 , 29] . Hence , we here explore the potential olfactory learning capacities of the l-ALT , i . e . the circuit from olfactory receptor neurons to the lateral horn via the antennal lobes , using a neural network model . We first focus on olfactory receptor models and modelled their responses to a panel of different odorants . The model reproduces realistic patterns of neural activity at the input level of the antennal lobes . We then implement a non-associative learning rule in the synaptic connections of local neurons to projection neurons of the antennal lobe . We show that exposing the model to different stimuli results in a rearrangement of the initial random inhibitory lateral connections , which then form a local connectivity pattern within the antennal lobes . This promotes separation of odour representations in the antennal lobes . Next , we incorporate VUMmx1 signalling and enrich our model with octopamine-modulated plasticity in the antennal lobes and the lateral horn to model associative olfactory learning ( i . e . learning of odour-sucrose associations ) . We compare the model output with behavioural data from different learning paradigms including elemental learning , configural discriminations , and olfactory generalization . We find that the neural circuit of the l-ALT model accounts for elemental learning and positive patterning discrimination , but not for negative patterning discrimination . In addition , the model can generalize a learned positive patterning discrimination to novel stimuli . The inability of our model to solve the negative patterning confirms the experimental finding that the mushroom bodies are necessary for some forms of configural learning . The model also supports the asymmetric nature of generalization between certain pairs of odorants reported for bees .
Odours are detected by olfactory receptor neurons , which are located within specialized structures called sensilla , distributed on the surface of the antennae . Axons of olfactory receptor neurons constitute the antennal nerve that project to the antennal lobe and provide odour information to this first olfactory processing centre . Since olfactory receptor neurons have selective but also overlapping odour-response-profiles [37] , an odour may activate more than one type of olfactory receptor . The odour-response profiles are modelled and described by using Eq 3 in the Methods Section , which allows generating dose-response curves . Each olfactory receptor neuron exhibits unique response curves , and saturates at a different ligand concentration [38] . These diversities are represented by a matrix of receptor affinity ( See Method and S1 Fig ) that controls the sensitivity of olfactory receptor neurons to different concentration levels ( S2 Fig ) [30 , 39] . We used a fixed affinity matrix throughout this study . We simulated spontaneous and evoked spiking activity of 36 types of olfactory receptor neurons during 1000 ms ( Fig 2 ) . The evoked activity was induced by an odour stimulus presented 250 ms after the onset of this period and which lasted 500 ms ( i . e . until 750 ms ) . Simulation continued during further 250 ms to complete the 1000 ms . The olfactory receptor neurons exhibit high spontaneous activity rate ( Fig 2A ) , which in turn maintains the high activity of projection neurons in the absence of stimuli . This allows a single projection neuron to code different odours at different concentrations by increasing or decreasing its firing rate from the spontaneous rate [38] . Multiple types of olfactory receptor neurons are activated by a single ligand . Fig 2B shows exemplary firing rates of three olfactory receptor neurons that are activated by the same input stimulus . The olfactory receptor neurons quickly respond to the olfactory stimulation and return to baseline activity after removing the stimulus [30] . Increasing odour concentration increases or decreases their responses from baseline level ( Fig 2C ) . Modelling responses of olfactory receptor neurons in this way reproduces the variable selectivity and sensitivity of real olfactory receptor neurons with different tuning responses ( S3 Fig ) [40] . The functional units of the antennal lobes are the glomeruli , where different types of neurons converge and connect to each other . Each glomerulus is made of synaptic contacts between excitatory afferent axons of olfactory receptor neurons , inhibitory local neurons , and excitatory projection neurons conveying the reshaped olfactory message to higher order centres . It has been shown that non-associative learning ( synaptic plasticity in the absence of reward , i . e . upon odour exposure ) changes neural activity in the antennal lobes [41] . Here , we modelled such non-associative learning by a symmetric inhibitory spike timing-dependent plasticity ( iSTDP ) in synaptic connections of local neurons to projection neurons . We then expose the antennal lobe model to a sequence of random odours in the presence of this iSTDP . Fig 3A illustrates weight matrices of the synaptic connectivity from 36 local neurons to 36 projection neurons throughout the simulation . Each matrix column shows the strength of an inhibitory local neuron connection to projection neurons . The initial random matrix is reformed to a structured local connectivity matrix between the glomeruli , which represents local connectivity within the antennal lobe ( see S1 Video ) [24 , 42] . Thus , the inputs to projection neurons are modified according to the state of activity across the antennal lobe . As a result , the correlation of projection neurons’ output approaches an uncorrelated diagonal matrix ( Fig 3B ) . This means that the activity of projection neurons is decorrelated as a result of non-associative learning by exposing the glomeruli to different stimuli [43] . In order to assess the decorrelation process of projection neurons , we quantified correlations by reduction of the entropy of projection neurons’ activity from their independent activity . This Entropy reduction was calculated by ER=12log ( ( 2πe ) 36|Σ| ) −12log ( ( 2πe ) 36 ) , where |∑| is the determinant of covariance matrix , ∑ , obtained from activities of 36 projection neurons ( Fig 3C ) . In addition , we examined the contribution of global inhibitory feedback neurons by changing their synaptic strength . Importantly , we found that the strength of global inhibitory neuron ( homogeneous local neuron ) regulates the redundancy reduction processing in the antennal lobes ( Fig 3C ) [32 , 44] . Thus , the structured lateral inhibition existing in the antennal lobe improves the capacity of linear detectors in the next layer to extract pattern identity [45] . We investigated the separation of odorant representations arising in the antennal lobes as a consequence of non-associative learning . To study the effect of inhibitory neurons within the antennal lobe on the output of projection neurons , an angular distance between two vectors , P1 , P2 that display the population activity of projection neurons for two odours was calculated by d=arccos ( P1 . P2|P1||P2| ) , where ‘ . ’ indicates the inner product between two vectors P1 , P2 , and | x | represents magnitude of the vector x . By measuring the angular distance between the activities of projection neurons in the antennal lobe for odours A and B , we found that the neural representation of the two different odours was more separated by exposure to different odorants ( Fig 4B ) . Here , we tracked an angular distance between the population activities of projection neurons across glomeruli for stimulus A , stimulus B and their mixture AB . ( Fig 4B ) . Odours A and B activate two different but overlapping sets of glomeruli ( neural representations of odours A and B in the antennal lobes ) . Given the proposed connectivity between local and projection neurons , presenting odours A and B together activates some of the projection neurons within glomeruli corresponding to both odours and activates a new set of projection neurons that were silent in presenting odours A or B . The new activated projection neurons constitute the neural response corresponding to the interaction between odours A and B . At strong activity of the global inhibitory neuron , the lateral inhibitory network pushes activity down and enhances the inter-glomerular contrast . This arrangement is compatible with observed data from the honeybee antennal lobes [46 , 47] . Moreover , strong inhibition across glomeruli has been reported for odour mixtures [47 , 48] . To compare the population sparseness within the antennal lobes during the simulation , we used the Treves-Rolls measures [49] of sparseness index , SI= ( ∑j=136rj/36 ) 2/ ( ∑j=136rj2/36 ) , where rj is the firing-rate of the jth projection neuron . We then compared the sparseness index of antennal lobes for an odour mixture AB with those corresponding to stimuli A or B alone . Fig 4C predicts the role of lateral inhibition in sparse coding , as demonstrated in sensory processing of various modalities [50] . This shows that the activity pattern of projection neurons gets sparser ( and the contrast of neural representation is enhanced ) when the global inhibitory neuron becomes stronger . It also indicates that fewer neurons are activated when the olfactory system experiences more odours , which is more energy efficient [51] . This result predicts that the sparse representation of odours in different regions of the antennal lobe might be adjusted by different distributions of inhibitory signals . Taken together , our assumptions on non-associative plasticity and local inhibition in the antennal lobe network can reduce the correlation between the responses of projection neurons and reproduce their tuning responses to different stimuli , both for m-ALT and l-ALT projection neurons . Since we were interested in the contributions of the l-ALT to different forms of olfactory learning and generalization , we defined a connectivity matrix between local interneurons and projection neurons with a strong inhibitory component and studied the capacity of this matrix to account for associative olfactory learning . We tested the performance of our model of the l-ALT pathway in a set of different learning paradigms . To this end , the network model was enriched with octopamine modulation of synaptic plasticity in the antennal lobe and lateral horn , consistent with octopamine-based signalling of sucrose reward in these regions via the VUMmx1 neuron ( Fig 1B , see Introduction ) . We first trained the model using a differential conditioning task , an elemental form of learning in which an odorant A is paired with sucrose reward ( CS+ ) during the stimulus presentation for 500 ms while another odorant B is delivered without reward ( CS- ) . Bees trained in this way easily learn the discrimination and extend the proboscis to A and not to B . Since the lateral horn is thought to be a premotor area [52–54] , the strong response of the lateral horn neuron to odour A would translate into proboscis extension response to this odorant . Hence , we assumed that the lateral horn neuron ( LHN ) acts as a decision neuron , showing a stronger response for CS+ than to CS- . After training , performance of the model is measured by an average firing rate of the lateral horn neuron obtained from presenting odour A and B ( for 3 times randomly ) without reinforcement . Fig 5A provides an example of the differential conditioning task , and shows that the firing rate of the lateral horn neuron increases after presentation of the CS+ and tends to decrease after CS- delivery . This figure implies that the maximum difference between responses of the lateral horn neuron to CS+ and CS- are obtained after only three presentation of CS+ . In order to replicate the learning task for different bees and different odours , we repeated the simulation , using different initial parameters and different odours . The firing rate of lateral horn neuron for the CS+ was significantly higher than that for the CS- ( p-value < 10−6 ) while there was no difference between them before training ( Fig 5B ) . This indicates that the model is able to reproduce the elemental discrimination learning underlying differential olfactory conditioning [3] . Conversely , the model with fixed random connectivity between local neurons and projection neurons within the antennal lobe cannot discriminate between positive and negative conditional stimuli ( p-value = 0 . 29 ) . A comparison between the results of these models reveals how the proper inhibitory connectivity between neurons in the antennal lobe can enhance the learning performance of the model . This emphasizes the importance of the structured connectivity that emerges in the antennal lobe by the non-associative learning in the performance of bees in olfactory learning tasks . To study whether our model can generalize from a learned odour to novel ones depending on odorant similarity , we conditioned the model following an absolute conditioning protocol , in which a single odorant ( A ) is paired with reward . The response of the lateral horn neuron for odour A reached a plateau before testing generalization . The model was then tested with two novel stimuli , one of which ( A’ ) was similar to and the other ( A” ) different from the odour A . The distance of LHN responses in firing rate for the odour A and novel stimuli are assumed to represent the perceived similarity between A and other odours . The model responses in the tests following conditioning ( Fig 6A ) resemble the olfactory generalization performances found in honeybees [6 , 9] , i . e . responses were higher for the odour A and decreased as a function of odour similarity . In order to evaluate the impact of odour similarity , we repeated our simulation but increased the number of test stimuli to six ( the CS+ and 5 novel stimuli ) , with different levels of physical similarity . We defined physical similarity based on the Euclidean distance between vectors of 6 odour stimuli ( see Method section ) . Fig 6B shows the similarity matrix , K , between the six odours . The colour of the element Ki , j denotes the firing rate of the lateral horn neuron for the jth stimulus when the model was trained to the ith stimulus . The performance of the model is consistent with experimental observations that showed asymmetrical generalization in honeybees [8]; for example , generalization from odour 3 to odour 5 is not the same as from 5 to 3 . Because of the modulated plasticity within the antennal lobe during conditioning , the connectivity structure between glomeruli is re-shaped according to the activity of odour 3 while this structure is different if we train the model with another odour ( odour 5 ) ( S4 Fig ) . Hence , the activity within the antennal lobe and LHN for odour 3 when the model was trained with odour 5 is different from the activity of odour 3 when the model was trained by odour 5 . Thus , an asymmetric similarity appears in the generalization matrix . We next focused on the capacity of our network to solve non-elemental learning discriminations . We chose two types of non-elemental learning discrimination , which have been thoroughly investigated in honeybees , the positive and the negative patterning tasks [6 , 15 , 16] . In both tasks , bees have to discriminate a mixture odour AB from its components ( A or B ) . In positive patterning , the odour components are non-rewarded and the compound is rewarded ( A- , B- , AB+ ) ; in negative patterning , the components are rewarded and the mixture is not ( A+ , B+ , AB- ) . We first focused on positive patterning and paired the mixture odour AB with the reinforcement while components A or B were always unrewarded . Although the firing rate of lateral horn neuron for CS-s increased during training , the response of the lateral horn neuron to the CS+ ( the rewarded compound AB ) was significantly higher than that to the CS- ( the components A/B; p-value = 0 . 003 ) ( Fig 7A ) . This differentiation shows , therefore , that our model can achieve a positive patterning discrimination . Focusing on negative patterning yielded , however , a different result . In this case , the model was unable to differentiate between the unrewarded odour mixture AB and its components A and B ( Fig 7B , p-value = 0 . 23 ) . Hence , the model can solve positive but not negative patterning , confirming that both tasks differ in complexity [6] and might thus involve different neural circuits [15 , 35] .
Young honeybees encounter a rich olfactory environment in the hive [55] , which shapes their olfactory system . It has been shown that such passive olfactory exposure increases the volume of the honeybee brain , and also leads to structural modification [56] . Accordingly , odour exposures at early ages , in particular if associated with food reward obtained within the hive , modify sensitivity of the bees , influence performance in behavioural tasks , and make sensory representations in the antennal lobes significantly different from each other [57–59] . Thus , early olfactory experiences are likely to have a strong effect on the bee olfactory circuit in adult life . However , it is unclear which synapses in the antennal lobes are changing , leading to the observed bee behaviour . Galizia et al . [41] suggested that synapses between antennal lobe local neurons and projection neurons change their properties upon odour exposure . To test this hypothesis , we applied the iSTDP learning rule between inhibitory local neurons and excitatory projection neurons . We confirmed that the non-associative plasticity in the antennal lobes can change the random connectivity between glomeruli and create specific connectivity . Interestingly , the connectivity created by exposing the model to different odours increases the separability of odour representations at the antennal lobes output [24 , 32 , 33 , 43] . Moreover , the strength of the global inhibitory feedback neurons can regulate redundancy reduction and connectivity in the antennal lobes . Neural activity patterns at the level of the antennal lobes change shortly after learning , or after a long time after differential conditioning tasks [60–64] . More specifically , Rath et al . recorded calcium signals after differential conditioning and showed that two-odour response patterns in the antennal lobe for CS+ and CS- become more separable after a classical conditioning paradigm [60 , 61 , 63] . Further , the strength of calcium signals of the corresponding glomeruli increased in a conditioning task [61] . In particular , associative learning improved detectability of the corresponding glomeruli to an odour mixture from background activity [65] . However , the mechanisms underlying activity changes in antennal lobe activity after the conditioning task still remain elusive . Antennal lobe activity may be changed by internal sub-circuitry that were re-shaped via local plasticity rules within antennal lobes; alternatively it may be changed by feedback signals from the mushroom bodies [66–68] . Octopamine released from VUM-mx1 in the antennal lobes influences local neuron synapses [31 , 69 , 70] . Since octopamine is the reinforcement signal in the olfactory system of bees , learning-dependent activity in the antennal lobes might be caused by modulated plasticity between antennal lobe local neurons . In this study , we assumed this type of plasticity between local neurons , and showed a modification of neural representation in the antennal lobes as observed in experiments . After conditioning to the CS , a bee is able to respond to a novel stimulus whose perceived similarity is close to the CS ( Fig 7A ) [71 , 72] . It appears that bees generalize odours based on the similarity between carbon chain lengths or whether they belong to the same functional group [8] . Our study showed that generalization is not symmetric for several pairs of odours as asymmetric generalization was found for six odours that were randomly selected from the set of odours ( Fig 6 ) . The possible reason for obtaining such asymmetric structure in our model could be effect of the modulated plasticity between antennal lobe local neurons with the reward . S4 Fig shows how the initially random connectivity matrix between local neurons changes to specific connectivity , depending on the two different conditioned stimuli that activate different projection neurons within glomeruli . This causes different neural representation in the antennal lobes after training , and yields an asymmetric generalization [8] . Elemental and non-elemental learning are intimately related to classical classification problems . Theoretically , differential conditioning , positive , and negative patterning are equivalent to , respectively , OR , AND , and XOR problems in classification theory , with different levels of complexity . For instance , AND and OR problems can be solved by a single layer ‘perceptron’ . It assigns different values to inputs of the network by discovering a linear plane [73] . However , single-layer perceptrons cannot solve the XOR problem because there are no planes that can be drawn across the space of inputs to separate the single components from their mixture . Numerous experimental studies have revealed better discrimination performance for positive patterning than for negative patterning [74–76] . However , a feed-forward network containing hidden units ( multiple layers ) can classify any inputs [77–78] . Thus , the medial pathway containing the Kenyon-cell layer connecting projection neurons and mushroom body extrinsic may allow bees to learn the negative patterning task [5 , 19] . Our model successfully solved the positive patterning task ( A- , B- vs . AB+ ) , but not the negative patterning discrimination ( A+ , B+ vs . AB- ) . This observation is interesting as it predicts that the former task could be solved just based on l-ALT circuitry , i . e . without mushroom body contribution . On the contrary , to solve a negative patterning task , the l-ALT circuitry would be insufficient and the downstream structure of the mushroom bodies would be required . This conclusion , however , contrasts partially with recent findings indicating that mushroom bodies are necessary both for positive and negative patterning discriminations [15] . Yet , these experiments relied on pharmacological blockade of mushroom bodies via procaine ( or PTX in the case of PCT neurons ) , which supports the notion that these structures are necessary , but not sufficient for these forms of non-elemental learning . However , our model predicts that the activities of projection neurons in the case of stimulation with the single odours A , B and the mixture odour AB ( as inputs of the decision neuron ) are more separable when the network has been sufficiently modified by exposure to very different stimuli in the environment ( Fig 4 ) . Consequently , this modified neural network can solve the positive patterning task without participation of mushroom bodies through a linear classifier that also applies to the differential conditioning task . This indicates that adult honeybees , after extensive training , might be able to solve the positive patterning task even after blocking their mushroom bodies . This difference underlines the different associative nature of these two patterning problems: despite their apparent similar complexity ( positive patterning may appear as a mirror-image discrimination with respect to negative patterning and vice versa ) , both tasks differ fundamentally in their difficulty . In fact , positive patterning discrimination could be solved through elemental learning because the associative strength of the non-rewarded components could be sub-threshold for the response but upon compound presentation they might result in a supra-threshold associative strength . Such a linear summation would yield higher associative strength and , therefore , higher responsiveness to the compound . This provides an elemental account of positive patterning , which is not possible in the case of negative patterning . Indeed , the negative patterning discrimination task can only be solved if the animal is able to process the mixture AB in non-linear terms . Otherwise , the sum of the excitatory strengths of the rewarded components upon compound presentation would always be greater than the strength of the single components . This difference may explain why fruit flies are able to master a positive patterning task but not negative patterning discrimination [18] . Fig 4B shows that the angular distance between glomeruli activated by the simultaneous presentation of odours A and B and those activated by the single odours A and B increases during non-associative learning . Hence , the number of antennal lobe neurons that fire for AB is greater than the number of neurons that fire for A and B . Some neurons fire selectively for the mixture AB but not for A or B . This should make synapses between these neurons ( selective for AB ) and the lateral horn neuron reinforced in positive patterning , leading to increased activity of LHN for the mixture AB . This result indicates that the spiking network acts as a linear classifier [15] . On the contrary , a specific neural circuitry has been recently identified as being necessary for negative pattern solving in honeybees [15] . PCT neurons which provide inhibitory GABAergic feedback to mushroom bodies are required for glomeruli in negative patterning [21 , 22] . These PCT neurons may reduce the activity of projection neurons in ventral region of the antennal lobe ( i . e . inputs of m-ALT ) for mixture odour AB . Therefore , the decision neuron in the next layer can discriminate odour components ( A+ , B+ ) from the mixture odour ( AB- ) . In summary , our model predicts that , given appropriate experience of different odors in their early life , bees with lesions in the mushroom bodies may be able to solve the non-elemental positive patterning task but that the circuits outside the mushroom bodies are not sufficient for the negative patterning task . There are substantial differences in the anatomy of the olfactory information processing system among different insect orders . Although the m-ALT is common to insects as diverse such as Orthopterans , Dipterans and Hymenoptera , the l-ALT is unique in the olfactory system of the latter [28] . Hence , it is possible that hymenopterans employ a different strategy for odour coding compared to other insects , thus enabling different olfactory learning abilities . Although it is important to understand the mechanism of the role of l-ALT as well as m-ALT in olfactory learning , our study focused on the l-ALT and its utility for olfactory learning . In modelling the l-ALT , we used only the anatomical and physiological evidence available for honeybees . One exception is the model of odour receptors , for which we used the neural properties of odour receptors of Drosophila , assuming that there are no significant differences between honeybees and fruit flies at the level of the peripheral odour encoding . Future studies must explore how the m-ALT and l-ALT interact during olfactory learning . Moreover , the learning performances of honeybees must be examined upon a specific lesion or blockade of the l-ALT , in particular this interface is performed at different places , i . e . before or after the lateral horn . Many studies suggest that projection neurons might employ temporal coding for odour representation in antennal lobes where the temporal delay between odour onset and spiking activity of projection neurons might complement rate-based coding in the olfactory system [25 , 79–81] . We did not investigate such temporal coding in the proposed network because the available studies reported their results by Ca2+ imaging with low temporal resolution . Moreover , latency coding was observed mostly in the responses of projection neuron belonging to the m-ALT , and reports of latency coding in projection neurons of the l-ALT are rare [81–82] . This might indicate that temporal coding is not as prevalent in the l-ALT as in the m-ALT . However , studying the spatial and temporal coding in the dual olfactory system of honeybees will be an attractive topic for future studies . Note that our result suggests temporal information of spikes may be used by means of iSTDP for better odour separation in antennal lobes ( Fig 4 ) . Our modelling predicts that non-associative learning changes the connectivity in the antennal lobes ( Fig 3 ) . Along with associative learning , our model further predicts that synaptic plasticity between local neurons and projection neurons in the antennal lobe may explain the individual difference in bee’s performance for the olfactory learning during their life ( Fig 5 ) . Behavioural and neurobiological investigations are needed to examine this prediction . We may compare learning performance of two groups of bees , one that explores different odours freely and another whose access to odours is limited in the early stages of their adult lives . We expect to find differences in the bees’ odour learning performance between the two groups in later life , and differently structured connectivity within the antennal lobes . Furthermore , one might discover distinct patterns of synaptic complexes within the antennal lobe for two groups of bees . In a natural environment , bees can detect some odour plumes immediately [83–84] . However , the activity of projections neurons in the ventral regions of the antennal lobe is delayed relative those projection neurons in the l-ALT [25] . Moreover , mushroom body extrinsic neurons encode the value of the stimulus approximately 20 ms after the representation of odours in the lateral horn [81] . This evidence indicates that information transmission in the m-ALT is slower than the processing through the l-ALT . Thus , it could be more efficient for the olfactory system to recognise the identity of the odour stimuli by using rapid processing in the l-ALT . Moreover , concentrations of odour stimuli are evaluated by honeybees [38] although they are less important than identity coding . Hence , it could be proposed that m-ALT has a principal function in encoding odour concentrations . Schmuker et al . [32] suggested that strong lateral inhibition is useful for odour discrimination whereas gain-modulation by means of weak feedback inhibition is suitable for concentration discrimination . Moreover , it has been reported that responses of antennal lobe projection neurons in the l-ALT to weak odour concentrations are stronger than responses of antennal lobe projection neurons in the m-ALT [50] . Hence we expect to find stronger inhibition in the l-ALT as a gain control mechanism . Our results showed that strong inhibition in the dorsal region of the antennal lobe increased the performance of the model in odour discrimination through l-ALT . Further neurobiological studies are needed to investigate the impact of different levels of inhibition in dorsal and ventral regions of the antennal lobe for the coding of odour concentration and identity . Computational models with different levels of complexities are critical to understand the bee olfactory system because a model can integrate biological evidence to link the function of neural networks to behaviour . Over the last decade , several models have been established to describe the characteristics of insect olfactory system from olfactory receptors to the mushroom bodies [5 , 19 , 34 , 44 , 85] . Many of the models focused on the role of antennal lobe networks in separating odour representations [32 , 44] . These studies showed that antennal lobe local neurons and their connectivity with projection neurons can improve performance of the classifiers that receive outputs from the antennal lobes [33] . A recent study also explored how synchrony between projection neurons can represent mixture odours differentially in the antennal lobes . Moreover , different coding strategies in the antennal lobe for odour identity and intensity were explained by the interaction of antennal lobe local neurons with a gain control neuron using a firing rate model [32 , 43] . Here , we reproduced similar results using a more realistic spiking neural network model , and additionally suggested how the specific connectivity between glomeruli emerges based on non-associative learning and different types of local neurons in the antennal lobe . Moreover , the current spiking-network model can be extended to investigate further questions in olfactory learning , for example , the effect of temporal separation between stimulus and reward presentation on learning performance [86–87] . Computational studies on the role of higher brain areas in insect cognition are scarce , and have mostly focused on mushroom bodies . For instance , Wessnitzer et al . developed a spiking model for Drosophila olfactory learning [19] . Their model followed Heisenberg’s approach [34] , which considers the mushroom bodies as a main centre for associative olfactory learning . A recent study by using a binary network of the medial olfactory pathway examined the capacity of the mush room bodies in the different types of learning [5] . This study showed that reward-depending modification of synapses between Kenyon cells and the extrinsic output neurons of the mushroom bodies and the high sparseness of Kenyon cells allow the learning of complex discriminations such as the negative patterning , but the mushroom bodies are not necessary for elemental learning [15 , 35] . Hence , here we provided a minimal spiking neural network model of the l-ALT capable of reproducing some types of the learning without mushroom-boy requirement . A comparison between our model and others models of the medial olfactory pathway [5 , 19] suggests that an additional layer of processing with high sparse response might be essential for solving the negative patterning task .
The model architecture of the honeybee lateral antennal lobe tract is shown in Fig 1B . Olfactory receptor neurons are activated by simulated odorant stimuli ( see Odorant stimulus section ) [30] . The olfactory receptor neurons then project to 36 glomeruli in the dorsal region of the antennal lobe , which is the primary site of olfactory processing in the l-ALT [28 , 35] . In each glomerulus , one projection neuron and one local neuron receive input from a single olfactory receptor neuron [88–89] . The glomeruli are laterally interconnected by the local neurons and projection neurons [90] . A local neuron in a glomerulus inhibits local neurons in the other glomeruli . The projection neuron in each glomerulus sends the excitatory signal to randomly selected local neurons in the other glomeruli . One global inhibitory neuron receives inputs from all projection neurons and sends a feedback signal to them . All projection neurons project to a neuron in the lateral horn called the lateral horn neuron ( LHN ) [29] , which is the output of the present model . Finally , a VUM-mx1 neuron ( shown in Fig 1B in yellow ) makes reward-modulated connections with all the antennal lobe local neurons and the LHN [31] . We describe odorant stimuli and the function of each neuron in detail in the next subsections . Odour molecules activate the initial stage of olfactory processing by producing nerve impulses in the olfactory receptor neurons . Odours contain a complex mixture of chemical compounds ( i . e . , ligands ) ; therefore each odour is specified using a high-dimensional space of ligands [30 , 91] . An odour consists of a few ligands in this space of various concentrations . In this study , we present an odour in a vector of 36 elements L = ( l1 , l2 , … , l36 ) . Each element’s value exhibits the concentration of a particular ligand of the odorant . Because an odour typically contains 2 to 5 ligands [30] , we randomly choose 2 to 5 elements , and assign the concentration values while unselected elements are fixed at zero ( Fig 1C ) . The odour concentration ranges from minimum 10−7 to a maximum of 1 indicating the proportion of dilution . These patterns are used for the inputs to olfactory receptor neurons . This model captures some of the variability of the odour stimuli in the environment . Since the responses of olfactory receptor neurons ( ORNs ) are highly dynamic ( i . e . , their spike rates peak quickly and then relax to a tonic level of activity ) , we simulated responses pattern of ORNs to a large set of odorants by employing the adaptive exponential integrate-and-fire model ( AdEx ) [92] . By combining the AdEx model with the self-organizing model of receptors [90] , we introduce a novel spiking neuron model that can generate the dynamic firing patterns of the olfactory receptor neurons . We constructed 360 olfactory receptor neurons composed of 36 different types ( 10 olfactory receptor neurons for each type ) . In this model , dynamics of sub-threshold membrane potential vi ( t ) of the ith olfactory receptor neuron ( i = 1 , … , 360 ) is described by the following two differential equations: CORNdvi ( t ) dt=−gL ( vi ( t ) −EL ) +gLΔTexp ( vi ( t ) −VTΔT ) −wi ( t ) +Ii+ϵi ( t ) , ( 1 ) τwdwi ( t ) dt=a ( vi ( t ) −EL ) −wi ( t ) , ( 2 ) where wi ( t ) is an adaptation variable , CORN is membrane capacitance and gL and EL are leaked conductance and leak reversal potential , respectively . ΔT ( slope factor ) is a time-scale of the adaptive threshold , and a ( adaptation coupling parameter ) and τw ( adaptation time constant ) are parameters for the adaptive membrane dynamics . The membrane potential vi is reset to v0 if it exceeds the threshold , VT . Moreover , the adaptation variable , wi , is changed by an amount b ( wi → wi + b ) . Here , the input to the model neuron is denoted as Ii , which is computed from odour stimuli by using the self-organizing model of olfactory receptors . The details are described below . Finally , we added a Gaussian noise ϵi ( t ) ∼ N ( 0 , σ ) to add randomness in the spiking activity of the olfactory receptor neurons . We set these parameters so that the model approximates characteristics of the olfactory receptor neurons [30 , 94] ( See S1 Table for parameter values used in the simulation ) . The input to the ith olfactory receptor neurons , Ii , is calculated , following the model proposed in [93] as follows . First , the response Ii , j of the ith olfactory receptor neurons to ligand concentration , lj , of a stimulus L = ( l1 , l2 , … , l36 ) is computed as: Ii , j=11+ ( Kijlj ) −mij , ( 3 ) where Kij is the binding affinity of the ith olfactory receptor neurons to lj . The parameter mij denotes the molecular Hill equivalent , which represents a width of the effective concentration range encoded in the response of each olfactory receptor neuron with respect to the ligand lj . The input to the ith neuron , Ii , is calculated as an average of the responses Ii , j to all ligands ( lj ) within the input stimuli L . The binding characteristic of the ith olfactory receptor neuron is thus specified by its affinity vectors , K= ( Ki1 , Ki2 , … , Ki36 ) and M= ( mi1 , mi2 , … , mi36 ) ( see S1 Fig ) . Here the vectors K and M exhibit the degree of sensitivity and selectivity of receptors to the stimulus . These vectors are generated randomly for 36 different types of olfactory receptor neurons as proposed in [19] . Although olfactory receptor neurons have high affinity for few ligands , most individual olfactory receptor neuron types respond to multiple ligands . The receptor responses saturate if concentrations of the active ligands are significantly high [30 , 95] . To realise the diverse selectivity and sensitivity of real olfactory receptor neurons , we assume that each olfactory receptor neuron exhibits a gradient of affinity to ligands while each olfactory receptor neuron possess a unique preferred ligand defined by the highest affinity . In this study , we construct 36 types of olfactory receptor neurons ( 10 olfactory receptor neurons for each type , 360 in total ) that converge onto 36 glomeruli in the antennal lobes . Further , we use a single projection neuron and antennal lobe local neuron for each glomerulus for simplicity . Since olfactory receptor neurons possessing similar response profiles to a ligand converge onto the same glomeruli , one projection neuron and one antennal lobe local neuron in each glomerulus receive an input from a single type of olfactory receptor neurons . This construction establishes the one-receptor for the one-glomerulus hypothesis for bee antennal lobes [89] . A local neuron in a glomerulus projects inhibitory connections to projection neurons in the other glomeruli . The same local neuron inhibits local neurons in the other glomeruli . We let the projection neuron in each glomerulus send weak excitatory signal to randomly selected local neurons in the other glomeruli ( not shown in Fig 1 ) as is reported in [4] . Finally , the projection neurons in the antennal lobes send an excitatory signal to the higher-order centre of the brain . The activation of projection neurons causes global inhibitory feedback to themselves through a single global inhibitory neuron ( GIN; a homogeneous local neuron ) that receives inputs from all projection neurons . In what follows , we explain in detail the spiking neuron models for these neurons . The subthreshold membrane potential of projection neurons ( uiPN ) and local neurons ( uiLN ) are described by the standard conductance-based leaky integrate-and-fire model . The membrane potential of projection neurons is given by: τmPNduiPN ( t ) dt=−uiPN ( t ) +RPNIiPN ( t ) , ( 4 ) where RPN and τmPN are resistance and membrane time constant of projection neurons respectively ( see S2 Table for parameters ) . The input current IiPN ( t ) represents synaptic inputs from olfactory receptor neurons , antennal lobe local neurons and a global inhibitory neuron as well as external noise . This input is written as IiPN ( t ) =∑j=1N∑fci , jORN→PNgiE ( t−tjORN ) ( VEPN−uiPN ( t ) ) +∑j=1M∑fci , jLN→PNgiI ( t−tjLN ) ( VIPN−uiPN ( t ) ) +∑fci , jgLN→PNgiI ( t−tf ) ( VgIPN−uiPN ( t ) ) +In ( t ) , ( 5 ) where N = 360 and M = 36 are the number olfactory receptor neurons and antennal lobe local neurons respectively . A positive scalar value ci , jORN→PN specifies the strength of a synaptic input from the jth olfactory receptor neurons to the ith PN . Similarly , ci , jLN→PN and ci , jgLN→PN represent a synaptic weight of the jth LN to the ith PN , and a synaptic weight of GIN to PNs . Here we assume that each input spike ( tjf; f: = olfactory receptor neuron , LN or GIN ) cause conductance changes given by giE ( t ) =e− ( t−tjf ) /τE ( t≥tjf ) for olfactory receptor neuron , and giI ( t ) =e− ( t−tjf ) /τI ( t≥tjf ) for local neurons and GIN . We use synaptic time constants τE = 5ms for projection neurons and τI = 10ms for local neurons and τI = 20 ms for the GIN . To implement randomness in the activity in the antennal lobes , we add independent Gaussian noise ϵi ( t ) ∼ N ( 0 , σ ) to the membrane potential of the projection neurons . Similarly , local neurons are modelled by a conductance-based leaky integrate-and-fire model as: τmLNduiLN ( t ) dt=−uiLN ( t ) +RLNIiLN ( t ) , ( 6 ) where IiLN ( t ) =∑j=1N∑fci , jORN→LNgiE ( t−tjORN ) ( VELN−uiLN ( t ) ) +∑j=1M∑fci , jPN→LNgiE ( t−tjPN ) ( VELN−uiLN ( t ) ) +∑j=1M∑fci , jLN→LNgiI ( t−tjLN ) ( VELN−uiLN ( t ) ) +ϵi ( t ) ( 7 ) Here , ci , jORN→LN determines synaptic strength from the jth olfactory receptor neurons to the ith local neuron . We assume random sparse connectivity from projection neurons to local neurons ci , jPN→LN and LNs to LNs ci , jLN→LN ( see S2 Table for parameters ) . The synaptic strengths from olfactory receptor neurons to local and projection neurons were adjusted so that the average activities of downstream local neurons to a stimulus become 40 Hz higher than the spontaneous spike rates [90] . In the subsequent method sections , we enrich the model by introducing non-associative and associative learning in the antennal lobe and lateral horn . There , the synaptic connections from local neurons to projection neurons are modified according to inhibitory spike-timing dependent plasticity ( STDP ) whereas the synaptic strength from LNs to LNs is modulated by octopamine during a learning procedure . We describe the details below . Recent studies revealed that non-associative plasticity modifies neural activities in the antennal lobes: neural representation of mixture odours is changed after bees are preferentially exposed to one component of the mixture without reward [95] . Further , the organization of the antennal lobes and honeybee's behavioural performances in learning tasks changes during the first week of their life apparently due to exposure to new stimuli [41] . The inhibitory synapses in the antennal lobes can be shaped by inhibitory spike-timing dependent plasticity ( iSTDP ) [96] . Here , we model non-associative learning by a symmetric iSTDP between presynaptic antennal lobe local neurons and postsynaptic projection neurons with a decay time constant τiSTDP . In the symmetric iSTDP , both temporal ordering of pre- or postsynaptic spikes potentiate the connectivity , and the synaptic strength of jth inhibitory local neurons onto ith projection neurons ( ci , jLN→PN ) is updated as follows . When we have a presynaptic event at time tjLN of the j local neuron , the synaptic change is given by Δci , jLN→PN=η ( xiPN−α ) , ( 8 ) where xiPN=∑fe− ( ti , fPN−tjLN ) /τiSTDP . Here , ti , fPN exhibits the time of the fth postsynaptic spiking of ith projection neuron that appears before the presynaptic event ( ti , fPN<tjLN ) . η is the learning rate . We added the depression factor α = 2 ρ0 τiSTDP ( ρ0 is a constant ) to control the target rate for the postsynaptic projection neuron [96] . When we have a postsynaptic event at tiPN of the ith projection neuron , the synaptic change is given by Δci , jLN→PN=ηxjLN , ( 9 ) where xjLN=∑fe−tj , fLN−tiPNτiSTDP . Here , tj , fLN exhibits the time of the fth presynaptic spiking of jth local neuron that appears before the postsynaptic event ( tj , fLN<tiPN ) . We assumed a random Gaussian connectivity matrix ( Fig 2B ) from LNs to PNs as an initial connectivity . This connectivity matrix is then modified according to the above procedure ( Eqs 8 and 9 ) until it converges to stable synaptic strengths . A population of projection neurons ( l-PNs ) transfer the olfactory signals to the lateral horn through the l-ATP . Further , the VUM-mx1 neuron releases octopamine in the antennal lobes , lateral horn and mushroom bodies [31] . Octopamine modulates synaptic changing for antennal lobe local neurons and the decision neuron ( LHN ) in the lateral horn [61 , 69] , which is thought to underlie the reinforcement learning during appetitive conditioning . The spiking neuron model of LHN is described by the standard leaky integrate-and-fire model ( see Eq 4 and S2 Table ) . In this study , we assume that all projection neurons convey their information to the single decision neuron ( LHN ) in the lateral horn . Synaptic strengths from the projection neurons to the LHN ( c1 , jPN→LHN ) are modified based on the STDP rule: STDP ( ∆t ) ={A+e−∆tτ+if∆t>0 , A−e∆tτ−if∆t<0 , ( 10 ) where Δt = tpost − tpre indicates the difference of spike times of presynaptic projection neuron ( tpre ) and postsynaptic LHN ( tpost ) . A+ and τ+ is the magnitude and time constant of the STDP function for synaptic potentiation whereas A− and τ− are constants for synaptic depression ( see S3 Table for parameters ) . This STDP learning rule is modulated by octopamine release . The strengths of synapses are limited based on their capacity in changing . Here the effect of the octopamine is modelled as follows: Δc1 , jPN→LHN=fd ( t ) STDP ( Δt ) , ( 11 ) where fd ( t ) =1ift<treward , otherwise1+de ( t−treward ) /τd is the eligibility trace function which modulates the STDP function after the reward signal at time treward . Here , d is the octopamine concentration and τd , which increases or decreases the sensitivity of plasticity to delayed rewards . This equation is a simplified plasticity rule of modulated STDP suggested as a distal reward protocol [97] . | The honeybee olfactory system offers the opportunity to study different levels of learning complexity in a small size network . Odour information is transferred from the antennae to the antennal lobes , and from there to the mushroom bodies and the lateral horn via parallel medial and lateral tracts of projection neurons . Although much progress has been made in understanding olfactory coding in the bee brain , the precise contribution of the lateral pathway in olfactory learning is still unclear . To understand the computational mechanisms underpinning the lateral antennal lobe tract , we modelled local computation and non-associative plasticity within glomeruli in the antennal lobe and the lateral horn where they were additionally modulated by octopamine to achieve associative learning . We establish that the connectivity within the antennal lobe ( that is shaped by non-associative plasticity ) can also be modified by inhibitory feedback neurons . This makes output patterns of antennal lobe more separable and sparser . Our modelling indicates that bees , using the lateral pathway , might learn to solve positive patterning tasks in addition to elemental learning and olfactory generalisation without the contribution of the mushroom bodies . Yet , the model cannot account for negative patterning , thus indicating that the mushroom bodies may be required for this discrimination . Our modelling approach allows us to link changes in structural organization of honeybees' antennal lobes with behavioural performance over the course of their life . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"learning",
"invertebrates",
"honey",
"bees",
"social",
"sciences",
"neuroscience",
"learning",
"and",
"memory",
"animals",
"synaptic",
"plasticity",
"cognitive",
"psychology",
"odorants",
"materials",
"science",
"neuronal",
"plasticity",
"bees",
"developmental",
"neurosc... | 2017 | Olfactory learning without the mushroom bodies: Spiking neural network models of the honeybee lateral antennal lobe tract reveal its capacities in odour memory tasks of varied complexities |
Nontyphoidal Salmonellae ( NTS ) are responsible for a huge burden of bloodstream infection in Sub-Saharan African children . Recent reports of a decline in invasive NTS ( iNTS ) disease from Kenya and The Gambia have emphasised an association with malaria control . Following a similar decline in iNTS disease in Malawi , we have used 9 years of continuous longitudinal data to model the interrelationships between iNTS disease , malaria , HIV and malnutrition . Trends in monthly numbers of childhood iNTS disease presenting at Queen’s Hospital , Blantyre , Malawi from 2002 to 2010 were reviewed in the context of longitudinal monthly data describing malaria slide-positivity among paediatric febrile admissions , paediatric HIV prevalence , nutritional rehabilitation unit admissions and monthly rainfall over the same 9 years , using structural equation models ( SEM ) . Analysis of 3 , 105 iNTS episodes identified from 49 , 093 blood cultures , showed an 11 . 8% annual decline in iNTS ( p < 0 . 001 ) . SEM analysis produced a stable model with good fit , revealing direct and statistically significant seasonal effects of malaria and malnutrition on the prevalence of iNTS disease . When these data were smoothed to eliminate seasonal cyclic changes , these associations remained strong and there were additional significant effects of HIV prevalence . These data suggest that the overall decline in iNTS disease observed in Malawi is attributable to multiple public health interventions leading to reductions in malaria , HIV and acute malnutrition . Understanding the impacts of public health programmes on iNTS disease is essential to plan and evaluate interventions .
Blood stream infection ( BSI ) caused by non-typhoidal Salmonella ( NTS ) is consistently reported as a major cause of morbidity and mortality in children across sub-Saharan Africa ( SSA ) , especially those aged between 6 and 30 months . [1] As pathogens such as Haemophilus influenzae type b ( Hib ) , Neisseria meningitidis serogroup A , and Streptococcus pneumoniae are targeted by highly effective protein-conjugate vaccines , life-threatening disease caused by invasive NTS ( iNTS ) is likely to become relatively more prominent [2] . The strong epidemiological association between malaria and iNTS disease has been documented in several African countries , including Malawi [1 , 3 , 4] , and there is increasing biological evidence that multiple malaria-induced immune defects predispose to iNTS disease , including iron release from haem , impaired neutrophil function and reduction in IL12 production [5–7] . In addition , there are many other factors which may influence host susceptibility to invasive NTS disease among children , including inadequate protective antibody [8 , 9] , malnutrition [10] and impaired cell-mediated immunity caused by HIV infection [11–13] . Recently , a temporal association between a decline in the incidence of malaria and a falling incidence of iNTS disease has been reported from both The Gambia and Kenya [14 , 15] . This has led to the suggestion that effective population-based malaria interventions might result in control of iNTS disease across the continent without the need for specific NTS-targeted measures . In Malawi , we have observed a fall in iNTS disease over 10 years of bacteraemia sentinel surveillance . Here , large-scale implementation of malaria control interventions gained considerable momentum in 2007; however , by 2010 these interventions had yet to impact on the incidence of mild and severe malaria [16] . Alongside this , there has been a highly effective roll-out of antiretroviral therapy ( ART ) since 2004 , with scale up of prevention of mother to child transmission ( PMTCT ) from 2006 [17 , 18] . In addition , a programme to subsidize fertilizer for subsistence farmers , which began in 2005 [19] , has contributed to reductions in all measures of child malnutrition between 2004 and 2010 [20] . Finally , we have previously described a strong seasonal relationship between rainfall and iNTS disease [21] . We hypothesised that the underlying risk factors of rainfall , malnutrition and HIV , in addition to malaria , would be associated , directly and/or indirectly , with the observed changes in iNTS disease incidence . Many of these risk factors are known to interact . To assess the complex interrelationships of factors associated with iNTS disease , we have used structural equation modelling ( SEM ) to analyse longstanding surveillance data collected at the largest government hospital in Malawi from 2001–2010 . The inter-relationships between the monthly numbers of malaria , malnutrition and HIV cases and their association with the corresponding monthly numbers of iNTS cases presenting over the same time period have been modelled in the context of rainfall levels which potentially have both direct environmental effects on NTS transmission through increased surface water and indirect effects on host susceptibility largely through malaria transmission and under-nutrition in the rainy season .
Queen Elizabeth Central Hospital ( QECH ) is a 1250-bed government funded hospital , serving a population of approximately 1 million . Approximately 50 , 000 children/year are assessed at QECH of whom approximately a quarter are admitted . Malaria transmission is endemic , with seasonal peaks during the rainy season [22] . Admissions for severe acute malnutrition ( SAM ) also peak during the rainy season , as increased infection risk coincides with the peak of the ‘hungry season’; when household food supplies are running low whilst the new season’s crops are growing . The prevalence of HIV in pregnant women was approximately 22% within Blantyre in 2001 , declining to 16% by 2010 ( Malawi Ministry of Health Quarterly HIV Programme Reports ) . Vertical transmission of HIV was estimated to occur in 17% of HIV infected pregnancies prior to 2008 and had declined to 13 . 5% by 2010 [23] . In Blantyre , 89% of HIV infected children died by the age of 3 years prior to the roll-out of ART [24] . The primary route of admission for children is the pediatric Accident and Emergency ( A&E ) Unit . All children presenting unwell to hospital with non-surgical illness have blood obtained for a thick blood-film malaria parasite examination , and , if slide-positive were diagnosed with malaria; no account was taken of severity in this study . Blood cultures are obtained from children in whom sepsis is suspected and the criteria for obtaining blood for culture did not change during the study period , nor did the numbers of cultures taken [25] . Automated blood culture was undertaken using a pediatric bottle ( BacT/Alert PF; BioMerieux , UK ) incubated at 37°C in air . Gram-negative isolates were identified using standard techniques; including over-night incubation on blood and MacConkey agar at 37°C in air , and if oxidase negative , identified by API 20E ( BioMerieux , UK ) . Salmonellae were then serotyped as S . Enteritidis , S . Typhimurium , S . Typhi or S . sp . by the following antisera; polyvalent O & H , O4 , O9 , Hd , Hg , Hi , Hm , and Vi ( Prolab Diagnostics , UK ) [26] . Numbers of P . falciparum-positive blood-slides are recorded at the end of each month at the paediatric A&E unit at QECH and an analysis of trends of slide-positivity and of severe disease , from January 2001 to December 2010 , has previously been undertaken [16] . To relate trends of malaria infection to indices of iNTS disease , all blood cultures collected from paediatric admissions during the same period were reviewed . Daily rainfall data ( mm ) were obtained for the Blantyre District from the Department of Climate Change and Meteorological Services , Malawi . As national nutritional data were only available for two time points , monthly admission numbers to the ‘Moyo’ Nutritional Rehabilitation Unit ( NRU ) at QECH were used as a proxy indicator of the incidence of SAM . The admission policy for the NRU did not change throughout the study period , following standard definitions of SAM: weight for height < 70% of the NCHS reference median and/or nutritional oedema , and/or a mid-upper arm circumference ( MUAC ) < 110 mm [27] . Admissions to the NRU almost all come through the paediatric A&E . Only rarely , if children are very sick or their malnutrition is initially missed , are they first admitted to the general wards and transferred later . As data were unavailable for malaria for the fourth quarter of 2004 and NRU admissions data were unavailable for 2005 , these data had to be estimated . As both variables are strongly seasonal , estimates were made for each missing month by calculating the mean of the corresponding month one year before and one year after . As there is evidence that risk of iNTS disease due to HIV-infection declines following effective ART , and that the vast majority of cases of paediatric iNTS disease occur in children under 3 years of age , an estimate was made for the number of children under 3 years in Blantyre with untreated HIV during the study period [28–30] . This estimate was made by taking the estimated number of HIV-infected pregnancies per year in Blantyre during the study period and multiplying this by the estimated incidence of vertical transmission , assuming a 1%/year fall from 18% in 2006 to 14% in 2010 ( Government of Malawi , Ministry of Health: Quarterly HIV Program Reports ( 2005–2014 ) https://www . hiv . health . gov . mw/index . php/our-documents ) [23] . Malawi Ministry of Health ART programme data were used to estimate the number of children in Blantyre on effective ART , based on the programme starting in 2006 and reaching 30% coverage by 2010 , and estimating that ART achieved effective protection against iNTS disease in 70% of recipients . Mortality was estimated at 30%/year in the first three years of life for those not on ART , based on published studies from Malawi ( S1 Table ) [24] . The total numbers of iNTS cases , malaria cases and admissions to the NRU were computed , along with the total rainfall ( in mms ) , for each month of the study period ( Fig 1 ) . As the predictor variables ( rainfall , malnutrition , HIV and malaria prevalence ) were inter-related , a series of structural equation models ( SEM ) were fitted to the data [31] . SEM allows a much more complex set of hypothesised inter-relationships to be explored between variables than is possible with standard multivariable linear regression methods . The latter requires the assumption that a set of predictor variables are independently associated with ( usually a single ) outcome variable . For the specific context of the objectives of this study , SEM methods construct a Bayesian network comprising nodes , representing the variables selected for investigation , linked by arrows indicating probabilistic relationships . A standardised regression coefficient for each line , calculated by the software , indicates the relative contribution of each relationship . This network allows the model to describe additional relationships between the multiple predictor variables . In addition , since the probability and regression coefficient represented by arrows in the model varies according to the direction of the arrow , SEM methods provide stronger evidence for interpreting variable parameters as causal effects , when analysing data from cross-sectional studies , than is the case with multivariable linear regression models . [32] . A simple graphical examination of the incidence estimates for iNTS , malaria , malnutrition and rainfall showed the expected strong monthly cyclical seasonal patterns along with year-on-year reductions , while HIV prevalence did not show a seasonal pattern ( Fig 1 ) . In the first SEM , the monthly numbers of iNTS cases , malaria cases , HIV cases , NRU admissions and rainfall levels were analysed in order to model cyclical monthly variations in these variables . In the second SEM , however , the data were smoothed by taking a 12-month rolling means in order to remove the effect of month-by-month seasonality and hence evaluate the impact of variables acting on iNTS disease over longer periods , in particular HIV-infection . Goodness-of-fit of each SEM was determined using a chi-square test and the root mean square error of approximation ( RMSEA ) statistic [33] . All SEMs were constructed using the IBM SPSS Amos ( release 20 . 0 ) software package . For this exploratory model , statistical significance was set at an alpha of 10% .
We hypothesised that the control of multiple conditions associated with increased risk of iNTS disease , including malaria , HIV and malnutrition , was likely to have led to a decline in iNTS disease . SEM were therefore constructed from the peak in of the epidemic in 2002 to 2010 ( Fig 2 and Fig 3 and S1 Table for complete monthly data ) . The relationships between variables , including seasonal variations , were explored by modelling monthly data describing culture-confirmed iNTS disease , slide-positive malaria cases , NRU admissions ( representing malnutrition ) , rainfall and HIV ( Fig 2 ) . This demonstrated statistically significant and direct contributions to iNTS disease from both malaria and malnutrition . There was also a non-directional correlation between malaria and malnutrition . In this model rainfall had no direct impact upon iNTS disease , but had a strong and significant impact on iNTS disease through its effects upon malaria and also upon malnutrition , lending biological plausibility to the model . This model suggested that whilst HIV had no direct effect on iNTS disease , it indirectly contributed to iNTS disease through its effect upon malnutrition . Time was found to have statistically significant negative relationships with some variables , suggesting that other factors outwith the model were contributing to the marked decline of malaria and HIV disease in Blantyre , as indicated by high standardised regression coefficients . The smaller negative effect of time upon iNTS suggests there were also other factors out-with the model contributing to the fall in iNTS , which we did not capture with our data . This model has good statistical strength ( chi-square ( 3 ) = 4 . 423 , p = 0 . 219; RMSEA = 0 . 067 ( 90% CI: <0 . 001–0 . 188 ) . In order to investigate long-term non-seasonal trends in iNTS disease , a second SEM was constructed , this time smoothing seasonality out of the data ( iNTS disease , malaria and malnutrition ) by taking a 12-month rolling mean ( Fig 3 ) . This enabled the effect of HIV prevalence , which is not seasonal and which changes over longer time periods , to be better evaluated . Rainfall did not contribute any statistically significant relationships once month-to-month seasonality was removed , and was therefore not included in this model . As is the case in the first model , both malaria and malnutrition continued to exhibit major direct relationships with iNTS disease , and the previous non-directional correlation also disappeared . In this model , however , HIV also demonstrated a direct , significant effect upon iNTS disease in addition to an indirect one through its contribution to malnutrition . Similar time effects on iNTS disease , malaria and HIV were seen in this model . In this model , the standardised regression coefficient for the association between time and numbers of iNTS cases presenting was approximately 50% of the corresponding coefficient for the first model , suggesting that there were fewer unexplained factors in this model acting on the observed decline in iNTS cases . Once again , in addition to the statistically significant interactions of the variables within the model , the overall model fit was good ( chi-square ( 2 ) = 1 . 121 , p = 0 . 571; RMSEA < = 0 . 001 ( 90% CI: <0 . 001–0 . 162 ) .
These observational data come from a single centre within Blantyre , and it is likely that some young children die from unidentified and untreated iNTS in the community . Furthermore , we do not have a longitudinal survey reflecting changes in patterns of health service utilisation from this surveillance period . Other factors such as changing availability of medications such as antibiotics or anti-malarials might also have exerted an unseen or proxy effect on our data . Our use of NRU admissions as a proxy indicator for population nutrition status means that nutrition-related observations must be interpreted with care as the NRU only admits cases of SAM . These do not necessarily reflect population distributions ( particularly after mid 2008 when community-based SAM treatment began in Blantyre district ) or the impact of the two other forms of malnutrition , underweight and stunting . Nonetheless , they do reflect the national data relating to all forms of under-nutrition—and as the models presented use actual hospital admission figures , they accurately measure disease burden on health facilities in the study area . Whilst our models have good statistical strength , the changes in iNTS disease that were directly attributed to time indicate that factors outwith the model also affect iNTS disease in Malawi . In particular , it is likely that access to drinkable water and to sanitation and hygiene ( WASH ) facilities affect iNTS disease and we have not been able to include data reflecting changes in provision of WASH facilities in Blantyre over the study period in the model . Our data suggest that the interaction between iNTS disease and malaria , HIV and malnutrition is complex , and that the observed decline in iNTS disease is likely to have been due to multiple public health interventions . We estimate that slightly less than half of this change is explained by a decline in malaria and that a similar proportion was explained by changes in the local epidemiology of HIV , both directly and through its impact on malnutrition . We illustrate the potential of modelling methodology and sentinel surveillance data to inform the use of public health interventions to reduce iNTS disease in SSA . The model indicates some gaps in the data , suggesting that there are other unknown factors that also appear to influence the incidence of iNTS disease . It is therefore likely that direct interventions such as NTS vaccines or improvements in WASH facilities will be required to match recent achievements seen in the control of other severe life-threatening bacterial infections such as Haemophilus influenzae Group b , pneumococcal and meningococcal disease . | Invasive nontyphoidal Salmonella ( iNTS ) disease is estimated to be responsible for 680 , 000 deaths/year and yet this is widely under-recognised by clinicians , epidemiologists and policymakers in Sub-Saharan Africa . Recently there have been reports of a decline in childhood iNTS disease from both Kenya and The Gambia that have been attributed to malaria control , and this has led to the unsubstantiated assumption that NTS disease will disappear . While this may be true in certain settings , numerous studies have also associated HIV and malnutrition with NTS disease . We therefore re-assessed this multifaceted relationship in Malawi , where we have previously reported little change in malaria cases , but where there have been highly successful antiretroviral and malnutrition programmes . Analysis of ~50 , 000 blood cultures and ~240 , 000 malaria slides demonstrates a significant decline in iNTS attributable to malaria , HIV and malnutrition , emphasising the complex inter-relationships between these factors and suggest that malaria interventions alone are unlikely to control iNTS disease . Our findings are highly relevant to the neglected field of bacteraemia in Africa and understanding the direct and indirect impacts of public health programmes on iNTS disease in Africa is essential for policy makers to plan and evaluate interventions to control this condition . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Modelling the Contributions of Malaria, HIV, Malnutrition and Rainfall to the Decline in Paediatric Invasive Non-typhoidal Salmonella Disease in Malawi |
Buruli ulcer ( BU ) is a progressive disease of subcutaneous tissues caused by Mycobacterium ulcerans . The pathology of BU lesions is associated with the local production of a diffusible substance , mycolactone , with cytocidal and immunosuppressive properties . The defective inflammatory responses in BU lesions reflect these biological properties of the toxin . However , whether mycolactone diffuses from infected tissues and suppresses IFN-γ responses in BU patients remains unclear . Here we have investigated the pharmacodistribution of mycolactone following injection in animal models by tracing a radiolabeled form of the toxin , and by directly quantifying mycolactone in lipid extracts from internal organs and cell subpopulations . We show that subcutaneously delivered mycolactone diffused into mouse peripheral blood and accumulated in internal organs with a particular tropism for the spleen . When mice were infected subcutaneously with M . ulcerans , this led to a comparable pattern of distribution of mycolactone . No evidence that mycolactone circulated in blood serum during infection could be demonstrated . However , structurally intact toxin was identified in the mononuclear cells of blood , lymph nodes and spleen several weeks before ulcerative lesions appear . Importantly , diffusion of mycolactone into the blood of M . ulcerans–infected mice coincided with alterations in the functions of circulating lymphocytes . In addition to providing the first evidence that mycolactone diffuses beyond the site of M . ulcerans infection , our results support the hypothesis that the toxin exerts immunosuppressive effects at the systemic level . Furthermore , they suggest that assays based on mycolactone detection in circulating blood cells may be considered for diagnostic tests of early disease .
Buruli ulcer ( BU ) is a cutaneous disease caused by Mycobacterium ulcerans , leading to the formation of progressive ulcers , with extensive skin and soft tissue destruction . The presence of a coagulative necrosis forming a nidus for colonies of bacilli , accompanied by minimal inflammation , are considered the most reliable features for the histopathological diagnosis of BU disease [1] . These hallmarks of BU lesions in fact reflect the dual properties of a macrolide toxin produced by M . ulcerans , mycolactone , which plays a critical role in bacterial virulence [2] , [3] . This unique polyketide has been shown to be a potent cytocidal molecule in vitro and in vivo [4] , [5] , [6] , [7] . In addition , mycolactone displays significant immunosuppressive properties at non-cytotoxic doses towards a wide range of immune cells [5] , [8] , [9] , [10] . Numerous studies have reported defective IFN-γ responses in BU patients , using assays of PBMC restimulation ex-vivo [11] , [12] , [13] , [14] , [15] , [16] . IFN-γ responses to M . ulcerans antigens are reduced in BU patients compared to healthy controls [11] , [12] , [13] , [14] , particularly during the early stage of the disease [15] , [16] . M . ulcerans infection-associated reduction of IFN-γ responses was initially thought to be restricted to mycobacterial antigens . In fact , systemic suppression of IFN-γ responses is not antigen-specific , and resolves after surgical excision of the lesion [17] . Notably , the optimal growth temperature of M . ulcerans is below 35°C [18] , and animal studies suggest that the bacilli remain essentially localized within ulcerative lesions in subcutaneous tissues [5] . The fact that immunosuppression in BU patients resolves after removal of infected tissues therefore strongly suggests that bacterial factors , such as mycolactone , may diffuse from the bacilli colonies and exert immunosuppressive effects at the systemic level [19] . In the present study , we have investigated the pharmacodistribution of mycolactone following injection in animal models by tracing a radiolabeled form of the toxin in vivo , and by directly assessing the integrity and the quantity of mycolactone in lipid extracts from internal organs and cell subpopulations . Our observation that mycolactone diffuses in blood and spleen , and concentrates within distinct immune cellular subsets , supports the notion that mycolactone permits M . ulcerans to establish long-term infections by remotely neutralizing the development of cellular immunity .
Six week old BalB/cByJIco and C57BL/6JIco female mice were purchased from Charles Rivers Laboratories . Mice were housed in a BSL-3 animal facility at the Institut Pasteur , in full compliance with French and European regulation and guidelines on experiments with live animal . Mu 1615 wt ( ATCC 35840 ) was obtained from the Trudeau collection . This strain produces a mixture of mycolactones A/B and C [20] . Bacteria were cultivated in Dubos medium complemented with oleic acid-albumin-dextrose 10% ( OADC , Becton Dickinson ) in spinner flasks at 32°C . To generate 14C-labeled mycolactone , exponentially growing cultures were supplemented weekly with 15 µl [1- 14C] propionic acid ( MP , 40–60 mCi/mmol ) and 15 µl [1 , 2- 14C] acetic acid ( MP , 50–120 mCi/mmol ) . After three weeks , mycolactone was purified from bacterial pellets as previously described [2] . The resulting 14C-radiolabeled mycolactone showed an activity of 300 cpm/µg . Its biological activity , as measured by the assay described below , was equivalent to that of the unlabeled molecule ( data not shown ) . BALB/C mice were injected with 300 µg radiolabeled mycolactone via either the sc , ip or iv routes . Animals were bled at the indicated time points , and serum radioactivity monitored by liquid scintillation counting . 24 h post injection , animals were sacrificed and spleen , kidney , liver , brain , fat tissues and intestine were harvested and ground with a Potter-Elvehjem homogeniser in a minimal volume of H2O . The radioactivity of the resulting suspensions was measured by liquid scintillation counting . Peripheral blood mononuclear cells ( PBMCs ) were isolated from whole blood by density gradient centrifugation using Lympholyte-Mammal ( Cedarlane , Ontario , Canada ) following the instructions of the manufacturer . For PBMC purification from lymph nodes and spleens , organs were first homogenized and passed through nylon filters with 100 µm-diameter pores ( Cell Strainer , BD Falcon™ ) , then submitted to density gradient centrifugation using Lympholyte-M ( Cedarlane , Ontario , Canada ) . Total lipids were extracted from homogenized organs or cell suspensions with chloroform-methanol ( 2∶1 , vol/vol ) for 2 h at room temperature . After separation from the aqueous phase , the organic phase was dried and the resulting material resuspended in ice-cold acetone . This acetone-soluble fraction was resuspended in methanol for analysis by ESI-LC-MS and ESI-LC-MS/MS ( Helium collision gas ) , using a Finnigan LCQ ion trap ( Thermo Finnigan , USA ) coupled with a HP1100 LC fitted with ThermoHypersil BDS C8 column ( 5 µm , 4 . 6×250 mm ) . Mycolactones were eluted with a 40-min gradient from 55 to 95% acetonitrile in water . The presence of m/z 765 . 5 ( mycolactone A/B ) and m/z 749 . 5 ( mycolactone C ) was determined by comparison of the MS/MS spectra with those from pure mycolactone preparations . As elution peak areas were directly proportional to mycolactone concentration , 5 µl of a 1 mg/ml reference solution of mycolactone A/B was analyzed with each set of biological samples and used as a standard . The elution peak areas of mycolactones A/B and C were combined to evaluate the total mycolactone concentration in each sample . The human T cell line Jurkat subclone E6 . 1 was cultured in RPMI with 10% FCS , 2 mM L-glutamine , 100 IU/ml penicillin and 100 µg/ml streptomycin . Cells were incubated in microtiter plates ( 105 cells/well ) with serum aliquots ( 2 , 5 µl ) for 6 h in the presence or absence of 400 ng/ml mycolactone A/B , then activated with 25 ng/ml PMA and 500 ng/ml ionomycin ( both from Calbiochem , La Jolla , CA ) for 16 h . Culture supernatants were assayed for interleukin ( IL ) -2 by ELISA ( R&D , Minneapolis , MN ) . For whole blood stimulation assays , 200 µl of pooled blood samples ( n = 6 ) were incubated with anti-CD3 and -CD28 antibodies ( both at 10 µg/ml ) for 24 h and the production of IL-2 measured by ELISA ( R&D , Minneapolis , MN ) .
To investigate the biodistribution of mycolactone , a radiolabeled molecule was generated by feeding M . ulcerans cultures with [1- 14C] propionic acid and [1 , 2- 14C] acetic acid . Purified 14C-labeled mycolactone was injected into mice using three alternative routes of administration , namely subcutaneous ( sc ) , intraperitoneal ( ip ) , or intravenous ( iv ) , and the radioactivity levels in circulating blood were then monitored during 24 hours , after which animals were euthanised ( Figure 1A ) . We found that the blood concentration of iv-delivered mycolactone declined progressively , likely reflecting a distribution outside the vascular system . In contrast , the blood concentration of mycolactone increased slowly following sc or ip delivery . No circulating mycolactone could be detected after 24 hours , for any of the administration routes . The levels of mycolactone in spleen , kidneys , liver , brain and fat tissues were then investigated 24 hours post-injection , by measuring the radioactivity of homogenized organs ( Figure 1B ) . Mycolactone was found in each of the organs except brain , irrespective of the injection route . The relative distribution of mycolactone in spleen , kidney and liver following iv or sc injection was examined . Interestingly , the levels of radioactivity found in the spleen were significantly higher than those of liver and kidney , suggesting that mycolactone displays a relative tropism for this organ ( Figure 1C ) . To determine whether mycolactone is present in internal organs as an intact molecule , or as degradation products , we repeated the experiment with equivalent doses of unlabeled mycolactone . Here total lipids were extracted from the various organs and their acetone-soluble fractions were analyzed by LC-MS/MS . In each lipid extract , the typical ion trace of intact mycolactone A/B ( m/z 765 ) and C ( m/z 749 ) was observed ( Figure 2 ) . No evidence that mycolactone had been metabolized by ( for example ) hydroxylation , methylation , demethylation , double hydroxylation or loss of the side-chain was observed . Furthermore , a quantitative analysis of mycolactone by LC-MS/MS confirmed the relative tropism of mycolactone for the spleen , suggesting that mycolactone may preferentially target this lymphoid compartment ( data not shown ) . The question of mycolactone's tropism was investigated by incubating mycolactone with either blood cells or splenocytes and subjecting the cell suspension to density gradient fractionation . This method allows the purification of mononuclear cells ( lymphocytes , monocytes , DCs and macrophages ) from extracellular medium , red cells and granulocytes . Strikingly , within 4 hours of incubation with whole heparinized blood , mycolactone ( total A/B and C forms ) distributed primarily in the mononuclear cell compartment , with only a marginal presence in the extracellular medium and in the red cell/granulocyte fraction ( Figure 3 ) . A similar distribution profile was observed when mycolactone was incubated with spleen cell suspensions for 1 hour , strongly suggesting that mycolactone has a particular affinity for mononuclear cell subsets . To evaluate the physiological relevance of this finding , C57BL/6 mice were infected experimentally with M . ulcerans , by injection of live bacilli into the tail via the sc route . We have shown previously that this mode of inoculation results in a progressive infection , with multiplication of the bacilli and development of skin ulcerations within 10 weeks following injection of 104 bacilli [21] . Importantly , although bacilli can disseminate to the draining lymph nodes in this model , they do not reach the spleen [5] . As ulcerative lesions developed , mice were sacrificed and the presence of mycolactone was assessed in spleen , brain , liver , kidney and intestine by LC-MS/MS analysis of their acetone-soluble lipid extracts . In accordance with our previous findings with mycolactone-injected mice , mycolactone was detected in spleen , kidneys , and liver but not in the brain ( Figure 4 ) . Importantly , mycolactone was structurally intact in these internal organs ( data not shown ) . We investigated the presence of mycolactone in the sera of infected mice indirectly , by taking advantage of its immunosuppressive activity on activation-induced IL-2 production by human lymphocytes [22] ( Figure 5A ) . Here C57BL/6 mice were infected by footpad injection of either the wild-type strain ( wtMu ) , or a mycolactone-deficient strain of M . ulcerans ( mup045Mu ) . Sera harvested at different pre-ulcerative stages of the disease were incubated with Jurkat T lymphocytes prior to cell stimulation . Control mouse sera caused a basal inhibition of activation-induced IL-2 production by T cells that was efficiently removed by a heat treatment ( Figure 5B ) . When assessed for their immunosuppressive properties , heat-treated sera from wtMu-infected mice did not differ from those of controls ( Figure 5C ) . Furthermore , no evidence that mycolactone distributes in the sera of wtMu-infected mice could be demonstrated by LC-MS/MS analysis of ethyl acetate extracts , at any stage of the disease ( data not shown ) . Having shown that mycolactone has a particular affinity for mononuclear cells in vitro , we assayed for its presence in this particular cellular subset in vivo . Blood samples , lymph nodes and spleens were collected from mice infected with wtMu , or mup045Mu as controls , and cell suspensions submitted to gradient density fractionation . Total lipids were then extracted from mononuclear cell pellets and their acetone-soluble fractions analyzed by LC-MS/MS . In parallel , the functional properties of peripheral blood lymphocytes ( PBLs ) were evaluated by a whole blood stimulation assay . After 6 weeks of infection with wtMu , that is 4 weeks before ulcerative lesions develop ( Figure 6A ) , mycolactone was detected in PBMCs and in mononuclear cells of the lymph nodes ( both draining and distant ) and the spleen ( Figure 6B ) . Diffusion of mycolactone into the blood of wtMu-infected mice did not alter the number or the viability of CD4+ and CD8+ blood T cells ( data not shown ) . However , it correlated with a decreased capacity of PBLs to produce IL-2 in response to stimulation . In contrast , systemic IL-2 responses of mice infected with mup045 Mu were stable and comparable to those of uninfected controls ( Figure 6C ) .
BU is a tropical disease receiving far less attention than TB and leprosy , although it is more common in some endemic regions of West Africa . In contrast to these other two mycobacterioses , BU is acquired from the environment following inoculation of M . ulcerans in the dermis by a mechanism involving aquatic niches and insect vectors although the exact mode of transmission remains unknown [3] . BU starts as a painless subcutaneous nodule , oedema or plaque , enlarging over time . As lesions progress , ulcers eventually form that are characterized by an extensive necrosis of subcutaneous tissues accompanied by minimal inflammation [1] , [23] . The pathology of the disease is closely associated with the production of a lipophilic toxin , mycolactone . This macrocyclic polyketide is highly cytotoxic to a variety of mammalian cells in vitro , and the injection of mycolactone into the skin of guinea pigs is sufficient to provoke ulcers [2] . Although the presence of mycolactone is reflected locally by its damaging effects on infected tissues any investigation of its diffusion outside the ulcerative lesion has been rendered difficult by the lack of a detection tool for this poorly immunogenic compound . In the present study , we have used a radiolabeled mycolactone to show that the toxin diffuses far beyond the sphere of its cytocidal action at the site of inoculation , or at the point of M . ulcerans infection . Mycolactone is cytotoxic and pro-apoptotic at micromolar concentrations , but in addition non-cytotoxic doses in the nanomolar range can efficiently suppress the functional biology of several types of mononuclear cells . Mycolactone was shown to inhibit the activation-induced production of IL-2 by human lymphocytes , and of TNF by monocytes and macrophages [5] , [9] . Mycolactone also blocked the capacity of dendritic cells ( DCs ) to prime cellular responses and to produce chemotactic signals of inflammation [8] . Lymphocytes , monocytes , DCs and macrophages compose the mononuclear cell fraction of blood and lymphoid organs . Together , these cell populations contribute to the generation of innate and acquired cellular immune responses , which are critical for protective immunity against mycobacterial infections . The fact that mycolactone targets mononuclear cells in mice infected with M . ulcerans strongly suggests that these cell subsets are immunosuppressed by the toxin in vivo . The organ distribution of mycolactone revealed a relative tropism of this molecule for the spleen , which is similar to that of other lipophilic immunosuppressive compounds such as rapamycin or FK506 [24] , [25] . However these two drugs are metabolized in the liver , which produces a large array of metabolites [26] , [27] . Surprisingly , mycolactone was preserved in all the organs ( including the liver ) that we analyzed by LC-MS/MS , and no metabolite could be identified . Further studies , for example the evaluation of toxin levels in bile , faeces and urine , will be required to determine how infected animals eliminate mycolactone . In contrast to FK506 and rapamycin , which concentrate primarily in erythrocytes ( >90% ) and only minimally in lymphocytes ( <1% ) in circulating blood , we found that mycolactone concentrates in mononuclear cells [24] , [28] . This is true both for whole blood and for splenocyte cell suspensions , which are richer in lymphocytes . FK506 and rapamycin are structural analogues binding the same intracellular receptor FKBP12 , although the resulting complex targets a different molecule . Their sequestration by red blood cells is explained by the high immunophilin levels of erythrocytes . For mycolactone the molecular target and pathway of action of the toxin are still unknown . Our findings , combined with the observation that mycolactone is a potent immunosuppressor of monocytes , macrophages , DCs and lymphocytes suggest that the molecular target of mycolactone may be expressed preferentially by mononuclear cells . We detected structurally intact mycolactone in PBMCs 6 weeks post infection with M . ulcerans , that is 4 weeks before ulcerative lesions start to develop in this mouse model . This is the first evidence that mycolactone diffuses outside the lesions of an organism infected with M . ulcerans and circulates via peripheral blood . BU is often diagnosed on the basis of clinical findings , because laboratory diagnoses based on smear examination , M . ulcerans cultures , or PCR detection require significant logistics and equipment . Simple and rapid diagnostic field tests for BU are urgently needed for this disease to be treated locally and inexpensively . Our results obtained in the mouse model suggest that assays based on mycolactone detection in circulating blood cells may be considered for diagnostic tests of early disease . We are currently trying to define ways to detect mycolactone directly in the blood cells of BU patients , which may be applicable to diagnosis in field conditions . | Mycolactone is a lipophilic molecule produced by Mycobacterium ulcerans , the causative agent of the skin disease Buruli ulcer ( BU ) . Mycolactone displays unique cytocidal and immunosuppressive properties that are reflected locally by massive tissue necrosis and minimal inflammation . Here we investigated whether mycolactone diffuses from infected tissues and exerts immunosuppressive properties at the systemic level . We used both a radiolabeled form of the toxin and a direct LC-MS/MS analysis of lipid extracts from internal organs and cell subpopulations to investigate the pharmacodistribution of mycolactone in vivo . Using a mouse model of subcutaneous infection with M . ulcerans , we show that mycolactone distributes far beyond the sphere of its cytocidal action . The toxin diffused from infected tissues into the blood and the spleen , where it concentrated in mononuclear cell subsets . Importantly , mycolactone was detected in circulating blood several weeks before ulcerative lesions develop . The presence of mycolactone in blood cells was associated with a decreased capacity of circulating lymphocytes to produce interleukin-2 upon stimulation . In addition to providing the first evidence that mycolactone targets key immune cell populations in infected hosts , this work suggests that mycolactone detection in peripheral blood cells may form the basis of diagnostic tests of early disease . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"pathology/immunology"
] | 2008 | Mycolactone Diffuses from Mycobacterium ulcerans–Infected Tissues and Targets Mononuclear Cells in Peripheral Blood and Lymphoid Organs |
Recent studies of the HapMap lymphoblastoid cell lines have identified large numbers of quantitative trait loci for gene expression ( eQTLs ) . Reanalyzing these data using a novel Bayesian hierarchical model , we were able to create a surprisingly high-resolution map of the typical locations of sites that affect mRNA levels in cis . Strikingly , we found a strong enrichment of eQTLs in the 250 bp just upstream of the transcription end site ( TES ) , in addition to an enrichment around the transcription start site ( TSS ) . Most eQTLs lie either within genes or close to genes; for example , we estimate that only 5% of eQTLs lie more than 20 kb upstream of the TSS . After controlling for position effects , SNPs in exons are ∼2-fold more likely than SNPs in introns to be eQTLs . Our results suggest an important role for mRNA stability in determining steady-state mRNA levels , and highlight the potential of eQTL mapping as a high-resolution tool for studying the determinants of gene regulation .
Genetic variation that affects gene regulation plays an important role in the genetics of disease and adaptive evolution [1] , [2] , [3] . However , unlike protein-coding sequences , we still know little about how to identify the DNA sequence elements that control gene expression . It is still difficult to predict with any confidence which SNPs are likely to affect gene expression , without performing targeted experimental assays . To address this gap , recent experimental and computational approaches have made progress on identifying elements that may be functional , for example through experimental methods that identify transcription factor binding sites [4] , [5] , by in vivo testing of possible enhancers [6] and by computational analysis of sequence data [7] , [8] , [9] . However , our understanding of the importance of different types of functional elements in gene regulation remains rudimentary . As a complementary approach , genome-wide studies of gene expression are now starting to provide information on genetic variation that impacts gene expression levels [10] . Recent studies in a variety of organisms have shown that levels of gene expression are often highly heritable [11] , [12] , [13] , [14] , and that for many genes it is possible to map cis- and trans-acting factors using linkage [13] , [15] , [16] , [17] , [14] or association mapping [12] , [18] , [19] , [20] , [21] . Recent studies of experimental crosses in yeast and mice have used the locations of SNPs within eQTL genes to provide further information about the identity of functional elements [22] , [23] . In studies of human lymphoblastoid cells , it has been reported that most strong signals of association lie within 100 kb of the transcribed region [12] , and that eQTLs cluster roughly symmetrically around the TSS [20] . In this study , we applied a new Bayesian framework to identify and fine map human lymphoblast eQTLs on a genome-wide scale . In effect , we treat the SNP data as a tool for assaying the functional impact of individual nucleotide changes on gene regulation . Our analysis focuses on the impact of common SNPs on gene expression levels . By using naturally occurring variation , we test the effects of several million variable sites in a single data set . Our results provide a detailed characterization of the types of SNPs that affect gene expression in lymphoblast cell lines .
For each of the 11 , 446 genes , we tested for putative cis-acting eQTLs by regressing measured mRNA levels against SNP genotypes , independently for each SNP in the cis-candidate region , using a standard linear regression model . Consistent with previous reports [20] , we found a substantial number of genes with strong evidence for containing at least one eQTL . A total of 744 genes ( 6 . 5% ) had at least one SNP with a p-value <7×10−6 . If the smallest p-value in each gene is treated as a summary statistic , this threshold yields a gene-level false discovery rate of 5% [28] . We also observed that , in many cases , the SNPs most strongly associated with mRNA levels for a particular gene lie in a restricted region , allowing relatively precise localisation of eQTLs . Figure 1 plots examples of p-values in three genes , illustrating both the strong association signal that is often achieved , and the relatively localised nature of many of the signals ( Figure S5 ) . Encouraged by the potential for these data to localise eQTLs , we next examined the distribution of the physical location of putative eQTLs within the cis-candidate region . For each gene with an eQTL ( defined as having at least one p-value <7×10−6 ) we took the position of the most significant SNP as an estimate of the location of the functional site . In practice , we expect that the most significant SNP will sometimes be the actual functional site , but usually it will not since ( 1 ) HapMap contains only ≈1/3 of common SNPs [25]; ( 2 ) some eQTLs may be due to SNPs in LD with nearby copy number variants , though in practice few of the copy number variants known to be associated with expression are well-tagged by SNPs in these data ( data not shown; [19] ) ; ( 3 ) a non-functional SNP in strong LD with the functional site may have a smaller p-value by chance . Using simulations we estimate that the median distance between the functional SNP and the most significant SNP in our data is 7 . 5 kb ( Figures S6 and S7 ) . As expected , local recombination rates are strongly inversely correlated with the distance between the functional SNP and the most significant SNP ( Figure S8 ) . Figure 2 shows histograms of the locations of the most significant eQTL SNPs , as a function of gene size . ( The plots incorporate a correction factor for the possibility of spurious signals due to undetected SNPs in the expression probes; see Methods . ) Several interesting features emerge . First , the distribution of the most significant eQTL SNPs is roughly centered on the transcribed region . Second , nearly all such eQTL SNPs lie close to genes: we find relatively few that are >50 kb from the corresponding gene . Third , as shown in Figure S9 , there is a significant enrichment of eQTL SNPs in exons compared to introns . We will return to this observation later in the paper . Finally , for all three gene sizes , the highest density of eQTLs is around the TSS and immediately upstream of the TES , as reported previously in yeast [22] . The TSS peak was reported in a previous plot of these data [20] , but in that previous analysis the TES signal peak was concealed due to the variability of gene lengths ( see Figure S10 ) . The TES signal is quite asymmetric: among genes with an eQTL , 10% ( 75 ) have the most significant eQTL in the 4 kb upstream of the TES , compared with just 4% ( 29 ) in the 4 kb immediately downstream . While Figure 2 reveals the broad distribution of eQTLs and makes few modeling assumptions , it does not easily enable formal model testing about which aspects of gene structure ( or other sequence features ) are most important for generating eQTLs . Moreover , since the most significant SNP is not always close to the functional site , this approach can be expected to flatten out the true peaks of eQTLs and to increase the numbers of eQTLs that appear to lie far from the target genes . Consequently , we next developed a Bayesian hierarchical modeling approach that solves many of these problems ( see the Methods for further details ) . We considered a collection of models in which the parameters predict the prior probability that any given SNP in the cis-candidate region will be an “eQTN” ( i . e . , the functional nucleotide that creates an eQTL ) . Each model incorporates information about the physical locations of SNPs and , in some of our models , additional functional annotation of the SNPs . ( Our calculations assume that the actual functional site is included in the HapMap genotype data; see below for further discussion . ) The model parameters are estimated by maximizing the overall likelihood of the expression data , across all genes . To implement our hierarchical approach , we switched to using Bayesian regression to test for association between SNPs and gene expression [29] ( Methods ) . For each SNP in the cis-candidate region around a gene , we computed a Bayes factor that measures the relative support for the alternative hypothesis ( the SNP is an eQTN ) compared against the null ( the SNP is independent of gene expression ) . For these data , the Bayes factors are highly correlated with p-values from standard linear regression . However , a key advantage of Bayes factors is that , combined with the prior probabilities specified by the model , they allow us to compute the posterior probability that each SNP is the actual eQTN . The hierarchical model shares information across all genes about the distribution of signals and this in turn allows better weighting of which SNPs in individual genes are most likely to be eQTNs . For example , consider a hypothetical gene in which two SNPs that are associated with expression are in perfect LD ( r2 = 1 ) . Suppose that one SNP is very close to the TSS , and the other is 30 kb upstream . In the p-value analysis , we would assign each of these SNPs 50% weight . In contrast , the hierarchical model downweights the upstream SNP because it is apparent from the overall data that eQTNs are much more abundant near the TSS , suggesting that the SNP near the TSS is much more likely to be responsible for the signal . Simulations show that the hierarchical model provides a more accurate profile of the distribution of eQTNs ( see Figures S5 and S11 ) . Of course , some degree of complication is added by the fact that current HapMap data do not yet contain all SNPs . Therefore , the sites that we infer to be “eQTNs” in this study surely include many SNPs that are tags of nearby functional SNPs that are not in HapMap . This effect will systematically reduce our estimates of the importance of any particular factor in predicting eQTNs . In the case of factors relating to physical location ( such as distance from the TSS ) simulations show that this has a modest impact on spreading out the signal peaks that we observe , and that the overall distribution of signals is still estimated very well ( see Figure S5 , S11 , and S12 ) . In contrast , in the case of factors relating to functional categories ( e . g . , whether a SNP lies in a conserved element ) we would expect the impact to be much more serious since functional elements are usually small and tag SNPs are unlikely to fall within the same element as a functional site . A second complication is caused by the possibility that undetected SNPs in the expression array probes might create spurious signals [30] . Our hierarchical model includes an explicit correction for this , using the 634 genes with a known SNP in the probe as training data . We first set out to get a more refined view of the distribution of eQTNs across the cis-candidate region . The basic versions of our hierarchical model described the positions of SNPs relative to a single “anchor” point such as the TSS . SNPs were grouped into discrete bins based on their distance upstream of the anchor , or downstream ( treated separately ) . The bins nearest the anchor point were just 1 kb wide , to accommodate rapid changes in the rate of eQTNs , while more distant bins were wider ( this improves the parameter estimates since the distant bins generally contain few eQTNs ) . Each bin was associated with a single parameter that relates to the proportion of SNPs in that bin that are eQTNs . The rationale for this model was that it would provide a good description of the data if , for example , the abundance of regulatory elements could be well predicted by distance from the TSS alone . We also considered models with pairs of anchor points ( e . g . , the TSS and the TES ) . In those models , each SNP belonged to two bins , each corresponding to the distance from one anchor point . This model treats the probability that a SNP is an eQTN as the sum of an effect due to the first anchor plus an effect due to the second anchor . Recall that our gene set includes only genes with a single annotated transcript , so that this analysis does not incorporate alternative transcription start or end sites . Table 1 compares eight different models using either a single anchor point ( e . g . , TSS or TES ) , or pairs of anchors ( TSS and one other anchor ) . We used AIC ( Akaike Information Criterion ) to penalize the two-anchor models for the extra parameters that they use . In summary , the results provide compelling support for a model including both the TSS and TES over all other models ( Table 1 ) . Two other two-anchor models ( namely TSS+probe location , and TSS+coding sequence end ) also performed well , presumably because the Illumina probes and the coding sequence end positions are usually near to the TES . However , given that the TSS+TES model had by far the strongest support , we use this model in our subsequent analyses . We next replotted the locations of eQTNs , using the posterior probabilities estimated by the hierarchical model ( Figure 3 ) . Compared to the p-value-based analysis , the two strong peaks of signal near the TSS and TES are considerably strengthened . Also , in the hierarchical model , the level of background signal upstream and downstream of the gene is greatly reduced , presumably because most of the background signal in the p-value analysis can be explained as resulting from LD with SNPs near the TSS and TES . The hierarchical model estimates that the total number of eQTLs is considerably larger than the number that we detected by linear regression at the rather stringent false discovery rate of 5% ( 1586 vs . 744 ) . This difference is partly because the hierarchical model adds fractional probabilities for eQTLs that have only partial support for being true eQTLs , and partly because the hierarchical model is more sensitive to signals in locations that are likely a priori . Another view of the hierarchical model results is shown in the cumulative plots in Figure 3 , which plot the cumulative distribution of eQTNs across the gene region . Most eQTNs lie close to the gene , with less than 7% of the detected cis-eQTNs located more than 20 kb outside the gene . Overall , there are about 3-fold more eQTNs in the upstream region of the gene ( 5′ of the TSS ) than downstream ( 3′ of the TES ) ( 30% vs . 9% ) . We next investigated the peaks of signal near the TSS and TES in more detail , using a finer bin partition close to the TSS and TES ( see Figure 4A and Methods ) . At this finer scale , the TES signal is extremely sharply peaked over a region of just ∼100 bp immediately upstream of the TES . The data strongly reject a model in which the signal is symmetric around the TES ( p = 3×10−7 ) . In contrast , the TSS signal is more spread out , and spans both sides of the TSS . There is no evidence of asymmetry in the TSS signal ( p = 0 . 34 ) . We also observed that the TSS and TES peaks both correspond with two parts of the typical gene structure that , averaging across all 11 , 446 genes , tend to be highly conserved across the mammalian phylogeny ( Figure 4B ) . The correspondence of the two eQTN peaks with the peaks of conservation suggests that there may be a causal link between these two types of signals . We propose that the sequence conservation reflects , at least in part , the roles of these two locations in regulating mRNA levels , though further work will be needed to verify the connection . Similarly , the TSS peak also matches up closely with the peak binding densities of a collection of transcription factors that are involved in transcription initiation ( reported previously by the ENCODE group , based on ChIP-chip data collected for a set of regions spanning ∼1% of the genome [4] ) . As might be expected , the ENCODE data identified almost no transcription factor binding near the TES . We return to these latter observations in the Discussion . We next used our hierarchical model to examine the impact of various types of functional annotation on the probability that a SNP is an eQTN . We first classified SNPs that lie inside genes into categories based on the exon/intron structure ( e . g , first , coding and last exons; first , internal , and last introns; Figure S13 ) . In order to make the model fully identifiable , we estimate the effect of each annotation relative to the abundance of eQTNs in internal introns ( as this category has the greatest number of SNPs ) . Since gene position is highly predictive of eQTN abundance , we controlled for SNP position using the TSS+TES model . In effect , the hierarchical model now tests whether the annotation adds any predictive value beyond the basic position information . As noted above , incomplete SNP ascertainment in HapMap means that we will generally underestimate–perhaps substantially–the impact of relevant annotations . The main result of this first analysis is that internal introns have a deficit of eQTNs , compared to coding exons , as well as first and last exons and introns ( Figure 5 , Table S1 ) . For example , SNPs in coding exons are ∼2-fold more likely than SNPs in internal introns to be eQTNs . First introns are also relatively enriched for eQTNs compared to internal introns ( controlling for position ) . However , since the total amount of sequence contained in introns vastly exceeds that in exons , 53% of genic eQTNs lie in internal introns compared to 10% in coding exons ( see Table S1 ) . SNP density differs slightly between exons and introns , but not nearly enough to generate a 2-fold difference in eQTN abundance ( Table S2 ) . Overall , the hierarchical model that includes the gene structure annotation as well as position effects relative to the TSS and TES is substantially better than the TSS+TES-only model ( by 7 units of AIC ) . We then considered the impact of a variety of other types of SNP annotation ( see Methods and Figure S14 ) . None of these annotations shows convincing signals of enrichment ( Table S3 ) . We estimate a 1 . 9-fold enrichment of eQTNs inside conserved noncoding elements , as might be expected if these identify functional elements , however the 95% confidence interval narrowly overlaps 1 . We also tested for an enrichment of eQTNs at computationally predicted microRNA binding sites , reasoning that SNPs in these binding sites might affect mRNA degradation . We found a suggestive , but non-significant , enrichment of eQTNs in these sites ( 1 . 4-fold ) . It is unclear whether the absence of significant effects in these analyses indicates that these types of annotation are not strongly associated with eQTNs or instead reflects the incompleteness of HapMap and the limitations of current functional annotations . Finally , based on ENCODE results showing that the promoter regions of genes with CpG islands tend to have more accessible chromatin and greater occupancy by transcription factors [4] , we predicted that CpG status might also provide relevant annotation . Indeed , we find that genes with a CpG island spanning the TSS are expressed at higher average levels , and are ∼50% more likely than genes without a CpG island on the TSS to have a cis-eQTN ( 15% vs 11% ) . This effect is consistent with the observation that genes with CpG islands are more likely to be expressed in a wide range of tissues than are genes without CpG islands [31] . After adjusting for the different overall rates of eQTNs , the distribution of signal locations in the two classes of genes is very similar ( Figure S15 ) .
Cells use a variety of mechanisms at the transcriptional and translational levels to regulate gene expression . Transcription initiation is controlled by the interaction between transcription factors and cofactors with a set of cis-acting regulatory elements including core and proximal promoters that lie close to the TSS , as well as enhancers , silencers and boundary elements that may act at a distance [32] , [33] , [34] , [35] . Initiation is also affected by epigenetic properties of the DNA such as chromatin condensation and DNA methylation . After transcription initiation , mRNA levels can also be regulated during mRNA elongation or splicing and by mRNA stability and degradation . However , for most genes , transcription initiation is usually thought to be the principal determinant of the overall mRNA gene expression profile [34] , [35] . Consistent with the importance of transcription initiation , we found a strong peak of eQTNs near the TSS , with 33% of eQTNs lying within 10 kb of the TSS . Many of these eQTNs are likely to be polymorphisms that change the binding strength of transcription factor binding sites , thereby affecting the rate of transcription [22] . We also found that eQTNs are distributed roughly symmetrically around the TSS , with the peak density in ∼1 kb on either side ( c . f . [20] ) . Our results at the TSS are consistent with recent observations by the ENCODE team that the peak density of transcription factor binding is centered on the TSS ( Figure 4C ) . These observations indicate that empirical scans for regulatory variants that only look upstream of the core promoter [e . g . , 13] , [36] may often miss important sites of regulation . In addition to the peak of eQTN signals near the TSS , we were intrigued to find a second , similarly strong peak near the TES , as seen previously in yeast [22] . This peak is more concentrated than the TSS peak , localizing immediately before the TES , and dropping extremely rapidly after the TES . We also found that , after controlling for position effects , SNPs in exons are consistently more likely than SNPs in internal introns to be eQTNs . These observations suggest that an important fraction of eQTNs may affect properties of the transcript , rather than of the DNA sequence . We hypothesize that these eQTNs are typically polymorphisms that affect transcript stability or the rate of transcript degradation [37] , [38] , [39] , [40] , [41] . In contrast to transcription initiation , mRNA stability has been less widely studied and we still have an incomplete picture of the mechanisms that determine transcript persistence . One such mechanism is the hybridization of microRNAs to single strand transcripts , thereby exposing them as targets for degradation . Hence a SNP that creates or disrupts a match between a microRNA and the transcript might affect the rate of degradation [40]; however we did not find a significant enrichment of eQTNs in predicted microRNA binding sites . An alternative explanation for the overrepresentation of eQTNs in exons is that in some cases these may cause alternative splicing of the exon containing the expression probe , thereby changing measured expression levels . In particular , SNPs in the last exon might sometimes affect the location of the TES [21] , perhaps even deleting the expression probe from the transcript . While this mechanism probably accounts for some of the data , we do not believe it is the main explanation for several reasons . First , we found that the TSS+TES model was significantly better than the TSS+probe model . If the effect was mainly due to SNPs that affect alternative splicing of the exon containing the probe , we anticipate that those SNPs would usually lie nearer to the probe than to the TES . In that case the TSS+probe model should have performed best . Second , in a separate analysis , we observed an enrichment of signals near the TES in Affymetrix exon array data when we combined data across probes from multiple exons ( results not shown , data from [21] ) . Third , the striking peak of sequence conservation near the TES ( Figure 4B ) indicates that this is a region with strong functional significance , presumably due to an important role in gene regulation . Our results also imply that surprisingly few eQTNs with large effects lie far upstream of the TSS ( or downstream of the TES ) : for example , just 5% of the eQTNs that we detected were more than 20 kb upstream of the TSS . These results are consistent with data showing that most transcription factor binding sites are near the TSS [4] . However , since our method focuses on the major eQTN in each gene , we may under-estimate the abundance of distant eQTNs if these typically have smaller effect sizes ( [12] ) . By focusing on SNPs , our analysis may miss the impact of other types of variation–such as copy number variation–that might plausibly exert effects over different physical scales . It is also possible that more distant elements are less likely to be disrupted by single nucleotide changes . Finally , it will be important to revisit the questions that we considered here in a range of other tissues . By studying cell lines , we may underestimate the importance of long-range enhancers that turn genes on or off depending on conditions outside the cell ( e . g . , during development ) . In summary , our results show that eQTL studies provide a remarkably high-resolution tool for identifying variants that affect gene expression . A major strength of the eQTL approach is that , unlike other experimental techniques that are more targeted , the eQTL approach is agnostic about the mechanism of action of the functional variants , provided only that they are encoded in the DNA sequence ( as opposed to epigenetic factors , for example ) . Hence , eQTL studies can provide a relatively unbiased view of the importance of different types of regulatory mechanisms . Moreover , as the cost of genome sequencing drops , it will soon be possible to conduct these analyses with nearly complete ascertainment of variation , potentially providing this approach with the resolution to study the sequence level determinants of gene expression . We anticipate that eQTL mapping will make an essential contribution to our understanding of human gene regulation .
We analyzed genotype and expression data from 210 unrelated individuals studied by the International HapMap project [24] , [25] . These include 60 Yoruba ( YRI ) and 60 CEPH ( CEU ) parents , and 45 unrelated Chinese ( CHB ) and 45 unrelated Japanese ( JPT ) individuals . We used the HapMap Phase II genotype data , release #21 ( phased and with missing data imputed ) . We used data from the 22 autosomal chromosomes only , giving a total of 3 , 304 , 587 SNPs . Since allele frequencies in CHB and JPT are extremely similar [24] , these two samples were treated as a single analysis panel of 90 Asians ( “ASN” ) . We used gene expression levels that were measured previously in lymphoblastoid cell lines from all 210 unrelated individuals , using Illumina's human whole-genome expression array ( WG-6 version 1 ) [19] . We downloaded the data that were normalized first by quantile normalization within replicates and then median normalized across all HapMap individuals [19] [ ftp://ftp . sanger . ac . uk/pub/genevar/] . Since mean expression levels at many loci differ between the HapMap populations [26] , [42] , [20] , [27] , there is a potential for spurious eQTLs in the combined sample due to population structure . To control for this effect , we applied a normal quantile transformation to the data for each gene , within each HapMap population ( ASN , CEU , YRI ) , prior to combining the samples . That is , for each gene , separately in each population , we transformed the rth largest gene expression value to the ( r−0 . 5 ) /nth quantile of the standard normal distribution , where n is the number of individuals with gene expression data from that population [29] . By forcing each population to have the same distribution of expression values , we avoid concerns about spurious associations due to allele frequency differences between the HapMap populations . ( Note that the overall results within populations are very similar; Figures S2 , S3 , and S4 . ) This normalization also reduces the effect of outlying expression values on the regression [29] . We used BLAT [43] to map the 47 , 294 Illumina array probes onto human RNA sequences from RefSeq ( hg18 ) [44] . The accession numbers of the RNA sequences were mapped against the Entrez Gene database and all probes that mapped with greater than 90% identity to multiple genes were discarded . Of the remaining probes we retained only those with exact matches to a unique gene , leaving us with 19 , 536 valid probes . Of these , we kept the 13 , 244 probes for which the gene has a single RNA accession in the RefSeq database . This was done to simplify the analysis by avoiding genes with multiple splice forms or multiple annotated start sites , etc . These 13 , 244 probes map to 12 , 227 unique autosomal genes . Of these 12 , 277 genes , 85% contained exactly one probe . For the genes with multiple probes , we analyzed only a single probe , selecting the probe nearest to the 5′ end of the gene . We selected this probe because overall the probes are strongly biased towards the 3′ end of the gene , and we wanted to reduce this bias as far as possible . Then , we removed 634 genes for which there was at least one HapMap SNP inside the probe since it is known that such SNPs can impact the measured expression level [30] . Finally , 147 very large genes ( size greater than 500 kb ) were discarded , leaving our core data set of 11 , 446 genes . Gene structure annotation was obtained from the RefSeq gene table [44] for human genome build 35 ( hg17 ) . For each gene the TSS and the TES genomic locations were obtained from the fields “Transcription start position” and “Transcription end position” of the RefSeq table , respectively . We checked the genomic positions of the TSSs against dbTSS , a database of experimentally-determined TSSs , [45] and found no differences among the 84% of gene transcripts in our data set that are also in dbTSS . We defined the CDS ( coding sequence ) to be everything between the translation start and stop positions defined by the fields “cdsStart” and “cdsEnd” , respectively , of the RefSeq Table . We then assigned every genic SNP to one of 8 mutually exclusive gene-related annotations ( see Figure S16 ) : We also included annotations that indicate whether a SNP is in the following special categories: SNP is in a ( 1 ) CpG island; ( 2 ) conserved noncoding region; ( 3 ) predicted cis-regulatory module; ( 4 ) predicted micro-RNA binding site; or that ( 5 ) a predicted binding site of the CTCF insulator protein lies between the SNP and the TSS . See the Supplementary Methods ( Text S1 ) for further details . Finally , note that in our analysis design , each SNP is tested for association with every gene that is within 500 kb . This means that typical SNPs contribute data to multiple genes . Our analysis treats these multiple tests as independent , which is likely a good approximation since we identified only five SNPs that are eQTLs for > one gene in cis . We present here an overview of the hierarchical model . Complete details on the models are provided in the Supplementary Methods section ( Text S1 ) . To compute the average sequence conservation as a function of position for Figure 4B , we estimated the average number of substitutions per site across the phylogeny of seven mammalian species ( human , chimpanzee , macaque , mouse , rat , dog , and cow ) , using data and alignments from the UCSC browser . This was done for the main set of 11 , 446 genes analyzed in this paper . For each gene , 5 kb on each side of the TSS ( and separately for the TES ) was split into non-overlapping 50-bp bins . We then concatenated all the sites across all genes that lay in the same bin . After excluding sites in coding exons we estimated the average number of substitutions at each site using baseml , a program in the PAML package [47] . We obtained results on transcription factor binding density using ChIP-chip data collected by the ENCODE project ( 4 ) . We used data for eight transcription factors that showed large numbers of binding fragments at a 1% false discovery rate in the ENCODE study . The left-hand panel of Figure 4C is essentially a replotting of data presented in Figure 5 of ( 4 ) , while the right-hand panel shows analogous data plotted with respect to the TES . | Individual phenotypes within natural populations generally exhibit a large diversity resulting from a complex interplay of genes and environmental factors . Since the advent of molecular markers in the 1980s , quantitative genetics has made a significant step toward unraveling the genetic bases of such complex traits , in particular by developing sophisticated tools to map the genomic locations of genes that affect complex traits . These regions are known as quantitative trait loci ( QTLs ) . More recently , these tools have been extended to the study of gene expression phenotypes on a massive scale . In this paper , we used a previously published dataset consisting of expression measurements of 11 , 446 genes in human cell lines derived from 210 unrelated human individuals that have been genetically characterized by the International HapMap Project . Our article develops and applies a framework for determining the genetic factors that impact gene regulation . We show that these factors cluster strongly near to the gene start and gene end and are enriched within the transcribed region . Our approach suggests a general framework for studying the genetic factors that affect variation in gene expression . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"genetics",
"and",
"genomics/complex",
"traits",
"genetics",
"and",
"genomics/gene",
"expression"
] | 2008 | High-Resolution Mapping of Expression-QTLs Yields Insight into Human Gene Regulation |
We quantified Leishmania infantum parasites transmitted by natural vectors for the first time . Both L . infantum strains studied , dermotropic CUK3 and viscerotropic IMT373 , developed well in Phlebotomus perniciosus and Lutzomyia longipalpis . They produced heavy late-stage infection and colonized the stomodeal valve , which is a prerequisite for successful transmission . Infected sand fly females , and especially those that transmit parasites , feed significantly longer on the host ( 1 . 5–1 . 8 times ) than non-transmitting females . Quantitative PCR revealed that P . perniciosus harboured more CUK3 strain parasites , while in L . longipalpis the intensity of infection was higher for the IMT373 strain . However , in both sand fly species the parasite load transmitted was higher for the strain with dermal tropism ( CUK3 ) . All but one sand fly female infected by the IMT373 strain transmitted less than 600 promastigotes; in contrast , 29% of L . longipalpis and 14% of P . perniciosus infected with the CUK3 strain transmitted more than 1000 parasites . The parasite number transmitted by individual sand flies ranged from 4 up to 4 . 19×104 promastigotes; thus , the maximal natural dose found was still about 250 times lower than the experimental challenge dose used in previous studies . This finding emphasizes the importance of determining the natural infective dose for the development of an accurate experimental model useful for the evaluation of new drugs and vaccines .
Leishmania are intracellular protozoan parasites that establish infection in mammalian hosts following transmission through the bite of an infected phlebotomine sand fly . Visceral leishmaniasis , caused by Leishmania donovani in the Old World and L . infantum in both the Old and New World , invariably leads to death if left untreated [1] . Despite the fact that parasites from the L . donovani complex are mainly associated with disseminated infection of the spleen and liver , it has been shown that L . infantum can also cause cutaneous lesions [2]–[5] . A novel focus of cutaneous leishmaniasis caused by L . infantum was recently described in the Cukurova region in Turkey [6] . During the natural transmission of Leishmania into the dermis , sand flies deposit pharmacologically active saliva [7] and egest parasite-released glycoconjugates , the promastigote secretory gel [8] . Both substances modulate the immune response of the bitten host and enhance the severity of infection ( reviewed by [9] ) . The ideal leishmaniasis model to test therapeutics and immunoprophylaxis candidates should reproduce the biological and immunological aspects of natural infection and disease . Different approaches regarding the parasite number and route of inoculation have been tested in order to develop an accurate experimental model for the L . donovani complex , most of them using subcutaneous , intraperitoneal or intravenous injections of millions of axenic promastigotes or amastigotes [10]–[11] . Although in some studies up to 107 parasites have been co-inoculated into the dermis with small amounts of sand fly saliva , is not clear how well these experiments mimic natural transmission [12]–[13] . The number of L . infantum parasites inoculated by infected vectors during natural transmission was not previously known , even though a determination of the natural infective dose is crucial for the development of an accurate experimental model to evaluate new drugs and vaccine candidates . In the L . major - P . duboscqi model , it was demonstrated that the number of promastigotes inoculated by individual sand flies ranged between 10 and 1×105 Leishmania [14] . The average number of L . infantum parasites egested was recently reported [15] , but the technique used ( feeding the pool of infected L . longipalpis through chick skin membrane on culture medium ) did not allow an evaluation of the variation in numbers delivered by individual sand flies . Thus , the main aims of this work were to determine the transmission rate and the number of promastigotes inoculated into the skin of mice by individual sand fly females . Phlebotomus perniciosus and Lutzomyia longipalpis , two main vectors of L . infantum in the Mediterranean basin and in the New World , respectively [16] , were experimentally infected by L . infantum dermotropic and viscerotropic parasites .
The L . infantum strains studied developed well in both P . perniciosus and L . longipalpis , producing heavy late-stage infection and colonizing the stomodeal valve of the vectors , which is a prerequisite for successful transmission . For both L . infantum strains , the average parasite load in the sand fly midgut is summarized in Table 1 . Quantitative PCR revealed that in P . perniciosus the intensity of infection was higher for the CUK3 strain ( p = 0 . 01 ) while L . longipalpis harboured more IMT373 parasites ( p<0 . 001 ) . However , in both sand fly species the number of parasites transmitted was higher for the dermotropic strain CUK3 ( p<0 . 001 ) ; see below . Out of 88 P . perniciosus , females that bit mice , 62 ( 70 . 5% ) were infected with CUK3; of these , 36 ( 58% ) delivered parasites into the skin of the mice on days 10–14 post infective blood meal ( Fig . 1a ) . Out of 114 biting L . longipalpis females , 86 ( 75 . 5% ) were infected and 56 ( 65% of those infected ) inoculated parasites into the mice on days 7–14 post infective blood meal ( Fig . 1b ) . Despite the fact that the intensity of infection was significantly higher in P . perniciosus ( p<0 . 01 ) , the percentage of transmission and number of inoculated parasites was comparable for both vectors . The parasite load delivered by P . perniciosus and L . longipalpis in the skin of mice ranged between 16 and 4 . 19×104 and between 4 and 1 . 11×104 , respectively . The average number of CUK3 parasites inoculated into the skin of mice and the percentages of transmission are summarized in Table 1 . In L . longipalpis , the feeding time was positively correlated with the number of CUK3 parasites delivered into host skin ( p<0 . 05 ) , while in P . perniciosus females no such correlation was observed . On the other hand , there was a significant correlation between the pre-feeding load inside both sand fly species' midguts and the number of parasites transmitted ( p = 0 . 0178 for L . longipalpis and p<0 . 001 for P . perniciosus ) . Out of 101 P . pernicious females that bit mice , 73 ( 72% ) were infected with IMT373 , and of these 24 ( 33% ) transmitted parasites into the mice's skin . Leishmania transmission occurred between days 9 and 16 post infective bloodmeal ( Fig . 2a ) . From 190 biting L . longipalpis females , 159 ( 84% ) were infected and 23 ( 14 . 5% on infected ones ) inoculated parasites into the mice between days 7 and 14 post blood meal ( Fig . 2b ) . In contrast to above , the intensity of infection was significantly higher in L . longipalpis ( p<0 . 001 ) , but the transmission rate ( i . e . percentage of transmitting females ) and the number of parasites transmitted were significantly higher in P . perniciosus ( p<0 . 01 ) . The number of parasites transmitted by P . perniciosus and L . longipalpis ranged from 8 to 513 and between 7 and 1240 promastigotes , respectively . The median number of IMT373 transmitted is summarized in Table 1 . For both sand fly species , there was no correlation between feeding time and the number of IMT373 parasites in each female ( p = 0 . 1594 ) , or between the time to take a blood meal and the number of parasites transmitted ( p = 0 . 6666 ) . Moreover , no correlation was observed between the pre-feeding load in each sand fly species and the number of Leishmania delivered ( p = 0 . 1340 for P . perniciosus; p = 0 . 6473 for L . longipalpis ) . For all Leishmania-sand fly combinations , ears were the preferential biting place for sand flies transmitting the parasites , followed by the paws and tail . A few specimens that fed in the nose and eyes were also able to transmit parasites . Table 2 summarizes the feeding times for both sand fly species: L . longipalpis transmitting IMT373 completed their bloodmeals in times ranging from 2 to 27 minutes , while those transmitting CUK3 parasites needed between 3 to 55 minutes . The maximum and minimum feeding times for P . perniciosus transmitting CUK3 and IMT373 parasites ranged between 4–33 and 1–32 minutes , respectively . Infected sand flies transmitting CUK3 needed more time to feed than those that were infected but non-transmitting , while no differences in feeding time were observed between transmitting and non-transmitting females with IMT373 parasites .
For the first time , we have quantified the number of parasites belonging to L . infantum dermotropic and viscerotropic strains transmitted to the dermis of experimental mice by individual sand fly females . The only previous attempt to calculate the number of transmitted L . infantum parasites was performed just recently [15] , with the average number of promastigotes inoculated by 63 L . longipalpis into culture medium through a chicken membrane skin being 457 parasites , with 95% ( 431 promastigotes ) of these corresponding to metacyclic parasites . However , these results do not allow us to take into consideration the individual variability of parasite transmission by a single specimen . The wide range of parasites inoculated per individual sand fly in our study ( from 4 up to 4 . 19×104 promastigotes ) is in accordance to data previously obtained with other Leishmania-vector combinations [14] , [17] , although the approach using microcapillaries as artificial feeding systems [17] could have interfered with the normal sand fly feeding behaviour . In our study , Phlebotomus perniciosus harboured more L . infantum dermotropic parasites of the CUK3 strain , while in L . longipalpis the intensity of infection was higher for the viscerotropic strain IMT373 . However , in both sand fly species the parasite load transmitted was higher for the strain with dermal tropism . All but one sand fly female infected by IMT373 strain transmitted less than 600 promastigotes , the exception being a L . longipalpis female that inoculated 1240 parasites . On the other hand , 29% of L . longipalpis and 14% of P . perniciosus infected with the CUK3 strain transmitted more than 1000 parasites . The majority of transmitting females inoculated less than 600 parasites . As most of these females were fully engorged by blood we may expect that their feeding pumps ( the cibarial and pharyngeal pumps ) and stomodeal valve were functioning normally . On the other hand , in those transmitting more than 1000 parasites there was a significant correlation between the pre-feeding load and the number of parasites transmitted . We suggest that these females with high dose deliveries regurgitated parasites because of impaired stomodeal valve function [18] . This would be consistent with previous studies [19] , [20] which have demonstrated an opened stomodeal valve due to the physical presence of a parasite plug and damage of the chitin layer of the valve by Leishmania chitinase . Infected sand fly females , and especially those that transmit parasites , feed longer on hosts than non-transmitting ones do . Lutzomyia longipalpis females transmitting dermotropic CUK3 strain parasites took an average of 1 . 5 times longer to complete a bloodmeal compared to specimens infected but not transmitting , and 1 . 8 times longer than uninfected females . Similarly , P . perniciosus infected by CUK3 and IMT373 take 1 . 5 and 1 . 2 times more time for a blood meal . Most of the infected sand flies exposed to anaesthetized mice did not demonstrate increased probing , but rather remained feeding for longer periods until either they were fully or partially engorged . This is in agreement with data previously published on the L . longipalpis-L . mexicana combination [21] . Although only one dermotropic and one viscerotropic L . infantum strains were evaluated , the significant variation in inoculum size between them allow us to hypothetise that the infectious dose delivered by vector sand flies may be an inherent character of each Leishmania strain . Moreover , the infectious dose might be a determining factor in the outcome of Leishmania infection . Local cutaneous lesions might result from a high-dose inoculum of dermotropic Leishmania resulting in a strong local immune response , whereas dissemination to internal organs might be the result of infected sand flies delivering a low number of parasites below the threshold required to produce/develop a localized and restraining immune response . This hypothesis corresponds with the data of Kimblin et al . [14] on the L . major-P . duboscqi combination . These authors evaluated the impact of inoculum size on infection outcome by comparing L . major infections with high ( 5×103 ) and low ( 1×102 ) dose intradermally inoculated by needle in the ears of C57BL/6 mice , and observed the rapid development of large lesions in the ears of mice receiving the high-dose inoculum . In contrast , the low dose resulted in only minor pathology but a higher parasite titre during the chronic phase [14] . Nevertheless , it will be necessary to evaluate more L . infantum strains with visceral and cutaneous tropism in order to determine if differences detected in our study were due to individual stock characteristics or if they are associated with parasite tropism in vertebrate hosts . In conclusion , we have demonstrated that individual sand flies transmit Leishmania parasites in a wide dose range . However , the maximal natural dose found was still about 250 times lower than the challenge dose used for the L . donovani complex in most previous experimental works . This finding emphasizes the importance of determining the natural infective dose for the development of an accurate experimental model , which is crucial for the evaluation of new drugs and vaccine candidates against leishmaniasis .
