Split (3)

train · 24.8k rows

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
Human co-infection with malaria and helmimths is ubiquitous throughout Africa . Nevertheless , its public health significance on malaria severity remains poorly understood . To contribute to a better understanding of epidemiology and control of this co-infection in Cameroon , a cross-sectional study was carried out to assess the prevalence of concomitant intestinal geohelminthiasis and malaria , and to evaluate its association with malaria and anaemia in Nkassomo and Vian . Finger prick blood specimens from a total of 263 participants aged 1–95 years were collected for malaria microscopy , assessment of haemoglobin levels , and molecular identification of Plasmodium species by PCR . Fresh stool specimens were also collected for the identification and quantification of geohelminths by the Kato-Katz method . The prevalence of malaria , geohelminths , and co-infections were 77 . 2% , 28 . 6% , and 22 . 1% , respectively . Plasmodium falciparum was the only malaria parasite species identified with mean parasite density of 111 ( 40; 18 , 800 ) parasites/µl of blood . The geohelminths found were Ascaris lumbricoides ( 21 . 6% ) and Trichuris trichiura ( 10 . 8% ) , with mean parasite densities of 243 ( 24; 3 , 552 ) and 36 ( 24; 96 ) eggs/gram of faeces , respectively . Co-infections of A . lumbricoides and P . falciparum were the most frequent and correlated positively . While no significant difference was observed on the prevalences of single and co-infections between the two localities , there was a significant difference in the density of A . lumbricoides infection between the two localities . The overall prevalence of anaemia was 42% , with individuals co-infected with T . trichiura and P . falciparum ( 60% ) being the most at risk . While the prevalence of malaria and anaemia were inversely related to age , children aged 5–14 years were more susceptible to geohelminthiasis and their co-infections with malaria . Co-existence of geohelminths and malaria parasites in Nkassomo and Vian enhances the occurrence of co-infections , and consequently , increases the risk for anaemia . In many afro-tropical countries , parasitic co-existence is common with increased potential for co-infection , which may adversely impact the outcome of the diseases they cause . [1] . Of all human diseases caused by protozoan parasites malaria has the greatest burden and is responsible for most deaths amongst young children in sub-Saharan Africa , accounting for 90% of all global cases [2] . Until the past decade , intestinal worms have been neglected due to insufficient knowledge of their impact on human life [3]–[6] . The fact that intestinal worms affect more than two-thirds ( 70% ) of humans has led to growing interest to understand their epidemiology and interactions with other parasitic infections [7] , [8] . In Cameroon , both malaria and helminth infections co-exist and are ranked amongst the major cause of parasitic mortality and morbidity . Plasmodium falciparum is the most prevalent and virulent of the malaria parasites [9]–[12] , while the geohelminths ( Ascaris lumbricoides , Trichuris trichiura , and Hookworm ssp . ) and Schistosoma mansoni are the major helminthic parasites [3] , [13]–[16] . Although there is much literature on the epidemiology of malaria and intestinal worms separately , little is known about the distribution and impact of their co-infections on the population across the country [11] , [17] . Due to the differences in the physiological , anthropological , genetic , immunological or geo-ecological factors , infections with multiple parasite species may not necessarily be independent within an individual , and could result in positive or negative associations in disease manifestation . The implications of concomitant malaria and helminth infections have been mainly explored and indicate that their interactions can decrease the course of malaria infection and disease [18] , [19] . On the other hand , co-infected school-aged children have been shown with different phenotypes in the pathogenesis of malaria with increasing risk of developing severe malaria [1] , [10] , [18] , [20]–[22] . These conditions could lead to increase malaria parasitaemia , increased risks of anaemia and malnutrition , modification of the immune response to malaria [12] , [23] , [24] , and increased mortality . Therefore , the acquisition of requisite information on co-morbidity and