The viscerotropic Leishmania infantum strain IMT373 MON-1 ( MCAN/PT/2005/IMT373 ) and the dermotropic L . infantum strain CUK3 ( ITOB/TR/2005/CUK3 ) were used in this study . CUK3 was isolated from Phlebotomus tobbi from a Cukurova focus of cutaneous leishmaniasis [6] while IMT373 was isolated from a dog with leishmaniasis and passaged through mice in order to keep its virulence [13] , [22] . Promastigotes ( with less than 12 in vitro passages since isolation ) were cultured at +26°C in M199 medium ( Sigma , USA ) containing 10% heat-inactivated foetal calf serum ( Gibco , USA ) , 50 mg/ml mikacin solution ( Bristol-Myers Squibb , Czech Republic ) and 1% sterile urine . Lutzomyia longipalpis ( originating from Jacobina , Brazil ) and Phlebotomus perniciosus ( originating from Murcia , Spain ) colonies were maintained in an insectary under standard conditions as described by Volf and Volfova [23] . Five to six-day old female flies ( 200 P . perniciosus and 150 L . longipalpis females per experiment , respectively ) were fed on heat inactivated rabbit blood containing promastigotes ( 107 parasites per ml of blood ) through a chicken-skin membrane . Blood-engorged females were separated immediately and maintained on a 50% sucrose diet in >70% relative humidity at +26°C . One group of females was dissected to study the development and localization of infection in the sand fly midgut two and ten days post blood meal , i . e . , during early and late stage infection , respectively . Individual midguts were placed into a drop of saline buffer , and parasite numbers were estimated under a light microscope at 200X and 400X magnifications by an experienced worker . Parasite loads were graded as previously described [24] into four categories: negative , 1–100 , 100–1000 , and >1000 parasites per gut . A second group of females from the same batch was used for transmission experiments and parasite quantification by Real-time PCR ( see below ) . Nine and six independent experiments were performed with P . perniciosus-IMT373 and P . perniciosus-CUK3 combinations , respectively , while six and four artificial infections were done with L . longipalpis-IMT373 and L . longipalpis-CUK3 combinations . One hundred and eight BALB/c mice ( 41 for P . perniciosus-IMT373 , 28 mice for L . longipalpis-IMT373 , 23 for P . perniciosus–CUK3 and 16 for L . longipalpis-CUK3 combinations ) older than 8 weeks of age were purchased from AnLab ( Czech Republic ) and housed at Charles University , Prague , under stable climatic and dietary conditions . Experiments were approved by the institutional Ethical Committee and performed in accordance with national legislation for the care and use of animals for research purposes . Mice were anaesthetized intraperitoneally with ketamine ( 150 mg/kg ) and xylazine ( 15 mg/kg ) . Sand fly females were allowed to feed on whole body of anesthetised mice in a rectangular cage ( 20×20 cm ) for about one hour at various days post infective blood meal ( 7–14 days for L . longipalpis and 9–23 days for P . perniciosus ) . Each mouse was placed individually into a cage together with about 50 P . perniciosus or 10 L . longipalpis females ( the difference was due to the fact that L . longipalpis were more aggressive and had higher feeding rate ) . Two people followed each experiment; one recorded biting place and feeding time while the second ensured that each sand fly female probed in different place and then collected engorged flies by an aspirator immediately after terminating their blood meal; the site of bite and time of feeding were recorded for each female . After exposure , mice were sacrificed , biting place was inspected under a stereoscope and excised . Both samples ( skin biopsies and corresponding fed sand flies ) were stored at −20°C until DNA extraction . Extraction of total DNA from each bite site and the corresponding sand fly were performed using a DNA tissue isolation kit ( Roche Diagnostics , Germany ) according to the manufacturer's instructions . DNA was eluted in 100 µl and stored at −20°C . qPCR for detection and quantification of Leishmania sp . was performed in a Rotor-Gene 2000 from Corbett Research ( St . Neots , UK ) using the SYBR Green detection method ( iQ SYBR Green Supermix , Bio-Rad , Hercules , CA ) . For adequate sensitivity , kinetoplast DNA was chosen as the molecular target , with primers as previously described [25] ( forward primer 5′-CTTTTCTGGTCCTCCGGGTAGG-3′ and reverse primer 5′-CCACCCGGCCCTATTTTACACCAA- 3′ ) . Two microliters of eluted DNA was used per individual reaction . PCR amplifications were performed in duplicate wells using the conditions described previously [26] . Briefly , 3 min at 95°C followed by 45 cycles of: 10 s at 95°C , 10 s at 56°C , and 10 s at 72°C . Reaction specificities were checked for all samples by melting analysis . Quantitative results were expressed by interpolation with a standard curve included in each PCR run . Mass cultures of L . infantum promastigotes were used to construct a series of 10-fold dilutions ranging from 105 to 1 parasite per PCR reaction . Diluted parasites were co-processed with mouse tissue or sand fly females for DNA extraction . DNA from uninfected sand flies and mice were used as a negative control . For sand fly females transmitting promastigotes into mouse skin , the pre-feeding midgut load was calculated as the sum of parasites in the midgut after feeding and the number of parasites transmitted . Statistical analysis was performed using the software STATISTICA . For each L . infantum strain , a nonparametric Kruskal-Wallis test was used to compare: ( i ) the intensity of infection in P . perniciosus and L . longipalpis , and ( ii ) the number of parasites transmitted by each sand fly species into mice's skin . Correlations between feeding time and the number of parasites ( i ) in each sand fly female and ( ii ) inoculated into the skin , as well as the correlation between pre-feeding load and the number of parasites transmitted , were determined by simple linear regression analysis . Differences were considered statistically significant for p values <0 . 05 . | Leishmaniasis is a disease caused by protozoan parasites which are transmitted through the bites of infected insects called sand flies . The World Health Organization has estimated that leishmaniases cause 1 . 6 million new cases annually , of which an estimated 1 . 1 million are cutaneous or mucocutaneous , and 500 , 000 are visceral , the most severe form of the disease and fatal if left untreated . The development of a more natural model is crucial for the evaluation of new drugs or vaccine candidates against leishmaniases . The main aim of this study was to quantify the number of Leishmania infantum parasites transmitted by a single sand fly female into the skin of a vertebrate host ( mouse ) . Two L . infantum strains , viscerotropic IMT373 and dermotropic CUK3 , were compared in two natural sand fly vectors: Phlebotomus perniciosus and Lutzomyia longipalpis . We found that the parasite number transmitted by individual sand flies ranged from 4 up to 4 . 19×104 . The maximal natural infective dose found in our experiments was about 250 times lower than the experimental challenge dose used in most previous studies . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"parastic",
"protozoans",
"medicine",
"infectious",
"diseases",
"veterinary",
"diseases",
"zoonotic",
"diseases",
"neglected",
"tropical",
"diseases",
"protozoology",
"biology",
"microbiology",
"veterinary",
"science"
] | 2011 | Experimental Transmission of Leishmania infantum by Two Major Vectors: A Comparison between a Viscerotropic and a Dermotropic Strain |
To support the Bangladesh National Kala-azar Elimination Programme ( NKEP ) , we investigated the feasibility of using trained village volunteers for detecting post-kala-azar dermal leishmaniasis ( PKDL ) cases , using polymerase chain reaction ( PCR ) for confirmation of diagnosis and treatment compliance by PKDL patients in Kanthal union of Trishal sub-district , Mymensingh , Bangladesh . In this cross-sectional study , Field Research Assistants ( FRAs ) conducted census in the study area , and the research team trained village volunteers on how to look for PKDL suspects . The trained village volunteers ( TVVs ) visited each household in the study area for PKDL suspects and referred the suspected PKDL cases to the study clinic . The suspected cases underwent physical examinations by a qualified doctor and rK39 strip testing by the FRAs and , if positive , slit skin examination ( SSE ) , culture , and PCR of skin specimens and peripheral buffy coat were done . Those with evidence of Leishmania donovani ( LD ) were referred for treatment . All the cases were followed for one year . The total population of the study area was 29 , 226 from 6 , 566 households . The TVVs referred 52 PKDL suspects . Probable PKDL was diagnosed in 18 of the 52 PKDL suspect cases , and PKDL was confirmed in 9 of the 18 probable PKDL cases . The prevalence of probable PKDL was 6 . 2 per 10 , 000 people in the study area . Thirteen PKDL suspects self-reported from outside the study area , and probable and confirmed PKDL was diagnosed in 10 of the 13 suspects and in 5 of 10 probable PKDL cases respectively . All probable PKDL cases had hypopigmented macules . The median time for PKDL development was 36 months ( IQR , 24–48 ) . Evidence of the LD parasite was documented by SSE and PCR in 3 . 6% and 64 . 3% of the cases , respectively . PCR positivity was associated with gender and severity of disease . Those who were untreated had an increased risk ( odds ratio = 3 . 33 , 95%CI 1 . 29–8 . 59 ) of having persistent skin lesions compared to those who were treated . Patients' treatment-seeking behavior and treatment compliance were poor . Improved detection of PKDL cases by TVVs is feasible and useful . The NKEP should promote PCR for the diagnosis of PKDL and should find ways for improving treatment compliance by patients .
Post-kala-azar dermal leishmanisis ( PKDL ) , is a skin disorder , usually develops in 10–20% and about 60% of patients with visceral leishmaniasis ( VL ) /kala-azar after treatment , respectively , in the Indian subcontinent and Sudan [1] . It has also been reported in individuals without prior history of VL and those undergoing treatment for VL [1]–[6] . The protozoan parasite Leishmania donovani ( LD ) is the only causative agent . Clinical manifestations of PKDL are macular , maculo-papular , and nodular rash in people who are otherwise well [1] and may be confused with leprosy . Since PKDL is the only interepidemic reservoir of anthroponotic VL , the existence of a few cases is sufficient to trigger a new epidemic of VL in a given community [1] , [4] , [7] . Thus , identification and treatment of PKDL is an essential strategy in eliminating VL . In 2005 , Health Ministers of Bangladesh , India , and Nepal signed a Memorandum of Understanding for the elimination of VL from the Indian subcontinent by 2015 [8] . Active VL and PKDL case detection and their proper management are two important strategies of the elimination program [8] . Until now , no definite method has been identified for active VL and PKDL case detection but one proposed plan includes a house-to-house search for cases by public-health workers . Such a method is expensive and requires alternative strategies for VL and PKDL active case detection [9] . In the Indian subcontinent , PKDL was first described by Brahmchari in 1922 [10] . Since then , nearly 90 years have passed and no gold standard diagnostic method could be developed for PKDL; diagnosis , thus , relies on clinical criteria [1]–[6] . PCR of slit skin scraping specimens has demonstrated high sensitivity for diagnosing PKDL in a laboratory-based study [11] . However , its use in clinically-diagnosed PKDL patients is unknown . PCR testing of peripheral blood buffy coat has been found to be a highly-sensitive method for the diagnosis of VL [12] , [13] , and theoretically , it may help also confirm PKDL . In Bangladesh , in the post-malaria-eradication era , the first reports on PKDL were from hospital- based studies [14] , [15] . Until now there has been limited information on the burden of PKDL in the VL-endemic communities of Bangladesh [4] . Preliminary results of an ongoing surveillance study of PKDL in Fulbaria , Mymensingh , Bangladesh , showed that the burden of PKDL was high and presents a challenge for the National Kala-azar Elimination Programme ( NKEP ) [4] . More information on the burden of PKDL will help the NKEP to develop adequate national strategies for controlling PKDL . We therefore studied the feasibility of using trained village volunteers ( TVVs ) for detecting suspected PKDL cases; estimated the prevalence of PKDL in Kanthal Union , Trishal , a VL-endemic area of Bangladesh; described the clinical features of these patients; evaluated the contribution of PCR for the confirmation of PKDL diagnosis in clinically-diagnosed PKDL cases; and investigated the patients' compliance to treatment .
We conducted a study in Kanthal union , Trishal sub-district under Mymensingh district , Bangladesh . During January-March 2008 , trained field research assistants performed a census in Kanthal union . During the census , field research assistants assessed household number and the number of people in each household . They also asked and recorded whether any member within the family suffered from kala-azar in the past and if any of them currently had skin rash . After consulting the local community leaders and obtaining their consent , one community volunteer was selected from each ward . The research team trained the volunteers on what was PKDL , how did a PKDL case looks like , and how to look for the suspected PKDL case ( see definition below ) . A two-day training was imported to the volunteers , and pictures of skin lesions from PKDL patients from published literature and textbooks were used . During April-May 2008 , nine TVVs visited each household at least once searching for suspected PKDL cases and , if found , referred the case to the study clinic . Household members who had a history of VL but were not present during the home-visits were invited to visit to the study clinic for assessment . Additionally , the study physician assessed patients with a history of VL and skin rash who lived in villages outside the study area . Most of these patients were directed to our study clinic from nearby public union health posts where rK39 tests were not available . The study physician examined the suspected PKDL cases and requested an rK39 strip test ( Kala-azar Detect ™ Rapid Test , InBios International , Seattle , WA , USA ) as needed . A trained field research assistant performed the rK39 strip test as per the manufacturer's instructions . If the results were positive , the patients were considered probable PKDL cases ( see definition below ) and were requested to undergo slit skin scraping and blood collection . A physician-microbiologist from the Bangabandhu Sheikh Mujib Medical University ( BSMMU ) in Dhaka performed slit skin scraping and collected skin specimens for staining , culture , and PCR test in the study clinic . The study physician collected up to 5 mL of venous blood in EDTA tubes and gently shook . After collection , the blood and skin specimens were transported to the Parasitology Laboratory of the International Centre for Diarrhoeal Disease Research , Bangladesh ( ICDDR , B ) within 3–4 hours maintaining cold chain . PCR tests were performed in the Parasitology Laboratory of ICDDR , B on the following day , and the study physician was informed about PCR results immediately after testing . We referred all cases positive for the LD parasite or positive for LD DNA by PCR to the Upazila Health Complex for treatment as per the national guideline for kala-azar elimination [16] . After referral , we followed up patients to find out whether they went to the hospital for admission , and if admitted , whether they completed full treatment courses . Further , we followed all the probable PKDL patients after one year from the date of their referral by household-visit and collected information on the status of their skin rash . Patients who were rK39 test-negative were referred to the Mymensingh Medical College Hospital for further medical consultation . Assuming a PKDL prevalence of 0 . 04% ( 95% CI 0 . 03–0 . 05 ) , the required sample-size needed to search for PKDL cases was 25 , 098 . A total population of 29 , 226 in the study area was sufficient for our study . All data were computed using the EpiInfo software ( version 3 . 2 . 2 . ) and were analyzed using the SPSS software ( version 11 . 0 ) through descriptive and analytical methods . Comparison between proportions was done by chi-square with Fisher's exact correction . Means were compared by ANOVA or Kruskal-Wallis where applicable . We used Kaplan-Meier survival curves to identify the median time for onset of PKDL after the completion of VL treatment . All p values were two-tailed , and a p value of ≤0 . 05 was considered significant . The Ethical Review Committees of both ICDDR , B and Tropical Disease Research/WHO approved the study . Written informed consent was obtained from each study participant and from parents/legal guardian for all child participants . Written informed consent also was obtained from the head of each household when performing the census .
The census performed by the Field Research Assistants revealed a total population of 29 , 226 people from 6 , 566 households: 11 , 298 people ( 38 . 7% ) were aged ≤15 years , and the median family size within each household was 5 ( IQR 4–6 ) . The Field Research Assistants also listed a total of 235 individuals with a past history of VL and 21 suspected PKDL patients based on the information obtained during their visits to households at the beginning of the study . However , the TVVs referred 52 suspected PKDL patients , including the 21 cases , listed by the Field Research Assistants from the study area . No cases from the study area were self- reported but 13 patients were self-reported from villages outside the study area . No new cases of suspected PKDL from the study area were found in the following months after the initial survey by the TVVs . Of the referred and self-reported suspected PKDL cases , 18 of 52 ( 34 . 6% ) and 10 of 13 ( 76 . 9% ) were respectively found to be probable PKDL cases . The referred and self-reported probable PKDL cases did not differ regarding age [mean years ±standard error ( SE ) , 24 . 1±3 . 7 for referred cases , 18 . 5±3 . 3 for self-reported cases , p = 0 . 33] , sex ( % of male patients , 55 . 6% for referred vs 40 . 0% self-reported , p = 0 . 69 ) , and duration of onset of PKDL ( mean years±SE , 46 . 0±5 . 1 for referred vs 36 . 0±23 . 1 for self-reported cases , p = 0 . 29 ) . PKDL was confirmed in 50% of the patients from each group ( 9/18 for referred and 5/10 for self- reported suspected PKDL patients ) . The calculated prevalence of probable PKDL in the study area was 6 . 2 per 10 , 000 people , slightly more common among the age-group of 15–43 years and without any relation with gender [Table 1] . We pooled all the 28 probable PKDL cases for further analysis . The median age of the patients was 21 . 5 years ( IQR , 10 . 7–29 . 0 ) , and the large majority ( 67 . 9% ) was aged ≥15 years . All the patients had previously been treated for VL with sodium stibogluconate for 28 consecutive days , except one ( 3 . 6% ) who had interruption in treatment . The median duration of onset of PKDL was 36 months ( IQR , 24–48 ) ( Fig . 1 ) . The duration of onset of PKDL was not related to gender ( 38 . 57±5 . 9 , n = 14 for females vs 46 . 71±5 . 9 , n = 14 for males , p = 0 . 34 ) but the younger group developed PKDL significantly earlier compared to the older age-group ( 26 . 67±3 . 8 months , n = 9 , vs 50 . 21±5 . 1 months , n = 19 , p = 0 . 006 ) . Hypopigmented macules were the most common skin lesions and most frequently appeared first on the face ( Table 2 ) . In most ( 94 . 4% ) cases , the skin lesions progressed gradually and were associated with itching in 25% of the cases . Eight of the 28 ( 29% ) cases , skin lesions were classified as grade I and the remaining 20 ( 71% ) as grade II at the time of presentation to the study clinic . The severity of the disease , as estimated by grading , showed no association with sex: 11 of the 14 females ( 78 . 6% ) and 9 of the 14 males ( 64 . 3% ) had grade II PKDL , p = 0 . 34 . Similarly , the severity of the disease had no association with age-group ( 78% vs 68% grade II respectively in the younger and older age-groups , p = 0 . 48 ) nor the duration of onset of PKDL ( r = 0 . 03 , p = 0 . 9 ) . All the patients were negative for M . leprae ( Table 3 ) . All the 28 cultures were negative for Leishmania parasites . Slit skin microscopy found LD bodies only in one of the 28 patients . Of the 28 patients 12 ( 42 . 9% ) were positive by slit skin scraping specimen PCR and 14 ( 50% ) by peripheral blood buffy coat PCR . The PCR method was positive in 64% of the patients , defined by any PCR test-positive ( Table 3 ) . The buffy coat PCR results were positively associated with the severity of the disease: 65% ( 13/20 ) of grade II PKDL patients were positive for peripheral buffy coat PCR compared to 12 . 5% ( 1/8 ) of the PKDL patients with grade I disease ( p = 0 . 012 ) . However , skin specimen PCR results did not show a similar relationship with PKDL grades: 45% ( 9/20 ) and 37% ( 3/8 ) respectively with grade II PKDL and grade I PKDL were positive by skin specimen PCR ( p = 0 . 72 ) . The female patients had significantly more peripheral buffy coat PCR-positive results compared to the male patients [10/14 ( 71% ) for female vs . 4/14 ( 29% ) for male patients , p = 0 . 023] . The female patients were also more likely to have a positive skin specimen by PCR ( 50% , 7/14 ) compared to the male patients ( 36% , 5/14 ) but the difference was not statistically significant ( p = 0 . 45 ) . Eighteen patients positive for the evidence of the LD parasite ( 1 by both LD body and LD DNA and 17 by LD DNA only ) by any laboratory method were referred to the sub-district hospital for treatment . Of these patients , 7 completed treatment , 6 partially completed treatment , and 5 refused treatment . The common reasons for incomplete treatment or not to be treated were concern about loss of daily wages , loss of school days , and insolvency ( Table 4 ) . All the patients with complete treatment , half of those with incomplete treatment , and none who refused treatment were cured at follow-up . Compared to those with complete treatment , the risk of not being cured with incomplete treatment or treatment refusal was three times higher ( relative risk 3 . 33 , 95% CI 1 . 29–8 . 59 , p = 0 . 004 ) . Ten of the 28 patients were not referred for treatment due to lack of evidence of the LD parasite in their skin or blood . At follow-up after one year , 8 of 10 patients had persistent skin lesions; skin lesions had deteriorated in 1 patient , and 1 was spontaneously cured .
The major findings of the present study were: the TVVs were useful for the active detection of PKDL case; the prevalence of PKDL was high in the study area; PCR was useful for the confirmation of PKDL diagnosis; and low treatment compliance and current treatment-seeking behavior of PKDL patients present a new challenge for the national kala-azar elimination programme in Bangladesh . The Government of Bangladesh is committed to eliminate VL by 2015 , and active detection of VL and PKDL cases is one of the main pillars of this elimination programme [8] , [16] . The active detection of VL and PKDL cases by the existing health facilities in rural Bangladesh may not be possible due to lack of human resources . We found that the TVVs were useful for finding suspected PKDL cases . Thus , the TVVs should be used by the NKEP for an annual active case search . We feel that it is sufficient to conduct an annual search , particularly for PKDL , since no new cases of suspected PKDL were found in our study after the initial active case search by the TVVs . The proportion of PKDL cases among the self-reported suspects was higher ( 76 . 9% ) compared to the referred PKDL cases ( 34 . 6% ) . The majority of the self-reported cases were directed to our study clinic from nearby public union health posts that lacked facilities for rK39 testing . This suggests that many people are concerned about signs and symptoms of possible PKDL but remain undiagnosed due to lack of diagnostic facilities , particularly rK39-based rapid tests , in the union health posts . The NKEP should thus , strengthen the existing public union health posts in the VL-endemic areas by equipping them with tools to diagnose PKDL . Clinical manifestations of the PKDL cases did not differ from those reported in the literature [1]–[6] , [18] . We found that the median duration of onset of PKDL was slightly longer compared to that reported recently [4] . This difference might be explained by the differences in the study design , and recall bias in our study population also might contribute to this difference . Itching was reported in 25% of the cases . This symptom among the PKDL patients was also reported by others , and inflammatory reactions in the dermis observed by Ismail et al might explain this symptom [19] . The prevalence of PKDL in our study was 6 . 2 per 10 , 000 people which was lower than that found by a recent study in another VL-endemic area of Bangladesh [3] . The limitation of our study was that the TVVs looked for the suspected PKDL cases based on the past history of VL and skin rash . Since PKDL may develop in individuals without a past history of VL , it was possible that the TVVs could not track all suspected PKDL cases . However , we believe that this risk was minimal because the percentage of PKDL without past history of VL was low ( about 10% of all PKDL cases ) in Bangladesh [4] . Additionally , no self-reported PKDL case from Kanthal union was presented to the study clinic or the Upazila Health Complex in the months following the study . Slit skin examination and culture methods were not very much useful for the diagnosis of PKDL since the LD parasite was found in only one by slit skin examination , and all the 28 cases were negative by culture method . Similar results were reported by others [18] . PCR of skin specimens and peripheral blood buffy coat was more useful for the detection of LD compared to slit skin examination and culture . PCR helped confirm the diagnosis of PKDL in additional 17 patients who were negative for LD by conventional methods . Seven of the 17 patients who received complete treatment were cured . Without this useful diagnostic tool , these 7 patients would not be considered PKDL patients because they were negative by SSE and culture and would continue to serve as a reservoir of anthroponotic VL , thus threatening the community's health . In Bangladesh , many tertiary hospitals are now equipped with PCR technology . Thus , the NKEP should consider the use of PCR for the diagnosis of PKDL . However , the feasibility and cost-effectiveness of such a diagnostic algorithm may be a concern for the NKEP and needs to be explored first . We observed a significantly-positive relationship between PKDL grade and PCR positivity , particularly with peripheral blood buffy coat PCR . The circulation of the LD parasites in peripheral blood increases with the severity of disease and indirectly indicated that the origin of LD DNA in blood was skin . So far , this is the first documentation of such a relationship between PKDL grade and PCR method . It also supports an earlier experiment that patients with PKDL may serve as a reservoir for the transmission of the LD parasite which circulates in their blood [20] . Although the severity of PKDL was equally distributed among the males and females , the positive PCR results were more common in the females , especially for peripheral buffy coat PCR . We have no explanation for this phenomenon . Ten of the 28 probable PKDL cases were not referred for treatment because we failed to demonstrate LD or LD DNA among them . At follow-up , only one had improvement in his skin rashes , which occurred spontaneously . Our current diagnostic capabilities limited our ability to determine whether these 10 cases were false-positive probable PKDL cases or if they were false-negative confirmed PKDL cases . The former explanation is less likely to be true because PKDL is typically not self-limited in the Indian subcontinent and less likely to be cured without treatment [1]–[6] . All except one patient had persistent skin lesions after a one year in this study . Thus , the development of more sensitive tests for the diagnosis of PKDL is urgently needed and should be encouraged by national and international researchers and funding agencies . The poor treatment-seeking behavior of and treatment compliance by patients with PKDL patients present a challenge for the NKEP . Only seven of the 18 cases ( 38 . 9% ) completed treatment after referral to the hospital . The reason behind not being treated and partially treated was respectively feeling otherwise healthy and concern about loss of daily wage or loss of school day . Intensive motivation and some financial support from the NKEP may improve this situation . In conclusion , the prevalence of PKDL is high in a VL-endemic area of Bangladesh . The use of TVVs feasible to actively detect PKDL suspects , and in conjunction with PCR techniques , this holds promise as an effective strategy for the NKEP to help meet goals for the elimination of VL , if the ways for improving treatment-seeking behavior and treatment compliance are found . | PKDL is a skin disorder which usually develops in 10–20% and about 60% of patients with visceral leishmaniasis ( VL ) after treatment respectively in the Indian subcontinent and Sudan . However , cases among people without prior VL have also been reported . Except skin lesion , PKDL patients are healthy and usually do not feel sick . However , persistence of a few PKDL cases is sufficient to initiate a new epidemic of anthroponotic VL . Thus , identifying and treating people with PKDL is a key strategy for the elimination of kala-azar . Diagnosis of PKDL relies upon clinical criteria and a serological test which is not specific for PKDL . The use of the existing laboratory diagnostic tools for confirmation of PKDL among PKDL suspects is unknown . In the Indian subcontinent , PKDL is not self-limited and needs to be treated with sodium stibogluconate injections for 4–6 months . No data are available relating to treatment compliance by patients , particularly in Bangladesh . The results of the present study showed that trained village volunteers were useful for identifying PKDL suspects , and diagnostic confirmation improved with the use of PCR . However , patients' adherence to prescribed treatment was poor . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"public",
"health",
"and",
"epidemiology",
"public",
"health",
"and",
"epidemiology/screening",
"public",
"health",
"and",
"epidemiology/health",
"policy"
] | 2010 | Enhanced Case Detection and Improved Diagnosis of PKDL in a Kala-azar-Endemic Area of Bangladesh |
The regulation of filopodia plays a crucial role during neuronal development and synaptogenesis . Axonal filopodia , which are known to originate presynaptic specializations , are regulated in response to neurotrophic factors . The structural components of filopodia are actin filaments , whose dynamics and organization are controlled by ensembles of actin-binding proteins . How neurotrophic factors regulate these latter proteins remains , however , poorly defined . Here , using a combination of mouse genetic , biochemical , and cell biological assays , we show that genetic removal of Eps8 , an actin-binding and regulatory protein enriched in the growth cones and developing processes of neurons , significantly augments the number and density of vasodilator-stimulated phosphoprotein ( VASP ) -dependent axonal filopodia . The reintroduction of Eps8 wild type ( WT ) , but not an Eps8 capping-defective mutant , into primary hippocampal neurons restored axonal filopodia to WT levels . We further show that the actin barbed-end capping activity of Eps8 is inhibited by brain-derived neurotrophic factor ( BDNF ) treatment through MAPK-dependent phosphorylation of Eps8 residues S624 and T628 . Additionally , an Eps8 mutant , impaired in the MAPK target sites ( S624A/T628A ) , displays increased association to actin-rich structures , is resistant to BDNF-mediated release from microfilaments , and inhibits BDNF-induced filopodia . The opposite is observed for a phosphomimetic Eps8 ( S624E/T628E ) mutant . Thus , collectively , our data identify Eps8 as a critical capping protein in the regulation of axonal filopodia and delineate a molecular pathway by which BDNF , through MAPK-dependent phosphorylation of Eps8 , stimulates axonal filopodia formation , a process with crucial impacts on neuronal development and synapse formation .
Deciphering the molecular mechanisms by which neurite extension and synaptogenesis occur during brain development is critical to understand the ontogenesis of the nervous system . In addition , the knowledge of the mechanisms through which initial synaptic contacts are established and modified during brain development may also shed light on synaptic remodeling and plasticity occurring in the adult brain . In the last years , evidence has accumulated indicating that filopodia , which are highly motile , narrow , cylindrical extensions emerging from both axons and dendrites , play important roles at initial stages of synaptogenesis [1] , [2] . Furthermore , the growth of new filopodia leading to new synaptic contacts has been suggested as a possible mechanism underlying long-term potentiation [1] , [3] , [4] . Although axonal filopodia have been investigated less systematically , it is now well established that filopodia extending at the tips of axonal growth cones mediate neurite navigation and axonal path finding [5] , whereas filopodia emerging from the shaft of distal axonal branches play a pivotal role in synapse formation [6] . Axonal filopodia differ from both the growth cone and dendritic filopodia since , in addition to bundles of filamentous actin , they also contain clusters of synaptic vesicles that undergo exo-endocytotic recycling [7]–[9] . These packages of vesicles move bidirectionally along the axonal shaft and the filopodia axis , and rapidly coalesce at the site of contact with the appropriate target cell to generate mature presynaptic sites [6] , [8] . Thus , filopodia emerging from axons are destined to differentiate into future presynaptic sites . The core structural and dynamic components of filopodia are actin filaments , whose dynamic formation and topological organization are controlled by ensembles of actin-binding proteins . These proteins play different functional roles in regulating actin dynamics , including binding and/or sequestering of actin monomers , nucleation of actin filaments , capping or anticapping of barbed ends , and severing , bundling , and anchoring of F-actin ( filamentous actin ) [10]–[12] . In order to elongate , filopodia must be protected from capping and becoming cross-linked into bundles [13] . Formins , which promote the linear elongation of actin filaments in a processive fashion , can also effectively compete with cappers [14]–[16] and have therefore emerged as key players in the generation of filopodia [13] . Alternatively , actin-binding proteins , such as Ena/VASP family members , which bind barbed ends , and bundle and elongate actin filaments [16]–[22] , have also been implicated in the formation and elongation of filopodia along neurite shaft and growth cone [17] , [20] , [23] , [24] . Remarkably , mice lacking all three Ena/VASP paralogs ( ENA , MENA , and VASP ) have defective neuritogenesis in the cortex , leading to a block of cortical axon-fiber tract formation . This defect was shown to arise from a failure of cortical neurons to form filopodia [23] . The neuritogenesis defect observed in Ena/MENA/VASP-deficient cortical neurons could be rescued through intrinsic factors , such as mDia2 or Myosin X , or extrinsic factors , such as laminin , that induce filopodia formation [23] . Instead , no evidence has been reported on the possible role of actin-capping proteins in controlling the formation of axonal filopodia in neuronal cells . Eps8 is the prototype of a family of proteins involved in the transduction of signal from Ras to Rac , leading to actin remodeling [25] . Eps8 directly controls actin dynamics and the architecture of actin structures by capping barbed ends and cross-linking actin filaments , respectively [26] , [27] . The barbed-end capping activity of Eps8 , which resides in its conserved C-terminal effector domain , is tightly down-regulated within the context of the holoprotein . Binding of Eps8 to one of its interactors , Abi1 , relieves this inhibition [28] . Conversely , Eps8 must associate with IRSp53 ( insulin receptor tyrosine kinases substrate of 53 kDa , also known as BAIAP2 for binding partner of the brain-specific angiogenesis inhibitor 1 ) [29]–[31] to efficiently cross-link actin filaments [27] . These multiple actin regulatory roles of Eps8 in vitro are reflected by the observation that in vivo Eps8 is required for optimal actin-based motility , intestinal morphogenesis , and filopodia-like extension [26]–[28] . The role of Eps8 in neuronal cells is , however , still elusive . Here , we show that Eps8 is enriched in the growth cone and filopodia of developing hippocampal neurons , where it down-regulates axonal filopodia formation through its barbed-end capping activity . We also show that in vivo , the actin-capping activity of the Eps8/Abi1 complex is modulated by MAPK-mediated phosphorylation of the protein at the residues ser624 and thr628 , which occurs in response to external cues , such as BDNF , thereby explaining the induction of axonal filopodia by BDNF treatment .
Eps8Ls family proteins are widely and concomitantly expressed in most tissues in developing and adult mice [25] . The brain , in which only the expression of Eps8 and Eps8L2 , albeit at lower levels , was detected , represents a notable exception [32] . This prompted a more detailed analysis of the functional role of this protein in neuronal cells . In 3–4-d-old primary hippocampal cultures , we found Eps8 expressed in neuronal cell body and neurites , and prominently enriched in the axonal growth cone ( Figure 1A , arrow ) . In both rat ( Figure 1A–1C ) and mouse ( unpublished data ) hippocampal neurons , Eps8 was also detected along finger-like protrusions emerging from the axonal shaft . These processes measured 3 . 27 µm±0 . 07 standard error ( SE ) ( in line with [33] ) , were highly dynamic ( Video S1 ) , contained actin filaments ( Figure 1B and 1D , see also [7] , [9] ) and the actin-binding proteins fascin ( Figure 1C ) and VASP ( Figure 1D ) , and were characterized by the presence of synaptobrevin/VAMP2-positive vesicle clusters ( Figure 2C and 2D , see also [8] , [9] ) , thus they are bona fide axonal filopodia . Immunoblotting of subcellular fractions of fetal rat cortices further indicated that Eps8 and its binding partner Abi1 were concentrated in growth cone particles ( GCPs ) ( Figure 1E ) , a preparation that allows the investigation , using a biochemical approach , of the expression and interaction of growth cone and axonal proteins during neuronal development [34] , [35] . GCPs are enriched in pinched-off , resealed , metabolically active growth cones [34] , [35] , as witnessed by the presence of the growth cone–specific marker GAP43 and of actin [36] , [37] . Conversely , the NMDA receptor subunit NR1 , which is uniformly distributed in the neuronal plasma membrane at this developmental stage [38] , is not enriched in the GCPs , whereas the proteolytic enzyme cathepsin D , as expected , is excluded from this fraction , consistent with its prevalent localization in the neuronal soma ( Figure 1E ) . To investigate the functional role of Eps8 in neuronal cells , we established primary hippocampal cultures from eps8 WT and deficient mice , and analyzed their morphology ( Figure 1F–1I ) . Strikingly , when compared to age-matched WT , eps8−/− neurons displayed a significantly higher number of processes emerging directly from the cell body ( primary processes ) ( Figure 1G–1I ) . Closer inspection of these cultures revealed that eps8−/− neurons properly formed one single axon , positive for synaptobrevin/VAMP2 , which is sorted to the axon at early stages of neuronal development ( Figure 2A and 2B ) . Conversely , a significantly higher number of axonal filopodia was evident in eps8 null when compared to control WT neurons ( Figure 2A , 2A′ , 2B , 2B′ , and 2C , and Video S1 ) . These protrusions contained synaptic vesicle clusters ( Figure 2C and 2D ) , were enriched in F-actin ( Figure 2D and 2E ) , were immunoreactive for fascin ( Figure 2E ) , and were significantly reduced by functional interference with VASP using a dominant-negative approach [17] ( Figure S1 ) . Both the percentage of neurons displaying filopodia emerging from the axons and the density of axonal filopodia were increased by either genetic ( Figure 2F and 2G ) or RNA interference ( RNAi ) -mediated ablation of Eps8 ( Figure S2A–S1E ) . On the other hand , no effect on filopodia length was observed ( unpublished data ) . Finally , dendritic filopodia also were augmented in eps8−/− neurons at early stages of dendrite formation ( 7 d in vitro [7DIV] ) ( Figures S2F and S1G ) . Thus , Eps8 is critical for the proper extension of neuronal filopodia , and its removal in neurons increases the formation of axonal filopodia . Next , we investigated whether the increase of axonal filopodia in eps8 null neurons depends on the protein actin barbed-end capping or cross-linking activity residing in its conserved carboxy-terminal effector domain [27] , [28] . To this aim , we restored Eps8 expression in eps8−/− cultures with a lentiviral vector encoding either a GFP-tagged Eps8 WT or an Eps8 mutant , Eps8MUT1 , specifically devoid of actin capping ( Figure S3A ) , but retaining actin-binding and -bundling activity ( Figure S3B and S3C ) . The infection of eps8 null neurons with the GFP-tagged Eps8 lentiviral vector led to a significant reduction in the number of axonal filopodia to WT levels ( Figure 3A–3D ) . Thus , the filopodia phenotype is specifically caused by the lack or reduction of Eps8 . Conversely , Eps8MUT1-infected neurons were morphologically indistinguishable from GFP-infected cells ( compare Figure 3F with 3G , see quantitation in Figure 3D ) and retained an increased number of axonal filopodia . It is of note that occasionally , and especially at high levels of ectopic expression of WT Eps8 , we observed thick , sometimes club-like actin-based protrusions along both axon and dendrites ( Figure 3C and insert , Figure 3E ) , never detectable in GFP-infected neurons , which showed a morphological appearance comparable to eps8−/− neurons ( Figure 3B , 3D , and 3G ) . Collectively , these results support a critical role of the barbed-end capping activity of Eps8 in regulating the balance between different types of actin-based membrane extensions , and more specifically , in controlling the formation of axonal filopodia protrusions . Eps8 , to function as capper , requires the binding of its interactors Abi1 or Abi2 [28] . Therefore , we analyzed the ability of these latter proteins to form a complex with Eps8 in brain lysates . Both Abi1 and the brain-specific Abi2 coimmunoprecipitated with Eps8 ( Figure S4A ) , indicating the existence of active capping complexes and supporting the importance of the capping activity in the nervous system . Conversely , in epithelioid HeLa cells , no evidence of a complex between Abi1 or Abi2 and Eps8 could be found ( Figure S4A ) , likely due to the fact that either Abi2 is expressed at lower levels than in brain ( A , right panel ) and/or these proteins are sequestered by interaction with other partners . These latter findings suggest that the actin-capping activity of Eps8 is critical for filopodia in neurons , but not in HeLa cells , as previously shown [27] . Eps8 can form a complex with Abi1 , Sos-1 , and PI3K [39] able to activate Rac , which , in turn , is involved in the formation of neurites and filopodia-like processes in neurons [40] and is activated by BDNF [41] . Thus , we examined the possibility that the increase in axonal filopodia observed after genetic removal of Eps8 could be due to reduced levels of Rac GTP . To this end , we performed two sets of experiments . In the first approach , we directly tested whether Eps8 removal impaired Rac activation . Measurement of RacGTP levels by affinity-based CRIB assays revealed that equal levels of RacGTP were present in lysates of brain cortex and hippocampus derived from WT and eps8 null mice ( Figure S4B ) , indicating that in brain , Eps8 is not required for Rac activation . As additional proof , the impairment of Rac activation , which might be predicted to occur as a consequence of Eps8 removal [42] , was mimicked by exposing hippocampal neurons to the specific Rac1 inhibitor , NSC23766 [43] . We demonstrated the efficacy of this compound by showing that NSC23766 administration ( 1 ) prevented PDGF-induced actin ruffles in NIH3T3 cells ( Figure S4C ) , a well-known process involving Rac activation [44] , and ( 2 ) decreased RacGTP levels of 3DIV hippocampal cultures ( Figure S4D ) . Notably , in hippocampal neurons , a reduction rather than an increase in the number of filopodia occurred upon Rac inhibition ( Figure S4E ) , as previously shown [45] . Thus , the pathways leading to Rac activation are not the underlying molecular cause of the enhancement of axonal filopodia extension observed in eps8 null hippocampal neurons [45] . Having established a critical role for Eps8 capping activity in down-regulating the formation of axonal filopodia , we next investigated how Eps8 biological functions might be regulated in vivo . Eps8 was originally isolated as a substrate of receptor tyrosine kinases activated in response to growth factors [46] . Therefore , we focused on the neurotrophic factor BDNF , which critically controls growth and differentiation processes in the brain during development , through the activation of Trk tyrosine kinase receptors . Consistent with this , and as described previously [47] , treatment of hippocampal neurons with BDNF induced the formation of axonal actin- and fascin-rich filopodia , mimicking genetic and RNAi-mediated eps8 removal ( Figure 4A and 4B ) . Remarkably , however , the BDNF-induced increase in filopodia density was not detected in eps8−/− cultures ( Figure 4A and 4B ) . Even if we cannot exclude the possibility that BDNF has no effect on filopodia in eps8−/− neurons simply because the density of filopodia is already maximal , this observation points to a role of Eps8 in BDNF-mediated filopodia formation . BDNF induced , as expected , MAPK activation in the GCPs ( Figure 4C ) , and MAPK-dependent phosphorylation of the synaptic vesicle protein synapsin I [37] , leading to its dispersion in the distal axonal shaft ( Figure 4D ) [37] , [48] . More importantly , the formation of filopodia induced by the neurotrophin was inhibited by pretreatment of the cultures with the specific MAPK inhibitor , PD98059 ( Figure 4E ) , indicating that MAPK activation is also critical in this pathway . One mechanism to control actin dynamics in response to extracellular cues is to ensure the correct , spatially restricted localization of critical actin-binding proteins , such as Eps8 and its interactors . We thus examined the distribution of Eps8 to the actin-rich , Triton X100–insoluble fraction of GCPs in response to BDNF . Treatment with BDNF increased the amounts of Eps8 ( Figure 4F ) and , slightly , of its activator Abi1 ( unpublished data ) , recovered in the Triton X100–soluble fraction , causing a reciprocal decrease in the Triton X100–insoluble fraction ( Figure 4F ) . Conversely , VASP was mainly found in the Triton-insoluble fraction , and its distribution was unaffected by BDNF stimulation ( Figure 4F ) . Finally , Eps8 redistribution to the Triton X100–soluble fraction correlated with BDNF-induced cellular relocalization of Eps8 , which disappeared from filopodia , persisting instead along the axonal shaft , upon BDNF stimulation ( Figure 4G and quantification ) . Notably , both Eps8 redistribution to the Triton-X100-soluble fraction and cellular relocalization by BDNF were prevented by preincubation with the MAPK inhibitor , PD98059 ( Figure 4F and 4G ) . Thus , collectively , these results indicate that BDNF controls , in a MAPK-dependent manner , the dynamic cellular redistribution of Eps8 in vivo . To investigate the molecular mechanisms through which BDNF and MAPK regulate Eps8 , its posttranslational modifications in synaptosomes were analyzed by two-dimensional electrophoresis . Seven Eps8 spots of similar molecular size , but different isoelectric points ( IP , Figure 5A , CTR ) , were detected by immunoblotting with two independently raised antibodies ( affinity-purified polyclonal antibody , Figure 5A; monoclonal antibody , unpublished data ) , suggesting the existence of posttranslational modified forms of Eps8 . Competition of Eps8 monoclonal and polyclonal antibodies with the respective immunogenic polypeptide further validated the identity of the immunoreactive spots ( unpublished data ) . More importantly , treatment of synaptosomes with BDNF induced a change in the electrophoretic pattern of Eps8 with the appearance of additional spots of increased acidic IP ( spots 6–9 in Figure 5A , BDNF ) , which were significantly diminished by pretreatment with MAPK inhibitor PD98059 prior to BDNF stimulation ( Figure 5A , PD+BDNF ) . Additionally , the five most acidic spots of Eps8 could no longer be detected after dephosphorylation of synaptosomes with alkaline phosphatase ( Figure 5A , BDNF+AP ) . Similar results were obtained in preparations of GCPs ( Figure 5B ) . Of note , Eps8 shows a different electrophoretic pattern in untreated GCPs with respect to untreated synaptosomes ( Figure 5B , “CTR , ” and 5A , “CTR” ) , suggesting that the protein is subjected to different posttranslational modifications during brain development . Remarkably , however , and at variance with stimulation with other RTK [46] , we could detect no Eps8 tyrosine phosphorylation after BDNF treatment of synaptosomes ( Figure S4F ) . Thus , collectively , the above results support the notion that MAPK or a MAPK-dependent distal kinase is responsible for Eps8 phosphorylation in response to BDNF . To directly assess this possibility , we carried out an in vitro phosphorylation assay . Purified Eps8 became phosphorylated by recombinant purified MAPK to a similar extent as a bona fide MAPK kinase substrate ( myelin basic protein [MBP] ) ( Figure 5C , first two lanes , and Figure S5A ) , thus indicating that Eps8 is a direct substrate of activated MAPK . To get insights into the residues specifically phosphorylated by MAPK , a structure–function analysis was carried out using different fragments of Eps8 . Eps8 ( 586–733 ) is the minimal fragment that is phosphorylated ( Figure 5C ) . However , the lack of phosphorylation of Eps8 ( 648–821 ) restricted further the phosphoresidue-containing region of Eps8 to amino acids 586–648 . This region contains two putative consensus sites ( S624 and T628 ) for MAPK ( Prosite [49] , Scansite [50] ) . Simultaneous mutation of these sites to A ( GST-535-821-SATA ) , either in the context of the C-terminal 535–821 fragment of Eps8 ( Figure 5D ) or in the full-length protein ( Figure S5A ) , abrogated MAPK-mediated phosphorylation , unequivocally identifying the Eps8 residues targeted by MAPK . To demonstrate the biochemical and physiological relevance of these phosphosites , we generated phosphoimpaired and phosphomimetic mutants of full-length Eps8 by substituting S624 and T628 with either A ( Eps8-SATA ) or E ( Eps8-SETE ) , respectively . We initially tested the effects of these mutations using in vitro assays of actin polymerization either in the context of the isolated C-terminal , constitutive active Eps8 ( 535–821 ) fragment , or with full-length Eps8 protein , whose activity is unmasked when in complex with Abi1 [28] . We detected no difference in the affinities for barbed ends when WT and Eps8 ( 535–821 ) mutants were compared , suggesting that these phosphoresidues do not interfere with the direct binding of the isolated Eps8 capping domain to actin barbed ends ( Figure S5B and S5C ) . Of note , phosphorylation of S624 and T628 did not affect the actin filaments side-binding capacity of Eps8 ( Figure S5D and S5E ) . Thus , when Eps8 C-terminal fragments are folded into a presumably “open” conformation that allows its association to actin barbed ends and filaments , we could evidence no detectable effects of its phosphorylation . Conversely , the affinity of the Eps8-SETE:Abi1 complex for actin filament barbed ends was about 10 times lower that of the Eps8-WT:Abi1 or EPS8-SATA:Abi1 complexes ( Figure 6A–6C ) . Importantly , Eps8-SATA and -SETE bound Abi1 with similar affinities as Eps8WT ( Figure 6D ) . Thus , phosphorylation of Eps8 significantly impaired the actin-capping activity of Eps8:Abi1 complex , most likely by facilitating Abi1's ability to promote the structural changes required to activate Eps8 . Next , we tested whether these posttranslational modifications of Eps8 had any effect on Eps8 distribution , association to actin-based structures , or BDNF-mediated relocalization . Remarkably , we found that in neuronal cells , Eps8-SATA was distributed to and colocalized with discrete actin-rich structures ( Figure 7A ) , whereas Eps8-SETE showed a more diffuse distribution ( Figure 7B ) , in line with the hypothesis that the latter , phosphomimetic construct is constitutively detached from actin filaments . Additionally , and more importantly , WT Eps8 ( Figure 7C ) , but not Eps8-SATA ( Figure 7D ) , was dispersed after BDNF stimulation . This is consistent with the possibility that , being unable to be phosphorylated , Eps8-SATA does not detach from actin even after BDNF treatment . Finally , the expression of Eps8-SATA ( Figure 7D and 7G ) , but not Eps8-SETE ( Figure 7E ) , significantly inhibited the formation of BDNF-induced axonal filopodia ( Figure 7G ) , acting in a dominant fashion . Collectively , these data indicate that phosphorylation of S624 , T628 residues controls the barbed-end capping activity of the Eps8:Abi1 complex and Eps8 subcellular localization in response to BDNF . Thus , one mechanism to activate the protrusive activity by BDNF in neurons is to down-modulate the activity of capping proteins , via MAPK phosphorylation , favoring actin filament elongation and axonal filopodia extensions .
In this study , we demonstrated through genetic and cellular biochemical approaches , a crucial role of the actin barbed-end capping activity of Eps8 in down-regulating the formation of VASP-dependent neuronal filopodia . We further show that this activity is controlled by phosphorylation of residues S624 and T628 of Eps8 , which occurs upon MAPK activation in response to BDNF , thus unraveling the molecular components of a novel signaling cascade converging on a capping protein and ultimately leading to the proper formation of axonal filopodia . Filopodia are formed by cross-linked actin filaments , whose initiation is controlled by the balance between capping and anticapping activities . Anticapping activity can be executed either by formins , which promote linear elongation of actin filaments from barbed ends competing with cappers , or by Ena/VASP proteins , albeit through multiple and diverse biochemical mechanisms . Indeed , VASP family proteins may directly [17] , [18] or indirectly antagonize capping proteins [22] , capture barbed ends [51] , and cross-link actin filaments [20] , [22] . Formins and Ena/VASP family proteins act independently of each other , as demonstrated by the finding that genetic removal of Ena , MENA , and VASP abrogates filopodia , which , however , can be restored by expression of the formin mDia2 [23] . Thus , there are likely multiple and independent mechanisms that give rise to filopodia . This is reflected by the fact that different models have been proposed to account for the initiation and elongation of filopodia with respect to other protrusions , such as lamellipodia . For instance , the convergent elongation model [52] proposes that the Arp2/3-dependent dendritic nucleation of actin filaments is the starting point of actin filament arrays in both lamellipodia and filopodia . The activity of capping proteins , according to this model , is then crucial to control the subsequent architectural organization of actin in these protrusions . When the capping activity is high , newly nucleated branched filaments become rapidly capped . The balance between de novo branched nucleation and elongation by Arp2/3 and blockade of filament growth by capping proteins , both competing for filament ends , would thus result in the generation of a dendritic actin network at the leading edge of lamellipodia . In this context , it is remarkable that Eps8 , when expressed at high levels , causes the formation of flat , actin-rich protrusions along axons , which resemble lamellipodia extensions ( Figure 3 ) . Conversely , when capping activity is low , such as after RNAi-mediated interference or genetic knockdown of Eps8 in neurons , filaments nucleated by Arp2/3 [53] can grow longer , uncapped . Factors such as VASP family members or components of the filopodia tip complex may then promote the transient association of actin filaments , which can be further stabilized by other cross-linkers , such as fascin , thus permitting the formation of bundles of sufficient stiffness to overcome buckling and membrane resilience [54] . Of note , filopodia initiation , rather than elongation , is affected by Eps8 . A corollary of this model is that Eps8-dependent filopodia should be , at least in part , mediated by VASP . Consistently , we show here that the increased axonal filopodia of eps8 null hippocampal neurons are significantly reduced upon interference of VASP functions through dominant-negative approaches , suggesting functional competition between the capping activity of Eps8 and the actin-regulatory properties of VASP . This is reminiscent of a situation similar to the one suggested by Mejillano and colleagues in nonneuronal cells , where depletion of the capping protein CP caused loss of lamellipodia and the explosive formation of filopodia , a phenotype which is counteracted by Ena/VASP proteins [21] . An alternative model of filopodia initiation questioned the relevance of Arp2/3-mediated nucleation ( reviewed in [13] ) , claiming instead that linear filament elongation at the leading edge of protrusions is the source of bundles in filopodia . Within this context , formins [16] , [55]–[57] or clustered VASP proteins [22] may be the culprits in promoting filopodia initiation . Also in this case , however , filaments must be protected from cappers , which have been shown , in the case of CP , to compete , either directly or indirectly , with VASP as well as with formins for barbed-end binding . Whether Eps8 can compete with clustered VASP or formins for barbed ends remains to be assessed . The fact that filopodia induced by removal of Eps8 are dependent on VASP argues , however , that a balanced activity of these two proteins may be critical for proper axonal filopodia formation . Notably , Eps8 is a multifunctional protein , whose actin-related activities , capping and bundling , are critically controlled by its binding partners , Abi1 and IRSp53 , respectively [27] , [28] . Both capping and bundling activities need to be coordinated and regulated for the initiation and maintenance of filopodia . In neurons , in which Eps8 is tightly bound to Abi1 or Abi2 , the capping function prevails , exerting a negative modulatory role on filopodia extension . Conversely , in HeLa cells , where the majority of Eps8 is associated with IRSp53 [27] , and no complex with Abi1 or Abi2 is detected , the Eps8:IRSp53 complex bundles actin filaments , positively contributing to the formation of filopodia . Thus , specific patterns of protein expression and different Eps8-based complexes may be the basis of the different roles that this protein exerts in different experimental systems . One of the major accomplishments of our study is the demonstration that extracellular cues regulate the subcellular localization and function of Eps8 through phosphorylation . This mode of regulation defines a novel mechanism through which signaling cascades , particularly those emanating from receptor tyrosine kinases , may regulate cappers . Indeed , how the activity of capping proteins is controlled in a signaling-dependent manner has remained a critical , only partially addressed issue . Phosphoinositides , whose generation is controlled through the balance of signaling-regulated lipid kinases and phosphates , can bind and inhibit both gelsolin and CP [58] , [59] . Albeit recent evidence obtained using yeast CP questioned whether phosphatidylinositol 3 , 4-bisphosphate ( PIP2 ) can access CP when it is bound to ends [60] , thus casting doubts that this lipid may effectively free barbed ends from capping . Protein:protein interaction is another way to control the activity of capping proteins . In this respect , Carmil , an adaptor protein , was shown to bind CP with high affinity and to inhibit capping , influencing the formation of lamellipodia [61] . Conversely , Abi1 functions as an activator of the otherwise inhibited Eps8 . We show here that BDNF , in particular , controls the barbed-end capping activity of the Eps8:Abi1 complex and modulates Eps8 interaction with the cytoskeleton , strengthening the importance of spatially restricted activation of proteins in the regulation of actin dynamics . Accordingly , BDNF-mediated detachment of Eps8 from actin cytoskeleton correlates with increased filopodia formation , which can be mimicked by genetic removal of this capper . At the molecular level , this is achieved through MAPK-mediated phosphorylation of Eps8 residues S624 and T628 , which controls the protein capping activity , with a 10-fold reduction in barbed-end binding of the Eps8:Abi1 complex , following S624 and T628 phosphorylation . Since Eps8 localization depends on both binding to Abi1 [62] and to actin , this reduction in actin binding reasonably contributes to Eps8 subcellular relocalization in response to BDNF . Of note , BDNF-induced detachment of Eps8 from cytoskeletal structures is not accompanied by that of VASP , which remains enriched in the Triton-insoluble fraction , where it may more efficiently associate to filament ends , promoting their elongation . The regulation of actin-capping activity of Eps8 may have consequences for the developing and adult brain . It is now widely accepted that presynaptic specializations , which are highly dynamic and are continuously added and eliminated throughout the lifetime of a neuron , play an active role in activity-dependent synapse formation and remodeling . In the mature neocortex in vivo , for example , filopodia and short axonal branches are frequently extended and retracted to form boutons onto postsynaptic structures , or to originate en-passant boutons , which are added and eliminated along the axonal shaft ( reviewed in [63] ) . Notably , eps8 null brains , at early postnatal stages , display a significantly higher number of presynaptic boutons ( unpublished data ) that might originate from the increased axonal filopodia during neuronal development , thus suggesting a role of the protein in controlling synapse formation in vivo . The dynamic regulation of eps8 actin-capping activity may therefore play an important role in presynaptic plasticity phenomena in the developing brain and , possibly , in the adult brain .
Cytomegalovirus ( CMV ) promoter–based and elongation factor-1 ( EF1 ) -promoter–based eukaryotic expression vectors and GST bacterial expression vectors were generated by recombinant PCR . Polyclonal and monoclonal antibodies ( abs ) against Eps8 and Abi1 were previously described [42] , [64] . Monoclonal abs against synaptobrevin/VAMP2 , synaptophysin , SNAP-25 , NR1 , and synaptotagmin I were from Synaptic System . Monoclonal abs against Gap43 , actin , phosphorylated ERK1 , 2 , and beta-tubulin were purchased from Sigma . Monoclonal ab against phosphorylated JNK was from New England Biolab . Polyclonal ab against synapsin I was a kind gift from P . De Camilli ( Yale University ) , ab against riboforin was from G . Kreibich ( New York University ) , and ab against human cathepsin D was from C . Isidoro ( Università di Novara ) . Antibody against fascin was from Dako: clone 55K-2 . Antibody against VASP was from BD Transduction Laboratories or from Dr . F . Gertler . GST-VASP was a gift from Dr . C . Le Clainche ( CNRS , Gif-sur-Yvette , France ) . Mouse monoclonal anti-IRSp53 were raised against full-length histidine-tagged recombinantly purified proteins . FITC-phalloidin and Texas Red phalloidin were purchased from Sigma and Molecular Probes , respectively . The secondary abs conjugated to FITC , Rhodamine , or horse radish peroxidase were obtained from Jackson Immunoresearch Laboratories . Recombinant human BDNF was obtained from Regeneron . PD98059 is from Sigma . NSC23766 is a kind gift from Y . Zheng ( Cincinnati Children's Hospital Medical Center ) Shrimp Alkaline Phosphatase was from Roche Diagnostics . Monoclonal anti-pY , activated Erk1 , and purified MBP were purchased from Upstate . The Eps8 S624A , T628A ( SATA ) , and S624E , T628E ( SETE ) mutants were generated by PCR ( SATA 1861-TCT GCC CCA ( T→ G ) CA CCC CCT CCA ( A→G ) CG CCA GCA CCC-1893; SETE 1861-TCT GCC CCA ( TC→GA ) A CCC CCT CCA ( AC→ GA ) G CCA GCA CCC-1893 ) . The Eps8-KO mice obtained as previously described [42] were backcrossed for more than ten generations to C57BL/6 mice . For preparation of hippocampal culture , littermates derived from heterozygous crosses were used; in some experiments , WT and knockout ( KO ) mice from F2N12 homozygous colonies were also used . All experiments were performed in accordance with the guidelines established in the IFOM-IEO Campus Principles of Laboratory Animal Care ( directive 86/609/EEC ) . Primary neuronal cultures were prepared from the hippocampi of 18-d-old fetal rats as described [65] or E16-P1 WT and eps8 null mice . The dissociated cells were plated onto glass coverslips coated with poly-l-lysine at densities ranging from 10 , 000 to 20 , 000 cells/cm2 . After a few hours , the coverslips were transferred to dishes containing a monolayer of cortical glial cells . The cells were maintained in MEM ( Gibco ) without sera , supplemented with N2 ( Gibco ) , 2 mM glutamine , and 1 mg/ml BSA ( neuronal medium ) . Three different double-strand small interfering RNA ( siRNA ) oligonucleotides ( Stealth RNAi; called 1525 , 1526 , and 1158 ) were designed against mouse eps8 by using RNAi Design Services from Invitrogen . Oligonucleotide sequences: 1525: 5′-GCCATGCCTTTCAAGTCAACTCCTA-3′; 1526: 5′-CCATGCCTTTCAAGTCAACTCCTAA-3′; 1158: 5′-GACAAAAGACACAGTTGATTTCTTAA-3′ . Stealth RNAi negative control with medium content of CG ( Invitrogen ) whose sequence is not homologous to any vertebrate sequence/transcriptome was used as control . WT mouse hippocampal neurons were transfected with Lipofectamine 2000 ( Invitrogen ) and with each oligonucleotide sequence at the dose of 200 nM the day of plating , fixed at 3DIV , and stained with anti-VAMP2 abs and phalloidin to evaluate filopodia density . GFP alone or GFP-eps8 WT or eps8MUT1 were subcloned in a pRRLsin . PPT . CMV lentiviral vector ( kind gift of L . Naldini , San Raffaele , Milano ) . The 293T human embryonic kidney cells were cotransfected with the lentiviral vector and the packaging plasmids ( pMDL , pRev , and pVSVG ) , providing the required trans-acting factors , namely , Gag-Pol , Rev , and the envelope protein VSVG , respectively , and viruses were obtained as described [66] . Primary cultures established from eps8−/− mice were infected 1 d after plating with GFP/GFP-Eps8/GFP-Eps8-MUT1 lentiviruses . Medium was changed after 6 h of incubation , and cells were fixed 48 h after viral infection . The cultures were fixed and stained as previously described [38] . The images were acquired using a BioRad MRC-1024 confocal microscope equipped with LaserSharp 3 . 2 software . Acquired images were analyzed using Metamorph Imaging Series 6 . 1 software ( Universal Imaging ) . The number of filopodia and the length of axonal shaft were then computed in order to obtain the filopodia density ( number of filopodia/micrometer ) . For each neuron , an axon length ranging between 30 and 80 µm was analyzed . Alternatively , the percentage of neurons with filopodia was evaluated . In this latter case , neurons with more than 0 . 04 filopodia/µm were considered as filopodia-bearing neurons . For each experimental condition , both types of analyses gave similar results . Statistical analysis was performed using SigmaStat 2 . 0 software ( Jandel Scientific ) . After testing whether data were normally distributed or not , the appropriate statistical test has been used . Detailed information is reported in the figure legends . Statistical significance is indicated in graphs as follows: a single asterisk ( * ) indicates p<0 . 05; double asterisks ( ** ) , p<0 . 001 . | Neurons communicate with each other via specialized cell–cell junctions called synapses . The proper formation of synapses ( “synaptogenesis” ) is crucial to the development of the nervous system , but the molecular pathways that regulate this process are not fully understood . External cues , such as brain-derived neurotrophic factor ( BDNF ) , trigger synaptogenesis by promoting the formation of axonal filopodia , thin extensions projecting outward from a growing axon . Filopodia are formed by elongation of actin filaments , a process that is regulated by a complex set of actin-binding proteins . Here , we reveal a novel molecular circuit underlying BDNF-stimulated filopodia formation through the regulated inhibition of actin-capping factor activity . We show that the actin-capping protein Eps8 down-regulates axonal filopodia formation in neurons in the absence of neurotrophic factors . In contrast , in the presence of BDNF , the kinase MAPK becomes activated and phosphorylates Eps8 , leading to inhibition of its actin-capping function and stimulation of filopodia formation . Our study , therefore , identifies actin-capping factor inhibition as a critical step in axonal filopodia formation and likely in new synapse formation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology/cell",
"signaling",
"cell",
"biology/neuronal",
"signaling",
"mechanisms",
"neuroscience/neuronal",
"signaling",
"mechanisms",
"neuroscience/neuronal",
"and",
"glial",
"cell",
"biology",
"cell",
"biology/cytoskeleton"
] | 2009 | Eps8 Regulates Axonal Filopodia in Hippocampal Neurons in Response to Brain-Derived Neurotrophic Factor (BDNF) |
The role of adult hippocampal neurogenesis in spatial learning remains a matter of debate . Here , we show that spatial learning modifies neurogenesis by inducing a cascade of events that resembles the selective stabilization process characterizing development . Learning promotes survival of relatively mature neurons , apoptosis of more immature cells , and finally , proliferation of neural precursors . These are three interrelated events mediating learning . Thus , blocking apoptosis impairs memory and inhibits learning-induced cell survival and cell proliferation . In conclusion , during learning , similar to the selective stabilization process , neuronal networks are sculpted by a tightly regulated selection and suppression of different populations of newly born neurons .
It was classically assumed that once the development of the central nervous system ended , “everything can die , nothing can regenerate and be renewed” [1] . This dogma , restricting neurogenesis to a developmental phenomenon has , however , been challenged by the discovery that new neurons are created in specific regions of the adult mammalian brain [2 , 3] . The dentate gyrus ( DG ) of the hippocampal formation is one of the few structures where adult neurogenesis occurs in mammals [4] , and it has been estimated that several thousand new cells are generated daily [5 , 6] . Neurogenesis in the DG is a complex , multistep process that starts with the proliferation of neural precursors residing in the dentate subgranular layer . Within a few days following their birth , at least 50% of the daughter cells die [7] . The adult-born cells that survive this initial period of cell death differentiate , for the most part into granule neurons , and survive for several months within the DG [8] . These new mature neurons receive synaptic inputs , extend axons along the mossy fiber tract , and exhibit electrophysiological properties very similar to those of mature dentate granule neurons [9–12] . The involvement of the hippocampal formation in memory has long been recognized [13] , and increasing evidence suggests that the production of adult-born neurons may contribute to memory processes . First , the rate of neurogenesis is positively correlated to hippocampal-mediated learning abilities [14] . Second , conditions that increase memory performance enhance neurogenesis , whereas conditions that decrease neurogenesis impair learning [15–18] . Third , spatial learning has been shown to increase both the survival of newborn neurons [19] and cell proliferation [16 , 20] . Remarkably , spatial learning in a water maze was also linked to a decrease in the number of newborn neurons in the DG [20 , 21] . Even more surprising , the decline in neurogenesis is correlated with spatial abilities , i . e . , rats with the lowest number of newly born cells have the best memory performances , indicating that learning , and not training , decreased the number of adult-born cells [20] . These complex results provide a puzzling picture in which increases and decreases in the number of newborn neurons are both correlated with learning . In order to solve this discrepancy , we hypothesized that spatial learning is accompanied by events that are similar to the selective stabilization process observed during development . Indeed , during brain development , many more neurons are produced than are actually needed , and the active and selective removal of the cells that have not yet established appropriate synaptic connections allows for the sculpting of the relevant and functional neural networks . In this study , we found that learning has three effects on neurogenesis . Learning promotes the survival of relatively mature neurons , induces the death of more immature neurons , and finally , stimulates cell proliferation . Cell death seems to be a pivotal event in this cascade because blocking learning-induced apoptosis inhibits the other two cellular events and impairs memory abilities . These results indicate that spatial learning involves a cascade of events similar to the selective stabilization process by which neuronal networks are sculpted by adding and removing specific population of cells as a function of their maturity and functional relevance .
In these experiments , rats were trained in a water maze , one of the most commonly used tests for spatial learning in rodents . In this task , the animals learn across daily sessions to find a hidden escape platform using the distal cues present in the surrounding environment . In a first experiment ( Figure 1A ) , animals trained in the water maze ( Learning group [L] ) were compared to two other groups . The first group was composed of animals that were transferred to the testing room at the same time and with the same procedures as the learning group except that they were not exposed to the water maze ( Control group [C] ) . The second group ( Yoked group [Y] ) contained animals that were submitted to the same procedures as the control group except that they were also exposed to the water maze for a time period equivalent to that of the Learning group but in the absence of the escape platform . Control and Yoked animals allowed us to control for the putative influence of stress and physical exercise ( batch 1 , see Figure 2 and Table S1 ) . At the end of the training , cell death was assessed in these three experimental groups by using two specific apoptotic markers , the active form of caspase 3 and the caspase 3–cleaved fragment of actin called fractin ( Figure 3A–3C ) [22] . Pyknosis and karyorrhexis were also used as morphological criteria to evaluate apoptotic cell death ( Figure 3A and 3B ) . It was found that spatial learning induced apoptotic cell death in the DG . Thus , although Control and Yoked animals did not differ in the number of apoptotic cells , a higher number was found in the Learning group ( Figure 1B , F2 , 16 = 4 . 70 , p < 0 . 05; Figure 1C , F2 , 16 = 11 . 01 , p < 0 . 001; and Figure 1D , F2 , 16 = 12 . 90 , p < 0 . 001 ) . Learning-induced apoptosis was associated with a decrease in the number of newly born cells that were produced during the early phase of training ( Figures 1E and 3D , F2 , 16 =7 . 38 , p ≤ 0 . 01 ) . Newly born cells were identified by injecting animals during the first 4 d of training with 5-bromo-2′-deoxyuridine ( BrdU ) . Consistent with these findings , learning-induced cell death occurred prominently in the subgranular layer , which is the part of the DG where neuronal precursors and immature neurons reside ( Table S2 ) . Furthermore , learning-induced cell death was specific to the DG; no changes were observed in the CA3 and CA1 regions of the Ammon's Horn of the hippocampus ( Table S3 ) . In a second experiment , we determined whether learning a task that does not require the hippocampus also increases cell death [23] . To this end , a fourth group of animals trained to find a visible platform ( Visible platform group [VP] ) were compared to the Control , Yoked , and Learning groups ( batch 2 ) . We found that cell death was specific to spatial learning . Thus , we found an increase in apoptotic cell death only in the Learning group , whereas the Visible platform , Control , and Yoked groups did not differ ( Control group = 65 . 00 ± 11 . 73 cells , Yoked group = 54 . 00 ± 8 . 12 cells , Visible platform group = 50 . 00 ± 5 . 92 cells , Learning group = 97 . 00 ± 9 . 03 cells , F3 , 16=5 . 67 , p < 0 . 001; Control group = Yoked group = Visible platform group < Learning group , all comparisons at least p ≤ 0 . 01 ) . In a third experiment , we then examined whether a particular phase of the learning process was responsible for the induction of apoptosis ( batches 3–6 ) . Indeed , during training in the water maze , two phases can be distinguished: an early phase during which performance improves rapidly , and a late phase during which performance stabilizes , reaching an asymptotic level [20] . In order to distinguish the effects of these two phases of learning on apoptosis , independent Control and Learning groups were sacrificed at different days of training ( Figure 4A , 4E , 4I , and 4M ) starting at day 3 . Since in the two previous experiments the Control and Yoked animals did not differ , only the Control group was used for this and subsequent experiments . It was found that only the asymptotic phase of learning induced cell death . Thus , an increase in cell death was seen starting from the fourth day of training ( Figure 4F , 4G , 4J , 4K , 4N , and 4O ) , which corresponds to the beginning of the asymptotic phase , but not at day 3 , which corresponds to the early phase ( Figure 4B and 4C ) . In order to further characterize the relationship between learning and changes in cell death , we correlated performance in the water maze with the number of apoptotic cells ( Figure S1 ) . A positive relationship was found; the animals that had the best learning of the task ( lower latency to reach the platform ) also had the highest number of dying cells . Finally in this experiment , we studied the effects of the different phases of learning on cell proliferation and on the number of newborn cells produced during the early phase of learning . Cell proliferation was analyzed by studying the expression of Ki67 [4] , a nuclear protein expressed for the entire duration of the cell cycle ( Figures 3E , 4D , 4H , 4L , and 4P ) . We found that the asymptotic phase of learning stimulates the production of new cells . However , this phenomenon appears 1 d after the start of cell death , i . e . , at the fifth day of training instead of the fourth ( Figure 4L , t21 = 7 . 02 , p < 0 . 001; and Figure 4P , t19 = 4 . 06 , p < 0 . 001 ) . In contrast , in this experiment learning had no effect on the number of newborn cells labeled with BrdU during the first 3 d of training ( Table S4 ) . The results of the previous experiments seem contradictory concerning the decrease in the number of newborn neurons . Thus , this phenomenon was observed during the first experiment but not the third . One possible explanation for this discrepancy is the difference in the age of the newly born neurons that were studied in the two experiments . In the first experiment ( batch 1 ) , in which a decrease in newborn neurons was found , BrdU-labeled cells were between 5 and 8 d old at the time of the sacrifice ( Figure 2 ) . In contrast , in the third experiment ( batches 3–6 ) , in which learning did not decrease the numbers of newly born neurons , BrdU-immunoreactive ( IR ) cells at the time of the sacrifice were less than 5 d old ( Figure 2 ) . These data suggest that learning promotes the death of cells that have reached a certain level of maturation and are older than 5 d . To test this hypothesis , additional groups of animals were trained in the water maze under conditions similar to those of experiment 3 , but were injected with BrdU either 3 or 4 d before the start of the behavioral training ( batches 7a and 7b , Figure 5A ) . In this way , BrdU-labeled cells would be either 7 or 8 d old at the end of the 5 d of training ( Figure 2 ) . We found that for both ages of BrdU-labeled cells , learning induced a decrease in the number of newly born cells ( Figure 5B , group × age interaction F1 , 20 = 0 . 01 , p = 0 . 90; group effect: F1 , 20 = 12 . 57 , p < 0 . 01 ) and , as expected , an increase in cell death ( Figure 5C , group × age interaction F1 , 20 = 2 . 82 , p = 0 . 11; group effect: F1 , 20 = 27 . 34 , p < 0 . 001; and Figure 5D , group × age interaction F1 , 20 = 1 . 83 , p = 0 . 19; group effect: F1 , 20 = 35 . 78 , p < 0 . 001 ) . To confirm that this decrease in the number of BrdU-labeled cells was due to learning-induced death of newly born neurons , we first evaluated the number of BrdU-labeled cells that expressed morphological manifestations of apoptosis . In comparison to healthy BrdU-IR cells with large nuclei ( ≈10 μm , Figure 3H ) , apoptotic newly born BrdU-labeled cells exhibited a small round shape and condensed , shrunken nuclei ( ≈3–4 μm , Figure 3I–3K ) . It was found that learning increased the number of BrdU-IR pyknotic cells ( Figure 5E , t12 = −3 . 83 , p ≤ 0 . 01 ) . We then studied the percentage of BrdU-labeled cells expressing the immature neuronal marker doublecortin ( Dcx ) . It was found that this population of cells was also decreased , most probably because the dying BrdU-labeled cells lost the Dcx labeling ( Figures 3G and 5F , t8 = −2 . 30 , p ≤ 0 . 05 ) . In addition , a quantitative evaluation based on the extrapolated total number of BrdU-labeled neurons showed that learning induces a decrease of 34% in this population of cells , which corresponds to a loss of approximately 1 , 600 newborn neurons ( Figure 5G , t8 = −3 . 31 , p ≤ 0 . 01 ) . We then examined whether learning-induced cell death plays a role in the stabilization of performances in the water maze . To address this issue , at the end of the fourth training session , i . e . , when learning-induced cell death becomes apparent , animals were infused in the lateral ventricles with either the pan-caspase inhibitor z-Val-Ala-Asp-fluoromethylketone ( zVAD ) or with vehicle [24 , 25] ( batch 8 ) . These treatments were repeated for two subsequent training days . Before the beginning of the treatment , the vehicle- and zVAD-infused animals did not differ in their latency to find the hidden platform ( unpublished data; F9 , 198 = 1 . 16 , p = 0 . 32 ) . However , after the infusions , zVAD-infused animals were significantly impaired , whereas vehicle-infused rats continued to stabilize their performances ( Figure 6A , F1 , 22 = 15 . 64 , p < 0 . 001 ) . This was particularly obvious in the first trial on day 5 , during which zVAD-infused animals exhibited the largest impairment , suggesting a deficit in retrieving what was learned on the preceding day . In contrast , zVAD-treated animals were able to learn , within the session , the position of the hidden platform . These data strongly suggest that inhibition of apoptosis disrupts the memory trace and not learning per se . To further evaluate the strength of the memory trace after zVAD treatment , a probe test was performed on the seventh day of training . The probe test consisted of exposing animals to the water maze in the absence of the escape platform and recording the time spent in the quadrant of the water maze that contained the platform during training ( target quadrant ) . zVAD-infused animals spent less time than vehicle rats in the target quadrant ( Figure 6B and 6C , t22 = 2 . 01 , p ≤ 0 . 05 ) . In addition , several indices used to measure the efficiency of the swim paths to reach the goal location were also impaired in zVAD-infused animals ( Table S5 ) . These results confirm that the inhibition of apoptosis when animals begin to master the task leads to an impairment of the memory for the platform location . We found that zVAD treatment efficiently prevented apoptosis ( Figure 6D , F3 , 39 = 9 . 61 , p < 0 . 001; and Figure 6E , F3 , 39 = 21 . 23 , p < 0 . 001 ) . However , we also found that zVAD treatment greatly reduced the increase in cell proliferation induced by the late phase of learning . In this experiment , cell proliferation was measured by Ki67 staining ( Figure 6F , F3 , 39 = 6 . 72 , p < 0 . 001 ) , and also by a second marker of cell genesis , the phosphorylated histone H3 [4] ( HH3 , Figures 3F and 6G , F3 , 39 = 6 . 89 , p < 0 . 001 ) . These results indicate that the increase in the production of new cells observed during the asymptotic phase of learning probably constitutes a compensatory phenomenon triggered by learning-induced apoptosis . Thus , as shown in experiment 3 ( Figure 4 ) , learning-induced cell proliferation follows learning-induced cell death and , as shown in the present experiment , this phenomenon is decreased by blocking cell death . In order to test for the specificity of the effects of zVAD , we performed several complementary measures and experiments . Thus , after the probe test , animals were tested for their ability to find a visible platform ( cued test ) , which allowed for a measure of visuomotor processes ( Table S5 ) . In this case , zVAD treatment had no measurable effects on behavioral performance . We also evaluated whether zVAD would modify the survival of newly born cells younger than 5 d , because these neurons are normally untouched by learning . This population of cells was labeled by injecting BrdU on day 1 to day 3 of training ( Figure 2 ) . Again , zVAD treatment had no measurable effects ( t22 = 0 . 87 , p = 0 . 39 ) . Then , in a subsequent experiment ( batch 9 ) , we infused zVAD during the first 3 d of training , a period during which learning has no influence on cell death ( Figure 4B and 4C ) . Consequently , if the effects of zVAD are mediated by a blockade of learning-induced cell death , this schedule of treatment should have no behavioral effects . Indeed , we found that under these conditions , zVAD infusions did not impair spatial memory ( Figure S2 ) . Finally , we performed physiological recordings in the Ammon's horn , a part of the hippocampus in which learning does not induce cell death ( Table S3 ) and which consequently should not respond to zVAD treatment ( batch 10 ) . Field recordings confirmed that zVAD did not alter excitatory synaptic transmission within this area ( Figure S3 ) . In conclusion , taken together , all these control experiments confirm that the disruption of spatial learning was due to a specific inhibition of apoptosis by zVAD . It has previously been shown that training in a water maze increases the survival of newborn neurons that were produced 1 wk before the start of the training [19 , 26] . It seems then that learning can induce increases in both the survival and death of newborn neurons . For this reason , in a final experiment , we studied the relationships between these two phenomena ( batch 11 ) . This experiment was performed by injecting the same animals with two thymidine analogs , 5-iodo-2′-deoxyuridine ( IdU ) and 5-chloro-2′-deoxyuridine ( CldU ) [27] ( Figure S4 ) . IdU and CldU were injected at different times in order to analyze in the same subject the fate of newborn cells of different ages ( Figure 2 ) . IdU was injected 7 d before training in order to label the newborn cells for which survival should be increased by learning . CldU was injected 3 d before the start of the training in order to label newly born cells that should die as a consequence of learning . Animals in these experiments were also infused either with vehicle or zVAD at the end of the fourth through sixth days of training . As found in the previous experiment , vehicle- and zVAD-infused animals did not differ in their latency to find the hidden platform during the first 4 d of training ( F9 , 135 = 0 . 52 , p > 0 . 05 ) . Similarly , an impairment was observed in zVAD-infused animals during the last two training days ( Figure 7A , F1 , 15 = 20 . 43 , p < 0 . 001 ) . In vehicle-treated rats , learning promoted the survival of IdU-labeled cells generated 1 wk before exposure to the task , and this effect was blocked by zVAD infusion ( Figure 7B , F3 , 24 = 6 . 65 , p < 0 . 01 ) . This prosurvival effect of learning on IdU-labeled cells was associated with a decrease in the number of IdU-IR pyknotic cells ( Figure 7C , F3 , 24 = 5 . 58 , p < 0 . 01 ) . Furthermore , as previously found here ( Figure 5B ) , learning decreased the survival of CldU-labeled cells that were born 3 d before the start of the training , an effect suppressed by zVAD infusion ( Figure 7D , F3 , 24 = 4 . 50 , p ≤ 0 . 01 ) . As expected , learning increased the number of CldU-IR pyknotic cells , and this effect was blocked by zVAD infusion ( Figure 7E , F3 , 24 = 40 . 55 , p ≤ 0 . 001 ) . These data show that learning-induced increases in survival and apoptosis of newborn cells are interrelated processes . Thus , blocking learning-induced apoptosis also blocks the increased survival of older neurons .
The results of the experiments reported here show that spatial learning promotes the survival of adult-born neurons that are relatively more mature , induces the death of cells that are more immature , and finally , stimulates proliferation of precursors . Blocking learning-induced cell death has shown an interdependency of these events and their involvement in learning . Thus , blocking learning-induced apoptosis inhibits cell survival and cell proliferation , and impairs memory abilities . These results indicate that spatial learning could involve a cascade of events similar to the selective stabilization process by which neuronal networks are sculpted by adding and removing specific populations of cells as a function of their maturity and functional relevance . Learning-induced apoptosis is a very specific phenomenon . It is selectively induced by a specific phase of spatial learning , the late phase , during which performances stabilize . In contrast , apoptosis of newborn neurons does not seem to be influenced by stress and/or physical activity: ( 1 ) animals were habituated to the pool before training in order to diminish its stressful component ( but see also [28] ) ; ( 2 ) learning did not induce cell death during the first 3 d of training , during which physical activity is at its highest; and ( 3 ) no modification in apoptosis was observed in Yoked animals exposed to the pool for 6 or 8 d . Furthermore , apoptosis was not influenced by hippocampus-independent learning in the water maze , such as cued learning of the platform position . Finally , the learning-induced increase in cell death is correlated with spatial abilities , i . e . , rats with the highest number of dying cells have the best memory performances . This observation confirms that spatial learning , and not training , physical activity , or stress , increases apoptosis . Spatial learning-induced apoptosis targets a population of young newborn neurons that are within a specific time window . Indeed , learning did not promote the death of newly born cells that were younger than 5 d or older than 13 d at the time of the sacrifice . In contrast , it promotes the death of cells that are 7 and 9 d old at the time of the sacrifice . These results are consistent with recent studies showing that the selective regulation of survival/death by input activity or the response to experience-specific modifications of adult-born neurons occur at a critical period during an immature stage [29 , 30] . We also showed that the administration of the antiapoptotic agent zVAD induces deficits in spatial memory . This is consistent with an earlier observation showing that administration of anti-caspases impaired spatial memory [31] . Here , we show that spatial memory impairment after caspase inhibition is due to the blockade of learning-induced neuronal apoptosis . The implication of apoptosis in learning seems quite specific . Thus , when the caspase inhibitor zVAD was infused during learning , but outside the window of learning-induced apoptosis , no effects on spatial learning were observed . In addition , zVAD injections per se did not alter the neurophysiological responsiveness of the hippocampus in a non-neurogenic area . Altogether , these data show that it is the learning-induced apoptosis in the DG that is involved in spatial memory . The relationship described here between learning-induced increases in survival , apoptosis , and proliferation of newborn cells provides a three-step picture of the relationship between neurogenesis and spatial learning ( Figure 8 ) . First , acquisition of the task induces an increase in the survival of newborn neurons generated 1 wk before the task and that consequently have reached an intermediate level of maturity . Second , once the task starts to be mastered , learning induces apoptosis of newborn neurons that are a few days younger than those for which survival has been increased . Third , learning-induced apoptosis is followed by an increase in cell proliferation that provides the hippocampus with a new pool of young neurons [16 , 20] . This homeostatic regulation of neurogenesis by learning is consistent with the selective stabilization theory according to which regressive events will stabilize a particular set of contacts among many others , thereby sculpting the precise circuits that are crucial for a given function [32] . It has been estimated that during development , after an initial proliferating phase during which a large number of newborn neurons are produced , at least half of the initial neuronal population is eliminated by apoptosis [33] . This neuronal elimination serves several functions , among which is the regulation of target innervation . Indeed , neural function depends upon a precise quantitative relationship between neurons: each axon innervates an appropriate number of target cells and each target cell is innervated by an appropriate number of axons . The decision for survival or death during development is governed by afferences and/or efferences [34 , 35] . In the case of hippocampal adult-born neurons , it might be hypothesized that those cells that are successfully connected , both in terms of efferent output and afferent input , are the ones that can be rescued by the stimuli generated in the course of learning . In favor of this hypothesis , it has been shown that enhanced synaptic activity enhances cell survival [36] . In contrast , apoptosis could constitute a trimming mechanism that suppresses more-immature neurons that have not been selected by learning . Their suppression could favor the integration of older cells that have been stabilized by activity-dependent stimuli generated in the course of learning . These regressive events could also , by clearing the network of nonspecific noise due to superfluous new neurons , enhance the signal-to-noise ratio . Supporting this idea , an improvement in the signal-to-noise ratio of motor cortex cells during motor skill learning has been linked to a practice-related improvement in behavioral performance [37] . The precise mechanisms by which learning promotes the survival or apoptosis of immature newborn neurons are currently unknown . However , analysis of the developmental pattern of newborn neurons provides a certain number of putative explanations . Newborn neurons follow a precise maturation of neuronal connectivity and function that requires about 1 mo . They extend their dendritic tree at variable times after mitosis , and by 3 wk , their dendritic arborization resembles that of mature neurons [12 , 38] . In addition , as soon as 10 d after birth , newly born cells extend axons into the CA3 subfield of the hippocampus [9 , 12] . After the first week of maturation , they also receive depolarizing GABA inputs [39–44] . Toward the end of the second week , GABA inputs become progressively hyperpolarizing , and the adult-born neurons begin receiving functional glutamatergic depolarizing afferents [11 , 29 , 40 , 41] , a process that occurs in parallel with the formation of dendritic spines . On the basis of this developmental pattern , it seems likely that newborn neurons that are younger than 5 d are not influenced by learning because they lack afferent inputs and have not yet reached projection territories . Neurons that are in the window during which learning induces apoptosis should have received functional depolarizing GABA inputs , although their dendritic tree would still be poorly developed , and these newborn cells should not have reached their target area . Thus , in response to learning-driven depolarization , this imbalance between input and output activity may impede the survival of these cells and lead to their death . Older neurons that survive as a consequence of learning have a more developed dendritic tree that receives depolarizing GABA inputs and starts to have some glutamatergic ones . Furthermore , these newborn neurons have also reached the CA3 subfield . It is then likely that these newborn neurons that have reached a higher stage of maturation , with balanced input/output connections , can benefit from the pro-differentiating effects of the activation by GABA and glutamatergic inputs by learning [45] . Whether or not the newly born neurons whose survival is increased by learning participate in the memory process remains an open question . Although newborn neurons need several weeks before reaching full functional maturation ( for review see [45] ) , they may participate in the processing of memory at immature stages due to their high plasticity level [46 , 47] . These peculiar properties may explain why immature neurons are responsive to life experiences within a critical time period [30] . Surviving newly born neurons having similar birthdates may induce the formation of functional neuronal assemblies in the CA3 subfield , and the resulting new circuits may store memory traces [48] . Alternatively , addition of these new circuits could encode the time of new memories [49] . However , recent studies have shown that although spatial behaviors preferentially activated new neurons in the dentate gyrus [50 , 51] , this recruitment did not occur until they were at least 4 wk old [50] . Thus , if the neurons whose survival is increased by learning are not recruited by the ongoing behavior , they may support a subsequent learning experience . Additional investigations are required to determine whether adult-born neurons exert a functional role in memory formation before or after reaching complete maturity Our observations also show that spatial learning is not onlybased upon the addition of new neurons or synaptic connections , but also upon regressive events that culminate in the removal of neurons from the cellular network of the adult central nervous system . An interplay between the addition and removal of adult-born neurons as a mechanism that sustains learned behavior has already been reported for adult songbirds [52 , 53] . Interestingly , our results show that relationships between learning , neurogenesis , and apoptosis are quite different in mammals and in birds . In the adult male canary , for example , neurogenesis is triggered by a wave of apoptosis of adult neurons within the higher vocal center [54] . The current interpretation of these processes is that the death of older neurons and their substitution by new ones allows the canaries to forget the song repertoire learned the previous year and replace it with a new one [52] . In mammals , during the encoding of new information , it is the apoptosis of younger neurons that facilitates the survival of older ones . As a consequence , whereas apoptosis in birds subserves the substitution of older learning for new , in mammals , apoptosis seems to allow the efficient adding up of new information . In conclusion , our results show that spatial learning involves a mechanism very similar to the selective stabilization process observed during brain development , in which the production of new neurons is followed by an active selection of some and removal of others . As a consequence , spatial learning is not only based upon additive processes , ranging from synaptic strengthening to the formation of new synapses and new neurons , but also upon regressive phenomena , such as neuronal apoptosis . This epigenetic specification of networks by removal of neurons in the adult brain provides evidence of an additional mechanism contributing to the establishment of memory formation in mammals .
Three-month-old male Sprague-Dawley rats were tested in a water maze according to a previously described method [14] . Briefly , animals were tested 2 wk following their arrival . The apparatus consisted of a circular swimming pool built of white plastic ( 180-cm diameter , 60-cm height ) filled with water ( 20 ± 1 °C ) that has been made opaque by the addition of a nontoxic white cosmetic adjuvant . Before the start of the training , the animals were habituated to the pool for 2 d for 1 min/d . During training , the Learning group ( L ) was composed of animals that were required to locate the submerged platform , hidden 1 . 5 cm under the water in a fixed location , using the spatial cues available within the testing room . They were all tested for four trials per day ( 90 s with an intertrial interval of 30 s and beginning from three different start points that varied randomly each day ) . If an animal did not find the platform , it was set on it at the end of the trial . The time to reach the platform ( latency in seconds ) was collected using a video camera fixed to the ceiling of the room and connected to a computerized tracking system ( Videotrack; Viewpoint , http://www . viewpoint . fr/en_EU/ ) located in an adjacent room that received the individual home cages of rats during testing . For the probe test ( 60 s ) , performances were assessed by the time spent in the target quadrant where the platform was previously located . For the cued test ( 90 s ) , performances were assessed by the latency to reach the visible platform located in a different quadrant than the one used for the nonvisible platform . In the first experiment ( Table S1 ) , two control groups were used: a Control group ( C ) consisting of animals that were transferred to the testing room at the same time and with the same procedures as the learning group but that were not exposed to the water maze , and a Yoked group ( Y ) , a control for the stress and motor activity associated with the water-maze training , composed of rats that were placed into the pool without the platform and were paired for the duration of the trial with the Learning animals . In the second experiment , three control groups were used: a Control group , a Yoked group , and an additional group of rats that were trained to find a visible platform ( VP ) in a fixed location . Animals in this group were all tested for four trials per day ( 90 s with an intertrial interval of 30 s and beginning from three different start points that varied randomly each day ) . Because in the first experiment the Yoked and Control groups did not differ for cell genesis or cell death , and because in the second experiment the Visual Platform , Yoked , and Control groups also did not differ , only the Control group was used for subsequent experiments . All experiments were performed in accordance with the European Union ( 86/609/EEC ) and the French National Committee ( 87/848 ) recommendations . Detailed analysis of swim paths made to reach the platform location during the probe test was performed using the Wintrack software ( see Protocol S1 ) . Guide cannulae were implanted , according to a previously described method [55] , above the rostral ventricle in order not to cause lesions in the hippocampus . Two weeks later , 6 μl ( per infusion site ) of vehicle ( Ringer's solution with 1% DMSO ) or of zVAD-fmk ( at a concentration of 1 μg of zVAD/μl of vehicle; Calbiochem , http://www . emdbiosciences . com/html/CBC/home . html ) [25] solutions were infused at a constant rate ( 3 μl/min ) in naive or in trained animals immediately after the last trial of the 4th–6th days ( batches 8 and 11 ) or of the 1st–3rd days ( batch 9 ) of training . Newly born cells were labeled by the incorporation of synthetic thymidine analog ( XdU [where X represents Br , Cl , or I]; Table S1 ) . Rats ( batches 1 , 3–8 , and 11 ) were injected with BrdU ( intraperitoneal ) . The Learning groups received one daily BrdU injection 30 min before the first trials or a single BrdU injection 3 or 4 d before the onset of training . Rats of the 11th experiment received a single injection of IdU and of CldU [27] , respectively , 7 and 3 d before the onset of training , both at equimolar doses of 50 mg BrdU/kg . The control groups were injected with XdU within the same period . Animals were sacrificed 1 d ( batches 1 , 2 , and 11 ) or 3 h after daily training session ( batches 3–7 ) , or after the probe test ( batches 8 and 9 ) . Free-floating sections ( 50 μm ) were processed in a standard immunohistochemical procedure in order to visualized BrdU ( 1/200; Dako , http://www . dako . com ) , IdU ( 1/1 , 000 , BD PharMingen #347580; BD Biosciences , http://www . bdbiosciences . com ) , CldU ( 1/1 , 000; Accurate Chemical and Scientific Corporation , http://www . accuratechemical . com ) , Ki67 ( 1/200 , Novocastra; Vision BioSystems , http://www . vision-bio . com ) , fractin ( 1/5 , 000 , BD PharMingen #551527; BD Biosciences ) , HH3 ( 1/2 , 000 , Cell Signaling #06–570; Upstate Biotechnology , http://www . upstate . com ) , and activated-caspase-3 ( 1/10 , 000 , BD PharMingen #551150; BD Biosciences ) [14 , 22 , 56] . Sections were counterstained with thionine in order to visualize pyknotic cells , characterized by a condensed nucleus of smaller size , and karyorrhexic cells displaying chromatin clumps . The number of immunoreactive ( IR ) cells throughout the entire granule and subgranular layers of the DG were estimated using the optical fractionator method [14] . To examine the phenotype of BrdU-IR cells , one in ten sections were incubated with BrdU antibodies ( 1/500; Accurate ) , which were revealed using a CY3–anti-rat antibody ( 1/1 , 000; Jackson Immunoresearch , http://www . jacksonimmuno . com ) . Sections were then incubated with anti-DCX antibodies ( 1/1 , 000; Santa Cruz Biotechnology , http://www . scbt . com ) , which were visualized with an Alexa-488 anti-goat IgG ( 1/1 , 000; Jackson ) . The percentage of BrdU-labeled cells expressing Dcx was determined throughout the DG using a confocal microscope with helium–neon and argon lasers ( DMR TCSSP2AOBS; Leica , http://www . leica-microsystems . com ) . To estimate the total number of BrdU-labeled neurons , the percent of BrdU-IR cells co-labeled with Dcx was multiplied by the total number of BrdU-labeled cells . Hippocampal slices ( 500 μm ) were prepared as described previously [57] from vehicle- and zVAD -infused rats by an investigator blind to the treatments . Slices were submerged in an oxygenated artificial cerebrospinal fluid ( ACSF ) comprising ( in mM ) : NaCl 123 , KCl 2 . 5 , Na2HPO4 1 , NaHCO3 26 . 2 , Cacl2 2 . 4 , MgCl2 1 . 2 , glucose 10 , bicuculline 0 . 02 ( pH 7 . 4; 295 mosmol . kg−1; room temperature ) . A concentric bipolar steel electrode was placed in the stratum radiatum to evoke ( 0 . 01 Hz ) field excitatory postsynaptic potentials ( fEPSPs ) recorded with a glass electrode filled with ACSF . Data were collected with a multiclamp 700A ( Axon Instruments , http://www . axon . com ) , filtered at 3 kHz , sampled at 10 kHz and analyzed offline using pClamp 9 ( Axon Instruments ) . The initial slopes of the fEPSPs were measured from approximately 10%–40% of the rising phase . Paired-pulse ratio corresponds to the slope ratio of the second fEPSP to the first fEPSP . All data ( mean ± standard error of the mean ) were analyzed by a Student t-test ( two tailed ) or by analysis of variance followed by Newman-Keuls test when necessary . Correlation analysis was performed using the Spearman test . | The birth of adult hippocampal neurons is associated with enhanced learning and memory performance . In particular , spatial learning increases the survival and the proliferation of newborn cells , but surprisingly , it also decreases their number . Here , we hypothesized that spatial learning also depends upon the death of newborn hippocampal neurons . We examined the effect of spatial learning in the water maze on cell birth and death in the rodent hippocampus . We then determined the influence of an inhibitor of cell death on memory abilities and learning-induced changes in cell death , cell proliferation , and cell survival . We show that learning increases the elimination of the youngest newborn cells during a specific developmental period . The cell-death inhibitor impairs memory abilities and blocks the learning-induced cell death , the survival-promoting effect of learning on older newly born neurons , and the subsequent learning-induced proliferation of neural precursors . These results show that spatial learning induces cell death in the hippocampus , a phenomenon that subserves learning and is necessary for both the survival of older newly born neurons and the proliferation of neural precursors . These findings suggest that during learning , neuronal networks are sculpted by a tightly regulated selection of newly born neurons and reveal a novel mechanism mediating learning and memory in the adult brain . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neuroscience",
"rattus",
"(rat)"
] | 2007 | Spatial Learning Depends on Both the Addition and Removal of New Hippocampal Neurons |
Synchronization between neuronal populations plays an important role in information transmission between brain areas . In particular , collective oscillations emerging from the synchronized activity of thousands of neurons can increase the functional connectivity between neural assemblies by coherently coordinating their phases . This synchrony of neuronal activity can take place within a cortical patch or between different cortical regions . While short-range interactions between neurons involve just a few milliseconds , communication through long-range projections between different regions could take up to tens of milliseconds . How these heterogeneous transmission delays affect communication between neuronal populations is not well known . To address this question , we have studied the dynamics of two bidirectionally delayed-coupled neuronal populations using conductance-based spiking models , examining how different synaptic delays give rise to in-phase/anti-phase transitions at particular frequencies within the gamma range , and how this behavior is related to the phase coherence between the two populations at different frequencies . We have used spectral analysis and information theory to quantify the information exchanged between the two networks . For different transmission delays between the two coupled populations , we analyze how the local field potential and multi-unit activity calculated from one population convey information in response to a set of external inputs applied to the other population . The results confirm that zero-lag synchronization maximizes information transmission , although out-of-phase synchronization allows for efficient communication provided the coupling delay , the phase lag between the populations , and the frequency of the oscillations are properly matched .
Brain function emerges from the collective dynamics of coupled neurons , the structural connectivity among which enables correlations between their firing activities . As a result of these correlations , effective neuronal networks function collectively on a mesoscopic scale , comprising thousands of coupled neurons that operate together , giving rise to emergent behavior . In awake animals , this collective dynamics takes the form of recurrent series of high and low network activity , corresponding with repetitive epochs of increased excitation over inhibition followed by boosts of inhibition . This leads to the appearance of rhythmicity at certain frequency bands . In particular , oscillations in the gamma-band ( ) are observed in several cortical areas in relation with cognitive tasks [1] . Synchronized oscillations can increase the functional connectivity between neural assemblies by coherently coordinating their firing dynamics . This hypothesis , known as communication through coherence ( CTC ) , was proposed [2] as a mechanism by which gamma-band synchronization could regulate routing of information between brain areas . Since neuronal oscillations are associated with the dynamics of the excitatory-inhibitory balance , they represent periodic modulations of the excitability of neurons , which are more likely to spike within specific time windows ( i . e . when inhibition is low ) . If two neuronal populations oscillate with a constant phase difference , then an effective transmission of information between the two groups of neurons is achieved provided the spikes sent by a population reach systematically the other population at the peaks of excitability . In that way , modulation of the relative phases of the emerging rhythms might dynamically generate functional cell assemblies [3]–[5] . A key requirement of the CTC mechanism is the existence of a constant phase difference between the two neuronal networks that reliably allows their binding , favoring communication . This coordination can be expected to arise from the synaptic coupling between the neurons of the two populations . But this coupling is not instantaneous , since propagation times between different cortical regions can reach up to several tens of milliseconds [6] . Previous CTC studies have mainly concentrated on describing the dependence of the coherence on the phase lag between the neuronal populations [2] , [3] , [7] , without examining systematically the relationship between the actual coupling delay and the phase lag at which the coherence is maximal . In fact , coupled nonlinear oscillators are known to adjust their phases upon frequency locking , leading under certain conditions to either in-phase ( zero phase lag ) or anti-phase synchronization ( -phase lag ) [8] . Anti-phase patterns in cortical networks , for instance , have been widely studied [9] . Zero-lag synchronization , in turn , has been experimentally observed between gamma oscillations emerging from separated brain areas [10]–[12] . The conditions leading to zero-lag synchronization in neuronal oscillations are however somewhat stringent , requiring non-trivial spiking dynamics [13] or complex network architectures [14] , [15] . In particular , zero-lag synchronization between two cortical areas has been shown to be possible even with long axonal delays [15] , [16] , provided the two areas interact through a third oscillator , which could correspond to the thalamus [17] , [18] . But in contrast with most nonlinear oscillators neuronal populations are highly complex , since they contain a very large number of degrees of freedom ( corresponding to the individual neurons ) , their oscillations are a pure collective phenomenon ( the individual neurons in the population do not oscillate ) , and they operate in a broad frequency range . Additionally , neuronal populations are connected by a large number of axons , and inhomogeneities in the properties of those axons affect differentially the propagation speed of action potentials and lead to a wide spectrum of axonal delays rather than a uniform distribution [19] . It thus becomes necessary to study systematically the conditions under which two such complex oscillators synchronize ( i . e . lock their frequencies ) , what is the resulting phase difference between them , how does this phase difference relate with the coupling delay ( and with the frequency band being considered ) , and how is the efficiency of the communication between the two cortical areas affected by the delayed coupling . We address these questions in what follows . As mentioned above , within the CTC scenario effective communication arises when spikes from the emitting neuronal population reach the receiver population during the windows of maximum excitability . For this to happen two conditions have to be met: ( ) the two coupled oscillators should be frequency locked , and ( ) the transmission delay , the oscillation frequency , and the phase difference between the two oscillations should match . In particular , if the networks and the inter-connectivity is symmetric the second condition should hold in the two directions of spike propagation . The time delay ( or rather , the distribution of time delays ) is fixed as given by the anatomical connectivity . Therefore , it is the frequency of the oscillation spectrum what determines the particular phase lag that meets the matching condition . We have investigated whether this condition only occurs at specific rhythms , or if it holds at all frequencies . To this aim , we have represented mathematically two reciprocally connected identical neuronal populations using conductance-based models for both excitatory and inhibitory cells , and have studied how the heterogeneous axonal delays between the populations affect their synchronization . We have characterized the collective dynamics through a variable comparable to the local field potential ( LFP ) recordings [20] . In agreement with experimental data , the power of the modeled LFP decays with increasing frequencies [21] . Here we have focused on the particular dynamical regime in which the collective oscillations show a prominent contribution in the gamma range arising from the inhibitory ( GABAergic ) synaptic decay time constants [22] . Lower frequency bands contain a strong component arising from the noisy Poissonian distribution of interspike intervals ( ISI ) , which affect the synaptic activation and hence do not reflect the intrinsic dynamics of the network . On the contrary , higher frequency bands of small amplitude reflect the fast dynamics of the action potentials , also affecting the synapse activation time course . The modeled neuronal networks exhibit other well-known features of cortical dynamics , such as coexistence of irregular firing at the single-neuron level with collective rhythmicity at the population level , arising from the synaptic recurrent connections between the excitatory and inhibitory neurons [23] ( see Figure 1 ) . The excitatory and inhibitory synaptic currents are balanced by compensating the higher number of excitatory neurons ( of the whole network ) with fast spiking inhibitory neurons and with strong inhibitory synaptic conductances . As a consequence , the neurons remain excitable but spent most of their time with a membrane voltage that fluctuates under the firing threshold . The gamma rhythm emerges from the periodic changes of this balanced synaptic current , which leads to periodic modulation of the distance to threshold . We have characterized the global activity of the network by means of averaging measures such as the aforementioned local field potential ( LFP ) and the multi-unit activity ( MUA ) . We first used these measures to quantify phase coherence between the oscillatory activity of the two delay-coupled populations at varying mean axonal delays , observing transitions between in-phase and anti-phase dynamics . We next used information theory to quantify the response of one population ( the receiver ) to a varying external input impinging originally on the other population ( the emitter ) . Our results show that information transmission is enhanced at zero-lag ( in-phase ) synchronization , and decreases at long delays for which communication occurs through anti-phase dynamics .
We start by considering an isolated population of neurons , of which are excitatory and are inhibitory . Each neuron forms on average random connections within the network , and all pairs of coupled neurons exhibit a certain time delay , taken from a gamma distribution whose scale and shape parameters are both equal to unity . All neurons receive an external Poisson-distributed spike train whose instantaneous firing rate follows an Ornstein-Uhlenbeck process with a mean value set to . This input and the excitatory recurrent synaptic activity are balanced by the recurrent inhibitory synaptic flow , since the GABAergic conductances are stronger than the glutamatergic AMPA ones . Furthermore , the inhibitory neurons fire at higher rates than the excitatory cells . Therefore , the membrane voltage of the neurons fluctuates below threshold , occasionally crossing it [24] . Despite the fact that the neurons fire sparsely and irregularly ( see Figure 1A ) , a rhythmicity emerges when considering the dynamics of multiple action potentials elicited by thousands of neurons [23] . These oscillations represent the transient synchronized activity of neuronal assemblies , and can be revealed by population measures such as the local field potential ( Figure 1B ) and the multi-unit activity ( Figure 1C ) , defined in the Materials and Methods section . In the computational model used throughout this work , the collective oscillatory dynamics exhibit a prominent gamma rhythm ( Figure 1D ) , whose period is mainly determined by the decay time constant of inhibition [23] , [25] , [26] . Another way of understanding the emergent gamma oscillations is by looking at the coupling between the MUA and the LFP . Since the LFP mainly captures the synaptic currents impinging on the pyramidal neurons ( see Materials and Methods section ) , it is a measure of the excitability of the network . Hence , at those intervals in which inhibition is low ( i . e . the inhibitory synaptic current fades away ) , the probability of firing is high . Due to the recurrent connections between the excitatory and inhibitory neurons , both the initiation and termination of the population bursts occur with a certain periodicity . Here this oscillatory pattern is around due to the inhibitory decay time constants [22] . The LFP and MUA are mutually locked to this frequency ( Figure 2A ) , and the spikes occur with higher probability close to the troughs of the LFP ( i . e . the minimum of inhibition , Figure 2B ) . We next consider two bidirectionally coupled neuronal networks of the type described above . Connections between the two areas are excitatory: of the excitatory neurons of each network project randomly to of the neurons belonging to the other pool . Although these parameter values cannot be generalized to any two separate brain areas , for which the specific connectivity might determine their interactions , it is known that the probability of connection decays with distance [27]–[29] . Here we assume that the connectivity within a network is -fold the connectivity across networks , neglecting heterogeneity across neurons . Moreover , in order to obtain a certain amount of phase coherence between the two networks , we consider that the majority of excitatory neurons project onto the other network . A stronger ( weaker ) coupling will lead to unrealistically higher ( lower ) phase coherence values [30] . We have introduced time delays in the coupling between networks , assuming that the inter-areal delays are larger than the intra-areal delays due to long-range connections . We also consider that the inter-areal delays are distributed heterogenously across the system [19] , following a gamma distribution whose mean and variance increase systematically with the mean delay [15] . This accounts for the variability of transmission delays through axons with heterogeneous properties ( see Materials and Methods for the definition of the gamma distribution parameters ) . The mean inter-areal delay shown in the figures , hereafter termed , accounts for the latency between the generation of a spike in a presynaptic neuron from one network and the elicitation of a postsynaptic potential in the other network . When coupled , the LFP power spectra of the two networks show the same gamma profile as in the absence of coupling , while the corresponding time series exhibit a substantial degree of correlation ( Figure 3A inset ) . We next asked whether the broad spectrum of these neuronal oscillations allows for partial phase coherence to arise in specific frequency regions . Our phase coherence measure , described in the Materials and Methods section , quantifies between and the reliability of the phase difference between pairs of oscillations , at a given frequency . Figure 3B shows the phase coherence between the LFPs of the two populations for instantaneous coupling ( ) . According to the regions of statistical significance observed experimentally [30] , we considered phase coherence values above , which mainly occurs within the gamma band around the peak of the LFP power spectrum ( horizontal gray bar in Figure 3B ) . This threshold corresponds to around four times the average phase coherence of the uncoupled case ( see black dashed line in Figure 3B ) . We have also computed the time lag between the two signals ( i . e . the time shift separating two equal phases of the coupled LFPs arising from each population ) for all frequencies ( Figure 3C ) , still in the case . This time lag is only meaningful for significant phase coherence values that lead to a consistent across trials ( red crosses in Figure 3B ) . The figure shows that for frequencies at which the phase coherence is significant , the LFP gamma rhythms of the two populations oscillate in phase ( ) , i . e . the two LFPs are synchronized at zero lag . The error bars in Figure 3B , C represent the standard deviation across trials of phase coherence and respectively , and are only shown for the region of significant phase coherence , since outside that region the phase distribution is very broad due to the variability across trials . Even within the significant region the standard deviation of can be seen to decrease with increasing values of phase coherence , which confirms the inverse relation between phase coherence and the broadness of the phase distribution . The fact that the two populations synchronize at zero lag when the coupling delay is zero is to be expected , and we now ask what happens in the presence of time delays . Figure 4 shows the phase coherence spectrum between the LFP oscillations for three different values of . While phase coherence is again significant only around the gamma band ( Figures 4A , C , E ) , the time traces look very different for small and large delays , with mostly in-phase dynamics for small delays ( Figure 4B ) , whereas the populations are mostly in anti-phase for large delays ( Figure 4F ) . For intermediate delays , interestingly , two coherence peaks appear ( Figure 4C ) , and the corresponding time series exhibit both in-phase and anti-phase episodes ( Figure 4D ) . These results indicate that in-phase dynamics seems to persist for non-zero coupling delays , eventually transitioning to an anti-phase regime with smaller , although still significant , phase coherence . Both types of dynamics seem to coexist for intermediate delays . In order to verify these conclusions , we have extended the analysis to a range of axonal delays , from to , calculating the phase shift for the frequencies corresponding to both the peak of the power and the phase coherence spectra , termed . Figure 5A shows the value of the frequency at which the power spectrum is maximum , , as a function of the coupling delay . Note that varying does not change the frequency peak of the LFP power spectrum , which remains around for all coupling delays . We have added a gray bar delimiting the maximum power spectrum range within the gamma band corresponding to the extent of this local peak , highlighting the fact that the LFP gamma rhythm expands over a range of frequencies between approximately . On the other hand , clearly affects the frequency at which phase coherence is maximal , as shown by Figure 5B . In particular , exhibits a discontinuous jump around a coupling delay , where two peaks of phase coherence coexist ( consistent with the result shown in Figure 4C ) . The phase coherence values themselves are shown in color code in Figure 5C for different frequencies ( vertical axis ) and for varying ( horizontal axis ) . We have superimposed in that plot the line shown in panel A , which marks the maximum of the LFP power spectrum ( black dashed line ) within the gamma range , , as well as the whole extent of the peak ( vertical gray bar ) . The local peaks of phase coherence ( black lines ) corresponding to panel B are also superimposed to Figure 5C . For ( as in Figure 3 ) the peak of phase coherence almost coincides with the peak of power spectrum . For increasing , below , only the coherence peak at the lower frequency is significant , whereas between and only the coherence peak at the faster frequency is above threshold . The transition between these two regimes involves a coexistence of the local coherence peaks . We also observe that in both branches the frequency at which phase coherence is maximum decreases with the axonal delay , becoming clearly smaller than the gamma frequency peak ( dashed black line in Figure 5C ) . Making greater than , which approximately matches the period of the power spectrum peak ( ) , a new branch of phase coherence appears , thus leading again to coexistence of the two regimes . This emerging pattern is shown in Figure 5C for large inter-areal axonal delays and it is not marked in Figure 5B because the phase coherence is under the threshold . Hence , as exceeds , the scenario of relative phases is repeated but now with cycle skipping . The phase coherence patterns shown in Figure 5C are affected by the inter-areal delay variability . If is fixed to a constant value , the region of coexistence between the in-phase and anti-phase coherence patterns increases , and for delays approaching the oscillation period the new peak emerging at ( detectable in Figure 5C and corresponding to in-phase dynamics in Figure 5E ) becomes significant . This is shown in Supplementary Figure S1C , which displays the phase coherence for constant ( blue line ) , in comparison with the case ( black line ) and the one with drawn from a gamma distribution with mean ( red line ) . According to Figure 5C , maximum values of phase coherence appear at different frequencies for each . Significant values of phase coherence at a certain frequency can occur provided that there is a certain amount of spikes being simultaneously and reliably sent between the two networks . Since , by construction , the two neuronal pools are identical , the information flow can only be symmetrically transmitted for an in-phase , , and/or an anti-phase , , relationship between the two LFPs . Therefore , for any we can obtain the corresponding that satisfies or . From this expression we can thus expect that larger leads to smaller and that the anti-phase configuration is given by equal to half the period corresponding to , not to be mistaken with , half the gamma period and equal to . To verify the aforementioned remarks we have next calculated the time shift between the two coupled LFPs as . Figure 5D shows that , at the peak of the LFP power spectrum ( here ) , is zero for low ( ) and large delays ( ) . On the other hand , for intermediate ( ) and large delays ( ) corresponds to half the period of the gamma rhythm ( ) ) . As mentioned before ( see Figure 2 ) , at frequency the MUA and the LFP in each population are frequency locked . Therefore , for any axonal delay , the presynaptic spikes arrive within the troughs of the postsynaptic LFP . We can interpret these sharp transitions from in-phase to anti-phase oscillations , appearing with a periodicity given by , as the way by which the system keeps the symmetry for any . Since the maximum of phase coherence does not coincide with , we have also obtained along the peaks of phase coherence . Figure 5E confirms that only two patterns arise: in-phase and anti-phase , which can simultaneously occur in the region between and . The lowest frequency branch corresponds to , and thus to zero-lag synchronization . On the other hand , the highest frequency branch corresponds to a value that matches half the period of the corresponding frequency , i . e . ( labeled by a red line in the plot ) , and thus corresponds to anti-phase synchronization . The full values of the time shift for all frequencies are shown in color code in Figure 5F . The region of zero-lag synchronization disappears as the delay increases , giving way to a region of anti-phase synchronization . Due to the oscillatory dynamics , for greater than , frequencies close to the gamma peak are again compatible with an in-phase pattern . However , it is important to note that phase coherence is strongly decreased as the cycle is repeated again ( ) , probably due to loss of temporal self-coherence of the oscillations themselves . Thus , provided that the LFP-LFP phase coherence is significant , an effective coupling exists at which the two populations oscillate with a constant phase difference , which depends on the frequency and on the axonal delay . In particular , only two possible phase shifts are allowed , namely zero-lag ( ) and an anti-phase ( ) synchronization . Figure 5C shows that the frequency at which maximum phase coherence occurs , , might not correspond to the predominant gamma rhythm at , although it is close to it and within the extent of the gamma peak ( gray vertical bar ) . Thus , phase coherence is bounded by the region in which spikes are still phase locked to the LFP ( Figure 2 ) . The separation between and is clear when varies between and . Phase coherence is achieved at slower rhythms that still reliably carry the action potentials . Hence , the spikes elicited by each population systematically reach the other one at its excitability windows . Moreover , lower implies larger excitability windows and allows the networks to be synchronized in phase . For larger , corresponding slower frequencies lying outside the gamma peak do not efficiently transmit spikes , due to the bounded region in which MUA is locked to the LFP . Therefore , at large the system moves towards an anti-phase configuration , where the time lag matches and compensates for the inter-areal axonal delay . The LFP oscillations studied so far are complex rhythms that convey a wide range of frequencies with a predominant component in the gamma range . We have seen before that the axonal delay determines the relative dynamics of the coupled neuronal pools , which fall in either an in-phase or an anti-phase pattern . The phase relationship set by the two LFP signals is proposed to regulate the effectiveness of communication [2] . In other words , a stable phase difference determines the response of a neuronal population to inputs perturbing directly another area . Therefore , depending on the phase difference between two coherent LFPs , the response of the unperturbed population will replicate to a certain extent the response of the other population to the perturbation . We next study how , in the two different synchronization scenarios described above , inter-areal axonal delays affect information transmission during temporal windows , in which the phase difference and the frequency cannot be independent of each other . Note here the difference between phase coherence and effective communication . Rigorously speaking , communication occurs whenever spikes from one population arrive to the other one , and this is guaranteed provided that there is some coupling across networks . In contrast , effective communication refers to a more specific situation in which information about the stimulus is being carried by the coupled populations . We can obtain a good proxy for communication by quantifying the response of a neuronal population ( the receiver ) to a perturbation that affects indirectly its dynamics via a second population coupled to it ( the emitter ) , and which receives directly the perturbation . The success in communication can be observed in the transient amplification of the neuronal oscillations of the receiver [31] . The perturbation simulates different stimulus features , and consists of increases in the mean firing rate of the background synaptic activity impinging on a subpopulation of the emitter . We have examined , at different inter-areal axonal delays , how well the LFP and MUA power spectra of the receiver convey information about the external stimulus being applied to the emitter . Since the connectivity within and between the two neuronal networks exhibits a certain degree of clustering , exciting a subpopulation of adjacent excitatory neurons from an area in the emitter population triggers a response in a well-defined subpopulation of neighboring neurons in the receiving population . We have chosen a set of different input intensities , , affecting long-range excitatory neurons from the emitter population during a -second time window . As a consequence of this perturbation , the amplitude of the LFP power spectrum increases with the strength of the perturbation ( Supplementary Figures S2A , B , S3A , B and S4A , B ) , without altering the position of the gamma frequency peak ( ) , consistent with the results were reported in [20] . Perturbing one of the populations breaks the symmetry of the system , since now the firing activity of the emitter is enhanced with respect to the receiver . As shown by the maps of phase coherence plotted in Figure 6 , an increase of the external firing rate boosts phase coherence between the two LFPs . Moreover , the two frequency bands where phase coherence is significant merge into a single region at larger values of concentrating closer to the gamma frequency peak . The corresponding values are shown in Figure 7 ( note the different ranges of the axes , which now concentrate on the significant values of phase coherence to better observe the transition to the out-of-phase regime ) . At the gamma frequency peak the system undergoes a transition from in-phase to anti-phase dynamics as increases . Small lead to time shifts of the emitter's LFP relative to the receiver's LFP ( Figure 7A-B ) and thus , the two signals oscillate approximately in phase . However , the route to the anti-phase configuration changes as is strengthened . In particular higher trigger smoother transitions and the anti-phase regime becomes narrower . Figure 8B shows values tracked at the gamma frequency peak . Here , larger leads to a leader-laggard configuration in which the emitter LFP precedes the receiver LFP by a time lag that equals the axonal delay ( see dashed black lines ) . Supplementary Figures S2C , D , S3C , D and S4C , D show the phase coherence and time shift for and ( the same delays as Figure 4 ) for the whole range of frequencies . The dependence of the phase coherence on for different values is shown in Figure 6A–D , corresponding to a shift from a symmetric to an effectively asymmetric coupling . As the extra perturbation is applied only to one of the populations , the effective coupling approaches an unidirectional connectivity , although the structural links are not changed . This can be further explained by carrying on the same analysis in a structural unidirectional scenario , in which only one population projects onto neurons from the other network . Supplementary Figure S5A shows that increasing the delay of the unidirectional transmission , the networks keep the phase difference constant at approximately the same frequency close to the power spectrum peak frequency . This represents a leader-laggard configuration and is similar to what happens in Figure 6D , where an over-excited subpopulation is driving the coupling between the two networks , still bidirectional but strongly asymmetric . The decrease of phase coherence with axonal delay is due to the variability in delay times: fixing to a constant value of leads to maximal phase coherence values comparable to the case of no delay ( Supplementary Figure S1A ) . Figure S5B shows that for increasing inter-areal axonal delays , the time shift between the two synchronized networks ( at frequencies corresponding to the significant phase coherence of Figure S5A ) increases as long as is smaller than half the period of LFP oscillation ( ) and then approaches zero , thus leading again to a transition from in-phase to anti-phase synchronization at frequencies close to that of the power spectrum peak . Phase coherence can influence the transmission of information between neuronal populations . As mentioned in the Introduction , the CTC hypothesis [2] suggests that phase relations between coupled areas modulate the response of a receiver area to presynaptic input coming from an emitter area . In order to maximize this response , the axonal delay , the frequency of the oscillations and the phase difference should verify . When this relationship holds , spikes fired in the emitting population at a specific phase of the signal ( for instance at the troughs of the LFP , which correspond to the maxima of excitability ) arrive at the receiving area at the same phase ( and thus at the same excitability maximum ) , triggering a maximal response in the receiving area . On the contrary , if does not fulfill the relationship given above ( or if it randomly varies ) , effective communication will not be achieved [31] . This condition is relevant at the frequencies at which the MUA and the LFP are phase locked ( Figure 2 ) . Otherwise , the troughs of the LFP do not correspond to intervals of maximum firing within the same population , and the peaks of MUA do not occur reliably with the same periodicity as the LFP . In order to quantify the efficiency of communication , we have computed the mutual information ( defined in the Material and Methods section ) between the power spectrum at a frequency of both the LFP and MUA of the receiver and the set of stimuli applied to the emitter . This definition of information does not require any assumption about the stimulus features being encoded by the neural signals [32] , [33] . quantifies the reduction of the uncertainty in predicting the applied stimulus given a single observation of the triggered response , and is measured in units of bits ( means a reduction of uncertainty of a factor of two ) . Several experiments have been done with this tool to characterize information transmission in the primary visual cortex of macaques in response to a naturalistic stimulus [33] . Several other studies have been performed using the LFP power spectrum as a measure of mutual information , showing the usefulness of this approach both experimentally and computationally [20] . The advantages of this approach are described in detail in [34] , [35] . To compute , we have run extensive simulations to properly estimate the conditional probabilities used in mutual information measures . The techniques adopted in order to reduce the bias error of the estimation of conditional probability due to the finite number of samples are explained in the Material and Methods section . Figure 9 shows that the mutual information is non-negligible only within the gamma range ( ; bootstrap test ) , in a narrow region around the peak of the power spectrum . This is consistent with the fact that the emitter encodes the different stimulus strengths in the gamma band , i . e . other regions of the LFP power spectrum are not affected ( Figure S2–S4A ) . Therefore , information transmission occurs within the gamma peak ( the mutual information spectrum of the two networks , computed from the LFP , for is plotted in Supplementary Figures S2E–S4E ) . Moreover , functional interactions between coupled neuronal populations can be maximized if their phase difference is close to zero , i . e . for short axonal delays . While is lower when computed for the LFP power spectrum ( Figure 9A ) than for the MUA power spectrum ( Figure 9B ) , it decreases monotonically in both cases for increasing axonal delays . This behavior contrasts with the one shown in Figure 5C , in which the maximum phase coherence in the absence of stimulus occurs at varying frequencies for different . Moreover , lies outside the significant mutual information spectrum . However , at large enough the phase coherence pattern ( Figure 6D ) closely resembles the mutual information dependency with ( Figure 9 ) , since here . Mutual information encoded in the power spectrum is bounded to the frequencies at which spikes are maximally frequency locked ( Figure 2 ) . Although this measure does not take into account the phase difference between the two LFP signals , their dynamics clearly rely on their relative time lag . Therefore , significant phase coherence is needed in order to reliably connect in time the excitability time windows of both networks , but it is not sufficient to achieve a maximal response of the receiver . In order to meet this second requirement , the frequency at which phase coherence is obtained needs to carry a precise timing of the action potentials , otherwise the presynaptic current will not elicit a postsynaptic response . Even the emitter population can only encode the stimulus strength as variations in the amplitude of the gamma frequency peak , since it is at that changes in the LFP represent synchronized alterations in the MUA . A symmetric coupling scenario allows for two emerging stable regimes , in-phase and anti-phase , while in an asymmetric regime the most excitable network leads the dynamics ( ) . Therefore , in the presence of axonal delays , the latter case is not compatible with the in-phase/anti-phase condition . The symmetry breaking allows for to follow , increasing phase coherence at the gamma rhythm and thus the receiver's response . In summary , efficient communication needs a sufficient locking between the spikes being transmitted and the LFP , the carrier of this information . This is maximized at the gamma frequency peak , here , at which changes in the power spectrum due to external stimuli become particularly evident . The coupling axonal delay modulates the level of phase coherence within all the gamma range , and strong driving of one of the populations precisely favors the frequency channel . As observed above , the variability of axonal delay affects the regions where the phase coherence maximum is significant . Supplementary Figures S6A , B show the LFP and MUA mutual information in the unidirectional case . As in the case of bidirectional coupling , the flow of information occurs at , where the MUA and LFP are frequency locked and the emitter encodes the stimulus strength . Specially , mutual information is higher at small , when the networks are synchronized in phase . In the unidirectional configuration the mutual information shows a strong dependence on , as in the case of phase coherence discussed above . This is due again to the variability of axonal delays . For a fixed time delay , the mutual information in the unidirectional coupling case does not show a substantial decrease for increasing ( Supplementary Figure S1B ) . The bidirectional configuration also exhibits a less significant decrease of the mutual information maximum for constant increasing ( Supplementary Figure S1D ) . This is consistent with the phase coherence peak corresponding to in-phase dynamics that persists for increasing constant delay ( Supplementary Figure S1C ) . Our results show that phase coherence cannot be taken as a precursor of information transmission . Phase coherence can be achieved in a broad range of frequencies around the gamma peak ( Figure 6 ) . Therefore , the spikes impinging on each network are able to bound the two populations in a constant phase relationship , constrained by the symmetry of the effective coupling . However , in order to communicate , presynaptic spikes must trigger a postsynaptic response . This requires that the presynaptic action potentials are synchronized in time to facilitate the integration of the synaptic currents . Hence , changes in the LFP and MUA amplitude occur precisely at and mutual information does the same ( Figure 9 ) . Stimulus that are able to modify the response of a population within a wider frequency range ( i . e . not frequency specific ) could possibly alter the situation here described .
Here we have examined how heterogeneous inter-areal synaptic delay influences the coupling between the collective dynamics of two neuronal populations . To that end , we first used population measures such as the local field potential and the multi-unit activity , by analogy with experimental studies , to capture the collective oscillatory dynamics of individual neuronal populations . In the presence of excitatory coupling , the LFP and MUA activities of two identical delayed neuronal networks oscillate in the gamma range , with a significant broad peak between and , which does not depend on the axonal delay . The emergence of this gamma peak in the isolated populations is due to the recurrence between excitatory and inhibitory synaptic activity , as revealed by the phase locking between the LFP and MUA signals ( Figure 2 ) . In contrast with the power spectrum , phase coherence is strongly affected by the axonal delays between the populations . We have seen that in-phase and anti-phase patterns occur at various frequencies for different ranges of , with high values of phase coherence occurring at frequencies below the gamma frequency peak ( Figure 5 ) . The phase coherence pattern shown in Figure 5C corresponds to a pure symmetrical connectivity , in which both the structural and functional coupling are equal in both directions ( in contrast with the unidirectional case of Figure S5 ) . The reciprocity between the feedback and feedforward pathways across cortical areas is not an unrealistic assumption [36] , although the specificity of such synapses might differ in each direction in order to account for the different effects of the top-down and bottom-up projections . Here we show that increasing axonal delays lead to either an in-phase or anti-phase synchronization with a vanishing maximal phase coherence at frequencies below although lying within the gamma peak . Hence , in basal conditions , there is always a certain reliable phase relationship , provided is small , relative to the period . The interesting point raised by the communication through coherence hypothesis [2] , is whether phase coherence can forecast efficient communication between two populations in the presence of a stimulus . According to the modulatory role of the top-down pathway , attention can determine which visual cues we are aware of [37] , [38] . In principle two situations can arise: either a stimulus catches our attention ( such as an unexpected noise or object ) or we are being attentive to an expected stimulus ( such as waiting the traffic light to turn green ) . In the first situation the communication outline between a primary cortical area and the associative areas is driven by the stimulus , while in the second case it is due to the internal cognitive state . The firing activity in visual areas has been shown to significantly increase even in the attentive state prior to the stimulus presentation [39] . Hence our results , in which we progressively increase the firing rate impinging on one population , could be viewed as arising from the alteration of the underlying attentional state . The experimental study of [38] shows that a neuronal cell assembly in V is phase coherent with an area in V that responds to a selected stimulus , while it is not with a V area that is not driven by the input . Here we have not modeled a competitive scenario between two networks . Instead we have focused on the mechanisms by which two neural pools can modulate their communication when they are simultaneously oscillating in the gamma band . We have quantified the efficiency of communication between the two neuronal networks as the ability of a population to encode information of an input which perturbs directly another coupled population . Mutual information measures between either the LFP or MUA power spectrum and the set of applied stimuli show that significant values concentrate around the gamma frequency peak ( ) . Mutual information decreases for long inter-areal axonal delays , and is slightly lower when the neural response is characterized by the LFP power spectrum than by the MUA power spectrum . Despite the fact that the LFP reflects the afferent and local synaptic currents within a given neuronal network , and that the MUA only captures the action potentials within this network , these two signals are closely related . As mentioned above , the gamma LFP rhythm reflects the dynamics of the excitatory balance . Increases in inhibition silence the spiking activity and therefore the MUA signal , although the GABAergic current is enhanced . Decreases in inhibition boost the spiking activity and therefore the MUA signal , although the GABAergic current is reduced . The peak at in the LFP-MUA phase coherence ( Figure 2 ) reveals this phase locking between the two signals . The arrival of each set of presynaptic spikes perturbs the postsynaptic LFP and might or might not elicit a postsynaptic suprathreshold response ( captured by the postsynaptic MUA ) depending on the mean distance to the excitatory threshold of the populations . Bursts of activity occur at each pool with a periodicity that fluctuates within the gamma band and are locked to the troughs of the LFP at this frequency . According to the CTC hypothesis , maximum communication requires that spikes from each population reach the peaks of excitability of the target area simultaneously in both coupling directions . Thus , efficient communication is restricted to the gamma peak , as revealed by the mutual information ( Figure 9 ) and preferentially at relatively small . This condition is only met for in-phase or anti-phase synchronization of the gamma rhythm: small axonal delays are able to tie two LFP troughs only at zero-lag synchronization , while larger require anti-phase synchronization . In principle , as increases zero-lag synchronization could again mediate communication by skipping one cycle . However , due to loss of phase consistence , mutual information decays with increasing . Here we show that phase coherence emerges spontaneously due to the excitatory coupling between areas without the need of further constrains ( Figure 5C ) . Higher stimulation of a particular population ( the emitter ) , which enhances the LFP power spectrum amplitude of the gamma peak , increases the range of phase coherence to larger axonal delays ( Figure 6D ) . The delay determines the phase shift between the two signals , with the emitter leading the oscillations . According to [38] , phase coherence is revealing communication in the sense of spike propagation , which in our case extends to frequencies within the gamma peak . However , efficient communication in the sense of the information encoding in the postsynaptic response , is restricted to a narrower band ( Figure 9 ) that maximizes spike synchronization . Adopting a spectrum of delays with increasing variability for increasing values of , instead of an ( unrealistic ) constant delay , affects quantitatively the results of phase coherence and mutual information but does not produce any strong qualitative change in the findings of the paper . However the effect of variability cannot be ignored , given the dispersion of axonal delays observed in experimental studies [19] . Figure 10 shows a schematic diagram of the two oscillatory LFPs filtered around the gamma frequency peak ( ) with the bursts of spikes locked at their troughs in agreement with Figure 2 . For a delayed coupling , zero-lag synchronization does not lead to a symmetric configuration demanding that the two oscillations are reciprocally influenced at the same phase . Therefore the system rapidly shifts toward an anti-phase synchronization at which roughly equals half of the period of the LFP ( Figure 10B ) . Importantly , when the symmetry of the system is broken ( for instance by perturbing just one of the populations ) , the possible stable solutions are no longer the in-phase or the anti-phase regime . In this case , phase coherence can be achieved through a leader-laggard configuration in which the time lag equals the inter-axonal delay . Without the symmetry constraint , this situation is achieved at the gamma frequency peak , for which the spikes of each population are preferentially locked to the LFP and changes in their power spectrum are maximized . In conclusion , we have studied two neuronal populations coupled synaptically with non-negligible delays . Our modeling results show that the populations organize their joint collective dynamics in patterns of in-phase or anti-phase synchronization , depending on the delay . Unidirectional couplings , either structural or functional , lead to a leader-laggard configuration with an out-of-phase synchronization determined by the axonal delay . Our study shows the dichotomy between phase coherence and communication . Whereas phase coherence arises due to LFP phase perturbations through the propagated spikes , communication is caused here by an increase in the firing response . The first occurs at different frequencies for every in order to conserve the functional connectivity . The second requires the spikes to be tightly locked to the LFP and at a faster frequency to enable spike integration , and hence a signal response that can be synaptically propagated .
We consider two populations of neurons , of which are excitatory while the remaining are inhibitory [40] . Each neuron connects on average with other cells through only chemical synapses . The structural connectivity is built according with the Watts-Strogatz small-world algorithm [41] . The rewiring probability is set to , so that the connectivity shows a certain degree of clustering , which favors the connections between neighboring neurons . Coupling between the two networks is mediated by of the neurons of one population making random long-range excitatory projections with of the neurons belonging to the other population . Here we assume that the connectivity within a network is -fold the connectivity across networks , neglecting heterogeneity across neurons . Moreover , in order to obtain a certain amount of phase coherence between the two networks , we consider that the majority of excitatory neurons project onto the other network . A stronger ( weaker ) coupling will lead to unrealistically higher ( lower ) phase coherence values [30] . We introduced a synaptic transmission delay within and among the networks , taken from a gamma distribution , assuming that internal delays ( taken from a gamma distribution whose scale and shape parameters are fixed to ) are smaller than the inter-area delays . The axonal delays , termed in the paper , stand for the time between the generation of a spike in a presynaptic neuron from one network and the elicitation of a postsynaptic potential in the other network . These delays are taken from a gamma distribution whose mean and variance increase with increasing . We choose the scale parameter of the distribution equal to unity , so that the shape parameter equals . In this way the coefficient of variation ( CV ) decreases for increasing mean as . In our analysis we systematically vary between and . Each neuron is dynamically described by the Hodgkin and Huxley ( HH ) model . The dynamics of the membrane voltage is given by: ( 1 ) where ( ) is the membrane capacitance for inhibitory ( excitatory ) neurons , the constants , , and are the maximal conductances of the sodium , potassium , and leakage channels , respectively , and , , and stand for the corresponding reversal potentials . According to the HH formulation , the voltage-gated ion channels are described by the following set of differential equations ( 2 ) where the gating variables , and represent the activation and inactivation of the sodium channels and the activation of the potassium channels , respectively . The voltage-dependent transition rates are ( 3 ) Given that activates rapidly , we replace it by its steady-state value . In Equation ( 1 ) is the synaptic current coming from the neighboring neurons impinging on a neuronal cell: ( 4 ) where is the synaptic conductance and is the reversal potential of the synapse . For positive values of the synapse is depolarizing or excitatory ( for glutamate receptors ) , otherwise it is hyperpolarizing or inhibitory ( for GABA receptors ) . In the equation ( 4 ) the synaptic conductance is described by: ( 5 ) where and are the decay and rise synaptic time , respectively , and is tuned in order to obtain a balance between excitation and inhibition . The constant is set to maintain the postsynaptic potential ( PSP ) amplitudes within physiological ranges . All parameters values can be found in [26] , [42] . In equation ( 1 ) represents an heterogenous Poisson train of excitatory presynaptic potentials with a mean event rate that varies following an Ornstein-Uhlenbeck process ( see Supplementary Figure S7 ) . This incoming external current mimics the direct input from any other area external to the network considered here . The instantaneous rate , , of the external excitatory train of spikes is generated according to an Ornstein-Uhlenbeck process , as considered in [20]: ( 6 ) where is the standard deviation of the noisy process and is set to . is set to , leading to a power spectrum for the time series that is flat up to a cut-off frequency . is a Gaussian white noise . The model has been integrated using the Heun algorithm [43] , with a time step of . All simulations represent seconds of activity . The connectivity , initial conditions and noise realization were varied from trial to trial . We quantified the activity of the network in two different ways . We calculated the multi-unit activity ( MUA ) as the total number of spikes per unit time in the population , and the local field potential ( LFP ) as the sum of the absolute values of the excitatory and inhibitory synaptic currents acting upon the excitatory neurons , averaged over this population [20]: ( 7 ) where denotes the average over all excitatory neurons . The term accounts for both the external excitatory heterogeneous Poisson spike train and the recurrent excitatory synaptic current due to the network , while corresponds to the recurrent inhibitory synaptic current . represents the resistance of a typical electrode used for extracellular measurements , here chosen to be . We computed the power spectral density of LFPs and MUAs using the Welch method: the signal is split up into point segments with overlap . The overlapping segments are windowed with a Hamming window . The modified periodogram is calculated by computing the discrete Fourier Transform , and then computing the square magnitude of the result . The modified periodograms are then averaged to obtain the PSD estimate , which reduces the variance of the individual power measurements . The code has been implemented in MATLAB . Spectral quantities are averaged over trials and phase coherence over trials . Phase coherence is calculated as in [30]: ( 8 ) where and denote the two signals , and is the cross-spectrum between them . Since in each trial the cross spectral density is normalized by its amplitude , each term of the sum is a unit-length vector representation of the phase relation . In other words , is the phase lag between the two signals at frequency in the data segment . Hence quantifies how broad is the distribution of within the 2π-cycle . Averaging across all data segments provides a mean angle . In our work is converted into a time shift , termed in the paper , dividing by the corresponding frequency . This quantity measures the time separation between an LFP maximum in one population and the following maximum belonging to the other population . An important mathematical tool to quantify information transmission in noisy systems is provided by information theory . We calculate the Mutual Information between the stimulus and the response as follows . The broadband LFP signal reproduces the variations in neural population activity over a wide range of time scales [44] . Thus LFPs signals are useful to qualitatively characterize mechanisms of information processing , because it is possible through them to verify if there are priviliged time scales for information processing . We can think that information is spread over all frequencies , or that each frequency contributes separately to the information representation . Given that we are interested in how the collective dynamics of the population carries information , we quantify the neural response as the power of either the LFP or the MUA at frequency , and we consider as stimuli different external firing rates impinging on one of the two populations . We compute the information between the stimulus and the response as: ( 9 ) where is the probability of presenting stimulus ( equal to the inverse of the total number of different external firing rates , namely of stimuli ) , is the probability of observing power across all trials in response to any stimulus , and is the probability of observing power at frequency in response to a single stimulus . quantifies the reduction of uncertainty about the stimulus that can be gained from observing a single-trial neural response , and we measured it in units of bits ( bit means a reduction of uncertainty of a factor of two ) [35] . This measure allows us to evaluate how well the power of either the LFP or MUA encodes the stimulus at a certain frequency . To facilitate the sampling of response probabilities , the space of power values at each frequency was binned into equipopulated bins [33] . We used seven different firing rates of the external Poisson-distributed input , for a time . An important issue to be solved regarding the calculation of the theoretical mutual information is that it requires knowledge of the full stimulus-response probability distributions , and obviously these probabilities are calculated from a finite number of stimulus-response trials . This leads to the so-called limited sampling bias , which corresponds to a systematic error in the estimate of information . We used the method described in [45] to estimate the bias of the information quantity and then we checked for the residual bias by applying a bootstrap procedure in which mutual information is calculated when the stimuli and responses are paired at random . If the information quantity is not zero ( it should be in the case of non finite samples ) , this is an indication of the bias and the bootstrap estimate of this error should be removed from the mutual information . After applying these procedures , the information quantity estimation could be defined as significant . Several toolboxes provide different bias-correction techniques , which allow accurate estimates of information theoretic quantities from realistically collectable amounts of data [46] , [47] . In order to accomplish those tasks , we used the Information Breakdown Toolbox ( ibTB ) , a MATLAB toolbox implementing several information estimates and bias corrections . It does this via a novel algorithm to minimize the number of operations required during the direct entropy estimation , which results in extremely high speed of computation . It contains a number of algorithms which have been thoroughly tested and exemplified not only on spike train data but also on data from analogue brain signals such as LFPs and EEGs [47] . | The correct operation of the brain requires a carefully orchestrated activity , which includes the establishment of synchronized behavior among multiple neuronal populations . Synchronization of collective neuronal oscillations , in particular , has been suggested to mediate communication between brain areas , with the global oscillations acting as “information carriers” on which signals encoding specific stimuli or brain states are superimposed . But neuronal signals travel at finite speeds across the brain , thus leading to a wide range of delays in the coupling between neuronal populations . How the brain reaches the required level of coordination in the presence of such delays is still unclear . Here we approach this question in the case of two delay-coupled neuronal populations exhibiting collective oscillations in the gamma range . Our results show that effective communication can be reached even in the presence of relatively large delays between the populations , which self-organize in either in-phase or anti-phase synchronized states . In those states the transmission delays , phase difference , and oscillation frequency match to allow for communication at a wide range of coupling delays between brain areas . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"computational",
"neuroscience",
"neuroscience",
"biology",
"and",
"life",
"sciences",
"computational",
"biology"
] | 2014 | Phase-Coherence Transitions and Communication in the Gamma Range between Delay-Coupled Neuronal Populations |
Myelin is essential for rapid saltatory conduction and is produced by Schwann cells in the peripheral nervous system and oligodendrocytes in the central nervous system . In both cell types the transcription factor Sox10 is an essential component of the myelin-specific regulatory network . Here we identify Myrf as an oligodendrocyte-specific target of Sox10 and map a Sox10 responsive enhancer to an evolutionarily conserved element in intron 1 of the Myrf gene . Once induced , Myrf cooperates with Sox10 to implement the myelination program as evident from the physical interaction between both proteins and the synergistic activation of several myelin-specific genes . This is strongly reminiscent of the situation in Schwann cells where Sox10 first induces and then cooperates with Krox20 during myelination . Our analyses indicate that the regulatory network for myelination in oligodendrocytes is organized along similar general principles as the one in Schwann cells , but is differentially implemented .
Myelin is a vertebrate invention and provides the basis for rapid saltatory conduction throughout the nervous system . It is produced by oligodendrocytes ( OL ) in the central nervous system ( CNS ) and Schwann cells in the peripheral nervous system ( PNS ) during terminal differentiation . In these cells myelin formation is controlled by a specialized regulatory network that – like myelin - is restricted to vertebrates . Much has been learned about this network in myelinating Schwann cells in which the transcription factor Sox10 plays a central role ( for review , see [1] ) . Sox10 is required during all developmental stages and causes the sequential induction of stage-specific transcription factors that cooperate with Sox10 in the next stage of development [2]–[4] . In pro-myelinating Schwann cells Sox10 teams up with its target Oct6 [5] to induce the zinc finger transcription factor Krox20 [6] . This sets the stage for terminal differentiation as the combination of Sox10 and Krox20 constitutes the main driving force for PNS myelination [7] , [8] . Transcription factors with roles in differentiation and CNS myelination have also been identified in OL ( for review , see [9] ) . Several key factors such as Olig2 , Olig1 and Nkx2 . 2 differ from the ones known to be important in Schwann cells [10]–[12] , and only few interactions have been defined so far [13] , [14] . As a consequence , it is not yet clear how these OL-specific transcription factors cooperate in their regulatory network to achieve myelination . One of the transcription factors with relevance for Schwann cell and OL differentiation is Sox10 . It therefore provides an excellent tool to study the myelin-specific regulatory network in both glial cell types . In the oligodendroglial lineage Sox10 expression starts early in the oligodendrocyte precursor cell ( OPC ) where it is already required in combination with its close relative Sox9 for survival and proper migration [15]–[17] . Its most obvious function , however , is during terminal differentiation and myelination . At the time of birth Sox10-deficient mice possess a full complement of OPC , yet exhibit little myelin gene expression [16] . This is in part attributed to direct activation of myelin genes by Sox10 [16] , [18] . Myelin Regulatory Factor ( Myrf , also known as Mrf and Gm98 ) is another transcription factor critically required for CNS myelination [19] . Unlike Sox10 , Myrf is not expressed in OPC and becomes expressed only upon cell cycle exit in promyelinating OL . Constitutive Myrf-deficient mice exhibit severe myelination defects . When Myrf is deleted conditionally in mature OL , oligodendroglial identity and the integrity of CNS myelin cannot be maintained [20] . Whereas myelin gene expression is strongly dependent on Myrf , Sox10 levels are not [19] . Considering the involvement of Sox10 and Myrf in OL differentiation and myelination , their genetic and functional relationship needs to be determined . As expression patterns and published evidence argue against a role for Myrf in Sox10 induction , we analyzed whether Sox10 is required for Myrf expression using mice in which Sox10 was selectively deleted in the CNS . These mice suffered from severe hypomyelination , but survived for three weeks after birth . Their analysis and accompanying studies show that Myrf is not only genetically downstream of Sox10 , but also represents a direct Sox10 target gene . Once induced by Sox10 , Myrf cooperates with its inducer to jointly activate myelination . Its function and relation to Sox10 is thus similar to Krox20 in Schwann cells . The present study defines a key circuit of the myelin-specific regulatory network in OL and emphasizes similarities and differences to Schwann cells .
Mice with constitutive Sox10 deletion are unable to breathe because of severe PNS defects and die immediately after birth [2] . At this time , OL differentiation and CNS myelination have barely started so that these mice do not allow a full evaluation of Sox10 in CNS myelination [16] . To sidestep perinatal lethality , we generated mice in which Sox10 was specifically deleted in the CNS using a floxed Sox10 allele [3] in combination with a Brn4::Cre transgene [21] . The resulting mice are henceforth referred to as Sox10ΔCNS . In these mice Cre expression is restricted to CNS and limb bud ectoderm . Within the CNS , Cre is already present throughout the early ventricular zone and allows efficient Cre-mediated gene deletion in neural precursors and all neurons and glia derived from them . On a gross morphological level , Sox10ΔCNS mice were indistinguishable from littermates during the first days after birth . However , starting from the second week Sox10ΔCNS pups shivered and exhibited an unsteady , shaky gait . Most Sox10ΔCNS mice died at the end of the third week . None survived past postnatal day ( P ) 24 . For analysis we concentrated on the spinal cord . In Sox10ΔCNS mice , Sox10 was already efficiently deleted by 12 . 5 days post coitum ( dpc ) so that OPC developed in the complete absence of Sox10 ( data not shown ) . As a consequence and in contrast to the wildtype , Sox10 was absent from the postnatal spinal cord of mutant mice ( compare Figure S1A–D , S to Figure S1E–H , S ) . Sox10 deletion did not dramatically alter oligodendroglial expression of its close relatives Sox8 and Sox9 ( Figure S1I–R ) . Sox9 was equally down-regulated in wildtype and Sox10ΔCNS spinal cord during the phase of active myelination ( Figure S1T ) . Sox8 remained expressed despite the absence of Sox10 ( Figure S1M–P ) . The fact that there were fewer Sox8-positive cells from P7 onwards ( Figure S1U ) reflects the reduced number of oligodendroglial cells in the mutant ( see below ) . To further characterize the phenotype of Sox10ΔCNS mice , we analyzed oligodendroglial lineage and differentiation markers including the OPC marker Pdgfra . At 18 . 5 dpc , the number of Pdgfra-positive OPC was very similar in wildtype and Sox10ΔCNS embryos as determined by in situ hybridization ( Figure S2A–C ) . This agreed well with previous findings on Sox10−/− mice [16] and argues that OPC numbers and - because they make up the vast majority of oligodendroglial cells in the late prenatal spinal cord - overall oligodendroglial cell numbers are not affected during embryogenesis . Extending the studies to the early postnatal period , we performed immunohistochemical analyses for the lineage marker Olig2 ( Figure S2D–L ) . At P3 , the number of Olig2-positive cells was as high in Sox10ΔCNS spinal cord as in wildtype ( Figure S2D , E , I ) again confirming that oligodendroglial cell numbers are first not affected by Sox10 loss . From P3 to P7 , the number of Olig2-positive cells further increased in the wildtype spinal cord before reaching its maximum at the end of the second week ( Figure S2D , F–H ) . This increase was not observed in Sox10ΔCNS mice ( Figure S2D , J–L ) . Instead , the number of Olig2-positive cells remained fairly constant in Sox10ΔCNS mice over the first three postnatal weeks . Myelin gene expression in the postnatal Sox10ΔCNS spinal cord was analyzed by in situ hybridization with Mbp and Plp probes . At P3 , Mbp- or Plp-positive cells were rarely seen in the mutant white matter , while present in substantial numbers in the wildtype ( compare Figure 1A , I to Figure 1E , M ) . At P7 , Mbp- and Plp-positive cells had increased in the Sox10ΔCNS spinal cord , but were still much fewer than in the wildtype ( compare Figure 1B , J to Figure 1F , N ) . This trend persisted: The number of Mbp- and Plp-expressing cells in Sox10ΔCNS mice continued to increase , but their number failed to catch up with the wildtype until time of death ( compare Figure 1C , D , K , L to Figure 1G , H , O , P ) . Histological and ultrastructural analyses of spinal cord white matter from Sox10ΔCNS mice at P14 failed to detect myelinated axons ( compare Figure 2A with Figure 2B , Figure 2D with Figure 2E ) . The only myelin present in Sox10ΔCNS mice was in PNS structures such as spinal nerve rootlets and produced by Sox10-expressing Schwann cells ( Figure 2B , C , F ) . Despite complete loss of myelin , residual myelin gene expression was still observed in Sox10ΔCNS mice . We have previously shown that the closely related Sox8 is co-expressed with Sox10 during OL development and participates in OL differentiation , although it cannot replace Sox10 or fully compensate its loss [22] , [23] . To test whether Sox8 is responsible for the residual myelin gene expression in Sox10ΔCNS mice we generated Sox10ΔCNS Sox8lacZ/lacZ mice , and analyzed myelin gene expression until time of death during the third postnatal week . After combined loss of Sox8 and Sox10 , Mbp and Plp expression was completely missing ( compare Figure 2G , H , K , L to Figure 2 . I , J , M , N ) . Immunohistochemistry for Olig2 on P7 confirmed that Olig2-positive cells were still present , although in slightly reduced number ( 29 . 8±4 . 0% of spinal cord cells in Sox10ΔCNS Sox8lacZ/lacZ mice as compared to 36 . 5±1 . 6% in the wildtype , corresponding to a reduction of 18% in the mutant ) . This confirms that residual myelin gene expression in Sox10ΔCNS mice is Sox8-dependent . In situ hybridizations were also carried out with a Myrf probe . Intriguingly , Myrf expression was dramatically reduced in Sox10ΔCNS mice . Myrf-positive cells appeared in substantial numbers not before P7 and increased slowly until P21 without ever coming close to wildtype numbers ( compare Figure 1Q–T to Figure 1U–X ) . Upon additional deletion of Sox8 , Myrf expression was completely absent at all times ( compare Figure 2O , P to Figure 2Q , R ) . The reduction in Myrf expression was not only observed on transcript but also on protein level ( compare Figure 3A–C to Figure 3D–F , and for quantitation Figure 3Y ) , and resembled the markedly reduced expression of the early oligodendroglial differentiation marker CC1 ( Figure 3G–L ) . Expression of 2′ , 3′-cyclic nucleotide 3′ phosphodiesterase ( CNPase ) was drastically reduced as well ( Figure 3M–R ) . Two markers of the promyelinating stage , in contrast , exhibited only mild ( Nkx2 . 2 ) or no ( Gpr17 ) alterations in Sox10ΔCNS mice ( Figure 3S-X- , Z and Figure S3A ) . RT-PCR experiments on RNA from P7 spinal cord confirmed a statistically significant downregulation of Mbp and Myrf ( Figure S3B ) . Zfp488 as a transcriptional regulator in already myelinating OL [24] was similarly downregulated , whereas Nkx2 . 2 expression was much less affected in good agreement with the immunohistochemical data . No alterations in transcript levels were observed for Olig2 , Pdgfra , Zfp191 , Tcf4 , Yy1 , and Hes5 in P7 spinal cord of Sox10ΔCNS mice relative to wildtype mice , while there was a trend towards upregulation for Id2 ( P = 0 , 09 ) and statistically significant upregulation for Id4 in the mutant ( Figure S3B ) . Similar results were also obtained at P14 – the main differences being that at this later stage Olig2 and Nkx2 . 2 levels were also significantly downregulated and levels of Id2 and Id4 were no longer increased ( Figure S3C ) . Considering that most OPC markers , and Nkx2 . 2 and Gpr17 as two markers of the promyelinating stage exhibited only mild changes in their expression , while markers of the myelinating stage were severely affected , it seems reasonable to assume that oligodendroglial development in Sox10ΔCNS mice is blocked at the transition from the promyelinating into the myelinating stage . Myrf appears to be one of the earliest , severely affected markers . To explore the relationship between Sox10 and Myrf , we first used primary oligodendroglial cultures . We transfected OPC with a Sox10-specific shRNA or a scrambled version in the presence of GFP , and analyzed the cells after one day in culture under differentiating conditions . While nearly all cells ( 98±2% ) transfected with a scrambled shRNA exhibited Sox10 expression , only very few ( 16±5% ) continued to be Sox10-positive in the presence of a Sox10-specific shRNA ( Figure 4A ) . At the same time , cells transfected with the Sox10-specific shRNA failed to induce Myrf or Mbp expression in contrast to untransfected cells and cells transfected with the scrambled shRNA version ( Figure 4B , C ) . This confirms that Myrf expression depends on Sox10 . To complement the loss-of-function studies in Sox10ΔCNS mice and oligodendroglial cultures , we performed gain-of-function studies in the early chicken embryo ( Figure 4D , E ) . We electroporated Sox10 into the neural tube of chicken embryos at Hamburger-Hamilton ( HH ) stage 11 . 24 h later , we detected Myrf on the electroporated side of the neural tube but not on the non-electroporated side ( Figure 4E ) or in control neural tubes that were electroporated with GFP only ( Figure 4D ) . Myrf induction occurred only in cells with very high levels of Sox10 whereas induction of the neural crest marker HNK-1 was seen under identical conditions in nearly all electroporated cells including those with low Sox10 levels [25] . While this argues that Myrf induction by Sox10 in the early neural tube is not physiological , it nevertheless shows that Sox10 is capable of inducing Myrf in the CNS in principle . In contrast , Sox10 overexpression failed to induce Myrf in S16 Schwann cells ( Figure S4A–F ) arguing that even high amounts of Sox10 cannot induce Myrf in this type of peripheral glia . In case of a direct effector-target gene relationship induction should be mediated by one or several regulatory regions of the Myrf gene . To identify such regions we searched for evolutionarily conserved non-coding sequences ( ECR ) within or near the Myrf gene . Using the ECR browser ( http://ecrbrowser . dcode . org/ ) we identified 13 ECR in the interval between the Myrf flanking genes Dagla and 1810006K21 Rik ( Figure 4F ) . Most ECR were located in the 11 kb long intron 1 . Although originally identified by their conservation among mammals , especially those in intron 1 were conserved down to chicken . All ECR contained at least one potential binding site for Sox proteins as defined by a perfect or near perfect match ( ≤1 mismatch ) to the 7 bp consensus binding motif 5′- ( A/T ) ( A/T ) CAA ( A/T ) G-3′ . However , it is very difficult to predict from the presence of such sites whether a Sox protein actually binds and functions through that site ( for review , see [26] ) . Considering the large number of potential Sox binding sites and the low predictive power of their presence , we decided to assess regulatory potential and Sox10-responsiveness of the ECR in luciferase assays . We combined each ECR with the minimal Hsp68 promoter and a luciferase reporter and transfected in rat 33B oligodendroglioma [27] . This cell line expresses Sox10 , Sox8 and Olig2 endogenously and thus bears resemblance to oligodendroglia ( data not shown ) . The proximal enhancer modules 1 and 2 from the Mbp upstream region [28] served as positive control and induced luciferase levels threefold above the levels obtained with the Hsp68 promoter alone ( Figure 4G ) . Of all ECR tested , ECR9 was the only one that reproducibly induced luciferase levels in 33B cells higher than the Mbp regulatory region ( Figure 4G ) . ECR9 also responded strongest to co-transfection of a Sox10-specific shRNA ( relative to co-transfection of scrambled shRNA ) by losing part of its activity ( Figure 4I ) . It is localized in intron 1 and exhibits conservation in mammals and birds . ECR9 similarly increased luciferase activity when combined with the minimal promoter of its own gene ( Figure 4H ) . This promoter had basal activity in 33B cells comparable to the Hsp68 minimal promoter ( data not shown ) and was refractory to the presence of a Sox10-specific shRNA ( Figure 4J ) . However , once combined with ECR9 , induction of the luciferase reporter was not only dramatically increased ( Figure 4H ) , but also became strongly sensitive to the presence of this shRNA ( Figure 4J ) . The strong negative influence of the Sox10-specific shRNA on ECR9 argues that it is responsive to Sox10 . To verify that Sox10 is bound to ECR9 in OL we next performed chromatin immunoprecipitation ( ChIP ) experiments ( Figure 5A–E ) . When chromatin was prepared from 33B cells and subjected to precipitation using antibodies directed against Sox10 , substantial amounts of ECR9 were found ( Figure 5B ) . Precipitates from parallel experiments with preimmune serum contained much less ECR9 . Such high and specific enrichment was not detected for the minimal Myrf promoter , ECR11 and ECR12 nor for control fragments located in the 5′ and 3′ flanking regions of the Myrf gene or in the adjacent Dagla gene ( Figure 5A , B ) . Interestingly , none of the fragments including ECR9 was enriched when chromatin precipitating antibodies against Sox8 were used for ChIP arguing that ECR9 is predominantly bound by Sox10 rather than Sox8 in 33B cells ( Figure 5C ) . Substantial and selective enrichment of ECR9 was also obtained when ChIP experiments were carried out with Sox10-specific antibodies on primary rat oligodendroglial cells cultured for one day in differentiation medium ( Figure 5D ) . When cells were kept instead in proliferation medium as OPC , no ECR9 enrichment was observed ( Figure 5D ) . In agreement with these ChIP data from cultured cells , ECR9 was also preferentially precipitated by Sox10-specific antibodies from chromatin prepared from spinal cord of P14 wildtype mice ( Figure 5E ) . When spinal cord was from Sox10ΔCNS mice , ECR9 enrichment was no longer visible ( Figure 5E ) . These results confirm the presence of Sox10 on ECR9 in an oligodendroglial cell line , in differentiating OL and in vivo . As ChIP experiments cannot distinguish between direct and indirect binding we searched for Sox10 binding sites within ECR9 . None of the 11 potential sites ( Figure S5A ) completely matched the Sox consensus binding site . Some were clustered so that eight oligonucleotides were sufficient to cover all sites in electrophoretic mobility shift assays ( EMSA ) . Site B and site C/C′ from the Myelin Protein Zero gene ( Mpz ) promoter [29] served as positive controls and marked the position of Sox10 monomers and dimers bound to DNA after electrophoresis in native gels ( Figure S5B ) . Surprisingly , only two of the eight oligonucleotides readily bound Sox10 . Oligonucleotide 3 preferentially bound Sox10 as a dimer , whereas oligonucleotide 4 preferred the monomer . To corroborate the binding site on the nucleotide level we introduced mutations . Oligonucleotide 3 contained three potential binding sites ( Figure S5C ) . Mutation of the first site in oligonucleotide 3a did not interfere with binding of a Sox10 dimer , while mutation of the second and third site in oligonucleotides 3b and 3c dramatically decreased overall binding of Sox10 and changed binding preference from dimer to monomer ( Figure 5F ) . Joint mutation of the second and third site in oligonucleotide 3bc abolished Sox10 binding altogether ( Figure 5F ) . This defines the dimer site in ECR9 as a composite element in which two imperfect Sox consensus sites are separated by four bp and arranged in a head-to-tail fashion ( Figure S5C ) . In oligonucleotide 4 , only one potential Sox10 binding site was present and its mutation in oligonucleotide 4a prevented binding completely ( Figure 5F ) . Both sites 3 and 4 exhibit strong sequence conservation in mammals ( Figure S5A ) . Interestingly , they were also bound by Sox8 in vitro despite the fact that ECR9 occupancy in vivo seems restricted to Sox10 ( Figure S5D ) . To validate the functional importance of the identified Sox10 binding sites , we introduced the 3bc and 4a mutations into ECR9 and analyzed the consequences on activity and ability to respond to Sox10 . In transiently transfected 33B cells , mutation of either dimer or monomer site led to a significant decrease in the ability of ECR9 to activate expression of a luciferase reporter under control of the Myrf minimal promoter ( 3bc and 4a in Figure 5G ) . Joint inactivation of both sites reduced activity even further ( mt in Figure 5G ) . Interestingly , the residual activity of an ECR9 variant with only one mutated site was still sensitive to the presence of a Sox10-specific shRNA . This sensitivity was lost after simultaneous mutation of monomer and dimer site ( Figure 5G ) . Both the wildtype ECR9 and the variant with inactivated monomer and dimer sites were combined with the Hsp68 minimal promoter and a lacZ reporter in a transgenic construct and used for the generation of transgenic animals by pronucleus injection ( Figure 6A ) . The Hsp68 minimal promoter was used because it has low activity and causes little ectopic expression in transgenic mice when combined with a lacZ reporter [30] , [31] . For wildtype ECR9 , eight transgenic animals were obtained and killed for analysis at P7 . X-gal staining revealed expression of the lacZ transgene in five of these animals ( Figure 6B ) . All five exhibited transgene expression in oligodendroglial cells despite some variability in expression levels and occurrence of additional expression sites ( Figure 6B , C ) . For four animals there was additional lacZ expression in a subset of spinal cord neurons and/or in PNS glia . One showed transgene expression in cartilage . None of these sites express Myrf normally [19] . The appearance of ectopic expression sites in this kind of analysis is common and likely results from the fact that additional regulatory sequences of the Myrf gene were missing or that the native chromatin context was not preserved in the transgenic construct . Oligodendroglial expression of the ECR9wt-lacZ transgene was restricted in all animals to differentiating cells . This is paradigmatically shown for transgenic animal 1 . As evident from X-gal staining , transgene expression in the thoracic spinal cord was strongly enriched in white matter ( Figure 6C ) . Immunohistochemistry furthermore confirmed that β-galactosidase colocalizes with Sox10 and Olig2 as oligodendroglial lineage markers ( Figure 6E , F ) and with Myrf as a marker of promyelinating and differentiating OL ( Figure 6G ) . In contrast there is virtually no overlap of β-galactosidase with Pdgfra as an OPC marker and Gfap as an astrocytic marker ( Figure 6H , I ) . In this particular transgenic animal there was furthermore no overlap between β-galactosidase and the neuronal marker NeuN ( Figure 6J ) . These data define ECR9 as an enhancer in differentiating OL . We also generated five transgenic animals in which the lacZ reporter was under control of the mutant ECR9 ( Figure 6B ) . None of these animals exhibited expression of the ECR9mt-lacZ transgene in the oligodendroglial lineage ( Figure 6D ) . Expression was only found in subsets of spinal cord neurons and cartilage . These results argue that the presence of Sox binding sites is essential for ECR9 activity in OL . Considering that transgene expression in spinal cord neurons and cartilage was observed similarly in the absence or presence of Sox10 binding sites and that the two cell types express little Sox10 [2] , [15] , it is unlikely that this ectopic expression is Sox10-dependent . Ectopic expression in peripheral glia may , however , rely on Sox10 as it was only observed in the transgene with intact Sox10 binding sites . This would further support the conclusion that ECR9 is Sox10-responsive . However , it would also argue that ECR9 in the context of the transgene is not able to reproduce the differential aspects of Myrf activation in OL as opposed to Schwann cells . Considering that both Sox10 and Myrf are required for terminal differentiation and myelination [16] , [19] , [20] , [23] they may cooperate once Myrf has been induced by Sox10 . To address this issue we performed immunoprecipitation experiments with anti-Sox10 antibodies on extracts of the OL-like OLN93 cell line . Sox10 was readily precipitated from OLN93 extracts as shown by western blot ( Figure 7A ) . The precipitate also contained a protein that reacted with an antibody specifically directed against the carboxyterminal region of Myrf arguing that both proteins interact at physiological concentrations . To confirm this result and map the interacting regions within both proteins , we first generated polypeptides corresponding to defined regions of the major Myrf isoforms ( Figure 7B ) and analyzed them for their ability to interact with fusion proteins between GST and several conserved Sox10 regions including the dimerization and HMG-domain , the central K2 domain and the carboxyterminal transactivation domain . These GST-pulldown experiments revealed that only the region encompassing dimerization and adjacent HMG domain of Sox10 interacted with Myrf ( Figure 7C ) . For Myrf , interaction with Sox10 was restricted within the carboxyterminal part to an approximately 90 amino acid long segment , 178 to 89 residues away from the carboxyterminus . To analyze whether physical interaction is paralleled by functional cooperation we performed transient transfections in N2a neuroblastoma cells with luciferase reporters under control of regulatory regions from several genes expressed in differentiating OL ( Figure 7D–I ) . These included the Connexin-47 ( Cx-47 ) 1b promoter , the Connexin-32 ( Cx-32 ) P2 promoter , the Myelin-associated glycoprotein gene ( Mag ) promoter , a 3 kb upstream region of the Mbp gene as well as the intronic WmN1 enhancer of the Plp gene and a recently identified Myrf-responsive region 17 kb upstream from the Mbp gene [8] , [18] , [28] , [32]–[34] . N2a cells lack endogenous Sox10 and Myrf . All regulatory regions were strongly activated by co-transfected Sox10 , while Myrf exerted at most a mild stimulatory effect under the assay conditions . Interestingly , the combination of Sox10 and Myrf led to a dramatic synergistic activation of the Cx-47 1b , the Cx-32 P2 and the Mag promoter ( Figure 7D–F ) . Other regulatory regions such as the Plp WmN1 enhancer and the 3 kb Mbp upstream region did not exhibit a synergistic response in the presence of Sox10 and Myrf ( Figure 7G , H ) . However , this does not mean that the corresponding genes are not jointly regulated by the two transcription factors as synergism may be mediated by other regulatory elements of the gene . In accord with such an assumption , the Mbp gene contains an evolutionary conserved element 17 kb upstream of the transcriptional start that has recently been identified as Myrf binding [34] . In contrast to the 3 kb upstream region , this region was synergistically activated by both Myrf and Sox10 ( Figure 7I ) . This provides evidence that the two transcription factors cooperate in the activation of at least some myelin genes that occur in OL , partly through known and partly through novel regulatory regions . Preliminary experiments did not yield any evidence , that Myrf is capable of potentiating Sox10 activity on ECR9 ( data not shown ) . In contrast to the OL-specific regulatory regions , the Mpz promoter and the MSE enhancer of the Krox20 gene as Schwann cell-specific regulatory regions [6] , [29] were even reduced in their activity by Myrf , and were less Sox10-responsive in the presence of Myrf ( Figure 7J , K ) . Whether this means that Myrf actively represses the expression of Schwann cell genes in OL , remains to be studied in future experiments . In any case , it supports the notion of Myrf as an activator of OL-specific gene expression .
While previous studies on the role of Sox10 in oligodendroglial development were restricted to prenatal stages [16] , [17] , [23] , CNS-specific deletion of a conditional Sox10 allele for the first time allowed a closer look postnatally . It confirmed our previous assumption that Sox10 is required for terminal OL differentiation and CNS myelination , and proved that these processes are not only delayed , but permanently disrupted in the absence of Sox10 . Interestingly , while some residual myelin gene expression was observed , there was no myelin formation . The residual myelin gene expression in the absence of Sox10 was completely lost when Sox8 was additionally deleted . Again this confirms previous findings that the two closely related Sox proteins participate in OL development and perform partially redundant functions [23] . Previous findings [22] had already established that the impact of Sox10 on OL development is stronger than the impact of Sox8 , and that both proteins have similar , but not identical functions . This agrees with our finding that Sox8 allows some minor degree of myelin gene expression in the absence of Sox10 , but no myelin formation . In Sox10ΔCNS mice Sox8 cannot sustain residual myelin gene expression at uniform levels in all oligodendroglial cells but rather allows fairly normal expression levels in a small subset . These cells do not contain higher amounts of Sox8 than adjacent ones which fail to turn on myelin genes ( data not shown ) arguing that the residual myelin gene expression in Sox10ΔCNS mice is not due to stochastic fluctuations of Sox8 amounts in oligodendroglia . It confirms previous findings that a replacement of Sox10 by Sox8 coding sequences fails to rescue terminal oligodendroglial differentiation in mice [22] , and argues for a heterogeneity among spinal cord OL with the majority mainly relying on Sox10 , and a minority being Sox8-dependent . Our present study also shows for the first time that Sox10 not only acts as a direct regulator of myelin gene expression during OL differentiation [16] , but is involved in initiating differentiation by inducing expression of Myrf as the central regulator of OL myelination [19] . Considering that Sox10 is expressed at all stages of oligodendroglial development this induction cannot be triggered by Sox10 alone . Most certainly it additionally requires cell intrinsic or extrinsic signals . These signals either lead to induction of transcription factors that then cooperate with Sox10 during the process , or alter Sox10 activities directly , for instance through posttranslational modifications . A candidate for a cooperating transcription factor is Nkx2 . 2 which is induced in promyelinating OL [35] . Myrf induction will likely involve one of the Olig proteins as major regulators of OL development as well ( for review , see [36] ) . Putative posttranslational modifications include sumoylation , phosphorylation and acetylation which all occur at multiple sites in Sox10 [37] , [38] ( data not shown ) . Whatever the underlying trigger , Sox10-dependent induction is direct , and at least in part mediated by a conserved intronic enhancer of the Myrf gene which we named ECR9 . We do not want to imply that ECR9 is the only oligodendroglial enhancer of the Myrf gene or that Sox10 acts solely through ECR9 , but evidence for the involvement of ECR9 in Sox10-dependent Myrf activation is manifold . It is bound in oligodendroglial cell lines and spinal cord by Sox10 . It increases expression from minimal promoters in oligodendroglial cells and responds in its activity to the presence of Sox10 . It furthermore contains a monomeric and a dimeric binding site for Sox10 that strongly contribute to enhancer activity . Most importantly , ECR9 is capable of directing reporter gene expression to differentiating OL in transgenic animals , again dependent on the presence of intact Sox binding sites . The fact that ECR9 exhibited substantial variability in its level of activity in transgenic mice and induced reporter gene expression at sites where Myrf is not normally expressed , may indicate that the enhancer reaches full and specific activity only in its normal genomic context or is only one of several critical enhancers for oligodendroglial Myrf expression . The fact that ECR9 contains both dimeric and monomeric Sox10 binding sites is not uncommon and has been observed for other Sox10-regulated enhancers ( for review , see [26] ) . Considering that Sox proteins act as architectural proteins [39] and that Sox10 dimers and monomers alter DNA topology differently [40] , type and location of each binding site are likely of functional importance . Once induced , Myrf interacts physically and functionally with Sox10 . The physical interaction involves a region in the carboxyterminal part of Myrf . In Sox10 , DNA-binding HMG-domain and preceding dimerization domain are involved in the interaction . This region has previously been identified as a hotspot for interactions with other proteins including transcription factors and chromatin remodeling complexes [25] , [41] . It is currently unknown to what extent these interactions can occur simultaneously or are mutually exclusive . There is also functional interaction between the two proteins . At least some regulatory regions from myelin genes are synergistically activated by the combination of Sox10 and Myrf . Others , however , are not , arguing that Sox10 does not necessarily work together with Myrf on every relevant regulatory region . It also argues against a model in which Myrf is simply recruited by its physical interaction with Sox10 . Myrf has a DNA-binding domain of its own , and its preferential binding site has recently been determined bioinformatically , but not yet confirmed by EMSA [34] . From all currently available data a scenario appears likely in which both Sox10 and Myrf have to bind separately and directly to the regulatory regions they jointly activate . Based on our analyses , we come up with a model in which Sox10 induces Myrf in the promyelinating OL and then teams up with Myrf to jointly activate the expression of myelin genes and other essential components of the terminal differentiation and myelination programs . Sox10 thus performs its function during OL development by a positive feed-forward mechanism . A comparable mode of action exists for Sox10 in differentiating Schwann cells of the PNS . Here Sox10 is first needed to induce Oct6 expression [3] , [5] before both transcription factors cooperate to induce Krox20 [6] . Krox20 and Sox10 then jointly coordinate terminal differentiation and myelination in Schwann cells [7] , [8] . The Sox10-Krox20 regulatory circuit in Schwann cells is thus functionally analogous to the Sox10-Myrf circuit identified here in OL . This argues that terminal differentiation of PNS and CNS glia and myelination is regulated along similar principles with both processes relying on Sox10 and feed forward mechanisms . However , the exact transcription factors that are induced by Sox10 and then cooperate with Sox10 in the differentiation program are different . In fact , they do not even share structural similarities . The functional equivalent of the Schwann cell differentiation factor Krox20 – a zinc finger protein – is Myrf , a factor with a Ndt80 DNA-binding domain in OL . This divergence of key components argues that both gene regulatory networks have been constructed independently around Sox10 as the common denominator and constitute the result of convergent evolutionary processes . Together with differences in the mode of myelination and with distinct ontogenetic origins of Schwann cells and OL from neural crest and neuroepithelial precursors , respectively , our findings give some support for a model in which the ability to myelinate arose separately in these two glial cell types of the CNS and PNS .
Several ECR from the Myrf genomic region ( see Figure 4G ) between positions 10 , 244 , 770 and 10 , 207 , 852 of mouse chromosome 19 ( mouse genome version mm10 ) were amplified by PCR and inserted as SacI/XhoI ( ECR2; ECR6; ECR7; ECR8; ECR9; ECR10; ECR11/12; ECR13 ) , EcoRV/XhoI ( ECR1 ) , or NheI/XhoI ( ECR4/5 ) fragments upstream of the Hsp68 minimal promoter ( positions −104 to +229 relative to the transcriptional start site ) into hsp68-luc . ECR positions relative to the transcriptional start site of the Myrf gene are as follows: ECR1 ( −4022 to −3639 ) , ECR2 ( −3332 to −2869 ) , ECR3 ( −2850 to −2131 ) , ECR4/5 ( −1141 to −144 ) , ECR6 ( +681 to +1376 ) , ECR7 ( +1966 to +2872 ) , ECR8 ( +4536 to +5045 ) , ECR9 ( +7620 to + 8432 ) , ECR10 ( +9081 to +9766 ) , ECR11/12 ( +10089 to +10885 ) and ECR13 ( +32523 to +32896 ) . Like hsp68-luc , the myrf-luc reporter plasmid was based on pGL4 . 10 ( Promega ) and contained the Myrf minimal promoter ( positions −309 to +61 ) . It was used to combine ECR9 in wildtype ( ECR9wt_myrf-luc ) or mutant versions ( ECR9 . 3bc_myrf-luc , ECR9 . 4a_myrf-luc , ECR9mt_myrf-luc ) with its own gene promoter ( Figure 5 ) . Other reporter plasmids used in this study carried the luciferase reporter under control of the Mbp 3 kb upstream region [16] , the Cx-47 1b promoter [18] , the Cx-32 P2 promoter [32] , the Mag promoter ( positions −614 to +12 , generated by PCR ) , the Mpz promoter [29] , the Plp WmN1 enhancer ( positions +3673 to +4844 , generated by PCR ) , the Krox20 MSE enhancer [42] and an evolutionary conserved region 17 kb upstream of the Mbp gene ( positions −17664 to −17033 , generated by PCR ) , with all enhancers being combined with the β-globin minimal promoter . The eukaryotic expression plasmids for Sox10 ( pCMV5-Sox10 and pCAGGS-Sox10-IRES-nls-GFP ) and carboxyterminally truncated Sox8 ( pCMV5-Sox8ΔC ) have been described [15] , [23] , [25] . Eukaryotic pCMV5-based expression plasmids were also generated for Myrf and several Myrf fragments ( see Figure 7B ) . Myrf fragments were fused to an aminoterminal Myc epitope . For knockdown experiments pSuper-Neo-GFP plasmids ( Oligoengine ) were used that expressed a Sox10-specific shRNA ( targeted region: 5′-CTGCTGTTCCTTCTTGACCTTGCCC-3′ ) or a scrambled control shRNA . ECR9-lacZ transgenes contained the ECR9 fragment in wildtype or mutant version ( Figure 6A ) in front of the Hsp68 minimal promoter followed by a lacZ reporter gene and polyA cassette . For conditional Sox10 deletion , Sox10loxP allele [3] and Brn4::Cre transgene ( bcre-32 line ) [21] were combined in Sox10ΔCNS mice . Sox10ΔCNS Sox8lacZ/lacZ mice additionally lacked Sox8 [43] . All mice were on a mixed C3H×C57Bl/6J background . Genotyping was performed by PCR [3] , [43] . Mice transgenic for the ECR9-lacZ transgene in wildtype or mutant form ( ECR9wt-lacZ and ECR9mt-lacZ , see Figure 6A ) were obtained by microinjecting a 4 . 7 kb NotI/XhoI fragment into the male pronucleus of fertilized F1 ( C57BL/6×CBA ) oocytes according to standard techniques . Pups were killed at P7 and genotyped by PCR using 5′-CCTGGCTTGAGTGTTCTGGT-3′ and 5′-AGTAGCTGTCAGCGTCTGGT-3′ as primers . In ovo electroporations were carried out on chicken embryos at HH stage 10–11 after injection of expression plasmids into the neural tube . Conditions for electroporation and procedures to obtain , process , and analyze material 24 h post electroporation at HH stage 19–20 were as described [25] . Material from staged mouse embryos to P21 mice and from electroporated chicken embryos was processed for light and transmission electron microscopy [3] , X-gal staining [16] , in situ hybridization with probes specific for Pdgfra , Mbp , Plp or Myrf , or for immunohistochemistry using primary antibodies against Sox10 , Sox9 , Sox8 , Olig2 , Myrf , Nkx2 . 2 , Pdgfra , CC1 , CNPase , Gfap , NeuN , GFP and β-galactosidase . With exception of Myrf and CNPase antibodies , source and working concentration of primary antibodies as well as fluorophore-labelled secondary antibodies were as described [17] . Antibodies against Myrf were generated in rabbit against a bacterially produced peptide spanning amino acids 1–45 and 248–386 in the aminoterminal part of Myrf ( NCBI accession number Q3UR85 . 2 ) . Antibodies against CNPase were from NeoMarker and used in a 1∶250 dilution . Nuclei were counterstained with DAPI . Spinal cord tissue was also used to prepare RNA [3] and sheared cross-linked chromatin [25] . RNA samples from mouse spinal cord were reverse transcribed and used to analyze expression levels by qPCR on a Biorad CFX96 Real Time PCR System . The following primer pairs were used: 5′-ACACAAGAACTACCCACTACGG-3′ and 5′-GGGTGTACGAGGTGTCACAA-3′ for Mbp , 5′-CCTGTGTCCGTGGTACTGTG-3′ and 5′-TCACACAGGCGGTAGAAGTG-3′ for Myrf , 5′-CTGCCTTGCTGATGCTGC GAGA-3′ and 5′-CCTGCCTGGGTCTGCTTGGG-3′ for Zfp488 , 5′-ACGACAGCAGCGACAACCCC-3′ and 5′-GCTTCCGCTTCTTGCCTGCG-3′ for Nkx2 . 2 , 5′-GAAGCAGATGACTGAGCCCGAG-3′ and 5′-CCCGTAGATCTCGCTCA CCAG-3′ for Olig2 , 5′-ACAGAGACTGAGCGCTGACA-3′ and 5′-CTCGATGGTCTCGTCCTCTC-3′ for Pdgfra , 5′-GCCGGTTTCTCTCCGTCGGC-3′ and 5′-GTTGAGAGTGCCGGGGCCTT-3′ for Zfp191 , 5′-AACGATGATGAGGACC TGAC-3′ and 5′-CAGCTTTCGGGTTCAGATTC-3′ for Tcf4 , 5′-GGTCACCATGTGGTCCTCGGATG-3′ and 5′-AGGGTCTGAGAGGTCAATGC CAGG-3′ for Yy1 , 5′-CCCAAGTACCGTGGCGGTGG-3′ and 5′-GCGGCGAAGGCTTTGCTGTG-3′ for Hes5 , 5′-CCCGGTGGACGACCCGATG-3′ and 5′-CAGATGCCTGCAAGGACAGGATGC-3′ for Id2 , and 5′-AGGCGGTGAGCCCGGT-3′ and 5′-CGGCCGGGTCAGTGTTGAGC-3′ for Id4 . Transcript levels were normalized to β-actin . Rat primary oligodendroglial cells were obtained by differential shake-off from mixed glial cultures [44] and grown in Sato proliferation or differentiation medium [45] . Transfection of primary oligodendroglial cells with pSuper-Neo-GFP-based expression vectors was by Amaxa nucleofection . 24 h post-transfection , cells were fixed and underwent immunochemical staining using the antibodies described above . HEK293 , 33B , OLN93 , S16 and N2a cells were maintained in DMEM containing 10% FCS . S16 cells were used for immunocytochemistry , HEK293 and OLN93 cells for extract preparation , and N2A and 33B cells for luciferase assays . In case cells were transfected , extract preparation was 48 h post-transfection [18] . HEK293 cells were transfected using polyethylenimine ( PEI ) and 10 µg pCMV5-based Sox10 expression plasmid per 100-mm plate . After verification of ectopic expression by western blotting , EMSA were performed [46] using 32P-labeled 25–26 bp double-stranded oligonucleotides containing putative Sox10 binding sites . For luciferase reporter gene assays , 33B and N2a cells were used . 33B cells were transfected with PEI on 24-well tissue culture plates using 500 ng of luciferase reporter and 100 ng of pCMV5- or pSuper-Neo-GFP-based expression vectors . N2a cells were transfected with Superfect ( Qiagen ) on 35 mm plates using 1 µg of luciferase reporter and 500 ng of pCMV5-based expression plasmids . Cells were generally harvested 48 h post-transfection except for knockdown experiments where analysis took place 72 h post-transfection . Luciferase activity was determined in the presence of luciferin substrate by detection of chemiluminescence . For GST pulldown assays , Myc-tagged Myrf fragments were produced in transfected HEK293 cells and used as whole cell extracts . Sox10 fragments were made in E . coli BL21 DE3 pLysS bacteria as GST fusion proteins and purified by affinity chromatography on glutathione-sepharose 4B beads . The bead-bound GST or GST fusion proteins were incubated with HEK293 cell extracts for 2 hours . For co-immunoprecipitation , OLN93 cell extracts were incubated overnight at 4°C with rabbit antiserum against Sox10 ( 1∶3000 dilution , [47] ) or control preimmune serum , and protein A sepharose CL-4B beads ( GE Healthcare ) . Bead-bound proteins in GST-pulldown assays and co-immunoprecipitations were eluted after repeated cycles of centrifugation and washing , and size fractionated on polyacrylamide-SDS gels . Detection was by western blot using mouse antibodies directed against the Myc epitope tag ( 1∶10000 dilution , Novagen ) , and rabbit antisera against Sox10 ( 1∶3000 dilution ) , or the carboxyterminal part of Myrf ( 1∶1000 dilution , generated against amino acids 852–1007 according to NCBI accession number Q3UR85 . 2 ) . Chromatin was prepared from 33B cells , primary oligodendroglial cells and spinal cord from P14 wildtype and Sox10ΔCNS mice after trypsinization , cross-linking of endogenous proteins to DNA and shearing [25] . Immunoprecipitation was overnight at 4°C using Sox10-specific and Sox8-specific antibodies or control immunoglobulins in the presence of pretreated protein A sepharose CL-4B beads . After washing , crosslink reversal , proteinase K treatment , phenol/chloroform extraction and ethanol precipitation , the amount of DNA from input and precipitated chromatin was quantified by qPCR using the Biorad CFX96 Real Time PCR system [25] . To detect various regions of the Myrf gene the following primer pairs were used: 5′-GGGTCTGGTATTCGTAGGTC-3′ and 5′-GTGTTCTCTGTACCTCTTGG-3′ to amplify positions -9415 to -9602 relative to the transcriptional start site of the Myrf gene in rat and 5′- GTAAGTGGGTCTCTGTGTGC-3′ and 5′- GTGGGTTCAGAATCTGCATAG -3′ to amplify positions -9392 to -9581 in mouse ( dagla ) , 5′- GAGTCCCAGAGTCTAGTAGG -3′ and 5′- CTGGTGTCCTGCCATGTCAG -3′ to amplify positions -3973 to -4152 in rat and 5′- GACAATGTGAATACCCAGTC -3′ and 5′- CCTGATGAACTGACAAGATG -3′ to amplify positions -2682 to -2891 in mouse ( fl5 ) , 5′- GTATTGTGCTAGGCCTGCAC -3′ and 5′- GGCAGAAGAAGGCAGTTCTC -3′ to amplify positions - 481 to - 654 in rat and 5′- GCCTTGCCTTAAAGTCTGTG -3′ and 5′- CTCCCTAACAAAGACCTGCC -3′ to amplify positions - 149 to - 364 in mouse ( myrf ) , 5′- CACGTGGCTGACGGGATTTC -3′ and 5′- CCACAGCTGTGGCTGCTGGC -3′ to amplify positions +7969 to +8136 in rat and positions +7586 to +8025 in mouse ( ECR9 ) , 5′- GCATTTGAAGAATGCTGAGCC -3′ and 5′- GACAGACTGACCATGTACAGC -3′ to amplify positions +10463 to +10666 in rat and 5′- GTCCAGGGCTTCTGATCATG -3′ and 5′- GGTCCTTCCTGCCTCAGTGG -3′ to amplify positions +10677 to +10885 in mouse ( ECR11/12 ) , 5′- CTATGCACACCTCTTGCCAC -3′ and 5′- GAGCCCATTGTTCTAAGAGAC -3′ to amplify positions +35839 to +36087 in rat and 5′- CCATGCCTACTTCTGCGTTG -3′ and 5′- GGAATTCCTTGCCACCACAC -3′ to amplify positions +32539 to +32737 in ( fl3 ) . Mice experiments were in accord with animal welfare laws and approved by the responsible local committees and government bodies . | In recent years it has become clear that complex developmental processes are not regulated by single transcription factors but rather by combinations of transcription factors that interact in intricate regulatory networks . Here , we analyze the regulatory network that drives terminal differentiation of oligodendrocytes , the cells of the vertebrate central nervous system that form myelin and thereby guarantee rapid saltatory conduction . We show that the transcription factor Myrf is directly activated by the transcription factor Sox10 , and map a Sox10-responsive enhancer to an evolutionarily conserved element in intron 1 of the Myrf gene . We then go on to show that once induced , Myrf physically interacts and functionally cooperates with its inducer Sox10 to activate myelin genes arguing that the two jointly drive terminal differentiation of oligodendrocytes . With this study we define an essential module in the myelin-specific regulatory network in the central nervous system . By comparing this module with the corresponding module in Schwann cells of the peripheral nervous system which consists of Sox10 and the Krox20 transcription factor we furthermore conclude that myelination in the two compartments of the vertebrate nervous system is regulated by similarly organized , but differentially implemented regulatory networks . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | The Transcription Factors Sox10 and Myrf Define an Essential Regulatory Network Module in Differentiating Oligodendrocytes |
Genome-wide association studies and follow-up meta-analyses in Crohn's disease ( CD ) and ulcerative colitis ( UC ) have recently identified 163 disease-associated loci that meet genome-wide significance for these two inflammatory bowel diseases ( IBD ) . These discoveries have already had a tremendous impact on our understanding of the genetic architecture of these diseases and have directed functional studies that have revealed some of the biological functions that are important to IBD ( e . g . autophagy ) . Nonetheless , these loci can only explain a small proportion of disease variance ( ∼14% in CD and 7 . 5% in UC ) , suggesting that not only are additional loci to be found but that the known loci may contain high effect rare risk variants that have gone undetected by GWAS . To test this , we have used a targeted sequencing approach in 200 UC cases and 150 healthy controls ( HC ) , all of French Canadian descent , to study 55 genes in regions associated with UC . We performed follow-up genotyping of 42 rare non-synonymous variants in independent case-control cohorts ( totaling 14 , 435 UC cases and 20 , 204 HC ) . Our results confirmed significant association to rare non-synonymous coding variants in both IL23R and CARD9 , previously identified from sequencing of CD loci , as well as identified a novel association in RNF186 . With the exception of CARD9 ( OR = 0 . 39 ) , the rare non-synonymous variants identified were of moderate effect ( OR = 1 . 49 for RNF186 and OR = 0 . 79 for IL23R ) . RNF186 encodes a protein with a RING domain having predicted E3 ubiquitin-protein ligase activity and two transmembrane domains . Importantly , the disease-coding variant is located in the ubiquitin ligase domain . Finally , our results suggest that rare variants in genes identified by genome-wide association in UC are unlikely to contribute significantly to the overall variance for the disease . Rather , these are expected to help focus functional studies of the corresponding disease loci .
Inflammatory bowel diseases ( IBDs ) are classified as chronic relapsing inflammatory diseases of the gastrointestinal tract . The two major forms of IBDs are Crohn's disease ( CD , OMIM 266600 ) and ulcerative colitis ( UC , OMIM 191390 ) . Both genetic and environment factors play a central role in the pathogenesis of the inflammatory response of IBDs [1] . Recent genome-wide association ( GWA ) studies and meta-analyses in IBD have shown great success , with the identification of 163 independent IBD risk loci . While some loci were shown to be specific to either CD or UC risk , most have been shown to impact on both diseases , supporting earlier claims that these diseases share genetic risk factors [2] . These recent studies have identified important disease pathways but the common SNPs identified , with generally modest effects , explain only 14% and 7 . 5% of disease variance for CD and UC , respectively [3] . Due to linkage disequilibrium in the genome and limitations of GWAS chip designs to date , genome-wide scans typically identify common variants that tag regions of variable sizes containing multiple candidate genes for disease susceptibility . Although there have been a few notable exceptions , most of the common associated SNPs do not clearly identify causal variants , and further studies are needed to highlight the causal gene in many associated regions [4]–[6] . Sequencing of exons within associated regions in order to identify rare variants with strong effect on disease has been proposed as a means to help identify the causal genes and to help explain a further portion of disease variance . We have recently performed a pooled next-generation sequencing study in Crohn's disease , and identified association to novel low-frequency and rare protein altering variants in NOD2 , IL23R , and CARD9 , as well as IL18RAP , CUL2 , C1orf106 , PTPN22 and MUC19 [7] . We opted to use a similar targeted pooled next-generation sequencing approach to study UC-associated regions from our recent meta-analysis of 3 independent genome-wide scans for UC [8] . Using this approach we identified putative causal variants significantly associated to UC in three of the 22 loci examined and identified variants of interest for an additional six loci .
We selected 200 ulcerative colitis cases and 150 healthy controls of French Canadian ancestry from among samples collected by the NIDDK IBD Genetics Consortium . Samples were pooled in batches of 50 cases or 50 controls and normalized in order for the DNA pool to reflect sample allele frequencies . We targeted 55 genes from 14 UC-associated regions , as well as 7 regions identified in CD showing nominal replication in our UC GWAS study and an additional candidate gene ( ECM1 ) reported in recent literature [6] , [8]–[10] . PCR amplification primers were successfully designed to capture a total of 508 amplicons for a total of 305 Kb or 70% of our original target sequences . Of these 508 PCR reactions , 472 ( 93% ) successfully amplified in each of the 7 sample pools and we used these to construct libraries for high-throughput sequencing on an Illumina Genome Analyzer II . This sequencing yielded large amounts of high-quality data for each pool , that captured 99% of our amplified target regions ( 283 Kb total; 117 Kb exonic sequences ) and achieved 1575× median coverage per pool ( corresponding to 31 . 5× per sample ) . We used the previously described variant calling method Syzygy , designed to accommodate pooled study designs , to identify rare and low-frequency single nucleotide variants in our pooled samples [7] . Syzygy detected 1590 high confidence variants in our target regions , including 309 coding region variants ( 189 missense , 114 synonymous , 2 nonsense and 4 essential splice junction variants ) with 56% of these already reported in dbSNP version 132 , a non-synonymous/synonymous ratio of 1 . 7 and a transition/transversion ratio of 2 . 38 ( Table S1 ) . These results are similar to those obtained from our recent re-sequencing study in CD , as well as those reported by the 1000 Genomes Project , and are indicative of a relatively high true-positive rate for our dataset . This was confirmed by genotyping the 350 discovery DNA samples for a random subset of 237 variants from the total of 1590 high quality variants ( Table S2 ) . After removal of variants that did not validate , variants observed only once in our sequencing dataset ( singletons ) and variants from the MHC region , 84 non-synonymous coding variants ( missense , non-sense and splicing variants ) , were used for subsequent analyses . Following removal of common variants ( frequency >5% ) and variants that did not design in our genotyping assays , we carried out follow-up genotyping for 42 of these variants . Genotyping was performed in 6 independent case-control cohorts totaling 7 , 292 UC cases and 8 , 018 HC ( Table S3 ) , and additional data was obtained for 7 , 143 UC cases and 12 , 186 HC from the International IBD Genetics Consortium ( IIBDGC ) Immunochip project for 14 of these variants [3] . Since our study focuses on infrequent and rare variants , we expect few non-reference alleles for these variants in each subcohort studied , which precludes the use of asymptotic statistics utilized in typical association studies of common variants . Also , given the low frequencies of the variants tested , population structure is likely to be a more substantial problem and thus requires a stratified analysis with strict population case-control matching . We used a previously described mega-analysis of rare variants ( MARV ) approach that provides a permutation-based estimate of significance , within each sub-cohort , and accommodates variable numbers of case-control samples in each independent population for single-marker analysis [7] . With a target set of 42 variants we can define a traditional corrected significance level of P = 0 . 0012 for our study . Three variants , located in the CARD9 , IL23R and RNF186 genes , reach this significance threshold suggesting that these could possibly be the causal genes/variants within these two loci ( Table 1 ) . Specifically , our results show that the c . IVS11+1G>C CARD9 splice variant confers significant protection to UC ( P = 1 . 47×10−11; OR = 0 . 39 [0 . 30–0 . 53] ) . We previously identified this splice variant in a sequencing project of CD loci and demonstrated that it leads to an alternatively spliced transcript that is missing exon 11 [7] . Our results also identify significant association to the valine to isoleucine substitution at position 362 ( Val362Ile ) in IL23R ( P = 1 . 18×10−03; OR = 0 . 79 [0 . 68–0 . 91] ) previously reported by a recent re-sequencing of positional candidates in Crohn's disease [7] , [11] . The significantly associated rare variant that we identified in RNF186 ( P = 8 . 69×10−4; OR = 1 . 49 [1 . 17–1 . 90] ) encodes an alanine to threonine substitution at position 64 ( Ala64Thr ) . RNF186 encodes a protein with a RING domain and two transmembrane domains . Importantly , the disease-coding variant is located in the RING domain , a domain with a predicted E3 ubiquitin-protein ligase activity ( Fig . 1 ) . Independence of effect between rare variants in IL23R and CARD9 and the reported common association signals in these genes has previously been shown [7] , [11] . For RNF186 , the Ala64Thr variant is mostly found on the protective haplotype background from the previously identified common variant , indicating that the reported association is not likely due to partial LD with the common variant . In addition , reciprocal conditional logistic regression analysis , using a subset of samples where both variants were genotyped ( 3548 UC cases and 3607 healthy controls ) shows that these are independent association signals ( data not shown ) . Given the challenge inherent in achieving corrected significance thresholds for rare variants , even with large cohorts , we expect that some of the other variants that we identified and found to have nominal significance ( 0 . 0012<P<0 . 05 ) are truly associated with UC . In fact with a target set of 42 variants included in follow-up genotyping , and supposing these are independent and under the null , we would expect <1 SNP to exceed P<0 . 01 ( with a probability of less than 1% to observe 3 or more associations at this level ) and ∼2 SNPs to exceed P<0 . 05 by chance alone ( with a probability less than 0 . 0001 to observe 9 or more association at this level ) , whereas we observe 3 SNPs with P<0 . 01 and 9 SNPs with P<0 . 05 , suggesting that there are additional true positives that have not met the more stringent threshold . Indeed , within the group of SNPs that we found to have nominal significance are two non-synonymous coding variants ( Gly149Arg and Val362Ile ) in IL23R that we and others have shown to be associated with protection from IBD ( Table 1 ) [7] , [11] . In addition to these previously-validated variants in IL23R , we have found variants that are nominally associated with UC in the genes encoding CEP72 , LAMB1 , CCR6 , JAK2 , and STAC2 ( Table 1 ) . Specifically , we identified two nominally associated rare variants in CEP72 ( Lys314Arg and Asp316Asn ) in perfect LD with each other that appear to protect from UC ( Table 1 ) . As we also sequenced the only other gene in this locus ( TPPP ) , but did not find any associated variants in it , this suggests that CEP72 is potentially causal . Similarly , we sequenced both genes in the LAMB1-DLD locus on chromosome 7 , with the nominally associated rare variant in LAMB1 ( Ile154Thr ) suggesting a role for this gene in risk to UC , especially as the associated allele is located in its DUF287 domain and is predicted to have a damaging effect [12] . All genes within the CCR6-FGFR1OP-RNASET2 locus were sequenced , with a single nominally-associated variant ( Ala369Val ) in CCR6 , consistent with this gene's probable role in the migration and recruitment of dendritic and T cells during inflammatory and immunological responses [13] . Within the JAK2-INSL6-LHX3 locus , we only sequenced JAK2 given its key role in signaling from the IL12R/IL23R , a biological pathway proven to be associated with IBD , and identified a nominally associated variant ( Arg1063His ) within its catalytic domain . STAC2 is within a locus with 16 other genes including ORMDL3 , which has been suggested to be the most likely causal gene based previous genetic and functional studies in IBD and asthma [8] , [14] . Although we find a nominally associated variant in STAC2 ( Lys302Arg ) and none in ORMDL3 , we have only sequenced 10 of the 17 genes within this locus ( Table S4 ) . Studies of each of these variants to determine their functional impact will be essential to prove causality .
Genome-wide association studies in IBD have been very successful in identifying genomic regions associated with CD , UC or both . Only infrequently have these GWA studies also directly identified the causal genes/variants , with NOD2 , IL23R and ATG16L1 being the few known examples . A recent targeted ( exons and exon-intron boundaries ) sequencing approach of known CD loci resulted in the identification of potentially causal variants in eight of the 36 loci examined [7] . The primary objective of the current study therefore was to use the same approach to identify likely causal variants within genes that were located in genomic regions associated with UC . While there are over 100 UC loci that have been identified and validated to date , we examined 22 UC loci that were known at the time of the initiation of this project . Of these 22 loci , the current study identified potentially causal variation in three of the loci: two protective alleles in CARD9 and IL23R , and an allele increasing risk in RNF186 . The identification of a rare variant ( Ala64Thr ) in RNF186 that shows significant association to UC strongly suggests that this is the causal gene within this locus . Importantly , the disease-coding variant is located in the RING domain , a domain with a predicted E3 ubiquitin-protein ligase activity . Ubiquitin ligases have been shown to regulate key adaptors of proinflammatory pathways [15]–[17] . We previously reported that RNF186 expression was higher in human intestinal tissues than in immune tissues [8] . We showed by immunostaining that the RNF186 protein was expressed at the basal pole of epithelial cells and lamina propria within colonic tissues . Using GEO public microarray datasets , we pursued a systematic follow-up analysis of expression profiles of epithelial cells in response to bacterial products , PAMPs/pathogens . We found that RNF186 gene expression was significantly up-regulated in small intestine epithelium and induced by Shigella infection in mice ( P = 4 . 21×10−8 ) ( Figure 1 , Panel A ) [18] , [19] . Both invasive ( INV+ ) and non-invasive ( INV− ) strains of Shigella induced significant overexpression of RNF186 in intestinal tissues of 4-day- and 7-day-old mice infected for 2 or 4 hours . To further identify putative transcriptional regulators of RNF186 expression , we employed a text-mining and network-generating analysis of human protein-protein , protein-DNA , protein-RNA and protein-compound interactions . Specifically , from our analyses we hypothesize that RNF186 is transcriptionally regulated in a two-step process by the transcription factor Hepatocyte Nuclear Factor 4 , alpha ( HNF4A ) ( Figure 1 , Panels B , C ) . Several studies have shown that HNF4A binds to the promoter region and up-regulates the expression of yet another transcription factor HNF1A [20]–[22] . Knockdown of HNF4A has been shown to down-regulate HNF1A gene expression [23] , [24] . HNF1A , in turn , regulates RNF186 and this interaction has been confirmed by chromatin immunoprecipitation and chip-on-chip assay [25]–[27] . Our own analysis of transcriptional profiles of HNF4A-Null colons recovered from HNF4AloxP/loxPFoxa3Cre and HNF4AloxP/−Foxa3Cre mice uncovered a significant up-regulation of RNF186 transcript [28] . Expression profiling of human tissues also supports this hypothesis , as HNF4A and RNF186 are clearly co-expressed in the small intestine and the colon ( Figure S1 ) . This putative interaction is particularly relevant given that HNF4A has previously been shown to be associated , with genome-wide significance , with risk to developing UC [9] . Our analysis now indicates a direct genetic interaction between two IBD susceptibility genes namely , HNF4A and RNF186 . While a singular loss-of-function mutation in HNF4A has already been shown to be associated with susceptibility to abnormal intestinal permeability , inflammation and oxidative stress , we speculate that a dual loss-of-function with additional mutation in RNF186 would further exacerbate one's susceptibility to develop chronic inflammation in the gut [29] , [30] . In addition to the variants in IL23R , CARD9 , and RNF186 , we also identified variants of interest in an additional five loci ( specifically within the CEP72 , LAMB1 , CCR6 , JAK2 , and STAC2 genes ) . While these latter six still require confirmation , we estimate that many will validate given that we observed an excess of nominally-associated variants . Examining the data from the current study along with the data derived from prior association and sequencing studies suggests that at a minimum , there currently is strong evidence of association to causal variation in IBD ( i . e . missense , nonsense or splice junction variants ) in the NOD2 , ATG16L1 , IL23R , MST1 , CARD9 , IL18RAP and RNF186 genes , and at least suggestive evidence for causal variation in the CUL2 , C1orf106 , PTPN22 , MUC19 , CEP72 , LAMB1 , CCR6 , JAK2 , and STAC2 genes ( Current study and references [4] , [5] , [7] , [11] , [31] ) . While only a small fraction of the recently identified 163 IBD loci have been sequenced ( 36 CD , 22 UC for total of 42 independent loci ) in IBD patients and controls , this would suggest that from ∼10% ( 15 of 163 total loci ) to ∼35% ( 15 of 42 loci sequenced ) of IBD loci have causal variation affecting the protein-coding or splice junctions . There are an additional 5 loci ( ITLN1 , GSDMB , YDGL , SLC22A4 , and FCGR2A ) for which there are non-synonymous coding or splice variants present in public databases ( dbSNP , 1KG ) that are correlated with the index SNP identified in the GWA studies that have yet been tested directly , thus potentially increasing the estimated number of IBD loci with causal variation within the coding and splice regions [3] , [32] . Furthermore , it should be noted that with the exception of a small number of variants with significant effect ( e . g . R702W , G908R , fs107insC in NOD2; R381Q in IL23R; IVS11+1G>C in CARD9; V527L in IL18RAP – all of which had 0 . 5>OR>2 ) most of the rare variants identified by targeted sequencing of loci from GWAS regions have relatively modest effect sizes that are comparable to those observed for the common variants identified by GWA studies . Consequently , very large sample sizes are required to detect statistically significant association . In the current study , for the majority ( 93% ) of variants with an observed minor allele frequency greater than 0 . 3% , we had more than 80% power to detect significant association if the OR is 2 or greater with the number of samples typed ( up to ∼14 , 000 cases and ∼20 , 000 controls ) ( see Table S5 ) . Moreover , should this observation not be limited to risk loci identified by GWA studies , this has implications with respect to future efforts for discovering risk loci . Specifically , if the occurrence of rare variants with large effects sizes is relatively infrequent , then this may favor the current paradigm of locus discovery by GWA followed by targeted sequencing rather than whole-exome or whole-genome sequencing for locus discovery as this would require even larger sample sizes . Alternatively , given the ever- growing size of public databases of common and rare variants , targeted genotyping of known variants within risk loci identified by GWA may prove to be an efficient approach . For example , all but two of the 22 candidate causal variants identified in the current study or that of Rivas and colleagues are now found in the Exome Sequencing Project database . Regardless of the study design , these results suggest that a significant proportion of IBD loci contain causal variants within exons or exon-intron boundaries . While these rare/infrequent variants may not account for what has been termed “the missing heritability” of common traits , discovering these variants certainly can provide focus for follow-up functional studies . For example , the current sequencing and follow-up genotyping of the chromosome 1p36 locus , which was first identified in a GWA study of UC , identified significant association to the Ala64Thr variant within RNF186 . While further studies will be required , the initial bioinformatics and experimental studies described above suggest that this ring finger protein with an ubiquitin-ligase domain may have an important role in the response to microbes/microbial products . Going forward , systematic evaluation of genes within risk loci via expression-driven functional studies in cellular models ( i . e . knock-down or over expression ) with sensitive high throughput/high content readouts may very well be a complementary approach given that at least a third of IBD risk loci appear to act via gene expression [3] .
We selected 200 ulcerative colitis patients and 150 healthy control of French-Canadian descent from the NIDDK IBD Genetics Consortium repository samples . The NIDDK IBDGC samples were collected under rigorous clinical phenotyping and control matching for the purpose of genetic studies [33] . Genomic DNA concentrations were measured by Quant-iT PicoGreen dsDNA reagent ( Invitrogen ) and detected on the Biotek Synergy 2 plate reader . All DNAs were normalized with at least two round of dilution and quantification down to a concentration of 10 ng/µl as described previously [7] . Equimolar amounts of samples were pooled together in batches of 50 cases and 50 controls for a total of 7 pooled groups . Target exonic sequences were selected based on the coding exons of 55 genes in 14 UC-associated regions and 7 regions identified in CD with nominal replication in our recent UC GWAS study , as well ECM1 identified from recent candidate-gene study in UC [6] , [8]–[10] , [34] . Specifically , amplicons were designed from genome build Hg18 using a web-base automated pipeline ( Optimus primer: Website ( http://op . pgx . ca ) ) that uses the Primer 3 design software and user defined parameters [35] . Design parameters included amplicon sizes between 400 and 600 base pairs , as well as the inclusion of Not1 tails for subsequent concatenation and shearing steps in library construction . PCR amplification reactions contained 40 ng of pooled genomic DNA , 1× HotStar buffer , 0 . 8 mM dNTPs , 2 mM MgCl2 , 0 . 4 units of HotStar Enzyme ( Qiagen ) , and 0 . 25 µM forward and reverse primers in a 10-µl reaction volume . PCR cycling parameters were as follows: one cycle of 95°C for 5 min; 30 or 35 cycles of 94°C for 30 s , 60°C for 30 s , and 72°C for 1 min; followed by one cycle of 72°C for 5 min . Each DNA pools were amplified for 508 PCR reactions; amplification products were then dosed by Quant-iT PicoGreen dsDNA reagent ( Invitrogen ) quantification and amplification specificity was validated by agarose gel electrophoresis . In total , 472 PCR amplicons ( 93% amplification success rate , capturing 283 Kb including 117 Kb of target exonic sequences ) ( Table S6 ) for each DNA pool were combined in equimolar amounts to obtain equal representation of all target in library construction . The combined PCR products from each pooled DNA group were concatenated using the NotI adapters and sheared into fragments as previously described [36] . Libraries were constructed according to Illumina single-end library protocol , with 150–200 bp gel size selection and PCR enrichment using 10 cycles of PCR , and then single-end sequenced with 36 cycles on an Illumina Genome Analyzer II . Each sample pool was sequenced using a single lane of Illumina GAII analyzer flowcell; 36-base pair reads were aligned to the genome using MAQ algorithm [37] and base qualities were recalibrated using GATK ( Genome Analysis ToolKit ) [38] . Finally , variant discovery was performed using the previously described Syzygy software , designed to analyze sequencing data from pooled DNA sequencing [7] . We randomly selected 237 high quality variants for validation in our 350 discovery DNAs samples using Sequenom MassARRAY iPlex200 chemistry . Genotyping assay designs were obtained from the Assay Designer v . 3 . 1 software , and genotyping oligonucleotides were synthesized at Integrated DNA Technologies . The correlation coefficient between observed minor allele frequencies and frequencies estimated from Syzygy for validated variants was calculated in order to evaluate the overall quality of our dataset ( Figure S2 ) . Eighty-four high quality non-synonymous coding variants ( missense , nonsense and splicing variants ( within 2 bp of a splice site ) ) remained after the exclusion of singletons from our sequencing results , variants that did not validate and variants within the MHC region . We then evaluated these variants in an independent cohort of North-American individual of European descent from the NIDDK IBD genetics consortium ( 754 cases and 1008 controls ) ; only variants detected in this independent cohort were kept for follow-up genotyping . Following assay design , 42 SNPs were genotyped using Sequenom MassARRAY iPlex200 chemistry in 6 independent follow-up case-control cohorts ( 7292 cases and 8018 controls ) ( Table S3 ) . Because of design constraints and assay failures , not all markers were examined in all follow-up sample sets . For a subset of these variants , further genotyping data was obtained from the International IBD Genetics Consortium Immunochip data ( 7143 UC , 12186 controls ) For all cohorts , UC was diagnosed according to accepted clinical , endoscopic , radiological and histological findings . Genotyping of the NIDDK IBDGC cohort , as well as the Italian and Dutch cohorts was performed at the Laboratory for Genetics and Genomic Medicine of Inflammation ( www . inflammgen . org ) of the Université de Montréal . NIDDK IBD Genetics Consortium ( IBDGC ) samples were recruited by the centers included in the NIDDK IBDGC: Cedars Sinai , Johns Hopkins University , University of Chicago and Yale , University of Montreal , University of Pittsburgh and University of Toronto . Additional samples were obtained from the Queensland Institute for Medical Research , Emory University and the University of Utah . Medical history was collected with standardized NIDDK IBDGC phenotype forms . Healthy controls are defined as those with no personal or family history of IBD . The Italian samples were collected at the S . Giovanni Rotondo “CSS” ( SGRC ) Hospital in Italy . The Dutch cohort is composed of ulcerative colitis cases recruited through the Inflammatory Bowel Disease unit of the University Medical Center Groningen ( Groningen ) , the Academic Medical Center ( Amsterdam ) , the Leiden University Medical Center ( Leiden ) and the Radboud University Medical Center ( Nijmegen ) , and of healthy controls ( n = 804 ) of self-declared European ancestry from volunteers at the University Medical Center ( Utrecht ) . Genotyping of the German cohort was performed at the Institute for Clinical Molecular Biology Christian-Albrechts-University in Kiel . German patients were recruited either at the Department of General Internal Medicine of the Christian-Albrechts-University Kiel , the Charité University Hospital Berlin , through local outpatient services , or nationwide with the support of the German Crohn and Colitis Foundation . German healthy control individuals were obtained from the popgen biobank . Genotyping of Swedish UC cases and controls was performed at Karolinska Institutet's Mutational Analysis core facility ( MAF ) . Swedish ulcerative colitis patients and controls were recruited at the Karolinska University Hospital , Stockholm , and at the Örebro University Hospital , Örebro , Sweden . Genotyping of the Belgian cohort was performed at the Genomics Core Facility at UZ Leuven , using a MassARRAY iPLEX ( Sequenom ) . Belgian patients were all recruited at the IBD unit of the University Hospital Leuven , Belgium; control samples are all unrelated , and without family history of IBD or other immune related disorders . All patients and control subjects provided informed consent . Recruitment protocols and consent forms were approved by Institutional Review Boards at each participating institutions . All DNA samples and data in this study were denominalized . Association analysis of follow-up genotyping data was performed using the previously described mega-analysis of rare variants ( MARV ) approach [7] . Briefly , this method evaluates significance of association from stratified sample , using within sub-cohort permutation of individual phenotypes to provide the test statistic . This approach is robust to population stratification and to deviation from Hardy-Weinburg equilibrium . We downloaded and analyzed several Gene Expression Omnibus ( GEO ) public microarray datasets including: ( a ) Expression data from newborn mice infected with Shigella flexneri; GSE9785 ( b ) Transcription profiles of colon biopsies from UC patients and healthy controls; GSE11223 ( c ) Steady-state gene expression data of Tuberculosis infected human primary dendritic cells; GSE34151 ( d ) PBMC transcriptional profiles in healthy subjects , patients with Crohn's Disease , and patients with Ulcerative Colitis; GSE3365 , ( e ) Transcription profiles of colon biopsies from Crohn's patients and healthy controls; GSE20881 , ( f ) Transcription profile of mouse small intestine epithelium vs . mesenchyme; GSE6383 , ( g ) Gene expression in HNF4 null mouse colons compared to control colons; GSE3116 , and ( h ) Microarray profiles of mouse epithelial colon harboring conditional knock out of HFN4A; GSE11759 . Each of these datasets was normalized using quantile normalization routine in MATLAB . Genes were tested for significant differences between pairs of control and stimulated/treated samples within each experiment . After selecting genes with nominal P<0 . 05 , estimated using an unpaired T-test , expression of RNF186 was evaluated whether it passed the significance threshold or not . The results of processing all these datasets are shown in Table S7 and Figures S3 , S4 , S5 , S6 , S7 , S8 , S9 , S10 , S11 , S12 , S13 , S14 . For transcriptional network analysis , we used Metacore's suite of network building algorithms to expand the sub-network around RNF186 . The algorithm searches through a manually curated knowledgebase of molecular interaction to identify bidirectional connectivity with genes , proteins and small molecules . The search was constrained to expand the overall network size up to 50 components . Given that the bioinformatic analyses suggested that HNF4A controlled the expression of RNF186 , we directly tested for their co-expression in a panel of RNAs from a variety of human tissues . Specifically , expression levels of RNF186 and HNF4A were evaluated using a custom expression array from Agilent , which was designed to include an independent probe for each exon of the genes tested ( Figure S1 ) . Briefly , total RNA from bone marrow , heart , skeletal muscle , uterus , liver , fetal liver , spleen , thymus , thyroid , prostate , brain , lung , small intestine and colon were purchased from Clontech Laboratories . A reference RNA sample was also included that consisted of an equal mix from 10 different human tissues ( adrenal gland , cerebellum , whole brain , heart , liver , prostate , spleen , thymus , colon , bone marrow ) . With the exception of the small intestine ( RIN = 7 . 6 ) , all RNAs had a RNA Integrity Value ( RIN ) value ≥8 ( range 8 . 0–9 . 3 ) as measured by Agilent 2100 Bioanalyzer using the RNA Nano 6000 kit ( Agilent Technologies ) . Labeled cRNA was then synthesized from 50 ng of each RNA sample using the Low Input Quick Amp WT labeling kit ( Agilent Technologies ) according to the manufacturer's protocol . Quantity and quality of labeled cRNA samples were assessed by NanoDrop UV-VIS Spectrophotometer . Sample hybridization was performed according to the manufacturer's standard protocol and microarrays were scanned using the Sure Scan Microarray Scanner ( Agilent technologies ) . An expression value was obtained for each gene in each replicate by calculating the geometric mean of all probes within the gene , followed by a median normalization across all genes on the array . A geometric mean and geometric standard deviation was calculated from at least 3 independent measurements for each tissue . | Genetic studies of common diseases have seen tremendous progress in the last half-decade primarily due to recent technologies that enable a systematic examination of genetic markers across the entire genome in large numbers of patients and healthy controls . The studies , while identifying genomic regions that influence a person's risk for developing disease , often do not pinpoint the actual gene or gene variants that account for this risk ( called a causal gene/variant ) . A prime example of this can be seen with the 163 genetic risk factors that have recently been associated with the chronic inflammatory bowel diseases known as Crohn's disease and ulcerative colitis . For less than a handful of these 163 is the causative change in the genetic code known . The current study used an approach to directly look at the genetic code for a subset of these and identified a causative change in the genetic code for eight risk factors for ulcerative colitis . This finding is particularly important because it directs biological studies to understand the mechanisms that lead to this chronic life-long inflammatory disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Deep Resequencing of GWAS Loci Identifies Rare Variants in CARD9, IL23R and RNF186 That Are Associated with Ulcerative Colitis |
The first reported Far East scarlet-like fever ( FESLF ) epidemic swept the Pacific coastal region of Russia in the late 1950s . Symptoms of the severe infection included erythematous skin rash and desquamation , exanthema , hyperhemic tongue , and a toxic shock syndrome . The term FESLF was coined for the infection because it shares clinical presentations with scarlet fever caused by group A streptococci . The causative agent was later identified as Yersinia pseudotuberculosis , although the range of morbidities was vastly different from classical pseudotuberculosis symptoms . To understand the origin and emergence of the peculiar clinical features of FESLF , we have sequenced the genome of the FESLF-causing strain Y . pseudotuberculosis IP31758 and compared it with that of another Y . pseudotuberculosis strain , IP32953 , which causes classical gastrointestinal symptoms . The unique gene pool of Y pseudotuberculosis IP31758 accounts for more than 260 strain-specific genes and introduces individual physiological capabilities and virulence determinants , with a significant proportion horizontally acquired that likely originated from Enterobacteriaceae and other soil-dwelling bacteria that persist in the same ecological niche . The mobile genome pool includes two novel plasmids phylogenetically unrelated to all currently reported Yersinia plasmids . An icm/dot type IVB secretion system , shared only with the intracellular persisting pathogens of the order Legionellales , was found on the larger plasmid and could contribute to scarlatinoid fever symptoms in patients due to the introduction of immunomodulatory and immunosuppressive capabilities . We determined the common and unique traits resulting from genome evolution and speciation within the genus Yersinia and drew a more accurate species border between Y . pseudotuberculosis and Y . pestis . In contrast to the lack of genetic diversity observed in the evolutionary young descending Y . pestis lineage , the population genetics of Y . pseudotuberculosis is more heterogenous . Both Y . pseudotuberculosis strains IP31758 and the previously sequenced Y . pseudotuberculosis strain IP32953 have evolved by the acquisition of specific plasmids and by the horizontal acquisition and incorporation of different genetic information into the chromosome , which all together or independently seems to potentially impact the phenotypic adaptation of these two strains .
Yersinia pseudotuberculosis is a bacterial pathogen that , with Y . pestis and Y . enterocolitica , causes worldwide infections in humans [1–4] . Y . pseudotuberculosis serotype O:1b is thought to be the direct evolutionary ancestor of Y . pestis , the causative agent of plague [4 , 5] . While these two species diverged from one another within the last 20 , 000 y , the Y . pseudotuberculosis and Y . enterocolitica lineages separated between 0 . 4 and 1 . 9 million y ago [6] . Y . pseudotuberculosis infections in humans are acquired through the gastrointestinal tract by the ingestion of contaminated food products and result in abdominal pain , fever , and occasionally diarrhea . Pathogenicity has been attributed to several key virulence factors , including the plasmid-borne Yersinia outer proteins that are delivered by a type III secretion system , the invasion adhesion molecule ( Inv ) , and the high pathogenicity island ( HPI ) [1] . Often , Y . pseudotuberculosis isolates from environmental and clinical sources harbor various plasmids ranging in size from 3–125 kb [7] , some of which have been linked to pathogenicity [8 , 9] . In 1959 , an epidemic of Y . pseudotuberculosis infections on the Pacific coast of Russia was called Far East scarlet-like fever ( FESLF ) , or scarlatinoid fever [10–17] for its clinical similarities to scarlet fever caused by group A streptococci [18 , 19] . Such atypical infections in Far East Asia are severe , and the clinical presentation includes erythematous skin rash , skin desquamation , exanthema , hyperhemic tongue , and toxic shock syndrome [10 , 11 , 18 , 19] . Y . pseudotuberculosis FESLF symptoms have been linked to the systemic expression of the superantigenic exotoxin Y . pseudotuberculosis–derived mitogen ( YPM ) [20] , as well as the presence of two uncharacterized plasmids , pVM82 and pIB [7 , 8] . Although no plasmid sequence was available , a 37 . 5-kb region of pVM82 was experimentally linked to increased immunosuppressive and antiphagocytic capabilities [21] . Here , we report the whole genome sequence analysis of serotype O:1b Y . pseudotuberculosis IP31758 that was isolated in 1966 from the stools of a patient presenting with FESLF in the Primorski region of the former Soviet Union . Intra- and interspecies comparisons with the genomes of the previously sequenced typical non-FESLF–causing Y . pseudotuberculosis strain IP32953 [22] , all published Y . pestis genomes [23–26] , and the more distantly related Y . enterocolitica strain 8081 [27] were performed in order to identify strain-specific genome characteristics of Y . pseudotuberculosis IP31758 intimately related to the atypical clinical FESLF manifestation . In addition , we tested for their distribution in a panel of geographically and phenotypically diverse Y . pseudotuberculosis and Y . pestis isolates . These analyses resulted in the identification of genetic traits potentially associated with the particular FESLF symptoms and led to a redefined model for the evolutionary history of the group .