interactions between malaria and helminthiasis would be invaluable to controlling malaria infection and clinical disease . The control of these parasitic diseases in Cameroon rely on the WHO strategies , which lay emphasis on a combined control approach for malaria infection [11] and deworming activities for helminth parasites [17] . However , implementation of such measures requires essential requisite information on the geo-ecological distribution and efficiency in infection transmission . Thus , this study sought to determine the prevalence of malaria and geohelminths co-infection and evaluate its impact on malaria and anaemia in Nkassomo and Vian , two rural communities of Cameroon . The findings from the study will provide useful information necessary to design strategies to effectively control and manage malaria in the context of co-infection . This study received ethical and administrative authorizations respectively from the IMPM institutional ethics review committee and the competent authorities of the Mfou health district . Participation was strictly voluntary and was dependent on informed written consent by the participant or assent by the parent/guardian for children . All positive cases for malaria and/or intestinal worms were treated for free according to the Cameroonian anti-malarial and anti-helminth treatment guidelines and if necessary referred to the Mfou district hospital for appropriate management . A single dose of Paracetamol and iron tablets was respectively given to participants with fever or anaemia . The study was carried out in two rural communities ( Nkassomo and Vian ) of the Mfou health district ( 4°27′N and 11°38′E ) , a forest area located in the Mefou-Afamba division , in the Centre region of Cameroon . The climate is typically equatorial with two discontinuous dry and wet seasons . The annual average rainfall measures 2 , 000 mm with an annual average temperature of 24°C . The population is made up of 71 , 373 inhabitants ( 4 , 100 in Nkassomo and 3 , 248 in Vian ) with about 82 . 7 inhabitants/km2 . Mfou is a multiethnic community made up of the Ewondo , Bané and Tsinga , with farming being the main economic pursuit . Commonly grown crops include cassava , maize and oil palm [25] . Houses are built in semi-dur with crevices and open joints serving as hide outs for mosquitoes . These villages lack access to potable water . Toilet facilities made up essentially of pit latrines , which are generally poorly constructed and insufficient for the members of each household . It is common to find pools of standing and dirty water close to some households that are prolific for mosquito . The study was conducted from February to March 2011 during the long dry season in the Mfou health district . Samples were collected from each household and only from individuals who willingly accepted to participate and provided written and signed informed consent and have lived in the village for at least the past six months . A total of 263 participants from the two rural communities ( 131 from Nkassomo and 132 from Vian ) aged 1 to 95 years old took part in the survey and were divided into 6 groups as follows: pre-school children ( <5 yrs ) , young school children ( 5–9 yrs ) , old school children ( 10–14 yrs ) , adolescents and young adults ( 15–24 yrs ) , adults ( 25–49 yrs ) , and aged people ( ≥50 yrs ) . Each participant was clinically examined by a medical doctor for febrile symptoms ( fever , axial T ≥37 , 5°C ) or any other clinical conditions ( headache , abdominal discomforts , etc . ) , and information on age and sex taken prior to specimen collection . Statistical analysis of the data was performed using SPSS 18 . 0 software ( SPSS Inc , Chicago IL , USA ) . Chi-square test and One Way ANOVA were used to set the difference in proportions and means respectively , whereas the Pearson logistic regression test was used to establish the correlation between variables ( Plasmodium densities , geohelminth densities , fever , and anaemia ) . Threshold for statistical significance was set at P<0 . 05 . Prevalence was defined as the proportion of individuals found harbouring the parasite in question to the sum total of the study population . Co-infection was defined as the simultaneous presence of at least one helminth parasite species plus malaria parasite in the same host , and was classified as either single or mixed , depending on the number of parasites found . Qualitative values were expressed in frequency or percentage , whilst quantitative values were expressed as geometric mean values ( minimum value; maximum value ) or arithmetic mean value ± Standard Deviation ( SD ) . Of the 263 study participants 131 ( 49 . 