The genome of Y . pseudotuberculosis IP31758 consists of a circular chromosome of 4 , 723 , 306 bp ( Figure 1A ) and two novel plasmids called pYpsIP31758 . 1 ( 153 , 140 bp; Figure 1B ) and pYpsIP31758 . 2 ( 58 , 679 bp; Figure 1C ) . Noteworthy , the highly conserved low-calcium response plasmid ( lcr ) pYV encoding the type III secretion apparatus , which can be found in many but not all Y . pseudotuberculosis and Y . enterocolitica isolates [28–30] , was not detected in Y . pseudotuberculosis IP31758 . The general genomic features of Y . pseudotuberculosis IP31758 are summarized and compared with those of Y . pseudotuberculosis IP32953 in Table 1 . Based on the level of sequence-read coverage in each assembly , it is estimated that the chromosome and the two plasmids are present in equal copy numbers . The combination of these two plasmids has not been reported in any other Yersinia strain , and no significant similarity has been found with the known Yersinia plasmid sequences from the public databases ( Table 1 ) [22–26 , 31] . However , the plasmid replication protein RepA ( YpsIP31758_B0136 ) of pYpsIP31758 . 1 displays 42% amino acid identity to the corresponding gene of the cryptic conjugative plasmid pYptb32953 ( pYptb0001 ) of the previously sequenced Y . pseudotuberculosis strain IP32953 [22] . Y . pseudotuberculosis IP31758 large plasmid , pYpsIP31758 . 1 , was identified as virulence plasmid pVM82 , named for its estimated molecular weight ( 82 kDa ) and previously reported in Y . pseudotuberculosis FESLF strains isolated from different areas of the former Soviet Union [7] . The prevalence of this plasmid in FESLF-causing Y . pseudotuberculosis strains and its association to virulence has been experimentally demonstrated [32] . Although no sequence data were available , a HindIII restriction map of pVM82 has been previously published [33] . A thorough comparison of pVM82 HindIII restriction map with that generated in silico from the sequence of pYpsIP31758 . 1 ( 153 , 140 bp ) revealed a few discrepancies in the number of restriction fragments and the order in which those were originally assembled ( Table S1 and Figure 1B ) , both of which could be explained by the insufficient resolution of the initial restriction fragment analysis by gel electrophoresis [33]: ( 1 ) the in silico HindIII digest of pYpsIP31758 . 1 resulted in two additional fragments ( I [1 , 678 bp] and II [7 , 188 bp] ) , the sizes of which were almost identical to other large fragments and hence would have been impossible to distinguish by gel electrophoresis ( fragment P [1 , 651 bp] , fragment I [7 , 057 bp] , and fragment J [7 , 098 bp]; Figure 1B and Table S1 ) ; ( 2 ) a small 44-bp fragment ( III ) was not previously reported; and ( 3 ) the size of the largest restriction fragment , measured at 25 kb , was underestimated and is 31 , 313 bp . Importantly , overall the sequenced plasmid restriction map is in agreement with the published restriction map ( Table S1 ) , including the presence of a 37 . 5-kb region of pVM82 ( fragment F ) , which was experimentally linked to increased immunosuppressive and antiphagocytic capabilities [21] . The updated HindIII restriction map based on the pYpsIP31758 . 1 plasmid sequence is shown as additional information in the outer circle of Figure 1B . Comparative genome sequence analyses between Y . pseudotuberculosis IP32953 and several Y . pestis isolates have shown that Y . pestis has an expanded number of insertion sequence ( IS ) elements [22] . These IS expansions observed in the Y . pestis lineage had a major impacts on the evolutionary process and speciation by introducing multiple recombinatorial hotspots [22] . Such recombinatorial hotspots account for the intrachromosomal rearrangements ( lack of synteny ) as well as the reductive evolution ( deletion of fragments flanked by IS elements and gene loss due to IS interruption ) in the Y . pestis lineage [3 , 25] . While Y . pseudotuberculosis IP31758 contains a greater number of IS elements than Y . pseudotuberculosis IP32953 ( Table 1 ) , both genomes contain far fewer IS elements than Y . pestis . A total of six IS families are present in both Y . pseudotuberculosis genomes , although strain-specific IS element distribution patterns and copy numbers are observed ( Table 1 and Figures 1A and S1 ) , probably resulting from the process of microevolution ( gene loss and acquisition ) as well as intrachromosomal IS duplications and translocations , as shown in Y . pestis [34] . The IS elements IS100 and IS1661 , both found in all sequenced Y . pestis strains and Y . pseudotuberculosis IP32953 [22 , 23] , were not detected in Y . pseudotuberculosis IP31758 ( Table 1 ) . The absence of IS100 has been previously linked in Y . pseudotuberculosis to sensitivity to pesticin and might indicate a more distant evolutionary and ecological relationship to Y . pestis [35] . Unlike the Y . pestis genome sequences , which display fragmented synteny patterns [25] , the two Y . pseudotuberculosis genomes are almost perfectly syntenic and have undergone very little rearrangement ( Figure 2A and 2B ) . A 665-kb inversion encompassing the origin of replication is the only major recombinatorial event that differentiates the two Y . pseudotuberculosis genome sequences as evidenced by the BLAST score ratio analysis ( Figure 2 ) [36] . On the other hand , the synteny at the interspecies level to the genomes of Y . pestis CO92 ( Figure 2C ) and Y . enterocolitica 8081 ( Figure 2D ) is partly resolved [26 , 27] . Similar results were obtained from comparison to all other published Y . pestis genomes . Minor synteny breakpoints are linked to horizontally acquired genomic regions , mainly due to the insertion of prophages , IS elements , and integrons that are specific to each individual Y . pseudotuberculosis strain ( Figure 1 , circle 5 and Figure S1 , circle 5 ) . Sequenced species belonging to the genus Yersinia harbor different types and numbers of restriction/modification ( R/M ) enzyme systems [37] . Noteworthy , our analysis shows that both enteropathogenic Y . pseudotuberculosis strains IP31758 and IP32953 harbor a unique type I R/M system , which is not present in all studied Y . pestis strains , and is composed of three genes , hsdRSM ( YpsIP31758_3536 to YpsIP31758_3538; Table S6 ) . The implications of this R/M system to Y . pseudotuberculosis genome evolution are still unresolved . Genomic rearrangements do not appear to have been facilitated by intrachromosomal recombination , as they are often flanked by undisrupted housekeeping or hypothetical genes and not by mobile elements or paralogous gene families . Our analysis did not reveal an obvious mechanistic basis for these rearrangements . Compared to the lack of genome-wide synteny found within Y . pestis , both sequenced Y . pseudotuberculosis strains IP31758 and IP32953 display a high level of genome conservation , which is emphasized by a high degree of nucleotide ( nt ) sequence identity of more than 95% over 94 . 8% of the length of the two chromosomes . Such level of nt identity , but conversely with poor synteny , is also observed between Y . pestis and Y . pseudotuberculosis [22] , as well as among Y . pestis genome sequences . For the Y . pestis lineage , other than the low degree of synteny , differences on the nt level were attributed to less than 100 single nucleotide polymorphisms [6 , 23 , 38] . A three-way comparison between both Y . pseudotuberculosis strains IP32953 and IP31758 and Y . pestis CO92 [26] using the BLAST score ratio analysis revealed a high level of protein similarity among all three predicted proteomes with 3 , 642 conserved gene products and also a more distant phylogenetic relationship of Y . pseudotuberculosis and Y . pestis to Y . enterocolitica ( Figure 3 and Table S2 ) [6] . The availability of a second Y . pseudotuberculosis genome sequence provides the opportunity to refine the set of species-specific genes for Y . pseudotuberculosis from 341 to 67 genes ( Table S3 ) , a number that is in agreement with the finding of a subtractive genomic hybridization approach , which discovered 112 Y . pseudotuberculosis species-specific small subtractive genomic hybridization fragments with reported insert sizes between 100 to 900 bp [39] . In addition , a total of 265 genes are unique to Y . pseudotuberculosis IP31758 ( Table S4 ) and 289 genes are unique to the previously sequenced Y . pseudotuberculosis IP32953 ( Table S5 ) . Examples of such genes include those on the 36-kb Yersinia HPI ( Figure S1 ) , which is not present in Y . pseudotuberculosis IP31758 . The HPI encodes the biosynthetic pathway for the siderophore yersiniabactin and has been shown to play a key role in the systemic spread of the Yersinia isolates that harbor this island ( all Y . pestis strains and subsets of Y . pseudotuberculosis and Y . enterocolitica ) [40] . Multiple regions potentially relevant to pathogenicity appear to have been horizontally acquired and are scattered throughout the Y . pseudotuberculosis IP31758 genome . These regions , comprising prophages , plasmid-like integrons , and genomic islands , are often characterized by a deviating GC content and are often inserted into tRNA genes ( Figure 1A ) . Mobile genetic elements such as those encoding phage-related integrases and IS elements frequently flank these unique regions and result from the specific mode of incorporation . A number of small insertions were most likely horizontally acquired by Y . pseudotuberculosis IP31758 but do not show or have lost their colocalization to mobile elements ( Table S6 ) . One example of a horizontally acquired virulence determinant is the Yersinia adhesion pathogenicity island ( YAPI ) that has always been found inserted into one of the two tRNAPhe genes and carries several mobility determinants , such as a phage integrase gene and IS elements ( Figure 1A ) . The YAPI was originally described in Y . pseudotuberculosis serotype I strain IP32777 [41] and is also present in Y . enterocolitica strain 8081 [27 , 42] . YAPIIP31758 is shorter than those previously described . Two large deletions in YAPIIP31758 correspond to api84–api56 and api52–api40 [41] , which code for unrelated metabolic functions and a R/M system , respectively ( Figure 4 ) . These deletions account for the difference in size between YAPIIP31758 ( 64 kb ) and YAPIIP32777 ( 98 kb ) . YAPIIP31758 contains several unique genes with no assigned function . All known Yersinia YAPIs harbor a polycistronic pilin gene cluster pilWVUSRQPONML . BLAST analysis of this gene cluster revealed that the best protein similarities outside this yersinial pathogenicity island are found to the respective genes of Photorhabdus luminescens TTO1 ( Figure 5D ) [43] . The YAPIIP32777 cluster has been experimentally shown to be critical for the virulence of Y . pseudotuberculosis IP32777 by mediating adhesion to the respiratory epithelium in a mouse model [41 , 44] . A comparison of the known YAPI revealed that , while genomic diversity exists in this island , the structure and composition of the pil gene clusters are conserved , strengthening its role in pathogenicity ( Figure 5 ) . YAPI-encoded surface exposed elements such as pilin might be associated with the severe host immune response observed in patients with FESLF . Supporting the role of pilin components in pathogenicity of Y . pseudotuberculosis IP31758 is the presence of two additional pilin gene clusters on each of the two plasmids . The pYpsIP31758 . 1-encoded pilin cluster is located in the pVM82 region previously thought to replace the pVM57 F fragment ( shown in yellow in Figure 1B; Table S1 ) and reported to be critical for pathogenicity [8] . The observed altered clinical manifestations as well as the conjugal transfer of pVM82 were attributed to the presence of this pVM82-specific region [21] . Unlike the entire YAPIIP31758 pil cluster , which is phylogenetically related to that of P . luminescens , different parts of the plasmid-borne pil clusters are most similar to several other bacterial species , including Escherichia , Salmonella , and Pseudomonas species , indicating a different phylogenetic origin than those of YAPIIP31758 . In contrast , Y . pseudotuberculosis IP32953 is YAPI negative and does not produce pilins . Another important virulence-associated factor identified in Y . pseudotuberculosis IP31758 is YPM [45 , 46] . YPM is a superantigenic toxin that belongs to a class of highly potent immune stimulatory proteins produced by a variety of Gram-positive bacteria and retroviruses [47] . Currently the Y . pseudotuberculosis mitogen is the only known superantigenic toxin identified in Gram-negative bacteria [20 , 48 , 49] . The YPM superantigen has been experimentally shown to interfere with the host immune system and is thought to be critical to the pathogenicity of FESLF-causing Y . pseudotuberculosis strains [41 , 50–53] . YPM may be associated with the particular scarlatinoid fever syndromes because it mediates an uncontrolled host immune system activation [20 , 54] . This is analogous to the role of superantigens in staphylococcal and streptococcal toxic shock syndromes [18 , 19] . The similarities in the clinical presentation of scarlet and scarlet-like fever suggest a direct role of YPM in the pathogenesis and the distinct clinical manifestation of Y . pseudotuberculosis isolates causing FESLF . Both superantigenic toxins , YPM and staphylococcal enterotoxin A , are implicated in scarlet-like and scarlet fever and have been shown to interact with multiple eukaryotic signaling pathways in a mouse model [51 , 52 , 55] . The ypm gene is found in a Y . pseudotuberculosis subgroup isolated predominantly in Far East Asia , and its presence or absence correlates with the different clinical manifestations observed between Far East Asia and Europe [56 , 57] . Furthermore , high anti-YPM antibody titers reported in patients with FESLF who have systemic infections suggest a direct role of YPM in pathogenicity [50] . In Y . pseudotuberculosis , three YPM variants encoded by ypmA , ypmB , and ypmC have been described [54 , 58] and shown to be integrated downstream of a conserved 26-bp motif known as Yersinia recombination site ( yrs ) ( Figure 6 ) . This motif is also present in the corresponding locus of the non-superantigenic strain Y . pseudotuberculosis IP32953 , which lacks the ypm gene and does not produce a superantigen . Comparison of these chromosomal loci showed a strong syntenic organization . In Y . pseudotuberculosis IP31758 , this locus is most similar to that of the ypmA-containing Y . pseudotuberculosis strain AH , with ypmA showing 100% identity at the nt level . The ypmA gene is predominantly found in clinical isolates of Y . pseudotuberculosis from Far East Asia , while the ypmB and ypmC loci are associated with environmental and animal isolates [54] . The HPI is present only in a subset of Y . pseudotuberculosis strains and may be lost by spontaneous excision from the chromosome [59 , 60] . Based on the presence or absence of HPI and ypmA , two subgroups can be established that reflect the geographical distribution of Y . pseudotuberculosis: The YPMA+ HPI− subgroup predominantly comprises far eastern pathogenic types , including those causing FESLF , while the YPMs− HPI+ subgroup contains European gastroenteric pathogenic types [57] . The absence of the HPI in Y . pseudotuberculosis IP31758 therefore most likely reflects its divergent phylogenetic branch rather than the secondary loss of this pathogenicity island . Similar to the staphylococcal enterotoxin A , which is thought to have been acquired through phage infection [61] , it has been speculated that the presence of the YAPIIP31758-encoded pilus might have favored the acquisition of the ypmA locus through phage infection by functioning as an attachment site [44 , 62] . In support of this hypothesis , a correlation exists between YAPI+ strains and YPM+ strains in Far East Asian Y . pseudotuberculosis isolates responsible for FESLF [53] . A 24-kb region ( YpsIP31758_0743 to YpsIP31758_0777 ) characterized by an unusual nt composition exhibits similarity and partial synteny to several reported Enterobacteriaceae pathogenicity-associated islands ( PAIs ) and is flanked by another copy of the YAPIIP31758 phage integrase gene ( YpsIP31758_0743 , 100% nt identity to YpsIP31758_3686; Figure 1A ) . This genomic island was reported to be a PAI and is predominantly found in uropathogenic E . coli strains and in several Shigella species [63–67] . The presence of the Enterobacteriaceae-related IS1 and IS630 elements further supports the phylogenetic origin of this genomic island . This region displays a mosaic composition of phage-like genes encoding integrases and structural components , and the plasmid-borne replication initiation genes repA and repB . Furthermore , a 21-kb region encoded entirely on the plus strand ( YpsIP31758_0312 to YpsIP31758_0327 ) is similar to and syntenic with other enterobacterial pathogenicity islands [63–67] ( Figure 1A ) . Most of the genes within these two islands encode conserved hypothetical proteins with no assigned functions , and orthologs of these genes are found within enterobacterial PAIs [63–67] . The flanking Rhs- and Vgr-related loci are often found to be recombinational hotspots in E . coli [68] . However , while these findings could suggest horizontal transfer , the nucleotide composition of this 21-kb region does not show any unusual pattern , no mobile elements are associated with the island , and the region is conserved in all published Yersinia species genome sequences . It is unclear if this island represents an ancient insertion event , is the remnant of a γ-proteobacterial ancestor genome , or has been transferred between Yersinia and other Enterobacteriaceae . Phages have been implicated in the evolution of bacterial pathogens [69] , and our analysis indicates that phage infections might have been responsible for the acquisition of several of the genomic islands implicated in FESLF pathogenicity [53] . The genome sequence of Y . pseudotuberculosis IP31758 contains several regions that were identified as prophage or phage remnants ( Figure 1A and Tables 1 and S6 ) . A large 41-kb prophage called PhiYpsI has been identified . It is encoded entirely on the minus strand and is inserted into tRNALeu-2 , which is part of a tRNALeu-2-Cys-1-Gly2 cluster , resulting in two imperfect direct repeats of 124 bp flanking the insertion site . PhiYpsI appears to be complete and potentially functional . PhiYpsI is similar to Enterobacteriaceae phages previously linked to pathogenicity in Salmonella and Shigella ( Table S6 ) [70 , 71] . The 10-kb P2-like phage PhiYpsII is found adjacent to PhiYpsI and is encoded entirely on the plus strand . These two phages appear to have inserted in tandem into the same target tRNA cluster . The PhiYpsII phage coding sequences ( CDSs ) display homology with CDSs of the large 122-kb phage of Y . pseudotuberculosis IP32953 ( Figure S1 and Table S6; YPTB1834–1840 , YPTB1741–1743 ) [22] . The similar target tRNA insertion site of these two phages may argue for the presence of a P2-like phage at this site in the ancestor of these two isolates , despite that PhiYpsII in IP31758 appears to have lost parts of this ancestral phage . Another 14-kb prophage , PhiYpsIII ( Figure 1A and Table S6 ) , displays similarity to the Burkholderia cenocepacia phage BcepB1A [72 , 73] . While the phages and their insertion sites can be identified , most of the CDSs encode hypothetical proteins whose relevance to pathogenicity cannot be evaluated , but is not excluded . Recently , the role of the unstable filamentous phage YpfΦ in the pathogenicity and fitness of Y . pestis was demonstrated [74] . The strain-specific prophage profile of the scarlatinoid and gastroenteric pathogenic strains Y . pseudotuberculosis IP31758 and IP32953 together with their unique gene content could potentially be used for the genotyping of clinical Y . pseudotuberculosis isolates . Besides the sheer presence or absence of virulence determinants in the Y . pseudotuberculosis strains IP31758 and IP32953 , genes of the shared genomic inventory revealed distinct polymorphisms , which may affect the pathogenic potential of the individual strains . Variation in length in their respective sets of adhesion genes may alter the adhesive and invasive capabilities of each Y . pseudotuberculosis strain during infection ( Figures 1A and S1 ) . The invasins are mediators of pathogenesis in some Yersinia species [75] and confer the ability to invade epithelial cells by binding to integrins , collagen , and fibronectin [76] . The invasin gene ( inv ) has been shown to be important in Y . enterocolitica pathogenesis , but its role in Y . pseudotuberculosis is not fully understood , and it plays no role in Y . pestis in which it is nonfunctional [77 , 78] . Both sequenced Y . pseudotuberculosis strains also encode the attachment invasion locus ( ail ) protein ( YpsIP31758_1160 , YPTB2867 ) [79] and a set of three invasins that show length variation ( YpsIP31758_0608 [2 , 795 aa] , YpsIP31758_2329 [941 aa] , YpsIP31758_4008 [4 , 953 aa] , YPTB1572 [1 , 075 aa] , YPTB1668 [985 aa] , and YPTB3789 [5 , 623 aa] ) . ? >The large plasmid pYpsIP31758 . 1 ( Figure 1B ) encodes several factors that could play a role in the pathogenicity of Y . pseudotuberculosis IP31758 . A detailed analysis of pYpsIP31758 . 1 revealed the presence of a type IVB icm/dot secretion system ( Figure 7 ) . The type IVB icm/dot system was initially discovered by examining Legionella pneumophila mutants defective in replication inside the macrophage and in the secretion of distinct effector molecules [80] . This system had previously only been found in Legionella and Coxiella species [81] and is reported for the first time in the genus Yersinia . The Y . pseudotuberculosis IP31758 icm/dot locus gene structure is most similar to that of C . burnetti , being contained within a single locus , whereas in L . pneumophila , this type IVB secretion system is comprised of two separate loci ( Figure 7 ) . In addition , the presence of a phage-like integrase ( YpsIP31758_B0092 ) within this cluster may indicate the acquisition in Y . pseudotuberculosis IP31758 via lateral gene transfer . In Legionella and Coxiella , these secretion systems have only been reported on the chromosome , while Y . pseudotuberculosis IP31758 represents the first instance of an icm/dot secretion system encoded on a plasmid as part of the mobile genome pool . The infectious process of both pathogenic Yersinia and Legionellales is thought to involve a temporary intracellular stage [4] . While this icm/dot secretion system is absent in all other sequenced Yersinia , it may mediate the intracellular survival of Y . pseudotuberculosis IP31758 in epithelial cells and trigger the host immune system response , both of which are features that may contribute to the unusual scarlatinoid-like clinical presentation [7 , 82–85] . Aside from DotA , none of the type IVB effector molecules reported for L . pneumophila and C . burnetti [86] are found in the genome of Y . pseudotuberculosis IP31758 . However , a number of hypothetical genes found interspersed within the cluster could be potential effector molecule candidates or a unique part of the secretion machinery . Such is the case in Legionella , where type IVB gene clusters include distinct hypothetical genes that are found at syntenic locations in different strains and are believed to be involved in the assembly of the secretion machinery . These diversified gene sets appear to be the result of strain-specific adaptation and specialization . This hypothesis is strengthened by the presence of polymorphisms in the secreted effector molecule DotA found in different Legionella isolates [87 , 88] . Y . pseudotuberculosis IP31758 DotA shows aa similarities of 52% and 54% to the respective homologs in L . pneumophila and C . burnetti . pYpsIP31758 . 1 encodes additional features that could potentially play a role in the pathogenicity and overall bacterial fitness of Y . pseudotuberculosis IP31758 . This includes a gene cluster ( tox ) similar to that of the biosynthetic operon of the phytotoxin toxoflavin initially described in Burkholderia glumae BGR1 . Toxoflavin has been shown to be critical to the pathogenicity and to the overall fitness of B . glumae [89 , 90] . In addition , a homolog of the E . coli umuDC operon that confers UV resistance is present on the plasmid and might contribute to the survival of Y . pseudotuberculosis IP31758 in the environment . The larger plasmid codes for three regulators , the Yersinia global negative regulator ( ymoA ) is found adjacent to the tox operon , the DNA-binding protein H-NS ( YpsIP31758_B0123 ) found upstream of the type IVB secretion machinery and the hemolysin expression–modulating protein Hha ( YpsIP31758_B0044 ) [91 , 92] . YmoA is a virulence-modulating regulator that controls multiple virulence-associated genes and is known to interact with the DNA-binding protein H-NS . In Y . pseudotuberculosis IP31758 , homologs of ymoA are present on both the chromosome ( YpsIP31758_3073 ) and pYpsIP31758 . 1 ( YpsIP31758_B0060 ) , displaying 89% aa similarity . One could speculate on a concerted role for these regulators in modulating plasmid- and chromosome-encoded virulence determinants [93] . pYpsIP31758 . 1 appears to lack a complete plasmid transfer system; however , such a system is present on the smaller plasmid pYpsIP31758 . 2 ( Figure 1C ) . The pYpsIP31758 . 2 transfer system is most similar to that of the Pseudomonas species IncP-1beta group pB3 plasmids [94 , 95] , and may also provide the transfer function for the large plasmid . pYpsIP31758 . 2 is replicated and maintained through a kil/kor system . Such a mechanism has not previously been reported in Yersinia , nor has the incompatibility surface exclusion protein also found on pYpsIP31758 . 2 ( YpsIP31758_A0016 ) [96 , 97] . Together with the chromosomally encoded pathogenicity determinants , the factors present on both pYpsIP31758 . 1 and pYpsIP31758 . 2 , including the two type IV pil gene clusters mentioned previously , might be key to the peculiar clinical presentations of Y . pseudotuberculosis IP31758 FESLF infections . Among the 67 Y . pseudotuberculosis species-specific genes in regard to Y . pestis , two loci were found to encode metabolic functions . These genes code for the methionine salvage pathway and the mdoCGH glucan biosynthetic cluster ( Figure 1A ) . Orthologs have been recently reported to be also present in the distantly related Y . enterocolitica strain 8081 [27] . Osmoregulated periplasmic glucans are intrinsic components of the Gram-negative bacterial envelope . This pathway was initially characterized in Erwinia chrysanthemi osmoprotectant-deficient mutants presenting hypersensitivity to bile salt and antibiotics , reduced enzymatic production , and even complete loss of virulence [98] . mdoG and mdoH are sufficient for glucan biosynthesis , and deletions in either abolish osmoregulated periplasmic glucans synthesis , whereas mdoC is dispensable and thought to succinylate the periplasmic glucan [99 , 100] . The number of deleterious point mutations observed in the two sequenced Y . pseudotuberculosis isolates suggests mdoC might not be functional . The methionine salvage pathway is present in both Y . pseudotuberculosis strains , IP31758 and IP32953 , although it is absent in all sequenced Y . pestis strains [101] . The methionine salvage cycle biochemical pathway maintains methionine levels by recycling methylthioadenosine , a product of the biosynthesis of polyamines such as spermine and spermidine into methionine . The presence of this pathway in the atypical Y . pestis subspecies pestoides F and Y . enterocolitica strain 8081 [27] suggests that this loci has been lost in Y . pestis and was present in the ancestral root of this lineage . This hypothesis is strengthened by the absence of deviating GC content or colocalization of mobile genetic elements at this locus that would indicate a recent or ancient acquisition . In addition , mtnN ( 5′-methylthioadenosine/S-adenosylhomocysteine nucleosidase ) , a component of the pathway located elsewhere in the genome , remains present in all sequenced Y . pestis strains . To expand the analysis , a panel of 46 geographically and phenotypically diverse Y . pseudotuberculosis and Y . pestis was screened for the presence of the identified unique chromosomal regions and plasmid content of Y . pseudotuberculosis IP31758 ( Figure 8 and Table S6 ) . We attempted to determine those genetic regions that differentiate the gastroenteric pathogenic type Y . pseudotuberculosis strain IP32953 from the FESLF-causing Y . pseudotuberculosis strain IP31758 and might be directly responsible for the peculiar clinical features of FESLF . The occurrence of such genes should be uniform within distinct Y . pseudotuberculosis FESLF isolates , while genes whose presence is variable within strains probably are not related to the clinical FESLF manifestation . The isolates selected encompass Yersinia genetic diversity ( serotype: I , II , III , IV , V; biotype: Antiqua , Medievalis , Orientalis ) and include 11 isolates from the time period of the first reported FESLF epidemic on the east coast of Russia [10] ( Figure 8 and Table S6 ) . We also tested for the prevalence of pYV within 12 other Russian isolates used in the study ( Carniel et al . , personal communication ) . This analysis revealed that 9 strains ( IP33117 , IP33215 , IP33125 , IP33223 , IP33156 , IP33199 , IP33208 , IP33218 , and IP33185 ) contained pYV , while 3 isolates ( IP33187 , IP33170 , and IP33111 ) lack pYV . It is not uncommon for pathogenic Yersinia to lose pYV in vitro , in particular when incubated at 37 °C , the temperature used for stool cultures in clinical microbiology laboratories ( Figure 8 ) and used for the isolation of Y . pseudotuberculosis IP31758 . Similarly , Y . pseudotuberculosis IP33187 and IP33170 , two pYV− isolates , were isolated from the stools of patients with FESLF . Furthermore , a number of Y . pseudotuberculosis and Y . enterocolitica isolates have been reported to be pathogenic while lacking pYV [28–30] . Because of the high sequence similarity between all Y . pseudotuberculosis pYV or Y . pestis pCD plasmids [9] , it is unlikely that pYV is responsible for the unique clinical manifestation of FESLF disease , but when present , pYV might contribute to the pathogenic potential of the isolates , such as IP33223 and IP33199 . The 36 Y . pseudotuberculosis isolates selected encompassed the main classical serotypes ( I to V ) found worldwide and included another 12 isolates from Russia , of which eight were isolated from human stools ( six of them from patients presenting FESLF symptoms: IP33223 , IP33170 , IP33187 , IP33199 , IP33156 , and IP33185 ) . The remainder included isolates from wildlife and environmental samples for which no clinical phenotypes were assigned . Five of the Russian isolates harbored all the loci tested but those of pYpsIP3158 . 2 , suggesting that they are genetically homogeneous . However , a broader diversity was found in the other isolates , some of which known to cause FESLF . This strengthens the findings that the genomic diversity in Y . pseudotuberculosis is greater than originally thought . Interestingly , two Y . pseudotuberculosis strains ( IP33208 and IP33199 ) isolated from stools of patients with FESLF appeared to be lacking three and four of the pYpsIP31758 . 1 loci tested , respectively . This result might indicate that in these isolates , either the sequence at these loci is missing or divergent from that of pYpsIP31758 . 1 , or the plasmid is lacking . The latter is not supported by previous experiments showing that pVM82 is critical for the pathogenicity of FESLF [8 , 21] . Overall , the tested markers are restricted to Y . pseudotuberculosis and narrowly distributed in Far East Asian isolates; they might therefore play a role in FESLF pathogenicity [102] . The genetic heterogeneity between Y . pseudotuberculosis isolates in far eastern and western countries is documented in our screening analysis [56 , 57] . The pattern linked to Y . pseudotuberculosis IP31758 dominates in Far East Asia , and the modern Russian strains still harbor the unique characteristics of the original strain . The superantigenic toxin ypmA was found in all FESLF-causing strains as well as in two Russian environmental isolates . However , a PCR product was also identified in non-FESLF–associated isolates from other parts of the world ( Figure 8 ) . This may indicate either that ypmA is not responsible alone for the scarlet-like symptoms , but it may be necessary in association with other genes , or that Russian and non-Russian isolates harbor different alleles with different activities . Most other Y . pseudotuberculosis IP31758 specific chromosomal genes were detected in several Y . pseudotuberculosis isolates of worldwide origins ( Figure 8 ) . Some of these genes have metabolic functions ( periplasmic glucans biosynthesis gene mdoG or glycerol phosphate transporter glpT ) and likely contribute to the overall bacterial fitness for the survival of Y . pseudotuberculosis in the environment . The small conjugative plasmid pYpsIP31758 . 2 was exclusively found in Y . pseudotuberculosis IP31758 , which argues against a role of this plasmid during FESLF infection . Nevertheless , the encoded adhesive pilin structure may contribute to the Y . pseudotuberculosis IP31758 strain-specific FESLF symptoms , and its conjugal transfer apparatus may interact with the coharbored pYpsIP31758 . 1 plasmid and facilitate its transmission and spread . Interestingly , the only non-Russian strain that carries pYpsIP31758 . 1 also harbors ten of the 12 IP31758-specific genes . This strain was isolated from the biopsy of an otter in Sweden ( Figure 8 ) . It may thus be speculated that derivatives of FESLF-associated Y . pseudotuberculosis isolates are spreading among wildlife in this part of the globe , and that human cases of FESLF may appear in previously unscattered countries neighboring Russia . The genome sequence comparison of two Y . pseudotuberculosis strains gives insights into the evolution of this important species and refines our understanding of genome reduction by lowering previous estimates of the number of genes lost in Y . pestis since emerging from Y . pseudotuberculosis . The genetic traits predicted to contribute to pathogenicity in Y . pseudotuberculosis IP31758 , including two novel plasmids , comprise the majority of the strain-specific gene pool . We have presented evidence demonstrating that most of the unique genes in each sequenced Y . pseudotuberculosis strain were laterally acquired , and not lost in the other Yersinia as previously thought . By reducing the Y . pseudotuberculosis species-specific gene pool to 67 , the number of putative genes lost in Y . pestis during the speciation process is also reduced ( 128 genes were found to be unique to Y . pseudotuberculosis IP32953 and Y . pestis CO92 [22] ) . Unlike the Y . pestis lineage that has undergone gene loss [22] , our analysis indicates that lateral gene acquisition is the predominant driver in the evolution of Y . pseudotuberculosis species . In the case of Y . pseudotuberculosis IP31758 , its unique gene pool was mainly acquired from Enterobacteriaceae and other soil-dwelling bacteria ( Figure S2 ) . The acquisition of a short DNA segment in a single event , such as observed for the inserted superantigenic toxin YPM or genes introduced by the novel plasmids pYpsIP31758 . 1 and pYpsIP31758 . 2 , may be a major evolutionary step in the evolution of a species and sufficient to transform a pathogenic bacterial strain into a more severe variant , causing a drastically different disease , regardless of the preexisting chromosomal background heterogeneity . The Y . pseudotuberculosis IP31758 genome contains only 21 degenerate genes , which is far less than reported for the published Y . pestis genomes [23–26] . Driven by different environmental selective pressures , the two sequenced Y . pseudotuberculosis isolates appear to have undergone niche specific microevolution that led to two different strains with vastly different pathogenic potential and unique physiological capabilities .
Y . pseudotuberculosis IP31758 ( serotype O:1b ) was isolated in 1966 from the stools of a patient presenting with scarlet-like fever in the Primorski region of the former Soviet Union and was sent in 1971 to the Institut Pasteur ( Paris , France ) by Dr . Timofeeva ( Antiplague Institute , Irkoutsk , Russia ) . The strain sequenced and analyzed in this study was subcultured once from that original 1971 stock culture for the purpose of this study . A collection of 46 geographically and phenotypically diverse Y . pseudotuberculosis and Y . pestis strains was screened for the presence or absence of 18 loci specific to Y . pseudotuberculosis IP31758 ( Figure 8 ) . Genomic DNA of Y . pseudotuberculosis IP31758 was subjected to random shotgun sequencing and closure strategies as previously described [103] . Random insert libraries of 3–5 kb and 10–12 kb were constructed , and 61 , 634 high-quality sequences of 837 nt average read length were obtained . A draft genome sequence was assembled using the Celera assembler [104] . An estimate of the copy number of each plasmid was obtained by dividing the coverage depth of the plasmid by the coverage depth of the chromosome . The chromosome and the two plasmids were manually annotated using the TIGR Manatee system ( http://manatee . sourceforge . net ) . For each of the predicted proteins of Y . pseudotuberculosis IP31758 , a BLASTP raw score was obtained for the alignment against itself ( REF_SCORE ) and the most similar protein ( QUE_SCORE ) in each of the genomes of Y . pseudotuberculosis IP32953 and Y . pestis CO92 . These scores were normalized by dividing the QUE_SCORE obtained for each query genome protein by the REF_SCORE . Proteins with a normalized ratio of <0 . 4 were considered to be nonhomologous . A normalized BLAST score ratio of 0 . 4 is generally similar to two proteins being 30% identical over their entire length [36] . The primer pairs are listed as supporting information in Table S7 . PCRs were performed with 1 U of Taq polymerase ( Roche , http://www . roche . com ) in the supplied buffer . PCR amplification reaction mixtures contained 10 μM of each primer and 1 mM dNTPs . The PCR program involved one step at 94 °C for 5 min , followed by 35 cycles of amplification of three steps: ( 1 ) 94 °C for 30 s , ( 2 ) 60 °C for 30 s , and ( 3 ) 72 °C for 7 min . PCR products were maintained at 72 °C for 7 min , separated by gel electrophoresis in 1% agarose gels , and stained with ethidium bromide . The chi squares and GC skews were computed according to Nelson et al . [103] . For the chromosomal chi square , a window size of 2 kb and a sliding window of 1 kb was used , while a window size of 1 kb and a sliding window of 0 . 2 kb were used for the two plasmids . GC skews were calculated using a window size of 1 kb for the chromosome and 0 . 2 kb for the two plasmids . The whole-genome alignment tool NUCmer [105] was used to calculate the overall gene identities to the respective Y . pseudotuberculosis and Y . pestis strains . Each of the 4 , 164 Y . pseudotuberculosis CDSs ( not including the RNA genes ) was blasted using BLASTP against the National Center for Biotechnology Information ( NCBI ) protein database ( E-value > 10−5 ) . The BLAST output was parsed using a custom Perl script that recorded the taxonomic affiliation of the BLAST best hit for each protein .
The sequences have been deposited in GenBank ( http://www . ncbi . nlm . nih . gov/Genbank ) under accession numbers CP000720 ( chromosome ) , CP000719 ( pYpsIP31758 . 1 ) , and CP000718 ( pYpsIP31758 . 2 ) . The genome assembly has been deposited in the NCBI Assembly archive ( http://www . ncbi . nlm . nih . gov ) under Assembly ID ( AI ) 1935 , and all sequencing traces are available from the NCBI trace archive under Contig ID numbers 280084 , 280085 , and 280086 . | We have analyzed the genome sequence of a Y . pseudotuberculosis isolate responsible for Far East scarlet-like fever ( FESLF ) . FESLF leads to severe clinical manifestations , including scarlet-like skin rash , from which this illness gets its name , and , most importantly , a toxic shock syndrome not seen in common pseudotuberculosis infections . The aim of this study was to catalogue the genomic inventory and get insights in the origin and emergence of this disease . The genus Yersinia comprises two other pathogens that cause worldwide infections in humans and animals: Y . enterocolitica , like Y . pseudotuberculosis , causes gastrointestinal disorders , while Yersinia pestis is the causative agent of plague , also known as the “Black Death . ” By comparing the genome of these three Yersinia species , we could identify several unique virulence determinants , many of which are known to trigger and modulate the host immune system response and may be intimately associated with the severe and atypical FESLF clinical presentations . We have shown that the reductive gene loss process that Y . pestis has undergone since emerging from the enteric pathogen Y . pseudotuberculosis is not as extensive as originally thought . On the other hand , our analysis indicates that gene acquisition is a major factor that influenced Y . pseudotuberculosis genome evolution . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] | [
"public",
"health",
"and",
"epidemiology",
"microbiology",
"evolutionary",
"biology",
"genetics",
"and",
"genomics",
"eubacteria"
] | 2007 | The Complete Genome Sequence of Yersinia pseudotuberculosis IP31758, the Causative Agent of Far East Scarlet-Like Fever |
In this paper we use a hybrid multiscale mathematical model that incorporates both individual cell behaviour through the cell-cycle and the effects of the changing microenvironment through oxygen dynamics to study the multiple effects of radiation therapy . The oxygenation status of the cells is considered as one of the important prognostic markers for determining radiation therapy , as hypoxic cells are less radiosensitive . Another factor that critically affects radiation sensitivity is cell-cycle regulation . The effects of radiation therapy are included in the model using a modified linear quadratic model for the radiation damage , incorporating the effects of hypoxia and cell-cycle in determining the cell-cycle phase-specific radiosensitivity . Furthermore , after irradiation , an individual cell's cell-cycle dynamics are intrinsically modified through the activation of pathways responsible for repair mechanisms , often resulting in a delay/arrest in the cell-cycle . The model is then used to study various combinations of multiple doses of cell-cycle dependent chemotherapies and radiation therapy , as radiation may work better by the partial synchronisation of cells in the most radiosensitive phase of the cell-cycle . Moreover , using this multi-scale model , we investigate the optimum sequencing and scheduling of these multi-modality treatments , and the impact of internal and external heterogeneity on the spatio-temporal patterning of the distribution of tumour cells and their response to different treatment schedules .
Chemotherapy and radiotherapy play important roles in the primary treatment of many cancers and in improving the survival after cancer surgery . Currently , numerous chemotherapeutic drugs and irradiation techniques are employed , which have evolved over several decades through empirical clinical usage . New treatments , such as a novel drug or a change in the scheduling of radiotherapy take many years to assess by conducting a clinical trial and clinicians would benefit greatly from having an alternative scientific approach to decide on how to improve current treatment strategies , but also in arriving at good decisions more quickly . Mathematical modelling of such complex , dynamic situations might provide one solution to this problem , and speed up delivery of efficacious treatments to patients while preventing the use of potentially sub-optimal treatment combinations . The effectiveness of these treatment protocols is considerably affected by internal tumour heterogeneities caused by perturbations in the intracellular pathways as well as by dynamical changes in the tissue microenvironment , in particular the distribution of oxygen [1] . Hence , it is important to consider such heterogeneity when studying various optimisation protocols , as this can help in improving the delivery of multi-modality treatments . A common treatment modality for cancer is chemotherapy . Its delivery is limited by toxicity to normal tissues , so is often delivered in cycles that allow recovery of normal cells , but also , unfortunately , of tumour cells , leading to treatment failure . Chemotherapeutic drugs function by killing the tumour cells through interfering with the cell-cycle mechanism , which regulates complex intracellular processes such as proliferation , cell division and DNA replication [2] . The cell-cycle mechanism is very dynamic in nature and is influenced by the surrounding microenvironment , which contributes to cell-cycle mediated drug resistance and poor treatment outcome [1] , [2] . One way of overcoming this is by using an appropriate combination of chemotherapeutic drugs that target the cell at various cell-cycle phase points , thus interfering with tumour cell division . Radiotherapy is curative for certain cancers when used as the sole treatment , but clinical trials conducted in the last thirty years suggest a synergistic effect of concomitant chemotherapy and radiotherapy . As with chemotherapy , the cell-cycle plays a vital role in mediating a cell's sensitivity towards radiation therapy , as the cell-cycle phase determines the cell's relative radiosensitivity [3] , [4] . Moreover , various studies have shown that the cells that are in G2-M phase are more sensitive to the radiation than those that are in G1 phase [3] . Furthermore , irradiation can also alter a cell's cell-cycle dynamics through the activation of various intracellular pathways including the p53 and p21 pathways [4] . Activation of these pathways and related cell repair mechanisms can delay the rate of progression of a cell's cell-cycle , causing a group of tumour cells to accumulate either in the G1 or G2 phase , preventing them from undergoing mitosis and making them progress in a synchronous manner [3] , [4] . The treatment-dependent perturbations of cell-cycle progression together with cell-cycle-dependent therapeutic sensitivity are some of the many rationales behind the use of kinetically-based administration protocols of chemotherapy and radiation therapy [3] , [5] , [6] . Studies have shown that both radiation therapy and chemotherapeutic drugs can induce a cell-cycle synchrony and arrest cells at a particular cell-cycle phase which improves the effectiveness of the next dose of radiation/chemotherapy [3] , [5] . For example , while the drug paclitaxel induces a cell-cycle arrest at the cell-cycle phase G2-M , Flavopiridol causes cells to accumulate in G1 and G2 phases , enhancing the radiation sensitivity [3] . Alternatively , radiation-induced cell-cycle delay can help various cell-cycle phase-specific drugs to induce a higher cell kill . Moreover , combination regimes can also provide benefit from spatial cooperation and tissue reoxygenation , which enhance therapeutic response . As the interdependency of such therapeutic protocols , the cell-cycle mechanism and the tumour microenvironment clearly affect a cell's response to therapy , it is important to carefully study optimal combination and sequencing of treatments in order to help clinicians design therapeutic protocols that improve survival rates , in which mathematical modelling can be very helpful . Clinically driven mathematical models can be used as powerful tools to understand , study , and provide useful predictions related to the outcome of various treatment protocols used to treat human malignancies . Although there are several models in the literature that study chemotherapy and radiation therapy , very few of them analyse the effect of the cell-cycle in treatment response [7]–[14] . Recently , Powathil et al . [15] developed a hybrid multiscale cellular automaton model incorporating the effects of oxygen heterogeneity and cell-cycle dynamics to study cell-cycle based chemotherapy delivery . They have shown that an appropriate combination of cell-cycle specific chemotherapeutic drugs could effectively be used to control tumour progression . Most of the mathematical models for radiation therapy are based on a linear quadratic ( LQ ) formulation [12] , [14] , [16]–[20] . A brief summary of various approaches in modelling tumour dynamics and radiotherapy can be found in the review by Enderling et al . [21] . Here , we use a discrete multiscale modelling approach to study the multiple effects of cell-cycle and radiotherapy . Ribba et al . [11] proposed a multiscale model incorporating a discrete mathematical model for cell-cycle regulation and cell-cycle phase dependent radiation sensitivity . Richard et al . [12] studied in vitro responses of cells with cell-cycle phase-specific sensitivity to targeted irradiation and analysed the so called “bystander effect” using a cellular automaton approach . In the present paper , we study the multiple effects of radiation therapy when applied in combination with cell-cycle specific chemotherapy in the control of malignant cell growth by using a previously developed hybrid multiscale cellular automaton model for tumour cell growth [15] . In particular , we are interested in studying the effects of cell-cycle regulation in radiation therapy and further , how radiation-induced cell-cycle heterogeneity can potentially be used to increase tumour control when radiotherapy is administered with chemotherapy ( “chemoradiotherapy” ) . Moreover , as the radiation sensitivity is also affected by the surrounding tumour microenvironment , especially the oxygen distribution , we use a modified linear quadratic model to study the effects of radiation in a changing microenvironment .
Cell-cycle phase-specific chemotherapeutic drugs are used in treating various human malignancies as they interfere with the rapidly proliferating mass of the cells by blocking their cell division cycle . Some of these chemotherapeutic drugs are S phase-specific as they interfere with its replication ( e . g . topoisomerase or thymidylate synthase inhibitors ) , resulting in cell death or cell-cycle arrest at the intra-S checkpoint or at the G2/M checkpoint . Some other drugs are M phase-specific as they damage the formation of the mitotic spindle or prevent it from disassociating ( e . g . taxanes , vinca alkaloids ) while some block phase transitions at G1/S or G2/M cell-cycle checkpoints ( e . g . CDKIs ) and other drugs are not necessarily phase-specific as they interact with the DNA irrespective of its cell-cycle phases . Here , for simplicity we consider two types of phase-specific chemotherapeutic drugs that are either G1 specific or G2-S-M specific . While the concept of phase-specific chemotherapy is useful , and although some drugs have specific effects on the machinery of mitosis ( e . g . ‘spindle’ poisons ) it is becoming clear that chemotherapy drugs may affect more than one aspect of the cell cycle , and so the concept of phase-specificity is somewhat of an over-simplification . The effects of cell-cycle specific chemotherapeutic drugs on solid tumours with intracellular and oxygen heterogeneities are described using the same cellular automaton framework used by Powathil et al . [15] . Using the mathematical model we have shown that the cytotoxic effectiveness of the cell-cycle phase-specific chemotherapeutic drugs is significantly dependent on the spatial distribution of the tumour cell mass , the timing of the drug delivery , the time between the doses of cytotoxic drugs , and also the cell-cycle and oxygen heterogeneity [15] . We have assumed that the diffusion coefficients and supply rates depend on the location of tumour cells within the tumour , as observed experimentally and hence , the cell-kill due to the chemotherapeutic drugs that are introduced affect the diffusion and supply rates of the drug and nutrients in a favourable manner and thus help to redistribute the subsequent doses introduced [15] . The study also highlighted the importance of considering intracellular and external heterogeneities while studying the potential effectiveness of chemotherapeutic drugs [15] . The effectiveness of radiation therapy significantly depends on the intracellular and extracellular dynamics of the targeted tumour . The key intracellular processes , such as cell-cycle dynamics and external factors including oxygen distribution play a vital role in determining the radiosensitivity of the cells that are irradiated [3] , [22] . In addition , the radiation fractions ( treatments ) that are delivered further dynamically change this radiosensitivity over time by redistributing the tumour cells within the cell-cycle , by inducing repopulation of the tumour cell mass , by allowing reoxygenation of the tumour , and by causing the need for repair of the DNA damage induced by treatment [3] , [4] , [22] . Clinically , a kinetically based administration of chemotherapy and radiation therapy is often used to achieve an improved therapeutic effect due to the processes of spatial cooperation , independent additive cell-kill and cellular , molecular and tissue level interaction between modalities [3] , [5] , [6] . However , most of these interactions are dependent on the type of drugs given and the temporal separation between the drugs and radiation fractions , and hence an appropriate combination of these therapeutic modalities is an essential requirement to achieve maximum survival [33] , [34] . Here , we show the analysis of the effects of four hypothetically-scheduled , clinically-used combinations ( adjuvant radiation , neo-adjuvant radiation , concurrent radiation and chemo-radiation-chemo ) of cell-cycle phase-specific chemotherapy and fractionated radiation therapy . A representative result showing the changing dynamics of cells in various cell-cycle phases for the adjuvant therapy ( radiation is given after the chemotherapy ) is given in Figure 4 and the figures for the rest of the combinations can be found in the Supplementary Material , in Figure S2 , S3 and S4 . Moreover , a comparison of the total number of cells for different combination protocols is given in Figure 5 . Figure 4 shows the sequencing of two types of chemotherapeutic drugs followed by radiation therapy . Two doses of cell-cycle phase-specific chemotherapy , specific to either G1 or G2 phases of the cell-cycle , are given at time = 340 h and 370 h , followed by 5 fractions of radiation , given with a daily dose of d = 2 . 5 Gy , starting at time = 400 h . The plots show that when the radiation therapy is given after the chemotherapy doses , the partial cell synchrony during the radiation is lost and a higher proportion of cells stay in G1 , except for the case where two G2 phase-specific drugs ( Figure 4b ) are combined . In Figure S2 , we plot the effects of combination treatments when two doses of the chemotherapy drugs are given after the radiation therapy . The radiation starts at time = 340 h with a similar dose as the previous case and the doses of chemotherapeutic drugs are given at times = 466 h and 496 h . As we have seen in the previous section , the radiation given before two doses of chemotherapy introduces a partial cell-cycle synchrony of cell distribution that remains until the end of the therapy . The cell-phase distribution for the case when the doses of chemotherapeutic drug are given before and after the radiation , is shown in Figure S3 . The figure shows that the administration of a G2-specific drug , which kills fewer number of cells compared to the G1-phase-specific drug , helps to keep the cells in synchrony throughout the treatment time ( similar to Figure 4 ) . In the last case , we studied the application of chemotherapy during the radiation schedule where the chemotherapeutic drugs are given at times 370 h and 400 h with the radiation , starting at time = 340 h . The plots in Figure S4 indicate that the increased cell-kill further reduces the cell-cycle synchrony with the number of G1 phase cells being dominant . The total number of cells for all four cases of radiation and chemotherapy sequencing are compared in Figure 5 . The plots show that in the absence of additional fractions of radiation and further doses of chemotherapy all the schedules perform in a similar fashion , although some give a better cell-kill . However , the effects of the sequencing are critically dependent on the number of cells in various phases of the cell-cycle , as this determines how sensitive they are to the various therapeutic strategies . In every combination except concomitant therapy ( chemotherapy given during radiation ) , we kept the total treatment time constant , if not identical , to compare the effects on tumour control . We have also used the same set of parameter values and doses for each phase-specific chemotherapy [15] . It can be seen from Figure 5 that when a G1-phase-specific drug is administrated , the cell-kill is usually higher than a G2-phase-specific drug . This is mainly because , in most of the cases , the percentage of cells that are in G1 phase is higher than those in G2 . The two factors that contribute to such a cell population distribution are hypoxia and space limitation , as hypoxic cells take a longer time to complete one full cell-cycle and the lack of space forces the cell to enter a resting phase . The cell will re-enter the active phase of the cell-cycle when conditions become favourable . In other words , resting tumour cells may , under favourable conditions created by the administration of the therapies , play a vital role in cell synchronisation and knowledge of this should inform the design of better combinations of cell-cycle phase-specific chemotherapy and fractionated radiation therapy . Moreover , it can be seen from Figure 5 that although the various combination regimes mostly show varying results immediately after the treatment , in the absence of further treatments , these lead to a similar end point with same number of cells as time increases as observed in most clinical situations . However , these differences in the proportion of cells in various cell-cycle phases immediately after each treatment protocol might be vital in designing further treatment plans for that individual patient . Additionally , another factor that plays an important role in therapeutic intervention is the spatial distribution of cells in the tumour mass and its blood supply as these determine the nature of the tumour microenvironment . The tumour microenvironment , in particular oxygen distribution , can significantly affect a cell's radiation sensitivity and thereby introduce cell-cycle heterogeneity throughout the tumour . On the other hand , the vascular distribution within and surrounding the tumour mass determines the effectiveness of the spatial distribution and supply of the chemotherapeutic drugs which determine the cell-kill . A representative figure showing the spatial distribution of the cells in various cell-cycle phases at different treatment time points is given in Figure S5 . In this section , we illustrate the potential of our multiscale computational model to compare different treatment regimens and test the predictions against what happens in a real tumour model in a prospective manner . In doing so we provide a ranking of each regimen in terms of overall treatment efficacy . This highlights the potential of our model to produce an optimal treatment regimen for a given patient . We have compared two different treatment protocols currently used in oesophageal cancer , namely Herskovic , modified Herskovic ( both currently used in clinical practice ) and also a third experimental protocol which is currently not used in clinical practice . Oesophageal cancer is in the “top 10” most common malignancies worldwide , and is the fifth highest in terms of mortality [35] . The treatment of oesophageal cancer used to be primarily surgical or palliative . In the 1980s , clinical trials started that were designed to look at radiation treatment of these tumours and determine whether better results could be obtained by combining radiation therapy with chemotherapy . A seminal trial that commenced in that decade ( “Herskovic” ) reported two groups of patients who received either radiation therapy alone or radiation therapy with chemotherapy [36]–[38] . Although the trial was not conducted with the same rigour of a modern-day trial ( numbers were relatively small and not all the patients were randomized , leading to the possibility of confounding the results ) , the authors reported that none of the radiotherapy alone group but a significant minority of the combined modality group was alive several years after treatment . The method of giving “chemo-radiotherapy” for non-operable cases was then adopted and a dose of radiation and chemotherapy using Cisplatin and 5-FU was used as most patients could tolerate this combination without severe or life-threatening complications . A later modification , not supported by any trial evidence , was to change the extra chemotherapy given in this regimen from adjuvant ( after the chemo-radiotherapy , to neo-adjuvant ( before the radiotherapy ) as this was better tolerated [38] . This “modified-Herskovic” regimen was then used for the best part of a decade or more , before further clinical studies were done to evaluate new chemotherapy agents and different radiotherapy planning techniques [38] . One of these , the SCOPE-1 study , a two arm , open , randomised multicentre Phase II/III trial , is designed to investigate the effect of the drug cetuximab on chemoradiotherapy [39] . This study has recently stopped recruiting and involves hundreds of patients . For the Herskovic treatment protocol , chemotherapy treatment is given on weeks 1 , 5 , 8 and 11 of the treatment period with Cisplatin 80 mg/m2 on day 1 ( D1 ) as a single dose , and 5FU , from day 1 to 4 ( D1-4 ) as a continuous infusion . Radiation therapy is given from weeks 1 to 5 , in 25 fractions of 2 Gy . Similarly , for the modified Herskovic regimen , chemotherapy is given on weeks 1 , 3 , 6 and 10 and radiation therapy is given from week 6 to 10 in 25 fractions of 2 Gy dose . Finally , for the experimental protocol , which has been designed for this comparison study , chemotherapy is given on weeks 1 , 3 , 6 and 9 and radiation is given from weeks 12 to 15 in 20 fractions of 2 . 5 Gy dose . Note that , in all these three treatment protocols the total amount of chemotherapeutic drug and the radiation dosage is kept the same ( although their biological effect are different ) . We simulated treatment with each of the three regimens using our multiscale model over a period of 17 weeks . The simulations were carried out using precisely the same set of parameter values for all three treatment regimens and the results showing the total cell kill over time are given in Figure 6 . First of all we note , as can be seen by comparing the green and blue curves , that our computational simulation results show that the modified Herskovic regimen gives a better final outcome than that of the Herskovic treatment plan as suspected clinically [38] . This gives a degree of confidence in our multiscale model , since this is effectively a “blind test” between the two different regimens . However , interestingly , as seen by comparing the red curve with the green and blue curves , our multiscale model predicts that the new experimental regimen is more effective than either the Herskovic or Modified Herskovic regimen . Although more clinical and experimental studies would be required to confirm these predications , these results highlight the predictive power of our model and its ability to distinguish between a number of different regimens and rank them in terms of overall efficacy or indeed even to predict an optimal treatment strategy . One of the crucial steps towards the successful delivery of anti-cancer treatment is the optimal scheduling and sequencing of different therapeutic modalities , in particular radiotherapy and chemotherapy . The roles of cell-cycle phases as well as tissue hypoxia are believed to be critical in determining the radiation sensitivity of cells and the action of several chemotherapeutic drugs [3] . Changes in the treatment of cancer are currently driven largely by the products emerging from the pharmaceutical industry and although some thought and time is devoted to understanding how best to schedule , combine , and deliver the anti-cancer treatments that are currently available in order to increase their effectiveness , it would be helpful to be able to make rapid rationale treatment choices when designing new treatments based on currently available knowledge rather than take several years to test two treatment regimens in the clinical setting , as currently happens with clinical phase III studies . We believe in silico experiments may help in this regard . In this paper , we have presented a hybrid multiscale cellular automaton model to study the effects of radiotherapy , alone and in combination with cell-cycle specific chemotherapeutic drugs , in controlling the growth of a solid tumour . We have also incorporated the heterogeneities in cell-cycle dynamics and oxygen distribution into our hybrid cellular automaton model as they play an important role in therapeutic effectiveness . The effect of radiation therapy is studied using a modified linear quadratic model that incorporates some of the important factors responsible for radiation sensitivity such as cell-cycle phase-specific radiation sensitivity , improved survival due to DNA repair , and hypoxia . The simulation results from the model showed very good agreement with previous biological experimental results in predicting the cell-cycle dynamics after the irradiation of the tumour cell mass with a single dose of the radiation . The results predicted an increase in the number of G2 phase cells and a possible scenario of partial synchronisation of the cell-cycle , while the control cell population remained in a more or less G1 phase cells dominated proportion . When the cells are irradiated with fractionated radiation , the results showed that the cell-kill enhances the reoxygenation of the tumour mass but also allows the re-entry of resting cells into the active cell-cycle . The study of various factors affecting radiation sensitivity indicated that cell-cycle phase-specific sensitivity and survival due to DNA repair mechanisms could play a vital role in improving radiation cell-kill . Using the present model we have also analysed various possible combinations of cell-cycle phase-specific chemotherapeutic drugs and fractionated radiotherapy . The results show that the sequencing and the type of the chemotherapeutic drugs can significantly affect the cell-cycle and oxygen heterogeneities of the tumour mass which will further affect the effectiveness of the entire therapeutic strategy when they are given in several doses and/or fractions and are consistent with various experimental results [33] , [34] . Overall , the results from the model show its potential usefulness in studying and understanding a kinetic administration of cell-cycle phase-specific chemotherapeutic drugs in combination with radiation therapy . The results from the current and previous studies [15] also confirmed the importance of temporally changing spatial dynamics to improve therapeutic strategies . In future studies , we would like adapt the current model to address the interactions between tumour cells and normal cells and to study how their combined spatial dynamics affect their therapeutic responses . Furthermore , our general computational model can be easily adapted to reflect the behaviour in real clinical scenarios by appropriate validation and comparison with experimental and clinical data . One such comparison on the treatment protocols used in the treatment of oesophageal cancer indicated a therapeutic benefit of the modified Herskovic treatment protocol over the previously used Herskovic protocol . Moreover , the simulations indicated that a suggested new experimental treatment protocol might be an even better strategy than currently used treatment options , but clearly more detailed studies are necessary to validate this prediction . As discussed earlier , the evolution of clinical treatment is slow and takes place over many decades , for the reasons that clinicians must be cautious when introducing new treatments in case of poor efficacy or excess and unexpected toxicity or intracellular and extracellular heterogeneities , and that once a preferred treatment route has been started on , it is generally modified in an incremental fashion . It is very unusual for clinicians to start a completely new way of treating tumours without evidence . The computaional simulation results of our multiscale mathematical model indicates a way for doctors to test the efficacy of new treatment strategies , to allow them to plan more adventurous treatments in silico , prior to beginning actual testing and long and costly clinical trials . This departure may help relieve some of the stagnation in treatment strategy for tumours that have a poor prognosis , and allow medicine to move forward to more innovative treatments that can be evaluated for potential efficacy , prior to clinical testing .
The computational model is simulated on a spatial grid of size grid points and each automaton element whether it is empty or occupied , has a physical size of , where , simulating a cancer tissue of area . If the element is occupied by a cancer cell , the evolution of this cancer cell is based on the decisions made by the cell-cycle mechanism within the cell . To model the cell-cycle dynamics within each cell , we use a very basic model originally developed by Tyson and Novak [40] , [41] that includes only the interactions which are considered to be essential for cell-cycle regulation and control , as given below . ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) ( 6 ) where are the rate constants and the values are chosen in proportional to those in Tyson and Novak [40] , [41] as given in Powathil et al . [15] . In addition to the cancer cells and the empty spaces , the spatial domain also consists of a random distribution of blood vessel cross sections with density , where is the number of vessel cross sections ( Figure S6 ) . This can be justified if we assume that the blood vessels are perpendicular to the cross section of interest and there are no branching points through the plane of interest [42] , [43] . Moreover , the tumour vessel network is irregular , chaotic and abnormal as compared to that of normal vascular network [44] , [45] .
Chemotherapy is a commonly used treatment for cancer . Chemotherapeutic drugs act on rapidly proliferating cells , such as cancer cells , by interfering with the cell-cycle and other cell-cycle specific targets . Hence , it might be more effective to use a combination of chemotherapeutic drugs that targets the cells in different phases of the cell-cycle . The distribution of chemotherapeutic drug type , can be modelled by a similar equation as that of oxygen distribution ( Eq . 7 ) , given by: ( 8 ) where is the diffusion coefficient of the drug type , is the rate by which the drug is take in by a cell ( assumed to be zero as it is very negligible when compared to oxygen uptake ) , is the drug supply rate by the pre-existing vascular network and is the drug decay rate [32] . Here , as similar to the oxygen distribution , the diffusion and the supply rate of the drugs are spatially varied depending on the location in the computational domain . The details can be found in [15] . Radiation therapy is often used in combination with chemotherapy with the intention of increased therapeutic gain for patients with cancer . The survival probability of the cells after they are irradiated are traditionally calculated using linear quadratic ( LQ ) model [46] , given by ( 9 ) where is the radiation dose and and are sensitivity parameters , taken to be and [32] . It has been observed that the radiation sensitivity varies with the cell's oxygenation status [47] , [48] and the effect of changing tissue oxygen levels on the radiation sensitivity can be incorporated into the LQ model ( Eq . 9 ) by using the concepts of an “oxygen enhancement ratio” or “oxygen modification factor” [32] , defined as ( 10 ) where is the oxygen concentration at position , OER is the ratio of the radiation doses needed for the same cell kill under anoxic and oxic conditions , is the maximum ratio and mm Hg is the pO2 at half the increase from 1 to [32] , [49] . Hence , the modified LQ model for the survival probability , incorporating the effects of oxygen distributions can be written as: ( 11 ) The relative radiosensitivity of an individual cell is also partially determined by the cell's cell-cycle phase and studies show that the cells are more sensitive when in the S-G2-M phase as compared with the G1 phase [3] . We have incorporated this varying sensitivity due to the changes in cell-cycle phase by an additional term in the equation for survival probability ( Eq . 11 ) [20] , which gives: ( 12 ) The parameter varies from 0 to 1 , depending on the individual cell's position at the time of the irradiation . Here , we assumed that the cells in S-G2-M phase has maximum sensitivity with while the cells in G1 phase and the resting phase has relative sensitivities of and , respectively . In the current model , although we are not considering the individual cell repair , the studies suggest that 98% of damage caused by the radiation is likely to be repaired within few hours of radiation , if they are treated with low dose radiation ( ) [50] , [51] . If the radiation dosage is higher , this repair mechanism may not be sufficient to repair all the DNA damage . Considering these aspects , Endering et al . [16] , introduced some correctional terms into the LQ model to accommodate these effects due to the cellular repair during low dose radiation treatment . Using these modifications ( allowing for less repair ) into the above LQ model ( Eq . 11 ) with the effects of hypoxia , we obtain the survival probability of the cell as: ( 13 ) This survival probability is then used to calculate the survival chances of each cell when they are irradiated with the radiation rays . To study this survival chance of an individual cell , a random number is drawn for each cell at every time when they are irradiated and compared against the calculated survival probability . The irradiated cell survives if the random number is smaller than the survival probability and die otherwise . Here , we also consider the effects of radiation on cell-cycle as irradiation results in a divisional delay , and , in particular , G2 phase delay/arrest in many cell lines [3] , [4] . Experimental results show that irradiated cells in G2 phase may take up to 9 hours longer to complete the cell-cycle due to the activation of several intracellular repair mechanisms induced by the radiation [4] . Radiation damage can also induce a cell-cycle delay in G1 phase , mainly through the activation of p53 and p21 pathways [3] . In the present model , we include this effect of irradiation induced delay by forcing the cells to stay in the same phase for an extra time duration of up to 9 hours [4] . This divisional delay might be an important factor to consider while studying the optimal sequencing of radiation therapy with cell-cycle phase-specific chemotherapy . The hybrid multiscale cellular automaton model is simulated using the rules and the parameters that are described in Powathil et al . [15] . Here , the position of the new daughter cells are determined by Moore and Von Neumann neighbourhoods alternatively to avoid the associated cell distribution patterns specific to each method . An overview of the equations and their simulation results as adapted from Powathil et al . [15] is given in the Figure 8 . We have incorporated the effects of radiation into the model using the equations ( 12 ) and ( 13 ) . The survival status of an individual cell is then determined using the calculated survival probability by comparing them against a random probability . | Anti-cancer treatments such as radiotherapy and chemotherapy have evolved through clinical trial-and-error over decades , and although they cure some cases and are partially effective in many , the majority of such cancers ultimately recur . Doctors turn to new , expensive drugs as they emerge , but perhaps fail to study and learn how to use the therapies they already have most effectively . This is partly because clinical trials are expensive to conduct , both in terms of time and money . The cancer cell is complicated , but many mechanisms that control its response to treatment are now understood . We show here how a mathematical model accurately reproduces the results of previous biological experiments of cancer treatment , opening up the possibility of using it to predict which combinations of drugs and radiotherapy would be best for patients . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods",
"Microenvironment",
"and",
"tumour",
"growth"
] | [
"oncology",
"medicine",
"mathematics",
"radiation",
"therapy",
"applied",
"mathematics",
"cancer",
"treatment",
"chemotherapy",
"and",
"drug",
"treatment",
"biology",
"computational",
"biology"
] | 2013 | Towards Predicting the Response of a Solid Tumour to Chemotherapy and Radiotherapy Treatments: Clinical Insights from a Computational Model |
High-throughput techniques for detecting DNA polymorphisms generally do not identify changes in which the genomic position of a sequence , but not its copy number , varies among individuals . To explore such balanced structural polymorphisms , we used array-based Comparative Genomic Hybridization ( aCGH ) to conduct a genome-wide screen for single-copy genomic segments that occupy different genomic positions in the standard laboratory strain of Saccharomyces cerevisiae ( S90 ) and a polymorphic wild isolate ( Y101 ) through analysis of six tetrads from a cross of these two strains . Paired-end high-throughput sequencing of Y101 validated four of the predicted rearrangements . The transposed segments contained one to four annotated genes each , yet crosses between S90 and Y101 yielded mostly viable tetrads . The longest segment comprised 13 . 5 kb near the telomere of chromosome XV in the S288C reference strain and Southern blotting confirmed its predicted location on chromosome IX in Y101 . Interestingly , inter-locus crossover events between copies of this segment occurred at a detectable rate . The presence of low-copy repetitive sequences at the junctions of this segment suggests that it may have arisen through ectopic recombination . Our methodology and findings provide a starting point for exploring the origins , phenotypic consequences , and evolutionary fate of this largely unexplored form of genomic polymorphism .
Structural rearrangements of the genome are generally defined to include insertions , deletions , inversions , copy number variants , translocations and transpositions greater than one kilobase [1] . While there is substantial interest in the functional and evolutionary role of structural rearrangements [2]–[5] certain classes of rearrangement have been historically difficult to study . In particular , rearrangements that are larger than a standard sequencing read ( ∼600 bp ) , yet smaller than can be detected cytologically by microscopy , have been very difficult to detect until recently [6]–[8] . Even now , most studies of genome structural polymorphism focus on unbalanced polymorphisms such as copy number variants , for the simple reason that they are straightforward to detect [9]–[12] . As a consequence , little is currently known about the frequency , mutational mechanisms , phenotypic consequences and evolutionary dynamics of balanced structural polymorphisms in any system , with the exception of those associated with transposable elements [13]–[15] . There is reason to suspect that structural variants might be an important source of natural variation . Altered gene expression with phenotypic consequences may arise through position effects [16] or rearrangements within regulatory regions [3] , [5] , [17] . Structural variation can also interfere with normal recombination by suppressing crossovers in structural heterozygotes or , more dramatically , by generating recurrent genomic lesions; the latter have been associated with a variety of disease phenotypes in humans [18] . Structural variation may also contribute to postzygotic isolation between incipient species through the production of genetically deficient hybrids [19] . Two striking examples of this have been reported recently . In crosses between Drosophila melanogaster and D . simulans , sterility in a fraction of hybrid males is caused by the absence of a gene , JYAlpha , that is present in both parental strains but located on the fourth chromosome of D . melanogaster and the third chromosome of D . simulans [20] . In Arabidopsis thaliana , recessive embryonic lethality has been observed in an intraspecific cross where the functional copy of an essential gene for histidine biosynthesis is located on different chromosomes in the two parents [21] . Structural rearrangements that affect gene order between closely related species are well known [20] , [22]–[24] . Though there are some intriguing individual cases [e . g . 25] , it is unclear whether such changes are generally adaptive or the consequence of the fixation of neutral and mildly deleterious mutations . One way to address this question would be to study the population genetics of a large and unbiased sample of naturally occurring gene order polymorphisms [26]–[28] . Such an experiment is currently a challenge , however , due to the lack of systematic methods for identifying such polymorphisms genome-wide . Here , we demonstrate an experimental methodology for identifying balanced structural polymorphisms genome-wide at kilobase resolution . It is based on the observation that if a given segment of DNA resides on different chromosomes in a diploid , random assortment of chromosomes during meiosis will cause that segment to appear as a duplication or deletion among a fraction of the resulting haploids ( Figure 1 ) . Thus , one can identify such unlinked transposed segments ( TS ) by using array-based Comparative Genomic Hybridization ( aCGH ) [6] to measure the copy number of small genomic intervals in the two haploid parents of a diploid , and in the haploid products of diploid meiosis . The ability to propagate Saccharomyces cerevisiae ( hereafter “yeast” ) strains as haploids allowed us to easily apply this strategy to identify transposed segments genome-wide . We analyzed a cross between polymorphic yeast strains S90 and Y101 . The small , well-annotated yeast genome allowed us to characterize the effects of structural variation with greater certainty than would be possible in more complex eukaryotic genomes . The two strains are phenotypically similar in culture and are sexually compatible , as evidenced by F1 tetrads with four viable spores . S90 is nearly identical in gene order and sequence to the sequenced reference strain S288C , while Y101 reportedly lacks ten genes present in S288C [29] , and differs from S288C by ∼0 . 5% at the nucleotide level [30] .
To identify transposed segments with confidence , we first excluded regions of the genome that differed in copy number or hybridization efficiency between the parental strains . Since a number of deletions in Y101 relative to S288C have been identified previously , these parental hybridizations were also used to assess our technical accuracy in identifying deletions . Genomic DNA from strains S90 and Y101 were separately hybridized to DNA microarrays . In both cases , genomic DNA from strain S288C was used as a hybridization reference . S288C is the original sequenced strain , and the strain from which the microarray probes were derived . The microarrays used in this study represent all coding and non-coding regions in the reference genome with an average probe size of ∼750 nucleotides [31] . We were able to identify all ten of the deletions identified previously in Y101 [29] , along with additional putative duplications and deletions ( Tables 1 and S1 ) . Some of these apparent deletions may reflect sequence polymorphisms relative to the probe sequence [32] , [33] . Having identified , and eliminated from further consideration , regions of putative copy number variation between the parents , we aimed to identify putative transposed segments . We comparatively hybridized DNA from each of the four spores of six different tetrads against S288C , and identified putative transposed segments as those genomic regions with copy-number differences among the spores ( Figure 1 ) . We would expect that , some fraction of the time , a transposed segment would result in a non-parental ditype ( NPD ) segregation pattern , in which two spores harbor a duplication and two spores harbor a deletion , or a tetratype ( TT ) pattern , in which one spore harbors a duplication , one spore harbors a deletion , and two spores harbor one copy each . The remaining tetrads would show the parental ditype ( PD ) pattern , in which each spore harbors one copy . The expected frequency of NPD and TT tetrads cannot be predicted in advance , since the expected frequency is a function of whether the loci are linked in the parental strain and the distance of each locus from the centromere . The measurement noise inherent in DNA microarray hybridizations prevented us from relying entirely on the presence of perfect NPD and TT tetrads to identify transposed segments . Therefore , we initially used a more relaxed criterion to identify genomic regions of potential interest . If any duplications or deletions of that region were observed among the spores , it was placed into one of three classes based on the degree of fit to the expected segregation pattern ( Table 1 ) . In Class 1 , a single duplication or deletion was observed in one of the tetrads; this class was the most liberal and included approximately 20% of the genome . In Class 2 , at least one duplication and at least one deletion were observed , but in independent tetrads; this class included approximately 4% of the genome . In Class 3 , the highest confidence category , at least one individual tetrad contained one or two duplications and one or two deletions; this class included 23 regions that covered 0 . 2% of the genome ( Figure 2 ) . While some of the Class 1 and Class 2 regions may be true transposed segments , we initially focused our attention on the higher-confidence Class 3 regions . Many of the Class 3 regions were adjacent or nearly so , suggesting that they were part of larger transposed segments . Thus , we grouped Class 3 regions that were located within five kilobases of each other , and that showed compatible segregation patterns , into six distinct putative transposed segments ( Figure 2 ) . Each transposed segment ( TS ) was named based on its chromosomal location in S288C , two on chromosome XV ( TS15 . 1 and TS15 . 2 ) and one each on chromosomes I , IV , VII , and XVI ( TS1 , TS4 , TS7 , and TS16; Figure 3 ) . They ranged in size from about 1 . 4 kb to 13 . 5 kb . Five of the six transposed segments were between 3 and 21 kb from a telomere and demonstrated the Class 3 pattern in at least four out of six tetrads . The lone exception , TS4 , was also the smallest putative transposed segment and exhibited the Class 3 pattern in just one tetrad . Collectively , the six transposed segments contained a total of 15 annotated genes ( seven ‘verified’ , six ‘uncharacterized’ , and two ‘dubious’ ) and one transposable element ( Table 2 ) . We sought to determine the endpoints of TS15 . 1 , the largest of the six segments , more precisely by manual inspection of the hybridization data ( Figure 4 ) . The segment was initially identified by eleven closely linked Class 3 regions , including 12–17 , 19 , 20 , and 22–24 . Probes 18 and 21 had been excluded from consideration initially because they were not present on the particular batch of microarrays we used . The positions of the TS15 . 1 endpoints were ambiguous on the basis of the aCGH data alone . While the most distal duplication ( relative to the centromere ) was at probe 12 , deletions continued all the way to the telomere in Y101 ( probes 0–11 ) , in addition to several of the spores . At the proximal end of TS15 . 1 , probe 25 had been excluded due to missing data from Y101 and probes 26 and 27 showed no evidence for copy number variation . While probes 28–31 had duplications in three of the spores , only one of these spores was duplicated in the core regions of TS15 . 1 . This ambiguity motivated us to further characterize the endpoints of TS15 . 1 by a PCR assay . We tested for amplification of an appropriately sized product from the parents and the spores using primers that corresponded to the ends of the microarray probes . While detection of duplications requires a more quantitative assay , our methodology could easily identify deletions . Amplification within a transposed segment should fail in a spore harboring a deletion while succeeding in both parents . Amplification could also fail if the primers span a transposed segment endpoint or if the primer sites have diverged between the two strains , but these cases can be distinguished from true deletions by a lack of amplification in Y101 , since the primers are designed to match the S288C reference sequence . Initially , we tested primer pairs corresponding to probes 11 through 32 and probe 36 in the two parental strains , the reference strain , and the four spores from tetrad 27 ( Figure 5 , Tables S2 and S3 ) . Amplification was obtained from all genotypes using primer pairs from probes 11 , 26–30 , 32 , and 36 . Primers corresponding to probes 13 through 25 failed to amplify products only in spore 27A , consistent with the hypothesis that this spore did not inherit TS15 . 1 from either parent . Probes 12 and 31 failed to amplify in spore 27A and D , and also in Y101 , indicating that segments 12 and 31 are candidate endpoints for TS15 . 1 . To map the right endpoint more finely , we designed new primers to split probe 31 into two halves , 31L and 31R . The results support 31L as being external to the transposed segment , and identify 31R as containing the endpoint ( as illustrated in Figure 6 ) . The pattern of amplification in probes 12 and 31R is consistent with the endpoints of the transposed segment occurring within these two intervals . However , since probes 26–30 amplified in all genotypes , they appear to be external to the transposed segment . To confirm this surprising result , we used primers corresponding to probes 11 , 14 , 16 , 19 , 21 , 27 , 30 , 32 , and 36 in the other five tetrads . In all cases , the amplification results were consistent with the aCGH-predicted duplications and deletions in probes 13–25 and the PCR-predicted endpoints in probes 12 and 31R ( data not shown ) . The presence of an apparent endpoint within probe 31R rather than probe 25 suggests that TS15 . 1 region differs not only in genomic location , but also in structure , between the two strains , possibly through an inversion of ∼6 kb . Thus , we conclude that TS15 . 1 comprises the segments of the genome covered by probes 13–25 and portions covered by probes 12 and 31R , for a total of about 15 kb of DNA originating approximately 12 kb from the left end of the chromosome XV reference sequence ( Figure 6 ) . We can infer the position of a transposed segment in S90 based on its position in the genome assembly of S288C , but its position in Y101 is unknown . To map the transposed segment in Y101 , we identified genomic regions that co-segregated with the transposed segment in F1 spores . We obtained parent-of-origin information for 6 , 215 open reading frame ( ORF ) -based probes in each spore using a second microarray-based technique , genomic mismatch scanning , or GMS [34]–[36] . With these data , we can also identify meiotic crossover events that may have occurred within the transposed segment . We focused on the two genes within TS15 . 1 , YOL158C ( probe 23 ) and YOL160W ( probe 16 ) . In the F1 haploids , alleles of YOL158C derived from Y101 perfectly cosegregated with eleven genes near the left telomere of chromosome IX , while alleles of YOL160W derived from Y101 perfectly cosegregated with three genes immediately adjacent to the same region on chromosome IX ( Figure 6 ) . To verify the predicted location of TS15 . 1 on chromosome IX in Y101 , the chromosomes of both S90 and Y101 were separated using pulsed-field gel electrophoresis ( PFGE ) and subjected to Southern blotting with a probe amplified from within TS15 . 1 . The probe hybridized to the band that includes both chromosomes XV and VII in S288C and S90 , while it hybridized to the chromosome IX band in Y101 ( Figure 7 ) . A secondary signal , perhaps the result of cross-hybridization to the rDNA repeats , was observed from chromosome XII in all strains . Thus , the PFGE and GMS-data are both in agreement that TS15 . 1 is located on chromosome IX in Y101 . Despite the location of TS15 . 1 on two different chromosomes in the parental strains , and evidence for structural heterogeneity between the two alleles , two meiotic crossovers appear to have occurred very near , possibly even within , the TS . One of the events was observed in tetrad 32 between gene YIL154C and gene YIL155C and one was observed in tetrad 55 between YIL158W and YIL157C ( the latter illustrated in Figure 6 ) . The segregation patterns leads us to infer that the orientation of the segment in Y101 is opposite to that in S90 , i . e . that YOL157C is distal to YOL161C ( results not shown ) . The TS15 . 1S288C arrangement is seen not only in S90 but in the genome assemblies of two other sequenced S . cerevisiae strains: the wine strain RM11-1a [37] and YJM789 , a strain isolated from the lungs of an AIDS patient with pneumonia [32] . While additional strains have been sequenced by the Saccharomyces Genome Resequencing Project [38] , [39] , the genome assemblies have used S288C as a template , and thus are not informative regarding structural differences relative to that template . Instead , we examined the sequence reads that spanned the breakpoints , segments 12 and 31R . We found no evidence for the null allele in any of the strains , suggesting that TS15 . 1S288C is by far the more common arrangement among present-day strains . To determine the ancestral state for TS15 . 1 , we examined the genome sequence of S . paradoxus and S . bayanus [24] , [40] . In the initial genome assembly , S . paradoxus “contig 539” contains homologs to the genes YOL157C ( probe 25 ) , YOL156W ( probe 27 ) and YOL155C ( probe 30 ) , which span the proximal endpoint of the transposed segment and are arranged in the same order and orientation as in S288C ( Figure 8 ) . Likewise , S . bayanus contig 223 contains homologs to the genes YOL163W ( probe 10 ) , YOL162W ( probe 11 ) and YOL161C ( probe 13 ) which span the distal endpoint of the transposed segment are also arranged in the same order and orientation ( Figure 8 ) . While it is not possible to compare the genome organization distal to the transposed segment due to the incompleteness and fragmentation of the assemblies in this region for S . paradoxus and S . bayanus , this nonetheless strongly suggests that TS15 . 1S288C is the ancestral state . To validate the aCGH results using an independent method , we obtained 28× coverage paired-end Illumina sequencing of Y101 . The ends of each mapped genomic fragment were separated by 243±35 ( mean±standard deviation ) base pairs . To detect transposition events , we looked for instances in which the two ends of a paired-end read mapped to a different chromosome , or cases in which they mapped more than 5 kb apart from each other on the same chromosome , relative to the S288C reference sequence . We detected 40 such discordant paired-end sequences , each represented by multiple independent sequence reads . The start and end points of these genomic segments are given in Table S4 . The end sequences can be used to locate and orient the position of the corresponding segment in Y101 , and to examine the location of the breakpoints in both strains at fine resolution . For instance , the location of TS15 . 1 on chromosome IX in Y101 is confirmed by the paired-end data . Eighteen of the 40 discordant paired-ends map to different chromosomes , and thus may represent transposed segments . Interestingly , the paired-end data do not support the idea , suggested by the Class 3 transposed segments , that such rearrangements occur predominantly within subtelomeric regions . There are 21 regions for which the paired-end sequences occur at two loci on the same chromosome separated by only 5 to 15 kb , indicative of local rearrangements . In only one case are the paired-ends found on the same chromosome at a distance greater than 50 kb . Overall , four of the six Class 3 TS were validated by the paired-end sequencing data: TS7 , TS15 . 1 , TS15 . 2 , and TS16 . For the remaining two Class 3 predictions ( TS1 and TS4 ) , the many independent paired-ends that spanned the junctions mapped to the same locus on the reference genome within the specified tolerance . Therefore , we conclude that TS1 and TS4 are false positives , yielding an overall specificity of 67% for calling TS by their Class 3 status . The paired-end sequencing data provide evidence for rearrangements in ten of the 529 Class 2 array features , yielding a specificity of only 1 . 9% for these lower-confidence predictions . The number of rearranged features identified by paired-end sequencing data among those categorized as Class 0 or Class 1 is 0 . 09% ( 10 rearranged features out of a total of 11 , 442 Class 0 and Class 1 features ) .
All four of the confirmed transposed segments found by aCGH were located near telomeres , which are known to be susceptible to internal rearrangement [44] , sister chromatid exchange [45] , and interchromosomal exchange [46] . This is consistent with the unexpected observation of illegitimate recombination between the transposed segment on chromosomes IX and XV . Interestingly , the genes concerned ( YIL154C and YIL155C ) have been previously identified as hotspots for double-strand breaks [47] . In addition , subtelomeric regions are commonly considered to be permissive of structural rearrangements [24] . For instance , Wei et al . [32] reported an 18 kb subtelomeric segment on chromosome VI in S288C that is found on chromosome X in YJM789 . The relatively few subtelomeric genes are seldom transcribed at a high level and are frequently silenced [48] , [49] . None of the genes within the confirmed Class 3 TS are essential for growth in rich medium [50] . Deficiencies arising from the segregation of the structural polymorphisms we identified would therefore not be expected to cause gross phenotypic effects in lab culture . While this is consistent with the viability of all the tetrad products examined in this study , the results are biased by the initial selection of complete tetrads for the original GMS study . Of the 312 tetrads originally assayed from the cross , the proportion found to have 0 , 1 , 2 , 3 , and 4 viable spores was 5 . 4% , 3 . 2% , 13 . 1% , 30 . 8% and 47 . 4% , respectively , yielding an overall viability of 77 . 9% ( J . McCusker , pers . comm . ) . On the other hand , relative to the overall gene density in yeast of 0 . 5 gene/kb [51] , we find higher-than-average density in three of the subtelomeric transposed segments ( TS7: 0 . 71 gene/kb; TS15 . 2: 1 gene/1 . 8 kb , TS16: 0 . 55 gene/kb ) and lower-than-average density in only one ( TS15 . 1: 0 . 37 gene/kb ) . Thus , the transposed segments do not , as a general rule , occur in gene-poor regions . In fact , many of the TS junctions , as inferred from paired-end data ( Table S4 ) , are located within the boundaries of annotated exons . Thus , both a high rate of subtelomeric structural mutation and relatively weak purifying selection may both contribute to the maintenance of polymorphism for the transpositions catalogued here . There are no obvious shared genomic features among the transposed segments apart from their subtelomeric positions , and even the subtelomeric bias is absent among the putative rearrangements that are observed only in the paired-end sequencing data . There is a fragmented TY retrotransposon sequence located inside TS15 . 1 ( YOLCDELTA2 ) , but retrotransposons are not found near any of the other transposed segments despite the ability of such elements to generate genomic rearrangements [13] , [52] . Alternatively , duplicate genes and other low-copy repeats can initiate ectopic exchange [53] . The two junctions of TS15 . 1 , as inferred from the paired-end sequencing data ( Table S4 ) , show strong sequence similarity . The chromosome XV junction is in the middle of YOL155C ( HPF1 ) , and the chromosome IX junction is in the middle of YIL169C , a homolog of HPF1 with high sequence similarity ( E = 2 . 6e−172 in a BLASTP search among annotated yeast proteins ) . Furthermore , both are adjacent to HXT genes that have 97% identity to each other at the nucleotide level . While repetitive gene families are known to be well-represented in subtelomeric regions [48] , this is the only pair of junctions among the four confirmed Class 3 TS for which we found evidence of sequence similarity . Thus , non-allelic homologous recombination appears to underlie at least some of the structural mutations observed , but we cannot exclude a role for other mechanisms such as repair of spontaneous double stranded breaks by non-homologous end-joining [54] , [55] . Further work will be necessary to determine the phenotypic consequences of these structural polymorphisms , and indeed of balanced structural polymorphisms in general . Interruption of functional genes and generation of fusion genes are likely to have phenotypic effects in some cases [54] . There is no direct evidence of such rearrangements within TS15 . 1 , but we do not have adequate data to rule them out in the other transposed segments . Position effects on expression are also likely [17] . Expression can be altered not only for the transposed genes , but also for neighboring genes [5] . Some effects on expression could be evolutionarily significant even if phenotypically subtle . For example , it has been proposed that selection favors the clustering of essential genes that are sensitive to stochasticity in expression levels [56] . Naturally occurring transposed segments provide an opportunity to study such context-sensitivity of expression for a large number of genes in a variety of genetic backgrounds . Other classes of phenotypic effects may be ephemeral on the population level , but nonetheless dramatic for affected individuals . Non-allelic homologous recombination between transposed segments at different loci can lead to recurrent genetic lesions [57] , and segregation of transposed segments in the population can lead to recurrent genetic duplications and deficiencies such as those observed among the tetrads studied here . If these phenomena affect organismal fitness , they will influence the frequency of balanced polymorphisms in natural populations . We would expect most transposed segments to be held at low frequencies by purifying selection ( due to the deleterious fitness effects of ectopic recombination and dosage imbalance ) or have dynamics that are governed largely by genetic drift . At the same time , it is likely that some small fraction of transposed segments are adaptive , as has been suggested for the compound structural rearrangement leading to the cluster of genes involved in allantoin degradation pathway in S . cerevisiae and S . castellii [25] . Since universally deleterious and adaptive rearrangements are unlikely to remain polymorphic for an extended period of time , balanced transpositions that are maintained at intermediate frequencies must either have a high recurrent mutation rate ( to counteract genetic drift ) or be subject to conditional fitness effects . The latter class of polymorphisms , typified by the classical inversion polymorphisms in Drosophila pseudoobscura [58] , would be particularly interesting to uncover . More recently , Feuk et al . [59] , reported that 13% of the inversions that they detected in a comparison of the human and chimp genome , with lengths between 1 Kb and 1 Mb , were segregating at appreciable frequency ( minor allele ≥5% ) within humans , though it is not yet clear if this small subset of inversions are neutral or deleterious polymorphisms still in transit , or actually being maintained by balancing selection . The derived TS15 . 1 chromosome has currently been observed in only one strain , although Schacherer et al . [60] , in a survey of diverse strains of yeast , report deletions of the genes within TS15 . 1 in as many as 19 ( 30% ) of the 63 strains surveyed ( Table 3 ) . They also report that genes within TS7 and TS15 . 2 may also be deleted with appreciable frequency . It is not yet clear whether the high frequency of deletions observed at these loci reflects segregation of the same alleles observed here , or whether these instead are derived from some number of independent rearrangement events . In the former case , it would be interesting to explore whether any of these three TS are being maintained by selection , while in the latter case , it would suggest the equally interesting conclusion that the frequency dynamics of these gene deletions are being driven by recurrent mutation . In conclusion , we have developed a systematic method to identify gene order differences between two divergent yeast strains . Our results show that the genomes of two divergent strains of S . cerevisiae are largely collinear , but do harbor a modest number of gene-containing subtelomeric transpositions that are several kilobases in size . These findings raise important questions about the phenotypic consequences and evolutionary dynamics of balanced structural polymorphisms , not only in yeast but also in other natural populations .
We examined six tetrads from a cross between two S . cerevisiae strains . One strain is YJM826/S90m , a spontaneous Gal+ derivative of S1 , which is isogenic with the S288C genome reference strain . The other strain is Y101 ( hoΔ MATα gal3 Mal Suc Bio ) , a haploid derivative of YJM627/Y55 ( HO gal3 ) said to have been isolated from a French vineyard in the 1930s by Oyind Winge [29] , [30] . Cultures of the parental strains and tetrads were kindly provided by P . O . Brown . For aCGH , twenty-milliliter cultures of each sample were grown in YPD medium ( yeast extract 10 g/L , peptone 20 g/L , and 2% dextrose ) to OD600 greater than one . Cells were collected and resuspended in TE ( 700 µL of 200 mM Tris , 50 mM EDTA , pH 8 ) and frozen overnight . To digest cell walls , the cells were resuspended in 535 µL of 20 mg/ml Zymolyase , 1 . 2 M Sorbitol , 20 mM HEPES ( pH 7 . 5 ) and incubated for 60 minutes at 37°C . Collected cells were resuspended in TE+10 mg/ml RNaseA at 65°C for 30 minutes . 200 µL 5 M potassium acetate was added and cells were incubated on ice for 1 hour . The lysed cells were then centrifuged at 14K rpm for 10 minutes to pellet debris . Genomic DNA was ethanol precipitated from the supernatant and resuspended in 200 µL TE ( pH 8 ) . The solution was then sonicated to fragment DNA to an average size of ∼500 bp . DNA was purified using a YeaStar Genomic DNA Kit ( Zymo Research , cat # D2002 ) and concentrations were determined by absorption spectroscopy with PicoGreen dye ( Invitrogen ) . DNA templates suitable for PCR were obtained as previously described [61] . For paired-end sequencing , five ml of YPD was inoculated with a single colony , incubated overnight at in 30°C , and centrifuged . The pellet was resuspended in 500 µl of lysis buffer ( 1 . 2 M sorbitol in 0 . 1 M KPO4 , pH 7 . 4 ) , 30 µl Zymolyase ( 20 mg/ml ) , incubated at 37°C for 45 minutes and centrifuged again . The pellet was resuspended in 500 µl of 50 mM Tris+10 mM EDTA , 1% SDS with 3 µg of RNase and incubated at 65°C for 25 min . 200 µl of 5 M potassium acetate was added , then the solution was incubated on ice for 40 min . The pellet was washed with isopropanol and 70% ethanol and left overnight at room temperature for drying . Libraries were prepared for paired end sequencing on an Illumina GA2 sequencer using standard Illumina protocols ( Illumina , San Diego , CA ) . Twenty microliters of 2 . 5× random primer/reaction buffer mix ( 125 mM Tris 6 . 8 , 12 . 5 mM MgCl2 , 25 mM 2-mercaptoethanol , 750 mg/mL random octamers ) were added to 21 µL of genomic DNA . The DNA was incubated at 100°C for 5 minutes and then placed on ice for 10 minutes . Five microliters of 10× DNTP mix ( 1 . 2 mM dATP , dGTP , dCTP , 0 . 6 mM dTTP , 10 mM Tris 8 . 0 , 1 . 0 mM EDTA ) were added to the genomic DNA . Test samples were labeled by addition of 3 µL 25 nmol of Cy5-dUTP ( Amersham ) to the reaction mix , and the reference sample , S288C , was labeled by addition of 3 µL 25 nmol of Cy3-dUTP ( Amersham ) to the reaction mix . One microliter 40 U/µL Klenow was added to each sample . The DNA was incubated at 37°C for two hours and the labeling reaction was stopped with 5 µL 0 . 5 M EDTA pH 8 . 0 . For each microarray , we purified 5 µg of labeled DNA using a DNA Clean and Concentrator System-5 ( Zymo Research , cat #D4004 ) . The yeast tiling microarrays were identical to those previously described [31] , [62] . Briefly , each gene probe extended from start codon to stop codon . Probes for intergenic regions and other features ( including rDNA , tRNA , transposons , transposon long terminal repeats , introns , telomeres , and centromeres ) conformed to the boundaries as annotated by the Saccharomyces Genome Database ( SGD ) in 2000 . The PCR products were printed on poly-L-lysine coated glass slides by a robotic arrayer as described [63] . DNA was hybridized to the microarrays as previously described [63] . In all cases , S288C was used as the reference for competitive hybridization . Images were acquired using a GenePix 4000B scanner and software ( Axon Instruments ) . Raw and processed aCGH data were loaded into the University of North Carolina Microarray Database ( http://genome . unc . edu ) . The hybridization signal of each probe was recorded as the log2 normalized ratio of the median pixel intensity for the sample relative to the reference . Only probes composed of pixels with consistent ratio values ( regression r2>0 . 6 ) and with gel-verified PCR products were used for analysis . The vector of values from each hybridization was normalized such that the median ratio of all probes relative to the S288C reference was zero . Duplications and deletions in the aCGH data were then detected in a stepwise manner using ChIPOTle version 1 . 0 [64] , using a 1-kb window and a 250 bp step size . First , deletions were identified by using the reciprocal of the ratios ( reference/sample ) . The deletions were then removed from the dataset , and the dataset was re-normalized such that the median ratio of all remaining probes was zero . ChIPOTle was then run on the re-normalized data to detect duplications . Segments for which p<0 . 001 after Bonferroni correction were classified as deletions or duplications , accordingly . A custom Perl script ( available upon request ) was used to identify potential transposed segments by classifying each ChIPOTle window according to the pattern of duplications and deletions observed in the parents and tetrads , as described in Results . Heatmap visualizations were generated using Java Treeview [65] . The initial primers for the PCR assay were the same as those used to amplify the probes for the microarray , with the exception of probes 14 , 19 , 22 , 30 , 31 , and 32 ( Table S2 ) . For probes 31L and 31R , primer pairs [31 forward and internal] and [31 internal and reverse] were used , respectively . Twenty-five µL PCR reactions were prepared at final concentrations of 1× Mg free buffer ( Promega ) , 0 . 1 unit/µL Taq DNA polymerase ( Promega ) , 0 . 25 mM dNTP , 2 mM MgCl2 ( Promega ) , 0 . 14ng/µL genomic DNA , 20 pmol amplification primers ( IDT ) , and 20 pmol positive control primers ( IDT ) . The reactions were carried out with an initial incubation at 92° for one minute , followed by 36 cycles of 92°C for 30 sec , 56°C for 45 sec , and 72°C for 3 . 5 min . GMS is a procedure for biochemical enrichment of genomic regions in which individuals share identical alleles , e . g . two parents and their progeny . GMS data for the six tetrads was obtained during the course of a companion experiment ( C . McCoach , K . Hayashibara , X . Cui , S . Elashoff , E . Ray , J . McCusker , J . DeRisi , D . Siegmund and P . Brown , unpubl . ) . The procedure was performed as described [35] . In brief , heterohybrid DNA molecules formed between genomic DNA fragments from two individuals were purified by differential methylation and endonuclease restriction . Heterohybrids containing mismatches were then removed by the E . coli mismatch repair enzymes Mut H , Mut L , and Mut S . This was done using heterohybrid DNA created from each of the 24 spore strains and the two parents ( a total of 48 samples ) . The hybridization pools for each spore were differentially labeled and comparatively hybridized against each parent to a microarray containing probes for 6 , 145 genes . The probability of inheritance of each gene from either parent was calculated from the GMS data using a hidden Markov model approach to be described elsewhere ( C . McCoach , K . Hayashibara , X . Cui , S . Elashoff , E . Ray , J . McCusker , J . DeRisi , D . Siegmund and P . Brown , unpubl . ) . Let XS90 , P and XY101 , A denote the S90 ( present ) and Y101 ( absent ) alleles , respectively , for a given transposed segment at position X; and let YS90 , A and YY101 , P denote the S90 ( absent ) and Y101 ( present ) alleles , respectively , at position Y . The transposed segment will be duplicated in spores with genotype XS90 , P/YY101 , P , deleted in spores with genotype XY101 , A/YS90 , A , and single copy in spores that inherit alleles from the same parent at both positions . Thus , to map the position of a transposed segment in one parental strain given knowledge of its position in the other , one can search for genes that show the requisite segregation pattern among spores in the GMS data for the known copy-number pattern in the aCGH data . This analysis was performed for TS15 . 1 with a custom Perl script ( available upon request ) that requires the user to input a gene that is close to , but not within , the transposed segment . DNA was prepared for pulsed-field gel electrophoresis ( PFGE ) as previously described [66] . PFGE was performed for 38 hours at a gradient setting of 5 . 0 V/cm on a CHEF Mapper system ( BioRad ) , with an initial and final switch time of 46 . 67 seconds and 2 minutes , 49 . 31 seconds , respectively . Chromosomes were visualized after ethidium bromide staining . The separated chromosomes were blotted onto a Hybond-XL membrane ( GE Healthcare ) and probed against a chromosome XV specific PCR fragment containing the YOL158C gene ( SGD coordinates 15:19490-21310 ) . The probe was labeled using Ready-to-Go DNA Labeling Beads ( -dCTP ) kit ( GE Healthcare ) and hybridized at 65°C for approximately 20 hours . The resulting signal was scanned and visualized using a Typhoon 9200 Variable Mode Imager ( GE Healthcare ) . Y101 and S90 genomic DNA were each run on a single lane of an Illumina GA2 sequencer for 36 cycles using the standard flow cell , yielding ∼28× coverage apiece . Here we report the results only for Y101 . Raw Illumina GA2 image data was phased and filtered for quality using default GERALD parameters for unaligned reads ( analysis: NONE , Use_Bases: 35 ) . Sequencing reads were mapped back to the S288C reference sequence with ELAND . A custom script was written to identify clusters of paired-ends where one paired end mapped to a substantially different location than the other ( available upon request ) . Both ends matched a unique string in the reference sequence for 61% of the reads in Y101 . Putative TS were identified by at least an order of magnitude increase , relative to background , in the number of paired-ends within a region that mapped to a chromosomal location greater than 5 kb away . Two additional patterns were used to confirm putative TS . First , we required that there be paired ends spanning the predicted deletion . Second , we required that there be sets of paired-ends that map to the two different chromosomal locations on either side of both breakpoints ( “reciprocity” ) . Genome assembly and gene order data was obtained for S . cerevisiae strains RM11-1a [37] and YJM789 [32] . Gene order for the TS15 . 1 region in S . paradoxus and S . bayanus [40] was determined using the tiling data available from the Broad Institute ( ftp://ftp-genome . wi . mit . edu/pub/annotation/fungi/comp_yeasts/S3b . Visualization_tiling/ ) . Raw microarray data and images are publicly available from the UNC Microarray Database ( UMD , https://genome . unc . edu ) , and microarray data and paired-end whole-genome sequencing data ( for both Y101 and S90 ) are available through accession number GSE14223 at the NCBI Gene Expression Omnibus and Short Read Archive , respectively . | Balanced structural polymorphisms are differences in the relative arrangement of genomic features within species that do not affect DNA copy number . Little is known about their prevalence or importance because they are difficult to observe . Here , we present a novel methodology for systematically identifying such polymorphisms based on the idea that single-copy DNA that occupies different genomic locations in two parents will segregate independently during meiosis and will therefore reveal itself as a copy number difference among a fraction of progeny . Comparative hybridization reveals multiple balanced structural polymorphisms that involve changes to gene order in two strains of yeast; the results are independently validated using paired-end whole genome shotgun sequencing . The longest transposed segment we identify comprises 13 . 5 kb near the telomere of chromosome XV in the S288C reference strain and contains several annotated genes . We map the location of this polymorphism in the non-reference strain using genome-wide genotypic data , which also reveals an appreciable frequency of ectopic recombination among transposed segment pairs . The breakpoints of the remaining polymorphisms are localized by the paired-end sequence data . Our work provides proof-of-principle for a very general approach to systematically identify all balanced genomic polymorphisms in two different genotypes and is a starting point for understanding the frequency , evolutionary origins , and functional consequences of this seldom-studied class of genomic structural variation in eukaryotes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/chromosome",
"biology",
"evolutionary",
"biology/genomics"
] | 2009 | Systematic Identification of Balanced Transposition Polymorphisms in Saccharomyces cerevisiae |
A major goal of systems biology is to understand how organism-level behavior arises from a myriad of molecular interactions . Often this involves complex sets of rules describing interactions among a large number of components . As an alternative , we have developed a simple , macro-level model to describe how chronic temperature stress affects reproduction in C . elegans . Our approach uses fundamental engineering principles , together with a limited set of experimentally derived facts , and provides quantitatively accurate predictions of performance under a range of physiologically relevant conditions . We generated detailed time-resolved experimental data to evaluate the ability of our model to describe the dynamics of C . elegans reproduction . We find considerable heterogeneity in responses of individual animals to heat stress , which can be understood as modulation of a few processes and may represent a strategy for coping with the ever-changing environment . Our experimental results and model provide quantitative insight into the breakdown of a robust biological system under stress and suggest , surprisingly , that the behavior of complex biological systems may be determined by a small number of key components .
Much of modern biology is inherently reductionist , seeking to enumerate interactions and components to elucidate the inner workings of cells and organisms . However , phenotypes often cannot be explained simply as the sum of the properties of the micro-components . Emergent phenomena [1] are not unique to biology; physical [2] , [3] , [4] , chemical [5] , and social [6] , [7] , [8] , [9] systems all have to contend with this challenge . Over the last several decades , thousands of studies have employed genetic and biochemical approaches to reveal the components of biological processes . High-throughput technologies have greatly accelerated discovery , generating detailed parts lists for cellular systems [10] , [11] , [12] . Such abundance of data facilitated development of fine-grained models that provided quantitatively accurate descriptions of signaling [13] , transcriptional regulation [14] , and the heat shock response [15] . Despite the success of this general approach , it cannot be used in circumstances when detailed understanding of molecules and processes is not available . While this limitation can be overcome by additional experimentation , fine-grained models have an intrinsic difficulty in connecting cellular phenomena to organismal behavior [1] , [16] , [17] , [18] , [19] . An alternative is to use macro-level modeling , which although omitting many specific details , could if properly constructed , describe the overall performance of complex systems [20] , [21] , [22] . Due to its easily quantifiable output , the reproductive system offers an attractive opportunity to bridge the molecular biology of a process and the emergence of dynamic , organismal-level phenotypes . Reproduction in Caenorhabditis elegans has been extensively studied using genetic [23] , [24] , [25] , [26] , [27] , [28] and biochemical [29] , [30] , [31] , [32] , [33] approaches . C . elegans hermaphrodites are self-fertile [34] . They first produce a finite cache of sperm [35] , and then irreversibly transition to oocyte production [36] , [37] , [38] , which occurs continuously until reproductive senescence [39] . The overall reproductive output is primarily determined by the availability of sperm [34] , [40] , because their number is set for the lifetime of an individual . Many of the specific molecular components involved in gametogenesis and later reproductive events have been characterized [41] , [42] , [43] , [44] , [45] , [46] , [47] . For example , a signaling mechanism directly couples oocyte maturation and ovulation to the presence of sperm [31] , [32] , [48] . Although considerable information is available about the components of the reproductive system , we are interested not in specific molecular interactions , but rather in understanding how individual animals reproduce . The distinction between these two questions can be compared to the difference between studying the molecular biology of neurons and human behavior [17] . Our goal here is to construct a parsimonious macro-scale model that is grounded in experimental data . If such a model could provide quantitatively accurate predictions , it would serve to identify a minimal set of biological components and processes necessary to endow the reproductive system with its characteristic dynamics . A time-tested approach to investigating macro-level processes is to perturb the environment in a controlled way and to measure the system's subsequent response . Temperature has often been used to probe dynamic behavior , as well as components and organization of biological systems [49] , [50] , [51] . This is because organisms are sensitive to environmental conditions and because temperature can be easily and precisely manipulated in the laboratory setting . Here , we analyzed the effects of chronic elevated temperatures on C . elegans reproduction to connect molecular processes to macroscopic phenotypes , particularly those involved in dynamic responses of organisms to a changing environment .
Compared to the well-understood heat shock response , less is known about how organisms respond to chronic , moderate temperature stress . It is well established that the average number of eggs laid by C . elegans hermaphrodites is dependent on temperature [35] . We asked whether reproduction is more temperature sensitive than other vital processes and how individual worms respond to temperature stress . We examined viability , movement , and reproductive output over a range of temperatures ( Table 1 , Table S1 ) . We developed an experimental protocol in which nematodes were reared at the commonly used cultivation temperature of 20°C , and then , just prior to the onset of reproduction , individually shifted to various elevated temperatures . This treatment—chronically exposing worms to temperatures between 20°C and 30°C—is qualitatively different from the standard acute heat shock experiments , which involve brief exposure to nearly fatal temperatures ( 33°C ) [52] . Whereas the average number of eggs laid at 28°C was substantially reduced compared to temperatures at which worms are routinely raised ( see below ) , at 30°C reproduction ceased completely ( Figure 1A ) . In contrast , neither viability nor motility was comparably affected ( Figure 1B ) . We documented the reproductive performance of 3 , 418 individual worms , which laid a total of 144 , 092 embryos ( Table 1 , Figure S1 , Text S1 ) . Importantly , we collected dynamic , time-resolved egg-laying curves , not simply overall brood sizes . The temperatures used in our studies ( 20–30°C ) are likely to be physiologically relevant because C . elegans have been isolated from tropical and equatorial locales [53] , [54] where temperatures routinely exceed 30°C . Furthermore , nematodes appear to dwell in compost and rotting vegetable matter [55] , [56] , where temperatures can be even higher than in the ambient environment [57] . Brood size of animals cultivated at 20 and 25°C were normally distributed ( Figures 2A , B , S2 , S3 , Text S1 ) . While the means of the brood size distributions varied with temperature , they had indistinguishable coefficients of variation ( p = 0 . 58±0 . 01 , permutation test ) . These results suggest that while the mean output of the reproductive system is temperature-dependent , increasing temperature does not lead to an appreciable increase in the individual-to-individual variability ( Figure S4 ) . At 28°C , however , we observed a qualitatively different behavior—there were more individuals laying low numbers of eggs than would be expected from a normally distributed population ( Figure 2C ) . This was accompanied by a coefficient of variation ( Figure S4 ) that was significantly higher at 28°C than at 25°C ( p = 2×10−4 , permutation test ) . Furthermore , these data could not be captured by a single normal distribution ( p<10−4 , Kolmogorov-Smirnov test ) , but could be well described by a mixture of two distributions ( Figure 2C ) . The relative proportion of animals laying a lower than expected number of eggs increased at higher temperatures ( Figure 2D ) , as evidenced by the increase in the coefficient of variation ( Figure S4 ) . These results suggest that whereas across a range of lower temperatures reproductive systems of all worms are robust , at higher temperatures , only a fraction of individuals continue to act in a robust manner , revealing an inherent heterogeneity in physiological response . We developed a macro-level model of the C . elegans reproductive system . Our model is both simple ( it includes a small set of essential features and parameters ) and falsifiable ( designed to be experimentally testable ) . The reproductive system ( Figure 3A ) can be abstracted as a pipeline for the serial maturation and subsequent fertilization of oocytes . We conceptualized it as a series of interconnected compartments—the gonad ( which is encapsulated by the gonadal sheath ) , spermatheca , and uterus—through which gametes flow ( Figure 3B ) . This process can be likened to a chemical reaction because transitions between compartments can be modeled as the conversion of precursors to products . We made two simple but plausible assumptions ( a list of major model assumptions is given in Table 2 ) . First , all gametes in the model are conserved and can be explicitly accounted for [58] . Second , all transitions between states obey mass-action kinetics . The latter is a typical assumption for dynamic systems , used in analysis of chemical reaction kinetics [59] . It states that a process proceeds at a rate that is proportional to the availability of each of its inputs . Although oocyte development and maturation involves a number of discrete steps and processes [48] , [60] , [61] , for simplicity , we subsume them into a single state . This mathematical abstraction simplifies the subsequent calculations and reflects the difference between a fine-grained molecular model and a macro-level approach . We represent the number of oocytes , that are generated de novo , as O . Experimental data suggest that the total number of germ cells in adults [62] and the rate of oocyte production [48] are constant . Therefore our model treats the rate at which oocytes are generated as a constant , subject to saturation that prevents O from increasing beyond an upper limit established by gonad size [48] . Together , these assumptions define the rate of oocyte creation ( Figure 3B ) , ( 1 ) where kg is a rate constant describing the generation of O , and ks is a rate constant pertaining to the carrying capacity of the gonad . Hermaphrodites of the standard laboratory strain ( Bristol or N2 ) of C . elegans produce approximately 300 sperm during development before the germline irreversibly transitions to oogenesis [34] . Because animals produce oocytes continuously until their cache of sperm is depleted , the number of sperm determines the overall fecundity [34] . A dedicated mechanism communicates the presence of sperm to the developing oocytes . Sperm release major sperm protein ( MSP ) into the proximal gonad [63] , where it induces meiotic maturation of the proximal oocyte [31] , [48] . Concomitantly , MSP promotes sheath cell contraction , leading to ovulation [32] . As the oocyte is pulled into the spermatheca , fertilization takes place [64] . After the spermatheca , the embryo passes to the uterus where it completes the first several cell divisions before being laid [24] . The dynamics of egg-laying are known to be bursty , but the time intervals between these bursts are typically on the order of minutes [65] , much shorter than the time intervals at which we counted eggs . Therefore we need not consider these dynamics in our model . The reproductive rate , while approximately constant early in adulthood , decreases as the animals age [66] . This decline in reproductive function likely has multiple causes . In the first several days of reproductive maturity it likely reflects the decreasing number of sperm and the coupling of ovulation to sperm number [63] , because mating during this period can produce substantially more progeny [67] , [68] . About 5 days after the onset on reproduction , oocyte quality becomes compromised [69] , [70] , and mating of week-old hermaphrodites does not increase their brood size [68] . At lower temperatures ( e . g . , 20°C ) , within 4–5 days of reproductive maturity nearly all of a hermaphrodite's sperm have been used to fertilize eggs [34] . However , it is reasonable to expect that chronic exposure to higher temperatures will result in gamete death . While developing oocytes are likely damaged by chronic temperature stress , they can be continuously generated , therefore their destruction is difficult to decouple from a decrease in their production rate . We thus captured this process by allowing net oocyte production rate in the model to vary with temperature . These assumptions , and their related mass action kinetics , yield expressions for the rate of ovulation and the rate of sperm death , ( 2a ) ( 2b ) where Sa is the number of active sperm , is a rate constant of ovulation , and is a rate constant of sperm death . Because O rapidly achieves a steady state [48] , we simplified the model specified in Equations 1 and 2 using a quasi-steady-state approximation [71] . We found that this reformulation results in a model that captures the system dynamics equally well ( see next section and Text S1 ) . We explicitly solved the steady-state mass balance equation to obtain ( see Text S1 ) . This allowed us to express the dynamics of the system using a smaller subset of parameters . In the interest of parsimony , we used the parameter kmax to summarize the intrinsic maximum rate of oogenesis , ( 3 ) where depends weakly on Sa , and can be treated as a constant ( see Text S1 ) . Together , these assumptions can be combined into a system of mass balance equations describing the dynamics of C . elegans reproduction , ( 4a ) ( 4b ) In our experiments , we observed substantial variability in both the overall fecundity and the dynamics of egg-laying among individuals . We hypothesized that this variability arises from differences in the intrinsic capacity ( kmax ) for oogenesis and the number of sperm produced by each animal , both of which we surmised are normally distributed ( Figures 2A , B , S2 , S3 , Text S1 ) . The rate of sperm production is approximately constant over time [72] , and high sperm count is associated with delayed onset of oogenesis [67] . To capture this , when simulating our model , the number of sperm of each individual and the timing of the onset of embryo production were determined by the same variable drawn from a normal distribution . Recalling the heterogeneity of brood sizes at higher temperatures ( Figure 2 ) , we reasoned that the fraction ( δ ) of animals that exhibit a non-robust reproductive output varies with temperature , and treated δ as a free parameter . Although the mean-field behavior of our model can be analytically solved ( Text S1 ) , we solved it numerically . We used maximum likelihood estimation [73] to determine the kinetic parameters for our model . Interestingly , our estimates of kmax were substantially different for the two classes . We used time-resolved , densely sampled egg-laying curves collected at 20 , 25 , and 29°C ( Table 1 , Figure 2 ) to train our model for both the robust and non-robust classes of animals . Noting the narrow range of relevant temperatures , we hypothesized exponential dependence of the model parameters on temperature . Because δ is only nonzero at 28°C and above , we used curves collected at 20 , 28 , and 29°C to estimate its value more robustly . The estimated coefficients of these exponential functions ( Figure 4A–C ) result in model predictions that closely recapitulate the empirical data ( Figure 4D ) . To obtain Equation 3 , we surmised that the dynamics of oocyte development are steady-state [48] , and the number of developed oocytes O is constant ( also see Text S1 ) . To ensure that this approximation does not lead to an overly simplistic model that fails to capture aspects of reproductive dynamics , we evaluated predictions for two distinct model formulations . The first assumed that O reaches a quasi-steady-state according to Equation 3 . This simplified model is fully described in Equation 4 . The second was more complicated , explicitly accounting for oocyte generation and development ( Equations 1 and 2a ) and allowing O to vary . Only subtle quantitative differences existed in the predictions of these two models , justifying the use of the parsimonious version ( Figure 5A ) . To ensure that the parsimonious model ( Equation 4 ) does not omit other details that could improve the description of the system , we constructed an alternative model with an additional component that plausibly exists in the reproductive system: oocyte death . In a model that explicitly included discrete states for dead oocytes ( Od ) ( Figure 5B ) , the rate of oocyte accumulation becomes , ( 5 ) where is the rate of oocyte death and is the rate constant of oocyte death . Reformulating Equation 5 , we obtain , ( 6 ) where . Because this expression is mathematically equivalent to Equation 4a , it is difficult to differentiate between this model that accounts for oocyte death from the more parsimonious model formulated above ( Equation 4 ) . Our modeling framework provides the basis for predicting the behavior of animals treated under different conditions and having different genetic backgrounds . As a first test , we generated predictions of the dynamics of reproductive output following chronic temperature shifts conducted under the same experimental protocol that was used to train the model , but at three different temperatures . At 23 , 28 , and 30°C , we observed a close correspondence between predicted values and experimental results ( Figure 6 ) . Predictions were obtained using parameters estimated from the training data ( Figure 4 ) ; the only additional information that was specified was the temperature to which the animals were exposed . Importantly , in addition to the quantitative matches obtained for the population means , we also observed a correspondence between predicted and experimentally measured animal-to-animal variances of brood sizes . As a second test , we probed the reproductive dynamics of two mutants , tra-3 ( e2333 ) [74] and cdc-48 . 1 ( tm544 ) [75] , that produce different numbers of offspring than the wild-type N2 strain ( Table S2 ) . In our experimental paradigm , at 20°C these two mutants produced 437±40 and 238±115 progeny , respectively . At least two lines of evidence suggest that availability of sperm is the limiting factor in C . elegans reproduction . First , self-fertile hermaphrodites continue laying unfertilized eggs once their cache of sperm becomes exhausted [34] , [76] . Second , hermaphrodites that are mated to males generate up to four times the number of progeny as their unmated counterparts because male ejaculate provides many more sperm than the number produced by a hermaphrodite [67] , [77] . Relevantly , the cdc-48 . 1 ( tm544 ) mutant animals lay approximately as many eggs as the wild type , but a substantial fraction of these oocytes are not fertilized [75] . We therefore reasoned that the number of progeny of individual animals accurately reflected the number of sperm they produced . Using these inferred sperm counts and the model parameters estimated from the training data ( Figure 4 ) , we predicted the dynamics of the reproductive output of the two mutants . At 20 and 25°C , predictions for the cdc-48 . 1 mutants matched the experimental results , as did predictions for the tra-3 animals at 20°C ( Figures 7A , B ) . At 25°C , however , the tra-3 mutants laid fewer embryos than predicted by our model ( Figure 7B ) . We investigated the plausible causes of this discrepancy . At 20°C the embryos of both the wild-type N2 and tra-3 animals were arranged in an orderly fashion within the uterus ( Figure 7C , D ) . At 25°C ( Figure 7E ) the embryos in wild-type animals were more numerous than at 20°C , but this effect was far more pronounced in the tra-3 mutants , which had retained embryos that were older than the age at which they are typically laid ( Figure 7F ) . The number of embryos retained by individuals correlated with the sperm count , such that retention in the tra-3 animals was substantially higher than in the wild-type ( Figure 7G ) . We interpreted this as an indication that our model over-predicted the number of eggs laid because it did not consider the accumulation of eggs in the uterus and its possible consequences . The total number of eggs laid and retained in the uterus of the tra-3 animals at 25°C was indistinguishable from that in the wild-type N2 animals under the same conditions . In contrast , at 20°C tra-3 mutants produced nearly 50% more offspring ( 437 vs . 302 ) reflecting a greater number of sperm . Together , these results suggest that a higher aggregate egg production rate at 25°C results in higher egg retention which causes a mechanical impediment to the passage of eggs and therefore disrupts reproduction . The accumulation of embryos inside the uterus led to a “bagging” phenotype [78] and eventual hatching within the parent ( Figure 7H , Table S3 ) . Significantly , the bagging phenotype of the tra-3 mutants was completely suppressed by an egl-19 ( ad695 ) mutation that causes constitutive egg-laying [79] . This suggests that the mechanical elements of the egg-laying apparatus were compromised by chronic heat stress , serving as a physical impediment to achieving the maximum rate of egg-laying and , therefore , the highest brood size given the number of available sperm .
We developed a macro-level , parsimonious model that , although it incorporates only a few of the known elements of the reproductive system of C . elegans , is sufficient to make quantitatively accurate predictions of the dynamics of reproduction under stress . Using detailed , time-resolved experimental data , we demonstrated that the model predicts reproductive dynamics of animals in a number of environmental and genetic backgrounds . The molecular details underlying reproduction undoubtedly are numerous and complex . Specifically , large numbers of genes are associated [80] with the following reproduction-related Gene Ontology terms: fertilization ( 23 ) , oviposition ( 394 ) , oocyte ( 60 ) , oogenesis ( 179 ) , and sperm ( 52 ) . We have shown that a minimal model of a process can be sufficient for capturing system dynamics . We were able to infer a minimum set of essential elements that are sufficient to describe the temperature-dependent dynamics of reproduction in C . elegans . The reproductive systems of individual C . elegans worms exhibited a heterogeneous response to temperature stress , manifested as more variable brood sizes . Several possible explanations can account for this phenomenon . Animals at higher temperatures might have an increased probability of a discrete failure event . This could plausibly give rise to two populations of animals—some reproducing at a relatively high rate , similar to ( although slower than ) that at lower temperatures—and some that have a broken reproductive system . Under this scenario , the combined distribution of brood sizes at a given temperature could be described as a mixture of a normal distribution , corresponding to robustly reproducing animals , and an exponential distribution , reflecting waiting time to a failure event ( Figure 8A ) . Alternatively , the observed heterogeneity could be indicative of a dichotomy of reproductive strategies ( Figure 8B ) . Phenotypic switching—the responsive or stochastic shift between two discrete modes of behavior—has been shown to be an important part of adaptation to environmental stress in unicellular organisms [81] , [82] , [83] . Our results are consistent with the possibility that animals adopt aggressive or conservative strategies by altering the rates of oocyte development . At higher temperatures , more worms shift from aggressive ( fast ) to conservative ( slow ) egg-laying behavior . In our model , the primary difference between these populations is kmax , the initial egg-laying rate before signal from sperm becomes rate limiting . It is conceivable that the observed heterogeneity could represent a bet-hedging approach in which some individuals in a population continue reproducing “expecting” conditions to become favorable soon , while others delay reproduction until conditions improve . Such a strategy may be advantageous for coping with the ever-changing environment [84] . Our results serve as a demonstration of the utility of macro-level modeling for understanding complex biological systems . We can envision the application of similar models to the understanding of other phenomena that involve mass transfer . Examples would include gas exchange in the respiratory system , filtration in the excretory system , and nutrient extraction in the intestine . More broadly , any system that consists of an orderly transition between defined compartments or states could be amenable to the type of analysis presented here . This would include development and tumorigenesis . Considerable , time-resolved experimental data are essential as are the knowledge of the initial conditions and the understanding of at least some interactions within the system . We believe that macro-level modeling can serve as a useful complement to more fine-grained approaches in the analysis of complex biological systems .
Caenorhabditis elegans Bristol wild-type N2 , as well as CB4419 ( tra-3 ( e2333 ) ) [74] , FX544 ( cdc-48 . 1 ( tm544 ) ) [75] , DA695 ( egl-19 ( ad695 ) ) [79] , and YR27 ( egl-19 ( ad695 ) /tra-3 ( e2333 ) ) mutants , were maintained at 20°C as described in Brenner [85] . CB4419 ( tra-3 ( e2333 ) ) is an allele of tra-3 that is not temperature sensitive and does not affect the somatic gonad [67] . This allele causes a delay in the switch from spermatogenesis to oogenesis and a concomitant increase in the number of sperm . FX544 ( cdc-48 . 1 ( tm544 ) ) is a deletion mutant of a gene that regulates tra-1 . In this mutant , the switch from spermatogenesis to oogenesis is premature and fewer sperm are produced [75] . DA695 ( egl-19 ( ad695 ) ) is a mutation in the α1 subunit of an L-type voltage-activated Ca2+ channel that causes myotonia and constitutive egg laying [86] . Mutant strains were obtained from the Caenorhabditis Genetics Center . To construct the double mutant , YR27 ( egl-19 ( ad695 ) /tra-3 ( e2333 ) ) , CB4419 males were mated to DA695 hermaphrodites . The progeny were allowed to self and double mutant candidates were selected on the basis of empty uterus and large brood size phenotypes . The genotype was confirmed by sequencing . To standardize the environment for nematode development , we prepared 60 mm NGM agar plates 48 to 62 h prior to experiments using 10 mL of medium per plate and seeded these plates with 100 µL of saturated OP50 culture 24 h before nematodes were transferred onto the plates . We prepared synchronized cultures of L1 larvae using hypochlorite treatment of gravid hermaphrodites [87] . The liberated eggs were left on a shaker for 18±3 h at room temperature ( 23–24°C ) in M9 buffer—sufficiently long for the population to arrest at the L1 molt . The L1 larvae were placed onto the plate in contact with bacteria to synchronize the sensing of food and the termination of L1 diapause . This transfer of L1 larvae corresponds to 0 h in relation to L1 arrest and serves as the benchmark for timing in the rest of the experiments . The developing nematodes were maintained at 20°C and microscopic examination of worms at 44 h post-L1 arrest confirmed that more than 92% of nematodes were late-L4 . Since a thin bacterial lawn with a small area increases both the density and visibility of laid eggs , we seeded new NGM plates with 5 µL of 1∶1000 dilution of saturated OP50 culture in Lysogeny broth ( LB ) 24±2 h after L1 arrest . We transferred single nematodes to the new NGM plates 1–2 h before the temperature shift . The time designated for temperature shift was determined for each strain by enumerating eggs in the proximal gonad and fertilized embryos in the uterus . At 48 , 50 , 52 and 54 hours post L1 arrest , we examined twenty-five worms from each strain and counted the number of mature oocytes in the anterior and posterior gonad arms as well as the number of fertilized embryos in the uterus . Compared to N2 , FX544 and CB4419 animals were delayed about three hours but otherwise appeared normal . The plates were moved into incubators at the experimental temperature shortly after the nematodes reached young adulthood: 48 h for N2 , and 51 h post-L1 arrest for FX544 and CB4419 mutants . We measured temperature in each of the incubators with recording thermometers and discarded any time courses in which fluctuations were greater than 1°C . We counted the total number of embryos on a plate manually using a dissecting microscope . We measured time courses at 2 h intervals for the first 12 h . For longer time courses at lower temperatures ( 20 and 25°C ) , we collected additional measurements every 12 h until egg-laying had ceased . To avoid unnecessary and possibly confounding temperature fluctuations for the time points recorded at 2 h intervals , we used new animals for each time point and discarded the plates after the number of eggs had been counted . To avoid the accumulation of offspring for time points recorded at 12 h increments , we removed the nematodes from the incubator , transferred them to new plates and returned them to their incubators within 10±5 minutes of their removal . Experiments for each temperature were replicated on different days at least four times with at least one experiment in both the Morimoto and Ruvinsky laboratories . Thermometers between laboratories were within 0 . 1°C . Analysis of the individual trials suggests that small variations in developmental timing at the onset of stress contribute significantly to the variation in the total eggs laid . Populations of nematodes were synchronized as described above with the following notable exceptions: ( i ) the worms were not transferred onto new plates before exposure to stress conditions; ( ii ) we stressed populations of 20–40 animals instead of using plates with single nematodes; ( iii ) we seeded the plates used for developing worms with 5 µL of 1∶1000 dilution of saturated OP50 culture instead of saturated OP50 culture . Viability and motility were assayed at 12 h increments by removing a different set of animals from the incubator at each time point , completing the measurements at room temperature , and discarding the worms . We touched animals with platinum wire to assess if they were motile or dead . Animals were scored as motile if they crawled at least one body length after gentle touch . Animals were scored as dead if they were unresponsive to touch and did not exhibit pharyngeal pumping . These experiments were replicated on different days at least three times in the Ruvinsky Lab for each temperature . An average of 164 and 235 animals were used for each time point at 30 and 31°C , respectively . Time points were counted by multiple lab members to limit operator error . Synchronized cultures of N2 , CB4419 , FX544 and DA695 were prepared and plated as for the egg-laying protocol described above . Twenty worms were singled for each temperature tested . At t = 0 ( 48 hours post L1 arrest for N2 and DA695 and 51 hours post L1 arrest for CB4419 and FX544 ) , the twenty plates were shifted to 20 , 25 or 28°C . After twenty-four hours of heat stress , the adult hermaphrodites were dissolved on the plate in 10 µL of alkaline hypochlorite solution and the eggs that had been retained in the worm were counted . Two trials were conducted for each strain . We used the permutation test [88] , a bootstrapping procedure , to compare distributions of brood sizes ( Figure 2 ) and coefficients of variation between brood sizes at different temperatures ( Figure S4 ) . For each comparison , the bootstrapping was repeated 106 times . The estimated probability that the data could be observed , given the null model is , is the fraction of bootstrapped results that is at least as extreme as d . | Dynamic response to changing conditions in the environment is an essential property of all biological systems . Whereas extensive research over the last several decades has elucidated numerous molecular responses to environmental stress , there is much less known how these translate into organismal-level responses . Two types of modeling approaches are often used to bridge this gap . Fine-grained models seek to explain phenomena as resulting from interactions of large numbers of individual components . This approach demands a highly detailed knowledge of the underlying molecular mechanisms and has an inherent difficulty in crossing spatial scales and organizational hierarchies . As an alternative , here we present a macro-level model of reproduction in C . elegans that uses fundamental engineering principles , together with a limited set of experimentally derived facts , to provide quantitatively accurate predictions of performance under a range of physiologically relevant conditions . One important finding is that individuals within a population display considerable heterogeneity in their response to heat stress . This could be a reflection of different strategies for coping with the ever-changing environment . Our study further demonstrates that dynamic behaviors of systems may be determined by a small number of key components that lead to the emergence of organismal phenomena . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"systems",
"biology",
"chemistry",
"biology",
"computational",
"biology",
"chemical",
"engineering",
"genetics",
"and",
"genomics",
"engineering"
] | 2012 | Macro-level Modeling of the Response of C. elegans Reproduction to Chronic Heat Stress |
Genome-scale metabolic reconstructions are currently available for hundreds of organisms . Constraint-based modeling enables the analysis of the phenotypic landscape of these organisms , predicting the response to genetic and environmental perturbations . However , since constraint-based models can only describe the metabolic phenotype at the reaction level , understanding the mechanistic link between genotype and phenotype is still hampered by the complexity of gene-protein-reaction associations . We implement a model transformation that enables constraint-based methods to be applied at the gene level by explicitly accounting for the individual fluxes of enzymes ( and subunits ) encoded by each gene . We show how this can be applied to different kinds of constraint-based analysis: flux distribution prediction , gene essentiality analysis , random flux sampling , elementary mode analysis , transcriptomics data integration , and rational strain design . In each case we demonstrate how this approach can lead to improved phenotype predictions and a deeper understanding of the genotype-to-phenotype link . In particular , we show that a large fraction of reaction-based designs obtained by current strain design methods are not actually feasible , and show how our approach allows using the same methods to obtain feasible gene-based designs . We also show , by extensive comparison with experimental 13C-flux data , how simple reformulations of different simulation methods with gene-wise objective functions result in improved prediction accuracy . The model transformation proposed in this work enables existing constraint-based methods to be used at the gene level without modification . This automatically leverages phenotype analysis from reaction to gene level , improving the biological insight that can be obtained from genome-scale models .
The advances in high-throughput sequencing techniques and genome annotation methods have enabled the construction of genome-scale models for hundreds of organisms [1] . At the same time , the constraint-based framework , with its wide variety of methods , has become a widely used tool to perform in silico experiments and predict cellular responses to different kinds of genetic and environmental perturbations [2 , 3] . Studies using constraint-based models cover a wide range of applications from biomedical research to industrial biotechnology , including the study of cancer metabolism [4] , drug target discovery for cancer cell lines [5] and pathogenic microorganisms [6] , and the design of microbial cell factories [7] and synthetic microbial communities [8] . Understanding the complex relation between the genotype and phenotype of an organism is a fundamental part of systems biology research . Unlike statistical approaches such as genome-wide association studies ( GWAS ) [9] , genome-scale reconstructions provide a mechanistic link between genotype and phenotype . The first component of this link is a list of gene-protein-reaction ( GPR ) associations that determines the set of metabolic reactions encoded in the genome . Another component is the stoichiometric matrix representing these reactions . This matrix is at the core of every constraint-based method , allowing the computation of the metabolic phenotype as described by metabolic fluxes at steady-state . Navigating back and forth in the space of genotype-to-phenotype relationships is hampered by the complex association between genes , enzymes and reactions . From the perspective of the central dogma of biology the simplest genetic mechanism is: one gene—one protein—one function . However , most GPR associations in a genome-scale metabolic network are quite complex due to the presence of enzyme complexes ( multiple genes—one protein ) , isozymes ( multiple proteins—one function ) and promiscuous enzymes ( one protein—multiple functions ) . Since most constraint-based methods do not explicitly account for GPR associations , they can only provide analysis at the reaction level . For instance , simulating a steady-state flux distribution predicts the rates of all metabolic reactions for a given phenotypic state , but fails to elucidate the contribution of individual genes/enzymes to that phenotype . GPR associations , typically implemented as Boolean rules , can be used to interpret the results of constraint-based analysis in an ad-hoc fashion . This is the case in rational strain design , where optimization procedures are used to find optimal interventions to maximize the production of a given compound [7 , 10] . With a few exceptions [11–13] , such methods can only compute reaction-based modifications that must be translated to gene-level modifications a posteriori , without guarantee that the optimality of the predicted phenotype is preserved . Undesired side-effects may arise if any of the target reactions involve promiscuous enzymes . In this work , we present a model transformation that generates a stoichiometric representation of GPR associations that can be directly integrated into the stoichiometric matrix . We show that the results obtained with the transformed model are consistent with those obtained from reaction-level models , and highlight the advantages of performing different kinds of analysis at the gene level . We also propose new variants of existing methods that take advantage of this representation to formulate gene-wise objective functions and test their predictive ability using experimental datasets .
Gene essentiality analysis consists of the identification of conditionally lethal gene deletions [30] . This type of analysis can be used to find drug targets for pathogenic microbes [31] and particular types of cancer cells [5] . It can also be used to improve model reconstructions by comparison with experimental data , and to exclude undesirable gene deletions from the search space of strain design algorithms . Gene essentiality analysis is usually performed by simulating the knockout of each gene in two stages , which requires evaluating the respective GPRs followed by FBA simulation to test the model for growth . With the extended stoichiometric matrix , gene essentiality analysis can be directly performed by flux variability analysis ( FVA ) ( see methods ) . For each gene , the flux range of the respective enzyme usage variables indicates the minimum and maximum amount of flux that can be carried in the given experimental conditions ( S1 Fig ) . In this case , any gene with a minimal enzyme usage above zero for a given minimal biomass production is essential . This approach can also be generalized to find synthetic lethal pairs by systematic computation of the minimum sum of fluxes of all pairs of enzyme usage variables . This analysis is more informative then traditional determination of essentiality ( binary test ) as it reveals the minimal ( and maximal ) flux that can be carried by each enzyme . With this approach one can also determine “blocked” genes ( i . e . genes encoding enzymes that cannot be used under any conditions ) , which can be used to guide the model reconstruction process . Furthermore , one can use the shadow price and reduced costs information to analyse the sensitivity of the results with respect to internal ( biological ) and external ( environmental ) constraints . For instance , the non-zero shadow prices for an essential gene represent the set of precursor metabolites that cause the essentiality , whereas the reduced costs of exchange reactions represent the effect of changing the medium composition with regard to essentiality . Random sampling of the flux solution space is a suitable strategy to analyse all possible physiological states described by a model [32] . Like FVA , it is an unbiased method to describe the flux solution space . However , while FVA only describes the admissible flux range for a given reaction , random flux sampling generates a probability distribution for each reaction , providing insight into the shape of the solution space . Flux sampling has been used to analyse global properties of metabolic networks [33] and to determine flux variation in perturbed conditions [34] . One limitation of flux sampling is that it does not account for the flux load distribution between isozymes or the overall flux carried by promiscuous enzymes . With the extended representation , it is now possible to analyse flux sampling results at the gene/enzyme level . An illustrative example for a model of core metabolism of E . coli [35] is presented ( S2 Fig ) . Flux sampling results are compared for two conditions: a wild-type phenotype and a succinate producing phenotype ( see methods ) . It is possible to observe significant differences between both phenotypes at the gene level . There is an increased flux in enzymes involved in lower glycolysis and the glyoxylate shunt , and an overall decrease of flux for enzymes in the pentose-phosphate pathway and the respiratory chain . Note that one can observe gene level differences that would not be captured by purely reaction-based sampling , such as the different utilization of Lpd relative to other enzyme subunits ( AceE , AceF , SucA , SucB ) given its simultaneous participation in different enzyme complexes . Flux sampling at the gene level can be used to guide rational strain design , since non-overlapping sampling distributions for a given gene between wild-type and the desired mutant indicate that the flux carried by the respective enzyme must necessarily change . We compared these results with those obtained by strain design methods that account for modulation of gene expression [25–28] . Some of the most significant changes observed ( deletion of sdh* and overexpression of frd* , ppc , and aceA ) are commonly proposed interventions to increase succinate production . It is also possible to observe some extent of agreement between our sampling results and gene expression measurements of succinate producing mutants [36] , most notably the down-regulation of aceE , aceF , icd , pykA and pykF . The continuous improvement of high-throughput techniques to measure different kinds of omics data has fostered the development of constraint-based methods that make use of these data to improve predictions . In a recent work , we evaluated several methods for integration of transcriptomics ( and proteomics ) data into constraint-based simulations , and observed that none of the methods resulted in consistent improvement of flux predictions compared to simple FBA simulation under the assumption of optimal growth and parsimonious enzyme usage [37] . This limitation arises from the underlying assumption of proportionality between gene expression and reaction rates , which does not seem to be generally valid [38 , 39] . It seems natural to reformulate some of these methods to take advantage of the flux simulation at the enzyme level . In this work , we propose gene-wise reformulations of two commonly used methods , GIMME and E-Flux [40 , 41] . In the reformulated versions , the expression level of a gene is mapped to its respective enzyme usage variable ( see methods ) . The original and reformulated versions of the methods were evaluated using two experimental datasets containing transcriptomics and fluxomics data [22 , 42] ( see methods ) . Similarly to our previous study , the results reveal that none of the transcriptomics-based methods outperforms pFBA ( Fig 6 ) . However , as observed earlier , gene-pFBA shows better performance than pFBA for the Ishii dataset . The gene-wise version of GIMME is generally more accurate than the original version in both datasets . This improvement can be attributed to the fact that the gene-wise formulation is less affected by the lack of correlation between gene expression and reaction rates . No improvement could be observed for the gene-wise version of E-Flux . Elementary mode analysis provides an unbiased description of the flux solution space of a metabolic network by determining all minimal pathways that can operate at steady-state , so-called elementary flux modes ( EFMs ) [43] . Elementary mode analysis reveals multiple properties of metabolic networks , including pathway yields , reaction usage frequency , and correlated reaction sets [44] . Common applications include analysis of cellular robustness [45] , detection of fragility points in metabolic networks as potential drug targets [46] , and elimination of undesired phenotypes to design optimal cell factories [47] . Since EFM computation does not account for GPR associations , they do not entirely reflect the topology of a metabolic network , disregarding that a promiscuous enzyme is a common link between different pathways and that isozymes provide alternative routes within the same pathway . Our stoichiometric representation of GPRs solves this problem by explicitly accounting for this complexity in the computation of EFMs . This concept is illustrated in Fig 7 . Although EFM computation algorithms differ with regard to specific implementation details , the manipulation of support vectors is a common denominator . Support vectors are binary representations of the minimal set of reactions included in an EFM . With the network transformation , the artificial enzyme usage reactions become part of the support vector of EFMs , being automatically computed by any EFM computation algorithm . This extended support vector contains a gene-wise representation of each EFM , denoting the genes that participate in the given pathway . We applied this analysis to a simplified central carbon model of E . coli ( Fig 8a ) . The model contains a total of 12 EFMs . After transformation the number of EFMs raises to 11085 . This drastic increase is caused by splitting isozymes into separate reactions , which leads to a large combination of possible routes . Fig 8b shows the gene participation in the set of EFMs . It can be observed that pgk , gapA , and eno participate in every pathway . These would be the best targets in a drug design application . On the other hand , pgi has the lowest participation ( 16 . 4% ) . The deletion of this gene would cause the least impact in the network . We also compared the frequency of each reaction in the original model and the transformed model ( Fig 8c ) . There is an overall increase in the frequency of reactions in the pentose-phosphate pathway due to the alternative routes created by the presence of isozymes . The frequency of glycolytic reactions remains the same , with the exception of PGI with a significant decrease ( from 67% to 16% ) . These results show that accounting for GPRs can shed a different perspective on the relative importance of different reactions , with a potential impact in methods that search for the most important pathway disruptions to block undesired phenotypes [46 , 47] . Although the increase in the number of EFMs hampers large-scale EFM computation , this approach is still amenable to the application of EFM-based methods that do not require complete enumeration of the full EFM set ( see discussion ) . Designing optimal cell factories for production of industrially relevant compounds is one of the most common applications of constraint-based modeling . Genome-scale models can be used to guide rational strain design by predicting the phenotype of mutant strains , which can be iteratively improved until economically viable product yields , titers and productivities are attained . The countless combinations of manipulations that could be tested require the implementation of powerful optimization methods to search the genetic design space [7 , 10] . Although a large number of methods ( ∼50 ) have been published so far , very few allow gene-based modifications [11–13] . The vast majority of methods determine optimal sets of reaction-based modifications ( deletions or up/down-regulations ) that must be a posteriori translated into gene-based designs for in vivo implementation . Given that enzyme promiscuity can affect a major fraction of the reactions in a model ( Fig 2 ) , it can be expected that many reaction-based designs will result in undesired side-effects when implemented at the gene level . MCSEnumerator is a recently published method that enumerates all minimal sets of reaction deletions up to a given size , so-called constrained minimal cut sets ( cMCSs ) , that are guaranteed to couple product formation to growth [48] . This method represents a breakthrough in the field , allowing unprecedented enumeration of the design space at the genome scale . We reproduced the results presented by the authors for growth-coupled ethanol production ( scenario 1 ) , and performed a deeper analysis of the feasibility of the strain designs when mapping the reaction-based solutions to gene-based ones ( Fig 9 ) . Given that any gene encoding a subunit of an enzyme complex can be deleted to disable the respective function , the number of potential designs significantly increases when converting reaction to gene deletions ( Fig 9c ) . This mainly results from the presence of reactions catalyzed by multiple complex isozymes . For instance , formate hydrogen lyase ( FHL ) can be catalyzed by two different complexes , with 11 and 7 subunits each , resulting in 77 possible combinations of gene deletions to disable this reaction . Other notable cases include the PTS system ( 57 combinations ) and ATP synthase ( 44 combinations ) . It can also be observed that the total number of required gene deletions can be significantly larger than the respective number of reaction deletions . For instance , a strain design of 4 reaction deletions may require up to 13 gene deletions ( Fig 9d ) . In order to test the feasibility of each design , accounting for possible side-effects , we calculated the actual set of reactions effectively disabled by the gene deletions required to implement a given cMCS . We then evaluated each phenotype and observed that only a small fraction of the original set of solutions ( ∼7% ) are valid with respect to the original production constraints ( Fig 9e ) . This drastic effect is mainly caused by the deletion of highly promiscuous enzymes ( such as those involved in transporters ) , which can result in the deletion of hundreds of side-effect reactions ( Fig 9f ) . The shortcomings of reaction-based design can be avoided by directly searching for gene-based designs . We applied MCSEnumerator to the transformed model and computed all minimal gene-based cut sets up to 8 deletions ( Fig 10 ) . It can be observed that the total number of gene-based designs is now much lower . In this case , all designs are necessarily feasible since all potential side-effects are implicitly accounted for . Nonetheless , we confirmed the feasibility of each design by testing the respective reaction deletions in the original model . Note that the total number of reaction-based designs is actually lower , since different gene cut sets generate the same reaction deletions ( Fig 10c ) . Furthermore , it can be observed that the number of deleted reactions is generally higher than the number of respective gene deletions without compromising the feasibility of the strain design ( Fig 10d ) . Finally , we tested cRegMCSs , a recent extension of MCSEnumerator that accounts for reaction up/down-regulation [29] . With our approach it is possible to apply constraints directly at the gene level , correctly accounting for the limitations discussed earlier and without any modification to the original method . Using a core metabolic model for E . coli and the same production goals as before , thousands of designs were found with as few as 3 gene manipulations ( see S3 Fig ) .
The integration of GPRs directly into the stoichiometric matrix enables bidirectional analysis between the gene and reaction levels . In one direction it is possible to observe the impact of gene perturbations on reaction fluxes . In the other direction one can perturb the environmental conditions and observe the required adaptations at the gene level . The complexity of GPR associations and their evolution has been recently analysed considering the role of environmental adaptation in driving enzyme specificity [59] . A recent reconstruction of the underground metabolism of E . coli revealed an even larger number of metabolic reactions available in the cell as a result of enzymatic side activities , playing an important role in the fitness landscape of the organism [60] . Our framework provides a mechanistic link between genotype and phenotype and should facilitate the development of new methods to integrate multi-omics datasets into genome-scale models , as well as methods to integrate metabolism with other biological processes . In this work , we explored the reformulation of previously published simulation methods with gene-wise constraints . It would be interesting to explore other suitable applications , such as the formulation of simulation methods that account for enzyme production costs [61 , 62] . A new generation of genome-scale models and simulation methods is on the rise [63] . This includes genome-scale models that account for gene expression and protein production [18 , 64 , 65] , models that account for protein structure [66] , and methods that predict the effect of genetic variation in protein function [67] . While such detailed models are not readily available for every organism , our method provides a suitable approach to leverage existing models to a new level . All the source code , models , and generated data are publicly available in the following repository: https://github . com/cdanielmachado/GPRTransform .
Unless otherwise stated , all simulations were performed using the iAF1260 genome-scale metabolic model for E . coli [14] and the Gurobi solver ( version 6 . 5 ) . The core metabolism version of this model [35] was used for random flux sampling and strain design with cRegMCS [29] . The model transformation to explicitly include GPRs in the stoichiometric matrix ( as exemplified in Fig 1 ) is defined as follows . Let S be the original stoichiometric matrix , v the steady-state flux vector ( after decomposition of reversible reactions and isozyme-catalyzed reactions ) , and ub the respective vector of flux bounds , such that S ⋅ v = 0 and 0 ≤ v ≤ ub . The extended stoichiometric matrix S′ , flux vector v′ , and flux bounds ub′ are defined as: S ′ = S0SgprIk v ′ = v u u b ′ = u b + ∞ where Sgpr is the stoichiometric representation of GPRs ( with si , j = −1 if gene i participates in reaction j ) , Ik is the identity matrix for k genes , and u is the enzyme usage vector . This transformed model can be readily used by any kind of constraint-based method with the general form: min/max f ( v′ ) s . t . S′·v′=00≤v′≤ub′ where f is a given objective function . The gene-wise reformulations differ from the original methods by expressing the objective functions and genetic constraints using the enzyme usage variables instead of reaction fluxes , and are defined as follows . The methods to predict mutant phenotypes ( pFBA [17] , MOMA [19] , linearMOMA , gene-pFBA , gene-MOMA , gene-lMOMA ) were tested and compared with fluxomics data from the Ishii2007 dataset [22] that includes 24 single deletion mutants in chemostat cultivation at D = 0 . 2 h−1 . In each case , the glucose uptake rate is constrained to the experimental value . The predicted fluxes are then compared to the experimental values , and the normalized prediction error is calculated as follows: error = | | v sim - v exp | | | | v exp | | where vexp are the experimental fluxes , vsim are the simulated fluxes for the experimentally measured reactions , and the vector norm is the l1-norm ( Manhattan distance ) . The methods for integration of gene expression data were tested using the transcriptomics and fluxomics data from two multi-omics datasets for E . coli [22 , 42] . For the Ishii2007 dataset , all experimental conditions were used ( wild-type at 5 different dilution rates and the 24 deletion mutants ) . The Gerosa2015 dataset includes data from shake flask cultivation under 8 different carbon sources . In this case , we constrained the uptake rate of the respective carbon source . The prediction error for transcriptomics-based methods was also calculated as described above . However , we observed that for this kind of methods , the degeneracy of the optimal solution can influence the prediction error , hampering the reproducibility of results . To address this problem , in all the methods we add a second step that , after each simulation , determines the optimal solution with the smallest norm . Gene essentiality for each gene i was determined by flux variability analysis of the respective enzyme usage variable as follows: min/max uis . t . S′·v′=00≤v′≤ub′vgrowth≥0 . 1·vgrowthmax where v growth max is the maximal growth rate determined by FBA . Random flux sampling was performed using the artificially centered hit-and-run ( ACHR ) method of the COBRA toolbox [68] . The wild-type phenotype was sampled under the assumption of a minimum biomass yield of 90% of the maximum theoretical value . The succinate producing mutant was sampled using a minimum of 10% for the biomass yield , and a minimum of 90% of the maximum succinate production ( at 10% of biomass yield ) . Each phenotype was sampled for 10000 flux distributions with a step size of 100 jumps per sample . EFMs for the simplified glycolysis model were computed using CellNetAnalyzer version 2015 . 1 [69] . MCSEnumerator [48] was tested using the API interface from CellNetAnalyzer . We used the same model that was applied in the original publication ( a version of iAF1260 customized for anaerobic growth ) . The list of targetable reactions was also defined as in the original publication . For the gene-wise version , the transformation was applied to the model , and the list of targetable reactions was defined to be the list of enzyme usage reactions . Computations were performed with CPLEX 12 . 6 . 3 using a 6-core Intel i7 processor with 64 GB of RAM . cRegMCS [29] was tested using the API interface from CellNetAnalyzer . The problem setup was performed similarly to MCSEnumerator , except in this case the E . coli core model [35] was used due to the higher computational cost of the method . The number of regulatory steps was set to 3 levels for every gene ( except those participating in futile cycles ) . | Genome-scale models of metabolism enable the exploration of the phenotypic landscape of an organism . Unlike probabilistic approaches such as genome-wide association studies , these models describe the mechanistic link between genotype and phenotype , predicting the response to genetic and environmental perturbations . However , this connection is hampered by the complexity of gene-protein-reaction associations . In this work , we implement a model transformation method that untangles this complexity by allowing gene-wise phenotype predictions using genome-scale models . The transformed model explicitly accounts for the individual flux carried by the enzyme or subunit encoded by each gene . Previously published simulation methods are automatically leveraged by this transformation , enabling new features such as the formulation of objectives and constraints at the gene/protein level . We demonstrate the application of different kinds of analysis and simulation methods , showing in each case how the gene-wise formulation can result in higher prediction accuracy in comparison to experimental data and improve the biological insight that can be obtained from available models . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"deletion",
"mutation",
"medicine",
"and",
"health",
"sciences",
"enzymes",
"enzymology",
"simulation",
"and",
"modeling",
"mutation",
"enzyme",
"metabolism",
"genome",
"analysis",
"isozymes",
"pharmacology",
"drug",
"metabolism",
"enzyme",
"chemistry",
"research",
"and... | 2016 | Stoichiometric Representation of Gene–Protein–Reaction Associations Leverages Constraint-Based Analysis from Reaction to Gene-Level Phenotype Prediction |
Vaccinia virus A33 is an extracellular enveloped virus ( EEV ) -specific type II membrane glycoprotein that is essential for efficient EEV formation and long-range viral spread within the host . A33 is a target for neutralizing antibody responses against EEV . In this study , we produced seven murine anti-A33 monoclonal antibodies ( MAbs ) by immunizing mice with live VACV , followed by boosting with the soluble A33 homodimeric ectodomain . Five A33 specific MAbs were capable of neutralizing EEV in the presence of complement . All MAbs bind to conformational epitopes on A33 but not to linear peptides . To identify the epitopes , we have adetermined the crystal structures of three representative neutralizing MAbs in complex with A33 . We have further determined the binding kinetics for each of the three antibodies to wild-type A33 , as well as to engineered A33 that contained single alanine substitutions within the epitopes of the three crystallized antibodies . While the Fab of both MAbs A2C7 and A20G2 binds to a single A33 subunit , the Fab from MAb A27D7 binds to both A33 subunits simultaneously . A27D7 binding is resistant to single alanine substitutions within the A33 epitope . A27D7 also demonstrated high-affinity binding with recombinant A33 protein that mimics other orthopoxvirus strains in the A27D7 epitope , such as ectromelia , monkeypox , and cowpox virus , suggesting that A27D7 is a potent cross-neutralizer . Finally , we confirmed that A27D7 protects mice against a lethal challenge with ectromelia virus .
Inoculation with vaccinia virus ( VACV ) protected against smallpox , a deadly disease caused by the related orthopoxvirus , variola ( VARV ) [1] . Its success pertains to its high infectivity and thermal stability , and the strong innate and B-cell immune responses that it triggers [2 , 3] . With the eradication of circulating variola virus from the human population , large-scale vaccination efforts against smallpox were ended [4 , 5] . As a result , the general population is no longer protected against orthopoxviruses . This lack of immunity is a concern due to the zoonotic risk of orthologous strains [6] such as monkeypox virus ( MPXV ) and specific strains of the cowpox species ( CPXV ) [7 , 8] , as well as their potential use as biological weapon [9] . It is because of the latter that certain professional groups , including military personnel are still getting vaccinated . The vaccinia virus smallpox vaccine used in the eradication campaign was highly effective , but was associated with adverse side effects . The frequency of side effects was acceptable at the time where smallpox was a major health threat but is unacceptable in the 21st century . More recently a vaccinia virus clonal isolate , ACAM2000 has been used , produced using modern cell cultures [10–12] . Highly attenuated vaccinia strains such as Modified Vaccinia Ankara ( MVA ) have been available for decades [13] . Large MVA clinical trials , and clinical use , have found that MVA has an outstanding safety profile , but this is accompanied by a decreased immunogenicity resulting in the need for two immunizations . Moreover , since MVA usage was predominantly after smallpox eradication , the protective efficiency of MVA toward variola virus was not proven . Research on new orthopox vaccines continues , both in response to these concerns and as a means of understanding why the vaccinia virus smallpox vaccine is so effective . The strategies leading to today's next generation candidate smallpox vaccines are diverse: they include , but are not limited to , the use of ( i ) vaccinia immunization in the presence of a small molecule inhibitor [14] , ( ii ) DNA immunization using select immunodominant antigens [15–18] , and ( iii ) soluble poxvirus proteins [19–21] . A limited number of immunodominant antigens [22] have been linked to successful protection: intracellular mature virion ( IMV ) antigens A27 [23 , 24] , D8 [25 , 26] , F9 [27] , H3 [20 , 28] , L1 [29] and extracellular enveloped virion ( EEV ) antigens A33 [30–32] , and B5 [33 , 34] . Maximal protection is obtained with vaccines combining recombinant membrane proteins from both forms of the infectious virus ( IMV and the EEV ) . The virus encodes the seven proteins A33 , A34 , A36 , A56 , B5 , F12 , and F13 , that are specific for the EEV membrane [32 , 34–38] . Among those , A33 is a 23 kDa , homodimeric type II transmembrane that undergoes both O- and N-glycosylation ( N125 and N135 ) [32 , 36 , 39] . Both N-linked glycosylation sites are used in vaccinia but variola virus and monkeypox virus lack the equivalent N125 site [40] . A33 controls the incorporation of A36 into the EEV particle and subsequently the production of actin tails . Therefore , A33 plays an important role in effective cell-to-cell spread within the host [41–43] . A33 is also required for proper trafficking of B5 to the EEV-specific membrane and proper formation of infectious EEV [43 , 44] . A33 contains a membrane-proximal cysteine on the A33 ectodomain that forms an intermolecular disulfide bridge . However , this cysteine is not required for the production of infectious extracellular virus [45] . The crystal structure of the ectodomain of A33 revealed an unusual C-type lectin-fold domain , similar in overall architecture to several NK cell ligands [40] . Antibodies targeting EEV proteins can protect against viral challenge . For example , anti-B5 mAbs are protective against lethal orthopoxvirus challenges in multiple small animals models [33 , 46 , 47] and anti-A33 antibodies can be protective in vivo [15 , 30 , 48 , 49] . Protection mediated by anti-B5 antibodies is not mediated by conventional direct virus neutralization , as EEV is highly resistant to direct neutralization by anti-B5 antibodies , anti-A33 antibodies , and other EEV targets [46 , 50 , 51] . However , protective anti-B5 mAbs are highly efficient at facilitating complement-mediated neutralization of EEV , complement-mediated lysis of infected cells , and Fc-dependent protective mechanisms [46 , 51 , 52] . Mechanisms of protective action of A33 antibodies are similar to those of B5 [52] . Using X-ray crystallography , we mapped the A33 epitopes for three different antibodies and characterized their VACV-neutralization potential in vivo . We further identified one antibody ( A27D7 ) that binds to A33 with high affinity , even when epitope residues are individually changed to other amino acids . We further tested the MAb binding to recombinant A33 , where epitope residues were replaced with those from other orthopoxviruses and observed high affinity binding for A27D7 . In an ectromelia virus ( ECTV ) infection model , A27D7 MAb was protective . In summary this suggests , that A27D7 is cross-species protective and able to neutralize a range of different orthopoxviruses in vivo .
We generated seven anti-A33 antibodies ( A2C7 , A26C7 , A17D7 , A25D11 , A27D7 , A25F2 , and A20G2 ) by VACV immunization of mice followed by boosting with recombinant A33 protein . DNA sequencing revealed that the seven MAbs are assembled from five distinct heavy chain ( HC ) and 4 different light chain ( LC ) sequences ( S1 Table ) : While A2C7 , A17D7 , and A25F2 have distinct HC sequences , A26C7 shares the HC with A27D7 and A25D11 shares the HC with A20G2 , with only one amino acid exchange in CDR L1 ( L to F ) . The LC sequences of A26C7 and A27D7 are nearly identical , while A25D11 is almost identical to A20G2 . Both LC pairs only differ by one amino acid substitution in CDR1 . A2C7 and A25F2 are derived from the same germline genes and are highly similar , while the A17D7 LC sequence could not be determined . In summary , antibodies A25D11 and A20G2 are identical except for one amino acid difference and the same is true for antibodies A26C7 and A27D7 , resulting in a total of five distinct antibodies , of which we have complete sequences for A2C7 , and A26C7/A27D7 , A25D11/A20G2 , and A25F2 . Peptide ELISA performed with 20mer peptides overlapping by 10 amino acids and covering nearly the entire amino acid sequence of A33 , indicated that all anti-A33 MAbs bind only to recombinant A33 but not to any linear peptide ( S1 Fig ) . Thus , we concluded that the seven MAbs recognized conformational epitopes on A33 . All anti-A33 MAbs were then tested for their ability to neutralize EEV in vitro by complement dependent and independent processes ( Fig 1 ) . None of the antibodies was able to neutralize EEV in the absence of complement ( Fig 1A ) . IgG1 anti-A33 mAbs A17D7 and A25F2 at 10 μg/ml exhibited no complement mediated EEV neutralization , consistent with the inability of recruiting complement by IgG1 MAbs in general ( Fig 1B ) . However , IgG2a ( A26C7 , A25D11 , A27D7 , and A20G2 ) and IgG2b ( A2C7 ) anti-A33 MAbs at 10 μg/ml exhibited strong complement-mediated EEV neutralization in the presence of complement ( Fig 1B ) . Anti-B5 MAbs B126 ( IgG2a ) at the same concentration show comparable neutralization activity as anti-A33 mAbs of appropriate isotypes . Irrelevant IgG1 anti-DNP MAbs ( DNP ) and anti-B5 MAbs B96 ( IgG1 ) at the same concentration showed no effect ( Fig 1B ) . These data support a model where anti-A33 MAbs can neutralize EEV in the presence of complement via opsonization of the EEV particle surface . To assess the interaction between the antibody and A33 we prepared the Fab portion of all five neutralizing antibodies ( A2C7 , A26C7 , A25D11 , A27D7 , and A20G2 ) and assessed Fab binding to recombinant A33 using size exclusion chromatography ( SEC ) ( Fig 3A ) . We observed two major peaks of different molecular weight ( MW ) , indicating that the Fab’s bound to A33 with different stoichiometries . Both Fab’s A26C7 and A27D7 eluted with a lower MW when bound to A33 , while the Fab’s of A2C7 , A25D11 , and A20G2 formed a higher MW complex with A33 . The SEC data indicated that a single Fab of A26C7 and A27D7 bound to one A33 dimer , while A2C7 , A25D11 , and A20G2 each bound with one Fab to one A33 subunit ( 2 Fab’s per A33 dimer ) . We next assessed the real-time binding kinetics with the three representative antibodies A2C7 , A27D7 , and A20G2 . Intact antibodies were immobilized on biosensors and binding to A33 in solution was measured by Biolayer interferometry ( BLI ) ( S2 Table and Fig 3B ) . A33 bound with very high affinity to the antibodies A2C7 and A20G2 ( KD ~65 pM ) , consistent with the SEC data , where the A33 dimer can simultaneously bind to two separate Fab’s . This binding mode is characterized by a very slow dissociation rate . A33 also bound with high affinity to the antibody A27D7 ( KD~14nM ) , yet roughly 400-fold weaker compared to A2C7 and A20G2 . Since the A33 dissociation rates were very slow for both A2C7 and A20G2 , which made calculation of the binding kinetics less robust , we repeated the binding assay but immobilized A33 instead of the MAbs ( S2 Table ) . Both A2C7 and A20G2 showed similar binding to chip bound A33 in this reversed orientation with a ~2 to 4-fold difference in KD each ( S2 Table ) . However , MAb A27D7 showed very slow dissociation from A33 , consistent with a model in which each Fab can bind to two separate A33 dimers on the sensor tip . As a result , the apparent binding affinity increased from 14 nM ( MAb immobilized ) to beyond 1pM ( A33 immobilized ) , as no dissociation could be reliably measured . This led to A27D7 being the antibody with the highest affinity in this setting . All subsequent kinetic binding studies using A33 mutants were performed using A33 immobilized on the sensor tip . This binding orientation also mimics the antibody binding during infection , where A33 is embedded in the EEV membrane . Of note , our recombinant A33 used for structural studies contained the three mutations Leu118Met , Lys123Ala , and Leu140Met that were previously incorporated and necessary for crystallization and structure determination [40] . We used the A33 variant Lys123Ala as “WT A33” as its mutation is outside the epitope of all tested MAbs ( see below ) . When analyzed , the binding kinetics of the Leu118Met/Lys123Ala/Leu140Met mutant does not differ significantly from the WT surrogate ( Lys123Ala ) , regardless of the choice of antibody ( S2 Table ) . To analyze the binding mode of each Fab to A33 we set out to determine the crystal structure of one representative of each of the three different neutralizing MAbs bound to A33 . To increase our chances of success we started out with all five neutralizing antibodies and finally determined the crystal structures of the complexes of A33/A2C7 , A33/A20G2 , and A33/A27D7 to 2 . 3 Å , 2 . 9 Å , and 1 . 6 Å resolution , respectively ( Table 1 and Figs 4 and 5 ) . To assess the impact of A33 epitope residues in antibody binding , we have further generated A33 variants with selected mutations in the epitope of each of the three crystallized antibodies and determined the antibody binding affinity using BLI ( S2 Table and Fig 6 and S3 Fig ) The A33/A2C7 complex shows two Fab molecules symmetrically bound to the A33 dimer , consistent with our SEC data ( Fig 3A ) . Interestingly , the asymmetric unit ( ASU ) of the crystal contained half of the biological assembly ( A2C7 Fab bound to one A33 monomer ) , which dimerized along a crystallographic symmetry axis . Each A2C7- Fab molecule elicits , therefore , identical contacts with a discontinuous and conformational epitope at the membrane-distal extremity of each A33 subunit ( Fig 4 ) . The buried surface area ( BSA ) between A33 and the LC is 365 . 1 Å2 , while the HC buries a total of 334 . 1 Å2 between A33 , leading to a total BSA of 699 . 2 Å2 between Fab and A33 , which corresponds to 4 . 9% of total protein surfaces ( S4 Table ) . This is considerably smaller than the typical range found in antibody/ protein antigen complexes ( 1400–1700 Å2 BSA , [54] ) . The epitope of A2C7 appears unusually small , yet it contains eleven A33 residues that are in contact with sixteen residues from the antibody , eight from each chain . In contrast to other available VACV antigen antibody complex structures , including D8/LA5 [50 , 55] , L1/7D11 [56] and L1/M12B9 [53] , the LC appears to be as important in antigen binding as the HC . Shape correlation ( Sc ) measures the geometric surface complementarity of protein-protein interfaces and reflects their specificity [57] . Both heavy and light chains appear to bind with high specificity to the antigen ( Sc = 0 . 67 and 068 , respectively ) . For antigen-antibody interfaces , Sc values of 0 . 64 to 0 . 68 [58] have been reported , illustrating that the A2C7/A33 interaction is very specific . This interface is held together by an extensive network of hydrogen bonds and salt bridges involving every CDR loop ( Fig 5 ) . Additionally , a number of water molecules mediate H-bonds between A33 and A2C7 ( S2 Fig ) . The presence of water molecules in the heavy chain interface may explain the slightly lower specificity observed for this chain , reflected by its Sc value of 0 . 67 vs . 0 . 68 for the light chain , since waters were not included in Sc calculations . Gln173 is a major contact residue of the A2C7 epitope and intersects with of almost every CDR loop . It forms multiple hydrogen bonds via its backbone and side chain ( H1 , H2 , H3 , and L3; Fig 5 ) : The Gln173 backbone carbonyl elicits a favorable hydrogen bond with H2:Tyr50 hydroxyl , while the side chain amide group elicits a hydrogen bond with H1:Tyr35 hydroxyl . Additionally , its side chain amide group elicits favorable hydrogen bonds L3:Ser91 and with L3:Trp96 . Because its backbone is involved in the interaction , we mutated this residue to an arginine instead of an alanine , on the basis that the bulky and positively charged side chain may induce steric clashes at the interface in the mutant and will therefore result in a decreased affinity . As expected , we observed complete loss of A2C7 binding to the A33 Gln173Arg mutant ( S2 Table and Fig 6 and S3 Fig ) . Interestingly , the Gln173Arg substitution is found in ECTV suggesting that A2C7 would be unable to protect against ectromelia challenge . A33 residue Asp170 is located on a flexible loop and is contacted by both H and L chain . It forms a salt bridge with L2:Lys55 and a hydrogen bond with L1:Tyr32 . However , the Asp170Ala mutation did not affect the binding affinity of A2C7-MAb , suggesting it is not a critical binding residue . It should be noted however , that for unknown reasons , the overall binding signal is much lower compared to the wild-type A33 ( S2 Table and S3 Fig ) . The ASU of the A33:A20G2 complex contains 2 Fab molecules bound to one A33 dimer ( 1 Fab per A33 subunit ) . Although the epitopes on each A33 monomer should be identical , subtle differences have been observed between both binding interfaces and will be reported where necessary . This could be due to a subtle influence of crystal packing . A20G2 binds a similar epitope compared to A2C7 . Sixteen residues of each A33 subunit are in contact with sixteen residues of the A20G2 Fab ( S3 Table ) . Both H and L chain grasp the tip of each A33 subunit similar to a pincer , resulting in the binding of two Fab molecules per A33 dimer ( Fig 4 ) . The binding stoichiometry is consistent with the SEC data ( Fig 3 ) . The Sc value is 0 . 68 and 0 . 69 for both antibody/antigen binding interfaces , while the BSA is 802 and 864 Å2 for each of the two Fab/A33 interfaces ( S4 Table ) . As such , the A20G2 epitope is slightly larger compared to A2C7 . Despite numerous amino acid differences in the CDRs , A20G2 approaches the antigen with an overall similar topology compared to A2C7 , with the exception of CDRs L2 and H3 . This is reflected in an overall change in binding angle ( Fig 4A and 4B ) . L1 plays a major role in both complexes , as it clamps onto one side of each A33 subunit ( A33 residues Phe119 , Asp155 , Gly156 , Asn157 , Asp170 , and Ser172 , while Ser154 and Asn157 are exclusive to A20G2 ) . L3 , H1 and H2 target a similar epitope at the top outward edges of A33 , centered around residue Gln173 in both A33 molecules . Similar to A2C7 , the Gln173Arg mutation also fully abrogates A20G2 binding ( Fig 6 ) . In A2C7 , the A33:Asp170-L2:Lys50 salt bridge is replaced against a hydrogen bond in A20G2 involving L3:Tyr96 ( Fig 5 ) . This is the result of L2:Lys50 in A20G2 being replaced with L2:Leu50 in A2C7 . However , the Asp170Ala mutation did not have any impact on binding of A20G2 to A33 , similarly to the binding of A2C7 . To compare the overall binding of A2C7 and A20G2 on A33 , we superimposed both the variable regions of the L as well as H chains separately . The L chain superimposed well even at the CDRs ( RMSD = 1 . 12 Å for Cα residues 1–119 ) . However , heavy chains did not superimpose well overall ( RMSD = 3 . 73 Å for Cα residues 1–115 ) but appeared to have a high structural homology when superimposing only residues until H3 ( RMSD = 0 . 77 Å for Cα 2–94 ) . We conclude that sequence divergences of these two antibodies had little effect over most of the antigenic interface , but that it induced major local topological differences in the CDR H3 region . This is not surprising since the H3 loop is much longer in A2C7 due to a four amino acid insertion ( H3:ARQWGGAMDY ) , compared to A20G2 ( H3:ARGMDY ) . Interestingly , the A33 mutation Leu118Arg that is located on both epitopes for A2C7 and A20G2 had no effect on the A20G2 binding kinetics , while they fully impaired A2C7 , likely due to a clash with L3 , which approached A33 from a different angle . Similarly , the mutation Asp168Ala also did not impair binding of A20G2 significantly ( S2 Table and Fig 6 and S3 Fig ) . In conclusion , A2C7 and A20G2 target similar but not identical epitopes mostly due to sequence differences within H3 and L3 . A27D7 MAb engages A33 in a drastically different manner compared to A2C7 and A20G2 . A single A27D7-Fab binds at the A33 dimer interface , with the L and H chains contacting both A33 subunits . This correlates well with the SEC data , which suggested that one Fab binds one A33 dimer ( Fig 3 ) . BSA for both L and H chains bound to A33 are larger compared to A2C7 and A20G2 ( L = 525 . 3 vs . H = 630 . 2 Å2 ) for a total of 1155 . 5 Å2 ( Fab:Ag ) ( S4 Table ) . Sc values of 0 . 71 and 0 . 74 are among the highest Sc values reported for any Fab:Ag interface and indicate a highly specific binding interaction . Eighteen residues of the H chain contact sixteen A33 residues , while fifteen residues of the L chain contact twelve A33 residues ( S3 Table and Fig 4 ) . CDR H1 interacts only with A33 subunit A ( Fig 4 ) . H1:Tyr32 forms hydrogen bonds with three A33 residues ( Glu149 , Thr160 , and Lys161 ) , the most favored of which is with Glu149 ( 2 . 53 Å ) . In addition , a salt bridge is formed between Lys161 and H1:Asp31 . Surprisingly , the Lys161Ala mutation did not impair A27D7 binding ( S2 Table and Fig 6 and S3 Fig ) . CDR H2 interacts only with A33 subunit B . A33 residue Asp115 is a central residue of the H2 interface , as it forms six potential hydrogen bonds with H2 residues Gly53 , Gly54 , Gly55 , and Thr56 . Interestingly , the Asp115Ala mutation only slightly affects A33 binding ( S2 Table and Fig 6 and S3 Fig ) . Three other hydrogen bonds link H2 residues Thr56 and Tyr58 to Gln117 , Met118 , and Asp121 . Van der Waals contacts are the most extensive of the interface , with forty-two contacts compared to only eighteen in the H1/A33 interface . Water molecules do not contribute largely to this interface ( S2 Fig ) . CDR H3 is centrally positioned in the A33 dimer groove , and contacts residues of both A33 subunits . H3 forms three distinct hydrogen bonds with A33 subunit B: Lys177 interacts with H3:Ala97 and H3:Ser98 . The third hydrogen bond is formed between Tyr116 and H3:Tyr99 . Water molecules are recruited to the interface and allow for water-mediated hydrogen bonds with A33 subunit A residues Glu149 and Asp150 . The H3 loop has a number of VdW contacts that is comparable to H1 ( 20 vs . 18 ) . Overall , H3 appears to make a modest contribution to the interface compared to H1 and H2 . The A27D7 light chain does not elicit any salt bridge with A33 , and its interaction with the antigen is mostly driven by van der Waals contacts . In particular L2 and L3 each interact specifically with either one of the A33 subunits . L1 is relatively short compared to group II MAbs A20G2 and A2C7 ( 12 vs . 16 residues ) . A single residue L1:Tyr32 , forms one hydrogen bond with Asp150 and also accounts for all VdW contacts with the A33 subunit B . L2 interacts solely with A33 subunit A , with a total of two hydrogen bonds and 30 VdW contacts . Moreover , four residues that elicited VdW interactions are involved in a water-mediated hydrogen bond network between L2 residues Asn53 , Leu54 , Ser56 , and Ala60 of A27D7 and A33 residues Gln173A , Asp168A , Ser172A , and Val175A . CDR L3 ( Ser92 and Leu94 ) forms an extensive hydrogen bond network with A33 residues Gln173B , Glu174B , Val175B and Arg176B . Binding affinity of A27D7-MAb to A33 mutant Gln173Arg , however , remained in the picomolar range ( KD = 20 pM ) , suggesting that the many contacts formed between A27D7 and A33 can compensate for the various single amino acid substitutions on A33 ( S2 Table and Fig 6 and S3 Fig ) . As the Fab of A27D7 does not form identical contacts with both subunits of A33 , some A33 residues will only be in contact with the Fab once . However , Tyr116 , Asp150 , and Val175 form contacts with the Fab in both A33 subunits , and among those , Tyr116 was chosen for single alanine scanning mutagenesis , as it might contribute more to the overall binding energy than the previously assessed A33 residues ( Fig 6 ) . The Tyr116Ala mutation led to a significant decrease in binding affinity ( S2 Table and Fig 6 and S3 Fig; KD = 0 . 54 nM ) . Next , we asked whether the moderately reduced binding affinity of the mutants Asp115Ala , Lys161Ala , Asp170Ala and Gln173Arg could be the result of a particularly mutation-tolerant A27D7/A33 binding interface , or because these residues are not making important contacts with the antibody . To answer this question , we created the double mutant Asp115Ala/Tyr116Ala and observed a synergistic effect of the mutations , as affinity decreased another ~20-fold compared to Tyr116Ala only ( S2 Table and Fig 6 and S3 Fig; KD = 12 . 47 nM ) . However , even combining both mutations still resulted in low nM binding affinity suggesting that a large number of energetically favorable interactions still remain intact and enable the specific binding interaction . The robustness of A27D7 binding to single alanine A33 mutants made this antibody an attractive candidate for cross-neutralization studies . We therefore explored the potential of our anti-A33 ( VACV ) MAbs to bind to recombinant A33 that mimics the epitope of the orthopox strains such as cowpox Brighton ( CPXV-Br ) , MPXV , and ECTV . Sequence alignment of A33 from different orthopoxviruses revealed that no more than three residues out of the thirty-eight possible variations differ in the A33 epitopes among the analyzed orthopox strains ( S4 Fig ) . The A2C7 epitope contains only two of those variations , located at positions 118 and 173 . A27D7 and A20G2 epitopes contain the same , plus an additional residue at position 117 . All residues that may vary within the epitopes are highlighted in red ( Fig 4 ) , regardless of the viral strain . Using site-directed mutagenesis we made two recombinant forms of A33 . The A33 variant Gln117Lys/Leu118Ser both mimics the epitope of CPXV and MPXV , while the A33 variant Gln117Lys/Leu118Ser/Gln173Arg is specific for ECTV . Both antibodies A2C7 and A20G2 failed to bind to the ECTV variant , while A20G2 still bound with a ~9-fold lower affinity to the CPXV/MPXV variant . As expected however , A27D7 was able to bind both A33 variants with high affinity ( KD of 47–117 pM ) ( S2 Table and Fig 6 and S3 Fig ) . Therefore , we next asked whether A27D7 could protect against ectromelia infection , since ectromelia A33 contains the most amino acid changes in the A27D7 epitope compared to the other orthpox species . In order to demonstrate the in vivo protective capabilities of antibodies A2C7 , A27D7 , and A20G2 we evaluated their efficacy following a lethal challenge with ECTV . Mice were treated with a single 100 μg IP injection of the antibodies at either T = -1 or T = +1 relative to challenge with 1000 PFU of ECTV . An additional group of mice was treated with the potent antiviral cidofivir ( CDV ) , which is known to be protective against this dose of ECTV . Following challenge , we observed that mice treated with antibodies A2C7 and A20G2 were not able to statistically protect mice against challenge when compared to mice treated with vehicle ( Fig 7A ) . Conversely , mice treated with the A27D7 antibody were fully protected against challenge when the antibody was administered at T = +1 ( P = 0 . 0004 ) and were 90% protected ( P = 0 . 006 ) when the antibody was administered at T = -1 ( Fig 7A ) . Although weight-loss was significant following the T = -1 administration , we found negligible levels of weight-loss when the antibody was administered at T = +1 ( Fig 7A ) . Indeed , at T = +1 we found that A27D7 protected against mortality and morbidity with the equivalency of the CDV-treated controls ( Fig 7B ) . The slightly reduced efficacy of the antibodies when administered at T = -1 is likely a reflection of the antibody half-life , which we have not addressed in this assay . At T = +21 days post-challenge , we re-challenged the mice with a 10 , 000 PFU inoculum of ECTV—all previously challenged mice were protected against subsequent mortality and morbidity , indicating that the antibodies did not impede the generation of memory and protective immunity .
In this study , we have generated a panel of anti-A33 MAbs and characterized their binding to recombinant A33 , as well as assessed their complement-dependent neutralization activity of VACV EEV . The three neutralizing MAbs , A2C7 , A20G2 , and A27D7 were further tested in a VACVWR challenge model , where they all demonstrated full protection . For each of these antibodies we determined the A33 epitope using X-ray crystallography and assessed the robustness of antibody binding to A33 variants with amino acid exchanges in the various epitopes . We further identified that A27D7 was unique in its binding to A33 and exhibited resistance to single alanine substitutions within the epitope . A27D7 further bound to engineered A33 mimicking other orthopox strains within its epitope and was protective against in vivo challenge with ECTV . In contrast , both A2C7 and A20G2 , which failed to bind the ECTV A33 mimic failed to protect against ECTV challenge as expected . The strong protection of A27D7 against ECTV challenge at day 1 post infection is surprising but we believe a reflection of its unique binding to A33 , in which the antibody-antigen interface can tolerate several amino acid exchanges and still bind with high affinity . There is slightly reduced protection when the A27D7 is administered 1 day prior to ECTV infection . As A27D7 fully protects against VACV when given at day -1 , it could be that the antibody half-life in vivo is more important during the ECTV challenge , especially since the A27D7 binds with reduced affinity to the ECTV A33 mimic compared to VACV A33 and a higher dose of antibody could therefore be needed for full protection . A former study showed that VACV-A33 vaccination protected mice against ECTV but interestingly not against CPXV-Br [31] . The reason for the lack of protection remains unclear , because VACV is more closely related to CPXV-Br than to ECTV ( S4 Fig ) . To address this conundrum , we have developed an in vitro assay to assess the potential cross-neutralization ability of a given antibody in three steps . First , we used single alanine mutagenesis to engineer variants of A33 in which epitope residues that were predicted to be important for antibody binding based on our structural data were targeted . Next , the real-time binding kinetics of the A33 variants to the different antibodies was tested . Antibodies that showed robust binding to different A33 variants ( e . g . A27D7 ) were further assessed in their ability to bind to A33 variants that have been engineered in their epitope to mimic the sequence of other orthopox strains , including ECTV and CPXV/MPXV . If high affinity binding to these A33 cross-species variants is still observed , we predict that the antibodies would be protective . Using such in vitro approach , rather than expressing all of the orthologous A33 antigens individually , is highly time and cost efficient and sufficient to predict the outcome of specific MAb-guided protection experiments , and does not require any stringent safety procedure associated to virulent strains . In our experiments the antibody A2C7 , which failed to bind to the A33 Gln173Arg ECTV mimic , also failed to protect against lethal challenge with ECTV , as expected . Position 173 is occupied by arginine in ECTV , while glutamine is conserved in all other strains ( S4 Fig ) . Position 173 is also common to all three epitopes and we expected Gln173Arg mutation to decrease binding of all three MAbs . It indeed resulted in abrogating A33 binding to A2C7 and A20G2 . However , its effect was opposite in A27D7 , as it reversed the loss of binding conferred by Gln117Lys/Leu118Ser mutations . Beside favoring A27D7 in terms of protectivity for the aforementioned reasons , our study also illustrates how divergent the properties of antibodies targeting a given Ag can be . In this study , we have ranked our representative MAbs using various properties , such as binding affinity the location and composition of the epitope , the binding stoichiometry ( 1 or 2 Fab per A33 dimer ) , and the tolerance to alanine scanning mutagenesis within the epitope . The most potent antibody ( highest affinity ) is not necessarily the top candidate in each category and each MAb may be raised and optimized for a particular property . For example , A27D7 is characterized by a >10-fold higher binding affinity for VACV-A33 compared to A2C7 and A20G2 . However , A2C7 and A20G2 present specific binding properties , such as increased MAb clustering resulting from A33 bivalency . This suggests that when considering a particular antigen , the efficiency of an antibody response most likely relies on the combination and the synergetic effect of different antibodies targeting a specific Ag . This added layer of complexity has been so far overlooked , and we hope that its study will be implemented in future protection models . This could be studied in an in vivo protectivity assay by comparing the responses obtained while using individual MAbs , to those obtained using cocktails combining MAbs with different features , such as discussed above . The data we obtained using single alanine scanning mutagenesis within the A27D7 antibody epitope may explain why a previous study observed that vaccination with recombinant Sindbis virus expressing A33 VACV conferred protection against ECTV but not CPXV [31] . In this case , vaccination might not have elicited anti-A33 antibodies that could bind the Gln173Arg variation found in cowpox . Residue 173 is adjacent in the A33 crystal structure to residues 117 and 118 and as such topographically part of the CPII region that the authors identified as containing the protective epitope . We observed a similar effect for A27D7 , which bound tighter to our A33 ECTV mimic than to the CPXV counterpart , after mutating this single residue from a Gln to an Arg . However , while we have not tested if A27D7 can protect against cowpox challenge , binding of the MAb to the cowpox A33 epitope variant is still in the nanomolar range , which we predict to be sufficient to confer protection .
The A33 expression vector ( VACV strain Acam2000 GenBank: AY313847 ) containing the two mutations Leu118Met and Leu140Met was graciously provided by David Garboczi ( NIH/NIAID ) . Recombinant A33 protein ( residues 89–185 ) was produced and purified as previously described [40] . Briefly , protein was expressed in BL21 CodonPlus E . coli cells ( Agilent ) and refolded by rapid dilution . Protein was then purified by Ni-NTA affinity chromatography using a 5 mL HisTrap FF column ( GE Healthcare ) with bound protein eluted in 0 . 2 M Imidazole , 20 mM Tris pH 8 . 0 , 0 . 3 M NaCl . The eluted protein was then further purified by gel filtration chromatography on a Superdex 200 size exclusion column ( GE Healthcare ) . Pure A33 eluted at a volume corresponding to a dimer . A biotinylated version of A33 ( BtA33 ) was engineered by fusing an Avitag to the N-terminus of A33 and recombinant A33 was enzymatically biotinylated according to manufacturers suggestion ( Avidity ) . BtA33 mutants were engineered by site-directed mutagenesis using the Quickchange II mutgenesis kit and verified by sequencing . Engineered mutations include the single mutants Lys123Ala , Leu118Arg , Tyr116Ala , Asp170Ala , Gln173Arg , Asp168Ala , Asp115Ala , Lys161Ala , the double mutants Asp115Ala/Tyr116Ala , Gln117Lys/Leu118Ser and the triple mutant Gln117Lys/Leu118Ser/Gln173Arg . For the generation of anti-A33 Mabs , BALB/C mice were infected intranasally with a sub-lethal dose ( ~0 . 5 x 103 PFU ) of VACV WR and subsequently boosted twice with intraperitoneal injection of ~100 μg of recombinant A33 protein . Hybridomas were generated from the immunized mice , and were screened for their specificity for VACV with an immunofluorescence assay , in which WR-infected HeLa cells were stained with culture supernatants of the hybridomas as described previously [59 , 60] . Anti-B5 Mabs , ( B126 or B96 ) were generated as described previously [46] . Rabbit-anti L1 sera were generated by immunizing rabbit with recombinant L1 protein as described in [61] . Rabbit anti-A33 polyclonal Abs ( NR-628 ) , pAbs , was produced by immunization of rabbits with recombinant A33R protein ( rA33 ) and was obtained through the NIH Biodefense and Emerging Infections Research Resources Repository , NIAID , NIH ( BEI Resources , Manassas , VA ) from Drs . Eisenberg RJ and Cohen GH . [49] . Total RNA from 300μL hybridoma cells in solution was isolated using the NucleoSpin RNA II kit according to manufacturer’s instructions ( MACHEREY-NAGEL ) . cDNA was amplified using the OneStep RT-PCR kit ( Qiagen ) . The reverse transcription PCR was performed using primers 5’MsVHE and 3’Cy2c outer ( for isotype IgG2a antibodies A26C7 , A25D11 , A27D7 , and A20G2 ) or 3’Cy2b outer ( for isotype IgG2b antibody A2C7 ) for the heavy chains , and primers 5’mVkappa and 3’mCĸ for the kappa light chains [62] . The cycling profile was slightly modified from manufacturer’s recommendations and set up as follows: 1 cycle of 30 min at 50°C and 15 min at 95°C; 40 cycles of 30 s at 94°C , 45 s at 60°C ( for heavy chains ) / 58°C ( for light chains ) , and 55 s at 72°C; followed by 1 cycle of 10 min at 72°C and a 12°C cool down . PCR products were verified by gel electrophoresis with a ~500bp product for heavy chains and ~450bp product for light chains . Afterwards , PCR products were purified using the QIAquick PCR Purification Kit ( Qiagen ) and subsequently sequenced by Retrogen ( provided with the respective 5’ primer for heavy and light chains ) . Sequences include V-D-J regions for heavy chains and V-J regions for light chains . Finally , antibody germ lines were determined using IMGT’s V-Quest server [63] . Alternatively , Illumina MiSeq library prep was performed by standard Illumina methods . Briefly , the used Nextera XT DNA sample preparation kit uses an engineered transposome to fragment and tag ( “tagment” ) input DNA simultaneously , adding unique adapter sequences in the process . A limited-cycle PCR reaction using those adapter sequences was performed to amplify the inserted DNA . The PCR also adds index sequences on both ends of the DNA , thus enabling dual-indexed sequencing of pooled libraries on the Illumina MiSeq instrument to generate approximately 100 . 000–300 . 000 paired-end reads per sample . Data was then analyzed via IMGT/HighV-QUEST . Biotinylated 20-mer peptides overlapping by 10 residues and covering nearly the entire amino acid sequence of A33 ( residues 1–180 vs . 1–185 ) were synthesized by A&A Labs ( San Diego , CA ) . Microtiter plates were coated with 100 μl of NeutrAvidin biotin binding protein ( 1 μg/ml ) diluted in PBS overnight at 4°C ( Thermo Scientific Pierce , Rockford , IL ) and linear peptide ELISA for the purified anti-A33 Mabs was done as described previously [55] . 100 μl of recombinant A33 protein ( A33 ) at 1 μg/ml in PBS was used to coat the plate as control . The secondary Ab was streptavidin-horseradish peroxidase-conjugated goat anti-mouse IgG ( Invitrogen , CA ) . MV of VACVWR ( Western Reserve ) strain stock was grown in HeLa cells in D-10 ( DMEM + 10% FCS + pen/strep/glutamine ) as described previously [61] . Fetal calf serum ( FCS ) used in all experiments was heat inactivated ( 56°C , 30 min ) prior to use to eliminate complement activity . Purified VACVWR stock was made via centrifugation through a sucrose cushion as described previously [61] . Virus was stored at -80°C . EEV stocks of VACV WR strain were prepared using HeLa cells grown in D-10 , and the medium containing EEV was harvested at 2 days after infection as previously described [46] [50] . Clarified supernatant was used immediately or stored at 4–8°C for a maximum of 3–4 weeks . EEV VACVWR stocks ( ~5 x 105 PFU/ml ) were untwined on VeroE6 cells in the presence of rabbit anti-L1 Abs to block contaminating MV or damaged EEV present in the EEV stock , as previously described [46] [50] . Briefly , VeroE6 cells were seeded at 1 . 5 x 105 cells/well into 24-well Costar plates ( Corning Inc , Corning , NY ) and EEV VACVWR neutralization was performed the following day using Mabs or pAbs samples at 10μg/ml ( final concentration ) or 1/100 ( final dilution ) respectively in the absence or the presence of baby rabbit complement 10% ( final concentration ) ( complement , Cedarlane Laboratories , Ontario , Canada ) as described previously [46] [50] . ( Rabbit anti-L1 ( 1:25 to 1:100 , final concentration ) was used to block the MV present in the EEV stock in all experiments . VACV EEV supplemented with anti-L1 antibody was regularly used in each assay with or without complement as a negative control . Female BALB/c mice were used at an age of 7–8 weeks . To infect the mice , a Pipetman was used to place 10 μl of VACVWR on each nare of an isofluorane-anesthetized mouse ( total volume , 20 μl ) , and the liquid was rapidly inhaled by the mouse . Mice were weighed daily to assess disease progression . Mice were euthanized if and when 25% weight loss occurred . A dose of 1x105 PFU of VACVWR was the standard dose given to 7-week-old BALB/c females . For A33 mAb protection studies , mice were treated by i . p . injection with 100 μg of antibodies 1 day before infection . Control mice received anti-A10 BG3 . 1 antibody , which is known to provide no protection . An additional group received anti-B5 B126 as positive control . A third group received anti-L1 M12B9 , which was expected to protect against death but not weight loss [53] . A2C7 ( mouse IgG2b ) , A27D7 , and A20G2 ( mouse IgG2a ) Mabs were digested with 2% ( wt/wt ) activated papain for 4hr at 37°C in digestion buffer ( A27D7: 50 mM NaOAc pH 5 . 5; A2C7 and A20G2: PBS pH 7 . 4 ) . Papain was activated by incubating 24 . 4 μL papain solution at 20 . 5 mg/mL ( Sigma ) in 1 mL solution containing 100 μL 10X Papain Buffer ( 1M NaOAc pH 5 . 5 , 12mM EDTA , and 10mM cysteine ) for 15 min at 37°C . The papain digestion was inhibited with 1 mL of 200 mM iodoacetamide ( IAA ) . A27D7 sample was then dialyzed in 2 L PBS pH 8 . 0 overnight at 4°C . For A2C7 and A20G2 , papain reaction mixtures were diluted in one volume of PBS pH 8 . 0 . All three samples were passed through a pre-equilibrated 1 mL protein A FF column in PBS pH 8 . 0 binding buffer . The purified Fab , contained in the protein A flowthrough , was concentrated to 1 mL using centrifugal filtration devices ( Pierce Concentrator; 9-kDa molecular mass cut-off ( MWCO ) –Thermo Scientific ) and purified to homogeneity by Size Exclusion Chromatography ( SEC ) on a Superdex 200 HiLoad 16/60 ( GE ) column in 20mM Tris pH 8 . 0 , 200mM NaCl running buffer . All Fabs eluted as a single and sharp peak ( Ve ~ 86 mL ) . A33 and Fabs were mixed together as homogeneous species at a molar ratio of 1:2 A33 dimer to Fab ( A2C7 and A20G2 ) or 1:1 ( A27D7 ) at a low concentration ( < 1mg/mL ) . Each A33:Fab complex was concentrated to a volume of 1 mL using centrifugal filtration devices and SEC-purified ( Superdex 200 HiLoad 16/60 ) in 20 mM Tris pH 8 . 0 , 200 mM NaCl to separate the A33-Fab complex from unbound A33 and/or Fab . Fractions corresponding to the complex were pooled and concentrated for crystallization . To assess binding kinetics of MAbs A2C7 , A20G2 , and A27D7 to WT-A33 , we performed BLI assays using the OctetRED instrument ( ForteBio , Inc ) . In a first experiment we immobilized MAbs on Anti-mouse Fc Capture ( AMC ) biosensors and tested binding to WT-A33 in solution . Biosensors were loaded with 10 μg/mL MAb diluted in 1X kinetics buffer ( PBS pH 7 . 4 , 0 . 002% Tween 20 , 0 . 01% BSA ) over 5 mins . MAb-loaded tips were tested against two-fold serial dilutions of WT-A33 analytes ( 0 . 78–50 nM ) . Binding kinetics were calculated using Octet Data Analysis 7 . 1 software ( Forte Bio , Inc . ) with curve-fitting statistics . The association steps were performed over 10 mins and dissociation steps performed over 20 mins . A negative control antibody ( EE11 ) that targets the VACV antigen D8 was run in parallel to all assays for subtraction of background binding signal . Curves were aligned to baseline . To test the effect of the A33 mutations on antibody binding , we immobilized biotinylated A33 ( BtA33 ) using streptavidin ( SA ) biosensors . SA biosensors were loaded with 1 . 5 μg/mL BtA33 protein in 1X kinetics buffer ( PBS pH 7 . 4 , 0 . 002% Tween 20 , 0 . 01% BSA ) over 5 mins . A33-loaded tips were tested against three-fold serial dilutions of IgG analytes ( 27 . 4 pM-20 nM ) . The association steps were performed over 10 mins and dissociation steps performed over 20 mins . Buffer control against biosensors loaded with WT-A33 was subtracted from raw data and curves were aligned to baseline . KD , kon , and koff were determined by global fitting of association and dissociation steps for all dilutions assuming a 1:1 binding model . Four- to six-week-old female C57BL/6 mice were obtained from Harlan Laboratories ( Indianapolis , IN ) , housed in filter-top microisolator cages , and fed commercial mouse chow and water ad libitum . The mice were housed in an animal biosafety level 3 containment area . The day before challenge , mice were treated i . p . with 100 μg of the specified antibody . Immediately before challenge , mice were anesthetized with 0 . 1 ml/10 g body weight of ketamine HCl ( 6 mg/ml ) and xylazine ( 0 . 5 mg/ml ) by i . p . injections . One thousand PFU of ECTV ( Moscow strain ) in PBS without Ca2+ and Mg2+ was slowly loaded into nares ( 5 μl/nare ) . Mice subsequently were left in situ for 2 to 3 min before being returned to their cages . Mice were monitored daily for mortality and morbidity , as measured by weight change . Animal husbandry and experimental procedures were approved by the Institutional Animal Care and Use Committee of Saint Louis University ( ectromelia challenge , protocol number 2082 ) and by the Department of Laboratory Animal Care and the Animal Care Committee of the La Jolla Institute ( VACV challenge , protocol number AP088-BP2-1112 ) . We follow PHS Policy on humane care and use of laboratory animals . The coordinates and structure factors of the A33:Fab complexes have been deposited in the Protein Data Bank ( www . rcsb . org ) with codes 4LQF , 4LU5 , and 4M1G . The nucleotide sequences of the anti-A33 antibodies have been deposited in GenBank , and accession numbers are listed in S5 Table . | Before the eradication of smallpox ( variola virus ) from nature , hundreds of million of people succumbed to the infection . The discovery of vaccinia virus ( VACV ) , the active ingredient of the smallpox vaccine , ultimately led to the eradiation of smallpox from the human population . Vaccination with VACV leads to a strong antibody response that protects against variola virus . As the protective antibodies recognize viral proteins that are highly similar in sequence between the different orthopox strains , such as A33 used in this study , several antibodies have the capacity to neutralize a larger breath of orthopx viruses . In this study we have identified an anti-A33 antibody from a larger panel that exhibits a unique binding mode to A33 . This antibody , A27D7 , is also resistant to single amino acid changes throughout the protein and binds to engineered A33 variants that mimic ectromelia and orthopox A33 in the antibody-binding site . As the antibody further protects against ectromelia infection of mice , this antibody appears to be a potent orthopox cross-species protective antibody with therapeutic potential . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Structural and Functional Characterization of Anti-A33 Antibodies Reveal a Potent Cross-Species Orthopoxviruses Neutralizer |
Turkey is located in an important geographical location , in terms of the epidemiology of vector-borne diseases , linking Asia and Europe . Cutaneous leishmaniasis ( CL ) is one of the endemic diseases in a Turkey and according to the Ministry Health of Turkey , 45% of CL patients originate from Şanlıurfa province located in southeastern Turkey . Herein , the epidemiological status of CL , caused by L . tropica , in Turkey was examined using multilocus microsatellite typing ( MLMT ) of strains obtained from Turkish and Syrian patients . A total of 38 cryopreserved strains and 20 Giemsa-stained smears were included in the present study . MLMT was performed using 12 highly specific microsatellite markers . Delta K ( ΔK ) calculation and Bayesian statistics were used to determine the population structure . Three main populations ( POP A , B and C ) were identified and further examination revealed the presence of three subpopulations for POP B and C . Combined analysis was performed using the data of previously typed L . tropica strains and Mediterranean and Şanlıurfa populations were identified . This finding suggests that the epidemiological status of L . tropica is more complicated than expected when compared to previous studies . A new population , comprised of Syrian L . tropica samples , was reported for the first time in Turkey , and the data presented here will provide new epidemiological information for further studies .
Leishmaniasis is a parasitic disease caused by intracellular protozoan parasite , Leishmania and transmitted by the bite of a certain female Phlebotominae sand flies . Leishmaniasis is classified as cutaneous , visceral and mucocutaneous by clinical manifestations and it is among the world’s six major tropical diseases . More than 70% of cutaneous leishmaniasis ( CL ) cases were reported from Afghanistan , Algeria , Colombia , Brazil , Iran , Syria , Ethiopia , North Sudan , Costa Rica and Peru [1] . In most countries cutaneous leishmaniasis is under-report , therefore it is difficult to estimate the real number of the cases . In Turkey , the first CL case was reported in 1883 , and 46 . 003 cases recorded between 1990–2010 [2] . Cutaneous leishmaniasis is a major health problem in Turkey . According to the Turkish Ministry of Health 45% of the CL patients originate from Şanlıurfa province located in the southeastern region of Turkey . To date , four different Leishmania species ( L . tropica , L . donovani , L . infantum and L . major ) were reported in Turkey to cause CL [3 , 4] . Rare cases were reported that L . tropica may be the causative agent of visceral leishmaniasis ( VL ) in Mediterranean Basin as well as in Turkey [5] . Syrian Ministry of Health reported that incidence of CL has increased during the last 15 years and peaked in 2011 with 58 . 156 cases [6] . Aleppo province , Syria is a hyper-endemic area for CL with 12 . 000 cases reported annually [7] . Species identification of the causative agent for CL in Syria revealed that the majority of cases ( 85% ) were caused by L . tropica , while only 15% were identified as L . major [8] . The ongoing Syrian civil war displaced more than 6 . 5 million people from Syria [9] . Millions of Syrian refugees have fled to neighboring countries with approximately three million refugees residing in Turkish camps alone . In Gaziantep , a city located in southeastern Turkey , reports have shown a dramatic increase in the observed number CL patients admitted to state hospitals [10] . In 2013 , the number of Syrian patients admitted to hospitals due to CL peaked with 76 positive cases , as opposed to only one positive case reported in 2012 [11] . Turkish L . tropica strains obtained from CL patients in Şanlıurfa during the 1995 outbreak were analyzed using MLMT in a previous study [12] . All 27 isolates from this outbreak showed an identical genotype , and formed a microcluster together with two strains from Adana . The MLMT profile of L . tropica strains isolated from Syrian refugees in Turkey have not been determined , and it is clear from published reports that many of these refugees will only pass through Turkey on their way to Europe where potential vector sand fly species have been reported in several countries [13 , 14] . Previous studies using Multilocus Enzyme Electrophoresis ( MLEE ) analysis of Turkish L . tropica isolates , conducted by Leishmaniasis Reference Center Montpellier , France , indicated that the majority of strains belong to zymodeme MON-304 . These strains were highly heterogeneous with 8 different zymodeme profiles ( MON-55 , MON-200 , MON-303 , MON-304 , MON-312 , MON-313 , MON-314 , and MON-315 ) noted [5] . The discriminatory power of the MLEE is limited , and not associated with geographic distribution . In addition , because large numbers of parasites are required , clinical samples cannot be used for MLEE analysis , which is important to understand the epidemiology of the disease . Another disadvantage of this method is that the enzyme panel may vary across the reference laboratories , and the obtained raw data cannot be compared directly among them . Microsatellite or short tandem repeats ( STRs ) are the neutral consecutive repeats of nucleotides ranging from 1 to 6 bases in the non-coding regions of the genome . MLMT is a powerful tool for discriminating among populations and sub-populations using a battery of markers . The population genetics of different Leishmania species were studied previously with various numbers of specific microsatellite markers . To date , 16 different polymorphic markers were designed to reveal genetic structure and dynamics of L . tropica , but subsequently only 12 specific microsatellite markers were used to reveal genetic structure of L . tropica [15 , 16 , 17] . L . tropica is the most widely distributed agent of CL in the Old World , and causes many different disease pathologies including chronic and recidivans skin disease . It has been identified as the main agent responsible for CL in Turkey [2 , 5] . In the present study , we examined the MLMT profiles of L . tropica from cultured strains and Giemsa-stained smears obtained from Turkish and Syrian patients in different geographic regions , thus updating the epidemiological status of CL caused by this parasite in Turkey .
The sample collection from patients was done according to the Ethical Committee approval ( No: 2016/266 ) of the “Dr . Ersin Arslan State Hospital” in Gaziantep . All adult subjects provided written informed consent , and a parent or guardian of any child participant provided informed consent on their behalf . A total of 38 previously isolated and cryopreserved strains , and 20 Giemsa-stained smears , obtained from Turkish and Syrian patients , were included in this study . The 38 isolates were selected from our Leishmania Cryobank according to their geographical origins . Fifteen strains isolated during the outbreak in Şanlıurfa in 1995 , 11 strains from Aydın province , two isolates ( one viscerotropic ) from Manisa province , five isolates from Syrian patients diagnosed in Turkey , one isolate from five different provinces ( Malatya , Muş , Muğla , Niğde , Izmir ) were used in the study ( S1 Table ) . Zymodeme ( MON ) analysis was conducted for all autochthonous strains used in this study in Montpellier , France . Additionally , 20 Giemsa-stained smears obtained from the lesions of CL patients were also included in the study . Twelve and eight of these samples were gathered from Syrian and Turkish patients , respectively , at the Dr . Ersin Arslan State Hospital in Gaziantep . Cryopreserved isolates were removed from liquid nitrogen and first cultured in NNN medium for two weeks . After the mass cultivation in RPMI+10%FCS medium , promastigotes were harvested and genomic DNA was isolated using Qiagen DNeasy® Blood & Tissue Kit ( Qiagen , Hilden , Germany ) . To isolate genomic DNA from patient samples , the material was removed from the Giemsa-stained smears by washing with molecular grade water and centrifuged at 8 . 000 rpm for 10 minutes . After centrifugation , the supernatant was discarded and pellet used for the DNA isolation . The isolation was performed using Qiagen DNeasy® Blood & Tissue Kit ( Qiagen , Hilden , Germany ) and the DNA was eluted in a final volume of 50 μl . Species typing of the parasite were done using the internal transcribed spacer 1 ( ITS1 ) real-time PCR as reported previously [5] . Only samples identified as L . tropica were included in this study . Previously reported twelve highly specific microsatellite markers were used to amplify the regions containing the repeat motifs [18] . After successful amplification of the DNA , PCR products were loaded to ABI PRISM 3130XL ( Applied Biosystems , USA ) sequencer and fragment sizes determined using Geneious R8 [19] . The reference strain of L . tropica , MHOM/PS/2001/ISL590 , was used in each run as a standard and the fragment sizes were noted for each marker . The repeat numbers were determined using the repeat motifs and numbers of the reference strain . After the normalization step of the raw fragment data , several modelling softwares were used to determine the population structure . The data was prepared as a text file and converted into the necessary input file formats using Microsatellite Analyzer 4 . 05 ( MSA ) software . In order to determine number of the populations , Bayesian clustering method was applied using STRUCTURE V . 2 . 3 . 4 [20] . The Markov Chain Monte Carlo iteration was set to 200 . 000 and length of burnin period to 20 . 000 . In order to determine most appropriate number of populations , 10 iterations were applied for each K value ( K1-10 ) . Delta K ( ΔK ) calculation was made using an online tool , Structure Harvester [21] . The distribution of genetic variation was evaluated using Factorial Correspondence Analysis ( FCA ) implemented in Genetix 4 . 05 [22] . Genetic distances based on Chord distance were calculated using POPULATIONS 1 . 2 software [23] . Phylogenetic network was created using SplitsTree software [24] . Genetic distances between populations were calculated using MSA 4 . 05 [25] . GDA 1 . 1 software was used to calculate number of alleles ( A ) , observed ( Ho ) , expected ( He ) and heterozygosity and the inbreeding coefficients ( FIS ) [26] .
Bayesian statistics were applied using STRUCTURE analysis and according to ΔK calculations , three main populations ( POP A , POP B and POP C ) were identified . Twenty-three isolates were clustered in POP A , while 19 were in POP B and 16 were in POP C . Three subpopulations were identified both for POP B and C , while no subpopulation was identified for POP A ( Fig 1 ) . The majority of the populations were correlated with their geographical origins but no clear difference was noted between samples from Şanlıurfa and Syria , where population exchanges occur frequently ( Fig 2 ) . In total , 23 L . tropica isolates were clustered in main POP A , which were isolated from three different geographical origins ( 15 Şanlıurfa , seven Syrian and one from Malatya ) . When POP B is analyzed separately using STRUCTURE , it split into three subpopulations , B1 , B2 and B3 . POP B comprised of 10 samples from Syrian , eight samples from Gaziantep and one samples from Niğde . All Syrian samples in POP B were clustered as sub-population B1 , while all two Syrian and five Gaziantep samples were clustered in sub-population B2 . The one sample from Niğde was clustered in sub-population B3 with three Gaziantep and two Syrian samples . Third main POP C comprised all samples originated from Mediterranean except one sample , which clustered in POP C . The sample was isolated from Muş city in eastern Turkey , which geographical origin does not correlate with other samples clustered in POP C . POP C was analyzed separately and three subpopulations were identified . However , the ΔK values suggest weak substructure and the sub-populations did not correlate with their geographical origins . Descriptive analysis per locus was performed and coefficient of relationship was noted positive for four markers ( 4GTG , LIST7033 , GA1 and GA2 ) ( Tables 1 and 2 ) . The FIS per locus was found positive in 11 out of 12 markers that suggest the large number of homozygous alleles in studied samples . Descriptive analysis per populations ( K: 3 ) was also performed and POP B was found to be the most heterogeneous group by having the highest number of alleles , He and Ho values . One of the L . tropica strains was viscerotropic while the rest were dermotropic . The MLMT repeat motif and Bayesian clustering method revealed that there are no significant differences between the viscerotropic isolate and the dermotropic isolates . The majority of the amplified markers ( 11/12 ) of the viscerotropic isolate were found to be highly similar to those in dermotropic isolates having the same geographical origins . The viscerotropic isolate was clustered in same subpopulation ( Sub-Pop C ) with those dermatropic isolates . The isoenzyme analysis ( MLEE ) was previously performed to 33 out of 38 autochthonous strains and six different profiles ( MON-55 , MON-200 , MON-303 , MON-304 , MON-312 and MON-315 ) were identified . POP A contained 16 strains belonging to MON-304 and seven Giemsa-stained smears for which MON typing was not available . POP B contains 18 smears with no MON profile and one sample belonging to MON-55 . Five different MON profiles ( MON-200 , MON-303 , MON-304 , MON-312 and MON-315 ) were clustered in population C . MLMT data from the previously studied 35 L . tropica samples [16] were combined with the 58 samples in the present study and all 93 samples reanalyzed using STRUCTURE software and ΔK ( K:2 ) was calculated . Two main populations were observed and they were named as Şanlıurfa and Mediterranean according to the origins of the majority of L . tropica strains . The populations , Şanlıurfa and Mediterranean were analyzed separately and found to form two and three subpopulations , respectively ( Fig 3 ) . Descriptive analysis per locus was performed and coefficient of relationship for three markers ( GA1 , LIST7033 and 4GTG ) was found higher among studied markers . According to descriptive analysis per populations , 57 isolates were clustered in Şanlıurfa population and 36 were in Mediterranean population ( Table 3 ) .
Şanlıurfa province is a well known highly endemic area for CL caused by L . tropica and Gaziantep is its neighbor city with 150 km in the west . The disease is present for centuries in this geographical area covering the cities located in northern Syria and southern Turkey . The altitude is between 510 and 800 meters in this plain area and there are no natural barriers like mountains [27] . Although L . major was detected rarely in recent years in Şanlıurfa , the main causative agent is L . tropica for most of the CL cases similar to Aleppo [3] . The inhabitants of Aleppo and Sanliurfa are relatives who visit each other frequently staying more than one week . In addition , the mass human migration between western , southeastern and Mediterranean regions in Turkey due to temporary summer works explains the wide spread of L . tropica in these regions . With Phlebotomus sergenti as the dominant and proven vector in the area with no animal reservoir detected so far [27] , the life cycle of L . tropica in the southeastern region of Turkey is considered anthroponotic . Similar findings were also reported from Syria [8] . For these reasons , most of our L . tropica strains and Giemsa-stained smears were obtained from southeastern region and analyzed in microsatellite level together with the strains isolated in other endemic areas of Turkey for better understanding their epidemiological origin and population structure . Giemsa stained smears obtained from Syrian patients showed similarities in marker level between Gaziantep and Şanlıurfa samples , while another five Syrian isolates were genetically identical to Şanlıurfa strains . The microsatellite profiles detected in this study are the reflection of real life related to mass human movements in the area as explained above . STRUCTURE analysis and Phylogenetic network clustering identified three main populations among the new isolates . The majority of the populations were congruent with their geographical origins . The oldest Şanlıurfa strain ( EP-40 ) used in the present study was isolated in 1999 after the outbreak in this city , and microsatellite typing was performed for the first time . The MLMT data was compared with a more recent strain isolated in 2007 from the same area ( EP-142 ) and no difference was found . One isolate ( EP-171 ) was viscerotropic and all markers are successfully amplified in the present study . No significant differences were found between other dermatrophic L . tropica isolates , which clustered in same population . One out of twelve markers was found to be heterozygous in this viscerotropic strain . This difference is statistically insignificant as suggested by STRUCTURE analysis . In order to evaluate the effects of clinical outcomes on microsatellite markers , further studies should be performed using more viscerotropic L . tropica strains . Totally , 20 Giemsa-stained smears were included in the present study and twelve of them were obtained from Syrian patients with CL . Unfortunately , no detailed information was available regarding their route of travel and whether the infection was acquired in Turkey or Aleppo , Syria . As seen in ( S1 Table ) and suggested by STRUCTURE ( Fig 3 ) analysis , Syrian smear samples clustered in two different populations and three subpopulations . Even though all Syrian samples and strains were reported to originate from Aleppo , Syria , the MLMT profile of these samples was not identical . The possible reason for this kind of minor changes in microsatellite markers may be the immune status and capability of the patients . Additionally , it can be due to the clonal isolation of the parasite by culturing as previously reported [28] . Our findings showed that the usage of smear samples have some advantages as follows; MLMT can be carried out more rapid without the need for culturing parasite and populations can be analyzed in more detail , which is also important in the meaning of epidemiology . In the present study , two of smear samples grouped in POP A together with 21 isolates from different areas while other 18 smear samples grouped in POP B together with only one isolate . There are two subpopulations consisting of only smear samples originated from POP B . We believe that the smear and strains obtained from same patients needs to be worked together ideally . In the present study , all of the cryopreserved isolates were previously typed by MLEE [5] . Three main populations were identified according to their zymodeme profiles , and the MON profiles were found to be partially congruent with Bayesian clustering method . Except one isolate , all MON-304 isolates clustered in POP-A , which are mostly Şanlıurfa isolates . Only one isolate MON-55 clustered in POP-B with other smear samples . In terms of MON profiles , POP-C is quite heterogeneous that contains five different groups . Microsatellite analysis of different MON profiles was studied in a previous study , and L . infantum MON-1 and non MON-1 group were successfully identified and found to be congruent with MLMT data [29] . However , another study conducted using Algerian strains of L . infantum reported no association between MON profiles and MLMT profiles [30] . Our results also support the idea that MLMT is the best candidate to be accepted as gold standard by having further discriminatory power [31] . Majority of the studied markers of the Syrian L . tropica samples were found similar to strains isolated from Şanlıurfa city . While Şanlıurfa has the longest border with Syria , and many people cross the border into Turkey in this region , only one out of 12 ( 0 . 8% ) Syrian isolates was found identical to the Şanlıurfa strains . Giemsa-stained smears studied from another border city with Syria , Gaziantep , showed that the microsatellite repeat numbers for the studied markers were highly similar to Şanlıurfa and the Syrian samples , but not identical . Slight differences in marker level were observed between the samples from Gaziantep , Syrian and Şanlıurfa , which was also supported by Bayesian clustering analysis ( S1 Table ) . The marker LIST7010 is noted to be highly variable among the studied Syrian samples . Interestingly , this marker was found heterozygote in all studied Syrian strains but homozygous in those smear samples obtained from Syrian CL patients . Analysis of all the Turkish L . tropica MLMT data available ( 39 samples ) , by STRUCTURE demonstrated the presence of two main populations , Şanlıurfa and Mediterranean . Each population further divided into two and three sub-populations , respectively . The Şanlıurfa population mainly consisted of isolates from southeastern Turkey , whereas the Mediterranean population was quite complex . Samples obtained from Syrian patients were clustered in Şanlıurfa-A sub-population , while sub-population Şanlıurfa-B comprised of samples only isolated from Şanlıurfa ( Table 4 ) . A previous MLMT study analyzed L . tropica strains from different regions revealed that two main parasite populations exist in Turkey . Isolates from Şanlıurfa created a unique subpopulation and all Turkish isolates were clustered in a different population , which differs from Moroccan , but correlates with previous isolates from Israel and Palestine [28] . The MLMT data of the isolates clustered in Şanlıurfa population correlate with findings of previous study and noted to be the major L . tropica population of Turkey [28] . One L . tropica strain ( TRO-35 ) isolated from a patient in the most western part ( Aydın city ) of Turkey clustered in Şanlıurfa population and it shows high similarities in the marker level to those strains isolated from eastern/southeastern part of Turkey . These kinds of discrepancies were also reported in previous studies and might have arisen from being infected with the parasite while visiting the other endemic regions for leishmaniasis [18] . In conclusion; in comparison to previous studies , our findings suggest that , the epidemiology of L . tropica is much more complicated . A new population , which comprised of Syrian L . tropica samples , was reported for the first time in Turkey and data available here will provide epidemiological knowledge to further studies . This study highlights the future role of L . tropica spread across Europe through the current wave of migration of infected people . | Turkey is one of the endemic countries for leishmaniasis , cutaneous and visceral . Cutaneous leishmaniasis caused by Leishmania tropica is a serious public health problem with more than two thousands of local cases each year . Moreover , with the civil war in Syria more than three million refugees were accepted to reside in different cities of Turkey . To date , several population genetic studies were performed using Turkish L . tropica isolates but all previously studied isolates were originating from the same geographical area thus population structure could not be revealed in detail for Turkey . The findings of this study suggests two new population structure for Turkey , which are consist of Syrian and Mediterranean isolates . Our findings provide important knowledge on epidemiology of L . tropica in Turkey and highlights the future role of L . tropica to Europe through the current wave of migration of infected people . | [
"Abstract",
"Introduction",
"Material",
"and",
"methods",
"Results",
"Discussion"
] | [
"turkey",
"(country)",
"heterozygosity",
"medicine",
"and",
"health",
"sciences",
"geographical",
"locations",
"tropical",
"diseases",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"alleles",
"microsatellite",
"loci",
"protozoans",
"leishmania",
"gene",
"types",
"ne... | 2017 | Epidemiological analysis of Leishmania tropica strains and giemsa-stained smears from Syrian and Turkish leishmaniasis patients using multilocus microsatellite typing (MLMT) |
Human visceral leishmaniasis ( VL ) caused by L . infantum and cutaneous leishmaniasis ( CL ) caused by L . tropica and L . infantum have been reported in Turkey . L . infantum is also responsible for canine leishmaniasis ( CanL ) and it is widely common in the country . The main aim of the present study was to design a real-time PCR method based on the internal transcribed spacer 1 ( ITS1 ) region in the diagnosis of all clinical forms of leishmaniasis in Mediterranean , and to identify the species directly from clinical samples . Totally , 315 clinical specimens , human/canine visceral ( blood , bone marrow , lymph node ) and cutaneous ( lesion aspiration ) samples , and 51 Turkish Leishmania isolates typed by isoenzymatic method were included in the study . For optimization , DNA samples of the 34 strains were amplified by conventional ITS1-PCR and then sequenced for designing the primers and probes , allowing the species identification . Following the validation with the isolates , the test was applied on clinical samples and melting temperatures were used for genotyping . A group of PCR products were further sequenced for confirmation and assigning the inter- and intraspecies heterogeneity . The diagnosis of leishmaniasis is successfully achieved by the new real-time PCR method , and the test identified 80 . 43% of human and canine VL samples as L . infantum and 6 . 52% as L . tropica; 52 . 46% of CL samples as L . infantum and 26 . 90% as L . tropica . In 13 . 04% of visceral and 20 . 62% of cutaneous samples , two peaks were observed . Hovewer , the higher peak was found to be concordant with the sequencing results in 96 . 96% , in terms of species identification . The real-time ITS1 PCR assay clearly identified the leishmanial species in 81 . 58% of all clinical samples . Genotypic variations of Leishmania parasites in Turkey within species and intraspecies were observed , and L . tropica is also found as causative agent of human and canine VL in Turkey .
Leishmaniases is a group of diseases caused by more than 20 species of the protozoan genus Leishmania in 98 countries and regions with 350 million people living at risk . The main forms of human disease are visceral ( VL ) , cutaneous ( CL ) and mucocutaneous ( MCL ) leishmaniasis [1] , [2] . Turkey is of special epidemiological interest because it lies at the crossroad between Asia and Europe and it comprises seven geographical regions with environmental and ecological differences . Human leishmaniasis , both visceral ( >40 cases yearly ) and cutaneous ( >2000 cases yearly ) have been reported for centuries in Turkey . Two species of Leishmania are prevalent in Turkey , causing anthroponotic CL ( Leishmania tropica ) , and zoonotic VL ( Leishmania infantum ) [3] , [4] . There is also a single report about the occurrence of L . major variants in specific regions of the country [5] . Human VL and CanL are endemic throughout Mediterranean , Ege , Marmara and Black Sea Regions of western Turkey while sporadic in other regions with higher infection rates in dog populations than human cases . There were two zymodemes of L . infantum ( MON1 and MON-98 ) in dog isolates while all human VL isolates were identified as L . infantum MON-1 by isoenzyme analyses [6]–[8] . Anthroponotic CL caused by L . tropica is highly endemic in the Southeastern Anatolia , East Mediterranean and Ege Regions of Turkey [3] , [6]–[9] . In the South Anatolia , L . infantum in addition to L . tropica has also been reported as a causative agent for human CL [6] , [10] , [11] . Leishmania major is known to be endemic in the countries bordering Turkey to the south: Syria , Iraq and Iran [6] . Parasites isolated from CL patients in Sanliurfa province located in Southeastern Region were all identified as L . tropica MON304 while in Aydin province located in western part , Ege region , were identified as L . tropica MON303 ( 87% ) and L . tropica MON304 ( 13% ) [unpublished data] . A universal PCR method targeting the internal transcribed spacer 1 ( ITS1 ) region between the SSU and 5 . 8S rRNA genes were described for the direct diagnosis of different clinical manifestations of leishmaniasis and parasite identification . This method is applicable where several more than one parasite species are aetiologically relevant . It is highly specific and sensitive detecting approximately 0 . 2 parasites per sample [12] . Most of the medically important Leishmania species are then readily distinguished by DNA sequencing or restriction enzyme analysis of the PCR product . ITS1 PCR restriction fragment length polymorphism ( RFLP ) are used for direct species identification in patient tissues , blood or other samples without prior parasite culturing , microscopic analysis or other technique [6] , [12]–[14] . In endemic areas , the presence of multiple Leishmania species with overlapping clinical features and geographical distribution requires the development of sensitive laboratory tests with Leishmania species identification in order to evaluate the prognosis of human and canine leishmaniasis and to choose appropriate therapies . Species identification will also contribute to better understanding the epidemiology of leishmaniases in Turkey [15] , [16] . Species identification of the agents of leishmaniasis in Turkey is crucial , since the country comprises seven geographical regions with environmental and ecological differences . ITS1 PCR RFLP was successfully performed for clinical samples collected from human leishmaniasis and CanL cases from Turkey [6] . The main aim of the present study was to design a real time PCR method based on internal transcribed spacer 1 ( ITS1 ) region in the diagnosis of all clinical forms of leishmaniasis and identifying of parasite directly from clinical samples ( human and canine ) or Leishmania isolates . We further sequenced the PCR products for confirmation and assigning the inter- and intraspecies heterogeneity .
The study was carried out in two steps; ( a ) designing primer/probes which are specific in genus and species level and optimization of a novel ITS1 real time PCR method using Turkish Leishmania strains previously identified by multilocus enzyme electrophoresis ( MLEE ) technique in Montpellier Reference Center and four international reference strains; ( b ) validation of the method using isolates and different types of clinical samples obtained from only confirmed human and dog leishmaniasis cases . Four international reference controls , L . infantum/chagasi ( MHOM/XX/1999/LRC-L774 ) , L . donovani ( MHOM/IN/1980/DD8 ) , L . tropica ( MHOM/IL/1990/LRC-L590 and MHOM/IL/1996/LRC-L691 ) and L . major ( MHOM/IL/2000/LRC-L779 ) are included . A total of 51 Turkish Leishmania strains were isolated from 5 VL , 38 CL , 8 CanL cases between 2000 and 2011 , and maintained by subcultures in NNN medium . They were identified by the isoenzymatic method previously . Leishmania promastigotes of all isolates were mass cultivated in RPMI+20%FCS medium and centrifuged in 5th day to obtain pellet in order to use DNA extraction The concentration of isolates was adjusted to 2 . 5–3×106 promastigotes/mL . A total of 315 clinical samples obtained from cases with human CL ( n = 223 ) , human VL ( n = 40 ) and CanL ( n = 52 ) originated from 31 different provinces ( mostly from Izmir , Aydin , Hatay and Şanlıurfa provinces ) of Turkey were included in the study . The samples were collected between April 2007 and May 2010 and they were all confirmed cases who are found positive/seropositive by different parasitological/serological methods in our laboratory ( Table 1 , Figure 1 ) . Blood and bone marrow samples were taken from hospitalized VL patients , and sent to our laboratory in tubes containing EDTA as anticoagulant or on slides . Two hundred microliter sample was used for DNA extraction . Tissue aspirates from CL patients were collected into syringes containing 0 . 5 mL of saline , a slide was prepared for DNA extraction and part of which was inoculated into NNN medium for isolating the parasite . All the prepared slides were washed with PBS and then this solution was transferred into a 1 . 5 ml eppendorf tube . The procedures were performed based on the steps mentioned on the DNA extraction kit ( Roche Applied Science ) for all samples . Quality and quantity of extracted DNA was analysed by spectrophotometry . Conventional ITS1-PCR was applied to 30 Turkish Leishmania strains ( obtained from 18 CL , 5 VL , 7 CanL cases ) and four international reference strains using the primer set ( forward – LITSR; reverse - L5 . 8S ) and conditions published by El Tai et al [17] , [18] . PCR products were sequenced commercially by RefGen ( http://www . refgen . com ) and compared by multialignment analysis within each other and with other ITS1 sequences published in BLAST using MultAlin program ( CLC Main Workbench ( v . 5 . 6 ) genetic program ) . The forward primer ( LITSR ) used in the initial experiment was kept but reverse primer ( ITS1R-TR1: 5′- GAAGCCAAGTCATCCATCGC -3′ ) and the probes ( Probe1 : CCGTTTATACAAAAAATATACGGCGTTTCGGTTT—FL; Probe 2: LC640-GCGGGGTGGGTGCGTGTGTG—PH ) were newly designed according to the variable region for detecting L . donovani complex , L . tropica and L . major , using LightCycler Probe Design Software 2 . 0 program [19] . The real time PCR method targeting ITS1 region between the SSU and 5 . 8S rRNA genes specific for Leishmania was first applied to 51 Turkish and international isolates to determine the melting temperatures ( Tm ) for each species . Then , the method was performed using clinical samples and it was repeated twice for each batch of samples . One positive and two negative controls were included for each PCR reaction . ITS1 real time PCR method was applied using samples containing 20–50 ng of genomic DNA , 400 nM of each primers , 200 nM of each probes , 2 mM of MgCl2 , 1 µl LightCycler FastStart DNA Master Hybridisation probe ( Roche Applied Science ) , and 1 , 5 µl PCR grade water ( Roche Applied Science ) to a reaction total volume of 10 µl . PCR amplification was performed as follows: one cycle of 10 minutes at 95°C , followed by 45 cycles consisting of denaturation at 95°C for 10 seconds , annealing at 50°C for 10 seconds , extension at 72°C for 20 seconds , and melting at 95°C for 0 second , 50°C for 10 second , 40°C for 10 second , 80°C for 0 second and cooling at 40°C for 30 seconds . Melting curves were analysed using channel 2 and 3 . A group of PCR products from 135 samples ( 121 human; 14 dogs ) including 93 clinical specimens and 42 Turkish Leishmania isolates were sequenced commercially by RefGen ( http://www . refgen . com ) for the confirmation of the results . The clinical samples group was consisted of DNA samples from 111 lesion aspiration samples of CL patients , 4 blood and 6 bone marrow samples of VL patients; 12 lymph node aspiration and 2 blood samples of dogs . Sequence data were analysed using ClustalW2 ( http://www . ebi . ac . uk/Tools/msa/clustalw2/ ) program and distance of molecular relationship in the group of samples detected as L . tropica and L . infantum were calculated . Phylogram was generated using CLC Main Workbench ( v . 5 . 6 ) genetic program . Univariant and Tukey analyses were performed using SPSS v . 15 program in order to compare the melting temperatures in the group of samples detected as L . tropica and L . infantum . Statistical significance degree was accepted as <0 . 05 . The study was approved by Local Animal Care and Ethics Committee of the School of Medicine and Ege University Medical School Clinical Research Ethical Committee , Izmir , Turkey .
The melting temperatures were detected as 68°C Tm for L . donovani complex; 62°C Tm for L . tropica and 53°C Tm for L . major . The sequences of variable region which were used for designing probes in Turkish Leishmania isolates were shown in figure S1 . The identification results of 51 Turkish Leishmania isolates were presented with the results of MLEE comparatively in table S1 . Result of the isoenzymatic method was used as gold standard . As indicated in the table S1 , species identification of 48 isolates well matched with MLEE results , and sensitivity of the method was found to be 94 . 11% . Out of three isolates which are not fully in aggrement with MLEE , two strains ( C010 and C056 ) are found to be L . tropica with PCR while they both were L . infantum by MLEE . One strain ( C078 ) gave two peaks with the higher related to L . tropica ( Figure 2 ) as concordant with the MLEE analysis ( L . tropica MON312 ) . Sequencing results of the 42 strains ( 10 L . infantum and 32 L . tropica ) were used to construct phylogenetic tree . L . infantum isolates constituted a single group while strains determined as L . tropica showed several different groups ( Figure 3 ) . The partial sequences of L . tropica ITS1 region representing MON200 , MON303 and MON304 zymodems were submitted to GenBank ( Accession numbers KC686338 , KC679052 and KC609747 , respectively ) . Totally , 315 clinical samples were analysed and all of them were diagnosed as positive by ITS1 real time PCR method in genus level . In species level , L . infantum and L . tropica were detected while no L . major was found among clinical samples . Genotyping identified 80 . 43% ( 74/92 ) of human and canine visceral leishmaniasis samples as L . infantum and 6 . 52% ( 6/92 ) as L . tropica; 52 . 46% ( 117/223 ) of cutaneous samples as L . infantum and 26 . 90% ( 60/223 ) as L . tropica . Fifty-eight ( 18 . 41% ) out of 315 clinical samples ( 12 of visceral and 46 of cutaneous samples ) gave two peaks and two melting temperatures were observed . The standart deviation of melting temperatures was found between 0 . 4512 and 0 . 5026 in one peak samples while it was between 0 . 3718 and 1 . 0858 in two peak samples . The sequencing was done for 102 samples ( 3 blood and 1 bone marrow samples of VL patients; 2 blood samples of dogs; 56 lesion aspiration samples of CL patients; 40 strains from CL , VL and CanL cases ) giving one peak and 33 samples ( 1 blood samples of VL patient; 1 dog blood sample; 29 lesion aspiration samples of CL patients and 2 strains from CL cases ) giving two peaks . The sequencing results were concordant with the real time ITS1 PCR results in 98 . 03% ( 100/102 ) samples . For the two peaks samples , a 96 . 96% ( 32/33 ) concordance were detected between the higher peak in the real time ITS1 PCR and sequencing results indicating that the higher peak can be decisive for the species identification . The ITS1 real time PCR results of clinical samples were classified into three groups which were clearly distinguishable as ( a ) 191 samples diagnosed as L . infantum including 29 visceral , 45 dog and 117 cutaneous samples; ( b ) 66 samples of L . tropica including 60 cutaneous and four visceral and two dog samples and ( c ) 58 samples of two peaks including 46 cutaneous , seven visceral and five dog samples . The working process and the results were summarized in Figure 4 . L . infantum and L . tropica were found to be causative agents of both clinical forms of human leishmaniasis as well as canine leishmaniasis in Turkey .
The various leishmaniases are caused by different species of Leishmania , some of which co-exist in the same region like several endemic areas in Turkey; therefore it is crucial to distinguish the species for diagnosis , treatment and epidemiological purposes . Microscopical examination of stained tissue preparations and culture of tissue aspirates for diagnosis , and multilocus enzyme electrophoresis ( MLEE ) for species identification are accepted as gold standards . However , all the conventional methods applied in the diagnosis of leishmaniasis have medium to low sensitivity and the amplification of DNA by PCR , using different genomic and kDNA targets which is shown to be more sensitive is gradually replacing the traditional methods for the diagnosis of leishmaniasis . Besides , molecular phylogeny studies of the parasite have increasingly suggested new approaches regarding treatment , prognosis of the disease , the distribution of Leishmania species in human and animal hosts , as well as in insect vectors for designing appropriate control measures [12] , [20]–[22] . Different PCR/PCR-RFLP based methods targeting kinetoplastic DNA , telomeric sequences , gp63 , miniexons , β-tubulin , or ribosomal RNA encoding genes ( particularly the internal transcribed spacers , ITS ) and recently microsatellites , have been proposed for species identification in Leishmania parasites using isolates and clinical samples [23] , [24] . The ITS1 of the ribosomal DNA repeat unit ( rDNA-ITS1 ) has previously been exploited for Old World Leishmania species discrimination using RFLP [12]–[14] , reverse hybridization assays [25] , and sequencing [26] . There are an estimated 20 to 200 identical copies in the Leishmania genome , making it a good target for analyzing low parasite quantities [13] , [14] , [27] . A new technique applying PCR-SSCP ( single-strand conformational polymorphism ) was published recently by Chargui et al . The authors notified that SSCP provides more resolution especially when PCR product has weak band on agarose gel and is less expensive than RFLP method [24] . In our study , we present a real-time ITS1-PCR method that can diagnose three Old World Leishmania species L . donovani complex , L . tropica and L . major , using newly designed probes to diagnose and simultaneously differentiate between Turkish species in clinical samples . The real time ITS1 PCR assay clearly identified the leishmanial species in 81 . 58% of all clinical samples . L . infantum was identified in 80 . 43% of human and dog visceral samples while L . tropica was detected in 6 . 52% . In five isolates ( C41 , C56 , C087 , C39 , C40 ) from 3 VL patient and 2 dogs , the agent was diagnosed as L . tropica . This finding was also supported by isoenzyme typing of the strain C87 , resulting as L . tropica MON-315 in Montpellier leishmaniasis reference center . This is the first report notifying L . tropica as a causative agent of human and canine VL in Turkey although , L . tropica has been reported to be isolated from human and canine VL cases in different countries [28]–[34] . Several studies and case reports published from neighbouring country , Iran , showed that L . tropica can rarely cause visceral leishmaniasis . Alborzi et al [35] identified only one L . tropica among 64 bone marrow/spleen samples while another group also found only one L . tropica out of 11 isolates obtained from dogs in Iran [36] . L . infantum is reported as a rare causative agent of CL most notably in the Mediterranean Basin countries such as Tunisia , Algeria , Morocco , Spain , Italy , Portugal , Greece and France [2] , [37]–[40] . Although L . tropica is the main causative agent of CL in Turkey , L . infantum has dominancy in the South Anatolia of Turkey , mainly in Hatay and Adana provinces according to identification of Leishmania strains by molecular techniques using miniexon gene PCR-RFLP [10] , [11] , kDNA-PCR [41] and ITS1 PCR-RFLP [6] assays . In our study , it is confirmed that both Leishmania species can be found in clinical samples from CL patients in this area . L . infantum is also isolated from proven vector sand fly species , P . tobbi by Svobodova et al . in the region [11] . So far only little is known about the population structures of the two Leishmania species in Turkey and the correlation with geographical origin , biogeographical parameters , clinical outcome , involved animal reservoir ( dogs ) and the transmitting sand fly species . Several pioneer studies were performed that used PCR-RFLP and/or sequence analysis of the ITS rDNA , minicircle kinetoplast DNA , miniexon , NAGT [5] , [6] , [10] , [42] mainly for species identification . In Turkey , nine zymodemes of L . tropica and four of L . infantum were described so far [43 , unpublished data] . The high degree of heterogeneity in L . tropica species has been reported [13] , [14] , [44] and shown by MLEE [30] . This variation was reported in different level in Old World Leishmania species by ITS PCR-RFLP analysis from highest to lowest in order of L . tropica>L . aethiopica>L . major>L . donovani [44] . A study using PCR-RFLP/sequencing based on ITS1 region was performed in Iran using clinical samples from CL patients and they showed six different genotype groups of L . tropica [45] . In our study , the heterogeneity of L . tropica was also observed in phylogenetic analysis . L . infantum isolates constituted a single group while strains determined as L . tropica showed several different groups . The main difficulty in the present study was to observe two peaks in some samples . It is probably due to genetic variety in ITS1 region . Genetic polymorphism in ITS region of different strains of same Leishmania species and possibility of heterogeneity in individual copies was described by El-Tai et al [17] . In a recent study , after digestion of the amplification product with the HaeIII , the ITS1 PCR assay clearly identified the leishmanial species in CL samples in only 72 . 3% and L . tropica was found to be the most dominant [20] . In our study , the failing of species identification in 18 . 42% of the samples could be either due to the minuteness of the DNA , possibility of a mix infection , hybridization of different species or intraspecies variations in Turkish Leishmania parasites as also commented by Kifaya et al . [20] . We also found that there are 6 copies of our probe region containing variable part in ITS1 sequence ( data not shown ) . Therefore , we evaluated that if all copies are identical in the DNA sample one peak was obtained; if not , several copies have one or two bases difference , two peaks were obtained but always one peak is higher . In the case of detection of two peaks , we decided to take higher one for species identification with the support of sequence analysis and it can be acceptable that the identification in species level was done in all clinical samples . Hereby , we can speculate that our assay could also propose intragenomic heterogeneity of particular isolates and/or Leishmania DNA samples obtained from clinical materials . The statistical analyses of melting temperatures in the samples having one and two peaks was also performed and standart deviation was found very low in one peak samples than two peak samples . After we compared these results with sequencing results , we also found highest variability in L . tropica group . Gelanew et al . performed PCR-RFLP/direct sequencing assay using L . donovani strains and the direct sequencing of both strands of ITS1 DNA showed the presence of multiple peaks in the chromatograms , which could possibly have resulted from: ( i ) the presence of multiple strains or clones of Leishmania; ( ii ) the presence of a hybrid genotype; ( iii ) intragenomic variation in the multicopy ITS1; or a combination of these [46] . This should be studied furtherly by multiple gene targets and innovative methods like Multilocus Sequence Typing ( MLST ) of the genes encoding the proteins used for MLEE and Multilocus Microsatellite Typing ( MLMT ) as well as experimental animal infection studies using Turkish isolates . Two aims were achieved through the analysis of real time ITS1 PCR products by sequencing of the reference and local strains as designing new probes and proving the sensitivity of the technique at the genus and species identification level . The results of isolates in the optimization step also showed that real time ITS1 PCR results are highly concordant with MLEE analysis ( 96 . 07% ) . We can also propose that the method can identify inter- and intraspecies variability based on ITS1 region but cannot differentiate L . donovani complex species within each other like other molecular assays using ITS1 region [26] . In this point , we would like to emphasize that the studies addressed to L . donovani complex species ( L . donovani and L . infantum ) identification should be planned . In conclusion , the proposed method presents a sufficient sensitivity for fast and correct diagnosis of leishmaniasis in all type of clinical samples but due to the samples giving two peaks the ability of the method for species identification is limited and needs further analyses . However , the higher peak was always very well matched with the results of sequence analysis . Genotypic variations based on ITS1 region of Leishmania parasites in Turkey within species and intraspecies were determined . The findings in this study were showed that L . tropica is one of the causative agents of human and canine visceral leishmaniasis in Turkey . | Leishmaniasis caused by Leishmania parasites are seen as cutaneous ( CL ) and visceral ( VL ) clinical forms in Turkey . Leishmania ( L . ) tropica and L . infantum were determined as CL agents , while L . infantum was incriminated for VL in the country . Canine leishmaniasis ( CanL ) is widely common throughout the country and L . infantum is the responsible agent of the disease . Related to Leishmania species diversity and different clinical forms in human and dogs in this geographical area , the identification of the parasite species prefers to be done during the time of diagnosis . Internal transcribed spacer region was chosen as the target area for developing a real-time PCR assay to use as a fast and standardized diagnostic method and species identification simultaneously . Clinical samples from parasitologically/serologically proven cases and isolates were included the study , and high positivity rate in species identification was obtained . The method can also determine the intragenomic heterogeneity in Leishmania tropica and L . infantum . The assay presents a sufficient sensitivity for fast and correct detection of leishmaniasis directly from clinical materials . L . tropica and L . infantum were found as causative agents of human CL , VL and CanL in Turkey . Knowledge about differences in the parasites is useful for future studies in Turkey . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"leishmaniasis",
"medicine",
"neglected",
"tropical",
"diseases",
"infectious",
"diseases"
] | 2013 | A Real-Time ITS1-PCR Based Method in the Diagnosis and Species Identification of Leishmania Parasite from Human and Dog Clinical Samples in Turkey |
Ticks are distributed worldwide and affect human and animal health by transmitting diverse infectious agents . Effective vaccines against most tick-borne pathogens are not currently available . In this study , we characterized a tick histamine release factor ( tHRF ) from Ixodes scapularis and addressed the vaccine potential of this antigen in the context of tick engorgement and B . burgdorferi transmission . Results from western blotting and quantitative Reverse Transcription-PCR showed that tHRF is secreted in tick saliva , and upregulated in Borrelia burgdorferi-infected ticks . Further , the expression of tHRF was coincident with the rapid feeding phase of the tick , suggesting a role for tHRF in tick engorgement and concomitantly , for efficient B . burgdorferi transmission . Silencing tHRF by RNA interference ( RNAi ) significantly impaired tick feeding and decreased B . burgdorferi burden in mice . Interfering with tHRF by actively immunizing mice with recombinant tHRF , or passively transferring tHRF antiserum , also markedly reduced the efficiency of tick feeding and B . burgdorferi burden in mice . Recombinant tHRF was able to bind to host basophils and stimulate histamine release . Therefore , we speculate that tHRF might function in vivo to modulate vascular permeability and increase blood flow to the tick bite-site , facilitating tick engorgement . These findings suggest that blocking tHRF might offer a viable strategy to complement ongoing efforts to develop vaccines to block tick feeding and transmission of tick-borne pathogens .
Ticks are distributed worldwide and affect human and animal health by transmitting diverse infectious agents . Ticks are considered to be second to mosquitoes as major vectors of human diseases [1] , [2] . For example , Ixodes spp . , transmit Borrelia burgdorferi ( the Lyme disease agent ) , Anaplasma phagocytophilum ( the cause of human granulocytic anaplasmosis ) , Babesia microti , and tick-borne encephalitis virus ( TBEV ) , among other pathogens [1] , [3] . Effective vaccines against most tick-borne pathogens are not currently available and there is an urgent need for the control of ticks and their associated pathogens [4] . Typical vaccines target microbes directly , using extracts of the organism , or recombinant antigens as the immunogen . For example , B . burgdorferi outer surface protein A has been extensively studied and resulted in an Federal Drug Administration-approved vaccine that was commercially available from 1998 until 2002 [5] , [6] . Currently one vaccine is approved and available for protection against a tick-borne pathogen – TBEV , which is transmitted by I . ricinus in Northern Europe and Asia [4] . The transmission of tick-borne pathogens can also theoretically be prevented by interfering with the ability of ticks to feed on a mammalian host [7] . A pilot study by Allen and Humphreys several decades ago , suggested that vaccines based on tick gut antigens successfully reduced Boophilus engorgement on cattle [8] . Recently , immunization of guinea pigs with a tick salivary antigen , sialostatin L2 , diminished the capacity of Ixodes scapularis nymphs to feed [9] . While reducing the ability of tick feeding , tick-based vaccines may have another equally important impact – to decrease the chance of transmission of tick-borne pathogens [10] . Immunization of cattle with Bm86 vaccines resulted in lower infestations as well as decreased incidence of babesiosis and Anaplasma marginale infection in some regions [1] , [7] . Repeated exposure of guinea pigs to ticks causes acquired resistance of the animals to subsequent tick bites [11] , [12] , and this development of “tick-immunity” can decrease tick-transmitted B . burgdorferi infection [13] . B . burgdorferi need to replicate within the ticks during blood feeding and are transmitted to the host after 36–48 h of tick attachment [2] , [14] , [15] , [16] . Thus , impairing I . scapularis feeding could be another useful strategy to reduce B . burgdorferi transmission . Tick saliva contains molecules that are important for formation and maintenance of the feeding cavity in the host dermis , as well as the transmission of tick-borne pathogens [17] , [18] . These activities include anti-hemostatic , anti-inflammatory and immunomodulatory effects , among others [17] , [19] . Histamine binding proteins are well characterized and suggested to be important to neutralize the inflammatory effect of histamine , which is secreted by host immune cells at the tick feeding site and critical for Ixodes ticks to successfully attach to the host [20] , [21] . Interestingly , Dermacentor variabilis ticks also express a protein in their saliva , which shares high homology with mammalian histamine release factor [22] . Given the deleterious effects of histamine on tick physiology , it is very surprising that ticks encode a histamine release protein that would presumably stimulate histamine secretion . The role of the tick histamine release factor in vivo during tick feeding is not understood and warrants detailed examination . Tick feeding can be divided into a series of 9 stages [2] beginning with host seeking , and culminating in engorgement on the host followed by detachment and dropping off the host . A feeding lesion is established about 24 h post attachment , and during this early phase of feeding there is minimal blood intake . Blood meal ingestion begins slowly around 48 h post tick attachment followed by rapid feeding to repletion around 72 h–96 h post tick attachment ( late stage ) . While it is recognized that the I . scapularis salivary gland proteome changes during these early and late phases of feeding [13] , a molecular understanding of these events remains to be elucidated . In this study , we have characterized a putative histamine release factor from I . scapularis , the predominant vector of B . burgdorferi , the agent of Lyme disease in North America . We invoke a pivotal role for I . scapularis HRF during the rapid phase of tick feeding , and address the vaccine potential of this antigen in the context of tick engorgement and B . burgdorferi transmission .