8% ) were from Nkassomo and 132 ( 50 . 2% ) from Vian . Of these , there were 123 ( 49 . 6% ) males and 139 ( 53 . 1% ) females . The median age in the study population was 26 . 1±22 . 4 yrs ( range , 1 to 95 yrs ) . Based on the WHO criteria , the median axial temperature was 37 . 3±0 . 5°C ( range , 35 . 5 to 41°C ) and 91 ( 34 . 6% ) participants had fever . Although children in the age group of 5–9 yrs had the most cases of fever no difference was observed based on age , sex and locality . A total of two hundred and eighteen participants ( 82 . 9% ) was found to carry at least one of the identified parasites ( Plasmodium falciparum , A . lumbricoides or T . trichiura ) . The overall prevalence of malaria was 77 . 2% ( n = 203 ) with Plasmodium falciparum being the only parasite species found . The mean parasite density in the population was 111 ( 40; 18 , 800 ) parasites per micro litre of blood ( P/µl ) . Based on the study locality , the prevalence of malaria in Nkassomo was 78 . 6% ( n = 103 ) and 75 . 8% ( n = 100 ) in Vian . This did not vary significantly in the prevalence . A similar trend was observed for the parasite density with an average parasite load of 125 ( 40; 14 , 760 ) P/µl in Nkassomo , and 98 ( 40; 18 , 800 ) P/µl in Vian . Sex did not influence the prevalence and intensity of P . falciparum in both localities . Although all age groups were affected , the peak of malaria prevalence was observed in children less than 9 years old . Only 32 . 5% ( n = 66 ) of malaria positive cases had fever and with the highest parasite densities ( Tables 1 and 2 ) . No gametocytes were found in the infected persons . Ascaris lumbricoides , Trichuris trichiura were the only geohelminth parasites detected . Of the 66 ( 28 . 6% ) participants positive for geohelminth infection , there were 50 ( 21 . 6% ) A . lumbricoides , 25 ( 10 . 8% ) T . trichiura . However , there were 9 ( 3 . 9% ) cases with both infections . The geometric mean of the parasite density for each species was 243 ( 24; 3 , 552 ) eggs per gram of faeces ( eps ) for A . lumbricoides , and 36 ( 24 , 96 ) eps for T . trichiura . There was no observed significant difference in the prevalence of the two parasites between the two localities . However , the parasite densities were higher in Vian with a significant difference observed for A . lumbricoides ( p = 0 , 001 ) . With regards to sex and age , the prevalence of A . lumbricoides was similarly distributed and was higher in children between the ages of 5 years and 14 years ( p = 0 . 002 ) , whilst no difference was observed for T . trichiura infection between the males and the females ( p = 0 , 071 ) and by age groups . Overall , school-aged children ( 5–14 yrs ) had the highest prevalence and parasite density of geohelminth infections . However , only the prevalence was significantly different when compared to other age groups ( p = 0 . 008 ) ( Table 1 ) . With regards to malaria and geohelminth co-infection , a total of 51 ( 22 . 1% ) participants were infected with at least one geohelminth parasite and P . falciparum . Two cases of co-infections were observed and classified as either single co-infections of A . lumbricoides and P . falciparum ( 38 [16 . 5%] ) , and of T . trichiura and P . falciparum ( 20 [8 . 7%] ) ; or mixed co-infections of the three parasites ( 7 [3 . 0%] ) . As observed with single infections of individual parasites , there was significant difference in the prevalence of co-infections between the two study localities . Co-infections with T . trichiura and P . falciparum was observably higher in males than females ( p = 0 . 009 ) . Similarly the overall co-infection was also higher in males than females ( p = 0 . 025 ) . The highest prevalence of co-infection was observed in children between 5 and 14 years old ( p = 0 . 010 ) . With regards to the parasite densities , single infections with only A . lumbricoides ( 179 [24; 1 , 920] eps ) and only P . falciparum ( 107 [40; 4 , 000] p/µl ) had generally lower parasite densities compared to when the two occurred as co-infections of A . lumbricoides ( 257 [24; 3 , 552] eps , p = 0 . 012 ) and P . falciparum ( 143 [40; 2 , 360] p/µl , p = 0 . 