To identify tick proteins that may be utilized by B . burgdorferi to facilitate transmission , 2-dimensional fluorescence difference gel electrophoresis ( DIGE ) was performed using extracts from B . burgdorferi-infected , and uninfected , I . scapularis salivary glands . Seventeen differentially expressed proteins ( 5-fold or more expression levels in Borrelia-infected salivary glands ) were selected for mass spectrometric analysis , and 4 I . scapularis proteins were unambiguously identified with significant MASCOT scores ( p<0 . 05 ) of 79 ( Table S1 ) . In this study , we characterize one of the most highly induced proteins , named tHRF because it shares high homology with a murine histamine release factor ( 57 . 1% similarity and 40 . 1% identity at amino acid level ) . tHRF mRNA levels were induced during I . scapularis engorgement , and significantly higher in B . burgdorferi-infected , than in uninfected , ticks ( p<0 . 01 ) . Immunoblots using tHRF antiserum further demonstrated a ∼2 . 5 fold up-regulation of tHRF in B . burgdorferi-infected ticks ( Figure 1 , A , C–D ) . tHRF was present in tick saliva , as well as in the salivary glands and midgut , indicating that it is a secreted protein ( Figure 1E ) . To analyze the potential role of tHRF in tick feeding , and also during B . burgdorferi transmission , tHRF-deficient I . scapularis nymphs were generated by RNA interference ( RNAi ) . Buffer-injected ( MOCK ) , SSRB ( another tick gene- Single Sequence Receptor Beta- found in our 2DIGE list , used as a control ) or tHRF double-stranded RNA ( dsRNA ) -injected B . burgdorferi-infected nymphs were allowed to engorge on mice . The silencing of tHRF and SSRB were confirmed by quantitative RT-PCR ( Figure 2A ) . After 3 days , the weighs of tHRF-deficient ticks were significantly lower than control ticks ( Figure 2B ) . Q-PCR revealed a decrease in spirochete levels in tHRF-deficient ticks , as well as in the skin of mice that were fed upon by tHRF-deficient ticks ( Figure 2 , C and D ) . At 3 weeks , when spirochetes have disseminated to diverse organs , the B . burgdorferi burden in the heart and joints was also lower in mice infected by tHRF-deficient ticks , compared to that in mice infected with control ticks ( Figure 2 , E and F ) . To further show that tHRF directly influences tick feeding , an RNAi study was performed with nymphs that were not infected with B . burgdorferi . Consistent with the results using B . burgdorferi-infected ticks , the tick weight was significantly decreased in tHRF-dsRNA-treated uninfected I . scapularis after feeding ( Figure 2G ) . These data show that tHRF is critical for tick feeding , regardless of whether B . burgdorferi are present within ticks . To further examine the importance of tHRF during tick feeding , and its influence on B . burgdorferi transmission , a passive immunization study was performed in naive mice . Groups of 5 mice were administered 200 µl of tHRF antiserum , or control sera ( normal rabbit serum or Salp25D antiserum; Salp25D is a tick salivary protein that does not influence tick feeding [23] ) . One day later , 6 B . burgdorferi-infected ticks were placed on each mouse and tick weights were assessed after 3 days of feeding . Ticks engorging on tHRF antiserum-treated mice weighed significantly less than ticks that fed on control mice ( Figure 3A ) . The spirochete burden in ticks was also substantially lower in I . scapularis that fed on tHRF antiserum-immunized mice ( Figure 3B ) . B . burgdorferi burden was also markedly reduced in tHRF antiserum-immunized mice . The spirochete load in murine skin at day 7 post-infection and in joints and hearts at 3 weeks post-infection was markedly lower in tHRF antiserum-immunized group compared to control serum immunized group ( Figure 3 , C–E ) . About 20–27% of the tHRF antiserum-immunized mice ( N = 15 ) were fully protected ( based on the absence of a detectable flaB signal in Q-PCR ) , while 100% of the control animals were infected ( N = 30 ) ( Figure 3 , C–E ) . Uninfected I . scapularis nymphs also fed less efficiently on tHRF antiserum-treated mice ( Figure 3F ) as seen by decreased engorgement weights compared to ticks fed on control antiserum-treated mice . We then assessed the ability of the ticks to feed on mice actively immunized with tHRF . Group of 5 mice were immunized with recombinant tHRF , or adjuvant ( control ) . Immunoblots confirmed that mice generated antibodies against tHRF following active immunization ( Figure 4A ) . The tick weights were significantly decreased when B . burgdorferi-infected , or uninfected , nymphs fed on tHRF immunized mice compared to ticks that fed on control mice ( Figure 4 , B and C ) . The spirochete load was also markedly reduced in ticks fed on tHRF-immunized mice ( Figure 4D ) and in murine skin ( at day 7 post-infection ) ( Figure 4E ) and in joints and hearts ( at 3 weeks post-infection ) in the tHRF-immunized group compared to that in control mice ( Figure 4 , F–G ) . 20–33% of tHRF immunized mice ( N = 15 ) were PCR negative , while 100% of the mice in the control groups ( N = 15 ) were PCR positive for B . burgdorferi flaB amplicon ( Figure 4 , E–G ) . Our above experiments focused on 72 h post tick attachment- a specific time point at which 30–40% of the ticks from the control groups successfully complete engorgement and drop off the mice , and the remaining ticks nearing engorgement . To address the role of tHRF on 72–96h post tick attachment , all the ticks were allowed to feed to repletion on tHRF antiserum immunized mice or control mice . While , ticks in the control group fed to repletion and detached around 72–84 h of attachment , ticks fed on tHRF-immunized animals fed to repletion around 96 h after attachment . Further , 10–20% of ticks from the tHRF group remained attached to the mouse even after 96 h , ( Figure S1A ) . The engorgement weights of ticks fed on tHRF-antiserum immunized mice were also significantly less than the engorgement weights of ticks fed on control mice ( Figure S1B ) , consistent with our data obtained from 72 h fed ticks ( Figure 2–4 ) . Mammalian histamine release factor binds to basophils and stimulates histamine release [24] . Since tHRF shares substantial homology with mammalian histamine release factors , we postulated that tHRF might also adhere to host basophils and induce histamine secretion . An in vitro binding assay was performed using a rat basophil cell line and recombinant tHRF ( tHRF shares 57% similarity with rat HRF at amino acid level , and rat HRF is 100% identical to mouse HRF ) . Flow cytometry and confocal imaging showed that recombinant tHRF bound to rat basophils ( Figure 5 , A and B ) . To examine the influence of tHRF on histamine release , basophils were incubated with recombinant tHRF , or nymphal tick salivary gland extracts ( T . SGE ) . Recombinant tHRF and tick salivary gland extracts stimulated histamine release from basophils ( Figure 5C ) . Ticks are sensitive to histamine during the early stage of blood feeding , and express histamine binding proteins to counteract this effect . However , tick sensitivity to histamine wanes after 3 days of attachment to a host [20] . Quantitative RT-PCR analysis showed preferential expressions of 3 histamine binding proteins in the salivary glands of I . scapularis nymphs at 24–48 h post tick attachment ( Figure 6A–C ) . However , tHRF was preferentially expressed at 48–72 h post tick attachment ( Figure 6 D ) . Since tHRF induces histamine release , histamine might play an under-appreciated role in the late/rapid phase of tick feeding . To confirm this , histamine or recombinant tHRF was injected into the skin - at the I . scapularis bite site - 60 h after tick-attachment . The tick weights at 72 h were significantly increased when I . scapularis nymphs fed on mice given histamine or recombinant tHRF compared to ticks fed on control mice ( Figure 6 , E and F ) . The B . burgdorferi burden was also higher in ticks fed on tHRF-treated mice compared to ticks that fed on control mice ( Figure 6G ) .
The incidence of tick-borne diseases has steadily increased over the past few years , and effective vaccines against most tick-borne pathogens are not currently available [4] . I . scapularis is the major vector of Lyme disease in the USA [2] , [15] . Further , I scapularis can serve as efficient vectors of A . phagocytophilum , B . microti , and Powassan virus ( a tick-borne encephalitis causing virus ) . The last decade has seen an increased functional understanding of tick salivary proteins and their critical interactions with the host and pathogen [17] , [18] , [25] . This information has also offered a new approach to develop effective vaccines against ticks and the pathogens they transmit by simultaneously targeting the pathogen and the tick [26] . The identification of tick proteins potentially involved in pathogen transmission is an important step in the development of effective tick vaccines [1] , [18] . The presence of B . burgdorferi within ticks may alter the expression level of selected genes that encode antigens in saliva [19] , [27] . One of best characterized genes is salp15 [27]; our recent study suggests that immunization with Salp15 could reduce the transmission of B . burgdorferi from infected ticks to mice , although Salp15 antibodies did not influence the ability of ticks to feed . The action is mainly due to the interaction between Salp15 antibody , Salp15 and Borrelia [26] . We performed a 2DIGE analysis to identify additional tick salivary proteins modulated by spirochetes . We found that tHRF was up-regulated in Borrelia-infected tick salivary glands . HRF is an evolutionally conserved multiple-function protein [28] , also a novel cytokine that provokes the release of histamine by both IgE-dependent and IgE-independent mechanisms from mammalian basophils and mast cells [29] . In addition to mammalian HRF , HRF homologs have also been identified in Plasmodium falciparum parasite [30] , Dermacentor variabilis [22] , [31] , [32] and Dermanyssus gallinae [33] . The latter study further indicated that antibodies against HRF increased the mortality of the mites after engorgement , suggesting its potential as a vaccine antigen [33] . Histamine , secreted by basophils in blood and mast cells in tissues , plays a deleterious role during tick feeding . Histamine is a mediator of the itch response and promotes the recruitment of pro-inflammatory cells to the tick bite site – and these immune response prevent tick attachment to the skin of the host [20] , [34] . However , the Ixodes tick encodes several histamine binding proteins ( HBPs ) to counteract the effect of histamine [21] , [35] . The elaboration of a histamine release factor in tick salivary glands therefore seemed counterintuitive , since such an activity would be detrimental to tick feeding . Mulenga et al . [22] suggested that ticks might need either HBPs or HRF , depending on its feeding phase . Tick feeding involves a complex series of 9 sequential stages [2] . Host seeking and engagement with the host precede actual tick attachment and establishment of the feeding lesion . The early phase of tick feeding that lasts about 24 h post attachment is sensitive to histamine [20] . We observed increased expression of Histamine Binding Proteins ( HBP ) in I . scapularis nymphal salivary glands during this early phase of tick feeding and might be critical to counter the effect of histamine ( Figure 6 ) . Ticks imbibe very little blood during this early phase of feeding . About 60–72h post tick attachment , which includes the rapid feeding phase , tick sensitivity to histamine significantly declines [20] , [34] , and ticks fully engorge . During this phase the expressions of HBPs appear to be significantly decreased , and the expression of tHRF increases ( Figure 6 ) . We speculate that this reciprocal expression of HBPs and tHRF might help increase the local concentration of histamine at the tick-feeding site during the rapid feeding phase . Increased histamine concentration might modulate the vascular permeability to enhance blood flow into the tick feeding site and facilitate tick engorgement . B . burgdorferi replicate after the tick begins to take a blood meal , and transmission to the host begins about 36–48 h post tick attachment , a time coincident with active spirochete replication and migration to the salivary glands [14] , [36] . Temperature and host blood are critical signals for B . burgdorferi replication and dissemination from the midgut . Since the feeding ability of the tHRF-deficient ticks was significantly impaired , the replication and dissemination of Borrelia inside the ticks was also significantly decreased ( Figure 2C ) . Consequently , the spirochete transmission from tick to mouse was also reduced ( Figure 2D ) , and 20–30% of the mice immunized with tHRF were fully protected from Borrelia infection based on the absence of a detectable flaB signal in Q-PCR ( Figure 3 and 4 ) . It is also conceivable that the vasodilatory effect of histamine , might additionally contribute to the efficient dissemination of Borrelia from the original tick-feeding site , where they are deposited , to distal sites . In summary , for the first time , we demonstrate that the I . scapularis salivary protein tHRF is critical for the tick engorgement , and consequently also facilitates Borrelia transmission to the murine host . We show that B . burgdorferi upregulates the expression of tHRF and immunization with tHRF significantly impairs tick feeding , and decreases B . burgdorferi burden in mice . Importantly , these observations underscore the dynamic nature of the temporal interactions between the vector , the host and the pathogen . While vaccine targeting of tHRF alone might not be sufficient to thwart tick feeding and spirochete transmission , blocking tHRF might offer a viable strategy to complement ongoing efforts to develop vaccines to block tick feeding and transmission of tick-borne pathogens .
An infectious and low passage isolate of B . burgdorferi N40 was used to generate B . burgdorferi-infected ticks . Larval , nymphal , and adult I . scapularis were maintained in our laboratory . Clean larvae were fed either on naïve C3H mice to generate naïve nymphs or on B . burgdorferi-infected C3H mice to generate infected nymphs . Female C3H/HeJ ( C3H ) mice , 4 to 6 weeks of age , were obtained from the Jackson Laboratory . Animals were housed and handled under the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The animal experimental protocol was approved by the Yale University's Institutional Animal Care & Use Committee ( Protocol Permit Number: 2008-07941 ) . All animal infection experiments were performed in a Bio-safety Level 2 animal facility , according to the regulations of Yale University . A quantitative analysis of the I . scapularis salivary gland proteome was carried out by 2D fluorescence differential gel electrophoresis ( DIGE ) at the W . M . Keck Facility at Yale University . Salivary gland extracts from 200 clean and Borrelia-infected I . scapularis nymphs fed for 66–72 h were suspended in a cell lysis buffer ( 7M urea , 2M thiourea , 4% CHAPS , 25 mM Tris , pH 8 . 6 at 4°C ) and equal amounts of protein ( 50 µg ) from Borrelia-infected and clean salivary gland extracts were then differentially labeled in vitro with Cy3 and Cy5 N-hydroxysuccinimidyl ester dyes as described in the Ettan DIGE manual ( GE Healthcare , NJ ) and electrophoresis and analysis performed essentially as described earlier [13] . The gel was sequentially scanned using the Typhoon 9410 Imager ( GE Healthcare , Piscataway , NJ ) and images exported into the DeCyder ( GE Healthcare , NJ ) software package to assess differentially expressed protein spots . The protein spots that were increased at least 5-fold in Borrelia-infected salivary glands were excised for identification . The gel spots of interest were robotically digested using trypsin prior to analysis on an Applied Biosystems 4800 MALDI-Tof/Tof mass spectrometer . The data was analyzed using the Applied Biosystems GPS Explorer software with Mascot analysis against the NCBI nr database , and a combined peptide mass fingerprint/MS/MS search was done . Spots identified with significant MASCOT scores ( P<0 . 05 ) of 79 were tabulated . Fed-nymph salivary gland cDNA was prepared as described [37] and used as template to amplify cDNA of tHRF ( GenBank accession no . DQ066335 ) , and SSRB ( Signal sequence receptor beta , another tick gene used as a control ) ( GenBank accession no . DQ066202 ) . The primer sequences are indicated in Table S2 ( P11 , 12 for tHRF; P13 , 14 for SSRB ) . The resultant amplicons were purified and cloned into the SacII-XhoI sites of the L4440 double T7 Script II vector [37] . dsRNA complementary to the DNA insert was synthesized by in vitro transcription using the Megascript RNAi kit ( Ambion , Austin , TX ) . The dsRNA was purified and quantified spectroscopically . The microinjection of dsRNA was performed as described previously [37] . Briefly , we injected ≈4 nl of dsRNA ( 1×109 molecules per nl ) or buffer alone ( MOCK ) into the ventral torso of the idiosoma of nymphal I . scapularis . The ticks were allowed to rest for 4∼6 hrs before feeding on mice . DNA was extracted from mouse tissues and ticks using a DNeasy tissue kit ( QIAGEN , Valencia , CA ) according to the manufacturer's protocol . The nymphal ticks ( unfed or fed for 24 , 48 , and 72 h ) were dissected under the microscope to get the tick salivary gland and midgut . Total RNA was extracted using RNeasy mini spin columns ( QIAGEN ) . RNA was converted into first-strand cDNA using random hexamers and Superscript III reverse transcriptase ( Invitrogen , Carlsbad , CA ) according to the manufacturer's protocol . All quantitative PCR ( Q-PCR ) assays were performed with an iCycler ( Bio-Rad Laboratories , Hercules , CA ) using gene-specific primers , and IQ SYBR green quantitative PCR system ( Bio-Rad ) or a Taqman quantitative PCR system ( Applied Biosystems , CA ) with a program consisting of an initial denaturing step of 3 min at 95°C and 45 amplification cycles consisting of 30 s at 95°C followed by 1 min at 60°C . The gene-specific primers ( and probes , for Taqman Q-PCR ) used for Q-PCR were indicated in Table S2 . The full open reading frame of tHRF was amplified from the tick salivary cDNA library using gene specific primers P15 , 16 ( Table S2 ) . The PCR product was subcloned into the pGEX-6P2 vector ( Invitrogen , CA ) and transfected into E . coli BL21/DE3 strain for protein expression . The recombinant tHRF was purified by GST sephorose 4B and the GST tag was removed by the precision protease on column according to the manufacturer's protocol . To make recombinant protein using the Drosophila S2 cell system , the full open reading frame of tHRF was subcloned into pMT/BiP/V5-His A vector ( Invitrogen ) using primers P17 , 18 ( Table S2 ) and transfected into Drosophila S2 cells ( Invitrogen , CA ) in combination with the hygromycin selection vector pCOHYGRO for stable transfection . The stable transformants were selected using 300 µg/ml hygromycin-B for 3–4 weeks . The recombinant tHRF with 6-His tag were induced and purified with Talon affinity column as described previously [38] . To generate polyclonal antisera , tHRF ( without the GST tag ) produced in E . coli was emulsified in complete Freund's adjuvant and injected into groups of 2–3 rabbits ( 100 µg/animal ) . The animals were boosted twice at 3-week intervals with the same dose of antigen in incomplete Freund's adjuvant , and the sera were collected 2 week after the second boost . A western blot was performed to analyze the protein expression of tHRF in adult tick saliva , nymphal salivary gland extract , midgut extract and whole nymphs . The tick saliva and tissue extract were prepared as described [22] . Protein preparations were separated on a 4–15% gradient poly-acrylaminde gel and transferred on to a PVDF membrane . The membranes were probed with polyclonal anti-tHRF antibody followed by HRP-conjugated anti-Rabbit IgG and detected with enhanced luminol-based detection ( ECL ) kit ( GE bioscience ) . Groups of 5 mice were passively immunized with 200 µl of normal rabbit serum , anti-Salp25D antiserum ( as controls ) or anti-tHRF antiserum , respectively ( Salp25D is a tick salivary protein that does not influence tick feeding [23] ) . 24 h after immunization , 6 B . burgdorferi-infected nymphal ticks or 10 non-infected nymphs were placed on each mouse . After 72h , the ticks were collected and weighed to analyze the feeding efficiency . For the B . burgdorferi-infected tick experiment , the Borrelia burden in ticks as well as in the localized skin specimen at 7 post tick repletion and in the murine heart and joints at 3 weeks post-infection were determined by measuring flaB copies using quantitative PCR . To address the role of tHRF on 72–96 h post tick attachment , all the ticks were allowed to feed to repletion on tHRF antiserum immunized mice or control mice ( normal rabbit serum immunized ) . After 60 h post tick attachment , the mice were examined every 12 h and the number of tick detached from the mice were recorded . The weights of ticks after repletion were measured as described above . In the active immunization study , groups of 5 mice were immunized by subcutaneously injecting 10 µg of purified recombinant tHRF suspended in complete Freund's adjuvant , or adjuvant alone ( mock control ) . Mice were boosted with 5 µg of antigen suspended in incomplete Freund's adjuvant every two weeks . Before tick challenge , mice were bled and the anti-tHRF antibody titer was analyzed by western blot . The tick challenge and pathogen burden analysis were performed using the same methods described above . To test whether tHRF binds to mammalian basophils , an in vitro binding assay was performed as described previously [24] . Briefly , a rat basophilic leukemia cell line RBL-2H3 was purchased from American type culture collection ( ATCC , Manassas , VA ) . Cells were cultured to confluence in a 6-well plate and then incubated with recombinant tHRF ( generated from E . coli ) or GST in 1% FCS or buffer alone at 4°C for 2hrs . After 3 washes with PBST ( PBS+ 0 . 1% Tween 20 ) , the cells were incubated with purified Alexa 488 labeled anti-tHRF IgG or Alexa 488-anti-GST IgG in 1% FCS plus 1% rat isotype IgG buffer . The IgG labeling was performed using the Alex488 easy labeling kit ( Invitrogen , CA ) according to manufacture's direction . After 3 washes with PBST , the cells were fixed with 4% PFA and permeablized with 1% Triton X-100 , and the nuclei were stained with TOPRO3 . The cells were then analyzed by microscopy and Flow Cytometry . To investigate whether tHRF can induce histamine secretion from basophils , a histamine release assay was performed using the method described [22] . The possible endotoxins from all test protein preparations were eliminated by passing the protein preparations through endotoxin-free columns ( PIERCE , Rockford , IL ) . Varying concentrations of 5 , 0 . 5 and 0 . 1 µg ml−1 of endotoxin-free HRF ( DES ) or GST ( as negative control ) were added to confluent RBL-2H3 cells ( in 2 ml media ) and incubated at 37°C for 30 min . To determine whether , native tHRF in tick tissue extracts could also induce histamine release , 1 . 0 µg ml−1 nymphal tick salivary gland extracts were assayed for histamine release . Substance C48/80 ( Sigma , St Louis , MO ) , a calcium ionophore was used at 0 . 5 µg ml−1 for positive control . A histamine ELISA kit purchased from Research diagnostic Inc . ( Flanders , NJ ) was used to determine histamine concentrations in culture supernatants . To investigate the role of histamine and tHRF on late stage of tick feeding , B . burgdorferi infected nymphal ticks were fed on naïve mice for 60h . Then 10mM of histamine or 10 µg of recombinant tHRF were injected into the mouse skin at the tick bite site ( usually around the ear ) . Control mice were given the same amount of PBS or recombinant GST . At 72h post-tick attachment , the tick weights were measured and B . burgdorferi burden in ticks were analyzed by Q-PCR . Results are expressed as the mean ± the SEM . The significance of the difference between the mean values of the groups was evaluated by Student's t test with StatView software ( SAS Institute ) . The GenBank accession numbers for the genes related with this study: tHRF/DQ066335; SSRB/DQ066202; Salp25D/AF209911; HBP1/DQ066014; HBP2/DQ066128; HBP3/DQ066002 . | Ticks are distributed worldwide and affect human and animal health by transmitting diverse infectious agents . Safe and effective vaccines against most tick-borne pathogens are not currently available . Typical vaccines target microbes directly , using extracts of the organism , or recombinant antigens as the immunogen; the transmission of tick-borne pathogens can also theoretically be prevented by interfering with the ability of ticks to feed on a mammalian host . In this study , we have characterized a putative histamine release factor ( tHRF ) from I . scapularis ticks , the predominant vector of B . burgdorferi , the agent of Lyme disease in North America . Our results suggested that tHRF is presented in tick saliva and critical for tick feeding; blocking tHRF markedly reduced the efficiency of tick feeding , and reduced the B . burgdorferi burden in mice . This finding provides novel insights into the molecular mechanisms of tick feeding and provides a potential vaccine target to block tick feeding and pathogen transmission . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/bacterial",
"infections",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis",
"immunology/immunity",
"to",
"infections",
"microbiology/immunity",
"to",
"infections"
] | 2010 | Tick Histamine Release Factor Is Critical for Ixodes scapularis Engorgement and Transmission of the Lyme Disease Agent |
Currently , there are three algorithms for screening of syphilis: traditional algorithm , reverse algorithm and European Centre for Disease Prevention and Control ( ECDC ) algorithm . To date , there is not a generally recognized diagnostic algorithm . When syphilis meets HIV , the situation is even more complex . To evaluate their screening performance and impact on the seroprevalence of syphilis in HIV-infected individuals , we conducted a cross-sectional study included 865 serum samples from HIV-infected patients in a tertiary hospital . Every sample ( one per patient ) was tested with toluidine red unheated serum test ( TRUST ) , T . pallidum particle agglutination assay ( TPPA ) , and Treponema pallidum enzyme immunoassay ( TP-EIA ) according to the manufacturer’s instructions . The results of syphilis serological testing were interpreted following different algorithms respectively . We directly compared the traditional syphilis screening algorithm with the reverse syphilis screening algorithm in this unique population . The reverse algorithm achieved remarkable higher seroprevalence of syphilis than the traditional algorithm ( 24 . 9% vs . 14 . 2% , p < 0 . 0001 ) . Compared to the reverse algorithm , the traditional algorithm also had a missed serodiagnosis rate of 42 . 8% . The total percentages of agreement and corresponding kappa values of tradition and ECDC algorithm compared with those of reverse algorithm were as follows: 89 . 4% , 0 . 668; 99 . 8% , 0 . 994 . There was a very good strength of agreement between the reverse and the ECDC algorithm . Our results supported the reverse ( or ECDC ) algorithm in screening of syphilis in HIV-infected populations . In addition , our study demonstrated that screening of HIV-populations using different algorithms may result in a statistically different seroprevalence of syphilis .
Syphilis is an ancient human disease caused by Treponema pallidum , which is mostly transmitted by sex activity . It remains a worldwide public health concern as there has been a global increase in the incidence of syphilis , especially among men who have sex with men ( MSM ) . MSMs are a unique population that experience disproportionately high rates of HIV infection . Although clinical profiling of symptoms is important , serologic tests are still considered the mainstay of syphilis diagnosis . Serological tests for syphilis can be categorized into two types: the non-treponemal tests ( NTT ) such as rapid plasma reagin ( RPR ) , toluidine red unheated serum test ( TRUST ) , and Venereal Disease Research Laboratory ( VDRL ) tests . Other treponemal tests ( TT ) include the T . pallidum particle agglutination assay ( TPPA ) , T . pallidum hemagglutination assay ( TPHA ) , treponemal ELISA , and chemiluminescence methodologies[1] . Currently , there are three algorithms for screening of syphilis . First , the traditional screening algorithm commences with a non-treponemal assay followed by a confirmation with a treponemal test . Second , the reverse algorithm starts with a treponemal assay , and a reactive treponemal screening assay is followed by a quantitative non-treponemal assay . Third , the European Centre for Disease Prevention and Control ( ECDC ) algorithm-a modified reverse algorithm: a reactive treponemal screening test is followed by a second ( and different ) treponemal test but is not accompanied by a non-treponemal test[2] . All testing algorithms possess certain advantages and limitations . Consequently , there is no generally recognized diagnostic algorithm[3] . For those infected with both syphilis and HIV , the situation is even more complex[4] . For example , unusual serologic responses such as the prozone and sreofast phenomenon have been observed in HIV-infected individuals[5] . To the best of our knowledge , no studies have analyzed the different algorithms for detecting syphilis in HIV-positive people . Therefore , this study aimed to compare the results of three syphilis screening algorithms in an attempt to evaluate their screening performance in this unique population . Moreover , we examined whether the different screening algorithms significantly influenced the seroprevalence of syphilis in HIV-positive patients .
We conducted a cross-sectional study to assess the impact of different syphilis screening algorithms in a HIV-positive population . Sample size was estimated to be 677 using N=Zα2P ( 1−P ) ⁄δ2 , assuming 19 . 8% syphilis prevalence in HIV infected patients[6] , with 3% precision and 95% level of confidence . We collected a convenience sample of discarded serum specimens from HIV patients undergoing serologic evaluation for HIV virus load in The First Affiliated Hospital , Medical College of Zhejiang University . The patients’ HIV infection status was confirmed by the detection of HIV antibodies in blood using enzyme-linked immunosorbent assay ( ELISA ) and western blot analysis . The following data abstracted from the hospital electronic medical record: age , sex , racial and ethnic identity , the route of HIV transmission , and the stage of AIDS ( the name were anonymized in the supporting information ) . This study was approved by the Institutional Ethics Committee of The First Affiliated Hospital , Medical College of Zhejiang University and complied with the Declaration of Helsinki guidelines . TRUST ( Rongsheng Biotech Co . , Ltd , Shanghai , China ) was used as the non-treponemal test , and TPPA ( Fujirebio INC , Tokyo , Japan ) and TP-EIA ( Wantai Biological Pharmacy Enterprise Co . , Ltd , Beijing , China ) were used as the treponemal tests . Every sample ( one per patient ) was tested by TRUST , TPPA , and TP-EIA simultaneously . All testing was performed according to the manufacturer’s instruction . The performing assay technician was unaware of the results of other testing , and all the results were reported independently . The results of syphilis serological testing were interpreted following different algorithms respectively . The definition of serological diagnosis of syphilis under different algorithms is illustrated in Fig 1 . In the traditional algorithm , samples were screened by TRUST test , and the positive samples would be checked by TPPA test . If the TPPA test also gave a positive result , the sample will be considered as positive for syphilis by serodiagnosis . In the reverse algorithm , samples were screened by TP-EIA test , and the positive samples would be referred to the results of TRUST test . If the TRUST test is positive , the sample is thought to be infected by syphilis . When an inconsistent result was got , the sample would be judged by TPPA test in addition . In the ECDC algorithm , samples were screened by TP-EIA test , and the reactive samples were confirmed by TPPA test . According to the results of TPPA assay , the positive percent agreement , negative percent agreement and total percent agreement , each with 95% confidence interval ( CI ) , of the TP-EIA and TRUST assays were calculated by standard 2 x 2 contingency tables . In addition to percent agreement , kappa coefficients were calculated as a secondary measure of agreement . The seroprevalence of syphilis using traditional and reverse algorithms were compared using McNemar's test for paired proportions . Statistical analysis was performed using SPSS , version 20 ( version 20; IBM Corp . , Armonk , NY , USA ) .
As shown in Table 1 , the 865 HIV infected individuals had a mean age of 40 . 7 ( range 17–81 ) years , and the male accounted for 82% . The majority of them ( 87 . 1% ) were of Han ethnicity . More than half ( 58 . 7% ) of the HIV infected individuals were transmitted by heterosexual . Among the 865 patients , 382 ( 37 . 9% ) patients were in AIDS stage and 1 patient’s stage of HIV infection was unavailable . The serological test results of syphilis are illustrated in Fig 2 . Overall , 123 subjects had TP-EIA +/TPPA+/TRUST+ results , and 602 subjects had TP-EIA −/TPPA−/TRUST− results . 90 patients were TP-EIA+/TPPA+/TRUST− . In order to exclude the prozone phenomenon , the TRUST tests for the 90 TP-EIA+/TPPA+/TRUST− subjects were repeated with serum samples diluted from 1:1 to 1:32 , and no subjects were found activate with TRUST after dilution . The total percentages of agreement and corresponding kappa values of each assay’s results compared with those of TPPA were as follows: for TP-EIA , 98 . 4% , 0 . 960; for TRUST , 85 . 2% , 0 . 566 . These data indicated that there was a very good strength of agreement between the TPPA test and the TP-EIA . Using the TPPA test as the standard test , the TP- EIA had 100% positive percent agreement and 97 . 9% negative percent agreement ( Table 2 ) . When the data was analyzed in the AIDS group and non-AIDS group , the results were similar to the total HIV infected individuals ( S1 Table ) . In the traditional algorithm ( Fig 1 ) , 161 ( 18 . 6% ) samples were reactive with TRUST . Of 161 TRUST positive samples , 123 ( 76 . 4% ) were confirmed as positive by TPPA and were suggestive of syphilis . 38 ( 23 . 6% ) were considered to be false-positive by the TRUST . The rate of serodiagnosis of syphilis was 14 . 2% ( 95% confidence interval [CI] , 11 . 9%– 16 . 6% ) using the traditional algorithm . In the reverse algorithm ( Fig 1 ) , 227 ( 26 . 2% ) samples tested positive with TP-EIA . 125 ( 55 . 1% ) of the 227 TP-EIA positive samples were TRUST-positive . Discordant samples ( n = 102 ) were tested with TPPA and 90 ( 88 . 2% ) tested positive . 12 ( 11 . 8% ) samples had negative TPPA results . The rate of serodiagnosis of syphilis was 24 . 9% ( 95% CI , 22 . 0%– 27 . 7% ) using the reverse algorithm . In the ECDC algorithm ( Fig 1 ) , 213 ( 93 . 8% ) of the 227 TP-EIA positive samples were confirmed by TPPA . The rate of serodiagnosis of syphilis was 24 . 6% ( 95% CI , 21 . 7%– 27 . 5% ) using the ECDC algorithm . Of the 213 samples diagnosed by ECDC algorithm , 123 ( 57 . 7% ) samples were active with TRUST . The reverse algorithm demonstrated significantly higher seroprevalence of syphilis than the traditional algorithm ( 24 . 9% vs . 14 . 2% , p < 0 . 001 ) in the 865 HIV infected patients . The 123 patients diagnosed by the traditional algorithm were also confirmed by the reverse screening algorithm , while the reverse screening algorithm detected an additional 92 patients that could not be detected using the traditional algorithm . Compared to the reverse algorithm , the traditional algorithm also had a missed serodiagnosis rate of 42 . 8% . The situation is similarly when compared the traditional algorithm with the ECDC algorithm ( with a missed serodiagnosis rate of 42 . 3% ) . Among the 92 patients , 90 patients were TP-EIA+/TPPA+/TRUST− and 2 patients were TP-EIA+/TPPA−/TRUST+ . The seroprevalence of syphilis screened by traditional algorithm , reverse algorithm and ECDC algorithm in AIDS stage and non-AIDS stage group was 13 . 1% vs . 14 . 9% ( p = 0 . 46 ) , 26 . 2% vs . 24 . 1% ( p = 0 . 48 ) , and 25 . 9% vs . 23 . 9% ( p = 0 . 50 ) , respectively . Both in AIDS stage and non-AIDS stage group , the traditional algorithm showed significantly lower seroprevalence of syphilis than the reverse algorithm and ECDC algorithm ( Fig 3 ) . The total percentages of agreement and corresponding kappa values of each algorithm’s results compared with those of reverse algorithm were as follows: for tradition algorithm , 89 . 4% , 0 . 668; for ECDC algirithm , 99 . 8% , 0 . 994 . Using the reverse algorithm as the standard test , the tradition algorithm had 57 . 2% positive percent agreement and 100% negative percent agreement ( Table 3 ) . Compared the traditional algorithm with the reverse algorithm , the positive percent agreement between the non-AIDS group and AIDS group had no statistically significant difference ( 62% vs . 50% , p = 0 . 08 , S2 Table ) .
Serological testing of syphilis remains an important component in the diagnosis of syphilis . Latent syphilis , which is without clinical symptoms , is mainly detected by the non-treponemal and treponemal serologic tests . Treponemal tests become positive in the 2–4 weeks after infection , and it can be detected after successful treatment , even persist lifetime . Non- treponemal tests become positive about 2 weeks later than Treponemal tests . Titers of non-treponemal tests are generally related to disease activity , and it can be declined to negative after successful therapy ( except for serofast phenomenon ) . Non-treponemal tests are mainly used to monitor disease activity and assess the response to treatment . Non-treponemal tests are not sensitive for latent , primary , tertiary syphilis and neurosyphilis , as well as successful treated syphilis . Our study showed a very good strength of agreement between TP-EIA and TPPA , while the TRUST only have a 57 . 7% positive percent agreement with TPPA ( κ = 0 . 566 ) in HIV positive patients . These results indicated the insensitive situations of TRUST in HIV infected individuals are common , especially in the AIDS group . Treponemal tests first or non-treponemal tests first ? Does the Order Matter ? That is the key difference between the tradition algorithm and the reverse algorithm . Nowadays , there is no uniform screening method for syphilis . Public health decisions on which algorithm should be employed depending on many factors , including disease prevalence , cost , ease of use , and suitability for automation . It is important to consider the screening abilities of different algorithms in the same population . Matthew[7]directly compared the traditional and reverse syphilis screening algorithms in a population with a low prevalence of syphilis . Their results showed that among 1000 patients tested , 6 patients were falsely reactive by reverse screening , compared to none by traditional testing . However , reverse screening identified 2 patients with possible latent syphilis that were missed by traditional testing . In HIV-positive individuals , the situation is more complex . This is the first direct comparison of the reverse and traditional syphilis screening algorithms in a HIV-infected population . Our present study found that among 865 patients tested , the reverse screening algorithm diagnosed an additional 92 patients that could not be observed using the traditional algorithm . The missed diagnosis rate of the traditional screening algorithm was 42 . 8% compared with the reverse screening algorithm , which is higher than the study by Tong [8] . That was a large survey conducted in an area with a high prevalence of syphilis ( 11 . 4% ) . Previous studies[9] have suggested that reverse screening can yield a high false-positive rate , while many early studies lacked parallel traditional screening on the same samples . Our study found the false-positive rate of reverse screening was lower than traditional screening ( 1 . 4% vs . 4 . 4% ) and our finding were consistent with Tong’s findings . The prevalence of syphilis of the participants may contribute to the difference . We and Tong’s study were carried out in a population with a high prevalence of syphilis . Our study showed there was a very good strength of agreement between the reverse and ECDC algorithm , and demonstrated that the seroprevalence of syphilis using the reverse algorithm ( or the ECDC algorithm ) was significantly higher than the traditional algorithm for HIV-positive individuals . The 92 patients missed by traditional algorithm contribute to this difference , of which , 90 TRUST−/TP-EIA+/TPPA+ patients were the majority . Patients with discordant TRUST and TP-EIA serological results are confirmed by TPPA . If TPPA is non-reactive , it is considered to be false-positive . When TPPA is reactive , there are 3 interpretations ( i ) successfully treated syphilis infection; ( ii ) early/late or latent syphilis , when the sensitivity of TRUST is low; ( iii ) the prozone phenomenon , especially in secondary syphilis . The prozone phenomenon in syphilis testing refers to a false-negative response resulting from an excess of antibody , which prevents visible agglutination in agglutination or precipitation tests . Beyond our expectation , no prozone phenomenon were found among the 90 TRUST−/TP-EIA+/TPPA+ patients in the present study , which is lower than Jeffrey’s [10]study ( 0 . 90% , 2/223 ) . May be it is due to the small sample size , and it needs to evaluate the rate of prozone phenomenon in HIV infected individuals in a larger sample size . There are several limitations to our study and the results should be interpreted with caution . First , all specimens were obtained from hospital patients and there is consequent sample selection bias . Second , the study was conducted from the perspective of serological diagnosis , and both the clinical diagnosis and prior history of syphilis were not analyzed . In conclusion , screening of HIV-populations using different algorithms may result in a statistically different seroprevalence of syphilis . When comparing the prevalence of syphilis in HIV-infected individuals from different surveys , it is important to assess which screening method is employed . Finally , we advocate the reverse algorithm ( or the ECDC algorithm ) approach for the screening of syphilis in HIV-infected populations , given its sensitivity for early/late and latent syphilis . The quantitative non-treponemal tests were recommended to determine serological activity of syphilis in ECDC algorithm . The tradition algorithm approach underestimates the prevalence of syphilis in HIV-infected individuals . | Syphilis remains a worldwide public health concern as there has been a global increase in the incidence of syphilis . Serologic tests are still considered the mainstay of syphilis diagnosis . Currently , there are three algorithms for screening of syphilis- traditional algorithm , reverse algorithm and European Centre for Disease Prevention and Control ( ECDC ) algorithm . But there is no uniform screening method for syphilis . Different surveys use different screening algorithm . Will the different screening algorithm influence the seroprevalence of serodiagnosis of syphilis in this unique population ? For those infected with both syphilis and HIV , the situation is even more complex . To the best of our knowledge , no studies have analyzed the different algorithms for detecting syphilis in HIV-positive people . Therefore , we compared the results of the three syphilis screening algorithms in an attempt to evaluate their screening performance in this unique population . Our results supported the reverse ( or ECDC ) algorithm in screening of syphilis in HIV-infected populations . In addition , our study demonstrated that the tradition algorithm approach underestimates the prevalence of serodiagnosis of syphilis in HIV-infected individuals . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"urology",
"hiv",
"infections",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"applied",
"mathematics",
"tropical",
"diseases",
"microbiology",
"simulation",
"and",
"modeling",
"algorithms",
"treponematoses",
"bacteria... | 2017 | The tradition algorithm approach underestimates the prevalence of serodiagnosis of syphilis in HIV-infected individuals |
Involuntary force variability below 15 Hz arises from , and is influenced by , many factors including descending neural drive , proprioceptive feedback , and mechanical properties of muscles and tendons . However , their potential interactions that give rise to the well-structured spectrum of involuntary force variability are not well understood due to a lack of experimental techniques . Here , we investigated the generation , modulation , and interactions among different sources of force variability using a physiologically-grounded closed-loop simulation of an afferented muscle model . The closed-loop simulation included a musculotendon model , muscle spindle , Golgi tendon organ ( GTO ) , and a tracking controller which enabled target-guided force tracking . We demonstrate that closed-loop control of an afferented musculotendon suffices to replicate and explain surprisingly many cardinal features of involuntary force variability . Specifically , we present 1 ) a potential origin of low-frequency force variability associated with co-modulation of motor unit firing rates ( i . e . , ‘common drive’ ) , 2 ) an in-depth characterization of how proprioceptive feedback pathways suffice to generate 5-12 Hz physiological tremor , and 3 ) evidence that modulation of those feedback pathways ( i . e . , presynaptic inhibition of Ia and Ib afferents , and spindle sensitivity via fusimotor drive ) influence the full spectrum of force variability . These results highlight the previously underestimated importance of closed-loop neuromechanical interactions in explaining involuntary force variability during voluntary ‘isometric’ force control . Furthermore , these results provide the basis for a unifying theory that relates spinal circuitry to various manifestations of altered involuntary force variability in fatigue , aging and neurological disease .
Involuntary fluctuations in muscle force are inherent to human motor control . Evidence suggests that this apparent ‘noise’ is functionally significant for movement execution and learning [1–5] . Furthermore , amplification of force variability or distortion of its frequency content is an almost universal phenomenon whenever neuromuscular control is altered , for example by aging [1 , 6] , fatigue [7 , 8] , and neurological diseases [9–13] . However , whether such phenomenon is caused by common or distinct factors is not known because the sources of involuntary force variability and their potential interactions are not well understood . By some descriptions , involuntary force variability is a manifestation of broad-band neural noise [3–5] . However , neural drive to muscles is known to have a highly structured frequency spectrum [14] . Accordingly , different neural sources of involuntary force variability , such as descending drive [15–18] and proprioceptive feedback [19–22] , are often described specifically in terms of their frequency content . Frequency-specific force variability can also stem from mechanical sources ( e . g . mechanical resonance ) , even if the neural drive itself contains no distinct oscillatory components [23 , 24] . Attempts to understand the relative contribution that each ‘source’ of involuntary force variability makes to the total have been difficult , given that they all act concurrently during muscle activation , and are difficult to experimentally isolate and manipulate . While different sources of involuntary force variability may be distinct , they are not likely to be independent . For example , there is recent evidence suggesting an inverse relationship between low ( 1-5 Hz ) - and high-frequency ( 5-12 Hz ) neural drive to muscles [25–27] . The high-frequency drive may originate from stretch-reflex circuitry [19 , 22] . The low-frequency drive ( the so-called ‘common drive’ ) does not have a known origin , but appears to be negatively influenced by Ia afferent feedback , since it is strongest in muscles which have low spindle densities [25] . Further , experimental conditions which increase high-frequency neural drive and H-reflex amplitudes also decrease low frequency neural drive [26 , 27] . Together , the clear implication is that Ia afferent feedback oppositely affects high and low frequency neural drive ( and thus force variability ) , but the mechanistic details are not yet understood . In this study , we establish how neural and mechanical sources of force variability interact to produce the structured force spectrum observed experimentally using a physiologically-grounded model of afferented muscle . Our simulation of an afferented musculotendon set inside of a closed-loop control scheme allowed us probe the mechanistic interactions that exist among an error correction mechanism for muscle force , proprioceptive feedback , and mechanical properties of muscle . Further , we describe these interactions in terms of their effects on involuntary force variability and on the behavior of a simulated pool of motor units . Our hypotheses were 1 ) neuromechanical interactions inherent to the closed-loop control of viscoelastic musculotendon would suffice to produce low-frequency force variability , 2 ) tuning of proprioceptive feedback ( i . e . , known modulation of fusimotor drive or presynaptic gains ) would impact the entire frequency spectrum of force variability , and 3 ) those changes in force variability would be reflected in motor unit synchronization . Our findings not only support these predictions , but ( i ) emphasize the importance of neuromechanical interactions to levels not previously recognized , and ( ii ) they describe how isolated changes in each proprioceptive pathway gain influences the full spectrum of involuntary force variability . This novel demonstration fills a critical gap in our understanding of how error correction mechanisms , proprioceptive feedback , noise , and musculotendon mechanics are interrelated , and our results emphasize the critical importance of investigating involuntary force variability within the context of closed-loop control . Our results are an important step towards a unifying theory that relates spinal circuitry to various manifestations of altered involuntary force variability in functional performance [1] , aging [1 , 6] , fatigue [7 , 8] and neurological disease [9–13] .
First , we investigated the interactions between mechanical properties of the musculotendon and broad-band neural noise using an open-loop input without any feedback ( Simulation 1 . 1 ) . For this simulation , our control input was simply the target trajectory ( i . e . , 1-sec zero input , 2-sec ramp-up and 32-sec hold at 20% MVC ) , with added signal-dependent noise . The coefficient of variation of force was 8 . 73% . This open-loop control resulted in force variability which fell almost entirely below 5 Hz , within the ‘common drive’ range ( red line in Fig 4A ) . It is worth noting that there was no distinct peak within this frequency range ( i . e . , 1-5 Hz ) . Accordingly , the neural drive produced in this simulation also caused a small degree of common drive , as measured by the ‘common drive index’ ( red boxplot in Fig 4B ) . A similar result was observed using motor unit coherence analysis . It is also important to note that high-frequency force variability ( 5-12 Hz ) did not arise from the interaction between mechanical properties of musculotendon and broad-band noise . We then ran the simulation in closed-loop condition using only the error correction mechanism ( i . e . , tracking controller ) ( Simulation 1 . 2 ) . The amplitude of overall force variability was 8 . 39% , which was not significantly different from the open-loop condition ( independent sample t-test using Yuen’s method , p = 0 . 19 ) . This addition of an operational tracking controller resulted in the generation of a peak at ∼ 1 . 8 Hz in the power spectrum of muscle force ( green line in Fig 4A ) . Also , the degree of motor unit synchronization in this range increased accordingly ( green boxplot in Fig 4B ) . These results altogether suggest that low-frequency force variability and common drive are primarily an emergent property of a close-loop control of muscle force . Also , these results show that high-frequency force variability does not emerge in the absence of proprioceptive feedback . Gain control of Ia afferent feedback at the spinal cord , often experimentally quantified by H-reflex amplitude , plays an important role in human motor control and learning to achieve a variety of movements [42 , 43] . Here , we examined how changes in the gain of Ia afferent feedback , modeled as presynaptic control input , influence force variability . We systematically altered the level of this presynaptic control input from the value of -0 . 5 to 0 while keeping the other gain parameters constant ( 70 pps for dynamic and static fusimotor drives and -0 . 3 for presynaptic control level of Ib afferent feedback ) . This range was set such that the mean input contribution of Ia afferent feedback to the neural drive spanned a range from 0 ( i . e . , no contribution from Ia afferent feedback ) to 30% of the maximum neural drive . The amplitude of force variability decreases as the presynaptic input level is increased and becomes minimal at the value of -0 . 15 ( Fig 5A ) . Further increases negatively affect the amplitude of force variability ( the presynaptic control level of -0 . 05 and 0 in Fig 5A ) . Analyses of force variability in the frequency domain show the change in force variability amplitude occurred across the frequency range ( p < 0 . 01 at all the frequencies between 1 and 12 Hz ) , but prominent peaks exist in the two distinct frequency ranges , namely the common drive range ( 1-5 Hz ) and physiological tremor range ( 5-12 Hz ) ( shown as blue and red bands for common drive range and physiological tremor range , respectively , in Fig 5B ) . These observations demonstrate that modulation of the strength of Ia afferent feedback is an important factor that influences overall force variability during ‘isometric’ force production . Further analyses on frequency-specific effects of Ia afferent feedback show increasing the gain of Ia afferent feedback reduces force variability within the common drive range ( Fig 6A ) while it increases the amplitude of physiological tremor ( Fig 6B ) . Excessive Ia gain led to excessive physiological tremor as suggested in previous studies [20 , 44] . As Ia afferent feedback increases , common drive decreases more than physiological tremor increases , after which physiological tremor dominates the spectrum and a monotonic increase in total force variability is observed . These observations suggest that the U shaped response comes from the relative contribution of common drive and physiological tremor to total force variability . Importantly , these concurrent changes in the common drive and physiological tremor are consistent with previous speculations [25–27] . These observations suggest that relatively faster excitation cycles of Ia afferent feedback can function as a negative feedback ( i . e . , withdrawal of Ia afferent input during muscle shortening and its excitation during muscle stretch ) , thereby interrupting the development of low-frequency force fluctuations , characteristic of a close-loop control of muscle force . Changes in motor unit synchronization in the common drive and physiological tremor ranges are shown in ( Fig 6C–6E ) . Stronger Ia afferent feedback reduces the degree of common drive ( Fig 6C and 6D ) . In contrast , it induces a higher degree of synchronization in the physiological tremor range ( Fig 6E ) . These results further confirm a previously suggested relationship between the strength of Ia afferent feedback and motor unit synchronization in the common drive and physiological tremor ranges [25–27] . Understanding the operation of the fusimotor system is hindered by the lack of techniques which can directly measure γ-motoneuron activities [45] . However , experimental evidence based on human group Ia and II afferent activities has suggested that humans have control over the fusimotor system which is independent of α-motoneuron drive , and which can be modulated by attention and task requirements [46–48] . Here , we postulate that fusimotor-induced changes in the dynamic sensitivity and static bias of Ia afferent activity will have profound effects on force variability as well . Therefore , we tested three scenarios; 1 ) co-modulation of γ dynamic and static fusimotor drives , 2 ) modulation of γ dynamic or 3 ) γ static fusimotor drive independently while the other is held constant , as done previously [33] . In this study , we varied them from 10 to 250 pps by increment of 20 pps . When γ dynamic or static fusimotor drive was varied independently , the other was kept at 70 pps . The presynaptic control levels of Ia and Ib afferent feedback were set at -0 . 15 and -0 . 3 . Results show that the amplitude of overall force variability depends on the levels of fusimotor drives ( Fig 8A–8C top figures ) . When both γ dynamic and γ static fusimotor drives are varied , the amplitude of overall force variability shows a similar response to the presynaptic manipulation of Ia afferent feedback ( Fig 8A top figure ) . Also , the changes again occur predominantly in the common drive and physiological tremor ranges ( p < 0 . 01 at all the frequencies between 1 and 12 Hz ) as indicated by prominent peaks in those ranges ( Fig 8A bottom figure ) . Independent modulation of the only γ dynamic fusimotor drive has comparably smaller effects on the amplitude of overall force variability ( Fig 8B top figure ) . On the contrary , modulation of γ static fusimotor drive produces effects similar to co-modulation of both fusimotor drives ( Fig 8C top figure ) . Again , their effects occur in the common drive and physiological tremor ranges ( Fig 8B and 8C bottom figures ) . These results show that the fusimotor system , especially γ static fusimotor drive , has profound effects on force variability in a frequency specific manner similar to presynaptic modulation of Ia afferent gain . This differential sensitivity to γ dynamic and static fusimotor drives might speak to differences in their functional significance during isometric force production . Also , it is important to note that too high levels of γ static fusimotor drives can lead to greater overall force variability accompanied by excessive physiological tremor , which might be similar to effects of fatigue [7 , 49] ( see Discussion ) . Further analyses in the two frequency ranges show greater fusimotor drives are associated with smaller force variability in the common drive range and larger physiological tremor ( Fig 9A and 9B ) . The effects of γ static fusimotor drive are substantially larger than those of dynamic fusimotor drive in both frequency ranges and the combination of those effects is illustrated in the case of co-modulation of γ dynamic and static fusimotor drives . These results are consistent with those from presynaptic Ia afferent feedback gain such that increased bias level ( mean input contribution ) of Ia afferent feedback , rather than the dynamic sensitivity of Ia afferent feedback , plays a more important role in shaping the power spectrum of force variability and generating physiological tremor [50] . Changes in motor unit synchronization correspond well to changes in force variability , as shown in Fig 9C–9E . Greater fusimotor drives result in lower CDI values and low-frequency coherence ( Fig 9C and 9D ) , as well as higher coherence in the physiological tremor range . These results suggest that modulation of γ dynamic and static fusimotor drives can also alter the degree of motor unit synchronization across the force-relevant frequencies . Given that Ib afferent feedback in general provides inhibition of α-motoneurons as a function of force level , one can easily expect that it helps stabilize force fluctuations [45 , 51] . However , exactly how such a feedback system influences either overall amplitude or frequency-specific components of involuntary force variability is unknown . Here , the presynaptic control value of Ib afferent feedback was varied from -0 . 5 to 0 , while the presynaptic control value of Ia afferent feedback was kept at -0 . 3 and dynamic and static fusimotor drives at 70 pps . This range corresponds to a Ib contribution of 0 to 45% of the maximum neural drive , respectively . The upper range of these values would be non-physiological as the Ib input contribution of 45% of the maximal neural drive , for example , means 45% total input is continuously inhibited and it requires other compensatory mechanisms through Ia afferent feedback and a tracking controller to maintain the target force level . Here , we merely try to fully characterize effects of Ib afferent feedback on force variability and thereby highlight differences between Ia and Ib afferent feedback . As expected , greater inhibition of α-motoneurons through Ib afferent feedback reduces the amplitude of overall force variability ( Fig 10A ) . However , excessive Ib gain can also lead to increased force variability at ∼4 Hz ( Fig 10A ) although it requires non-physiologically large Ib input contributions . In the frequency domain , changes in force variability occur across the frequencies ( p < 0 . 01 at all the frequencies between 1 and 12 Hz ) , but mainly in the common drive as indicated by peaks appearing only in that range ( Fig 10B ) . The slightly lower frequencies at which the second peak occurs compared to those of Ia afferent feedback might result from the longer loop delay of Ib afferent feedback ( Fig 10B ) . These results highlight that Ib afferent feedback can regulate force variability much like presynaptic/fusimotor modulation of Ia afferent feedback , but its effects are mostly confined in the common drive range . Increasing the strength of Ib inhibition results in smaller force variability in the common drive range ( Fig 11A ) , but excessive Ib inhibition can lead to excessive force fluctuations in this range as shown in ( Fig 10A ) . Its effects on physiological tremor are considerably smaller than presynaptic/fusimotor modulation of Ia afferent feedback ( Fig 11B ) . These results highlight the differences in cross-frequency interactions between Ia and Ib afferent feedback pathways , which has not been reported previously . As before , the frequency-specific effects of presynaptic Ib modulation on force variability are also reflected in motor unit synchronization ( Fig 11C–11E ) . Higher Ib feedback gain is associated with lower synchronization in the common drive range and higher synchronization in the physiological tremor range . Interestingly , CDI and coherence in the common drive range respond differently to excessive force fluctuations at ∼4 Hz seen with excessive Ib inhibition ( Fig 11C and 11D ) , suggesting that these two measures have differing sensitivity to synchronization at different frequencies within 1-5 Hz .
A series of closed-loop simulations of an afferented muscle show that many cardinal features of involuntary force variability emerge from closed-loop neuromechanical interactions . Our results reveal that closed-loop control of a viscoelastic musculotendon unit , combined with the tuning of proprioceptive feedback gains , naturally generate both low-frequency ( 1-5 Hz ) force variability and high-frequency oscillations analogous to physiological tremor ( 5-12 Hz ) . Moreover , we show that these low- and high-frequency phenomena are in fact mechanistically related to each other—which suggests novel and fruitful directions for future research . This study is , to our knowledge , the first to directly confirm mechanistic links between low- and high-frequency force variability , as was proposed earlier [25–27] . Finally , we also used the emergent time histories of closed-loop net neural drive ( ‘ND’ in Fig 1 ) to drive the model of a motor unit pool . We find that these inputs suffice to produce motor-unit synchronization compatible with experimental findings [25–27] . Involuntary force variability at low frequencies ( 1-5 Hz ) can arise from various sources , including low-frequency variability in the neural drive to muscle ( the so-called ‘common drive’ ) [36] . As such , the amplitude of this common drive is a contributor to error during voluntary control of precision forces [14 , 15 , 36 , 52] . Although common drive has been studied for over 30 years , its origins remain debatable [10] . Our results are significant because they suggest that common drive can emerge due to a combination of factors inherent to any neuromuscular control loop . Foremost among them is the viscoelasticity of the musculotendon , which acts as a mechanical low-pass filter that naturally allows the preferential conversion of low frequencies in the neural drive into muscle force as previously shown in [53–55] . It is this low frequency component ( 1-5 Hz ) of muscle force that would be selectively reinforced by any imperfect physiological error correction mechanism . Thus , our results demonstrate that low-frequency force variability emerges naturally when controlling viscoelastic muscles—and do not require the presence of proprioceptive feedback . This is a novel alternative to other peripheral explanations . For example , Watanabe and Kohn suggested that high-frequency neural drive can be demodulated into lower frequencies [18] , which still remains to be tested . In fact , our results are congruent with previous evidence for peripheral mechanisms , such as the fact that common drive persists even after disruption of the cortico-spinal tract , as in capsular stroke [41] . Another component of force variability is oscillations in the 5-12 Hz range , often called ‘physiological tremor . ’ Physiological tremor may arise from multiple factors [56] . One of the earliest and most well-supported mechanisms is cycles of excitation around the stretch reflex loop [19 , 20 , 22] . The first important implication of our results is that , in contrast to common drive , physiological tremor does require the proprioceptive feedback in order to arise as shown experimentally [19 , 22] and in computational simulations [20] . Thus mechanical resonance of musculotendons as proposed by [23 , 24] , did not suffice . In fact , we could not elicit physiological tremor via interactions between broad-band noise and the mechanical properties of musculotendon using an open-loop input which consisted of the target trajectory and signal-dependent noise . This result is consistent with previous experimental evidence and simulation [19 , 20 , 22] . Moreover , our simulations allowed us to characterize how physiological tremor amplitude is modulated by proprioceptive pathway gains . Those include both presynaptic control levels of inhibition/disinhibition ( ‘PCIa’ & ‘PCIb’ in Fig 1 ) and ‘descending’ γ fusimotor drive to muscle spindles ( ‘fusimotor drive’ in Fig 1 ) . This detailed characterization was not possible in the previous simulation study by Stein and Oguztoreli [20] and added a new insight that physiological tremor amplitude is mostly determined by the bias level ( i . e . , mean input contribution ) of Ia afferent feedback , not dynamic sensitivity of muscle spindle . Importantly , excessive Ia afferent gains could produce excessive oscillations primarily in the physiological tremor range in Figs 5 and 8 , similar to what has been shown previously in animal models [44] . Interestingly , excessive Ib afferent gains could lead to excessive oscillations in the lower frequency range ( 3-5 Hz ) possibly due to the longer delay along this pathway ( Fig 10 ) . These findings are particularly important to design hypotheses about how peripheral mechanisms interact with descending neural drive to produce physiological and other kinds of tremor in healthy and pathological conditions [16 , 17 , 57 , 58] . Although we find that proprioceptive feedback is not strictly necessary to generate common drive , we do find that it can influence its strength . This is compatible with experimental findings [25–27] . Specifically , De Luca and colleagues report a negative correlation between the degree of common drive and muscle spindle density [25] . Further , Laine and colleagues showed that heightening the perception of task-related errors during a force tracking task led to increases in physiological tremor and H-reflex—while common drive decreased [26 , 27] . Their interpretation was that the changes in common drive and physiological tremor both stemmed from the tuning of proprioceptive gains due to alterations in psycho-sensory state [46–48 , 59–61] . These lines of experimental evidence , however , could not test a mechanistic link between common drive and physiological tremor . Here , we show that increasing the strength of proprioceptive feedback ( via ‘PCIa’ and ‘γ static fusimotor drive’ ) increases physiological tremor but concurrently decreases common drive ( Fig 6 ) . Thus , our results demonstrate that peripheral mechanisms suffice to reproduce those experimental findings . This close link between the amplitude of involuntary force variability and proprioceptive pathway gains ( in Figs 5 , 8 and 10 ) may explain many experimental findings . For example , removing proprioceptive feedback leads to greater overall involuntary force variability ( i . e . , smaller values of ‘PCIa’ , ‘PCIb’ and ‘γ static fusimotor drive’ in Figs 5 , 8 and 10 ) . This is similar to what has been seen in patients with deafferentation [12] . Moreover , we show that excessive proprioceptive pathway gains result in greater overall force variability and excessive physiological tremor ( i . e . , larger values of ‘PCIa’ , ‘PCIb’ and ‘γ static fusimotor drive’ in Figs 5 , 8 and 10 ) . Interestingly , fatigue can produce similar effects on force variability and physiological tremor [7 , 8]; however , a precise mechanism for this phenomenon has not been established . The enhancement of physiological tremor in fatigue can be attenuated by blocking Ia afferent feedback [7] . Further , the sensitivity of stretch/tonic vibration reflex responses is enhanced during fatigue [49] . An emerging picture is that Ia afferent feedback gains are increased during fatigue , but it is not clear how this occurs ( i . e . , via presynaptic inhibition or fusimotor modulation ) , and it is not clear why fatigue influences overall force variability rather than just physiological tremor . Biro and colleagues suggested that augmented Ia afferent feedback during fatigue reflects a fusimotor-dependent compensation for reduced descending drive [49] . This suggestion was based on previous findings in cat where 1 ) the activity of fusimotor system is enhance by activation of group III and IV afferents [62 , 63] , which respond to an accumulation of metabolites during fatigue [64 , 65] , 2 ) Ia afferent firing rates increase accordingly during fatigue contractions [62 , 66] , and 3 ) group III and IV afferents , on the contrary , enhance presynaptic inhibition of Ia afferent feedback [67] . Since presynaptic inhibition would reduce Ia afferent feedback gain , only the increased fusimotor activation seems a plausible compensatory mechanism . Thus it is important to mention that , when we tested the effects of increased fusimotor drive in our simulation , the results of γ static fusimotor drive ( ‘γ static fusimotor drive’ in Fig 8 ) accurately predicted changes in force variability , as might occur during fatigue . Our findings therefore may provide a mechanistic link between several complementary lines of investigation related to fatigue . As demonstrated in the cases of deafferentation and fatigue , the close link between our results and experimental findings may represent an important step in developing a unifying theory of human sensorimotor control that further relates spinal circuitry to manifestations of altered involuntary force variability under various neuromuscular conditions such as aging [1 , 68 , 69] , stroke [10] , cerebral palsy [9] , Parkinson’s disease [13] , and essential tremor [70] . For example , we show that increased Ia afferent feedback gains result in increased force variability below 0 . 5 Hz ( i . e . , larger values of ‘PCIa’ and co-modulation and ‘γ static fusimotor drive’ in Figs 5 and 8 ) . This might provide a link between increased force variability below 0 . 5 Hz seen in patients post stroke [10 , 15] and their heightened Ia afferent feedback gains [71] or lower reflex threshold [72] . Thus , a unifying principle emerges . Namely , that the task-specific tuning of proprioceptive pathway gains in spinal circuitry—or its disruption—produces characteristic changes in the spectra of neural drive . Importantly , these can be quantified by measuring force variability . Our results highlight the significance of considering closed-loop control of afferented muscle in the generation and modulation of involuntary force variability in motor control research . Historically , the force fluctuations have been considered as manifestation of ‘neural noise’ that is intrinsic to neural drive [73] . Despite the fact that such noise ( e . g . , signal dependent noise ) is usually not frequency-specific , involuntary force fluctuations tend to be highly structured [14] . Our results now show that neuromechanical interactions impose structure onto noisy neural drive , and thus involuntary force variability and ‘noise’ are not independent , as is often assumed [15] . This idea may be significant in formulation of theoretical frameworks in motor control . For example , the ability of the proprioceptive feedback system to regulate the amplitude of overall involuntary force variability provides a neural mechanism to minimize it , as suggested by some [3 , 74] . It is important to discuss how the limitations of our model do not affect our conclusions . Our afferented muscle model was not intended to represent the full complexity of the spinal cord circuitry . We used a simplified version of a previously described model of a spinal-like regulator [75 , 76] that can replicate experimental behavior . Specifically , we did not include Renshaw inhibitory interneurons , which are known to provide recurrent inhibition of α-motoneurons and inhibition of Ia inhibitory internuerons [77] . However , in our simulation of a single muscle , the role of Renshaw inhibitory interneurons would be restricted to recurrent inhibition and therefore have effects similar to that of Ib inhibitory feedback , which we did include . Secondly , our model did not attempt to replicate the exact biophysical structure of α-motoneurons and sensory afferents . Rather , we used a single-input/single-output structure to describe the population behavior of each system . We believe this simplification is reasonable because 1 ) the population response of an α-motoneuron pool is linear with respect to its common/shared synaptic input , since noise and non-linear properties of individual neurons get canceled out in the overall population behavior [14 , 78] , and 2 ) the common input to an α-motoneuron pool is the ‘effective’ neural drive , that is , the input that is actually translated into muscle force [14] . Therefore , it was appropriate for the contractile element in the afferented muscle to be modeled as a single input-output element . Another outcome of using a lumped parameter model of muscle is that force is not generated by the summation of twitches from progressively recruited motor units . However , neither physiological tremor nor ‘common drive’ is thought to relate directly to this aspect of physiological force generation [79] . It is also worth noting that since we simulated constant-force contractions , the number of units recruited/derecruited during each trial would have been very small and therefore would have only minor influence on the overall amplitude of force variability . Similarly , the population behavior of muscle spindles can be appropriately modeled as a single element , as muscle spindles are in general believed to distribute their synaptic inputs widely across a motor unit pool [80] . While potential non-uniformity of Ia projections has been suggested [81] , this remains to be validated , and confirmed across different muscles . Thirdly , we did not include modulation of α-motoneuron excitability through various neuromodulatory inputs arising from the brainstem , which can influence reflex sensitivity [82] . Such neuromodulatory effects would be widespread and more difficult to interpret , while also greatly increasing the complexity of our analyses . Finally , our simulation was limited to that of a single muscle during isometric contraction , which is a valuable and informative experimental paradigm [1 , 9 , 26–28] . As in those experimental studies , it is difficult to extrapolate our findings to complex actions involving movement and coordination among multiple muscles . Still , we believe that our results help establish a strong basis for future study of peripheral and neuromechanical factors influencing the control of muscle force . Lastly , we demonstrate that the modulation of involuntary force variability via proprioceptive pathway gains gives the nervous system a certain degree of control over involuntary force variability . Properly regulating those gains is important if disruptive tremor is to be avoided [44] . Our ability to understand and modify these relationships will be instrumental to providing insights into the neural mechanisms and circuits associated with functional performance [1] , aging [1 , 6] , fatigue [7 , 8] and neurological disease [9–13] . Finally , our approach of combining experimental observations with a computational simulation should provide a springboard for future investigation of neuromechanical interactions and task-dependent tuning of sensorimotor integration and proprioceptive mechanisms during voluntary actions in healthy development and aging; and disease .
We used a closed-loop simulation of an afferented muscle model , which is an extension of a previously published model [28] , to identify the sources of frequency-specific force variability and to characterize interactions among them . The schematic diagram of this model is provided in Fig 1 . The afferented muscle model is comprised of a musculotendon unit [29–32] , muscle spindle [33] , and Golgi tendon organ ( GTO ) [34] , which is controlled by a tracking controller [28] . The model was implemented in the MATLAB environment ( The MathWorks Inc . , Natick ) . Full details of each model are given in the corresponding references and only brief descriptions are provided here . All model parameters were taken from the corresponding references except for musculotendon architecture as described below . We simulated a force tracking task using a closed-loop simulation of the afferented muscle model . Each trial consisted of 1-s zero input phase , 2-s ramp-up and 32-s hold at 20% MVC . The last 30 s of each trial were used for further analysis . We simulated 20 trials for each condition described below . In the first set of simulations ( Simulation 1 ) , we examined the frequency spectrum of output force variability arising from interactions among musculotendon , broad-band neural noise , and error correction mechanism . First , we simulated the force tracking task without the tracking controller and proprioceptive feedbacks ( i . e . , open-loop control ) to characterize the interaction between mechanical properties of musculotendon and broad-band neural noise ( Simulation 1 . 1 ) . In this open-loop control condition , the neural drive consisted of open-loop input ( i . e . , 1-sec zero input , 2-sec ramp-up and 32-sec hold at 20% MVC ) and signal dependent noise . The amplitudes of the open-loop input and noise were adjusted such that the mean force level was at 20% MVC and coefficient of variation of force equaled to that from the closed-loop condition described below . We made these adjustments so that 1 ) the same force/input level is used in both conditions , thus facilitating comparisons of force variability and motor unit synchronization , and 2 ) because normalizing the total force variability across conditions makes comparison of their spectral characteristics more straightforward . Then , we replaced the open-loop input with the tracking controller ( i . e . , closed-loop control ) to investigate how an error correction mechanism interacts with force variability arising from mechanical properties of musculotendon ( Simulation 1 . 2 ) . In this closed-loop control condition , proprioceptive feedback was removed by setting presynaptic control inputs to the value of -0 . 5 for each pathway . In Simulation 2-4 , we examined effects of proprioceptive feedback on force variability . To do so , we ran a set of simulations varying one of the proprioceptive pathway gains ( i . e . , presynaptic control level of Ia afferent feedback ( Simulation 2 ) , dynamic and static fusimotor drives ( Simulation 3 ) , and presynaptic control level of Ib afferent feedback ( Simulation 4 ) ) from its minimum to maximum values , while keeping the other gains constant . The minimum value corresponded to the value at which the contribution of that particular proprioceptive feedback is completely removed . The maximal value was determined empirically by the presence of non-physiological high-frequency force variability . The minimum and maximum values of fusimotor drive were set at 10 and 250 pulse per second ( pps ) , respectively . At each parameter value , we run 20 trials . Following Simulation 2 , we ran two additional sets of simulations to investigate the mechanism through which increases in Ia afferent feedback gain lead to reductions in low-frequency force variability . In the first set of simulations ( Simulation 2 . 1 ) , we tested whether reductions in the relative strength of the tracking controller in response to increased excitatory input from Ia afferent feedback could explain that observation . To do so , we ran a set of simulations ( 20 trials ) where the presynaptic control level of Ia afferent feedback was set at -0 . 5 and we added a constant excitatory input whose amplitude corresponded to the average input contribution from Ia afferent feedback at its presynaptic control level of -0 . 1 . This presynaptic control value was chosen because we observed the smallest amplitude of low-frequency force variability . All the other gain parameters were held at 70 pps for dynamic and static fusimotor drives and -0 . 3 for presynaptic control level of Ib afferent feedback . In the second set of simulations ( Simulation 2 . 2 ) , we quantified the frequency response of the closed-loop afferented muscle to investigate how addition of Ia afferent feedback changes the dynamics of this closed-loop system . To obtain the frequency response , we ran two sets of simulations using presynaptic control levels of Ia afferent of -0 . 5 and -0 . 1 , while presynaptic control level of Ib afferent , dynamic and static fusimotor drives were kept constant at -0 . 3 , 70 , and 70 , respectively . In these simulations , we removed the signal dependent noise and injected a set of sinusoids ( 0 . 5 to 15 Hz in steps of 0 . 5 Hz ) with amplitude of 1% of the maximum neural drive 5 sec after the initiation of simulations . We quantified gain and phase of output force in response to an input sinusoid . The gain was computed as the ratio of the amplitude of output force to the amplitude of the input sinusoid . Phase was calculated as a phase difference between mean phase of the output force and input sinusoid during each trial . We performed secondary simulations for all conditions tested ( Simulations 1-4 ) to investigate whether or not changes in the involuntary force variability produced in the closed-loop simulations can be detected as synchronization of motor unit activities within a pool . To do so , we used our simulated neural drive as a common input to a simulated motor unit pool obtained from [39] . While this model is simplistic by comparison with more biophysically-nuanced compartment-based models [34 , 92 , 93] , it is nonetheless entirely sufficient to describe the basic phenomenon of motor unit entrainment by common input [94 , 95] for the following reason . The ‘effective’ ( i . e . , force-generating ) neural drive is common to all motor units within a pool [14] . Thus , the population activity is a linear transformation of the common input , even though individual motor unit responses to that input are nonlinear [96] . As a result , specific non-linearities present in each motor neuron response such as plateau potentials , adaptation , resonance , accommodation , etc . , which can be modeled by those more complex models , are not an important consideration for our present application . This motor unit model describes the orderly recruitment of motor units and rate coding in response to a common excitatory input . The motor unit pool consisted of 120 motoneurons , whose recruitment threshold to excitatory input had exponential distribution with a greater number of motoneurons with low thresholds and a small fraction of motoneurons with high thresholds as described previously [39] . The range of recruitment thresholds was set such that the highest value was 30-fold that of the lowest . Firing rate of a motoneurons was linearly scaled to excitatory input with a constant minimum firing rate of 8 imp/s . All of these properties were same as those in the original study [39] . We added discharge rate variability ( 5% CoV ) to spike trains of individual motor units , indicated as IN in Fig 3 , to simulate the effects of independent noise . In this simulation , we also included a previously modeled intrinsic property of motor unit known as persistent inward current [97] . | Involuntary fluctuations in muscle force are an unavoidable consequence of human motor control and underlie movement execution errors . Amplification and distortion of involuntary force variability are common phenomena found in various neurological conditions and in fatigue . However , the underlying mechanisms for this are often unclear . We investigated the generation and modulation of involuntary force variability arising from different sources , as well as their interactions . We used a closed-loop simulation which included a physiologically-grounded model of an afferented musculotendon and an error-controller . We show that interactions among neural noise , musculotendon mechanics , proprioceptive feedback , and error correction are critical components of force control , and by taking these into account , our model was able to both replicate and explain many cardinal features of involuntary force variability previously reported experimentally . Also , our results suggest previously unrecognized pathways through which force variability may be altered in fatigue and in certain neurological diseases . Finally , we emphasize the potential for important clinical and scientific information to be extracted from relatively simple , non-invasive measurements of force . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"classical",
"mechanics",
"myoclonus",
"pathology",
"and",
"laboratory",
"medicine",
"fatigue",
"engineering",
"and",
"technology",
"signal",
"processing",
"biomechanics",
"noise",
"reduction",
"simulation",
"and",
"modeling",
"s... | 2018 | Cardinal features of involuntary force variability can arise from the closed-loop control of viscoelastic afferented muscles |
The cytoplasmic TRIM5α proteins of certain mammalian lineages efficiently recognize the incoming capsids of particular retroviruses and potently restrict infection in a species-specific manner . Successful retroviruses have evolved capsids that are less efficiently recognized by the TRIM5α proteins of the natural hosts . To address whether TRIM5α contributes to the outcome of retroviral infection in a susceptible host species , we investigated the impact of TRIM5 polymorphisms in rhesus monkeys on the course of a simian immunodeficiency virus ( SIV ) infection . Full-length TRIM5α cDNAs were derived from each of 79 outbred monkeys and sequenced . Associations were explored between the expression of particular TRIM5 alleles and both the permissiveness of cells to SIV infection in vitro and clinical sequelae of SIV infection in vivo . Natural variation in the TRIM5α B30 . 2 ( SPRY ) domain influenced the efficiency of SIVmac capsid binding and the in vitro susceptibility of cells from the monkeys to SIVmac infection . We also show the importance in vivo of the interaction of SIVmac with different allelic forms of TRIM5 , demonstrating that particular alleles are associated with as much as 1 . 3 median log difference in set-point viral loads in SIVmac-infected rhesus monkeys . Moreover , these allelic forms of TRIM5 were associated with the extent of loss of central memory ( CM ) CD4+ T cells and the rate of progression to AIDS in the infected monkeys . These findings demonstrate a central role for TRIM5α in limiting the replication of an immunodeficiency virus infection in a primate host .
The cytoplasmic tripartite motif protein 5α ( TRIM5α ) has been shown to restrict the replication of a broad range of retroviruses in a species-specific manner [1]–[5] . For example , TRIM5α of rhesus monkeys mediates an early , post-entry block of HIV-1 replication but only a modest block of SIV replication in vitro [4] , whereas human TRIM5α restricts N-MLV potently but HIV-1 only weakly [1]–[3] , [5] . Although primate TRIM5α variants share a similar domain organization , TRIM5α sequences are highly polymorphic in different species; this interspecies sequence variation of TRIM5α is associated with differences in the viral specificity of TRIM5α-mediated restriction [3] , [5] , [6] , [7] . Multiple TRIM5 alleles have been recently identified in Old World primates , including rhesus monkeys and sooty mangabeys [8] , [9] . While a few of these polymorphic forms of TRIM5α have an effect on the susceptibility of transduced cells to in vitro retrovirus infection ( HIV-1 and N-MLV ) , no significant association has been observed between any of these TRIM5 alleles and either the control of SIV replication in vivo or AIDS pathogenesis . Although they are resistant to infection with HIV-1 , rhesus monkeys support the replication of certain strains of SIV and develop an AIDS-like disease following infection with these isolates . The SIV-infected rhesus monkey has become an invaluable model for studying AIDS pathogenesis and evaluating AIDS vaccine strategies [10]–[12] . Interestingly , high levels of viral replication are achieved following infection with SIV in some rhesus monkeys , whereas other rhesus monkeys appear to be more resistant to infection . Certain MHC class I alleles have been associated with efficient control of SIV replication in vivo on the basis of particularly effective SIV-specific cytotoxic T lymphocyte responses [13] . Factors other than adaptive immune responses also have been implicated in the control of SIV replication , since the relative permissivity of a monkey's peripheral blood mononuclear cells ( PBMCs ) for SIV replication in vitro predicts the level of SIV replication that occurs in vivo when that monkey is infected with SIV [14] . The present study was initiated to explore the contribution of intraspecies TRIM5 allelic polymorphisms to the control of SIV replication in Indian-origin rhesus monkeys .
To characterize the TRIM5α variants of rhesus monkeys , full-length TRIM5 cDNAs were derived and sequenced from 79 unrelated Indian-origin rhesus monkeys . Six to 15 independent cDNA clones from each monkey were sequenced . Genomic DNAs from these same animals were also sequenced to confirm the identity of these alleles . We identified a 2-amino acid deletion and 14 single nucleotide polymorphisms ( SNPs ) , each of which was present in at least two animals in this cohort of rhesus monkeys ( Fig . 1 ) . The 2-amino acid deletion and most of the nonsynonymous SNPs ( nsSNPs ) clustered in the coiled-coil ( CC ) and the B30 . 2 ( SPRY ) domains of the molecule . An allele was identified with a nsSNP in the L2 linker , and 4 TRIM5 alleles were also characterized that had synonymous polymorphisms clustered in the CC and B30 . 2 ( SPRY ) regions of the molecule . Based on the distribution of the 10 nsSNPs , the rhesus monkey TRIM5 coding sequences were grouped into 12 distinct alleles . These alleles and their frequencies in a cohort of 79 monkeys are shown in Table 1 , ordered according to increasing numbers of nsSNPs . The predicted amino acid sequence of allele 1 was identical to the published rhesus monkey TRIM5α sequence ( AY625001 ) , and the predicted sequence of allele 11 was most divergent from allele 1 . TRIMCyp , a molecule in which the TRIM5 RBCC domains are fused to cyclophilin A ( CypA ) [15]–[18] , was cloned from cDNAs prepared from the rhesus monkeys and is designated as allele 12 . The coincident expression of two different alleles was observed in many of the monkeys . We then sought to assess the impact of this rhesus monkey TRIM5 allelic diversity on the susceptibility of monkey cells to SIV infection . Since TRIM5α restricts the replication of retroviruses by interfering with a post-entry step in their life cycle , we chose first to evaluate the effect of specific TRIM5 allelic products by assessing the in vitro susceptibility of rhesus monkey B-lymphoblastoid cell lines ( B-LCLs ) expressing defined TRIM5 alleles to infection with a vesicular stomatitis virus G ( VSV-G ) -pseudotyped SIVmac239-GFP construct . The relative in vitro susceptibility to SIVmac239 infection of rhesus monkey B-LCLs expressing homozygous or heterozygous TRIM5 alleles 1–11 is shown in Fig . 2A and B . A significant variation in susceptibility to SIVmac239 infection was observed among these genetically diverse B-LCLs ( P = 0 . 0164 ) . We then used a Dunn's post test to compare the relative permissiveness for infection of each group of B-LCLs to that of the B-LCLs that expressed only TRIM5 allele 1 ( Table 2 ) . Rhesus monkey B-LCLs homozygous for TRIM5 alleles 7 , 9 and 10 were significantly more permissive for SIVmac239 infection . The TRIM5α proteins encoded by these alleles share identical B30 . 2 ( SPRY ) domain residues 328–497 . These amino acid residues are also shared by TRIM5α products of alleles 6 , 8 and 11 . Our inability to demonstrate statistically significant increased permissiveness of alleles 6 , 8 , and 11 for SIVmac239 infection may be a consequence of the small number of B-LCL populations for evaluation . In fact , when these B-LCLs were assigned to 3 groups on the basis of homozygosity or heterozygosity for expression of alleles 1–5 or 6–11 , a statistically significant difference in the permissivity of the cells for SIVmac239 infection was observed ( Fig . 2C ) . The observation above suggested that the permissiveness of rhesus monkey B-LCLs for SIVmac239 replication was associated with amino acid changes clustered in the B30 . 2 ( SPRY ) domain of TRIM5α . The TRIM5α proteins encoded by the more restricted alleles 1–5 are identical in amino acid residues 328–497 in the B30 . 2 ( SPRY ) domain . These sequences are also identical to that previously described as rhesus monkey TRIM5α AY625001 . Interestingly , alleles 6–11 are all identical in residues 328–497 of the B30 . 2 ( SPRY ) domain , but differ from alleles 1–5 because of a 2-amino acid deletion ( 339–340 ) and 3 nsSNPs ( A333S , P341Q and S422L ) . The results suggest that amino acid changes in the region including residues 328–497 of the TRIM5α B30 . 2 ( SPRY ) domain determine the efficiency of TRIM5α-mediated restriction for SIVmac239 replication in B-LCLs . A number of other complementing observations were made in studying the permissiveness of rhesus monkey B-LCLs assigned to 2 groups on the basis of homozygosity for expression of alleles 1–5 or 6–11 to retrovirus infection . We found that B-LCLs expressing only TRIM5 alleles 1–5 efficiently restricted SIVsmE543 and HIV-1 infection , whereas B-LCLs expressing only TRIM5α molecule from the group of alleles 6–11 were more permissive for infection by these virus constructs ( Fig . 3A , B ) . The overall infectivity of HIV-1 was substantially lower than that of SIVmac239 in the rhesus monkey B-LCLs . This likely reflects the fact that , even in B-LCLs bearing alleles 6–11 , rhesus monkey TRIM5α provides a significant barrier to HIV-1 infection . In contrast to this finding , no significant differences in susceptibility to infection by other retroviruses such as equine infectious anemia virus ( EIAV ) , feline immunodeficiency virus ( FIV ) , N-tropic murine leukemia virus ( N-MLV ) and B-tropic murine leukemia virus ( B-MLV ) were found between B-LCLs expressing one or the other groups of TRIM5 alleles ( Fig . 3C–F ) . These results suggest the rhesus monkey TRIM5α polymorphisms specifically influence restriction potency for a subset of retroviruses that includes the primate immunodeficiency viruses ( PIVs ) . The contribution of specific rhesus monkey TRIM5 alleles to the inhibition of SIVmac infection was also assessed in TRIM5-transduced cell lines . Feline renal fibroblast ( CRFK ) cells were stably transduced to express TRIM5α variants encoded by alleles 1 , 5 , 7 or 11 . To produce recombinant viruses , we co-transfected 293T cells with the pLPCX vectors expressing each TRIM5α variant with pVpack-GP and pVPack-VSV-G packaging plasmids . CRFK cells were transduced using the resulting viruses and then selected in G418 . These transduced cells were infected with a VSV-G-pseudotyped SIVmac239-GFP construct and subjected to flow cytometric analysis . Consistent with the studies of the B-LCLs , we found that cells transduced with allele 1 or 5 supported a lower level of infection by SIVmac239-GFP , whereas cells transduced with allele 7 or 11 supported a higher level of SIVmac239 infection ( Fig . 4A ) . We also assessed the infectivity of VSV-G-pseudotyped GFP-expressing SIVsmE543 , HIV-1 , EIAV , FIV , N-MLV and B-MLV viruses these same transduced cells . The restriction of SIVsmE543 and HIV-1 infection was more efficient in cells transduced with allele 1 or 5 than in cells transduced with allele 7 or 11 ( Fig . 4A ) . While transduction of cells with the TRIM5 alleles inhibited infection by EIAV , FIV and N-MLV , all evaluated TRIM5 alleles inhibited the infection with comparable efficiency ( Fig . 4B ) . Expression of these alleles did not restrict B-MLV infection ( Fig . 4B ) . We verified that the HA epitope tag at the C terminus of the TRIM5α protein did not affect the ability of transduced cells to restrict SIVmac239 infection ( data not shown ) . We also confirmed that downregulation of TRIM5α expression by siRNA specifically rescued the infectivity of the restricted SIVmac239 ( data not shown ) . Since the level of TRIM5α protein expression in each of the transduced cells was comparable as determined by Western blotting , differences in TRIM5α protein expression level could not account for the observed differences in the ability of the cells to restrict SIVmac239 infection . The TRIM5α proteins encoded by allele 1 and 7 , and by alleles 5 and 11 , differ only in the sequences of the B30 . 2 ( SPRY ) domain ( Table 1 ) . Therefore , as in the studies of B-LCLs , the permissiveness for SIVmac replication in the transduced cells was associated with amino acid changes clustered in the B30 . 2 ( SPRY ) domain of TRIM5α . We then explored the mechanism accounting for the differences in efficiency of restriction of SIVmac239 infection mediated by these different TRIM5 alleles . A direct interaction of TRIM5α with the retroviral capsid is required for restriction of retrovirus replication [19]–[21] . The recognition of the capsid by TRIM5α is mediated by the B30 . 2 ( SPRY ) domain . Because the phenotypic differences among rhesus monkey TRIM5α variants grouped according to the sequence of the B30 . 2 ( SPRY ) domain , we hypothesized that the two groups of TRIM5α variants might exhibit differences in the ability to bind SIVmac capsids . Two TRIM5α proteins encoded by alleles 1 and 7 were studied for capsid-binding ability . TRIM5 allele 1 represents the major allele and encodes a protein identical in sequence to that of the previously reported rhesus monkey TRIM5α [4] . The TRIM5α protein encoded by allele 7 differs from that encoded by allele 1 only in the sequence of the B30 . 2 ( SPRY ) domain ( Table 1 ) . Therefore , TRIM5α variants 1 and 7 were chosen as representatives of the two groups of TRIM5α B30 . 2 ( SPRY ) domain variants and tested for their ability to bind SIVmac capsid . For this purpose , we adapted an in vitro capsid-binding assay , which has been used to study TRIM5α binding to the HIV-1 capsid [22] . In our assay , SIVmac239 capsid-nucleocapsid ( CA-NC ) complexes were generated in vitro , and then incubated with serial dilutions of 293T cell lysates containing TRIM5α allelic variants 1 and 7 . After incubation , TRIM5α protein bound to CA-NC complexes was separated from unbound TRIM5α by pelleting through a sucrose cushion . The amount of TRIM5α in both input and pellet was analyzed by Western blotting and densitometry . As expected from previous experience with the HIV-1 capsid-nucleocapsid binding assay [23] , both TRIM5α allelic variants bound the SIVmac239 CA-NC complexes in a dose-dependent manner with a sigmoidal pattern . Importantly , however , binding of the variant encoded by TRIM5 allele 7 was less efficient than that of the TRIM5α allele 1 variant ( Fig . 5A , B ) . This finding is consistent with the possibility that differences in the efficiency of binding the SIVmac239 capsid contribute to the observed differences in SIVmac239 restriction potency , with the less efficiently binding allelic variant 7 mediating the less potent restriction of viral replication . To determine the ramifications of these in vitro findings for the control of viral replication in vivo , a study was done to determine whether the expression of particular TRIM5 alleles by Indian-origin rhesus monkeys was associated with the in vivo control of SIVmac251 replication . We tested the specific hypothesis that variation in residues 328–497 of the TRIM5α B30 . 2 ( SPRY ) domain influences the course of SIVmac251 infection in vivo . The monkeys were divided into 3 groups: in one group , only TRIM5 alleles 1–5 were expressed; in the second group , one copy of alleles 1–5 and one copy of alleles 6–11 were expressed; and in the third group , only alleles 6–11 were expressed . Monkeys were selected for this study that did not express the MHC class I alleles Mamu-A*01 , -B*08 , or -B*17 , as their expression is associated with particularly efficient control of SIV replication [24]–[26] . Eliminating these monkeys from the evaluated animals eliminated an already defined source of variation in virus control . Plasma SIV RNA levels were assessed in this cohort of rhesus monkeys through day 178 after SIVmac251 challenge . We measured viral replication for each monkey by doing an area-under-the curve calculation for the plasma SIV RNA levels between days 1 and 70 following infection . A 0 . 1 log median difference in these values was observed between the allele 1–5 homozygous and the allele 1–5/allele 6–11 heterozygous monkeys , and a 0 . 6 log median difference in these values was observed between the allele 1–5 homozygous and the allele 6–11 homozygous monkeys ( Fig . 6A ) . Plasma virus RNA levels were also assessed in this cohort of rhesus monkeys on days 14 ( peak ) and 70 ( set-point ) following the intravenous inoculation of SIVmac251 . Consistent with the decreased in vitro antiviral activity of the products of TRIM5 alleles 6–11 , the expression of TRIM5 alleles 6–11 by the rhesus monkeys was associated with higher levels of SIVmac251 replication in vivo . A 0 . 4 log median difference in plasma virus RNA levels at the time of peak viral replication was observed between the allele 1–5 homozygous and the allele 1–5/allele 6–11 heterozygous monkeys , and a 0 . 6 log median difference in these values was observed between the allele 1–5 homozygous and the allele 6–11 homozygous monkeys ( Fig . 6A ) . A log median difference in set-point virus RNA levels was 0 . 7 between the allele 1–5 homozygous and the allele 1–5/allele 6–11 heterozygous monkeys , and 1 . 3 between the allele 1–5 homozygous and the allele 6–11 homozygous monkeys ( Fig . 6A ) . We then assessed the clinical consequences of infection in these monkeys . First , we evaluated the loss of peripheral blood total CD4+ T cells and CM CD4+ T cells in the monkeys following SIVmac251 inoculation . We sought to determine whether the expression of particular TRIM5 alleles was associated with particular immune sequelae of infection . We displayed the data in two ways: dividing the monkeys into 3 cohorts as described above and dividing the monkeys into 2 cohorts , one group of animals expressing only alleles 1–5 and the other expressing at least 1 allele of the group 6–11 . The expression of TRIM5 alleles 6–11 by the rhesus monkeys was associated with a rapid loss of both total CD4+ T cells and CM CD4+ T cells ( Fig . 6B ) . Then we assessed the effect of the expression of TRIM5 alleles 6–11 on the survival of the monkeys following infection ( log-rank test of equality , Kaplan-Meyer survival curves ) . The monkeys expressing only alleles 1–5 maintained a statistically significant survival advantage over the monkeys expressing at least 1 allele of the group 6–11 ( Fig . 6C ) . Therefore , the expression of TRIM5 alleles 6–11 was associated with both less efficient control of SIVmac replication in vivo and an increased rate of disease progression .
We demonstrated TRIM5 allele-determined relative resistance to SIV infection in vitro using both B-LCL and transduced fibroblast target cells . However , the variable resistance or permissiveness for infection was more readily apparent in the B-LCL than in the fibroblasts . This difference between target cell populations may reflect differences in TRIM5 expression under normal physiologic control in B-LCLs and in transduced CRFK cells . Thus , overexpression in CRFK cells may partially mask the relative inefficiency of some of the TRIM5 alleles to restrict SIV and HIV replication in vitro . A number of genes have previously been implicated in the control of HIV-1 replication . These include selected MHC class I alleles and loci , including HLA B*5701 , HLA B*27 [27] and HLA C [28] , as well as HLA Bw4-8OI in association with KIR3DS1 [29] . Expression of an allelic form of the HIV-1 co-receptor , the delta 32 form of CCR5 , is also associated with HIV-1 containment [30]–[32] . These genes and gene loci are thought to contribute to HIV-1 control through classical immune effector mechanisms or viral entry . The present study shows that naturally occurring TRIM5α B30 . 2 ( SPRY ) polymorphisms affect immunodeficiency virus infections . These observations demonstrate the importance of SIV/B30 . 2 ( SPRY ) interactions in vivo . The finding that TRIM5 allelic products are associated with both the permissiveness of cells to HIV-1/SIV infection in vitro and clinical sequelae of SIV infection in vivo expands the known genetic determinants of HIV-1/SIV susceptibility and implicates a novel intracellular mechanism in that virus control . Our data suggest that the co-expression of restrictive and permissive TRIM5α variants results in a less effective antiretroviral state than the expression of two restrictive TRIM5α variants . Differences in SIVmac restriction potency and capsid binding map to the B30 . 2 ( SPRY ) domain of the rhesus monkey TRIM5α variants . As the functional TRIM5α moiety is thought to an oligomer [33]–[36] , co-expression of TRIM5α variants with different capsid-binding affinities would be expected to result in heterodimers with decreased avidity for the capsid . Dominant-negative effects of less functional TRIM5α variants on more potently restricting TRIM5α variants have been observed [4] , [37]–[39] . A number of groups have evaluated the contribution of common TRIM5α polymorphisms to AIDS susceptibility in HIV-1-infected humans [40]–[44] . These various studies have defined associations of nsSNPs in the RING , B-box 2 , Coiled-coil domains and Linker 2 region of the molecule with very modest effects on HIV-1 susceptibility and AIDS clinical progression [40]–[43] . The single SPRY domain ns SNP in human TRIM5α , H419Y , was also shown to have only a modest effect on these clinical endpoints [44] . The absence of evidence for a robust association between a SNP in TRIM5α and a substantial HIV-1-associated clinical endpoint in humans is likely a consequence of the fact that the common variant forms of this molecule in humans do not have severely impaired interactions with retroviruses . The human TRIM5 SNPs are for the most part clustered in the B-box 2 and Coiled-coil domains of the molecule , regions of TRIM5α that are not thought to contact the retrovirus capsid . In contrast , as we have shown in the present study , variants of TRIM5α in the rhesus monkey that cluster in the B30 . 2 ( SPRY ) domain of the molecule have a significant impact on the interaction of TRIM5α with the capsid of SIV and , as a result , exert a significant effect on SIV replication in vitro and in vivo . The findings in the present study indicate that TRIM5α has a function beyond restricting cross-species transmission of retroviruses . We show that TRIM5α variants with an impaired ability to interact with the retrovirus capsid restrict pathogenic retroviruses in a susceptible host species less efficiently both in vitro and in vivo . Therefore , TRIM5α proteins can exert antiretroviral effects ranging from modifications of viral load to complete suppression of infection .
HEK293T and feline renal fibroblasts ( CRFK ) were obtained from American Type Culture Collection and grown in RPMI/10% FBS . Recombinant SIVmac239-GFP virus was produced by cotransfection of HEK293T cells with pSIVmac239Δenv-GFP , pVSV-G and pRev using the Lipofectamine 2000 ( Invitrogen ) . HIV-1-GFP , SIVsmE543-GFP , EIAV-GFP , FIV-GFP , B- and N-MLV-GFP viruses were prepared as previously described [7] . At 48h after transfection , cell-free supernatant was collected . The titer of virus in the supernatant was determined by infection of HEK293T cells and assessment for % of cells expressing GFP by flow cytometric analysis . An SIV p27 antigen ELISA was performed to measure the p27 gag gene product of SIV . B-LCLs were generated from the same cohort of 28 uninfected rhesus monkeys and 32 rhesus monkeys previously infected with SIVmac251 as described [45] . Briefly , B cells were transformed by incubating peripheral blood mononuclear cells isolated from rhesus monkeys with Herpesvirus papio . B-LCLs were then propagated in RPMI 1640 supplemented with 20% FBS and penicillin-streptomycin . Total RNA was isolated from H . papio-immortalized rhesus monkey B-LCLs and PBMCs by using the RNeasy Mini kit ( Qiagen ) . Untagged , full-length TRIM5α cDNA clones were generated by RT-PCR with primers TRIM5F1 ( 5′-CAGACGAATTCCACCATGGCTTCTGGAATCCTG-3′ ) and TRIM5R1 ( 5′-GGACGTTCGAAATAGAAAGAAGGGAGACAGC-3′ ) by using the SuperScript One-Step kit ( Invitrogen ) . PCR products were cloned directly into the TOPO Blunt vector ( Invitrogen ) . Six to 15 independent cDNA clones from individual rhesus monkeys were subjected to automated sequence analysis ( GENEWIZ ) . Genomic DNA was isolated from lymphocytes from the same cohorts of rhesus monkeys by using QIAamp DNA kit ( Qiagen ) , and sequenced . Primer selection was facilitated by the use of the computer program Beacon 3 . Primer sequences used to amplify and sequence TRIM5 exons or TRIM5 cDNAs are shown in Table S1 . To generate TRIM5α C-terminally tagged with hemagglutinin ( HA ) , genes were amplified with primers TRIM5F1 and TRIM5HA-R ( 5′-CCACCGGTGGCTCAAGCGTAGTCTGGGACGTCGTATGGGTAGCCGCCAGAGCTTGGTGAGCACAGAG-3′ ) . B-LCLs were selected for study that had previously defined TRIM5 alleles . A VSV-G pseudotyped recombinant virus constructs were used for single-round infection . 4×104 cells per well were seeded into 96-well plates and cells were infected with serial dilutions of virus by spinoculation for 2 h at 1 , 900×g . At 48 h after infection , these B-LCLs were evaluated for their relative susceptibility to retrovirus replication by assessing for the % of cells expressing GFP as determined by flow cytometric analysis . Feline renal fibroblast ( CRFK ) cells stably expressing TRIM5α variants encoded by allele 1 , 5 , 7 or 11 were generated as described by Stremlau et al . [4] . Briefly , 293T cells were co-transfected with the pLPCX vectors containing each TRIM5α cDNA with pVpack-GP and pVPack-VSV-G packaging plasmids ( Stratagene ) to produce recombinant viruses . CRFK cells were then transduced using the resulting viruses and selected in G418 ( 0 . 5mg/ml , Sigma-aldrich ) . For infection , CRFK cells expressing TRIM5α variants encoded by allele 1 , 5 , 7 or 11 , or control cells transduced with the empty pLPCX vector were harvested and subsequently seeded into 48-well plates ( 1 . 2×104 cells per well ) . Cells were incubated with a VSV-G-pseudotyped recombinant virus construct for 48 h and then subjected to flow cytometric analysis . Feline renal fibroblast ( CRFK ) cells were transduced as described above and harvested . Cells were then lysed in NP-40 lysis buffer ( Boston BioProducts ) . Total protein was measured by using a BCA assay kit ( Pierce ) and equal amounts ( 20 µg ) were separated by SDS/PAGE . HA-tagged TRIM5α was detected with either monoclonal anti-HA antibody ( Roche ) and HRP-conjugated either anti-rat or anti-rabbit IgG secondary antibody , respectively ( Sigma-Aldrich ) . Levels of β-actin were also assessed as a control for the loading of total protein by using anti-β-actin antibody ( Sigma-Aldrich ) . The SIVmac239 CA-NC fusion protein was expressed in Escherichia coli and purified with minor modifications from the previously described purification method used for the HIV-1 CA-NC protein [22] . The purified SIVmac CA-NC protein was mixed with ( TG ) 50 DNA oligonucleotides in 50 mM Tris-HCl ( pH 7 . 0 ) and 500 mM NaCl solution and incubated at 37°C . The resulting SIVmac CA-NC complexes were negatively stained and examined under the electron microscope , confirming that the size and shape of the CA-NC complexes are very similar to those HIV-1 CA-NC complexes . Approximately 1×107 293T cells were transiently transfected to express the appropriate TRIM5 allele or with a control vector , and then harvested at 24 h post-transfection . The cells were lysed in 1 ml of hypotonic lysis buffer ( 10 mM Tris-HCl , pH 7 . 0/10 mM KCl ) . After brief centrifugation at 4°C to remove cell debris , the supernatant was spun at 110 , 000×g for 1 h at 4°C in a Beckman ultracentrifuge . After the pre-cleaning spin , different amounts of the supernatants containing the TRIM5α variants were aliquoted in separate tubes and were complemented with the supernatants from the vector-transformed cells to achieve a final volume of 200 µl . To these mixtures , the same amount of SIVmac CA-NC complexes were added and the final concentration of NaCl was adjusted to 200 mM before incubation at 30°C for 1 h . After incubation , 20 µl of the mixtures were saved for further analysis as an input and the rest of mixtures were layered onto 60% sucrose cushion ( prepared in 1× PBS ) and centrifuged at 110 , 000×g for 1 h at 4°C in a Beckman ultracentrifuge . The pellet was resuspended in 1× SDS sample buffer and subjected to SDS/PAGE and Western blotting . Peripheral blood mononuclear cells ( PBMCs ) were stained with MAbs specific for cell surface molecules , including CD3 , CD4 , CD28 , and CD95 , and then subjected to flow cytometric analysis . Peripheral blood CD4+ T lymphocyte and central memory CD4+ T lymphocyte counts were calculated by multiplying the total lymphocytes count by the percentage of CD3+CD4+ T cells , and CD95+CD28+ T cells , respectively , determined by MAb staining and flow cytometric analysis [46] . The two-sided nonparametric Mann-Whitney test was used to determine statistical significance . Levels of P <0 . 05 were considered statistically significant . For multiple comparisons , non-parametric one-way ANOVA Kruskal-Wallis test was used with Dunn's multiple comparison test . | The cytoplasmic TRIM5α restricts the replication of a broad range of retroviruses in a species-specific manner . In the present study we show that TRIM5α is more than a species barrier for retroviruses . We show that naturally occurring B30 . 2 ( SPRY ) polymorphisms affect retrovirus infection . These observations demonstrate the importance of SIV/B30 . 2 ( SPRY ) interactions in vivo . These findings are the first demonstration of the importance of such a pathogen/host protein interaction in vivo . Importantly , the striking variability in the clinical course of HIV-infected individuals has long puzzled the biomedical community . A large number of investigators have devoted considerable effort to determine what genetically determined factors might contribute to the containment of HIV replication , reasoning that an understanding of the determinants of effective control of HIV spread will provide important targets for both drug and vaccine development . Our demonstration in the present study that B30 . 2 ( SPRY ) polymorphisms have a dramatic effect on the clinical outcome of an AIDS virus infection highlight the extraordinary importance of TRIM5α on the control of an AIDS virus infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"virology/mechanisms",
"of",
"resistance",
"and",
"susceptibility,",
"including",
"host",
"genetics",
"virology/immunodeficiency",
"viruses",
"virology/animal",
"models",
"of",
"infection",
"virology/host",
"antiviral",
"responses"
] | 2010 | TRIM5α Modulates Immunodeficiency Virus Control in Rhesus Monkeys |
Assembling of the membrane-bound viral replicase complexes ( VRCs ) consisting of viral- and host-encoded proteins is a key step during the replication of positive-stranded RNA viruses in the infected cells . Previous genome-wide screens with Tomato bushy stunt tombusvirus ( TBSV ) in a yeast model host have revealed the involvement of eleven cellular ESCRT ( endosomal sorting complexes required for transport ) proteins in viral replication . The ESCRT proteins are involved in endosomal sorting of cellular membrane proteins by forming multiprotein complexes , deforming membranes away from the cytosol and , ultimately , pinching off vesicles into the lumen of the endosomes . In this paper , we show an unexpected key role for the conserved Vps4p AAA+ ATPase , whose canonical function is to disassemble the ESCRT complexes and recycle them from the membranes back to the cytosol . We find that the tombusvirus p33 replication protein interacts with Vps4p and three ESCRT-III proteins . Interestingly , Vps4p is recruited to become a permanent component of the VRCs as shown by co-purification assays and immuno-EM . Vps4p is co-localized with the viral dsRNA and contacts the viral ( + ) RNA in the intracellular membrane . Deletion of Vps4p in yeast leads to the formation of crescent-like membrane structures instead of the characteristic spherule and vesicle-like structures . The in vitro assembled tombusvirus replicase based on cell-free extracts ( CFE ) from vps4Δ yeast is highly nuclease sensitive , in contrast with the nuclease insensitive replicase in wt CFE . These data suggest that the role of Vps4p and the ESCRT machinery is to aid building the membrane-bound VRCs , which become nuclease-insensitive to avoid the recognition by the host antiviral surveillance system and the destruction of the viral RNA . Other ( + ) RNA viruses of plants and animals might also subvert Vps4p and the ESCRT machinery for formation of VRCs , which require membrane deformation and spherule formation .
Plus-stranded ( + ) RNA viruses replicate by assembling membrane-bound viral replicase complexes ( VRCs ) consisting of viral- and host-coded proteins in combination with the viral RNA template in the infected cells . Although major progress has recently been made in understanding the functions of the viral replication proteins , including the viral RNA-dependent RNA polymerase ( RdRp ) and auxiliary replication proteins , the contribution of many host proteins to VRC assembly is far from complete [1]–[7] . The host proteins contributing to VRC assembly likely include translation factors , protein chaperones , RNA-modifying enzymes , and cellular proteins involved in lipid biosynthesis [8]–[14] . Other host proteins , such as the ESCRT proteins , reticulons and amphiphysins could be involved in membrane deformation occurring during VRC assembly [15]–[17] . However , the actual functions of the majority of the identified host proteins involved in VRC assembly have not been fully revealed . To assemble their VRCs , RNA viruses take control of cell membranes by interfering with intracellular lipid metabolism , protein regulation , targeting and transport [7] , [18] . Viral polymerases of many ( + ) RNA viruses interact with membranes and build functional VRCs in spherules that are single-membrane vesicles with a narrow opening to the cytosol . Spherules form as invaginations in a variety of cell organelles [7] , [18] , [19] . Tubulovesicular cubic membranes , double membrane vesicles ( DMV ) and planar oligomeric arrays are some other classes of membranous structures that can harbor VRCs as documented in the literature [18] . TBSV is a small ( + ) RNA virus that has recently emerged as a model virus to study virus replication , recombination , and virus - host interactions using yeast ( Saccharomyces cerevisiae ) as a model host [7] , [20]–[23] . Several systematic genome-wide screens and global proteomics approaches have led to the identification of ∼500 host proteins/genes that interacted with the viral replication proteins or affected TBSV replication and recombination [9] , [11] , [24]–[32] . Subsequent detailed analysis revealed the functions of the two viral replication proteins ( i . e . , p33 and p92pol ) , the viral RNA , the host heat shock protein 70 ( Hsp70 ) and the eukaryotic elongation factor 1A ( eEF1A ) , sterols and phospholipids in the assembly of the tombusvirus VRCs [30] , [31] , [33]–[41] . Hsp70 and eEF1A proteins have been shown to bind to the viral replication proteins [1] , [33] , [42] . The auxiliary p33 replication protein , which is an RNA chaperone , recruits the TBSV ( + ) RNA to the site of replication , which is the cytosolic surface of peroxisomal membranes [43]–[48] . The RdRp protein p92pol binds to the essential p33 replication protein that is required for assembling the functional VRC [22] , [34] , [45] , [49] , [50] . Interestingly , the genome-wide screens and a proteome-wide over-expression approach for host factors affecting TBSV replication in yeast [9] , [11] , [26] have led to the identification of 11 ESCRT ( endosomal sorting complexes required for transport ) proteins involved in multivesicular body ( MVB ) /endosome pathway [51] , [52] . The identified host proteins included Vps27p ( ESCRT-0 complex ) ; Vps23p and Vps28p ( ESCRT-I complex ) , Vps25p and Vps36p ( ESCRT-II complex ) ; Snf7p and Vps24p ( ESCRT-III complex ) ; Doa4p ubiquitin isopeptidase , Did2p having Doa4p-related function; Bro1p ESCRT-associated protein and Vps4p AAA+ ATPase [9] . The identification of ESCRT proteins led to a model that tombusvirus replication depends on hijacking of ESCRT proteins to the peroxisomal membrane . It has been suggested that the protection of the viral RNA is compromised within the VRC assembled in the absence of cellular ESCRT proteins . Altogether , the formation of membranous spherule-like replication structures in infected cells might require co-opted ESCRT proteins . However , the actual functions of the subverted ESCRT proteins in TBSV replication are currently unknown . A set of 20–30 ESCRT proteins is important for the endosomal/multivesicular body ( MVB ) protein-sorting pathway in eukaryotic cells , which down-regulates plasma membrane proteins via endocytosis; and sorts newly synthesized membrane proteins from trans-Golgi vesicles to the endosome , lysosome or the plasma membrane [53]–[55] . The ESCRT proteins are involved in membrane invagination and vesicle formation during the MVB pathway . Defects in the MVB pathway can cause serious diseases , including cancer , early embryonic lethality and defect in growth control [53]–[56] . Also , enveloped retroviruses ( such as HIV ) , ( + ) and ( − ) RNA viruses ( such as filo- , arena- , rhabdo- and paramyxoviruses ) redirect cellular ESCRT proteins to the plasma membrane , leading to budding and fission of the viral particles from infected cells [51] , [52] . The MVB pathway starts with the recognition of monoubiquitinated cargo proteins in the endosome by Vps27p ( ESCRT-0 complex ) , which serves as a signal for proteins to be sorted into membrane microdomains of late endosomes [54]–[57] . Vps27p in turn recruits Vps23-containing ESCRT-I complex and then the ESCRT-II complex , resulting in grouping the cargo proteins together in the limiting membranes of late endosomes and deforming the membranes that leads to membrane invagination into the lumen [58] , [59] . Then , components of the large ESCRT-III complex are recruited from the cytosol , followed by sequential assembly of ESCRT-III monomers into helical lattice on the membrane that leads to the scission of the neck of the invaginated membrane , giving raise to vesicle budding into the lumen of endosome and to MVB formation [60] , [61] . Then , Vps4p recycles the ESCRT proteins , whereas Doa4p recycles the ubiquitin , leading to budding of multiple small vesicles into the lumen [57] . The ESCRT-III and the Vps4p AAA+ ATPase together comprise a conserved membrane scission machinery . The ATP-dependent function of Vps4p is to disassemble and remove the ESCRT-III components from the membranes ( i . e . , recycling them back to the cytosol ) [59] , [62] . Vps4p is a member of the AAA+ ATPase family , which uses ATP to remodel macromolecular structures in various biological processes , such as protein disaggregation , microtubule severing and membrane fusion . The N-terminal part of Vps4p , termed MIT domain , binds to the ESCRT-III components , while the ATPase domain is involved in ATP hydrolysis . Vps4p is present as an inactive dimer in the cytosol and during activation , Vps4p likely forms a dodecamer with two parallel rings and the ESCRT-III components are likely pulled across the central hole of these rings during the Vps4p-driven recycling event [62] . Our previous works have demonstrated that tombusviruses co-opt selected components of the cellular ESCRT machinery for replication via interaction of the p33 replication protein with Vps23p ( similar to Tsg101 in mammals ) and Bro1p ( ALIX ) [15] , [63] . The recruitment of these cellular ESCRT proteins to the sites of tombusvirus replication is assumed to lead to the sequential recruitment of additional selected ESCRT factors , such as ESCRT-III and Vps4p AAA+ ATPase . Indeed , in the absence of these cellular factors in yeast or the expression of dominant negative versions of these proteins in N . benthamiana host plant , tombusvirus replication is decreased by 10-to-20-fold [15] , [63] . Based on the known functions of the ESCRT proteins , it was suggested that the ESCRT proteins are recruited by TBSV to aid the formation of VRCs , which require membrane deformation to induce spherule-like structures . However , the presented data do not explain how the “neck” of the spherule-like replication structure is stabilized ( to maintain an opening towards the cytosol ) and why the recruitment of ESCRT factors does not lead to enclosure of the replicase complex ( i . e . , the VRCs are not converted rapidly into vesicles via scission of the necks of the spherules that bud inside the peroxisomes ) . The latter event is predicted if TBSV would take advantage of the canonical functions of the ESCRT proteins , which always result in budding of the newly formed vesicles away from the cytosol [57] , [59] . In this paper , we identify Vps4p AAA+ ATPase as a major host factor in TBSV replication . We show that the p33 replication protein interacts with Vps4p and three other ESCRT-III proteins . Surprisingly , we find that Vps4p is a permanent member of the tombusvirus VRCs and it also interacts with the viral RNA . EM images revealed that Vps4p is localized in a compartment also containing the viral double-stranded ( ds ) RNA . Altogether , we propose that the interaction of p33 and Vps4p is critical for spherule formation and efficient tombusvirus replication . These data are consistent with the model that TBSV co-opts ESCRT proteins for its replication and these ESCRT proteins play noncanonical functions in aiding VRC formation and TBSV replication .