018 ) . Although there was a similar observation for mixed co-infections for A . lumbricoides , T . trichiura was associated with low densities of P . falciparum , but the difference was not significant ( Table 3 ) . On the other hand , logistic regression analysis ( Figure 1 ) showed a positive association between the parasite densities of A . lumbricoides ( independent variable ) and P . falciparum ( dependant variable ) in co-infected individuals ( R = 0 . 406 , Correlation coefficient = 0 . 381 , p = 0 . 175 ) . According to the WHO classification , out of the 263 participants in the study , 110 had haemoglobin levels less than 11 g/dl of blood , giving an overall anaemia prevalence of 42% . However , most the anaemic cases were classified as moderate anaemia with a mean haemoglobin level of 9 . 8±0 . 8 g/dl . Only one case of severe anaemia ( 0 . 9% ) was observed with a total haemoglobin level of 6 . 7 g/dl . Fifty three ( 40 . 5% ) and 57 ( 43 . 5% ) participants were found to be anaemic in Nkassomo and Vian , respectively . Although anaemia did not differ by locality and sex , the prevalence of anaemia decreased significantly with the age groups ( p = 0 . 003 ) ( Table 1 ) . Overall , all anaemic cases were found to carry at least one of the detected parasitic infections . There were 13 ( 52% ) , 95 ( 46 . 8% ) , and 19 ( 38% ) cases of anaemia amongst those infected with T . trichiura , P . falciparum , and A . Lumbricoides , respectively . The prevalence however was observed to increase by up to 71% in co-infection cases ( Figure 2 ) . Although anaemia was more frequent in those infected with T . trichiura and P . falciparum as co-infections , A . lumbricoides did not show any influence on the prevalence of anaemia in co-infections . Exploratory multiple linear regression analysis depicted positive associations between the parasite densities ( P . falciparum and T . trichiura ) considered as the independent variables with anaemia and fever , considered as the dependent variables . However , these infections were not significant predictors of anaemia and fever in the area . The findings suggest that malaria is hyperendemic in the study localities , and co-exists with geohelminths with their co-infections common amongst schoolchildren . This co-infection constitutes an important risk to anaemia while exacerbating malaria intensity . Though further epidemiological studies are needed to support these observations and assess the mechanisms involved in such interactions , the results provide useful information necessary to design control management strategies for malaria in the context of co-infection .	Co-infection of malaria and intestinal helminths causes significant and additive problems against the host . In order to contribute to a better understanding of the epidemiology and control of concomitant intestinal geohelminthiasis and malaria infections in Cameroon , a cross-sectional study to assess its prevalence and evaluate its influence on malaria and anaemia was carried out in two rural communities ( Nkassomo and Vian ) of the Mfou health district . Fresh stool and finger prick blood samples from 263 participants aged 1–95 years were analysed to identify and quantify the geohelminths and malaria parasites respectively . Anaemia was assessed by measuring the haemoglobin levels . Whereas Plasmodium falciparum ( 77 . 2% ) was the only malaria parasite species found in the study population , Ascaris lumbricoides ( 21 . 6% ) and Trichuris trichiura ( 10 . 8% ) were the only geohelminths detected . The prevalence of malaria and geohelminth co-infections was 22 . 1% with co-infections of A . lumbricoides and P . falciparum being the most frequently encountered and associating positively . The prevalence of anaemia was 42% , with the most affected being those with co-infections of T . trichiura and P . falciparum . Schoolchildren had significantly higher risk of geohelminthiasis , and their co-infection is a cause for concern and should strategically be considered in the designing and implementation of their control .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "biotechnology", "inflammatory", "diseases", "medicine", "and", "health", "sciences", "animal", "models", "of", "disease", "tropical", "diseases", "parasitic", "diseases", "parasitology", "gastroenterology", "and", "hepatology", "molecular", "biology", "techniques", "extr...	2014	Co-infections of Malaria and Geohelminthiasis in Two Rural Communities of Nkassomo and Vian in the Mfou Health District, Cameroon
Cryptococcus neoformans is one of the leading causes of invasive fungal infection in humans worldwide . C . neoformans uses macrophages as a proliferative niche to increase infective burden and avoid immune surveillance . However , the specific mechanisms by which C . neoformans manipulates host immunity to promote its growth during infection remain ill-defined . Here we demonstrate that eicosanoid lipid mediators manipulated and/or produced by C . neoformans play a key role in regulating pathogenesis . C . neoformans is known to secrete several eicosanoids that are highly similar to those found in vertebrate hosts . Using eicosanoid deficient cryptococcal mutants Δplb1 and Δlac1 , we demonstrate that prostaglandin E2 is required by C . neoformans for proliferation within macrophages and in vivo during infection . Genetic and pharmacological disruption of host PGE2 synthesis is not required for promotion of cryptococcal growth by eicosanoid production . We find that PGE2 must be dehydrogenated into 15-keto-PGE2 to promote fungal growth , a finding that implicated the host nuclear receptor PPAR-γ . C . neoformans infection of macrophages activates host PPAR-γ and its inhibition is sufficient to abrogate the effect of 15-keto-PGE2 in promoting fungal growth during infection . Thus , we describe the first mechanism of reliance on pathogen-derived eicosanoids in fungal pathogenesis and the specific role of 15-keto-PGE2 and host PPAR-γ in cryptococcosis . Cryptococcus neoformans is an opportunistic pathogen that infects individuals who have severe immunodeficiencies such as late-stage HIV AIDS . C . neoformans is estimated to infect 278 , 000 individuals each year resulting in 181 , 000 deaths [1 , 2] . C . neoformans infection begins in the lungs where the fungus is phagocytosed by host macrophages . Macrophages must become activated by further inflammatory signals from the host immune system before they can effectively kill C . neoformans [3 , 4] . When this does not occur C . neoformans proliferates rapidly intracellularly and may use the macrophage to disseminate to the central nervous system leading to fatal cryptococcal meningitis [5–9] . Eicosanoids are an important group of lipid inflammatory mediators produced by innate immune cells such as macrophages . Eicosanoids are a diverse group of potent signalling molecules that have a short range of action and signal through autocrine and paracrine routes . Macrophages produce large amounts of a particular group of eicosanoids called prostaglandins during microbial infection [10 , 11] . Prostaglandins have a number of physiological effects throughout the body , but in the context of immunity they are known to strongly influence the inflammatory state [12] . The prostaglandins PGE2 and PGD2 are the best-studied eicosanoid inflammatory mediators . During infection , macrophages produce both PGE2 and PGD2 to which , via autocrine routes , they are highly responsive [12] . In vertebrate immunity , the synthesis of eicosanoids such as PGE2 is carefully regulated by feedback loops to ensure that the potent effects of these molecules are properly constrained . Exogenous sources of eicosanoids within the body , such as from eicosanoid-producing parasites [13] or tumours that overproduce eicosanoids [14 , 15] , can disrupt host inflammatory signaling as they are not subject to the same regulation . It is well known that C . neoformans produces its own eicosanoid species . These fungal-derived eicosanoids are indistinguishable from those produced by vertebrates [16–18] . Only two Cryptococcus enzymes are known to be associated with cryptococcal eicosanoid synthesis—phospholipase B1 and laccase [18 , 19] . Deletion of phospholipase B1 reduces secreted levels of all eicosanoids produced by C . neoformans suggesting that it has high level role in eicosanoid synthesis [19] , perhaps fulfilling the role of phospholipase A2 in higher organisms . Deletion of laccase results in reduced levels of PGE2 but other eicosanoids are unaffected suggesting that laccase has putative PGE2 synthase activity [18] . C . neoformans produces eicosanoids during infection , these eicosanoids are indistinguishable from host eicosanoids so it is possible that C . neoformans is able to manipulate the host inflammatory state during infection by directly manipulating host eicosanoid signaling . It has previously been reported that the inhibition of prostaglandin E2 receptors EP2 and EP4 during murine pulmonary infection leads to better host survival accompanied by a shift towards Th1/M1 macrophage activation , however it was not