The formation of spherule-like structures during TBSV replication on the peroxisomal membrane surfaces and the effect of various ESCRT proteins on tombusvirus replication [15] , [48] , suggest that some of the co-opted cellular ESCRT proteins might play noncanonical functions . To gain insights into the functions of the co-opted ESCRT proteins during tombusvirus replication , first we analyzed if p33 replication protein could interact with ESCRT-III components or the Vps4p AAA+ ATPase . The split-ubiquitin-based yeast two-hybrid assay ( membrane-based MYTH assay ) between the tombusvirus p33 and the yeast ESCRT-III components or Vps4p revealed strong interaction between p33 and Vps4p ( Fig . 1A ) . This is surprising , since Vps4p is known to interact with the ESCRT-III components to recycle them from the endosome [59] , [62] , while recycling of the peroxisome-bound p33 to the cytosol is unlikely to happen and not yet documented . In addition , we observed good interactions between p33 and Vps2p , Vps20p , and Vps24p ESCRT-III factors ( Fig . 1A ) . The most abundant ESCRT-III factor , namely Snf7p , whose deletion greatly affected TBSV replication [15] , and Did2p interacted only weakly with p33 ( Fig . 1A ) . Affinity-based co-purification experiments from the membrane fraction of yeast confirmed that Vps4p strongly interacted with p33 ( Fig . 1B , lane 12 ) , while the interaction of Vps2p and Vps20p with p33 was also detectable ( Fig . 1B , lanes 2 and 8 ) . Unlike in the MYTH assay , the co-purification experiments suggested strong interaction between p33 and Vps24p ( Fig . 1B , lane 4 ) . We could not co-purify Snf7p and Did2p with p33 ( Fig . 1B , lanes 6 and 10 ) . Altogether , Vps4p showed consistently the strongest interaction with the p33 replication protein and this interaction is unexpected and could play a direct role in TBSV replication . To confirm that the interaction between the viral replication protein and Vps4p also occurs in plants , we co-expressed the Arabidopsis thaliana AtVps4 in Nicotiana benthamiana leaves together with the FLAG-tagged tombusvirus p33 replication protein ( Fig . 1C ) . After isolation of the membrane-bound replicase from the leaves and solubilization of the membrane fraction with nonionic detergent , we FLAG-affinity purified p33 , followed by Western blotting . This approach revealed co-purification of AtVps4 with the tombusvirus p33 ( Fig . 1C , lane 1 ) , while AtVps4 was missing after purification in the sample prepared from leaves lacking p33 ( lane 2 ) . Thus , similar to yeast , the tombusvirus p33 replication protein also interacts with AtVps4 in plant leaves , suggesting subversion of Vps4 for viral activities . Since Vps4p showed strong interaction with the tombusvirus p33 replication protein , we tested if Vps4p is part of the tombusvirus VRC . Interestingly , the affinity-purified tombusvirus replicase contained Vps4p ( Fig . 2A , lane 2 versus 1 ) . This finding highlights the possibility that Vps4p is a permanent component of the VRC ( i . e . , not rapidly recycled ) , such as Hsp70 [40] . To test this , we added cyclohexamide to yeast to prevent new p33 and p92pol translation , thus formation of new VRCs . Then , we measured the level of Vps4p in the affinity-purified replicase at various time points to study if Vps4p is released from the VRCs . As a control , we used Ssa1p Hsp70 , whose amount did not change within 150 min , confirming that Ssa1p remained stably associated with the existing tombusvirus VRCs ( Fig . 2B , lanes 10–12 ) . In contrast , the amount of Pex19p peroxisomal shuttle protein , which is involved in the delivery of p33 and p92pol to the peroxisomes [64] before its getting recycled to the cytosol , decreased by 60% after 150 min of incubation in the presence of cyclohexamide ( Fig . 2B , lanes 2–4 ) . Interestingly , the amount of Vps4p did not change significantly in the affinity-purified replicase preparations during this time period ( Fig . 2B , lanes 6–8 ) , suggesting that Vps4p is likely a permanent component of the tombusvirus VRCs . Obviously , this is different from the canonical role of Vps4p , which is quickly recycled from the endosomal membranes to the cytosol after the disassembly of the endosome-bound ESCRT-III structures [59] , [62] . Additionally , the N-terminally truncated Vps4p carrying the ATPase domain , but lacking the MIT domain , which is responsible for interaction with the ESCRT-III proteins , was recruited to the VRCs ( Fig . 2A , lane 4 ) , suggesting unique interaction between p33 and Vps4p . To map the binding sites in Vps4p , we used the split ubiquitin assay that revealed that the N-terminal MIT domain , which binds to the ESCRT-III proteins [59] , [62] , bound efficiently to the full-length p33 ( construct 1–100 versus the full-length Vps4 construct 1–437 , Fig . 3A ) . We also observed weaker , but substantial binding between the C-terminal ATPase domain and p33 ( Fig . 3A ) . We confirmed the interaction using the C-terminal ATPase domain and compared with the full-length Vps4p in a pull down assay with p33 ( Vps4-ΔMIT , Fig . 3B ) . The interaction between the ATPase domain and p33 is not abolished by addition of ATP ( not shown ) . Overall , the binding of p33 to Vps4p is different from the binding between Vps4p and the cellular ESCRT-III components that target only the MIT domain and leads to the Vps4p-driven recycling of the ESCRT-III components from the membranes back to the cytosol . We suggest that the unique interaction between the p33 and Vps4p subverts Vps4p for viral replication , leading to association of Vps4p with the membrane-bound replicase , and likely altering the canonical function of Vps4p . Detailed mapping of p33 sites interacting with the Vps4p or the ATPase domain revealed that the very C-terminus of p33 , which contains the p33:p33/p92 interaction sites , is involved in binding to Vps4p ( Fig . 3C–F ) . Binding of the ATPase domain of Vps4p to p33 mostly overlaps with that of the full-length Vps4p in this assay . Altogether , the binding between Vps4p and p33 involves unique interactions that will require high-resolution structural studies . We have developed an EM-based assay to visualize the tombusvirus-induced spherule-like structures in yeast , which are known to form in infected plant tissues [15] , [48] . EM images revealed the presence of single-membrane vesicle-like structures of ∼25–50 nm ( Fig . 4A–C ) that were missing in wt yeast not expressing the tombusvirus replication proteins ( Fig . S1A–B ) . These vesicles likely represent the spherules seen in TBSV-infected plant tissues [15] . To show if Vps4p is localized to the tombusvirus VRCs , we used immuno-gold EM of yeast cells co-expressing HA-tagged Vps4p , and MT ( Metal-binding protein metallothionein ) -tagged p33 that was visualized by Metal-Tagging Transmission Electron Microscopy ( METTEM ) [65] and replicating the TBSV repRNA , which was detected through using a dsRNA-specific antibody [65] ( Fig . 4D–G ) . These samples were processed in the absence of osmium tetroxide that would destroy most protein epitopes and would mask the 1 nm nano-clusters associated to p33-MT . We frequently observed Vps4p in the close vicinity of the dsRNA ( present in the VRCs ) based on using different sized gold particles for immuno-gold EM . It was difficult to detect Vps4p in the areas of the yeast cells lacking dsRNA ( Fig . S2 ) . We suggest that Vps4p is likely concentrated in the VRCs due to binding to p33 , thus facilitating detection of Vps4p in yeast membranous compartments replicating TBSV . For an adequate visualization of gold nanoclusters bound to p33-MT , cells in Fig . 4D and 4E were processed in the absence of osmium tetroxide and contrasting agents as described . Under these conditions intracellular membranes are invisible; however , some membranes can be visualized if these ultra-thin sections are stained with uranyl acetate and lead citrate ( Fig . 4G–I ) ( see Materials and Methods ) . Stained cells showed that Vps4p and the viral dsRNA were surrounded by membranes ( Fig . 4G–I ) , supporting the model that Vps4p is part of the functional membrane-bound tombusvirus VRCs . To enhance the visualization of Vps4p , we over-expressed His6-MT-tagged Vps4p in yeast cells replicating TBSV repRNA . This approach facilitated the frequent observation of Vps4p in the close vicinity of the TBSV dsRNA by immuno-gold EM ( Fig . 5 ) . Interestingly , the spherule-containing membranous structures are frequently connected to form a complex compartment where Vps4p and dsRNA are clustered together ( Fig . 5 ) . Altogether , the co-localization ( proximal location based on immuno-EM ) of Vps4p and the tombusviral dsRNA in yeast cells suggests that Vps4p is recruited to the VRCs actively involved in TBSV RNA replication . If Vps4p is part of the tombusvirus VRCs , it is likely associated with the neck structure of tombusvirus-induced spherules , which is predicted to serve as exit places for the newly synthesized ( + ) RNA from the VRCs . Therefore , we tested if Vps4p is in contact with the tombusviral RNA . Affinity purification of Vps4p from the membrane fraction of yeast containing the tombusvirus VRCs resulted in co-purification of the TBSV ( + ) repRNA ( Fig . 6A , lane 3 ) . Interestingly , affinity purification of Vta1p accessory protein , which facilitates the formation of the functional Vps4p rings with ATPase function from the inactive Vps4p dimers [62] , also resulted in co-purification of the TBSV ( + ) repRNA ( Fig . 6A , lane 4 ) . As a positive control , we also used the tombusvirus p33 replication protein ( purified via FLAG-tag ) , which also resulted in co-purification of the TBSV ( + ) repRNA ( Fig . 6A , lane 5 ) . The Ssa1p Hsp70 chaperone , which is another permanent component of the tombusvirus VRCs also resulted in co-purified ( + ) repRNA , although in a lesser amount than p33 or Vps4p ( Fig . 6A , lane 6 versus 3 and 5 ) . The negative control ( HF , a peptide sequence containing His6-FLAG sequence expressed from the empty expression plasmid ) did not contain any detectable TBSV ( + ) repRNA , excluding that the viral RNA bound nonspecifically to the affinity resin or the FLAG-antibody . We also tested the presence of ( − ) repRNA ( which might be present in a dsRNA form within the VRCs ) in the above samples . As expected , the p33 replication protein preparation contained ( − ) repRNA ( Fig . 6A , lane 15 ) , while Vps4p , Vta1p , Ssa1p and the negative control samples did not contain ( − ) repRNA ( Fig . 6A ) . Based on these data , we suggest that the Vps4p/Vta1p complex might be in direct contact with the TBSV ( + ) repRNA . This contact is unlikely via p33 replication protein since both ( + ) and ( − ) repRNAs were co-purified with p33 . To obtain evidence that Vps4p is involved in the formation of tombusvirus VRCs , we took advantage of a cell-free TBSV replication assay based on TBSV-free yeast cell-free extracts ( CFE ) prepared from wt or vps4Δ yeast [41] . The CFE supports the assembly of the VRCs when purified recombinant p33 and p92 and ( + ) repRNA are added . Micrococcal nuclease was also added for 15 min ( after which it was inactivated ) to the assay at various time points to test the nuclease-sensitivity of the viral repRNA within newly assembled membrane-bound VRCs ( Fig . 7A ) . The CFE from wt yeast was able to assemble nuclease-insensitive VRCs in ∼45–60 min ( Fig . 7B , lane 5 ) . In contrast , the VRCs assembled with CFE from vps4Δ yeast produced only small amounts of repRNA in the presence of nuclease , suggesting that Vps4p is required for the assembly of the nuclease-insensitive tombusvirus VRCs in vitro . The nuclease-insensitivity of the viral RNA requires cellular membranes and ATP/GTP-dependent VRC assembly and is not due to protein coating of the viral RNA , as shown previously [40] , [41] . To visualize the tombusvirus VRCs in yeast lacking Vps4p , we performed EM imaging of vps4Δ yeast replicating TBSV repRNA . Interestingly , unlike the characteristic tombusvirus-induced spherule-like structures in wt yeast ( Fig . 8A and B ) , the membranes from vps4Δ yeast expressing the p33 and p92 replication proteins and the repRNA showed different deformations ( Fig . 8C and D ) . These crescent-shaped structures likely represent incompletely formed spherules with wide openings containing viral replicases . Similar structures ( either spherules or open crescent-shaped structures ) were not visible in control yeasts not expressing the viral proteins ( not shown ) . Similar crescent-shaped membrane-structures were also visualized by EM in vps24Δ yeast replicating TBSV repRNA ( Fig . 8E ) , suggesting that ESCRT-III components are also likely needed for proper deformation of membranes and VRCs formation . We interpret these data that the tombusvirus replication proteins could not induce the formation of complete spherule-like structures in vps4Δ or vps24Δ yeasts . To visualize the tombusvirus p33 replication protein in yeast subcelluar membranes , we used Metal-Tagging Transmission Electron Microscopy ( METTEM ) [66] with yeast expressing MT-tagged p33 replication protein . The MT-p33 molecules were present in elongated structures in vps4Δ yeast ( Fig . 9A ) , while they could form round vesicle-like structures in wt yeast ( Fig . 9B ) . These data are consistent with the model that Vps4p is required for the formation of tombusvirus-induced spherule-like structures in yeast cells . Labeling with anti-dsRNA antibodies revealed weak to moderate signals in vps4Δ yeast ( Fig . 9C ) . These results suggest that , in the absence of Vps4p , the abnormal VRCs are still able to support some repRNA replication . However , the level of repRNA replication is low as shown in the CFE-based replication assay ( ∼75% less repRNA replication in this strain than in wt ) [15] . Thus , replication of repRNA is inefficient in vps4Δ yeast and the repRNAs is less protected in wide open VRCs and much more sensitive to degradation .
In this paper , we document a novel , noncanonical role for the Vps4p AAA+ ATPase during TBSV replication . We propose that tombusviruses not only recruit the cellular ESCRT machinery for assembling the membrane-bound VRCs , but the virus could alter the activities of Vps4p and the ESCRT-III proteins to create new functional structures ( spherules acting as VRCs ) and possibly new activities . The recruitment of Vps4p and additional ESCRT proteins are needed for the assembly of the replicase complex , which could help the virus evade recognition by the host defense surveillance system and/or prevent viral RNA destruction by the gene silencing machinery .
To study co-purification of host ESCRT proteins with p33 replication protein , the yeast strain BY4741 was transformed with plasmids pGBK-FLAGp33-CUP1 ( or pGBK-His33-CUP1 as control , see plasmids used in the Supplementary material ) plus the pYES plasmids expressing 6×His-tagged ESCRT proteins from the GAL1 promoter ( see supplementary Material and Methods S1 for additional information ) . The transformed yeasts were pre-grown in SD minimal media plus 2% glucose for 20 h at 29°C , then transferred to SD minimal media plus 2% galactose for 16 h at 29°C to induce over-expression of ESCRT proteins from the GAL1 promoter . Then , the cultures were supplemented with 50 µM CuSO4 to induce expression of FLAG-p33 or His6-p33 from the CUP1 promoter . The yeasts were collected by centrifugation after 8 h , washed in phosphate buffer saline ( PBS ) and then incubated in PBS plus 1% formaldehyde for 1 h on ice to cross-link proteins . Afterwards , the formaldehyde was quenched by adding glycine to a final concentration of 0 . 1 M . Then , yeasts were washed in PBS and processed to purify the FLAG-tagged p33 protein as described [31] . The FLAG purified fractions were eluted from the FLAG M2-agarose columns with SDS-PAGE loading buffer ( without 2-mercaptoethanol ) . Then the eluted fractions were supplemented with 2-mercaptoethanol ( 5% ) and boiled for 30 minutes to reverse cross-linking . To study co-purification of His6-Vps4p and His6-Vps4101–437 , the yeast strain BY4741 was transformed with plasmids pGBK-FLAGp33-CUP1/DI72-GAL1 ( or pGBK-Hisp33-CUP1/DI72-GAL1 as control ) , pGAD-His92-CUP1 [67] and pYC-NT-VPS4 or pYC-NT-vps4101–437 . Transformed yeasts were pre-grown in SD minimal media plus 2% glucose for 20 h at 29°C , then transferred to SD minimal media plus 2% galactose and 50 µM CuSO4 for 24 h at 29°C . The yeasts are then treated with 1% formaldehyde as above and processed for FLAG-p33 protein purification . To analyze the time dynamics of host proteins association with p33 , BY4741 was transformed with plasmids pGBK-FLAGp33-CUP1/DI72-GAL1 ( or pGBK-Hisp33-CUP1/DI72-GAL1 as control ) , pGAD-His92-CUP1 [67] and either pYES-PEX19 , pYES-VPS4 or pYES-SSA1 . Transformed yeasts were pre-grown in SD minimal media with 2% glucose for 20 h at 29°C and then transferred to media with 2% galactose for 24 h at 29°C . The cultures were supplemented with 50 µM CuSO4 for 2 . 5 h to induce expression of the viral proteins . Then cycloheximide was added ( 100 µg/ml ) to stop protein translation and samples were taken immediately ( time 0 ) , 60 and 150 minutes afterwards . Yeast cultures were treated with formaldehyde and processed for FLAG-p33 purification as above . Purified FLAG-p33 was detected by western blot using anti-FLAG antibody followed by anti-mouse antibody conjugated to alkaline phosphatase . Co-purified His6-tagged host proteins were detected with anti-His antibody followed by anti-mouse antibody conjugated to alkaline phosphatase . Detection was performed with NBT and BCIP . The pMAL c2x-derived plasmids described in the Supplementary materials were transformed into Epicurion Bl21-codon-plus ( DE3 ) -R1L cells ( Stratagene ) . Expression of MBP-tagged proteins was induced by IPTG as described [68] . pGEX-His-VPS4 and pGEX-His-vps4101–437 were also transformed into Epicurion Bl21-codon-plus ( DE3 ) -R1L cells . Expression of GST-His6-tagged Vps4p and Vps4101–437 was essentially done as described [68] with the exception that cultures were incubated at 23°C during IPTG treatment . The cells were broken by sonication as described [68] . The lysates were incubated with GST resin ( Novagen ) for 1 h at 4°C . The GST resin was washed four times with column buffer [68] and then incubated for 3 h at 21°C with column buffer plus 1 mM 2-mercaptoethanol + 1 mM CaCl2 + 1 U Thrombine ( Novagen ) to cleave the His6-Vps4p and His6-Vps4101–437 from the column-bound GST protein . The amylose columns containing the MBP or MBP-tagged p33 portions were then incubated with 15 µg of the purified His6-Vps4p or His6-Vps4101–437 for 1 h at 4°C . The columns were then washed five times with column buffer and the MBP-tagged proteins were eluted with column buffer plus 10 mM maltose . The MBP-tagged proteins were analyzed by SDS-PAGE electrophoresis followed by coomassie staining . Bound His6-Vps4p or His6-Vps4101–437 were detected with anti-His antibody followed by alkaline phosphatase- conjugated anti-mouse and NBT/BCIP . The yeast membrane two-hybrid assay , based on the split-ubiquitin system ( Dualsystems ) has been described before [31] . The plasmid pGAD-BT2-N-His33 was co-transformed with pPR-N-RE derived constructs into the reporter yeast strain NMY51 . Transformed yeasts colonies were suspended in 100 µl of water and serially diluted ( 10-fold ) in water . 6 µl of each dilution were spotted onto TLHA- plates , to score for interaction , or TL- plates , as growth controls . The yeast strain BY4741 was transformed with plasmids pGBK-Hisp33-CUP1/DI72-GAL1 , pGAD-His92-CUP1 and either pYC-HF , pYC-HF-VPS4 , pYC-HF-VTA1 or pYC-HF-SSA1 . Additionally yeast was transformed with pGBK-FLAGp33-CUP1/DI72-GAL1 and pGAD-His92-CUP1 . Transformed yeasts were pre-grown in SD minimal media with 2% glucose for 20 h at 29°C , then transferred to SD minimal media with 2% galactose for 24 h at 29°C . The cultures were then supplemented with 50 µM CuSO4 for 3 h to induce expression of viral proteins . The yeasts were collected by centrifugation and subjected to formaldehyde cross-linking and FLAG purification as described above . The FLAG purified fractions , eluted in SDS-PAGE loading buffer , were treated with phenol/chloroform and ethanol precipitated to recover co-purified RNA . For detection of DI-72 ( + ) RNA , RT reactions were done with SuperScript II ( Life Technologies ) and primer #22 ( GTAATACGACTCACTATAGGGCTGCATTTCTGCAATGTTCC ) , followed by PCR with primers #1165 ( AGCGAGTAAGACAGACTCTTCA ) and #927 ( TAATACGACTCACTATAGG ) . For detection of DI-72 ( − ) RNA , the primer used for RT was #18 ( GTAATACGACTCACTATAGGAGAAAGCGAGTAAGACAG ) , followed by PCR with primers #1190 ( GGGCTGCATTTCTGCAATG ) and #927 . The yeast strains BY4741 and vps4Δ::hphNT1 were transformed with plasmids pGBK-FLAGp33-CUP1 and pGAD-FLAGp92-CUP1 [67] . Transformed yeasts were grown and cell free extracts prepared as described [49] except that the cultures were supplemented with 50 µM CuSO4 1 . 5 h before harvesting , to induce expression of p33 and p92 , and incubated at 37°C for 45 min before harvesting . Replicase reactions were carried out as described [49] . RNA protection was tested by adding 1 mM CaCl2 and 50 ng micrococcal nuclease at selected time points followed by 15 min incubation and then inactivation of the nuclease by addition of 2 . 5 mM EGTA . All reactions were incubated for a total time of 3 h . Yeast strains ( see Supplementary materials ) were pre-cultured from plated single colonies by inoculation in 2 ml of YPG ( yeast extract peptone galactose ) and shaking overnight at 250 rpm at 30°C . For inducing and maintaining viral replication , yeasts cells were grown for 24 h in YPG at 23°C and shaken at 250 rpm . When OD600 was around 2 , cells were collected , centrifuged for 5 min at 4000 g , resuspended in TSD reduction buffer ( Tris-sulfate DTT , pH 9 . 4 ) and either chemically fixed for electron microscopy ( see below ) or processed for removing the cell wall and obtaining spheroplasts . For obtaining spheroplasts , yeast cells were incubated for 10 min at room temperature and treated at 30°C with 0 . 1 µg/µl zymolyase 20T ( AMS Biotechnology ) in spheroplast medium A ( 1× yeast nitrogen base , 2% ( w/v ) glucose , 1× amino acids , 1 M sorbitol , 20 mM TrisCl , pH 7 . 5 ) for 5 or 15 min , depending on the yeast strain . After zymolyase treatment , cells were centrifuged for 5 min , at 1000 g and 23°C and washed once with spheroplast medium B ( 1× yeast nitrogen base , 2% ( w/v ) glucose , 1× amino acids , 1 M sorbitol ) and twice with spheroplast medium A . For ultrastructural studies , yeast cells and spheroplasts were processed . Compared to whole yeast cells , spheroplasts are well infiltrated with fixatives and resins allowing an optimal visualization of intracellular compartments . Cells were first fixed for 20 min in suspension at room temperature with 8% paraformaldehyde and 1% glutaraldehyde followed by a second fixation step of 1 h at room temperature with 4% paraformaldehyde and 0 . 5% glutaraldehyde in HEPES ( pH 7 . 4 ) ; fixed cells were then processed by conventional embedding in the epoxy-resin EML-812 ( Taab Laboratories ) following procedures for an optimal preservation of cell endomembranes [65] , [69] , [70] . Cells were post-fixed for 1 h at 4°C with 1% osmium tetroxide and 0 . 8% potassium ferricyanide in water , washed with HEPES , and 40 min with 2% uranyl acetate at 4°C . During post-fixation , samples were protected from light . Cells were submitted to dehydration steps for 20 min each with increasing concentrations of acetone ( 50 , 70 , 90 , and twice in 100% ) at 4°C and incubated with acetone-resin ( 1∶1 ) with gentle agitation at room temperature . Cells were infiltrated overnight with pure resin for 1 day and polymerized at 60°C for 3 days . Ultrathin ( 50–70 nm ) sections were collected in 300 mesh cooper grids ( G300-C3 , Taab ) with a plastic layer of 0 . 25% formvar in chloroform . Then , grids were stained for 20 min with saturated uranyl acetate and for 2 min with lead citrate following standard procedures . Samples were studied in a Jeol JEM 1011 electron microscope operating at 100 kv . For visualization of MT-tagged p33 in cells and immunogold labeling , cells were processed by embedding in the acrylic resin LRWhite following procedures for an adequate preservation of protein epitopes and optimal visualization of small nano-clusters [66] . Spheroplasts were incubated for 75 min with 0 . 2 mM HAuCl4 ( SIGMA-ALDRICH ) in spheroplast medium A . This treatment builds gold nano-clusters in MT-tagged proteins allowing detection of protein molecules in cells with high sensitivity and at molecular scale resolution [66] , [71] . Cells were washed with spheroplast medium A before fixation with 4% paraformaldehyde and 0 . 2% glutaraldehyde in PHEM ( 20 mM PIPES , 50 mM HEPES , 20 mM EGTA and 4 mM MgCl2 , pH 6 . 9 ) ( 1 h at room temperature ) . Cells were submitted to short dehydration steps of 10 min each in increasing concentrations of ethanol ( 30 , 50 , 70 , 90 and twice with100% ) at 4°C . Spheroplasts were incubated in mixtures of ethanol- LR White acrylic resin ( 2∶1 , 2∶2 , 1∶2 ) with gentle agitation and protected from light and embedded in 100% resin for 24 h . Samples were polymerized for 48 h at 60°C . Ultra-thin sections were collected in 300 mesh Quantifoil holey carbon grids ( R 3 . 5/1 Cu/Rh , Quantifoil Micro Tools ) and studied without staining . For immunogold labeling sections of cells embedded in LR White acrylic resin were processed as described [70] . Briefly , sections were incubated for 6 min with 1% BSA ( Bovine serum albumin ) in PBS , with primary antibodies diluted in 1% BSA and with secondary antibodies conjugated with 5 or 10 nm colloidal gold particles ( from BB International ) and diluted in 1% BSA . Dilutions of antibodies were rabbit anti-HA 9110 from ABCAM ( 1∶200 ) and mouse anti-dsRNA MAb J2 from English & Scientific Consulting ( 1∶200 ) . Secondary antibodies were diluted 1∶40 . | Replication of positive-stranded RNA viruses depends on recruitment of host proteins and cellular membranes to assemble the viral replicase complexes . Tombusviruses , small RNA viruses of plants , co-opt the cellular ESCRT ( endosomal sorting complexes required for transport ) proteins to facilitate replicase assembly on the peroxisomal membranes . The authors show a surprising role for the ESCRT-associated Vps4p AAA+ ATPase during tombusvirus replication . They show that Vps4p is recruited to and becomes a permanent member of the replicase complex through its interaction with the viral replication proteins . Also , EM and immuno-EM studies reveal that Vps4p is required for the formation of single-membrane vesicle-like structures , called spherules , which represent the sites of tombusvirus replication . The authors propose that Vps4p and other ESCRT proteins are required for membrane deformation and replicase assembly . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"model",
"organisms",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"microbiology",
"molecular",
"cell",
"biology",
"agriculture",
"research",
"and",
"analysis",
"methods"
] | 2014 | Noncanonical Role for the Host Vps4 AAA+ ATPase ESCRT Protein in the Formation of Tomato Bushy Stunt Virus Replicase |
One of the fundamental questions in biology is how cooperative and altruistic behaviors evolved . The majority of studies seeking to identify the genes regulating these behaviors have been performed in systems where behavioral and physiological differences are relatively fixed , such as in the honey bee . During colony founding in the monogyne ( one queen per colony ) social form of the fire ant Solenopsis invicta , newly-mated queens may start new colonies either individually ( haplometrosis ) or in groups ( pleometrosis ) . However , only one queen ( the “winner” ) in pleometrotic associations survives and takes the lead of the young colony while the others ( the “losers” ) are executed . Thus , colony founding in fire ants provides an excellent system in which to examine the genes underpinning cooperative behavior and how the social environment shapes the expression of these genes . We developed a new whole genome microarray platform for S . invicta to characterize the gene expression patterns associated with colony founding behavior . First , we compared haplometrotic queens , pleometrotic winners and pleometrotic losers . Second , we manipulated pleometrotic couples in order to switch or maintain the social ranks of the two cofoundresses . Haplometrotic and pleometrotic queens differed in the expression of genes involved in stress response , aging , immunity , reproduction and lipid biosynthesis . Smaller sets of genes were differentially expressed between winners and losers . In the second experiment , switching social rank had a much greater impact on gene expression patterns than the initial/final rank . Expression differences for several candidate genes involved in key biological processes were confirmed using qRT-PCR . Our findings indicate that , in S . invicta , social environment plays a major role in the determination of the patterns of gene expression , while the queen's physiological state is secondary . These results highlight the powerful influence of social environment on regulation of the genomic state , physiology and ultimately , social behavior of animals .
Behavior is a complex phenotypic trait , which results from the interactions of multiple intrinsic and extrinsic factors that associate in a nonlinear , often unpredictable fashion [1] . Intrinsic factors include the genetics , the physiology or the phenotype of an organism , while the most typical extrinsic factor is the external environment . In social systems like insect societies , environmental cues primarily are the result of the social environment , i . e . the nature and patterns of interactions among individuals within the colony [2] . The “nature-versus-nurture” debate has long been the major driver of the discussion as to whether internal state of an animal or the external environment ( e . g . , the social environment ) regulates gene expression more [3] . Regardless , extrinsic and intrinsic factors clearly are reciprocally interconnected: the social environment influences the neurogenomic state of the animal , which is responsible for the social behavior performed [4] , [5] . A hallmark of advanced social behavior is altruistic behavior , which is achieved through a reproductive division of labor in which few individuals develop into the reproductive caste while most of the colony members become non-reproductive workers and perform all tasks related to colony maintenance and growth . Both fixed ( developmental pathways ) and plastic ( behavioral strategies ) factors contribute to this division of labor ( reviewed in [6] ) . Consequently , there has been great interest in studying genes and biological processes that regulate the reproductive and worker divisions of labor [7] . In the advanced eusocial systems examined thus far , differences between queens and workers are largely the result of developmental factors , while differences among workers are often triggered by social signals [8] . However , primitively social systems display reproductive division of labor between females that are anatomically , physiologically and genetically very similar and this reproductive division of labor seems to be primarily established and maintained by social environment . The genes underlying this process have not yet been examined , and potentially may function as core genes associated with sociality . Variation in colony founding among ant queens is an ideal model to examine the interplay between genes and social environment that has shaped the evolution of cooperative behavior in primitively social systems . Colony founding can occur in two modalities: haplometrosis , where a single queen independently starts a new colony , and pleometrosis , where multiple queens associate and cooperate to start a new colony [9] . Pleometrosis is a fascinating example of cooperative behavior that is not fostered by kin selection , because these groups often comprise unrelated individuals ( reviewed [10] ) . Among social insects , pleometrosis exists in halictine bees [11] , termites [12] , paper wasps [13] , [14] , [15] and in several species of ants [8] . In ants , pleometrosis is known to be associated with division of labor in the leaf-cutter ant Acromyrmex versicolor [16] and in the harvester ant Pogonomyrmex californicus [17] . Pleometrotic associations produce a complex social environment , where individuals simultaneously are in cooperation and conflict , and social and reproductive dominance hierarchies are established . These associations represent relatively primitive social systems in which individuals with equivalent anatomical and physiological features develop a division of labor through their behavioral interactions . Thus , identification of the genes underlying establishment of these hierarchies will not only provide insight into the effects of social environment on an individual , but also into the evolution of social behavior . The red imported fire ant Solenopsis invicta is an excellent system for studying the genes associated with haplometrotic and pleometrotic behaviors , because queens from the monogyne social form ( characterized by a single egg-laying queen per nest once established ) can adopt either approach for colony founding depending on multiple factors , e . g . the density of newly mated queens in nesting sites [18] . However , monogyne fire ants ultimately only tolerate a single reproductive queen such that the initial cooperation among unrelated pleometrotic cofoundresses slowly transitions to competition and rivalry , which will inevitably produce only one winner and one or multiple losers [19] , [20] , [21] . Once the first workers emerge , pleometrotic queens engage in open fights where they injure or kill rival cofoundresses and workers actively participate in this process until all the queens are executed but one ( see Movie S1 ) . In both haplometrosis and pleometrosis , founding queens initially face a critical period ( claustral period ) where they are sealed in their nest and must defend it from enemies and competitors , e . g . other fire ant colonies that populate the same area [22] , [23] . During the claustral period , fire ant queens rely exclusively on their body mass reserves to produce the first generation of workers . There are physiological and behavioral differences between haplometrotic queens , pleometrotic “winners” and pleometrotic “losers” . Haplometrotic queens lose more weight during the claustral period , and produce more brood per individual than queens in pleometrotic associations [24] , [25] . In pleometrotic associations , winners tend to have larger head size , lose less weight [20] , [21] , [26] and occupy the top of the brood pile while losers are usually found outside the nest chamber [19] , attempting to avoid any interaction with the winner or with workers . However , nothing is known about the genes and molecular pathways that underlie these processes . We performed two separate experiments to characterize the genomic basis for haplometrotic and pleometrotic founding behavior in fire ants . We developed a microarray platform using the official gene set of the fire ant genome [27] plus a set of ESTs obtained from assemblies of the fire ant transcriptome to examine genome-wide expression patterns across founding queens . In the first experiment , we compared whole body gene expression patterns among haplometrotic queens and paired pleometrotic winners and losers that were collected shortly after emergence of the first workers ( but prior to execution of the loser ) . We predicted that haplometrotic queens would be more similar to pleometrotic winners than to pleometrotic losers , because they both will serve as the single queen for the mature colony . For the comparison between winners and losers alone , our expectations were less well-defined: on one hand , we expected to find substantial differences given that their physiology , behavior and fate differ significantly , but on the other hand , winners and losers are not anatomically distinct and there is only weak correlation between morphological measures and outcome of the conflict [20] . In the second experiment , we manipulated queen rank in pleometrotic pairs to determine how changing social rank and social environment would affect an individual's gene expression patterns . This was accomplished by pairing pleometrotic queens with a new partner at the end of the claustral period in order to switch putative winners to losers and vice-versa . Controls in which partners were altered ( and social environment was changed ) but social rank remained the same were also included . We hypothesized that final social rank would be the primary regulator of gene expression patterns . However , for both experiments our results indicated that social environment ( pleiometrosis vs . haplometrosis , switched rank vs . maintained rank ) was a much greater driver of gene expression changes than social rank itself , suggesting that social environment , and not reproductive state , is a key regulator of gene expression , physiology and ultimately , behavior .
Haplometrotic queens ( haplo ) and paired pleometrotic winners ( win ) and losers ( los ) were collected shortly after emergence of the first workers ( see methods , N = 8 haplo , 8 win and 8 los ) . Microarray analysis of gene expression patterns ( see Methods for design and validation of microarrays ) in whole bodies of these queens revealed that 4080 of the 9388 transcripts included in the analysis were differentially regulated at FDR<0 . 001 ( Table S1 ) . A principal components analysis ( PCA ) of the differentially regulated transcripts revealed that the social environment is more important than social rank in driving the patterns of gene expression in founding queens ( Figure 1A ) . Differences between haplometrotic and pleometrotic queens accounted for 91 . 8% of the variation in gene expression while differences between win and los accounted for only 8 . 2% . Pairwise comparisons of transcripts differentially regulated ( FDR<0 . 001 ) among the three groups of fire ant queens demonstrated that expression patterns in haplo are more similar to win than to los , since there are fewer genes differentially regulated uniquely between haplo and win ( 404 ) than haplo and los ( 477; Nominal Logistic Fit: df = 1 , ChiSquare = 6 . 78 , P = 0 . 0092; Figure 1B ) . We performed Gene Ontology analysis on the 3003 differentially regulated transcripts ( out of the initial pool of 4080 ) that have Drosophila orthologs with FlyBase annotations using DAVID [28] , [29] . 517 GO terms were significantly enriched at a p-value<0 . 05 ( Functional Annotation Chart , see Table S2 for the complete list of GO terms ) . Additionally , 6 KEGG molecular pathways ( Kyoto Encyclopedia of Genes and Genomes , [30] ) were significantly enriched ( P<0 . 05 ) : aminoacyl-tRNA biosynthesis , basal transcription factors , dorso-ventral axis formation , endocytosis , RNA degradation and ubiquitin mediated proteolysis ( Table S2 ) . To cluster the GO categories into larger functional groups , the 517 significantly enriched GO terms were mapped to the GO_slim2 file in CateGOrizer [31]: 440 GO terms were assigned to one of the ancestor terms by single count ( Figure 2 ) . The functional groups containing the greatest number of GO terms were metabolism ( 19% of the significantly enriched GO terms ) , cell organization and biogenesis ( 11% ) and development ( 10%; for a complete list of all ancestor terms represented in this analysis see Table S3 ) . To further characterize the genes that were differentially regulated between haplo and pleometrotic queens ( pleo ) , we examined the overlapping set of 3192 transcripts ( of which 2541 had Drosophila orthologs with FlyBase annotations ) that were differentially regulated between both haplo vs . win and haplo vs . los ( Figure 1B ) . For clearer graphical presentation , we used k-means Clustering in GENESIS [32] to separate these transcripts into two large clusters according to expression patterns: 2280 transcripts that were upregulated in haplo and downregulated in pleo ( cluster 1 , Figure S1 ) and 912 transcripts downregulated in haplo and upregulated in pleo ( cluster 2 , Figure S2 ) . We performed GO analysis on both groups , using Functional Annotation Clustering with medium stringency . For cluster 1 ( 1925 FlyBase matches ) , we obtained 88 significantly enriched GO terms ( see Table S4; P<0 . 05 ) and 1 KEGG pathway ( basal transcription factors , P = 0 . 01 ) . Several of the GO terms were related to aging ( determination of adult life span , death ) , immunity ( immune system development , JNK cascade , hemopoiesis ) , reproduction ( reproductive developmental process , oocyte development , eggshell formation , morphogenesis of follicular epithelium , regulation of oocyte development ) , response to stimuli ( response to stress , regulation of response to stimulus , negative regulation of response to stimulus , response to ecdysone ) , lipid biosynthetic process , locomotion and neurological system processes ( neurotransmitter secretion , neurogenesis , central nervous system development , regulation of nervous system development ) . In cluster 2 ( 616 FlyBase matches ) , 34 GO terms ( Table S5 , P<0 . 05 ) and 1 KEGG pathway ( glycerophospholipid metabolism , P = 0 . 01 ) were significantly enriched . Many GO terms were similar to those in cluster 1 , and included determination of adult life span , olfactory behavior , lipid metabolic process , detection of light stimulus , as well as several related to morphogenesis or development of organs and apparatuses like sensory organ , muscle , limb , wing disc , gut and respiratory system . Interestingly , only 43 transcripts were differentially regulated between win and los queens . A GO analysis performed on this small set of transcripts revealed that fatty acid and hormone metabolic processes were significantly enriched GO terms ( Functional Annotation Clustering , P<0 . 001 and P<0 . 01 , respectively , Table S6 ) . Several transcripts in this group have interesting functions ( Table 1 ) . Transcripts upregulated in win included: G protein-coupled receptor kinase 2 ( Gprk2 ) , which is involved in the Toll signaling pathway during the response against Gram positive bacteria [33]; endosulfine ( endos ) , which functions in the insulin-signaling pathway during oogenesis [34]; Pheromone-binding protein-related protein 3 ( Pbprp3 ) , a member of the odorant binding proteins responsible for chemoreception [35]; and bubblegum ( bgm ) , which is involved in the metabolism of very long-chain fatty acids and prevents neurodegeneration [36] . Transcripts upregulated in los relative to win included: I'm not dead yet ( Indy ) , associated with aging [37]; pale ( ple ) , which plays a role in the response to wounding [38] and in the metabolism of dopamine [39]; desat1 , a major regulator of cuticular hydrocarbon biosynthesis involved in pheromone emission and detection [40]; and juvenile hormone acid methyltransferase ( jhamt ) , a key enzyme in the biosynthesis of JH , the major endocrine regulator in insects [41] . GO categories related to aging and longevity were significantly enriched in sets of transcripts that were differentially regulated in haplo vs . pleo ( clusters 1 and 2; see Tables S4 and S5 ) . Out of the 129 genes included in the Drosophila aging GO term , oligos representing 93 putative orthologs were present on the fire ant microarray . Of these , 90 were expressed at high enough levels to be included in the microarray analysis and 46 were significantly differentially regulated across the three groups of queens ( Figure 3A ) . The majority of these genes ( 34 ) were upregulated in haplo: this number was significantly higher than expected by chance ( Nominal Logistic Fit: df = 2 , ChiSquare = 29 . 58 , P<0 . 0001 ) . In addition to their role in regulating aging pathways in Drosophila , several genes in this group have been linked to queen vs . worker caste differentiation and behavioral maturation in honey bees [42] . These include: forkhead box ( foxo ) and target of rapamycin ( TOR ) , two major players in the insulin-signaling pathway which is associated to caste determination in honey bees [43] and to the workers' transition from nursing to foraging behavior [44] , [45] , and Peroxiredoxin 5037 ( Prx3 ) , which is associated with enhanced learning ability when expressed at higher levels in honey bee workers [46] . To further investigate the patterns of expression of aging genes in haplo and pleo queens , we compared our study to a study on aging in Drosophila [47] . In this study , which was aimed at investigating the temporal and spatial ( tissue-specific ) transcriptional profiles in Drosophila , the authors listed all the age-related GO terms that were significantly enriched and classified them based on the tissue where they were expressed and on their directional expression . We found that 106 GO terms were upregulated in fire ant haplo queens and old Drosophila , while only 36 were shared between old flies and pleo: however , the difference was not statistically significant ( Fisher's Exact Test: P = 0 . 67 ) . When we compared downregulated GO terms , we found that 11 were shared between haplo and old flies while 67 were shared between old flies and pleo: the difference was statistically significant ( Fisher's Exact Test: P = 0 . 0029 ) . Most of these 67 overlapping GO terms ( see Table S7 for details ) encompassed genes that were regulated in the gut ( 32 ) and fat bodies ( 23 ) , followed by brain ( 13 ) , muscles ( 11 ) , malpighian tubules ( 7 ) and accessory glands ( 6 ) . Immune-related GO terms were significantly enriched in cluster 1 ( genes upregulated in haplo vs . pleo , see Table S4 ) . To better examine the overall expression profiles of genes involved in immune pathways , we obtained a list of significantly enriched GO terms for both cluster 1 and cluster 2 ( Functional Annotation Chart in DAVID , P<0 . 05 , see Tables S8 and S9 ) and we mapped these lists to the list of “Immune system gene classes” available on the CateGOrizer website ( http://www . animalgenome . org/bioinfo/tools/catego/slims . html ) . Thereafter , we compared the relative proportions of the parent/ancestor immune categories between the two groups ( Figure 4 ) . This analysis confirmed a significant overrepresentation of immune-related classes in cluster 1 relative to cluster 2 ( Nominal Logistic Fit: df = 1 , ChiSquare = 61 . 16 , P<0 . 0001 ) , clearly visible in terms of total number of immune categories present and number of GO terms within common categories . These results suggest that haplometrotic queens overall have higher expression levels of immune-related genes and therefore may be better equipped in terms of immune response . Next , we examined the expression of the fire ant orthologs of the 177 canonical immune-related genes annotated in honey bees [48] . Orthologs for 83 of these genes were included in our array; 82 were expressed at high enough levels to be included in the analysis , and 34 were within our list of 3003 significantly differentially regulated transcripts ( Figure 3B ) . Expression levels of these genes are not strongly coordinated , with similar numbers of up- vs . downregulated genes in haplo vs . pleo queens . Several genes in the Immune-deficiency ( IMD ) pathway were differentially regulated , including Inhibitor of apoptosis 2 ( Iap2 ) , TGF-beta activated kinase 1 ( Tak1 ) , immune deficiency ( imd ) , bendless ( ben ) and Death related ced-3/Nedd2-like protein ( Dredd ) . Furthermore , several members of the Immunoglobulin ( IG ) Superfamily were differentially regulated , including bent ( bt ) , turtle ( tutl ) , sidekick ( sdk ) and Down syndrome cell adhesion molecule ( Dscam ) . We examined gene expression levels of candidate genes that were included in one or more GO terms that were significantly enriched in our GO analyses ( Figure S3 and Table S10 ) . Expression patterns of all 13 candidate genes were consistent with what observed for haplo and los in the arrays , and these expression differences were significant for 11 genes . We validated Indy and Sod2 for determination of adult life span . In the arrays , Indy was downregulated in haplo and Sod2 was upregulated in haplo: qRT-PCR analysis confirmed these trends and the difference between groups was statistically significant for both genes ( P<0 . 001 ) . For immune response , we validated Dredd and kay , which were both upregulated in haplo in the arrays: qRT-PCR analysis confirmed this trend and the difference between the two groups of queens was statistically significant for kay ( P<0 . 05 ) but not for Dredd ( P = 0 . 29 ) . Desat1 , ifc and putative fatty acyl-CoA reductase CG5065 were analyzed for their involvement in the synthesis and metabolism of fatty acids . In the arrays , desat1 and putative fatty acyl-CoA reductase CG5065 were downregulated in haplo while ifc was upregulated in haplo: these trends were confirmed by qRT-PCR analysis and the difference in the expression levels was statistically significant for putative fatty acyl-CoA reductase CG5065 ( P<0 . 001 ) and ifc ( P<0 . 05 ) but not for desat1 ( P = 0 . 11 ) . Reproductive genes included br and Btk29A , both upregulated in haplo in the arrays: this was confirmed by qRT-PCR ( P<0 . 001 ) . We validated Sema-5c and Mer because they play a role in olfactory behavior: these genes were downregulated in haplo in the arrays and in the qRT-PCR analysis ( P<0 . 001 and P<0 . 05 , respectively ) . Finally , we analyzed fru for aggressive behavior and woc for neurogenesis: these genes were upregulated in haplo in the arrays and in the qRT-PCR analysis ( P<0 . 05 and P<0 . 001 , respectively ) . We further examined the role of social rank on gene expression patterns by manipulating social rank of individuals in pleometrotic pairs . We swapped winners and losers between groups to generate four groups of queens: winners switched to losers ( win/los ) , losers switched to winners ( los/win ) , continuing winners ( win/win ) and continuing losers ( los/los ) . Very few transcripts were differentially regulated among these groups , with a total of 616 transcripts at a relatively low significant threshold ( FDR<0 . 1 , see Table S11 for a listing of these transcripts ) . Principal components analysis demonstrated that 48% of the variation in gene expression was associated with switching social rank ( win/los and los/win were clustered relative to win/win and los/los ) , 37% of the variation was associated with the final rank ( i . e . , win/los and los/los were clustered ) , while 15% was associated with the initial rank ( i . e . win/win and win/los were clustered; Figure 5 ) . We performed GO analysis with Functional Annotation Clustering on the 527 differentially regulated transcripts that have Drosophila orthologs with FlyBase annotations . 21 GO terms were significantly enriched at p-value<0 . 05 and three survived Benjamini correction: cellular metabolic process , cellular ketone metabolic process and maintenance of protein location ( see Table S12 ) . Among the other GO terms , of particular interest was lipid metabolic process , which includes several genes involved in the metabolism of fatty acids such as Helix loop helix protein 106 ( HLH106 ) [49] , scully ( scu ) [50] and two putative fatty acyl-CoA reductases . Additional genes with significant differences in expression included Coenzyme Q biosynthesis protein 2 ( Coq2 ) , which plays a role in the response to pathogens , aging and in the insulin-signaling pathway [51] , [52] , juvenile hormone acid methyltransferase ( jhamt ) , which was also significantly differentially regulated between win and los in experiment 1 ( see above ) , and radish ( rad ) , which is involved in learning and memory [53] . Finally , the GO term “response to stress” was significantly overrepresented , which includes key immune response genes such as immune response deficient 5 ( ird5 ) [54] , Ras-related protein Rac1 ( Rac1 ) [55] , Hemolectin ( Hml ) [56] , Argonaute 2 ( AGO2 ) [57] and caspar ( casp ) [58] . Ninety-three transcripts were differentially regulated both between win and los in experiments 1 ( 548 transcripts , FDR<0 . 1 ) and in experiment 2 ( 616 transcripts , FDR<0 . 1 ) : this is significantly more than expected by chance ( Hypergeometric Test: Representation factor: 2 . 2 , p<1 . 009−13 ) . These transcripts corresponded to 80 Drosophila orthologs , which were used to perform a GO analysis using Functional Annotation Clustering: 6 GO terms appeared to be significantly enriched , including lipid metabolic process ( P<0 . 001 ) and regulation of hormone levels ( P<0 . 05; Table S13 ) . The expression patterns of the 9 differentially regulated transcripts involved in lipid metabolic process across the two experiments are shown in Table 2 .
Differences in gene expression between haplometrotic and pleometrotic queens were likely due to differences in the physiological demands placed on singly- vs . multiply-founding queens and differences in the costs associated with social environment , where pleometrotic queens are more likely to incur in higher levels of stress due to the establishment of social ranks . We found that genes involved in core physiological processes , including metabolism , cellular processes , development , morphogenesis and biosynthesis were significantly differentially regulated between these groups of queens . Haplometrotic queens produce more eggs and lose more weight than pleometrotic queens during the claustral period of colony founding [20]: this seems to be due to queen-queen reciprocal reproductive inhibition and oophagy/cannibalism of larvae in pleometrotic associations [62] . Genes associated with reproductive functions ( including development of reproductive tissues and production of oocytes and eggs ) were upregulated in haplometrotic queens . Furthermore , in order to produce eggs , newly mated queens degrade wing muscle tissues and metabolize fat body storage proteins to produce free amino acids [63] . We found 58 protein-related GO terms and 10 amino acid-related that were upregulated in haplometrotic queens versus 5 and 0 , respectively that were upregulated in pleometrotic queens ( Functional Annotation Chart , see Tables S8 and S9 ) . Genes associated with stress response were differentially regulated between haplo and pleo queens . Stress tolerance may be achieved by reducing the production of reactive oxidant species ( via improved regulation of mitochondrial processes ) and/or by increasing the antioxidant activity [64] , [65] . In our study , we found that two mitochondria-related GO terms , namely mitochondrial electron transport , NADH to ubiquinone ( 15 genes ) and mitochondrion organization ( 18 genes ) were upregulated in haplo and none in pleo . Moreover , 9 antioxidant genes were upregulated in haplo , including two superoxide dismutases ( Sod2 and CCS ) , two Peroxiredoxins ( 6005 and 5037 ) , Glutathione S transferase S1 and PTEN-induced putative kinase 1 ( Pink1 ) , which plays an essential role in maintaining neuronal survival by preventing neurons from undergoing oxidative stress [66] . These results suggest that haplo queens may experience lower levels of oxidative stress either by producing less ROS or by keeping the levels of antioxidants high . Higher stress levels in pleo queens could be correlated to their social environment , dominated by queen-queen aggressive interactions and competition . Stress tolerance is positively correlated with lifespan [67] and this trait has been used as a proxy for long-lived phenotypes in studies that examine the genetic basis of lifespan [68] . Only SOD was upregulated in pleo . Interestingly , overexpression of SOD has been correlated to increased organismal longevity in Drosophila [69] , but this was not confirmed in Lasius niger , where long-lived queens expressed lower levels of this gene than short-lived males and workers [70] . It is evident that the effect of SOD on longevity is highly dependent upon the sex and genetic background [71] and also the social environment [72] . The overrepresentation of GO terms associated to biosynthesis and metabolism ( in particular those related to lipids ) prompted us to look closer at the nutritional state of founding queens . Nutrition is closely linked to fertility and longevity [73] . In insects , the insulin-signaling pathway regulates nutrient-sensing [74] while juvenile hormone and ecdysone mediate reproductive processes [75] . In honey bees , long-lived queens have low levels of insulin and juvenile hormone , while they have high levels of FOXO , vitellogenin and ecdysone; opposite patterns are found in sterile short-lived workers [76] . Our results show that haplo had higher levels of FOXO and of the ecdysone receptor . Haplo queens also presumably had lower levels of JH , since levels of juvenile hormone acid methyltransferase ( jhamt ) , an enzyme that converts inactive precursors of JHs to active JHs [41] , were downregulated , and levels of juvenile hormone epoxide hydrolase 2 ( Jheh2 ) , involved in juvenile hormone catabolic process , were upregulated . There is no clear prediction about which group of queens should have a longer life-span . Our analyses show that a large set of aging-related GO terms was upregulated in haplometrotic queens , while a smaller set was upregulated in pleometrotic queens . This result is not sufficient to establish which group of queens is expected to have longer life-span , since ageing is a quantitative trait determined by both environmental and genetic components . Previous studies of the genetics of longevity in Drosophila melanogaster , identified sets of genes in which upregulated expression either accelerates or decelerates the aging process [65] . However , in our study , genes from both categories were equally up- and downregulated across haplo and pleo queens ( see Figure 3A ) . Therefore , the knowledge of the genetics of longevity in the insect model D . melanogaster cannot be transferred directly to our study system . Immune-related genes were overexpressed in haplometrotic vs . pleometrotic queens . Most of the overrepresented immune-related GO terms were associated to cellular immunity: endocytosis , phagocytosis , cell adhesion , apoptosis , cytokinesis , the cascade regulating mitogen-activated protein kinase ( MAPKKK ) and the c-Jun amino-terminal protein kinase ( JNK ) cascade . In particular , the JNK pathway controls the rapid up-regulation of cytoskeletal genes in response to infection and plays a major role in wound healing [77] . Key genes in the JNK pathways [78] were upregulated in haplo , namely kayak , hemipterous , misshapen , anterior open and Cdc42 . Hemopoiesis is the process that is responsible for production and differentiation of immune cells [79]: two key genes involved in this process , serrate and serpent , were upregulated in haplo . Haplo queens may have better constitutive immune responses perhaps because they experience less social stress than pleo queens do: in fact , once initial cooperation transitions into open competition , pleo queens frequently engage in reciprocal aggressions which can lead to serious injuries or death . It is hypothesized that there is a trade-off between reproduction , nutrition and immunity [80] , suggesting that highly reproductive haplo queens should have overall reduced immune responses during colony foundation period when food sources are limited . However , previous studies in honey bees demonstrated that reproductive queens have higher expression of immune genes than non-reproductive workers [81] , [82] , and thus this trade-off may not exist in social insect queens , perhaps because queens have more energy resources than workers . Only 43 transcripts were significantly differentially regulated between winners and losers in couples of pleometrotic queens from experiment 1 . Although surprising , this result might be explained by the small phenotypic differences between the two types of queens . Previous studies showed that some phenotypic traits such as head width are weakly correlated with the reproductive investment and survival ( hence the rank ) of pleometrotic cofoundresses [20] . It has been suggested that the relatively weak association between these parameters stems from selection to maintain cooperation [20] . If phenotypic differences strongly correlate with the chances of surviving , smaller queens with lower fighting abilities would be selected not to cooperate and feed the brood in the colony . Thus , the small differences at the genomic level between winners and losers may reflect selection for a system where differences between cofoundresses is sufficiently small so that all of them have a chance of surviving , and thus an interest to cooperate with unrelated individuals . The two GO terms that were differentially regulated between winners and losers were related to metabolism of lipids and metabolism of hormones . Four transcripts , bubblegum , desat1 , Dmel_CG17374 and Dmel_CG31522 , which function in fatty acid metabolism , were differentially regulated . Long-chain fatty acids are the precursors of cuticular hydrocarbons in insects , which can function as nestmate recognition cues and social pheromones in many insect species ( reviewed [83] ) . Interestingly , bgm , which encodes a very long-chain fatty acid CoA ligase [36] , was downregulated in losers relative to winners and haplo: thus , this gene may be involved in regulating chemical cues related to dominance . Similarly , desat1 , which is expressed at higher levels in losers than in winners or haplo functions in pheromonal communication [40] . Altered bgm expression has also been associated with infection ( and correlated with changes in cuticular hydrocarbon profiles ) in honey bees [84] , while desat1 appears to play a role in autophagic responses [85] . Thus , these genes may also be involved in signaling infection , nutrient deprivation or other stress responses . Behavioral manipulation of the social rank in pairs of pleometrotic queens demonstrated that manipulation of social environment ( i . e . , conditions in which the social rank of the individual changes ) had a much larger effect on gene expression than the initial or final social rank of the individual . Note , however , that all individuals in the study switched social partners , which may have elicited additional ( undetected ) changes in gene expression . Studies in vertebrates have demonstrated that social interactions and changes in the social environment can be one of the most potent stressors [86] . Indeed , genes associate with ‘response to stress’ were significantly enriched , with a set of 30 transcripts differentially regulated among the four groups of manipulated queens ( see Results and Table S12 ) . The effects of restructuring social ranks have not been considered broadly in other species [87] , but decreased social rank in dark-eyed junco birds is associated with increased metabolic rates , while increased social rank results in a much lower physiological change [88] . Similarly , for dominant , but not for subordinate , birds there is a measurable metabolic cost to joining a new social group [88] . In both experiments , genes involved in lipid biosynthesis and metabolism were differentially regulated , suggesting that these processes play a key role in mediating fire ant founding behavior and foundress associations . Lipids such as cuticular hydrocarbons play a role in advertising the fertility state in many ant species: these compounds are usually more abundant in reproductive queens and egg-laying workers ( reviewed in [89] ) . Indeed , ‘lipid biosynthetic/metabolic process’ was differentially regulated in haplo vs . pleo and in win vs . los in experiment 1 ( Tables S4 , S5 and S6 ) and in experiment 2 ( Table S12 ) . These results support the hypothesis that lipids ( and in particular fatty acids ) are of great importance in regulating social interactions between queens and among nestmates in general . In fire ant pleometrotic associations , the pheromones and nestmate recognition chemicals derived from these fatty acids are most likely an important component of the individual's chemical profile , which is used by nestmate queens to decipher the physiological condition and thus the social rank of the partner . We used newly developed genomic tools to examine the gene expression patterns associated with complex social behaviors involved in colony founding by fire ant queens . Our results suggest that social environment ( haplometrotic vs . pleometrotic , switched vs . maintained social rank ) is more important than the social rank or internal condition of the individual in regulating gene expression patterns , and presumably downstream physiological and behavioral traits . Furthermore , because the process of pleometrotic colony founding in fire ants has all the features of a primitively social system in which morphologically , physiologically , and genetically similar individuals perform cooperative behavior to form social groups of unrelated individuals , this is an excellent model to examine the genes that underlie these social behaviors . We found that several core processes were significantly differentially regulated , including metabolism , stress response , aging , reproductive processes , and immunity . Interestingly , lipid metabolic processes were regulated across experiments; these may play a role in both nutrient storage/mobilization and chemical communication . In the future , it will be interesting to investigate whether the molecular pathways characterized in this study also are operating at earlier stages of the co-founding process ( e . g . , before the emergence of workers ) . Such studies will help elucidate the mechanisms responsible for the transition from cooperation to conflict in pleometrotic founding queens . Finally , fire ants also display genetically distinct monogyne ( colonies headed by a single queen ) and polygyne ( colonies headed by multiple queens ) social forms . It will be of great interest to determine if the same genes that regulate haplometrosis and pleometrosis also are involved in regulating queen number in mature colonies .
A total of 787 fire ant queens were collected immediately after a nuptial mating flight on May 5th , 2010 in a large parking lot ( Target , 3970 SW Archer Rd , Gainesville , FL ) and weighed . Since the area of collection has a high prevalence of monogyne colonies , we expected these queens to belong to the monogyne social form; this was subsequently confirmed by screening 108 queens for social form using Gp-9 as a marker following the protocol as described in [90] . The remaining 679 queens were split into two groups: 308 queens were set up in pairs ( pleometrosis ) based on having similar weights ( range ±0 . 2 mg ) and paint-marked with different colors , while 371 queens were set up individually ( haplometrosis ) and paint-marked as well . All the queens were provided with a nesting chamber consisting of a glass tube half-filled with water , which was covered by a cotton ball and a layer of dental plaster: this keeps the chamber moist but avoids an excess of water which is deleterious for the young brood . Tubes were sealed with a loose cap to provide air flow . Specimens were reared in the dark at 28°C , 70% relative humidity under claustral conditions ( no food and no water ) for 1 month . After eclosion of the first batch of workers ( minims ) , incipient colonies were provided with water , sugar water and frozen crickets . Glass tubes were set open in pencil boxes coated with Fluon to prevent escape . Queens were subsequently monitored daily until it was possible to identify the social rank of the two cofoundresses in pleometrotic couples . Previous studies have found that the initial cooperation between the two cofoundresses turns into conflict after the emergence of minims , resulting in the execution of one queen [19] . Queens that survive the competition ( winners ) are usually located at the top of the brood pile within the nest chamber and they are generally tended by workers; conversely , queens that will be executed ( losers ) are normally seen outside the nest chamber , hiding from workers in order to avoid being attacked ( Figure S4 ) . We used these observations to establish the social rank of the two pleometrotic queens , i . e . winner and loser . We collected 25 pleometrotic couples and 25 haplometrotic queens in dry ice and stored them at −80°C to be later processed . This assay was performed with 34 couples of pleometrotic queens from the same pool of newly mated queens as experiment 1 . The queens were paired and placed in nesting chambers as before . After emergence of minims , queens' behavior was monitored as before . Once the behavioral observation revealed the social rank of the two cofoundresses , queens were weighed again and re-paired with a different partner . We created the following three groups of queens: a ) winner+winner ( similar weight ) , b ) loser+loser ( similar weight ) , and c ) winner+loser ( different weights ) . Again , we monitored the behavior until the social rank of the newly coupled specimens was evident and we collected 4 new behavioral phenotypes in the same way as above: a ) winners switched into losers ( win/los , N = 7 ) , b ) losers switched into winners ( los/win , N = 11 ) , c ) continuing winners ( win/win , N = 12 ) and d ) continuing losers ( los/los , N = 5 ) . Individual fire ant queens were thawed and dissected under cold RNAlater ( Qiagen , Valencia , CA ) to confirm the mating status: unmated queens were not included in the analysis . Total RNA was extracted using the RNeasy Plus kit ( Qiagen ) combined with a RNase-Free DNase step ( Qiagen ) to remove any possible contamination by genomic DNA . Subsequent steps in the microarray analysis were performed at the Penn State Genomic Core Facility . RNA concentration and purity were assessed using NanoDrop and Qubit and RNA quality was assessed using RNA Nano Chips on the Agilent Bioanalyzer . 1 µg of each sample was amplified using the Ambion ( Life Technologies ) Amino Allyl MessageAmp II aRNA Amplification Kit ( AM1753 ) . 15 µg of aRNA were dyed with Cy3 or Cy5 ( GE Health Care #RPN5661 ) and subsequently purified according to the Ambion Kit instructions . 1 . 5 µg of a Cy3 labeled sample were combined with 1 . 5 µg of a Cy5 labeled sample and fragmented using RNA Fragmentation Reagents ( Ambion AM8740 ) according to the manufacturer's instructions . Samples were hybridized with mixing in a MAUI hybridization instrument overnight at 42°C . Arrays were scanned using Axon GenePix 4000B . For the first microarray developed to validate the efficiency of probe sequences , we pooled RNA samples ( 2 µg total ) from different castes , developmental instars and social forms as follows: 3 female alates , 15 workers , 5 larvae and 5 pupae from both monogyne and polygyne social forms and 5 males from monogyne colonies only . The fire ant genome includes an official gene set of 16 , 569 protein-coding genes that were generated by a combination of ab initio , EST-based , and sequence similarity-based methods [27] . For our microarray studies , we combined the official gene set with a set of ESTs obtained from assemblies of the fire ant transcriptome for a total set of 63 , 436 sequences ( “transcripts” ) . We successfully designed 60-mer probes for 51 , 531 of these transcripts ( Roche NimbleGen , Inc . , Madison WI ) . These sequences/probes were grouped into four categories: ESTs with gene models ( EWGM , 7433 transcripts ) , ESTs without gene models ( EWOGM , 40 , 613 ) , gene models ( GM , 3246 ) and gene models redundant with other models ( GMRWOM , 239 ) . We developed and used a first microarray ( 1-plex 385 , 000 probe capacity , Roche NimbleGen , Inc . , Madison WI ) to validate the probe design and test multiple probes per transcript . On average , we designed 7 probes per transcript for a total of 355 , 930 probes . Each probe was tested for both the red ( Cy5 ) and the green ( Cy3 ) dyes . For transcripts with only one probe ( N = 296 ) , we verified that the probe had acceptable intensities for both dyes . For the other transcripts we examined the performance of the probes with the green dye only , because these showed consistently higher intensity compared to the red dye . Probes were ranked in the follow manner: a ) if there were only 2 probes per transcript ( N = 230 ) , we selected the one with higher intensity; b ) if there were 3 to 6 probes ( N = 744 ) , we calculated the ratio “probe intensity/median intensity of all probes for that transcript” and selected the probe with highest ratio if the value was <3 , otherwise we selected the probe with the second highest ratio; c ) for transcripts with 7 probes ( N = 50 , 261 ) , we followed the procedure as in “b” but , in case the probe with the highest ratio was >3 , we removed that probe , calculated new ratios and selected a new probe with highest ratio . This procedure allowed us to select the probes with highest intensity that were not outliers . Selected probes were printed in pairs on two 12-plex microarrays ( each array had a 135 , 000 probe capacity , Roche NimbleGen , Inc . , Madison WI ) . We used a loop design with dye swaps incorporated , allowing us to hybridize 24 RNA samples to each array . For experiment 1 we hybridized 8 haplometrotic queens , 8 pleometrotic winners and 8 pleometrotic losers ( Figure S5 ) and for experiment 2 we compared 6 win/los , 6 los/win , 6 win/win and 5 los/los ( Figure S6 ) . Any spots with an intensity of less than 300 ( the background level on the arrays ) were removed from the analyses , as were spots present on less than 20 out of 24 arrays . Expression data were log-transformed and normalized using mixed-model normalization ( proc MIXED , SAS , Cary , NC ) with the following model:where Y is expression , dye and block are fixed effects , and array , array*dye and array*block are random effects . Transcripts with significant expression differences between groups were detected by using a mixed-model ANOVA with the model:where Y represents the residual from the previous model . Treatment , spot and dye are fixed effects and array is a random effect . P-values were corrected for multiple testing using a false discovery rate of <0 . 001 for experiment 1 and <0 . 1 for experiment 2 ( proc MULTTEST , SAS ) . Because the number of differentially regulated transcripts for experiment 1 was very high ( ∼13 , 000 out of ∼50 , 000 ) , and to avoid an excess of redundancy among the different groups of transcripts , we included only probes corresponding to GM and EWGM ( see above ) . Hierarchical clustering , using the Ward method , and principal component analysis ( PCA ) for global patterns of gene expression were performed in JMP 9 . 0 . 2 ( SAS , Cary , NC ) . We used Genesis 1 . 7 . 6 ( Graz , Austria ) to cluster differentially regulated genes based on average linkage and to perform k-means clustering in experiment 1 . Gene Ontology analysis was performed using functional annotation chart/clustering in DAVID version 6 [28] , [29] using DAVID default population background and a cutoff of p<0 . 05 . For all Gene Ontology ( GO ) analyses , fire ant genes were matched to their Drosophila orthologs in FlyBase ( http://flybase . org/ ) . CateGOrizer [31] was used to count the occurrences of significantly enriched GO terms within each of the pre-defined set of parent/ancestor GO terms . The array data were deposited on the ArrayExpress website according to MIAME standards ( ArrayExpress accession: E-MEXP-3886 for experiment 1 , E-MEXP-3898 for experiment 2 ) . We compared the results from experiment 1 to the following studies: We performed overlaps between list of transcripts and GO terms with Venny [91] . In the first comparative study we overlapped fire ant transcripts directly , while in the second study we used Drosophila orthologues ( FlyBase numbers ) to compare fire ant transcripts to the genes of the fruit fly . Statistical significance of the overlap was calculated using a hypergeometric test ( http://nemates . org/MA/progs/overlap_stats . html ) . Selected GO analyses based on study overlap were performed in DAVID as above . In the second study , to test for the significant agreement in the patterns of expression between two studies we performed Fisher's Exact Tests in JMP . We examined gene expression levels of the following candidate genes ( Table S10 ) : Indy and Sod2 ( determination of adult life span ) ; Dredd and kay ( immune response ) ; desat1 , ifc and Putative fatty acyl-CoA reductase CG5065 ( synthesis and metabolism of fatty acids ) ; br and Btk29A ( reproductive functions ) ; Sema-5c and Mer ( olfactory behavior ) ; fru ( aggressive behavior ) and woc ( neurogenesis ) . We used the total RNA extracted from fire ant queens for the microarray analysis and compared gene expression between haplo and los on an ABI Prism 7900 sequence detector ( Applied Biosystems , Foster City , CA , USA ) . cDNA was made using SuperScript III First-Strand Synthesis System for RT-PCR ( Invitrogen-Life Technologies , Carlsbad , CA , USA ) and Random Hexamers according to the manufacturer's protocol . The cDNA was then diluted 2 ( x ) with ultra-pure water . Amplification was performed in a 10 µl reaction mixture containing 5 µl of 2× SYBR Green Master Mix ( Applied Biosystems-Life Technologies , Carlsbad , CA , USA ) , 1 µl of each primer ( 10 µM ) and 2 µl of cDNA at the following conditions: 50°C for 2 min , 95°C for 10 min , 40 cycles of 95°C for 15 sec and 60°C for 1 min , a dissociation step of 95°C for 15 sec and 60°C for 15 sec . We used 8 queens per group: triplicate reactions were performed for each of the samples and averaged for use in statistical analysis . Expression levels of candidate genes were normalized to the geometric mean of two housekeeping genes , Rp-9 and Rp-37 [27] . Negative control ( cDNA reaction without RT enzyme ) was also used . Primer sequences were developed in Primer3Plus ( http://www . bioinformatics . nl/cgi-bin/primer3plus/primer3plus . cgi ) and primer efficiency was first validated using standard curves . Statistical analysis was performed with nonparametric Kruskall-Wallis rank sums in JMP 10 ( SAS , Cary , NC ) . The data were shown normalized to the haplo group . | The characterization of the genomic basis for complex behaviors is one of the major goals of biological research . The genomic state of an individual results from the interplay between its internal condition ( the “nature” ) and the external environment ( the “nurture” ) , which may include the social environment . Colony founding in the fire ant Solenopsis invicta is a complex process that serves as a useful model for investigating how the interplay between genes and social environment shapes social behavior . Unrelated , newly mated S . invicta queens may start a new colony as a group , but ultimately only one queen will survive and gain full reproductive dominance . By uncovering the genetic basis for founding behavior in fire ants we therefore provide useful insights into how cooperative behavior evolved in a context that might be considered primitively eusocial , because newly mated queens in a founding association are morphologically , physiologically and genetically very similar and display no evident division of labor . Our results suggest that social environment ( founding singly or in pairs , switching dominance rank vs . maintaining rank ) is a much greater driver of gene expression changes than social rank itself , suggesting that social environment , and not reproductive state , is a key regulator of gene expression , physiology and ultimately , behavior . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genome-wide",
"association",
"studies",
"genome",
"expression",
"analysis",
"evolutionary",
"ecology",
"functional",
"genomics",
"genome",
"evolution",
"integrated",
"control",
"gene",
"function",
"genome",
"analysis",
"tools",
"chemical",
"ecology",
"genome",
"databases... | 2013 | Sociogenomics of Cooperation and Conflict during Colony Founding in the Fire Ant Solenopsis invicta |
An approach combining genetic , proteomic , computational , and physiological analysis was used to define a protein network that regulates fat storage in budding yeast ( Saccharomyces cerevisiae ) . A computational analysis of this network shows that it is not scale-free , and is best approximated by the Watts-Strogatz model , which generates “small-world” networks with high clustering and short path lengths . The network is also modular , containing energy level sensing proteins that connect to four output processes: autophagy , fatty acid synthesis , mRNA processing , and MAP kinase signaling . The importance of each protein to network function is dependent on its Katz centrality score , which is related both to the protein’s position within a module and to the module’s relationship to the network as a whole . The network is also divisible into subnetworks that span modular boundaries and regulate different aspects of fat metabolism . We used a combination of genetics and pharmacology to simultaneously block output from multiple network nodes . The phenotypic results of this blockage define patterns of communication among distant network nodes , and these patterns are consistent with the Watts-Strogatz model .
Systems biology explores the emergence of patterns , structures , and properties in biological systems that cannot be understood by examining individual components . One of the goals of this discipline is to discover how the topological arrangements of proteins within signaling networks endow the networks with features that are important for their cellular functions . Due to the availability of extensive proteomic data and the existence of strain collections bearing deletion mutations for most of its genes , the budding yeast Saccharomyces cerevisiae is an excellent system in which to discover topological principles governing the design of signaling networks [1 , 2] . Some network analyses in yeast have examined all of the proteins identified by genome-wide proteomic methods [3–15] , while others have focused on essential genes that encode highly connected proteins , referred to as hubs , that are characterized by a lethal phenotype when removed [16–18] . There are disadvantages associated with using either of these approaches to analyze the relationships between network topology and function . First , although proteomic data define connections among proteins , not all connections made by a given protein are relevant when that protein performs its functions in a specific cellular process . Second , lethality can be produced through many different mechanisms , so genes and proteins required for viability do not necessarily have related functions . Third , the contributions of essential genes to survival can only be scored as viability or lethality . Most biological processes , however , exhibit variations in output strength , and incorporation of this information can add value to network models . Fourth , due to the lethal phenotype of these genes , networks of essential genes usually do not provide information about their relationships to the products of interacting nonessential genes . Here we show that molecular mechanisms used for regulation of fat storage in yeast provide an excellent system for network analysis . First , the mutant phenotype , an alteration in fat levels , is specific enough to suggest that there should be molecular relationships among many of the proteins in the network . Second , the severity of the fat storage defect when a fat level-regulating protein is removed can be quantitatively assessed , and this can be used to determine the protein’s importance to network function . Third , since the loss of a fat storage-regulating gene usually does not cause lethality , mutants selected for quantitative changes in fat content can also be assayed for alterations in other aspects of fat metabolism , such as lipid droplet ( LD ) morphology and the ability to use different carbon sources for fat synthesis . By using a system-wide approach that combines genetic , proteomic , pharmacological , mathematical , and physiological analysis , we have identified and characterized a physically interconnected network of 94 proteins that regulates fat storage in budding yeast . The fat regulation network is not scale-free , and is best approximated by the Watts-Strogatz model [19] , which generates “small-world” networks with high clustering and short path-lengths . Such networks have many features that are useful for biological control . The importance of a protein to network function is dependent on a particular kind of topological centrality , and the use of this centrality measure may provide a guideline for future analysis of proteins in other biological networks . We were also able to validate the network model by experimentally blocking function of multiple network nodes and showing that the patterns of internode communication predicted by this analysis are consistent with the small-world architecture of the network .
We developed a quantitative 96-well plate assay to screen the viable Saccharomyces cerevisiae deletion collection for alterations in fat content . In this assay , stored fat levels in fixed yeast cells were assessed by staining with the lipid dye Nile Red together with the nuclear dye DAPI and measuring the Nile Red/DAPI fluorescence ratio . Positive mutants were confirmed using a thin layer chromatography ( TLC ) assay to measure triglycerides , as described by [20] ( Fig 1A ) and by histological staining of fixed cells with another fat-specific dye , BODIPY 493/503 . Mutations in 86 genes caused statistically significant increases in fat content ( Fig 1 and S1 Table ) . 54 of the 86 genes identified in this screen have metazoan orthologs or relatives . Of these , brahma ( a chromatin remodeling protein orthologous to SNF2 ) , histone H4 ( HHF1 ortholog ) , cdk12 ( a Cdk family kinase orthologous to CTK1 ) , and me31b ( an RNA helicase orthologous to DHH1 ) , were identified in one or both of the published screens of cultured Drosophila cells for LD morphology phenotypes[21 , 22] . Extensive proteomic data exist for budding yeast ( see [10–13 , 23] ) . These data were obtained by a variety of methods , including the two-hybrid system[3 , 9] , the protein fragment complementation assay[4] , affinity purification and co-precipitation[7 , 8] , and analysis of global protein phosphorylation patterns[5 , 6] . We assembled current data on physical interactions for proteins encoded by these genes from the Saccharomyces Genome Database ( www . yeastgenome . org ) and BioGRID Database ( biogrid . org ) . Remarkably , superimposition of our genetic data on these proteomic data showed that 91% of the mutations we identified affect genes encoding proteins that have physical connections to one another , forming a densely interconnected network ( Fig 1B ) . Most of the proteomic interactions shown in the figure were defined by affinity purification studies , indicating that they define stable and abundant complexes that exist in vivo . The network defined by screening the viable deletion collection necessarily contains only nonessential genes . However , proteins encoded by essential genes are also known to be involved in regulation of fat synthesis and storage , or are components of pathways identified by the screen of nonessential genes ( see next section for discussion of network pathways ) . We thus added eight proteins encoded by essential genes to the network . Fas1 and Fas2 are subunits of fatty acid synthetase , and Acc1 is acetyl-coA carboxylase . These enzymes are required for de novo synthesis of long-chain fatty acids . Kog1 and Lst8 are components of the TOR complex 1 ( TORC1 ) , which we identified as part of the network through the identification of two other TORC1 subunits , Tor1 and Tco89 . Ste12 is a transcription factor activated by a MAPK signaling cascade , and Cdc42 is a small G protein that regulates MAPK signaling . Many of the genes in the network encode MAPK pathway components . Cdc39 is a component of the CCR4-NOT core complex , and we identified many other components of this complex in our screen [24–28] . These essential proteins were added in order to create a more complete and biologically meaningful network , not to artificially increase its connectedness by adding “hubs” . In fact , only two of these eight proteins , Cdc39 and Cdc42 , have more than four connections to other proteins in the network . Moreover , as described below , we have shown that removal of all of these essential proteins from the network does not significantly affect its topological properties . There are 94 proteins ( nodes ) and 203 total connections ( edges ) in the complete network ( S2 Table ) . To evaluate the significance of this network , we first wished to determine whether the extensive interconnections among its proteins reflect a common biological function and are likely to have been selected by evolution . To address this issue , we asked whether randomly selected collections of similar numbers of proteins from the yeast proteome would display similar connection densities . We generated 200 random collections of approximately 98 proteins ( nodes ) each and annotated the interactions within each collection that had been identified in published experiments . Of the 200 networks thus defined , most had fewer than 30 total connections ( edges ) , with an average of 24 . 8 and standard deviation of 11 . 5 . Only one of the 200 randomly selected networks has more than 50 connections . This network contains a ubiquitin protein , UBI4 , which forms a hub that makes 43 of the 73 total connections in that network ( S1 Fig ) . In summary , then , the 203 edges in our network make the connectivity in our network ~15 . 5 standard deviations from randomly selected networks of yeast proteins with a similar number of nodes . This corresponds to a p-value of 10–54 . The network contains extra- and intracellular energy/glucose detection mechanisms controlled by Mth1 and the AMP-activated protein kinase complex ( AMPK ) encoded by the SNF1 , SNF4 , and SIP1 genes and their regulators TOS3 , ELM1 , and SAK1 . High levels of extra- or intracellular glucose cause degradation of Mth1[29 , 30] , while high AMP/ATP ratios cause activation of AMPK [31] . A reduction in signaling through either system is likely to produce a perceived surplus of energy that can stimulate fat storage . Mth1 and AMPK are connected to a set of transcriptional regulators and to four processes that may represent outputs . These are: 1 ) autophagy , which involves the TORC1 and vacuolar H+-ATPase ( Vma ) pathways [32]; 2 ) de novo fatty acid and sterol synthesis pathways; 3 ) MAP kinase ( MAPK ) pathways; 4 ) mRNA degradation , elongation , and initiation pathways involving the CCR4-NOT complex and its associated proteins . The MAP kinase pathways are involved in sugar and amino acid starvation responses , while the CCR4-NOT complex is known to repress the translation of mRNAs encoding proteins necessary for utilization of non-fermentable carbon sources and for glucose neogenesis [24 , 26] . We next examined if proteins that participate in the same process or pathway exhibit a higher tendency to have connections with each other than with other network proteins using a walk-trap algorithm . This function finds densely connected communities via random walks from one node to another [33] . The idea is that short random walks will tend to stay in the same community . This analysis showed that the network contains seven different communities ( modules ) composed of three or more proteins , and that each module is enriched for proteins that are involved in the same process ( Fig 1C ) . We performed an analysis of the Gene Ontology ( GO ) terms associated with each of the proteins in the network . GO includes three categories: Biological Process , Molecular Function , and Cellular Component . S3 Table shows a list of the GO terms that are enriched for network proteins relative to random yeast proteins , with p-values for the statistical significances of the enrichments . Many of these highly enriched categories are very general , with the top three being “growth” , “biological regulation” , and “protein phosphorylation” . Others overlap with the output processes described above ( Fig 1 ) , and are properties of the signaling pathways , biological processes , and cellular components represented in the network . Four of the modules within the network defined by the walk-trap algorithm ( Fig 1C ) are highly enriched for specific GO terms . These are terms related to MAPK signaling ( green circles ) , vacuolar acidification and proton transport ( dark blue circles ) , mRNA catabolic processes ( light blue circles ) , and positive regulation of transcription ( red circles ) . It has been previously shown that GO categories can correspond to clusters within the global yeast protein-protein interaction map[34] . The fact that proteins required for filamentous growth are overrepresented in the network ( p = 1 . 8 x 10–8 ) is interesting , In limited nutritional conditions , yeast can adopt a filamentous growth pattern that permits a non-motile colony to explore its surroundings for additional nutrients . The induction of this state requires the action of the AMPK , MAP , and Tor kinase pathways [25] , which form a considerable portion of the fat regulation network . We hypothesize that some yeast mutants that store excess fat do so because they are in a state of perceived energy excess , which is the opposite of the conditions in which filamentous growth would be favored . In yeast , MAP kinase pathways are organized as cassettes composed of specific combinations of MAPKKKs , MAPKKs , and MAPKs , and these cassettes can cross-regulate one another [28] . Because we found mutations affecting multiple MAP kinase pathways , there may be redundancy between the cassettes with regard to control of fat storage . Consistent with this , we observed that a double mutant ( fus3 kss1 ) lacking two MAPKs accumulates more fat than either single mutant ( S2 Fig ) . We also note that mutations in TORC1 units and not TORC2 units cause an increase in fat storage , thus implicating only TORC1 in yeast fat storage regulation . This was confirmed by the fact that rapamycin ( Rap ) , a selective inhibitor of the TORC1 complex[27] , causes an increase in fat storage in wild-type yeast ( S2 Fig ) . We did not isolate any ‘lean’ mutants in our screen . This could be due to the fact that our growth media does not contain fatty acids that could be converted to phospholipids necessary for membrane synthesis during cell division , or , alternatively , to genetic redundancy . Indeed , in cases where a low fat storage yeast strain has been reported , a phenotype was only detected when more than one fatty acid synthesis gene is removed [35 , 36] . Most of the genes we identified have not been previously implicated in fat storage . Earlier screens were done for LD morphology defects , but only ~20% of the 171 genes previously reported to affect LDs showed an increase in fat storage in our assays [37–39] . This is not surprising , because these genes were identified in screens by altered LD morphologies in live mutant cells , not by measuring fat content . Because the goal of our screen was to find mutants with altered fat storage levels , not to study LD dynamics , we fixed cells in order to deactivate pumps that can affect dye uptake and block vesicular traffic to increase the specificity of the dyes to lipid droplets , and used quantitative assays to evaluate fat content [20 , 40 , 41] . The fact that the interconnection density of the fat storage regulation network is much greater than that of any randomly generated network of yeast proteins ( S1 Fig ) implies that the majority of the connections within the network have biological relevance . This makes yeast fat regulation an ideal system in which to examine whether a biological network conforms to a mathematical model for network design . See S1 Text for detailed information on methods for mathematical analysis used in this paper . Many real-world networks , such as the Internet , power grids , and social networks , have been studied , and they tend to have certain features in common . A social network , for example , contains many different local clusters of people that are linked to each other by mutual acquaintances , so that any person within the cluster can be reached from another person by a small number of steps . Social networks also contain celebrities ( “hubs” ) , who have exponentially more followers than does an average member of the network . Similarly , biological signaling networks tend to contain modules and have hub-like proteins that make many more connections than other proteins . These features of real-world networks have been simulated using graph theory , which is the study of systems of objects , referred to as nodes , and their relationships , referred to as edges or connections . The properties of networks generated by graph generation models can be characterized by a variety of different parameters ( for review see [42] ) , including degree distribution P ( k ) , global clustering coefficient Cg , modularity M , and path length L . P ( k ) is the probability that a given network node has a certain number of connections , while Cg is a measure of the degree to which nodes in a network tend to cluster together [43] . M is a measure of the division of a network into communities or modules , produced by the tendency of some nodes to form connections primarily within the community to which they belong . Finally , L is the average number of steps along the shortest paths that connect all possible pairs of network nodes . The simplest model used to generate networks is the Erdos-Renyi model [44] , characterized by an equal probability of forming connections between any two nodes ( Fig 2A ) . Erdos-Renyi networks have a degree distribution similar to a Poisson distribution ( Fig 2B ) , and have low Cg , M , and L values . Most real-world networks are not approximated well by this model , because Erdos-Renyi graphs lack local clustering and hubs . The Watts-Strogatz model was designed to produce better models of real-world networks by remedying the lack of clustering in Erdos-Renyi networks . Watts-Strogatz graphs are produced by randomly moving the edges of a regular ring lattice , thus producing “criss-cross” connections across the ring[19] . This does not imply that a real network approximated by the Watts-Strogatz model contains a structure resembling a ring lattice . Moving edges on a ring lattice is simply the algorithm by which these graphs are generated . Similarly , Erdos-Renyi and scale-free graphs ( see below ) are also generated by specific algorithms , but this does not imply that real networks with similar topologies are built using these algorithms . Watts-Strogatz networks are expected to have a degree distribution that depends on the rewiring probability β , which is related to the number of new connections between nodes that are introduced into the ring lattice . β varies between 0 ( a regular ring lattice ) and 1 ( a lattice where so many connections are changed that it approximates an Erdos-Renyi graph ) . For moderate to high rewiring probabilities , the degree distribution of the network is a binomial distribution , which in the limit ( β = 1 ) approximates a Poisson distribution . For a low rewiring probability ( β near 0 ) , the distribution resembles the P ( k ) of a regular ring lattice , which is a delta function . In all cases , graphs of P ( k ) as a function of k for such networks have a distinct peak . Networks generated by the Watts-Strogatz model have local clustering and small-world properties , with high Cg and M values relative to those of Erdos-Renyi networks[19] ( Fig 2A ) . Scale-free networks have a P ( k ) distribution that follows a power law , meaning that a few hub nodes make exponentially more connections than other nodes in the network . A graph of P ( k ) as a function of k for a scale-free network resembles an exponential decay curve ( Fig 2B ) . The best-known model for generating such networks is the Barabási–Albert model , where the network evolves via the addition of new nodes that preferentially form connections to highly-connected nodes[45] . This model can account for the hubs found in real-world networks , but not for local clustering . Hierarchical scale-free networks are composed of modular units of nodes and connections that are combined in an iterative manner , while non-hierarchical scale-free networks have hubs but are not arranged in an organized pattern ( Fig 2A ) . Some have argued that protein networks within cells are likely to be non-hierarchical scale-free networks[46] , but others have found that real biological networks cannot be made to fit a power-law distribution[47 , 48] . The P ( k ) distribution of the experimental network has a distinct peak , and resembles a Poisson distribution more closely than an exponential decay curve , which indicates that the network is not scale-free ( Fig 2B ) . To quantify this statement , we considered the SSE ( standard sum of squares due to error ) between a binomial distribution and the observed P ( k ) distribution of the experimental network . The minimal SSE obtained between a binomial distribution and the degree distribution for the experimental network is 0 . 012 . The mean SSE on a selection of 200 , 000 simulated random ( Erdos-Renyi ) networks with the same expected distribution was 0 . 0099 , with a standard deviation of 0 . 0057 . This tells us that the SSE value of our network is less than half a standard deviation from the expected value . By contrast , the SSE obtained by fitting the experimental network to a power-law distribution was over two standard deviations from this expected mean . This provides strong evidence that the fat storage regulation network is much more likely to be approximated by an Erdos-Renyi model or a Wattz-Strogatz model than by a scale-free model ( Fig 2B and S4 Table ) . The experimental network has a relatively small L value ( 3 . 18 ) , which is essentially the same as the expected L values for simulated random ( Erdos-Renyi ) networks ( 3 . 19 ) . This does not distinguish between the Watts-Strogatz and Erdos-Renyi models , since both produce networks with low L values . However , the Cg and M values for the experimental network ( 0 . 22 and 0 . 567 , respectively ) are higher than the expected values for simulated Erdos-Renyi networks ( 0 . 048 and 0 . 44 ) , but are in agreement with the Watts-Strogatz model for a moderate rewiring probability ( β = 0 . 238 ) . Its high Cg value and low L value characterize the experimental network as small-world . This small-world topology is a distinctive feature of the Watts-Strogatz model . In a scale-free network , loss of peripheral nodes will have little effect on network parameters , but loss of hub nodes can produce major changes in path length L . An analogy to this is that shutdown or dysfunction of a hub airport ( e . g . , O’Hare in Chicago ) due to bad weather will cause many travelers to have to take additional flights to reach their destinations . However , shutdown of a small regional airport will have no effect on the global pattern of air travel . To evaluate whether our network is vulnerable to deletion of any nodes , we determined the contribution made by each protein to the topological parameters of the entire network . This was done by calculating L , Cg , and M for the network and then recalculating those parameters ( L-P , Cg-P , and M-P ) when each network protein and its connections were removed . Only minor changes to these values were observed in response to removal of any one protein ( <10%; S2 Table ) . This is of interest , because it argues that the Watts-Strogatz topology of the network may contribute to the robustness of its topological parameters to deletion of any node . Unlike scale-free networks , it lacks essential hub nodes for which elimination of function by mutation or damage would dramatically alter the entire network . We also simultaneously removed all eight of the proteins encoded by essential genes that were added to the network ( see above ) , to ensure that these proteins were not required for network properties . This produced no significant changes in L or M , and only a 10% decrease in Cg . To evaluate the relationships between the experimentally determined fat storage regulation network and manually curated GO categories , we generated networks of protein-protein connections for the proteins in yeast Biological Process categories , since fat storage regulation is more like a Biological Process than it is like a Molecular Function or Cellular Component . These categories range in size from a few proteins to more than 250 . We then calculated edge density ( the number of edges/number of possible edges ) , Cg , and M for these networks and compared these numbers to the values for the fat storage regulation network . S3 Fig shows that the edge density for our network is around the median for the GO categories . Cg is below the median , and M is above the median . These data indicate that the experimental fat storage regulation network has similar properties to those of networks formed from the proteins within GO Biological Process categories . Proteins in a ‘typical’ Biological Process category are extensively interconnected , and exhibit more clustering but less modularity than the proteins in the fat storage regulation network . Because the fat regulation network was defined by a quantitative genetic screen , it has the useful property that the importance of a given protein to network output can be defined by the severity of the fat storage phenotype ( the amount of fat added above wild-type levels ) for a deletion mutation in the gene encoding that protein . This is graphically depicted in Figs 1 and 2 , where the sizes of the circles in the network diagrams are proportional to the strength of the phenotype produced by loss of the corresponding protein . This allows us to evaluate the relationship between a node’s position in the network and its importance for network function . If such a relationship can be established , it can be used by future investigators to find the essential elements in other biological networks that are found to fit the Watts-Strogatz model , but for which quantitative information on phenotype is not available . In network analysis , the centrality of a node , C ( v ) , refers to indicators that identify the most important nodes [49] . For example , centrality analysis has been used to find the most influential person in a social network . There are many types of centrality , each of which emphasizes a certain quantifiable attribute of a given node . We examined the five standard centrality types ( Degree , Betweenness , Closeness , Eigenvector , and Katz centralities ) [49–51] to measure these attributes . For degree centrality , the value is evaluated by counting the number of directly connected nodes . Betweenness centrality is evaluated by counting the number of shortest paths that pass through a given node . Closeness centrality is a sum of the inverses of the shortest path lengths . Eigenvector centrality is an implicitly defined measure evaluated by asserting that the centrality of a single node is proportional to the sum of the centralities of all nodes it is connected to . For Katz centrality , the importance of a node is determined by how many nodes it is path-connected to , with a penalty that increases exponentially with the path length between those nodes . The colored heat-map diagrams in Fig 2C–2G show the values of each type of centrality for each node in a small toy network , with red indicating the highest centrality . The diagrams below superimpose centrality values , indicated by the sizes of the red circles , on the grey circles that indicate phenotypic severity . We assessed the relationships between the centrality of a given node and the severity of the fat storage defect exhibited when the gene corresponding to that node is removed , using both independence and correlation testing ( Fig 2C–2H and S4 Table ) . Dependence is any statistical relationship between two random variables , while correlation is a special type of dependence that can be used to predict the magnitude of the change that will occur in one variable in response to changes in a linked variable . Two random variables that are independent necessarily have low correlation , but variables with low correlation can be dependent . For example , driving while intoxicated and fatal car accidents are clearly not independent of each other , but the correlation between the two is only moderate , since only 32% of such accidents involved an intoxicated driver and most intoxicated drivers do not have an accident . Dependence is a more general and arguably more useful measure , as it is able to detect nonlinear relationships , whereas correlation assumes a simple linear dependence . The only centrality measure that passed five separate independence tests ( Blomqvist β , Goodman-Kruskal , Hoeffding D , Kendall tau , and Spearman Rank ) was Katz centrality ( Fig 2E ) , for which the p-value ( probability of independence ) was <0 . 027 for all 5 tests . There are good reasons that Katz centrality gives the best dependence score , which are based on how the centrality measures are defined . First , a good measure of the importance of a protein node should be global , and not just depend on proteins that are directly connected to it . Second , those nodes related by longer paths should have a smaller effect on each other’s function than those related by shorter paths . These attributes are held by closeness , eigenvector and Katz centrality measures . The effect of adding a node on the closeness centrality depends on the reciprocal of path lengths , for eigenvector centrality it is loosely related to degrees of the node , and for Katz centrality , the effect will become exponentially smaller the as the path length increases . In many models , it is reasonable to assume that only a fraction of a signal passes through a node to each of the other nodes to which it connects , and this assumption predicts an exponential drop-off . Although the statistical analysis shows a clear relationship between the severity of the fat storage phenotype and Katz centrality , the moderate correlation between these parameters ( 0 . 24 ) suggests that this is not necessarily a linear relationship . Fig 2E displays this graphically . It shows that many of the larger grey circles have corresponding large red circles , but there are some large grey circles with very small red circles , indicating low Katz centrality . This shows that the quantitative impact on fat levels caused by removal of a network protein cannot always be predicted from its Katz centrality score in a linear way . In particular , mutations eliminating transcriptional regulators that have low Katz centrality scores often have large impacts on fat levels ( e . g . , Spt10 ) . This may be due to the fact that the importance of a transcriptional regulator to the network is more likely to be related to the number of gene targets whose expression it controls than to the number of proteomic connections it makes . Most of the transcription factors that regulate fat storage levels bind to the promoter regions of genes that are themselves part of the network , creating potential feedback loops [52] ( Fig 2H ) . Ino2 , Srb5 , and Ctk1 bind to the promoter regions of up to 40% of network genes . In total , 67% of network genes have network transcription factors other than the global regulators Hhf1 and Spt10 that bind to their promoter regions . The number of network genes that a given network transcriptional regulator binds to is an excellent predictor ( correlation = 0 . 787; probability of independence <0 . 01; S4 Table ) for the severity of the fat storage defect when the gene that encodes it is removed . Another useful feature of yeast fat regulation is that mutants selected for increased fat content can also be examined for related phenotypes , such as LD morphology and metabolic alterations . We can then ask whether the genes and proteins that share these “sub-phenotypes” also form networks , and examine the relationships between these subnetworks and the larger fat regulation network . Yeast LDs are composed of triglycerides and sterol esters surrounded by a monolayer of phospholipids . LDs may increase in size and/or number when trigylceride levels increase . A wild-type fixed yeast cell usually has three to eight LDs that are around 0 . 4 μm in diameter ( Fig 3A ) . We examined all the mutants for LD morphology phenotypes , grouped them into three phenotypic classes , and created subnetworks encompassing all the proteins for each class . Class I contained mutants with small but numerous LDs , class II contained mutants with mixed populations of normal and small LDs , and class III contained mutants that have giant LDs ( Fig 3 ) . Each of the three subnetworks was highly interconnected , indicating that members of the same morphological class tend to have mutations affecting proteins that are connected to each other ( S5 Table ) . Fat levels are influenced by the rate of fat storage utilization , the rate of de novo fatty acid synthesis , and the level of caloric intake . We examined fat storage utilization by subjecting all of the mutants to glucose starvation for three days . ~85% of the mutants showed no reduction in fat storage or a reduced rate of fat storage depletion compared to wild-type when starved , indicating defects in the ability to use stored fat to meet energy demands ( Fig 4A and S6 Table ) . To be used as an energy source , fats have to be broken down to glycerol and fatty acids , which are used by mitochondria for ATP production [53 , 54] . We evaluated the overall status of mitochondria in all mutants with reduced fat utilization rate by growing them on glycerol . ~60% of mutants showed either slowed or no growth on glycerol , suggesting defects in mitochondrial function . All of these mutants also failed to grow on palmitic acid and lard . Another 26 mutants did grow on glycerol , but failed to grow on either palmitic acid or lard or both . Mutations that produce a similar growth defect on a given carbon energy source tend to affect proteins that are connected to one another , forming subnetworks ( S6 Table and Fig 4B and 4C ) . These results show that a significant portion of the network is dedicated to maintaining normal mitochondrial function . Indeed , nearly half of the genes we identified were previously implicated in mitochondrial function[53] . In our growth conditions , cells were provided with glucose and a mixture of amino acids that can be used for de novo fatty acid synthesis . To examine if any of the mutants have alterations in this process , we grew them on media containing either 14C-labeled D-glucose or 14C-labeled L-aspartic acid . About 30% of the mutants showed an increase in the conversion of either D-glucose or L-aspartic acid to fat . Mutants that exhibited increased conversion of either nutrient to fat tend to encode proteins that are part of a connected subnetwork ( Fig 5 and S7 Table ) . To test the hypothesis that genes for which mutants exhibit similar physiological profiles would encode proteins that formed subnetworks , we examined all of the potential subnetworks encoded by subsets of genes with similar detailed sub-phenotypes , in the same way that we had examined the complete network . We found that these subnetworks also exhibited Watts-Strogatz small-world topology , with relatively high clustering coefficients ( 0 . 179–0 . 28 ) and short path lengths ( 3 . 09–3 . 85 ) . To determine if the subnetworks represented functional subsets of the larger network , we then compared each subnetwork to randomized simulations of 104 networks with identical degree distribution and vertex counts . In all cases , the experimentally observed subnetworks had much higher clustering coefficients than their simulated equivalents ( red and green numbers on the right sides of the panels of S5–S7 Tables and Figs 3–5 ) , indicating that the subnetworks were selected by evolution , and do not represent randomly chosen subsets of the complete network . We also observed that in all cases , these subnetworks are not confined to a single module , but span modular boundaries . These results suggest that the larger network that controls fat storage contains within it smaller networks that govern different aspects of physiology related to a cell’s decision whether to store fat or metabolize it for energy . Having established that the fat regulation network fits the Watts-Strogatz model , we then conducted experiments in which we perturbed the functions of multiple network nodes and measured the effects of these perturbations on fat content . This allowed us to evaluate whether the patterns of communication among distant nodes in the real network are consistent with the Watts-Strogatz topology . We call these patterns “signal propagation” , because , although they are not measured dynamically , they reflect the movement of information through the network . They thus represent signal flow along signal transduction pathways and crosstalk between these pathways . We conducted “chemogenomic” experiments in which we simultaneously perturbed the functions of multiple network nodes by treating yeast bearing a deletion mutation in each network gene with a set of drugs that block the functions of specific network pathways , followed by measuring the effects of these perturbations on fat content . The approach of adding drugs to single mutants to block multiple nodes was chosen in order to avoid the slow growth or lethality frequently observed for double mutants ( see [55] ) , as well as the necessity to construct all possible double mutants . The chemogenomic approach has often been employed in the yeast system to map synergistic and antagonistic relationships between drug targets and other genes ( for recent reviews see [56 , 57] ) . We selected five drugs that blocked signal propagation through specific network pathways . The first drug was U0126 , an inhibitor of mammalian MAPKKs [58] that has been shown to block reporter expression controlled by the MAPK mating factor response pathway [59] . The second drug was Rap , an inhibitor of the mammalian and yeast TORC1 complex[27] . The third and fourth drugs were chloroquine ( ChQ ) and concanamycin A ( Conc . A ) ; these are both known to block the acidification of vacuoles by the Vma pump [60–63] . The fifth drug was cerulenin , which inhibits both yeast and mammalian fatty acid synthase [64 , 65] and thereby eliminates fat synthesis; cerulenin also served as a control to ensure that the effects of the mutations were dependent on network output . The specificity of the drugs we chose was confirmed by the fact that they did not produce increases in fat levels in mutants missing their potential targets ( since in those mutants signals from the drug targets are already absent ) , and by the observation that they produce fat levels similar to mutations that remove these targets ( see Supplemental Materials and Methods ) . Regardless of whether the drugs were completely specific for particular targets , they clearly caused perturbations in network function , as indicated by their ability to alter fat levels in wild type yeast ( Fig 6B ) . The effects of these perturbations on mutant networks , each of which lacks a single node , can therefore be used to analyze internode communication within the network . We defined three types of signaling relationships between network proteins that are removed by mutation and those whose activities are blocked by a given drug . First , the proteins could be in independent signaling pathways . In such cases , the application of the drug would produce an additive effect , such that the amount of additional fat would equal the sum of the amount of fat added by the mutation and the amount added to wild-type yeast in response to drug treatment . This type of interaction is like the negative interactions between deletion mutations ( or between deletion mutations and drugs ) that are often observed in double mutant or chemogenomic analyses of yeast growth phenotypes ( reviewed by[56] ) . The strongest form of this type of negative genetic interaction is synthetic lethality , in which double mutant cells ( or mutant cells treated with drug ) die , while single mutants or drug-treated wild-type cells are viable . Second , the proteins could be part of the same pathway . In such cases , treating the mutant with drug would produce fat levels that are no higher than those in wild-type yeast treated with drug , since pathway signaling had already been eliminated by the mutation . Third , the proteins affected by the drug and by the mutation could be components of two pathways that relay part , but not all , of their signals through each other . We refer to these as synergistic pathways . In such cases , addition of a drug affecting pathway 1 to a mutant affecting pathway 2 would produce an increase in fat levels that is less than the sum of the amount of fat added by the mutant and the amount added by the drug , since the portion of the signal relayed through the action of both pathways was already blocked by the mutation , and the drug could therefore only affect the fraction of the signal that was still active and available for inhibition . In other words , part of the fat level increase observed for the mutation alone would be due to partial blockage of the drug-affected pathway ( Fig 6A ) . During the course of our analyses , we noticed that none of the drug-mutant combinations produced fat levels that approach the levels of our “fattest” mutant , spt10∆ , indicating that our analysis was not limited by cells reaching their maximum fat storage capacity . We also never observed a situation in which the amount of fat in given drug-mutant combination was greater than the sum of the amount added by the drug alone and the mutation alone . These observations indicated that the behavior of the network could be understood using our method . First , we found that , for each of the four different drugs that increase fat , 12–15% of network proteins behave as if they are in pathways that are independent of the pathway affected by the drug ( e . g . , they have strong synthetic negative interactions with the drug target ) . Second , 58% of network proteins had either “same pathway” or “synergistic pathway” relationships with all four drug target pathways ( S8 Table and Figs 6 and S4 ) , indicating that there is extensive communication across the network and that network outputs reflect integration among multiple signaling pathways . The fact that proteins involved in such relationships can be in different regions of the network is consistent with the idea that the short path length characteristic of small-world networks facilitates signal propagation between distant parts of the network . Third , proteins that had a “same pathway”-type relationship with a given drug target sometimes were in different network communities from the drug target . Some of these relationships suggested hitherto unknown interactions between pathways . For example , components of the CCR4-NOT complex involved in mRNA processing had “same pathway” type relationships to the MAPK node affected by U0126 , showing that the output of CCR4-NOT was relevant to regulation of fat by the MAPK pathway ( Fig 6D ) . Proteins with “same pathway” type relationships to a given drug could be grouped into subnetworks that had Cg values that were much higher than those of simulated random networks of the same size and connectivity ( 0 . 22–0 . 35 ) , indicating that they have been selected by evolution . These subnetworks did show significant overlap with one another ( sharing some of the same proteins ) , but each one had a unique combination of proteins that was specific for a given drug ( S9 Table and Figs 6 and S4 ) . Fourth , there is a “sub-additive” response ( that is , an increase in fat content that is less than the sum of the increase produced by the drug and that produced by the mutation ) to loss of function of two nodes for most node pairs . Fifth , eight proteins had a “same-pathway” type relationship with all four drugs , qualifying them as points of convergence for network signals ( Fig 6E ) . This might suggest that these are hub proteins within a scale-free network . However , the hypothetical power-law distribution shown in Fig 2B predicts that in order for our network to be truly scale-free , the number of hubs should not have exceeded two . Furthermore , of these eight proteins , only Ste5 had more than four connections , and hubs should make many more connections than other proteins in the network . We compared our results to some recent analyses of double mutant and drug-mutant growth and metabolism phenotypes[55 , 66 , 67] . These are genome-wide studies , so only a small percentage of mutants ( ~5% ) exhibited a double mutant interaction or an interaction with a given drug . In our study , we investigated a subset of genes that we had defined as critical for fat storage regulation , and examined interactions with drugs that we had selected due to their ability to affect fat content , so we observed that most mutants displayed interactions with all of the drugs . In the genome-wide studies , the frequency of synthetic ( negative ) interactions was much higher for genes within the same GO Biological Process category than for genes in different categories . For most categories , this frequency was between 10 and 18%[55] , which is in the same range as the frequency of genes within the fat regulation network that have strong negative interactions with a given drug target ( 12–15%; “independent pathways” category ) . Thus , the experimentally defined fat storage regulation network , which has interconnection density and modularity values that are similar to those of some Biological Process categories ( S3 Fig ) , also behaves somewhat like a Biological Process category with respect to chemogenomic interactions . Finally , about 1/3 of interactions observed in genome-wide studies were positive , but we did not detect any positive interactions . This is due to the fact that we only isolated genes for which mutation increases fat contes , so all of our phenotypes have the same sign . Cellular signaling pathways , however , contain both positive and negative regulators , and mutation of regulators with opposite sign can produce phenotypes of opposite sign . We also compared the results of Rap treatment of fat storage mutants with a study that examined interactions between Rap and all nonessential genes for a different TORC1-dependent phenotype , expression of a DAL80 reporter that is induced by TORC1 inactivation by Rap or starvation[68] . Of the 63 Rap-specific genes they identified , only 6 ( Bck1 , Las21 , Slt2 , Srb5 , Swi4 , Swi6 ) were identified in our screen as mutations that increase fat content , suggesting that fat storage regulation and DAL80 induction are not closely related processes . However , two of these common genes ( Srb5 and Swi6 ) , increase induction of the DAL80 reporter by Rap[68] and also synergize with Rap to increase fat content ( Fig 6 and S8 Table ) .
By screening the yeast ( Saccharomyces cerevisiae ) viable deletion collection for mutations affecting fat content , we discovered a densely interconnected network of 94 proteins that regulates fat storage ( Fig 1 ) . From a computational analysis of this network , we derived three major conclusions . First , the network is not scale-free , and can be best approximated by the Watts-Strogatz model , which has not been previously applied to biological signaling networks ( Fig 2 ) . This model can account for the high degree of clustering observed in the experimental network . Second , the importance of an individual protein to network function is dependent on a particular measure of centrality , Katz centrality , which is influenced by the topological relationships between the test protein and the other proteins within its network community or module ( Fig 2 ) . Physiological analysis showed that the fat regulation network is divisible into connected subnetworks which span community boundaries and affect specific aspects of lipid metabolism ( Figs 3–5 ) . Finally , by combining drug perturbations with genetics ( chemogenomics ) , we showed that there is extensive cross-talk between the different signaling pathways represented in the network ( Fig 6 ) . The small-world topology endows the network with useful features . Its short path length property allows distant nodes to communicate with each other through a small number of steps . Its topological parameters are robust with respect to removal of any one protein . In scale-free networks , the removal of highly connected hubs will cause substantial changes in network topology and may fragment them into unconnected subnetworks [69–71] , while the absence of modular structure in the Erdos-Renyi model makes it inconsistent with the formation of biological networks that must receive different types of inputs and couple them to unique outputs . Our results also show that the importance of a protein to network function is dependent on its Katz centrality score ( Fig 2 ) . The Katz centrality of a node is determined by the number of shortest paths that pass through it to all other nodes in the network , with penalties assigned to connections to distant nodes ( nodes that are connected to the node of interest only through a proximal node ) . The reason that this type of centrality gives the best prediction of a protein’s importance to the function of the fat regulation network may be due to the fact that it combines the local attributes of a node within its community with the position of that community within the network as a whole . As such , it is suited to modular/community based networks , and we predict that future analysis of other such networks will reveal that some of the most important proteins will be identifiable by their Katz centrality scores . In the fat regulation network , the importance of some of the proteins with high Katz centrality can be explained based on their known roles in the module to which they belong . Two examples are Bem1 , which is a MAP kinase pathway scaffolding protein [72 , 73] , and Snf1 , which is the catalytic subunit of the AMPK complex [29 , 31] . By analyzing other phenotypes associated with alterations in fat storage , such as LD morphology and utilization of carbon sources , we further demonstrate that the network can be divided into subnetworks that span molecular categories and modules , but affect specific aspects of lipid metabolism ( Figs 3–5 ) . All subnetworks contain proteins that have connections that span modular boundaries . To examine whether the topological features described above are associated with biologically relevant properties , we examined signal propagation within the network by combining genetics and pharmacology ( Fig 6 ) . The results indicate that the network has a sub-additive response to perturbation , because the blockage of two pathways or proteins by a drug and a mutation usually produces effects that are smaller than the sum of those caused by the drug and the mutation individually . Of course , the network is not immune to alteration; the removal of a protein from the network does produce an increase in fat levels , since that is how the network was defined . Nevertheless , the attenuated response of the network to blockage of multiple nodes is consistent with the idea that network structure buffers it against external or internal perturbations . The fact that proteins in distant parts of the network often interact with each other , as indicated by the drug-mutant experiments , are in agreement with the short path length and small-world properties of the Watts-Strogatz model . Unlike Erdos-Renyi networks , Watts-Strogatz networks are modular . In neural networks , modularity is known to facilitate multifunctionality , which can divide large tasks into smaller compartmentalized subtasks that can be executed efficiently [74–76] . Multifunctionality also allows neural networks to integrate different inputs and generate diverse output responses . The yeast fat regulation network has properties that are analogous to multifunctionality in modular neural networks , because it takes in multiple inputs , processes them , and generates multiple outputs in response . The integrated input of the network represents an evaluation of the available nutrients and energy stores . It is provided by the glucose level detection function of Mth1 and the AMP and starvation detection mechanisms of MAP kinase and AMPK pathways . The network then generates anabolic and catabolic outputs that are appropriate to those inputs . Our analysis of mutant phenotypes indicates that these outputs affect many aspects of cell physiology , such as LD morphology , mitochondrial function , and fat store utilization . Biological signaling networks often contain hub-like elements , and some researchers have proposed that biological networks with hubs are best described by scale-free models[46] . However , real interaction network datasets that have been tested do not fit power-law P ( k ) distributions , which are diagnostic of scale-free networks [47 , 48] . The P ( k ) distributions of proteins in yeast proteomic networks generated by the two-hybrid method have been subjected to mathematical analysis , and a variety of models were tested , including Poisson and binomial distributions ( characteristic of Erdos-Renyi and Watts-Strogatz networks ) and power-law distributions ( characteristic of scale-free networks ) were tested . None of the models could be definitively proven or ruled out [18 , 77] . Almost all of the proteomic connections used to define the fat storage regulation network were identified by affinity purification/coprecipitation methods , which provide a reliable means to identify abundant and stable protein complexes that exist in vivo [7 , 8 , 10 , 12 , 78 , 79] . Our computational analysis clearly shows that this network’s topology is best approximated by the Watts-Strogatz small-world model ( Fig 2B ) . This topology has significant regulatory advantages that are correlated with biologically relevant features , and we suggest that mathematical analysis of other proteomic networks defined by a combination of genetics and affinity purification methods may reveal that many signaling networks within cells are of this type .
A collection of haploid MATa nonessential yeast deletion strains was purchased from Thermo Scientific . The strains were transferred into 150 μl YPD media in a 96 well-plate and grown for three days at 30°C . 5 μl of the resulting culture was respectively transferred into 150 μl synthetic complete dextrose media [80] in a 96 well assay plates ( black wall with transparent bottom ) and grown for two days at 30° C . After the growth period , formaldehyde was added to each well to a final concentration of 4% , followed by incubation at room temperature for 20 minutes . The plates were then spun down at 3000 rpm for two minutes and the supernatant was discarded . The pellets were then resuspended in 150 μl PBS containing 0 . 125 μg/ml Nile Red and 0 . 003 μM DAPI , followed by incubation at room temperature for 20 minutes . Fluorescence was measured using a spectrophotometer for both Nile Red ( Ex485/Em590 ) and DAPI ( Ex 358/Em440 ) . Results were plotted as a ratio between the Nile Red and DAPI signals . Any positive lines were regrown in eight replicas , of which 4 were stained as above; the others were processed in the same manner , except that Nile Red and DAPI were not added . The latter group was used to measure autofluorescence , the value of which was later subtracted from the final reading . Only lines that remained positive were taken to the second round of selection . Lines that passed both the 96-well plate assay and the histological examination were subjected to the TLC assay . In this assay cultures were grown in complete synthetic 2% dextrose media for two days at 30° C , and 20 ml normalized culture of OD600 = 1 was obtained . Protein from 1ml of OD600 = 1 culture was extracted and measured and reading was used for additional normalization step , and we rarely observe a situation in which cultures with the same OD600 ending up having a different protein measurements . The twice-normalized culture was then spun down , and the resulting pellets were suspended in 200 μl 2:1 chloroform: methanol mixture , and three glass beads ( 2 mm ) were added to each tube . The samples were subjected to continuous agitation for one hour , and vortexed three times ( 1 minute each ) during that period . The samples were then spun down for 2 min . at maximum speed and the lower phase was isolated and transferred to a fresh tube . The isolated lower phase was spun down again and run on a TLC plate as described [20] . For confocal microscopy , yeast were grown on synthetic complete 2% dextrose media at 30° C for two days . The samples were then spun down and processed for microscopy [81] . For additional information see S1 Text ( Supplementary Materials and Methods ) , which contains methods for starvation studies , growth on different carbon sources , drug treatment , and mathematical analysis . | We discovered a large protein network that regulates fat storage in budding yeast . This network contains 94 proteins , almost all of which bind to other proteins in the network . To understand the functions of large protein collections such as these , it will be necessary to move away from one-by-one analysis of individual proteins and create computational models of entire networks . This will allow classification of networks into categories and permit researchers to identify key network proteins on theoretical grounds . We show here that the fat regulation network fits a Watts-Strogatz small-world model . This model was devised to explain the clustering phenomena often observed in real networks , but has not been previously applied to signaling networks within cells . The short path length and high clustering coefficients characteristic of the Watts-Strogatz topology allow for rapid communication between distant nodes and for division of the network into modules that perform different functions . The fat regulation network has modules , and it is divisible into subnetworks that span modular boundaries and regulate different aspects of fat metabolism . We experimentally examined communication between nodes within the network using a combination of genetics and pharmacology , and showed that the communication patterns are consistent with the Watts-Strogatz topology . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Experimental and Computational Analysis of a Large Protein Network That Controls Fat Storage Reveals the Design Principles of a Signaling Network |
First described in humans in 1964 , reports of co-infections with dengue ( DENV ) and chikungunya ( CHIKV ) viruses are increasing , particularly after the emergence of chikungunya ( CHIK ) in the Indian Ocean in 2005–2006 due to a new variant highly transmitted by Aedes albopictus . In this geographic area , a dengue ( DEN ) outbreak transmitted by Ae . albopictus took place shortly before the emergence of CHIK and co-infections were reported in patients . A co-infection in humans can occur following the bite of two mosquitoes infected with one virus or to the bite of a mosquito infected with two viruses . Co-infections in mosquitoes have never been demonstrated in the field or in the laboratory . Thus , we question about the ability of a mosquito to deliver infectious particles of two different viruses through the female saliva . We orally exposed Ae . albopictus from La Reunion Island with DENV-1 and CHIKV isolated respectively during the 2004–2005 and the 2005–2006 outbreaks on this same island . We were able to show that Ae . albopictus could disseminate both viruses and deliver both infectious viral particles concomitantly in its saliva . We also succeeded in inducing a secondary infection with CHIKV in mosquitoes previously inoculated with DENV-1 . In this study , we underline the ability of Ae . albopictus to be orally co-infected with two different arboviruses and furthermore , its capacity to deliver concomitantly infectious particles of CHIKV and DENV in saliva . This finding is of particular concern as Ae . albopictus is still expanding its geographical range in the tropical as well as in the temperate regions . Further studies are needed to try to elucidate the molecular/cellular basis of this phenomenon .
Dengue ( DEN ) and chikungunya ( CHIK ) are two mosquito-borne viral infections transmitted by mosquitoes of the genus Aedes . Dengue viruses ( DENV-1 , -2 , -3 , -4 ) belonging to the Flaviviridae family , Flavivirus genus [1] are of primarily concern as they are responsible of the most important arboviral disease widely distributed in the tropical world [2] . Dengue infection may be unapparent or induce an undifferentiated febrile illness , a classic dengue fever ( DF ) , or a dengue hemorrhagic fever ( DHF ) . The highest prevalence of DEN is observed in South-East Asia and South Americas with approximately 50–100 million cases and 25 000 deaths per year [3] . The transmission is mainly ensured by the highly anthropophilic Aedes aegypti in urban areas [4] . However , Aedes albopictus may act as a secondary vector in rural areas and even as the main vector when Ae . aegypti is not present or too scarce as observed in some localities in China , Japan , Hawaii , and Seychelles [5] . Ae . albopictus was indeed the only vector in the recent dengue outbreaks observed in the Indian Ocean on La Reunion Island [6] and Madagascar [7] . Chikungunya virus ( CHIKV ) , first isolated in Tanzania in 1953 [8] , belongs to the Togaviridae family , Alphavirus genus [9] and is endemic to Africa , India and South-East Asia . In Africa , the virus is maintained within a sylvatic cycle with wild mosquitoes ( Aedes furcifer , Aedes luteocephalus , Aedes taylori , Aedes africanus ) feeding preferentially on primates [10] , [11] . In Asia , CHIKV is mainly transmitted within an urban cycle in an inter-human transmission achieved essentially by the human-biting Ae . aegypti and the less anthropophilic Ae . albopictus , which prefers suburban and rural areas where it colonizes both artificial and natural containers [12] , [13] . CHIKV mainly induces high fever and severe arthralgia , and had limited impacts on public health before its emergence in the Indian Ocean in 2005 . This major epidemic started in the Comoro Islands in January 2005 then spread rapidly to the other islands of the region , Mayotte , Seychelles , La Reunion and Mauritius [14] . In April 2006 , one third of the population in La Reunion Island had been in contact with the virus [15] . Surprisingly , the vector in this epidemic was not Ae . aegypti , only present as residual populations on the island , but Ae . albopictus [16]–[18] . This latter species was proved to be a very efficient vector of a mutated strain CHIKV harboring a switch from an alanine to a valine in the E1 glycoprotein , mutation that appeared in the course of the outbreak and was then selected as a major epidemic genotype [19] , [14] . CHIK outbreaks spread rapidly and caused several million clinical cases in the Indian Ocean Islands and India , where outbreaks had been absent for 32 years [20] , [21] . One consequence was an increasing overlapping in the distribution of the two arboviral diseases , DEN and CHIK leading to the detection of a higher number of co-infections in humans . Co-infection by DENV and CHIKV in patients has been known for a long time . First described in 1964 in South India [22] , it has been more frequently reported since the re-emergence of CHIK: Sri-Lanka [23] , India [24] , [25] , Malaysia [26] where the main vector involved is Ae . aegypti and in Gabon [27] , Madagascar [7] and La Reunion Island [28] where viral transmission is achieved by Ae . albopictus . On La Reunion Island , the CHIK outbreak was preceded by a small outbreak due to DENV-1 with 228 cases reported between March and July 2004 [6] . Despite the limited impact of the DEN outbreak , in January 2006 , suspected cases of co-infection DENV-CHIKV in patients were reported [28] and the same phenomenon was also observed in Madagascar in January 2006 [7] . Dual arboviral infections in humans can occur following the bite of two mosquitoes , each infected by one virus , or the bite of one mosquito infected by the two viruses . If isolations of either virus from mosquitoes collected in the course of an epidemic have been already noticed , to our knowledge , doubly infected mosquitoes have never been described . A recent study even concluded on the failure to prove co-infection of Ae . aegypti by these two viruses [29] . Here , we orally infected in a single blood-meal , Ae . albopictus from La Reunion Island with autochthonous viral strains isolated during the 2004–2005 outbreak of DEN-1 and the 2005–2006 CHIK outbreak . Dissemination of both viruses within the vector was checked by immunofluorescence assay on female head squashes . Furthermore , saliva was collected from each female to check their ability to deliver simultaneously both DENV and CHIKV infectious particles . Additionally , we performed secondary infections by inoculating first DENV-1 then providing CHIKV in an infectious blood-meal .
All experiments on live vertebrates were performed in compliance with French and European regulation and according to the Institut Pasteur guidelines for laboratory animal husbandry and care . Ae . albopictus Providence ( ALPROV ) were collected as eggs in 2006 on La Reunion Island and provided by the DRASS ( Direction Régionale des Affaires Sanitaires et Sociales ) . The F5 or F6 generation was used for experimental infections . Colonies were maintained at 28±1°C with a light:dark cycle of 16 h:8 h and a 80% relative humidity . Larvae were reared in pans containing 1 yeast tablet in 1 liter of tap water . Adults were provided with 10% sucrose solution ad libitum and fed three times a week on anaesthetized mice . The CHIKV 06 . 21 isolated in November 2005 from a new-born male from La Reunion Island presenting meningo-encephalitis symptoms was used for all experiments . This strain harbored the A->V mutation at the position 226 in the E1 glycoprotein ( E1-226V ) [14] . Viral stock used was a third passage on Ae . albopictus C6/36 [30] stored at −80°C in aliquots . Procedure for C6/36 cell infections and passages are described elsewhere [19] . The DENV-1 185/04 was isolated in May 2004 from the plasma of a patient in La Reunion Island . The strain belonged to the Brazil group of the Pacific genotype which was the main genotype isolated during the outbreak ( GenBank: DQ285558 . 1 ) . The virus was provided as a second passage on C6/36 cells . DENV-1 production on mosquito cell cultures being insufficient to allow mosquito oral infections , the virus stock was produced by inoculating intra-thoracically mosquitoes with the viral strain [31] . Inoculated mosquitoes were incubated 10 days at 28°C and their infectious status checked by indirect immunofluorescent assay ( IFA ) on head squashes [32] . Bodies were then pooled and triturated in heated ( 56°C for 30 min ) FCS ( Fetal Calf Serum ) . The supernatant fluid recovered after low speed centrifugation was used as a source of virus in mosquito blood-meals . Both viruses were provided by the French National Reference Center for Arboviruses at the Institut Pasteur which had obtained the verbal consent from patients or parent's patients who provided blood sera . Adult females were inoculated using the protocol described by Rosen and Gubler [31] , each mosquito receiving 0 . 17 µL ( i . e . 102 . 8 PFU/mL ) of the DENV-1 strain . Infection assays were performed with 7 day-old females which were allowed to feed for 15 min through a chicken skin membrane covering the base of a glass feeder containing the blood-virus mixture maintained at 37°C . The infectious blood-meal was composed of a virus suspension diluted ( 1∶3 ) in washed rabbit erythrocytes isolated from arterial blood collected 24 h before the infectious blood-meal [33] . A phagostimulant , ATP , was added at a final concentration of 5×10−3 M . Fully engorged females were transferred to small cardboard containers and maintained with 10% sucrose at 28±1°C for 14 days . Viral suspension provided in the blood-meal contained one or two viruses . For the secondary infection experiment , the blood-meal with CHIKV yielding 106 FFU ( foci forming unit ) /mL was provided 7 or 13 days after inoculation with DENV-1 . As control , females were inoculated with DENV-1 alone or orally infected with CHIKV alone . For the two trials where both viruses were provided by oral route , titers of the blood-meals were respectively: 106 FFU/mL for CHIKV in both trials and 104 , 5 FFU/mL for DENV-1 in the trial 1 and 105 , 9 FFU/mL for DENV-1 in the trial 2 . At day 14 post-infection , females were chilled , and their wings and legs removed , the stress inducing a forced salivation . Proboscis was inserted into 1 µL micropipette ( microcaps® , Drummond Scientific Company , USA ) filled with FCS . After 45 min , medium containing the saliva was expelled under pressure into 1 . 5 mL tubes containing Leibovitz L15 medium supplemented with 10% FBS . To allow a specific detection of both viruses , each sample was inoculated in two wells , one for the detection of CHIKV and one for the detection of DENV-1 . 20 µL of each sample were added to monolayers of C6/36 cells in 24 wells plaque to detect infectious particles by the foci forming technique using an immunoperoxydase assay . Cells were incubated 3 days for CHIKV and 5 days for DENV-1 at 28°C under an overlay consisting of 50% of Leibovitz L-15 medium supplemented with 10% FBS and 50% of carboxyl methyl cellulose . Cells were then fixed with 3 . 6% formaldehyde at room temperature ( RT ) for 20 min and a immunoperoxydase assay staining was performed to detect foci . After a first incubation of 4 min with PBS 0 . 1% Triton X-100 ( Sigma ) at RT , cells were incubated 20 min at 37°C with a mouse ascitic fluid at a dilution of 1∶1000 for CHIKV and 1∶100 for DENV-1 ( both ascetic fluids were provided by the French National Reference Center for Arbovirus at the Institut Pasteur ) . After a wash in PBS 1X , cells were incubated at 37°C for 45 min with peroxydase-conjugated goat anti-mouse IgG antibody ( Pierce biotechnology , Rockford , USA ) at a 1∶100 dilution in PBS 1X . After final wash in PBS 1X , Fast 3 , 3′ Diaminobenzidine ( Sigma ) was used to reveal the staining and foci were counted . The titer of infectious particles per saliva was expressed as FFU/mL ( mean ± standard deviation ) . After salivation , females were tested for the presence of CHIKV and DENV-1 by IFA on their head squashes [32] . CHIKV and DENV-1 antigens were detected with the same mouse ascitic fluids used for saliva titration . Head squashes being performed between two slides , infection status of females fed with both viruses , could be checked for both antigens by using one slide for each IFA . Mosquitoes inoculated with CHIKV and DENV-1 were used as positive controls , negative controls were inoculated with cell culture media .
As shown on Table 1 , only few females inoculated with DENV-1 were willing to feed on an artificial blood-meal in the BSL-3 insectarium . When a blood-meal was proposed at day 7 post-inoculation , eight females out of 106 fed and among them , three survived until day 13 post-inoculation . By IFA on head squashes , we found that all 3 females had disseminated both viruses . Besides , when a blood-meal was offered at day 13 post-inoculation , eight females out of 54 females fed and four survived until day 20 post-inoculation . These four females had disseminated both viruses . Control females inoculated or orally infected by only one virus were all positive . Females were exposed to both viruses in a same blood-meal and disseminated infection rates were estimated at day 14 post-infection ( pi ) ( Table 2 ) . In the trial 1 , 71 . 6% of females have only disseminated CHIKV , 0% only DENV-1 , 18 . 6% both viruses and 9 . 8% did not disseminate any virus . In the trial 2 , 30 . 8% of females have only disseminated CHIKV , 7 . 7% only DENV-1 , 50 . 8% both viruses and 10 . 7% did not disseminate any virus . When providing a higher titer of DENV-1 in the blood-meal ( trial 2 ) , a higher proportion of females were co-infected with both viruses . For females having disseminated both viruses , saliva was collected at day 14 pi and titrated . Relative transmission of the two viruses are shown in Table 3 . In the trial 1 , out of 19 saliva , 4 contained simultaneously CHIKV and DENV-1 , 4 only CHIKV , 3 only DENV-1 and 8 no virus . In the trial 2 , out of 33 saliva , 9 presented concomitantly CHIKV and DENV-1 , 8 only CHIKV , 2 only DENV-1 and 14 no virus . Mean titers , expressed as FFU per saliva , and standard deviation are shown on Table 4 .
We report here the ability of Ae . albopictus from La Reunion Island to replicate simultaneously autochthonous strains of DENV-1 and CHIKV provided in the same blood-meal and to deliver both infectious viral particles in their saliva . To our knowledge , such co-infection has never been shown neither under laboratory conditions nor in the field . Lastly , we succeeded in inducing a secondary infection with CHIKV 7 or 13 days after a first infection with DENV-1 virus . CHIKV and DENV are both transmitted by Ae . aegypti and Ae . albopictus , the former being considered the major vector and the latter , the secondary vector . However Ae . albopictus is able to sustain DEN outbreaks in the absence of Ae . aegypti [5] . Indeed , in the Indian Ocean , Ae . albopictus was predominant in Seychelles and in La Reunion Island where the species took part of DEN outbreaks in 1976-77 [34] , [35] and in 2005 [6] . On La Reunion Island , Ae . aegypti populations are scarce and do not exhibit a high anthropophily [16] , [17] while Ae . albopictus has favored the emergence of a new CHIKV strain harboring a substitution ( alanine → valine ) at the position 226 of the E1 glycoprotein during the 2005–2006 outbreak . This variant presents high levels of replication in Ae . albopictus [19] and a short extrinsic incubation period as the virus could be found in saliva as early as two days after infection [36] . Subsequent outbreaks due to the new CHIKV variant were often related to transmission by Ae . albopictus corroborating an adaptative mutation in response to a requirement for transmission by this species . Moreover , co-infections with both DENV and CHIKV have been detected in some patients from La Reunion Island [28] and Madagascar in 2006 [7] . Co-infections CHIKV-DENV in patients have been first described in 1967 and since the emergence of CHIKV in the Indian Ocean , reports of co-infections are increasing: in the Indian Ocean , as mentioned above , but also in Sri-Lanka [23] , Malaysia [26] , in India [24] , [25] where the main vector involved is Ae . aegypti and in Gabon [27] . Except for the Americas still free of CHIK infection , the geographic range of CHIK is now largely overlapping that of DEN . Furthermore , the emergence of CHIK outbreaks due to the new variant coincided with the recent establishment of Ae . albopictus in Central Africa , in Cameroon [37] , [38] and Gabon [39] , [40] . In 2007 , patients with co-infections DENV-CHIKV were indeed detected for the first time in Africa in Gabon [27] . Co-infection of a mosquito vector by two different viruses can occur by the way of two successive infectious blood-meals taken on two different viremic hosts or by one blood-meal taken on a co-infected host . We chose to mix both viruses in the same meal since Ae . albopictus females from La Reunion Island were reluctant to feed twice at 7 or 13 days interval on an infectious meal in the BSL-3 conditions . If Ae . albopictus females can be readily fed on an artificial meal in a BSL-3 laboratory it is very difficult to make them take a second artificial meal in the same conditions after 1 or 2 weeks of incubation and even to make then take a first meal if they have been kept more than 48 h in the depression conditions of the BSL-3 . Temperature and humidity were optimal , females had the opportunity to lay their eggs if needed , were starved prior blood-feeding as usual . Therefore , we were unable to offer viruses in sequential meals and choose to test the possibility of a superinfection by first injecting one virus ( DENV-1 ) then offering an infectious blood meal with the second virus ( CHIKV ) 7 or 13 days latter . Even then , we had very few individuals blood-fed but could however detect replication and dissemination of both viruses . Ae . albopictus from La Reunion Island were already known to be a very efficient vector for the new CHIKV variant and able to sustain DEN outbreaks . We demonstrated that replication of both viruses can take place simultaneously and vectors become able to transmit the two viruses in a single bite . In Ae . albopictus , CHIKV infectious particles can be found in saliva from two days after oral infection [36] when this delay is much longer for DENV at least 10 days [41]–[42] . We performed the saliva analysis 14 days after feeding the females on a co-infected blood-meal but were not able to detect infectious viral particles in all saliva collected from females presenting a disseminated infection of both viruses . As CHIKV was not detected , when it should have been considering its rapid excretion in saliva [36] , females with negative saliva must have been females unable to salivate using our technique [36] or excreting a very small amount of infectious particles that we could not detect due to the low sensitivity of the technique of viral detection . It would be interesting to perform a kinetic study of excretion of infectious particles from females exposed to one or both viruses while controlling the presence of saliva in the collected samples . We performed our study with Ae . albopictus from La Reunion Island with two viral strains isolated in the same area during the DENV-1 outbreak of 2005 and the CHIKV outbreak of 2005–2006 . This combination vector/virus fits well with the natural context strengthening the findings of this work . Nevertheless , it seems difficult to infer to other epidemiological contexts as shown by the failure of Rohani et al . [29] to prove co-infection of Ae . aegypti by CHIKV and DENV . The mechanism of co-infection in Ae . albopictus by two different arboviruses needs to be further investigated and tested using additional trials with different viral strains to determine if this phenomenon is an exception due to the particularly well adapted partners or a quite common mechanism . The still ongoing expansion of this species , particularly in Africa where numerous arboviruses are transmitted , in of particular concern and , if the co-infection was a quite usual phenomenon could have great implication on human health . It should be noted that superinfection is possible in mosquitoes as well as in cells infected with heterologous viruses ( i . e . , different genus ) and not with homologous ones [43]–[47] . Little is known about the molecular and cellular basis of co-infection which should be explored .
We are grateful to the DRASS in the Reunion for providing the ALPROV mosquito strain and to the French National Reference Center for Arboviruses for the E1-226V CHIKV and the DENV-1 strains . We also wish to thank Sara Moutailler for critical reading of the manuscript . We are grateful to Michèle Bouloy and Félix Rey for their support . | Dengue ( DEN ) and chikungunya ( CHIK ) are two mosquito borne infections transmitted by Aedes mosquitoes in the tropical world . Ae . albopictus has been shown to efficiently transmit the new variant of CHIK virus ( CHIKV ) that emerged in the Indian Ocean region in 2005 . At the same time , this vector is able to sustain outbreaks due to DEN virus ( DENV ) . Since this CHIK emergence , co-infections DENV-CHIKV in humans have been regularly reported . This phenomenon , known for a long time , may be due to two consecutive bites from two mosquitoes infected by one virus or by the bite of a mosquito infected by both viruses . We used two viral strains isolated in La Reunion Island , DENV-1 in 2004 and CHIKV in 2005 , and co-infected an autochthonous strain of Ae . albopictus , testing experimentally one of the possible ways to get co-infections in humans . We were able to show the ability of Ae . albopictus to replicate simultaneously both arboviruses and , furthermore , to deliver both infectious viral particles concomitantly in their saliva . This finding is of particular interest since Ae . albopictus is now widely distributed all around the world and still expanding its geographical range . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Acknowledgments"
] | [
"infectious",
"diseases/neglected",
"tropical",
"diseases"
] | 2010 | Orally Co-Infected Aedes albopictus from La Reunion Island, Indian Ocean, Can Deliver Both Dengue and Chikungunya Infectious Viral Particles in Their Saliva |
Brucella are highly infectious bacterial pathogens responsible for a severely debilitating zoonosis called brucellosis . Half of the human population worldwide is considered to live at risk of exposure , mostly in the poorest rural areas of the world . Prompt diagnosis of brucellosis is essential to prevent complications and to control epidemiology outbreaks , but identification of Brucella isolates may be hampered by the lack of rapid and cost-effective methods . Nowadays , many clinical microbiology laboratories use Matrix-Assisted Laser Desorption Ionization–Time Of Flight mass spectrometry ( MALDI-TOF MS ) for routine identification . However , lack of reference spectra in the currently commercialized databases does not allow the identification of Brucella isolates . In this work , we constructed a Brucella MALDI-TOF MS reference database using VITEK MS . We generated 590 spectra from 84 different strains ( including rare or atypical isolates ) to cover this bacterial genus . We then applied a novel biomathematical approach to discriminate different species . This allowed accurate identification of Brucella isolates at the genus level with no misidentifications , in particular as the closely related and less pathogenic Ochrobactrum genus . The main zoonotic species ( B . melitensis , B . abortus and B . suis ) could also be identified at the species level with an accuracy of 100% , 92 . 9% and 100% , respectively . This MALDI-TOF reference database will be the first Brucella database validated for diagnostic and accessible to all VITEK MS users in routine . This will improve the diagnosis and control of brucellosis by allowing a rapid identification of these pathogens .
Brucella are important pathogens in medical and veterinary context . These Gram-negative bacteria can be transmitted from their animal reservoir to humans , usually by ingestion of contaminated milk products or direct contact , causing brucellosis . This zoonosis causes a severely debilitating illness characterized by intermittent fever , chills , sweats , weakness , myalgia , osteoarticular or obstetrical complications and endocarditis . This disease is largely unreported and the true incidence of human brucellosis is thus unknown [1] . According to the World Health Organization ( WHO ) , half a million new cases are reported each year , most of them in the poorest rural areas of the world [2] . Indeed , while the disease has been successfully prevented in most industrialized countries , it remains a significant burden in the Mediterranean region , all over Asia , sub-Saharan Africa , and certain areas in Latin America . Approximately half of the human population worldwide is considered to live at risk of exposure [3] . Moreover , due to the low dose required to cause infection ( 10–100 colony-forming units ) and the potential for aerosol dissemination , Brucella was considered a potential bioterrorism agent early in the 20th century [1] and its possession and use is still strictly regulated in many countries . Currently , the Brucella genus consists of eleven recognized species plus several isolates that have not yet been officially designated . The major zoonotic species are B . melitensis , B . abortus and B . suis which are subdivided into biovars by a set of phenotypic characteristics including lipopolysaccharide ( LPS ) epitopes , phage sensitivity , dye sensitivity and a battery of biochemical tests . These three species are also the most common in domestic livestock . B . melitensis is responsible for the majority of human cases in the Mediterranean basin , the Arab peninsula , Latin America countries and Asia , while B . abortus is more prevalent in the United States , Northern Europe and Africa [4] . B . suis and B . canis infections are more sporadic in humans . Very rare human infections have also been reported with B . inopinata [5 , 6] , B . ceti [7 , 8] and B . neotomae [9 , 10] . Clinical microbiology laboratories play a key role in the diagnosis and management of human brucellosis and should be able to provide a rapid and exact identification of Brucella spp . Currently , the most suitable tool for identification of bacteria is Matrix Assisted Laser Desorption/Ionization-Time of Flight Mass Spectrometry ( MALDI-TOF MS ) . This method provides rapid , sensitive and cost-effective identification and is currently replacing phenotypic microbial identification . Its accuracy however largely depends on the coverage of the database of the commercially available MALDI-TOF MS systems . With regards to Brucella , identification was not possible because this genus was not represented in the databases of the two main MALDI-TOF MS system manufacturers ( i . e . bioMérieux and Bruker ) [11–13] . Only the Bruker Security Relevant ( SR ) database , or custom databases developed in some laboratories , can identify these highly pathogenic bacteria , but access to these databases is not possible in some countries due to export restriction regulations [13–15] . Moreover , only B . melitensis is included in the SR database .
The bacterial strains used for the construction of the database are listed in Tables 1 and S1 . Each of these strains was cultivated on several different media ( S1 Table ) . The bacterial isolates used for the external evaluation and their culture conditions are listed in Tables 2 and S2 . All strains used in this study were previously characterized using an established workflow ( phenotypic assays , Multiple-Locus Variable number tandem repeat Analysis -or MLVA- , whole-genome sequencing ) [16] . Samples used to build the spectra database were prepared according to a previously established inactivation protocol [19] consisting in resuspending two full loops of bacteria ( i . e . multiple colonies ) in 200 μL of solvent mix , vortexing ( 10 sec ) , centrifuging ( 10 , 000 g , 2 min ) at room temperature , removing 190 μL and resuspending in the 10 μL of solvent left in the tube . For the external evaluation study , this protocol was simplified by suspending only one loop of bacteria in 100 μL of solvent mix , vortexing ( 10 sec ) and incubating at room temperature ( 20–25°C , 3 minutes ) . Bacteria were efficiently inactivated by this method and the biomass concentration of the samples allowed identification by MALDI-TOF MS , demonstrating that the centrifugation step in the original protocol was not required . One μL of each sample was applied to a single well of a disposable , barcode-labeled target slide ( VITEK MS-DS , bioMérieux ) , overlaid with 1 μL of a saturated solution of alpha-cyano-4-hydroxycinnamic acid matrix in 50% acetonitrile and 2 . 5% trifluoroacetic acid ( VITEK MSCHCA , bioMérieux ) then air dried . For the database construction , several independent measurements were recorded for each strain ( see S1 Table for the different culture conditions ) . For instrument calibration , an Escherichia coli reference strain ( ATCC 8739 ) was directly transferred to designated spots on the target slide using the procedure recommended by the manufacturer . Mass spectra were acquired using a VITEK MS Plus ( bioMérieux , Marcy l’Etoile ) and the Launchpad v2 . 8 software program ( Kratos , Shimadzu group Compagny , Manchester , UK ) . Dendrograms showing taxonomic relationships between strains were constructed using the SARAMIS software ( bioMérieux , Marcy l’Etoile , France ) . The database was built as previously described [20] . Briefly , peak lists were binned by assigning each peak within the mass range of 3 . 000–17 . 000 Da to one of 1 , 300 bins . A predictive model was then established for each species using the Advanced Spectra Classifier ( ASC ) algorithm developed by bioMérieux ( La Balme les Grottes , France ) . The outcome of this procedure provided an assignment of a dimensionless weight for each bin and for each species . As a result , a specific pattern of weights for the 1 , 300 bins was obtained and combined for all species in a weighted bin matrix . For optimization , the spectral data were partitioned into 5 complementary subsets . One round of cross-validation involved a learning phase on 4 subsets ( “training set” ) and a validation of the predictive model on the remaining subset ( “testing set” ) . Five rounds of cross-validation were performed by permutation , and the results from the five rounds combined . To assess the accuracy of the database and calculate its performance in cross-validation , individual spectra were re-used as template for identification . The ASC algorithm compares the acquired spectrum to the specific pattern of each organism/organism group in the database and calculates a percent probability , or confidence value ( %ID ) , which represents the similarity in terms of presence/absence of specific peaks between spectra . A perfect match provides a %ID of 99 . 9% . %ID >60 to 99 . 8% are considered as good . Scores <60% are considered to have no valid identification . The VITEK MS system renders the following types of identification results: “Single Choice” , when the spectrum acquired presents a high level of similarity ( %ID >60 to 99 . 9% ) with only one specific pattern in the database; “Low discrimination” , when the spectrum acquired presents a high level of similarity with 2 to 4 specific patterns in the database; or “No Identification” , when the spectrum acquired either does not match with any pattern in the database , or presents a high level of similarity to more than 4 specific patterns . During cross-validation , identification was considered as correct when the result was consistent with the reference identification . Low discrimination results were considered as correct if the expected identification was included in the matches . A misidentification was defined as discordant organism identification between the cross-validation result and the reference identification . External spectra were generated from bacteria cultivated with different growth conditions ( media , incubation time , etc ) to mimic possible inter-laboratory variations . To reflect clinical laboratory practice , inactivated samples were spotted in duplicate , and analyzed with the updated database . If only one of the two spectra allowed a correct identification , the isolate was considered correctly identified . The cut-off for identification confidence was as described above .
To update the MALDI-TOF MS VITEK database , we used 84 Brucella strains , either reference strains or well characterized clinical/veterinary isolates ( Tables 1 and S1 ) , to generate independent spectra covering the Brucella genus . After initial selection based on quality criteria such as peak resolution , signal to noise ratio , number of peaks , absolute signal intensity , and intra-specific similarity , 590 spectra were retained and submitted for biomathematical analyses using an iterative system ( bioMérieux patented ASC algorithm ) . Using an optimization process , we next evaluated the possibility to discriminate between different Brucella species and biovars . Discrimination between the different species was obtained , with the exception of B . ceti and B . pinnipedialis . These two species could not be clearly separated , as illustrated by the intertwining of their spectra on a dendrogram ( Fig 1 ) . Distinguishing the different biovars of B . melitensis and B . abortus was not possible . Discrimination between several of B . suis biovars was obtained ( S1 Fig ) , but biovars 1 and 4 gave cross-identifications . Classes representing the different Brucella species were thus created by grouping together the different biovars of B . melitensis , of B . abortus and of B . suis , and the two species B . ceti and B . pinnipedialis . The eight species represented in the MALDI-TOF database are thus: B . melitensis ( biovar 1 , 2 or 3 ) , B . abortus ( biovar 1 , 2 , 3 , 4 , 5 , 6 or 9 ) , B . suis ( biovar 1 , 2 , 3 , 4 or 5 ) , B . canis , B . ovis , B . ceti/B . pinnipedialis , B . inopinata and B . papionis . After optimization , cross validation was performed to evaluate the performance of the updated database , which contains 37 , 902 spectra covering 1 , 095 bacterial species including Brucella . This mathematical method is used to assess how accurately the database can perform . Correct identification at the genus level was obtained in 97 . 29% of cases ( Table 3 ) . Importantly , the remaining 2 . 71% of results corresponded to “no ID” , but never to an incorrect identification . At the species level , the performance varied between the different classes . For the three main zoonotic species ( B . melitensis , B . abortus and B . suis ) , correct identification was obtained with 96 . 06% , 100% or 89 . 34% of spectra , respectively . Finally , as an external validation , the database was challenged with the MALDI-TOF spectra from 48 independent Brucella isolates , and 2 strains of Ochrobactrum , which are “near neighbors” of the Brucella genus ( Tables 2 and S2 ) . The implemented database allowed correct identification at the genus level in 88 . 4% of cases , all the other results being “No-identification” but never misidentification as another genus ( Tables 4 and 5 ) . At the species level , the performances varied . For B . melitensis , B . abortus , and B . suis , correct identification was obtained for 100% , 92 . 3% or 100% of strains , respectively . It should be noted however that only one extra B . suis isolate was available to be tested in the external validation . Interestingly , the rare clinical isolate 02/611 , described as B . ceti-like after molecular characterization [7] , was indeed identified within the B . ceti/B . pinnipedialis class . Also , both the Bullfrog ( B13-0095 ) and the Australian rodent ( NF2637 ) isolates were identified as B . inopinata , in agreement with previous work showing that these belong to the atypical Brucella clade of this genus [27 , 28] . The two isolates belonging to “B . abortus biovar 7” , a rare biovar of this species , were identified as B . abortus . Finally , the recombinant 16M strain overexpressing the green fluorescent protein ( GFP ) was correctly identified as B . melitensis using this database . Moreover , using different culture conditions for 16M did not affect its identification by MALDI-TOF MS ( S3 Table ) .
A major asset of this MALDI-TOF MS database is its ability to identify Brucella isolates at the species level , which is essential for following epidemiological outbreaks . Obtaining such a resolution was very challenging for this genus , as highlighted in previous studies [29] , because of the high similarity between species at the genetic level [30] . Discrimination between species was made possible using a patented approach to differentiate closely related species using internal calibration and a two-step algorithm . This was not sufficient to distinguish the two species of Brucella from marine mammals ( B . ceti and B . pinnipedialis ) . This is in agreement with a recent Multi-Locus Sequence Analysis ( MLSA ) showing that the taxonomy is inconsistent with the phylogeny of these two species , and that taxonomic rearrangement should be envisaged [31] . This MALDI-TOF MS database is however able to discriminate eight different Brucella species , which include the most common in human or animal disease . The updated database allowed correct identification of Brucella isolates at the genus level in 88 . 4% of cases . It is important to mention that none of them was identified as Ochrobactrum spp . , a misidentification that is common with other standard identification methods [32–34] and recently reported using the VITEK MS database currently available [35] . Analysis at the species level gave only one discordant result , corresponding to cross identification between two Brucella species . Such result would have no consequence for human medicine , as identification at the genus level is sufficient to prescribe the appropriate treatment . As for all MALDI-TOF databases , the limitation of this system is its inability to identify non-clinically validated species or species not included in the database . However , the large coverage of the Brucella genus ( in particular the most common species ) in this database makes this risk is very minor . Diminution of the performance at the genus and/or species level was due to “no ID” results for some rare and/or atypical Brucella spp . ( B . neotomae strain 5K33 , B . microti strain CCM4915 , and the rodent isolate NF2653 ) , several strains from marine mammals , and the vaccine strain B . abortus RB51 . These results were not due to the quality of MALDl-TOF spectra , which was good ( based on the number of spectral peaks , Table 5 ) . In the spectra for RB51 , we found that several masses characteristics of the B . abortus class were less frequently present , in particular the masses of 5 , 920 . 63 , 6 , 040 . 32 and 7 , 467 . 89 Da were present in only 14 . 3% of spectra ( vs . in 75–95% of the spectra of other B . abortus isolates , with a tolerance of 800 ppm ) . The only discordant result in our assay was obtained with B . canis Mex51 , which was identified as B . suis . This was due to the presence in its spectra of additional masses that are common with the B . suis class in addition to the major peaks characteristics of the B . canis class . This finding is consistent with an exhaustive MLSA showing that B . canis strains are very close to B . suis biovars 3 and 4 [31] . Importantly , the MALDI-TOF database allowed the correct identification as Brucella of several recently discovered “atypical” isolates [5 , 6 , 28 , 36] . These strains represent a serious problem for diagnosis laboratories , as they are not identified as Brucella using classical phenotypic tests . It is possible that similar strains have been isolated in the past but misidentified . Very little is known concerning the ability of these new species to cause disease in humans or livestock . The possibility to identify these isolates as Brucella will thus be important for both human and animal health . Overexpression of an exogenous protein ( GFP ) did not affect the identification of B . melitensis 16M . This is important since recombinant Brucella strains are common tools in research laboratories and could potentially infect lab workers . Moreover , the use of such Brucella strains as vaccines was proposed , since the presence anti-GFP antibodies would allow distinguishing vaccinated animals from naturally infected ones [37] . In conclusion , this updated MALDI-TOF MS database is a new diagnostic tool that allows the identification of Brucella . It combines precision of identification ( broad coverage of the Brucella genus together with species-level identification ) and widespread availability . After integration in the VITEK MS ( v3 . 2 ) , this will be the first Brucella database validated for diagnostic with CE accreditation and accessible to all users in routine . This will allow accurate diagnosis and timely treatment in brucellosis . These highly infectious pathogens also causing one of the most frequent laboratory-acquired infection [38] , their rapid identification by MALDI-TOF MS will decrease the risk of accidental infection of laboratory workers . A paradox of global health however is that the countries where brucellosis is endemic may not have access to MALDI-TOF MS . This could be circumvented by the use of the in-tube inactivation method described earlier [19] , which will allow the shipment of erstwhile infectious samples to mass spectrometry platforms . | Brucella are bacteria that mainly infect animals . They can also be transmitted to humans and cause a serious disease called brucellosis . Half the world's population is considered exposed , especially in the poorest rural areas . Experts agree that prompt identification of Brucella isolates is essential to provide appropriate treatment to patients and to control epidemiological outbreaks . Mis-identification of these highly infectious pathogens may lead to delays in diagnosis , but also to increased risks of accidental exposure for laboratory workers . MALDI-TOF mass spectrometry is now the first line of bacterial identification in many routine diagnostic laboratories . However , not all clinical mass spectrometers can identify Brucella . In this work , we updated a database with Brucella spectra to improve the performance of MALDI-TOF mass spectrometers . These instruments will now be able to identify accurately Brucella isolates . This will greatly improve the diagnosis of brucellosis . | [
"Abstract",
"Introduction",
"Material",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"ruminants",
"pathogens",
"tropical",
"diseases",
"microbiology",
"vertebrates",
"animals",
"mammals",
"brucellosis",
"marine",
"biology",
"bacterial",
"diseases",
"... | 2018 | A MALDI-TOF MS database with broad genus coverage for species-level identification of Brucella |
The function of most proteins is not determined experimentally , but is extrapolated from homologs . According to the “ortholog conjecture” , or standard model of phylogenomics , protein function changes rapidly after duplication , leading to paralogs with different functions , while orthologs retain the ancestral function . We report here that a comparison of experimentally supported functional annotations among homologs from 13 genomes mostly supports this model . We show that to analyze GO annotation effectively , several confounding factors need to be controlled: authorship bias , variation of GO term frequency among species , variation of background similarity among species pairs , and propagated annotation bias . After controlling for these biases , we observe that orthologs have generally more similar functional annotations than paralogs . This is especially strong for sub-cellular localization . We observe only a weak decrease in functional similarity with increasing sequence divergence . These findings hold over a large diversity of species; notably orthologs from model organisms such as E . coli , yeast or mouse have conserved function with human proteins .
Understanding the relation between gene evolution and function is perhaps our only hope of bringing functional annotation in line with the furious pace of genomic sequencing . Indeed , despite developments in high-throughput experimental techniques , propagation of functional knowledge from evolutionarily related genes remains the procedure that scales best and appears most dependable [1] . The simplest model for this assumes that function is conserved among homologs , which motivates a process that assigns function by sequence similarity . A canonical refinement of this model distinguishes orthologs from paralogs [2] , [3] . As gene duplication is considered an important source of functional innovation , the “standard model” posits that orthologs tend to have a conserved function , whereas paralogs tend to diverge in function [4] . Yet , large-scale studies corroborating this standard model are surprisingly scarce [5] . Furthermore , sequence similarity seems to be a better predictor of function conservation than orthology [6] . This suggests an alternative model , that orthologs versus paralogs might not be the primary clue to functional similarity . With the recent availability of genome-wide reliable orthology predictions on the one hand , and systematic , standardized functional annotations on the other , we now have the ability to test these models on a broad and representative sample of biological data . Recently , Nehrt et al . [7] have proposed such a test of the “ortholog conjecture” ( i . e . , the “standard model” ) , using human and mouse functional annotations . Surprisingly , they find that paralogs appear more functionally similar than orthologs . In the present study , we investigated the functional similarity of 395 , 328 pairs of orthologs and paralogs with experimental GO annotations [8] for both genes , from 13 genomes ( see Materials and Methods ) . After controlling for confounding factors which we describe in detail below , we find that—contra Nehrt et al . [7]—current experimental annotations do support the “ortholog conjecture” , albeit not as strongly as might have been expected .
GO annotations—even restricting to experimentally supported ones—are heterogeneous in many ways , such as type of function described , level of specificity , applicable species , method of investigation , or curation practices [9] . Therefore , to meaningfully compare GO annotations , it is essential that potential confounding factors be controlled . In this section , we describe and address four confounding factors ( Fig . 1 ) : ( i ) authorship bias , ( ii ) variation of GO term frequency among species , ( iii ) variation of background similarity among species pairs , and ( iv ) propagated annotation bias . To our knowledge , the effect of these factors has not been clearly reported previously . Correcting for the biases described above , we first restricted our comparison to experimental annotations with no common investigator from the two yeast species , Saccharomyces cerevisiae and Schizosaccharomyces pombe . They were chosen because ( i ) they form the pair of species with the most ortholog pairs which both have experimentally supported GO annotations; ( ii ) they are quite similar in biology , and are studied by scientific communities with similar interests; and ( iii ) since horizontal gene transfer is relatively rare in eukaryotes , the distinction between orthology and paralogy is conceptually straightforward . Thus , we hope to minimize organism specific annotation biases . Function similarity was computed using an information-theoretic measure taking into account the variation of annotation coverage among species ( Fig . 1B ) , and normalizing with respect to the background similarity of random gene pairs ( Fig . 1C; for details , see Materials and Methods ) . The first observation is that at similar levels of sequence divergence , one-to-one orthologs do have significantly more similar experimental GO annotations than paralogs , and that one-to-many and many-to-many orthologs ( referred to as “other orthologs” in the remainder of the text ) are somewhat intermediary ( Fig . 2A ) ( Kruskal-Wallis test between homology types , p<2 . 2 10−16; t-test of 1∶1 orthologs vs . other homologs , p<2 . 2 10−16 ) ; this is consistent with the ortholog conjecture . The difference of excess similarity between one-to-one orthologs and other homologs is considerable ( average functional similarity of 0 . 36 vs . 0 . 20 ) . Also consistent with expectations , there is almost no difference between same-species paralogs and different-species paralogs ( t-test , p = 0 . 029; difference of 8% ) . It is difficult to tell whether this small difference is biologically relevant , or whether it corresponds to some residual species-specific annotation bias . On the other hand , the difference between orthologs and paralogs is not as important as might have been expected under a naive interpretation of the ortholog conjecture: orthologs are far from having almost the same function . This might stem in part from the differences between experiments performed by different investigators . Most surprising , the decrease in annotation similarity with protein divergence is very weak ( Spearman correlation between sequence identity and GO similarity over all homologs: ρ = −0 . 019 , p = 0 . 009 ) . This contradicts the predominant notion that “sequence divergence is generally accompanied by higher likelihood of divergence in function” [13] . We have verified these results with a number of additional controls: using different metrics of GO annotation similarity ( Fig . S3 ) ; using different metrics of protein divergence ( Fig . S4 ) ; and using only gene quartets with orthologs and paralogs in both species ( Fig . S5 ) . In all cases , we recover the higher functional similarity of orthologs than paralogs , and the low correlation between annotation similarity and protein divergence . Furthermore , to assess the significance of the difference between orthologs and paralogs for each level of sequence divergence , we performed bin-by-bin non-parametric Mann-Whitney U tests ( Fig . S6 ) . All tests that are significant at the 99% confidence level showed an excess of similarity of orthologs over paralogs ( Table S3 ) . The GO is composed of three orthogonal ontologies , which we have analyzed separately for the two yeasts . The Cellular Component ontology shows the most marked pattern , with a very clear excess of similarity between one-to-one orthologs , relative to all other homologs ( Fig . 2B; t-test , p<2 . 2 10−16; difference of 57% ) . Orthologs are also very significantly more similar for Biological Process ( Fig . 2C; t-test , p<2 . 2 10−16; difference of 41% ) , whereas for Molecular Function the difference is weaker ( Fig . 2D; t-test , p = 1 . 6 10−7; difference of 30% ) . The difference is a bit stronger for Molecular Function if all orthologs are contrasted to all paralogs ( t-test , p = 3 . 9 10−12 ) , but it remains weaker than for the other two ontologies . One inherent limitation of two-species analyses is that all pairs of orthologs started diverging at the same time ( the speciation event between the two species ) , with almost all paralogs being either older ( the “out-paralogs” ) or younger ( the “in-paralogs” ) than the orthologs . By considering sequences from many different gene families—some of which faster evolving , other slower evolving—we can compare orthologs and paralogs that have similar levels of sequence divergence , but inevitably , slow-evolving orthologs will tend to be compared with in-paralogs , while fast-evolving orthologs will tend to be compared with out-paralogs . To avoid the potential bias that this might introduce , we need to look at data from multiple species . We performed the same comparisons between all possible pairs of the 13 species with sufficient experimental GO annotations . Results are widely consistent with the yeast only study ( Fig . 3; Fig . S7 ) : at similar levels of sequence divergence , orthologs , and especially one-to-one orthologs , are more similar in GO annotations than paralogs , although the absolute difference is modest . Likewise , the difference between same-species paralogs and different-species paralogs is still quite modest ( t-test , p<2 . 2 10−12; difference of 10% ) . We also confirm that the excess similarity of orthologs vs . paralogs is strongest for the Cellular Component ontology . With this larger and more diverse dataset , the excess similarity of orthologs is also highly significant for the Molecular Function ontologies ( all orthologs vs . all paralogs , t-test , p<2 . 2 10−16 ) , as for the Biological Process ( Fig . 3 ) . To assess the significance of the difference between orthologs and paralogs for each level of sequence divergence , we also performed bin-by-bin non-parametric Mann-Whitney U tests ( Fig . S8; Table S3 ) . They were significant and consistent with the general trend of orthologs more functionally similar than paralogs for all but very divergent sequences ( 10–20% range of sequence identity ) , where there is a slight excess of similarity for paralogs ( difference of 0 . 0250 , p-value = 0 . 00022 ) . But it should be noted that 10–20% identity is well into the twilight/midnight zone , where even homology calling is difficult , let alone orthology/paralogy calling . As additional controls , we confirmed that our results are not sensitive to the choice of bin size ( Fig . S9 ) , function similarity measure ( Fig . S10 ) , or overrepresented gene families ( Fig . S11 , 12 ) . Furthermore , the results are also supported by analyses performed on Ensembl compara data , an alternative source of orthologs/paralogs sequence pairs ( [14]; Fig . S13 ) . Like for the yeast study , there is little correlation between functional similarity and protein sequence identity ( Spearman ρ = −0 . 023 , p<2 . 2 10−16 ) . The correlation with species divergence time is also very weak ( Spearman ρ = −0 . 052 , p<2 . 2 10−16 10−12; computed only on orthologs; Fig . S14c ) . A potential confounding effect is that only well-conserved proteins can be detected as homologs between distantly related organisms . To control for this effect , we compared annotations of orthologs conserved among triplets of genomes ( Fig . 4 ) . For the human-mouse-fly triplet , functional similarity is stronger between human and mouse than with fly . But for triplets involving yeast or E . coli , functional similarity is the same between human or mouse and the third genome , as between human and mouse . Of note , the GO similarity of human or of mouse to the outgroup is always extremely similar , despite using independently generated annotations . Enzyme commission ( EC ) numbers are an alternative source of functional annotations . The relation between EC numbers and sequence divergence has already been studied extensively ( e . g . , [15] ) , especially before GO supplanted EC as the main source of functional annotations , but is restricted to genes with catalytic activity . In relative terms , the functional similarity of orthologs and paralogs in terms of EC numbers behaved like experimental GO annotations , with 1∶1 orthologs showing the highest level of similarity , other orthologs a somewhat lower level , and paralogs the lowest level ( Fig . S13 ) . In absolute terms , however , average functional similarity for all categories was generally higher and decreased much more distinctly with decreasing percentage identity ( Spearman ρ = 0 . 45 p<10−16 ) . As we note above , computational propagation of functional annotations inflates functional similarity in absolute terms . Since there is no evidence code used for EC annotations , most of the comparison is based on computational propagation . This could also explain the stronger decrease , as propagation preferentially takes place between homologs close in sequence .
The distinction between orthologs and paralogs has been a central concept of phylogenomics [3] . And yet , it is only recently that the functional relevance of this distinction has been treated as a hypothesis to be tested . To date , several indirect , sequence-based studies have failed to support this classical model , rather supporting an alternative model of uniform functional divergence , independent of duplication [reviewed in 5] . Recently , Nehrt et al . [7] have compared the functional annotations of orthologs and paralogs between human and mouse . Surprisingly , they report the strongest functional similarity for paralogs , which is expected neither under the classical model nor under the uniform model . Directly comparing functional annotations is complicated , because they are derived from a variety of sources and by a variety of procedures . The best-known bias is that computationally derived annotations ( IEA code ) are generally believed to be less reliable than experimentally derived annotations . The computational annotations reflect the algorithms used to propagate annotations [16] , and thus are shared preferentially among proteins with high sequence similarity , among orthologs , or among proteins sharing well-defined domains . Any analysis including these GO annotation will recover the impact of these algorithms , which is indeed what we find when we use all GO annotations . Much of the older literature on function divergence used the EC nomenclature as a measure of function , and thus mixed indiscernibly electronic and experimental annotations . Thus it is probable that most results based on the EC nomenclature are biased by electronic annotations ( i . e . , Fig . S15 ) . Even limiting ourselves to experimentally derived annotations , there remains a great deal of complexity and bias in the data of functional annotation . First , different model organisms are studied by different scientific communities , for different purposes , which bias the types of experiments conducted and reported . Moreover , each organism is predominantly annotated by one Model Organism Database team , which differs from others in its data curation and annotation practices . Indeed , we observe significant differences in background functional similarity , depending on the species compared . While part of this variation might be due to biological differences among the species , these differences appear to be mostly due to the artifacts outlined above . Here , we have compared 13 organisms spanning the tree of life ( Fig . S17; Table S4 ) , and we have corrected each comparison by the background frequencies of annotations from the relevant genomes . Moreover , we show that results limited to two yeasts are consistent with results averaged over all organisms . Second , each experiment is performed and reported by a given team of investigators , who have a scientific focus and a manner of reporting which are specific to them . This induces a strong bias towards similar annotations derived from the same paper , which mostly affects same-species paralogs . Importantly , there is a bias towards similar annotations even when considering different papers which share at least one co-author . Unless accounted for , this confounding factor leads to a large spurious excess of similarity between same-species paralogs [similar to the results of 7] . Controlling for it leads to the opposite conclusion: a weak excess of similarity between orthologs ( Fig . S16 ) . This observation is also corroborated by a recent rebuttal from the GO consortium , which reexamined two case studies from Nehrt et al . 's paper and concluded that the difference in function similarity computed between orthologs and paralogs was mainly due to bias in annotations , not in the underlying functions [9] . While GO annotations are complex and biased , it nevertheless appears possible to identify and correct these biases , and to detect biologically significant signal . We feel that the use of 13 different species , with diverse annotation levels and evolutionary distances , contributes to the robustness of our results . Once the biases identified above are accounted for , the signal which emerges can be summarized in three major points: ( i ) Consistent with the “ortholog conjecture” , or “standard model of phylogenomics” , overall functional similarity is highest between one-to-one orthologs , lowest between paralogs , and intermediate between other orthologs . ( ii ) There is at best a very weak relation between protein sequence similarity and functional similarity . ( iii ) The difference between orthologs and paralogs , although consistent with the ortholog conjecture , is weaker than expected under a naive understanding of that model; this is especially true when Molecular Function and Biological Process are considered separately . The standard model of higher functional similarity among orthologs than paralogs at similar levels of sequence divergence could not be supported until it was explicitly tested [5] . Several recent studies have performed such tests , and found some measure of support for the standard model . On a structural level , there appears to be higher conservation of intron position [17] , of protein structure [18] , and of domain architecture [19] between orthologs . Presumably more relevant to biological function , the conservation of expression patterns appears higher between orthologs than between paralogs , in mammals [20] . On the other hand , Nehrt et al . [7] have found that the expression correlation of human/mouse inparalogs is significantly higher than that of orthologs ( but not outparalogs ) . And a study of the evolution of sub-cellular localization in yeasts did not find any difference between orthologs and paralogs [21] . These contradictory results might be due in part to the overall modest difference between orthologs and paralogs , and in part to differences between different aspects of function . An intriguing pattern in our results is that we find strong conservation of Cellular Component annotations among orthologs . Contrary to the two other ontologies , sub-cellular localization is an aspect of function which leaves little room for divergent interpretation . Moreover , experimental results are easier to report in similar terms in different species . These factors might allow better detection of the excess conservation of orthologs . Thus , of the 3 ontologies , our results on cellular components are arguably the most conclusive . As for the two other aspects of protein function captured by the Gene Ontology—Molecular Function and Biological Process—they have more subtle patterns . Molecular Function shows an excess of conservation between orthologs which is weaker than for Cellular Component , but which is strongly significant over all 13 genomes analyzed . This is the aspect of function for which there was previously the most evidence for the “uniform model” of no significant difference between orthologs and paralogs; with the available data , this can now be rejected . This is also the aspect of function for which the absolute value of excess similarity ( i . e . , excess similarity of homologs over random pairs ) is strongest—for both orthologs and paralogs . Thus , Molecular Function appears to be strongly conserved between even distant homologs , which supports the received wisdom of predicting this type of annotation on the basis of conserved protein domains . Biological Process also has a significant excess of function conservation among orthologs , although weaker than for the Cellular Component . This is surprising , given the wide differences in biology between the species compared . Indeed , throughout the entire range of sequence divergence , orthologs are considerably more similar in function than even same-species paralogs . Of note , the biases which amplify apparent similarity between paralogs are strongest for this aspect of function: not correcting for the sampling bias of orthologs or paralogs detected between species can lead to a spurious excess of conservation of same-species paralogs . Our results contradict the concept of the evolution of cellular context set forth by Nehrt et al . to explain the apparent higher similarity of function of in-paralogs between human and mouse [7] . This concept was also related to the weak relation between protein sequence divergence and functional divergence . Nehrt et al . [7] speculated that protein function might evolve more as a function of the divergence of cellular context than as a function of protein sequence . They suggested that a comparison of orthologs of different ages might recover an effect of divergence age on functional divergence . Our analysis includes species divergences spanning the range from 36 Mya to 3300 Mya , yet we still do not find a strong relation between functional divergence and protein divergence , nor with species divergence time . These observations suggest that protein function evolves in a very non-clock-like manner . Indeed , clock-like evolution is an expected pattern for neutrally evolving characters [22] , whereas selection is expected to be the major force shaping the evolution of protein function . The low impact of evolutionary time on average protein function conservation is also apparent if we compare humans to model organisms with very different divergence times . Indeed , the extent of functional similarity of one-to-one orthologs is similar between human and E . coli , human and yeast , human and fly , or human and mouse . This supports the strong relevance of these various species for understanding human biology . In fact , the average similarity over all available one-to-one orthologs is even higher for the more distant E . coli and yeast , than for fly or mouse . This is probably due the fact that only proteins with very strong function conservation are kept as detectable one-to-one orthologs over such long evolutionary spans . We verified this by comparing only proteins which are detected as one-to-one orthologs in triplets of these species . For human-mouse-fly , we do recover a stronger similarity for more closely related species . But for the triplets with yeast or E . coli , this is not the case . In terms of evolutionary biology , this shows that , to some extent , protein function does diverge with time . Yet there is a class of proteins , conserved beyond animals , which conserve their function , irrespective of divergence time , on average . In terms of annotation procedures for databases , and even design of new experiments , these results show that if a protein is conserved between two species , as one-to-one ortholog , then its function is probably mostly conserved , even if the divergence time is very large . In conclusion , our analyses corroborate the central tenet of the standard model of phylogenomics—that at similar levels of sequence divergence , orthologs are in general more similar in function than paralogs . But although significant , the difference is modest , and is uneven among different aspect of function ( among different ontologies ) . Furthermore , our results expose other trends unexplained by the standard model , such as differences among subtypes of orthology and paralogy ( also observed in other contexts , such as intron conservation [17] ) , or the lack of interaction between sequence and function divergence . Hence , the standard model has validity , but is of only limited practical use . To further progress in our understanding of the relation between gene evolution and gene function , we need to move beyond the orthology/paralogy dichotomy .
We selected 13 genomes with highest coverage in GO annotations backed by experimental data ( evidence codes EXP , IDA , IEP , IGI , IMP , and IPI ) . The annotations were retrieved from the GOA database [16] release 73 and Ensembl [23] release 54 . We used orthologs and paralogs inferred by OMA [24] , [25] . For each species pair , we extracted all the one-to-one orthologs , all other orthologs ( one-to-many and many-to-many ) , all out-paralogs ( within and between the species pair ) and all inparalogs ( in this context by definition only within the same species ) . Only gene pairs with more than 10% identity in amino-acid sequence were kept . This yielded a total of 9 , 564 , 666 pairs of genes . Of those , 395 , 328 had experimental GO annotations for both genes . We computed the similarity of these experimental annotations using several measures ( see below ) , with evolutionary distance in percent sequence identity computed over the total protein sequences as independent variable . We used the EC number assignments of the ENZYME database , maintained by Swiss-Prot [26] . We used orthologs and paralogs induced by Ensembl Compara gene trees ( version 65 ) [23] together with GO annotations from GOA ( release 2012-01-21 ) as an alternative dataset ( Fig . S13 ) . The comparison of gene annotations requires a measure of semantic similarity . In recent years , several measures have been proposed ( for review , [27] ) . In the present context , 3 aspects of these metrics are most relevant: ( i ) how to compute the similarity between two GO terms , ( ii ) how to deal with multiple terms for a given gene , and ( iii ) how to normalize the measure across species . For each GO annotation an evidence code and a reference identifier is recorded . In the case of experimental annotations ( EXP , IDA , IEP , IGI , IMP and IPI ) , this reference id is usually a PubMed identifier or a reference id from a model organism database ( MOD ) . We extract authors associated with a given GO annotations by first mapping non-PubMed reference ids to PubMed ids using publicly available mapping files from the MODs . Second , for each PubMed id we extract the authors of that publication from the PubMed webpage . | To infer the function of an unknown gene , possibly the most effective way is to identify a well-characterized evolutionarily related gene , and assume that they have both kept their ancestral function . If several such homologs are available , all else being equal , it has long been assumed that those that diverged by speciation ( “ortholog” ) are functionally closer than those that diverged by duplication ( “paralogs” ) ; thus function is more reliably inferred from the former . But despite its prevalence , this model mostly rests on first principles , as for the longest time we have not had sufficient data to test it empirically . Recently , some studies began investigating this question and have cast doubt on the validity of this model . Here , we show that by considering a wide range of organisms and data , and , crucially , by correcting for several easily overlooked biases affecting functional annotations , the standard model is corroborated by the presently available experimental data . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"gene",
"function",
"genome",
"analysis",
"tools",
"phylogenetics",
"sequence",
"analysis",
"comparative",
"genomics",
"biology",
"proteomics",
"evolutionary",
"systematics",
"biochemistry",
"genetics",
"genomics",
"evolutionary",
"biology",
"computational",
"biology",
"gen... | 2012 | Resolving the Ortholog Conjecture: Orthologs Tend to Be Weakly, but Significantly, More Similar in Function than Paralogs |
Plant organ growth is controlled by inter-cell-layer communication , which thus determines the overall size of the organism . The epidermal layer interfaces with the environment and participates in both driving and restricting growth via inter-cell-layer communication . However , it remains unknown whether the epidermis can send signals to internal tissue to limit cell proliferation in determinate growth . Very-long-chain fatty acids ( VLCFAs ) are synthesized in the epidermis and used in the formation of cuticular wax . Here we found that VLCFA synthesis in the epidermis is essential for proper development of Arabidopsis thaliana . Wild-type plants treated with a VLCFA synthesis inhibitor and pasticcino mutants with defects in VLCFA synthesis exhibited overproliferation of cells in the vasculature or in the rib zone of shoot apices . The decrease of VLCFA content increased the expression of IPT3 , a key determinant of cytokinin biosynthesis in the vasculature , and , indeed , elevated cytokinin levels . These phenotypes were suppressed in ipt3;5;7 triple mutants , and also by vasculature-specific expression of cytokinin oxidase , which degrades active forms of cytokinin . Our results imply that VLCFA synthesis in the epidermis is required to suppress cytokinin biosynthesis in the vasculature , thus fine-tuning cell division activity in internal tissue , and therefore that shoot growth is controlled by the interaction between the surface ( epidermis ) and the axis ( vasculature ) of the plant body .
The epidermis is formed from the outermost L1 layer in the shoot apical meristem ( SAM ) and functions as an important interface with the environment . However , recent studies have shown that it also plays an essential role in the establishment and maintenance of the plant body . Arabidopsis mutants with defects in epidermal cell specification exhibit disorganized morphology [1]–[3] . Biophysical manipulation of the epidermis revealed that it generates mechanical constraints on inner layers , thus restricting plant growth [4] , [5] . Another report showed that epidermis-specific expression of brassinosteroid receptor ( BR ) or brassinosteroid biosynthesis enzyme rescued plant growth in dwarf mutants , indicating that a BR-generated signal from the epidermis promotes the growth of ground tissue [6] . These results suggest that the epidermis participates in both driving and restricting growth via inter-cell-layer communication . However , it remains an open question as to whether the L1 layer can send signals to internal tissue to control cell proliferation during development . A characteristic feature of the epidermis is that it is covered with a hydrophobic barrier , the cuticle , which prevents plants from transpiring and protects tissues from pathogen attack [7] . The cuticle is mainly composed of cutin matrix and cuticular wax; cutin is a plant-specific lipid polymer that consists of long-chain fatty acids ( LCFAs ) with an acyl chain length of 16 or 18 carbons , whereas cuticular wax contains very-long-chain fatty acids ( VLCFAs ) with fully saturated unbranched hydrocarbon chains ( ≥20 carbons ) . Plant VLCFAs are synthesized in the endoplasmic reticulum by sequential addition of 2-carbon moieties to the 18-carbon LCFA , which is made in the plastid . The carbon donor malonyl-CoA is synthesized from acetyl-CoA by acetyl-CoA carboxylase and used for each cycle of the elongation reaction . VLCFA synthesis consists of four enzymatic steps: ( 1 ) condensation of acyl-CoA with malonyl-CoA catalyzed by ketoacyl-CoA synthase ( KCS ) , ( 2 ) reduction of 3-ketoacyl-CoA by 3-ketoacyl-CoA reductase ( KCR ) , ( 3 ) dehydration of 3-hydroxyacyl-CoA by 3-hydroxy acyl-CoA dehydratase ( HCD ) , and ( 4 ) reduction of enoyl-CoA by enoyl-CoA reductase ( ECR ) . VLCFAs are also components of seed storage triacylglycerols and sphingolipids; in yeast and mammalian cells , the latter function as signaling molecules controlling cell proliferation , stress response , and programmed cell death [8] . Arabidopsis mutants with defects in VLCFA synthesis display cuticular deformation , leading to alteration of pathogen-plant interactions [9] , post-embryonic organ fusion [10] , [11] , and retardation of plant growth with abnormal morphology [12] . PASTICCINO2 ( PAS2 ) is the Arabidopsis gene encoding HCD , one of the enzymes involved in VLCFA synthesis [13] . A loss-of-function mutant of PAS2 displays embryo lethality , and the leaky mutant pas2-1 , which contains reduced amounts of VLCFAs , cuticular wax , and sphingolipids , exhibits severe morphological defects [13]–[17] . However , cuticular deformation cannot explain all of these phenotypes; in particular , the cause of defective overall growth is not well understood . Here we show that VLCFA synthesis in the epidermis is essential for plant growth , and that it suppresses cell proliferation by targeting cytokinin biosynthesis in the vasculature , thus fine-tuning cell division activity in determinate growth . Our results suggest that the epidermis sends non-autonomous signals to the vasculature and suppresses overproliferation .
pas2-1 mutant seedlings exhibit various morphological defects to a variable extent in individual plants; for example , true leaves are fused ( Figure 1A ) [14] , [15] , [17] , and hypocotyls are swollen and possess more cortical cell layers ( Figure 1B and 1C ) [14] . Previous reports also noted that mutant leaves sometimes produce a callus-like structure as a result of increased cell proliferation [14]–[17] . We therefore observed the SAM , which gives rise to organs like leaves and flowers . We found that more cells accumulate in the rib zone ( RZ ) —a region below the self-renewing stem cell pool that contributes to the meristem pith—and that the vasculature was disorganized ( Figure 1D ) . We monitored the expression pattern of PAS2 using the ProPAS2:β-glucuronidase ( GUS ) reporter gene ( a fusion of the ∼2 . 0-kb PAS2 promoter and the GUS gene ) . GUS signal was detected in mature embryos , cotyledons and true leaves of seedlings , the inflorescence stem , and pistils and anthers of flowers ( Figure S1 ) . In tissue sections , GUS expression was observed only in the L1 layer of the SAM and in the epidermis of young leaves and the inflorescence stem ( Figure 2A–2C ) . In situ RNA hybridization also indicated L1-specific expression ( Figure 2D ) . To examine protein-level expression , we generated the ProPAS2:PAS2–GUS reporter gene ( the same promoter and the full-length PAS2 coding region fused in-frame to GUS ) ; the functionality of the PAS2–GUS fusion protein was tested as described below . GUS expression was again detected in the L1 layer , and faint expression was noted in the vasculature ( Figure 2E ) . To test whether PAS2 expression in the epidermis is necessary and sufficient for normal plant development , we downregulated PAS2 by RNAi using the ATML1 promoter , which drives L1-specific expression [18] . The expression level of PAS2 was reduced in the transgenic plants compared to wild-type ( Figure 3A ) . Although the phenotypes were highly variable , we could find pas2-1-like phenotypes in eight of the 44 transgenic lines , such as swollen hypocotyls , fused leaves , and retarded growth ( Figure 3B ) . Moreover , in transgenic lines showing no macroscopic phenotype , we observed overproliferation of vasculature cells and enlargement of the RZ , the latter of which appeared to be mainly due to enhanced cell expansion ( Figure 3C ) . As mentioned above , the ProPAS2:PAS2–GUS reporter gene showed faint expression in the vasculature; thus , to verify that PAS2 expression in the epidermis is sufficient for proper development , we introduced the RNAi construct under the procambial ATHB8 promoter [19] . As a result , no pas2-1-like phenotype was found among 79 transgenic lines , supporting the epidermis-specific role of PAS2 in plant development . We then introduced ProATML1:PAS2–GUS into the pas2-1 mutant . GUS expression was specifically observed in the L1 layer ( Figure 3E ) , and five of the six transgenic lines displayed fully rescued phenotypes ( Figure 3D ) , indicating the functionality of PAS2–GUS . On the other hand , when ProATHB8:PAS2–GUS was introduced into pas2-1 , none of the 19 homozygous mutants were rescued . These results demonstrate that VLCFA synthesis in the epidermis is essential and sufficient for proper plant development . In pas2-1 , the cuticular wax content is severely reduced [13]; it is thus difficult to distinguish the outcome of defective cuticular formation from other effects arising from low VLCFA content . Therefore , we examined dose-dependent phenotypes in wild-type seedlings using the synthetic inhibitor cafenstrole , which blocks the first step of VLCFA elongation reactions by targeting KCS [20] . Our recent study showed that cafenstrole treatment of Arabidopsis seedlings reduced the content of C22 and C24 fatty acids , although our experimental conditions did not allow us to detect C26 or longer fatty acids [21] . Seedlings treated with 3 µM cafenstrole displayed severe growth retardation with swollen hypocotyls and fused leaves , as observed in pas2-1 ( Figure 4A and 4B ) . Those treated with 30 nM cafenstrole did not show growth inhibition , but instead produced larger leaves with thicker hypocotyls ( Figure 4A ) . Measurements of leaf area and cell size showed that , in 12-d-old seedlings , leaf blade area increased 1 . 7-fold after 30 nM cafenstrole treatment ( 14 . 4±2 . 5 mm2 for the control and 24 . 0±5 . 0 mm2 for the cafenstrole treatment; mean ± standard deviation [SD] , n≥11 ) , whereas cell size did not change significantly ( 830±55 µm2 for the control and 852±158 µm2 for the cafenstrole treatment ) ( Figure S2 ) . Cell number also increased 1 . 7-fold in cafenstrole-treated leaves ( 17 , 265±2 , 467 for the control and 29 , 297±6 , 765 for the cafenstrole treatment ) ( Figure S2 ) , accounting for the 1 . 7-fold enlargement of leaf blades . In shoot apices , 30 nM cafenstrole caused more cells to accumulate in the vasculature ( Figure 5A ) , and , as a result , cell number in the vasculature of the hypocotyl dramatically increased ( Figure 4C ) . To examine cell division activity , we used the ProCDKB2;1:NT–GUS reporter ( comprising the CDKB2;1 promoter and the first CDKB2;1 exon [NT] fused in-frame to GUS ) , which monitors mitotic cells during the G2 and M phases [22] . The number of GUS-stained cells increased when the cafenstrole concentration was elevated , especially in the region along the vasculature ( Figure 5A ) . These results demonstrate that a mild reduction of VLCFA content ( with 30 nM cafenstrole or in PAS2 RNAi plants ) enhances proliferation of vasculature cells , while a severe reduction ( with 3 µM cafenstrole or in pas2-1 ) causes overall growth retardation with impaired cuticular formation , as described below . Transmission electron microscopy revealed that an electron-dense cuticular layer disappeared in pas2-1 , and that only a trace of cuticular layer was formed in wild-type seedlings treated with 3 µM cafenstrole ( Figure 4D ) . On the other hand , a thicker cuticle was formed in the presence of 30 nM cafenstrole ( Figure 4D ) ; thus , it is unlikely that cell proliferation was enhanced as a consequence of reduced cuticle synthesis . This idea is supported by the observation that Arabidopsis mutants specifically impaired in cuticular formation did not display enhanced cell proliferation , as described later . Moreover , the expression of the L1-specific reporter ProPDF1:GUS [23] retained its L1 specificity in pas2-1 ( Figure S3A ) , indicating that epidermal identity is maintained under low-VLCFA conditions . We quantified phytohormone content , and found that levels of the cytokinins isopentenyladenine ( iP ) and trans-zeatin ( tZ ) , and of their ribosylated and phosphorylated precursors ( iPR , iPRPs , tZR , and tZRPs ) , increased in pas2-1 and in wild-type treated with 30 nM or 3 µM cafenstrole ( Table 1 ) . This indicates that active cytokinins are highly synthesized in pas2-1 and after cafenstrole treatment . Indeed , expression of the primary cytokinin response marker ARABIDOPSIS RESPONSE REGULATOR 6 ( ARR6 ) [24] was stimulated by cafenstrole treatment in vascular bundles ( Figure 5B and 5C ) . Moreover , 30 nM cafenstrole did not enlarge leaves , but instead slightly reduced the cell number and the leaf size , in ipt3;5;7 triple mutants , in which cytokinin levels are severely decreased because of defects in cytokinin biosynthetic isopentenyltransferases ( Figure 6 ) [25] . This finding suggests that cytokinin is associated with the cafenstrole-induced activation of cell division . A higher level of cytokinin would also explain the previously observed hypersensitivity of pas2-1 to cytokinin treatment [14] , [17] . On the other hand , the content of indoleacetic acid ( IAA ) and gibberellins ( GA1 and GA4 ) did not increase , except that IAA became elevated in the presence of 3 µM ( but not 30 nM ) cafenstrole ( Table 1 ) . To further examine whether low VLCFA content is responsible for higher cytokinin level and enhanced cell proliferation , we next used Arabidopsis mutants with defects in LCFA and VLCFA synthesis ( Figure 7A ) . In mutants of PAS3 and PAS1 , which encode acetyl-CoA carboxylase and a scaffold protein for the elongase complex , respectively , VLCFA content is dramatically reduced and , as a result , organ growth is severely inhibited [14] , [26]–[28] . As observed in pas2-1 , these mutants contained higher amounts of tZ and iP compared to wild-type , and more cells accumulated in the RZ ( Figure 7B and 7C ) , indicating an enhancement of cell division . A recent report demonstrated that glossyhead1 ( gsd1 ) , another mutant allele for PAS3 , did not show severe growth inhibition [29] . However , overproliferation of vasculature cells was observed in the shoot apex of gsd1 , as in the case of 30 nM cafenstrole treatment and PAS2 RNAi plants ( Figure S4 ) . FIDDLEHEAD ( FDH ) /KCS10 encodes one of the 21 KCSs in Arabidopsis , and is thus associated with VLCFA synthesis; however , in the fdh-13 mutant [2] , only a mild leaf phenotype and a small reduction in C24 fatty acids were reported , probably due to redundancy in KCS genes [30] . Correspondingly , we detected a small increase of tZ level in seedlings , and a mild enhancement of cell proliferation in the RZ , but these phenotypes were less prominent than those in pas mutants ( Figure 7B and 7C ) . Note that the RZ in the control ( Ler ) was already larger than that in Col-0 ( Figure 7C ) . We also found a similar trend in the leaky mosaic death1 ( mod1-1 ) mutant , in which the activity of the LCFA-synthesizing enzyme enoyl-ACP reductase was reduced by half ( Figure 7A ) [31] . Although LCFA and VLCFA content in mod1-1 have not been reported so far , we noticed that tZ and iP levels increased slightly and that cell proliferation was enhanced in the RZ ( Figure 7B and 7C ) . This suggests that VLCFA content might be reduced as a result of decreased LCFA synthesis , but not as severely as in pas mutants , leading to modest effects on cytokinin level and cell division . The above results indicate that VLCFA synthesis in the epidermis is responsible for suppressing cytokinin biosynthesis and cell proliferation . We also observed three mutants with defects in cuticular wax formation from VLCFAs , cer4-1 , wax2 , and mah1-3 , which display impaired synthesis of primary alcohols , aldehydes , and secondary alcohols and ketones , respectively ( Figure 7A ) [11] , [32] , [33] . However , we found neither an increase of cytokinin level nor an enhancement of cell proliferation in these mutants ( Figure 7B and 7C ) . To identify the cause of higher cytokinin production under low-VLCFA conditions , we conducted microarray analyses and examined the expression levels of cytokinin biosynthesis genes . ( Microarray data have been deposited in the ArrayExpress database under accession number E-MEXP-3315 . ) In pas2-1 , the mRNA levels of IPT3 and CYP735A2 were 3 . 9- and 6 . 6-fold higher , respectively , than in wild-type ( Table S1 ) . IPT3 encodes one of the nine adenosine phosphate-isopentenyltransferases ( IPTs ) , which catalyze the first and rate-limiting step of cytokinin biosynthesis to produce isopentenyladenine riboside phosphates ( iPRPs ) [34] . CYP735A2 converts iPRPs to trans-zeatin riboside phosphates ( tZRPs ) [35] . We examined cytokinin levels and IPT3 expression in the pas2-1 mutant carrying ProATML1:PAS2–GUS , which rescued pas2-1 phenotypes ( Figure 3D ) . As described above , levels of tZ and iP were highly elevated in pas2-1 compared to those in wild-type ( Table 1 ) , but in pas2-1 carrying ProATML1:PAS2–GUS , no such increase of cytokinin content was detected ( tZ , 0 . 57±0 . 06 pmol/g fresh weight for Col-0 and 0 . 71±0 . 13 pmol/g for the transgenic line; iP , 0 . 49±0 . 03 pmol/g for Col-0 and 0 . 51±0 . 03 pmol/g for the transgenic line; mean ± SD , 7-d-old seedlings [n = 3] ) . The elevated level of IPT3 transcripts in pas2-1 was also reduced to the wild-type level by PAS2–GUS expression in the epidermis ( the relative mRNA level , with that for wild-type set to 1 , was 3 . 77±0 . 15 for pas2-1 and 0 . 81±0 . 09 for pas2-1 carrying ProATML1:PAS2–GUS; mean ± SD , 7-d-old seedlings [n = 3] ) . These results indicate that VLCFA synthesis in the epidermis is required to suppress not only cytokinin biosynthesis but also IPT3 expression . We then monitored IPT3 expression in 5-d-old seedlings using the promoter:GUS reporter . Consistent with a previous observation of IPT3 expression in the phloem [19] , we detected the GUS signal in vascular bundles ( Figure 8 ) . Cafenstrole treatment increased the intensity of the GUS signal and extended the expression domain in shoot apices and leaves; a similar expression pattern was also observed in pas2-1 ( Figure 8A and 8B ) . In cafenstrole-treated leaves , expression of the procambial marker ProATHB8:GUS [21] was restricted to vascular bundles ( Figure S3B and S3C ) , but IPT3 expression extended to spongy mesophyll cells ( Figure 8C ) . This indicates that low-VLCFA conditions increase IPT3 expression in the vasculature and cause ectopic expression in non-vascular cells . To examine whether increased cytokinin synthesis is a cause or a consequence of the overproliferation phenotype , we monitored IPT3 expression after transfer of 3-d-old seedlings to a medium containing 30 nM or 3 µM cafenstrole . We also observed the ProCYCB1;2:NT–GUS reporter ( comprising the CYCB1;2 promoter and the N-terminal region of CYCB1;2 [NT] fused in-frame to GUS ) , which monitors G2/M phase cells [36] . As shown in Figure S5 , higher IPT3 expression was noted after 6 h and 12 h for 3 µM and 30 nM cafenstrole , respectively , compared with the non-treated control . By contrast , CYCB1;2 expression increased from 12 to 24 h in the SAM and in young true leaves regardless of cafenstrole treatment ( Figure S5 ) , suggesting a general activation of cell division at this developmental stage . Expression was even higher after 48 h for both 30 nM and 3 µM cafenstrole , but not for the control ( Figure S5 ) , demonstrating that cafenstrole-induced overproliferation occurred later than 24 h . Measurement of cytokinin content revealed that cytokinin precursors , especially iPR and iPRPs , increased after 6 to 12 h of 3 µM cafenstrole treatment , and that iP and tZ increased after 24 h ( Table S2 ) . These results indicate that cafenstrole induces cytokinin synthesis , which is then followed by activation of cell division , implying that enhanced cytokinin synthesis is the cause of overproliferation triggered by low-VLCFA conditions . The above results suggested an interesting hypothesis , namely , that VLCFA synthesis in the epidermis is required to confine cytokinin biosynthesis to the vasculature and prevent cells from overproliferating . To test this hypothesis , we examined whether the effect of cafenstrole is suppressed by reducing cytokinin levels . We expressed the gene for Venus-fused cytokinin oxidase 1 ( CKX1 ) , which degrades active forms of cytokinins [37] , under the control of ATML1 and ATHB8 promoters . Venus fluorescence showed that the ATML1 and ATHB8 promoters conferred epidermis- and vasculature-specific expression , respectively ( Figure S6A and S6B ) . We measured leaf area in four independent lines for each promoter construct , and found that 30 nM cafenstrole enlarged leaves in wild-type and ProATML1:CKX1–Venus , but that no such enlargement occurred in ProATHB8:CKX1–Venus ( Figures 9A , 9B , and S6C ) . The latter effect was due to the suppression of cafenstrole-induced enhancement of cell proliferation in leaves ( Figure 9B ) . Enhanced cell accumulation in the vasculature , expansion of hypocotyl width at the base of cotyledons , and an increase in vascular cell number in hypocotyls were also suppressed by CKX1–Venus expression in vascular bundles ( Figures 9C , S6D , and S6E ) . We also expressed CKX1–Venus in pas2-1 , but the macroscopic phenotype of the mutant was not suppressed with either promoter , probably owing to severely impaired cuticular formation . However , when shoot apices were observed microscopically , we found that the enhanced cell accumulation in the RZ and disorganization of the vasculature were partially suppressed with ProATHB8:CKX1–Venus , but not with ProATML1:CKX1–Venus in pas2-1 ( Figure 9D ) . These results indicate that , under low-VLCFA conditions , an increase of cytokinin biosynthesis in the vasculature is the major cause of overproliferation . The inability of ProATML1:CKX1–Venus to suppress overproliferation suggests that non-cell-autonomous factors ( other than cytokinins ) act from mesophyll cells to the epidermis to promote cell division , as reported previously [38] .
In this study , we showed that a higher concentration of cafenstrole ( 3 µM ) caused severe growth defects , notably swollen hypocotyls and fused leaves , which are similar to those observed in the leaky pas2-1 mutant . More cells accumulated in the RZ of pas2-1 , and the number of cortical cell layers increased in the hypocotyl . In contrast , at a lower concentration ( 30 nM ) , seedlings showed neither overall growth inhibition nor organ fusions; rather , the leaves were enlarged due to increased cell number . Enhanced cell proliferation was also observed in the vasculature in shoot apices , resulting in a dramatic increase of cell number in the vasculature of hypocotyls . When PAS2 expression was specifically downregulated in the epidermis , we could again observe disorganized vasculature due to enhanced cell proliferation . It is likely that , in pas2-1 and in wild-type plants treated with 3 µM cafenstrole , the stimulatory effect on cell division might be difficult to observe macroscopically , except for the swollen hypocotyl , due to impaired cuticular formation and consequent growth defects . However , a common feature was observed following mild or severe inhibition of VLCFA synthesis , namely , enhanced cell proliferation in the vasculature or in the RZ , respectively . It is noteworthy that , in pas2-1 , cell accumulation was prominent in the RZ but not in the vasculature . One possible explanation for this observation is that the amount of cytokinin in the vasculature may be so high that cell division is actually inhibited . The lower level of cytokinin in the RZ than in the vasculature may efficiently enhance cell proliferation . It is also probable that faster accumulation of RZ cells in pas2-1 suppresses cell division in the vasculature by intertissue communication . Further studies are needed to examine such possibilities . PAS2 has been identified as an antiphosphatase , which interacts with tyrosine-phosphorylated cyclin-dependent kinase A ( CDKA ) and prevents it from being dephosphorylated and activated [39] . However , in a pas2 mutant carrying phosphomimic mutations in CDKA , the phenotype of phosphomimic CDKA plants was not epistatic to the pas2 phenotype; rather , the two phenotypes were additive , indicating that PAS2 functions in parallel to CDKA [40] . This is supported by the fact that PAS2 is exclusively localized to the endoplasmic reticulum , whereas CDKA is distributed in both the nucleus and the cytoplasm [13] , [41] . Here , we observed enhanced cell proliferation in the RZ or in the vasculature not only in pas2 , but also in other VLCFA-related mutants , such as pas1 , pas3 , gsd1 , and fdh , as well as by treatment with the KCS inhibitor cafenstrole . Therefore , the overproliferation phenotype is not specifically linked to PAS2 functions , but instead is caused by inhibition of VLCFA synthesis . We revealed that , in pas2-1 and under cafenstrole treatment , IPT3 expression in the vasculature was elevated and its domain expanded to the spongy mesophyll cells . Indeed , the content of active cytokinins increased prior to the activation of cell division , whereas the overproliferation phenotype was suppressed in ipt3;5;7 mutants and by vasculature-specific degradation of cytokinins . These results demonstrate that VLCFA synthesis in the epidermis confines cytokinin biosynthesis to the vasculature and restricts cell proliferation . The idea that IPT3 is a possible target of epidermis-derived signals is supported by a previous report that overexpression of IPT3 in Arabidopsis resulted in a 3 . 4-fold increase of cytokinin content , and enlarged leaves with increased cell number [42] . However , it is also likely that CYP735A2 , whose expression increased 6 . 6-fold in pas2-1 , is another target . In Arabidopsis and rice plants , impairment of VLCFA synthesis elevates the expression of KNOTTED-like homeobox ( KNOX ) genes [17] , [43] . Moreover , overexpression of class I KNOX ( KNOXI ) genes is known to promote cytokinin synthesis in Arabidopsis [44] . However , it is unlikely that VLCFA synthesis in the epidermis restricts the cytokinin level by controlling KNOXI expression , because we observed overproliferation not only in the SAM but also in leaves , where KNOXI genes are not expressed . It is known that cytokinin induces expression of the KNOXI genes KNAT1/BP and STM [45] , suggesting that KNOX upregulation under low-VLCFA conditions results from increased cytokinin synthesis in the shoot apex . Our results indicate that epidermis-derived signals fine-tune cell division activity in internal tissue , suggesting that shoot growth is controlled by the interaction between the surface ( epidermis ) and the axis ( vasculature ) of the plant body ( Figure 10 ) . Indeed , perturbing this regulation by lowering VLCFA synthesis increased leaf size , demonstrating that non-autonomous signals are essential to restrict organ size . Arabidopsis mutants with defects in cuticular wax formation from VLCFAs ( cer4 , mah1 , wax2 ) did not exhibit phenotypes similar to those observed in pas mutants or cafenstrole-treated wild-type plants . Although we cannot exclude the possibility that some level of wax classes synthesized in these mutants suppresses the overproliferation phenotype , it is more likely that VLCFA derivatives function as signaling molecules to control cytokinin biosynthesis and cell division ( Figure 7A ) . In yeast and animals , sphingolipids made from VLCFAs act as signaling molecules controlling cell proliferation , cell death and stress responses [8] . Although Arabidopsis mutants defective in sphingolipid biosynthesis are impaired in cell growth , and in severe cases die [46] , [47] , some types of sphingolipids may control cell division by affecting cytokinin synthesis . It is also possible that VLCFA-containing lipids may function as mediators or ligands that control gene transcription , as suggested in mammals , yeast , and bacteria [48] . Indeed , arachidonic acid is known to induce stress-related gene expression and elicit defense signaling in Arabidopsis [49] . It is also likely that some metabolites , whose levels change depending on VLCFA synthesis , confine cytokinin biosynthesis to the vasculature . In Arabidopsis , 21 genes have been identified for KCS , which catalyzes the first step of VLCFA elongation reactions . Some of them are expressed predominantly in the epidermis , as observed for PAS2 [50] . Indeed , PAS2 , KCS11 , KCS16 , and KCS20 have one L1-box , and KCS6 , KCS9 , KCS10/FDH , and KCS18 have two L1-boxes , in their promoter regions , at which transcription factors ATML1 and PDF2 bind and control the L1/epidermis-specific gene expression [1] . Four kcs mutants , kcs6 , kcs10/fdh , kcs2 , and kcs20 , exhibit a glossy appearance and/or organ fusion , but no overproliferation phenotype was described [51]–[54] . However , we observed mildly enhanced cell proliferation in the RZ of fdh-13 , suggesting that other kcs mutants may also accumulate more cells in the RZ ( or in the vasculature ) than wild-type plants . Alternatively , specific VLCFAs may be required to suppress cell proliferation , because the distribution of various VLCFA species may depend on the substrate preference of each KCS ( e . g . , for a particular carbon chain length ) . HCD is unlikely to have such substrate preference , as it is encoded by the single-copy gene PAS2 in Arabidopsis; thus , the pas2 mutation reduces the overall level of VLCFAs including those that are responsible for suppressing cell proliferation . Further studies will reveal which VLCFA species or VLCFA-containing lipids are associated with non-autonomous signals to suppress cell proliferation in tissues . Previously , cyclin-dependent kinase inhibitor genes were ectopically expressed in the L1 layer , and meristem organization was investigated [55] . Cell number in the epidermis was reduced , while that in the cortex and mesophyll was the same as in wild-type . These and our present observations indicate that cell proliferation in shoot growth is not coordinated between the L1 and the inner layers; rather , it is controlled by VLCFA-derived signals that act on cytokinin biosynthesis in the vasculature . A likely benefit of this system is that plants can coordinate cuticular wax formation and organ growth , and can thus maintain proper development under various environmental conditions . Several KCS genes are induced by abiotic stress , such as salt , dehydration , and osmotic stress [50]; the transcription factor MYB30 , which activates expression of VLCFA biosynthesis genes , is also induced by pathogen infection [56] . Therefore , plants may deploy mechanisms to actively form cuticular wax and suppress cell proliferation to minimize energy consumption under stressful conditions . It is also known that BES1 , a downstream transcription factor in the brassinosteroid signaling pathway , directly interacts with and activates MYB30 [56] , [57] . While BES1 is involved in generating cell growth-promoting signal ( s ) from the L1 to inner layers [6] , brassinosteroid signaling in the epidermis may also control cell proliferation by activating MYB30 and VLCFA synthesis . Identification of non-autonomous signals will reveal how plants limit organ growth and adapt to changing environments by controlling cell growth and proliferation .
ProPDF1:GUS [23] and ProIPT3:GUS lines [34] , pas1-3 [27] , pas2-1 [14] , pas3-1 [14] , fdh-13 [2] , mod1-1 [31] , gsd1 [29] , and ipt3;5;7 [27] were described previously . Seeds of ProARR6:GUS ( N25262 ) , ProATHB8:GUS ( N296 ) , mah1-3 ( SALK_133155 ) , and wax2 ( SALK_020265 ) were obtained from the Arabidopsis Biological Resource Center . Seeds of cer4-1 ( N34 ) were obtained from the European Arabidopsis Stock Centre . All Arabidopsis plants used were in the Columbia ( Col-0 ) background , except that cer4-1 and fdh-13 were in the Landsberg erecta ( Ler ) background . To isolate the pas2-1 mutant , a genomic DNA fragment was amplified by PCR using a set of primers ( 5′-TCCACTGGTATCAGGGGAG-3′ and 5′-CTACTGAGAAGGAACCAATGATT-3′ ) , and treated with MvaI to observe the digestion pattern . Arabidopsis plants were grown in Murashige and Skoog ( MS ) medium ( 1× MS salts , 1× MS vitamins , 2% [w/v] sucrose , and 0 . 8% agar [pH 6 . 3] ) under continuous light conditions at 23°C . Cafenstrole ( HPLC standard grade , Wako Chemical ) was dissolved in dimethylsulfoxide at appropriate concentrations , and diluted 1 , 000-fold into the media . The 2-kb promoter fragment of PAS2 was PCR-amplified and cloned into the SalI-BamHI site of the pBI101 . 2 binary vector ( Clontech Laboratories ) to generate a fusion construct with GUS ( ProPAS2:GUS ) . The promoter and the coding region of PAS2 were PCR-amplified from genomic DNA and cloned into the Gateway entry vector pDONR221 ( Invitrogen ) by a BP reaction . An LR reaction was performed with the destination vector pGWB3 [58] to generate a binary vector carrying the fusion construct with GUS ( ProPAS2:PAS2–GUS ) . The 3 . 4-kb promoter fragment of ATML1 and the coding region of PAS2 were PCR-amplified from genomic DNA and cloned into the EcoRI and SmaI sites , respectively , of the pBluescript II KS ( - ) vector ( Stratagene ) . The resultant plasmid was digested with BamHI and HindIII , and the fragment was cloned into the HindIII-BamHI site of the pBI101 binary vector ( Clontech ) to generate ProATML1:PAS2–GUS . To make the PAS2 RNAi construct , the region encompassing nucleotides 6 to 641 of the PAS2 ORF was cloned into the EcoRI-KpnI and BamHI-HindIII sites of the pHANNIBAL vector [59] . The 35S promoter region in the vector was then replaced by the 3 . 4-kb ATML1 promoter , and the resultant ProATML1:PAS2RNAi fragment was cloned into the NotI site of the pART27 binary vector [60] . To express the CKX1–Venus fusion protein , the ORF of Venus and the genomic fragment comprising the coding region of CKX1 were PCR-amplified and tandemly cloned into the SalI site of pAN19 , a derivative of the pUC19 vector ( Invitrogen ) , to be in-frame with each other . The resultant CKX1–Venus construct was then PCR-amplified and cloned into pDONR221 by a BP reaction . The 3 . 4-kb and 1 . 7-kb promoter fragments of ATML1 and ATHB8 , respectively , were PCR-amplified and cloned into the Gateway entry vector pDONRP4-P1R ( Invitrogen ) by a BP reaction . An LR reaction was conducted with the destination vector pGWB501 [61] and the above-mentioned entry vectors to generate a binary vector carrying each promoter fragment fused to CKX1–Venus . To express PAS2–GUS and the PAS2 RNAi construct under the ATHB8 promoter , the fragments of PAS2–GUS and the PAS2 RNAi construct were PCR-amplified using the above-mentioned binary vectors with the ATML1 promoter , and cloned into pDONR221 by a BP reaction . These entry clones were used for an LR reaction with the destination vector pGWB501 and the entry vector pDONRP4-P1R carrying the ATHB8 promoter fragment . Primers used for plasmid constructions are listed in Table S3 . GUS staining and tissue sectioning were performed as described previously [62] . For counter-staining , sections were incubated with 0 . 05% ( w/v ) toluidine blue O . In the case of GUS-stained samples , sections were incubated with 0 . 05% ( w/v ) ruthenium red . Arabidopsis tissues were fixed in FAA ( 50% [v/v] ethanol , 5% [v/v] acetic acid , and 3 . 7% [v/v] formaldehyde ) , and 8-µm paraffin sections were hybridized with digoxygenin-labeled probes according to the protocol from the manufacturer ( Roche ) . The PAS2 probe was the antisense strand corresponding to the region 6 to 506 of the PAS2 ORF . For measurements of leaf blade area , healthy first leaves were harvested and fixed in a solution of 2 . 5% glutaraldehyde , and stored at 4°C . The area of edited microscopic images was measured using the image analysis program NIH ImageJ 1 . 43u ( http://rsb . info . nih . gov/nih-image/ ) . To measure cell size and cell number , data were collected by scanning images of the abaxial epidermis located at 50% of the distance between the tip and the base of the leaf blade , halfway between the midrib and the leaf margin . Images that included at least 40 cells in focus were edited using Photoshop Elements 6 ( Adobe , http://www . adobe . com/ ) . Epidermal cells in the edited image were counted , and the area of the edited image was measured with ImageJ . The average cell area was determined on the basis of these measurements . The total number of epidermal cells on the abaxial side was estimated on the basis of the average cell area and leaf blade area . For detection of Venus fluorescence , plant seedlings were embedded in 7% agarose and sliced manually with a razor , and sections were observed with a confocal laser scanning microscope ( LSM710; Carl Zeiss ) . Samples were fixed with 2 . 5% glutaraldehyde in phosphate buffer ( pH 7 . 0 ) at 4°C overnight , and then postfixed with 1% osmium tetroxide in the same buffer at 4°C for 1 h . Fixed samples were dehydrated in an ethanol series and embedded in Spurr resin , and polymerized at 73°C . Ultrathin sections were prepared with a diamond knife , stained with uranyl acetate and lead citrate , and observed with a JEOL 1200EX microscope . Sampling of about 100 mg of fresh whole seedlings was repeated three times . Extraction and determination of hormones were performed as described previously [63] . Data were processed by MassLynx software with QuanLynx ( version 4 . 0 , Waters ) . Total RNA was extracted from 3-d-old whole seedlings using TRIzol ( Invitrogen ) and purified with an RNeasy microkit ( QIAGEN ) as described in the manufacturer's instructions . GeneChip analyses were independently performed twice with the Arabidopsis ATH1 Genome Array ( Affymetrix ) as described in the GeneChip Expression Analysis Technical Manual ( Affymetrix ) . Probe synthesis was performed with the GeneChip 3′ IVT Express kit ( Affymetrix ) following the manufacturer's protocol . Hybridization and washes were performed as described in the GeneChip Expression Analysis Technical Manual . Signal detection and global normalization were performed using GeneChip Operating Software ( Affymetrix; version 1 . 4 ) with standard parameters . | The epidermis functions as an important interface with the environment , but in plants it is also essential for establishing and maintaining the primary plant body . Recent studies have shown that the epidermis participates in both driving and restricting plant growth via inter-cell-layer communication . However , it remains an open question as to whether the epidermis can send signals to internal plant tissues to control cell proliferation during development . Here we report that the synthesis of very-long-chain fatty acids ( VLCFAs ) in the epidermis is essential for the proper control of cell proliferation in the plant Arabidopsis thaliana . We find that defects in VLCFA synthesis cause cells in the vasculature or in the rib zone of shoot apices to overproliferate . When VLCFA levels decrease , we observe that the synthesis of the phytohormone cytokinin increases in the vasculature . We also find that when cytokinin is degraded by the expression of cytokinin oxidase in the vasculature , enhanced cell proliferation in internal tissues is suppressed , indicating that VLCFA synthesis in the epidermis is required to suppress cytokinin biosynthesis and thus cell overproliferation . Our results demonstrate that shoot growth is controlled by interactions between the surface ( epidermis ) and the axis ( vasculature ) of the plant body , and highlight a role for VLCFAs in this interaction . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"plant",
"science",
"model",
"organisms",
"plant",
"biology",
"biology"
] | 2013 | Synthesis of Very-Long-Chain Fatty Acids in the Epidermis Controls Plant Organ Growth by Restricting Cell Proliferation |
We are interested in how intragenic recombination contributes to the evolution of proteins and how this mechanism complements and enhances the diversity generated by random mutation . Experiments have revealed that proteins are highly tolerant to recombination with homologous sequences ( mutation by recombination is conservative ) ; more surprisingly , they have also shown that homologous sequence fragments make largely additive contributions to biophysical properties such as stability . Here , we develop a random field model to describe the statistical features of the subset of protein space accessible by recombination , which we refer to as the recombinational landscape . This model shows quantitative agreement with experimental results compiled from eight libraries of proteins that were generated by recombining gene fragments from homologous proteins . The model reveals a recombinational landscape that is highly enriched in functional sequences , with properties dominated by a large-scale additive structure . It also quantifies the relative contributions of parent sequence identity , crossover locations , and protein fold to the tolerance of proteins to recombination . Intragenic recombination explores a unique subset of sequence space that promotes rapid molecular diversification and functional adaptation .
The ubiquity of sex and recombination suggests a significant role in evolution , yet their benefit is still debated [1] , [2] . Intragenic recombination events generate chimeric genes , which are believed to make important contributions to allelic diversity in natural populations [3]–[6] . Laboratory experiments clearly demonstrate the benefits of recombining homologous proteins: intragenic recombination generates new proteins that are functionally diverse while still having a high probability of folding properly and functioning [7] , [8] . We have developed techniques for the design , construction , and characterization of libraries of chimeric proteins generated by site-directed recombination of homologous sequences [9]–[12] . Briefly , libraries are designed ( i . e . crossover sites are selected ) to minimize the number of novel residue contacts that are generated upon recombination ( we call this number ‘SCHEMA disruption’ ) , which tend to be deleterious to protein function . The sequence fragments chosen this way are then shuffled to generate a combinatorial library of chimeric proteins . The resulting proteins have no random point mutations; all the ‘mutations’ are homologous , that is , to amino acids that already appear in at least one of the parent sequences . These libraries can be used to explore the nature of recombination , without the high levels of random mutations typically found in protein libraries made by DNA shuffling [7] and other , similar methods for homologous recombination in vitro . To date , this laboratory has constructed and tested eight such recombination libraries consisting of chimeric bacterial -lactamases ( lac13 and lac ) , bacterial cytochrome P450s ( P450 ) , fungal family 5 cellulases ( Cel5 ) , bacterial family 48 cellulases ( Cel48 ) , fungal class I cellobiohydrolases ( CBHI ) , fungal class II cellobiohydrolases ( CBHII ) , and human arginases ( Arg ) ( Table 1 ) . Each library , which typically consists of thousands of new sequences , provides a glimpse of the protein fitness landscape that is accessible by recombination , which we refer to as the recombinational landscape . Since every member of the library can be generated by recombining other members , the genetic diversity in these libraries has similarities to that of a sexually reproducing population . Studies of these libraries have highlighted the enrichment of functional sequences in the recombinational landscape: SCHEMA-designed libraries contain a significant proportion ( 20–50% ) of functional sequences , despite having a high average mutation level ( i . e . average distance of a chimera sequence from its closest parent ) . For comparison , random mutation libraries with the same number of mutations are estimated to contain 10–20 orders of magnitude fewer functional sequences [13]–[15] . Whereas random mutations cause the probability a sequence remains functional to decrease exponentially , mutation by recombination always moves towards other functional sequences and is therefore significantly more conservative [16] . For this reason , intragenic recombination effectively explores functional ridges through a protein sequence space that is mostly nonfunctional . These libraries have also revealed significant variation in thermostability [17] , [18] and other properties [19]–[21] within the recombinational landscape . We observed that most of this variation can be explained by additive effects [17] , [18] , [20]–[22] . That is , the stability , for example , of a chimeric protein can be expressed as the sum of contributions from each of its sequence fragments . This additivity can be used to efficiently engineer highly optimized chimeric proteins for a variety of applications [17] , [20] , [22] , [23] . The additive structure , or lack of epistasis , within the recombinational landscape may provide an abundance of adaptive pathways for natural protein evolution . We would like to understand the features of the recombinational landscape that contribute to its extreme enrichment in functional sequences and its additive structure . Since the details of the protein recombinational landscape are unknown , we develop a random field model which captures its statistical properties . Random field models are effective at describing statistical features of uncertain , spatially-organized functions , with applications ranging from geostatistics to image analysis [24]–[26] . This versatile class of models has also been used to describe fitness landscapes [27] , the best known example being Kauffman's -model [28] . Our random field model for the recombinational landscape uses a physics-inspired energy function to describe the sequence-fitness relationship and is parametrized with experimental data . Using this model , we derive approximations for the fraction of functional sequences within a recombination library and the degree of landscape additivity , and we relate these quantities back to experimental observations . We discuss how the structure of the recombinational landscape contributes to the utility of recombination in evolution .
We use a pairwise , residue-level energy function to describe the large number of intramolecular interactions that stabilize protein structures ( Figure 1 ) . Such simplified contact potentials have been used in the past for protein folding simulations and structure prediction [29]–[31] . Assuming a fixed structure ( set of residue-residue contacts ) , the energy of any sequence is the sum of energy terms associated with the sequence's specific residue combinations at every pair of contacting residues . For chimeric proteins we distinguish between two types of contacts: parental ( P ) contacts , which are residue pairs observed in at least one of the parents , and novel ( N ) contacts , which are not ( Figure 1 ) . With this model , the energy of any chimeric protein is given by summing the contact energies ( 1 ) where is the energy term associated with parental contact , is the energy term associated with novel contact , and and are binary variables which indicate the specific residue pairs for each contact in chimeric protein . Since the specific energy values of and are unknown , we define the independent and identically distributed random numbers and , distributed with means and variances ( 2 ) ( 3 ) Substituting these random variables into equation 1 defines a random energy function associated with any chimeric protein ( 4 ) This random energy function is defined over the parental subspace , the set of all sequences that can be generated by recombining the parent sequences , which specifies the random field ( 5 ) The expected value of the random field at chimeric protein is ( 6 ) and the covariance between any two sequences is ( 7 ) Importantly , this covariance structure expresses how pairs of sequences are related and captures our intuitive notion of protein similarity: proteins with similar sequences have similar structures and therefore similar properties . This random field model provides a statistical description of the recombinational landscape . To parametrize the random field model , we must determine the mean energy and variance of parental contacts and the equivalent parameters and for novel contacts . Using a large binary functional status data set from a library made by recombining three bacterial cytochrome P450 enzyme heme domains [32] , these four parameters were estimated by maximizing a marginalized likelihood function ( see Methods ) . If we assume the functional status depends on a sequence's Gibbs free energy difference from the nonfunctional state , these estimated parameters can be interpreted as Gibbs free energy differences in units because the two-state Boltzmann distribution is identical to the logistic likelihood function . As expected , parental contacts are slightly stabilizing ( ) , novel contacts are significantly destabilizing ( ) , and both classes of contacts show significant variation relative to their means ( and ) . Estimating these parameters on recombination data from other protein families yields qualitatively similar relationships ( Figure S1 ) . This is not surprising , considering that most proteins are marginally stable [33] and mutations ( novel contacts ) tend to be deleterious to protein function [13]–[15] . In the following sections , we use this parametrized random field model to interpret experimental observations from protein recombination libraries . Previously , we compared the effects of random versus homologous amino acid substitutions [16] . Whereas the fraction of functional sequences declines exponentially with increasing random mutations [13] , [14] , that fraction varies log-parabolically with the number of substitutions taken from another functional parent . For two parents , the log-parabolic behavior appears because accumulating homologous substitutions must eventually convert one functional parent sequence into another functional parent sequence . Random mutagenesis of -lactamase indicated a probability that a single random mutation will preserve function ( neutrality ) of 0 . 54 , whereas recombination experiments on the same enzyme indicated the probability a homologous substitution will preserve function ( recombinational tolerance ) was 0 . 79 [16] . A recombinational tolerance significantly larger than the neutrality indicates that homologous substitutions tend to be more conservative than random ones . Here , we evaluate the effects of homologous substitutions using the random field model and compare the results to this previous analysis . Analyzing a library of chimeric -lactamases ( lac13 ) [34] , the probability of functioning for each chimera was estimated by evaluating the logistic function at the expected value of the random field ( equation 6 ) . These probabilities were averaged within 15 groups of chimeras binned by their number of homologous substitutions . The same analysis was also performed on simulated random substitutions , where a novel contact was any residue pair not present in the two -lactamase parents . With two parents , at least 18/19 random mutations will result in non-parental amino acids and therefore novel interactions with any contacting residues . The random field model results show excellent agreement with the experimental results of substitutions generated by recombination and randomly ( Figure 2A ) . As observed previously , the fraction of functional sequences undergoes a steep exponential decline with random mutations , while functionality displays a log-parabolic dependence on homologous substitutions . With the random field model , we can now explore how key recombination parameters , such as parent sequence identity or the number of sequence crossovers , influence the shape of the recombination curve shown in Figure 2A . As the sequence identity shared by the parents decreases , the curve stretches to a higher level of mutation ( more mutations are possible for a fixed sequence length ) and to a lower fraction functional ( Figure 2B ) , as was shown previously using lattice protein simulations [16] . Here we see that homologous substitutions from more-distant parents tend to be more deleterious to protein function than substitutions from less-distant proteins . This happens because distant proteins are more likely to have their contact networks composed of different residues , and these networks are therefore less compatible when recombined . We also see that as the number of sequence crossovers decreases , the log-parabolic recombination curve shifts towards a higher fraction functional ( Figure 2C ) , necessarily approaching a flat line when there are no crossovers . This happens because each crossover event creates opportunities to generate deleterious interactions . This improvement to the previous analysis allows us to see how recombinational tolerance depends on the number of sequence crossovers . To estimate the effects of homologous amino acid substitutions independent of the number of crossovers , we sampled random homologous substitutions and calculated the average probability of folding at each level of mutation ( Figure 2C ) . The effects of random homologous substitutions still follow the log-parabolic curve , although this curve dips over five orders of magnitude lower than the curve obtained from the -lactamase library experiments [34] . Fitting the log-parabolic equation [16] , we estimate the recombinational tolerance of random homologous substitutions to be . The recombinational tolerance is still significantly greater than the neutrality ( 0 . 54 ) , but to a lesser degree than previously estimated . The only difference between random homologous substitutions and those generated by recombination ( Figure 2C ) is how the mutations are distributed throughout the sequence and structure . Random homologous substitutions are distributed uniformly throughout the sequence , while those generated by recombination occur in contiguous stretches of sequence . By making mutations in groups , recombination preserves many local interactions . From this analysis , we propose an updated model for the conservative nature of intragenic recombination which includes contributions from homologous substitutions ( as shown previously ) as well as groups of coevolved residues that vary simultaneously . The latter effect is expected to be particularly important in natural evolution , where the number of intragenic crossover events per generation is likely to be small . It is interesting that the random field model for the recombinational landscape also works reasonably well to describe the effects of random mutations . Random mutations frequently result in a non-parental amino acid and therefore cause deleterious novel interactions with all contacting residues . This simplified model recapitulates the exponential decline in functional sequences that was observed upon random mutagenesis of -lactamase ( Figure 2A ) and in other mutational studies [13]–[15] . In addition , this model trivially captures the well-known fact that surface mutations tend to be less deleterious than mutations in the protein core , because core residues tend to have many more contacts . With a single model to explain the effects of both random and homologous substitutions , we can understand their differences in terms of residue contacts . The number of deleterious contacts generated by a homologous substitution is less than or equal to the number generated by a random mutation at the same position , with equality rarely being achieved . This is consistent with the explanation that homologous substitutions are conservative because they have been previously selected to be compatible with the protein fold [16] . The factors that determine a particular protein family's tolerance to recombination are unknown . Table 1 reports the fraction of functional sequences in eight recombination libraries , representing protein families of different functions , sizes , and fold classes . Seven of these libraries were designed with the intent of maximizing the fraction of functional sequences , yet there is significant variation ( 2–3 fold ) in this fraction between libraries . While some of this variation is likely due to experimental differences in classifying functional versus nonfunctional sequences for different enzymes , we expect a significant proportion of this variation to arise from differences in parent fold , parent sequence identity , and the specific crossover locations chosen in the library design . Using the random field model , we derive an approximation for the expected value of the fraction of functional sequences in a recombination library and use this to understand how these factors contribute to a protein's tolerance to recombination . Consider a recombination library generated by recombining sequence fragments from parental sequences at crossover sites . We refer to the sequence fragments between crossover sites as ‘blocks’; therefore the library is composed of sequence blocks ( ) . All sequence fragments in these libraries are roughly the same length , and therefore , with the random field model , we can assume that each fragment's energetic contribution is an independently and identically distributed Gaussian random variable . With this assumption , the distribution of sequence energies within recombination library is Gaussian and can be described by its mean ( 8 ) and variance ( 9 ) The fraction of functional sequences within library is given by evaluating the Gaussian cumulative distribution function at zero ( 10 ) where is the error function . Since the specific energy terms that shape the recombinational landscape are unknown , we use the random field model to calculate the expected value of the fraction of functional sequences by integrating over all possible energy terms and . The expected value of the library mean is given by ( 11 ) where is the total number of contacts and is the number of novel contacts in chimera . The expected value of the library variance is given by ( 12 ) More specific details of , , and are given in Text S1 . With these two expectations , the expected value of the fraction of functional sequences can be approximated with a leading-order Taylor series expansion about and ( 13 ) The expected value of the fraction of functional sequences within a library shows quantitative agreement with the experimentally determined values , as shown in Figure 3A . With the random field model , both parental and novel contacts contribute to the distribution of sequence energies within a recombination library and therefore to the fraction of functional sequences . The deleterious novel contacts dictate the mean energy of the library ( ) , while parental contacts , which typically outnumber novel contacts 50–100-fold , dominate the variance ( ) . This suggests recombination events can cause loss of function by two independent mechanisms: ( 1 ) by introducing new deleterious interactions between sequence fragments , or ( 2 ) by introducing sequence fragments that already contain deleterious interactions . To better understand the variation in the fraction of functional sequences in the different recombination libraries , we sampled random libraries , calculated , and estimated the contribution from protein fold , specific breakpoints , and parent sequence identity . For each protein fold , we sampled 100 random two-parent sequence alignments with sequence identity ranging from 10–90% , and for each of these alignments we sampled 100 random 7-crossover libraries , for a total of 90 , 000 libraries . A three-way analysis of variance shows the protein fold ( ) , specific breakpoints ( ) , and parent sequence identity ( ) all make significant contributions to the . Estimating the variance components , we find parent sequence identity to be the main determinant of ( contributing 92% of the variance ) , followed by specific crossover locations ( 4% ) , and protein fold ( 2% ) . This strong dependence on parent sequence identity is the result of the approximately exponential increase in the number of ( deleterious ) novel contacts as parent sequences diverge . Interestingly , the parent sequence identity also dictates the mechanism of chimeric protein inactivation . When the parent sequence identity is low , most of the nonfunctional chimeric proteins are the result of new deleterious interactions between sequence fragments . However , when the parent sequence identity is high , nonfunctional sequences are usually the result of inheriting sequence fragments which already contain deleterious interactions . This is consistent with the observation of high mutual information between a chimeric protein's functional status and its number of novel contacts for the -lactamase library ( low parent sequence identity ) and the low mutual information observed for the P450 library ( high parent sequence identity ) [35] . In the -lactamase library , the number of new interactions between fragments ( novel contacts ) is predictive of the functional status of chimeras . However , in the P450 library , the number of novel contacts is not predictive , suggesting other mechanisms must be responsible for chimera inactivation ( i . e . acquisition of deleterious sequence fragments ) . Perhaps the most surprising finding from protein recombination experiments has been the additive structure of the recombinational landscape [17] , [20] , [22] , [23] , [36] . Linear models are able to explain a majority of variation in stability as well as some other properties , suggesting that sequence elements make largely independent , additive contributions to a protein's overall properties . In quantitative genetics , this is referred to as additive genetic variance , which according to Fisher's fundamental theorem of natural selection determines a population's response to selection [37] , [38] . Additive landscapes are easier for evolving populations to climb because they are not stymied by rugged , epistatic features . This additivity has been especially useful for engineering optimized chimeric proteins in the laboratory , because a small sampling of sequences provides sufficient information to make accurate predictions across the entire library [17] , [22] , [23] . Here , we use the random field model to understand the origin of the additive structure within the recombinational landscape . Within the recombination library described in the previous section , the total variance can be partitioned into additive and epistatic components ( ) . We define the landscape's additivity as the fraction of the total variance that is explained by additive effects ( 14 ) This dimensionless quantity , which ranges from 0 to 1 , describes the smoothness of the landscape and is inversely related to the landscape ‘ruggedness’ defined in [39] . For four of the recombination libraries , there are sufficient data to calculate the additivity of the thermostability landscape ( see Methods ) . These results are presented in Table 1 . The additive variation can be understood by considering how individual mutations contribute to variation in the library . A mutation that occurs at a position with a fixed structural context , such as a mutation within a structural fragment inherited from one parent or a mutation surrounded by conserved positions , will always have the same effect throughout the library and therefore contributes entirely to additive variation . However , a mutation can have different effects in different sequences if it occurs at a position whose environment varies . The effects of these mutations can only be expressed with an epistatic ( non-additive ) model , but their additive contribution can be found by averaging their effects over all structural environments within the library . An additive energy function can be defined by accounting for purely additive and averaged epistatic effects ( Text S1 ) . This additive energy can be used to calculate the expected value of a library's additive variance using the same equations as the total variance ( previous section ) . With this , the expected value of the additivity can be approximated with a Taylor series expansion about and ( 15 ) The expected value of the landscape additivity shows close agreement with the experimentally determined values ( Figure 3B ) . While the correlation is not statistically significant , due to the limited data , all the s are large and within the experimentally observed ranges . In addition , the four uncharacterized libraries also have large expected additivities ( lac13 = 0 . 44 , lac = 0 . 67 , Cel5 = 0 . 65 , Arg = 0 . 82 ) , suggesting this additive structure within the recombinational landscape may be quite general . Despite being generated by a purely pairwise energy function , which is by definition epistatic , a majority of the variation within these recombination libraries can be explained by additive effects . This surprising result can be attributed to two factors: sequence conservation among the parents and the partitioning of interactions into structural modules . Epistatic interactions that are conserved among the parents will not contribute to the variation of any property within the library , and those interactions involving one conserved position will only contribute to additive variation . Epistatic interactions that are partitioned into structural modules will vary together , and therefore contribute to only additive variation . Of the thousands of contacting residues within a chimeric protein , only a small fraction ( 5% ) actually contribute to epistatic variation . The additivity exhibited by the random field model does not hold for chimeric proteins that adopt alternate structures ( as described by a contact map ) . For example , nonfunctional sequences , which account for a significant proportion of chimeras , will clearly not display additivity in properties involving protein function . For many properties , such as thermostability ( retention of function at elevated temperatures ) , where we have observed additivity , the experimental measurements require the chimeras be enzymatically active , which greatly increases the likelihood that they will adopt the same or very similar structures . The subset of sequences that adopt the same structure is referred to as a neutral network [40] , [41] , and this may define the domain of additivity within the recombinational landscape . By using a statistical description of the protein recombinational landscape , we can study the behavior of an astronomical number of sequences–insight which could not be obtained experimentally or even by analyzing homology-based structural models . A probabilistic contact potential was used to specify the mean energy of individual chimeric proteins and how the energy of any sequence is expected to co-vary with others ( equations 6 and 7 ) , defining a multivariate probability distribution over all sequences accessible by recombination . While this random field model provides little information about specific sequences , it does reveal the large-scale structure of the recombinational landscape , which we used here to interpret experimental results from past recombination libraries . Within this random field , the expected values of various library properties show excellent agreement with experimental values across multiple protein families . This striking correspondence may arise because a library's properties depend on a large number of interactions , and the cumulative effects of these interactions converge toward the expected value . The random field model was used to study the enrichment of functional sequences within the recombinational landscape . As shown previously , we found the tolerance of proteins to recombination to be influenced by the conservative effects of homologous substitutions , which have been previously selected to be compatible with the protein fold [16] . However , a more significant contribution comes from groups of coevolved residues varying together . This is especially relevant for understanding natural evolution , where the number of crossover events is relatively low . Evaluating the random field model across protein families , we found parent sequence identity to be the primary determinant for tolerance to recombination , while the specific crossover locations and parent fold make statistically significant , but minor contributions . Using the random field model , we explored the origins of the additive structure of the recombinational landscape . Both sequence conservation among the parents and the partitioning of epistatic interactions into structural modules make significant contributions to this additivity . The results presented here are for a random field that describes a protein's free energy difference between the functional and non-functional states , which is closely related to protein stability . However , this additivity is generally true for any landscape that is generated from local interactions ( including higher order ) , because sequence conservation and structural partitioning will still be present . This suggests the additivity may apply to numerous biophysical quantities such as binding affinity or substrate specificity . Previous studies of protein fitness landscapes have highlighted the abundance of nonfunctional sequences [42] , [43] and neutral sequence changes [13] , [14] , [44] , suggesting a surface which is mostly flat and filled with holes [45] . In contrast to this full landscape , the recombinational landscape contains orders of magnitude fewer ‘holes’ ( non-functional sequences ) . The functional variation displayed within recombination libraries reveals the large-scale structure of the recombinational landscape , which arises from the cumulative effects of multiple mutations . In addition , most of this functional variation can be explained by additive effects , and additive variation determines a population's response to selection [37] , [38] These results were observed in SCHEMA-designed libraries , which tend to be optimized for both functional sequences and additivity . This emphasizes the evolutionary preference for some crossover sites over others , which could explain the presence of recombination hotspots in natural genes [6] , [46] , [47] . The picture of the recombinational landscape that emerges from the random field model is a surface enriched in functional sequences , which can display locally-epistatic behavior but still has an overall additive structure . The evolutionary benefit of intragenic recombination may arise because mutation and recombination effectively traverse different landscapes [48] . While climbing the landscape by point mutations , evolution encounters a large number of nonfunctional sequences in addition to epistatic landscape features . In contrast , recombination explores sequences which are much more likely to be functional , in a landscape with an abundance of adaptive pathways . Recombination can provide faster adaptation than point mutation because it generates functional sequences with a large number of substitutions . Recombination may also find sequences that are inaccessible by adaptive point mutation , by simultaneously incorporating multiple coupled mutations , essentially ‘jumping over’ epistatic landscape features . A similar effect was recently described for recombination at the genome level [49] , where the authors describe how landscapes arising from high epistasis within genes and no epistasis between genes strongly favors recombination . Running simulations on these ‘modular’ landscapes , the authors found recombination to provide an efficient route to genotypes that were inaccessible by point mutation . Intragenic recombination is a powerful molecular diversification mechanism . The ubiquity of intragenic recombination in nature and experimental evidence from protein recombination libraries show that it provides distinct advantages over point mutation . In naturally evolving populations , these two genetic variation mechanisms work together . Mutation provides new diversity , while recombination efficiently sorts through the large combinatorial space of existing diversity . A better understanding of how to balance mutation and recombination could assist in engineering highly optimized proteins .
Since multiple structures have been solved for each protein family tested , we decided to use all available structures to generate the residue-residue contact map . The contact map for each library was determined by identifying all protein chains within the Protein Data Bank that share at least 50% sequence identity with any parent . Also included were three unpublished P450 structures , for a total of 88 lac13 , 173 lac , 91 P450 , 39 CBHI , 24 CBHII , 6 Cel5 , 21 Cel48 , and 143 arginase chains . For each chain , a residue pair was considered contacting if they contained any heavy atoms within 4 . 5 Å . The final contact map for each library is composed of residue pairs that are contacting in more than 50% of all chains . We believe this ‘averaged’ residue-reside contact map should provide a more complete description of the protein family's fold , but the use of any single structure does not change the results presented above . The number of functional and nonfunctional chimeric proteins was retrieved from previously published results: lac13 [34] , lac [50] , P450 [32] , CBHI [22] , CBHII [23] , Cel5 ( unpublished data ) , Cel48 [21] , Arg [20] . The fraction of functional chimeras was estimated using maximum likelihood , and 95% confidence intervals were calculated using the Clopper-Pearson method [51] . We could not accurately estimate the fraction of functional sequences for the CBHI library due to the extreme bias in chimera sampling [22] . The results from the lac13 library were reanalyzed to account for library construction errors ( see below ) . The additivity of the P450 , CBHI , CBHII , and Cel48 libraries was calculated using published thermostability data [17] , [21]–[23] . For each library , a block-based linear regression model [17] was parametrized on all the available data . The resulting predictions are unbiased , so the total variance can be partitioned into explained and residual components . The ratio of the explained variance to total variance is the additivity , and in this case is identical to the regression model's coefficient of determination . Given a data set which maps contact information to binary functional status , we want to estimate the mean energy and variance of parental contacts and the mean energy and variance for novel contacts . The true energy terms and can be integrated out to give the marginalized likelihood function ( 16 ) where is the binary functional status and for notational simplicity all parental energy terms are combined in the vector , all novel energy terms are combined in the vector , and all binary indicator variables ( and ) are combined into the matrix . The mean and variance of parental and novel contacts can be estimated by maximizing this marginalized likelihood function . Since is composed of binary data , we assume that it is generated from a Bernoulli process whose proportion is determined by the energy of a sequence . With this assumption , the first term in the integrand is given by the logistic likelihood function ( 17 ) where is the logistic sigmoid function given by , is the binary functional status of chimera , is a vector containing all , and is a vector containing all . Assuming the energy components are Gaussian distributed , the second and third terms of the integrand are given by multivariate Gaussian distributions . Since the integral in equation 16 is analytically intractable , we can approximate it using Laplace's method [52] . First we approximate the integrand with a multivariate Gaussian about a stationary point and then we evaluate the Gaussian integral to yield ( 18 ) where and are the stationary points , is the fixed number of contacts , and is the Hessian matrix evaluated at the stationary points . The stationary points were found using Newton's method and the marginalized likelihood function was maximized using the Nelder-Mead method . The 13-crossover -lactamase library ( lac13 ) was assembled from synthetic fragments and had a significant number of construction errors [34] . Sequencing of unselected chimeric genes found 9 of 13 to have frame shift mutations [16] , which almost certainly result in inactive proteins . Since a majority of frame shifts are incorporated at the PCR step during library construction , it is likely these errors are present throughout all constructed chimeras [11] . The maximum likelihood estimate for the proportion of correctly constructed chimeras is , with 95% confidence intervals between 0 . 09 and 0 . 61 using the Clopper-Pearson interval [51] . The sequencing data indicate there may be one to three sequence fragments ( chimera blocks ) that contain frameshift mutations . Assuming all frame shifts cause inactivation and exhaustive library coverage ( over twelvefold sampling ) , the fraction of functional chimeras can be estimated as the number of functional chimeras divided by the number of correctly constructed chimeras . With these assumptions , we estimate the fraction of functional sequences to be with 95% confidence intervals between and The same modification can be performed on chimeras binned by the number of homologous substitutions ( Figure 2A ) because the construction errors display little relation to the level of mutation . | Mutation and recombination are the primary sources of genetic variation in evolving populations . The relative benefit of these two diversification mechanisms and how they complement each other has been a long-standing question in evolutionary biology . While it is clear what types of genetic diversity these two mechanisms can create , a significant challenge is relating these sequence changes to changes in fitness . The fitness landscape , which describes this mapping from genotype to phenotype , is extraordinarily complex and defined over an incomprehensibly large space of sequences . Here , we develop a model of the landscape that relies not on the details of this mapping , but rather on the statistical relationships between sequences . By studying the expected values of landscape properties , we can gain insights into the structure of the landscape that are independent of the details of how genotype dictates phenotype . We use this random field model to understand how recombination explores a functionally enriched and diverse subset of protein sequence space . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"evolutionary",
"biology",
"evolutionary",
"modeling",
"protein",
"structure",
"biology",
"computational",
"biology",
"macromolecular",
"structure",
"analysis"
] | 2012 | Random Field Model Reveals Structure of the Protein Recombinational Landscape |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.