determined if PGE2 was derived from the host or the fungus [20] . Therefore , a key aspect of C . neoformans pathogenesis remains unanswered: do eicosanoids produced by C . neoformans manipulate host innate immune cells function during infection ? We have previously shown that the eicosanoid deficient strain Δplb1 has reduced proliferation and survival within macrophages [21] . We hypothesised that eicosanoids produced by C . neoformans support intracellular proliferation within macrophages and subsequently promote pathogenesis . To address this hypothesis , we combined in vitro macrophage infection assays with our previous published in vivo zebrafish model of cryptococcosis [22] . We found that PGE2 was sufficient to promote growth of Δplb1 and Δlac1 independent of host PGE2 production , in vitro and in vivo . We show that the effects of PGE2 in cryptococcal infection are mediated by its dehydrogenated form , 15-keto-PGE2 . Finally , we determine that 15-keto-PGE2 promotes C . neoformans infection via the activation of the host nuclear transcription factor PPAR-γ , demonstrating that 15-keto-PGE2 and PPAR-γ are new factors in cryptococcal infection . We have previously shown that the C . neoformans mutant strain Δplb1 has impaired proliferation and survival within J774 murine macrophages in vitro [21] . The Δplb1 strain has a deletion in the PLB1 gene which codes for the secreted enzyme phospholipase B1 [23] . The Δplb1 strain is known to produce lower levels of fungal eicosanoids indicating that phospholipase B1 is involved in fungal eicosanoid synthesis [19] . It has been proposed that the attenuation of this strain within macrophages could be because it cannot produce eicosanoids [19] . A previous study has identified PGE2 as an eicosanoid that promotes cryptococcal virulence and manipulates macrophage activation , however this study did not determine if PGE2 was produced by the host or C . neoformans [20] . We hypothesised that PGE2—or other phospholipase B1 derived eicosanoid species—are produced by C . neoformans during infection and promote macrophage infection . To test if PGE2 promotes the intracellular growth of C . neoformans we treated Δplb1 infected J774 macrophages with exogenous PGE2 and measured intracellular proliferation over 18 hours . The addition of exogenous PGE2 to J774 macrophages infected with Δplb1 was sufficient to recover the intracellular proliferation of Δplb1 compared to the H99 ( parental wild type strain ) and Δplb1:PLB1 ( PLB1 reconstituted ‘rescue’ strain ) strains ( Fig 1 p = 0 . 038 ) . These findings support our initial hypothesis and identify PGE2 as a mediator of cryptococcal virulence during macrophage infection . We have previously shown that Δplb1 has reduced intracellular proliferation due to a combination of reduced cell division and reduced viability within the phagosome [21] . In our intracellular proliferation assay the number of intracellular Cryptococcus cells is quantified by lysing infected macrophage and counting the number of Cryptococcus cells with a hemocytometer . Due to their structurally stable fungal cell wall , dead or dying Cryptococcus cells do not look noticeably different on a hemocytometer from viable cells . To quantifiy the viability of Cryptococcus cells retrieved from the phagosome we diluted the lysate to give an expected number of CFUs ( in this case 200 CFU ) , spread the diluted lysate on YPD agar and count the actual number of CFUs produced–a difference between the expected CFU count ( 200 CFU ) and the actual CFU count indicates a loss of Cryptococcus cell viability . In this case viability assays showed that exogenous PGE2 produced no significant increase in the viability of Δplb1 cells within the phagosome ( S1A Fig ) . Our in vitro data showed that PGE2 promoted the intracellular proliferation of C . neoformans within macrophages . To confirm this in vivo , we injected 2-day post fertilisation ( dpf ) zebrafish larvae with Δplb1-GFP ( a constitutively expressed GFP tagged version of the Δplb1 generated for this study ) . One of the advantages of this model is that the fungal burden can be non-invasively imaged within infected larvae using fluorescently tagged C . neoformans strains and we able measure the growth of the fungus at 1- , 2- and 3-days post-infection ( dpi ) . We found that Δplb1-GFP infected larvae had significantly lower fungal burdens at 1- , 2- and 3-days post infection ( Fig 2D and S1B Fig ) compared to the parental strain H99-GFP ( Fig 2A and S1B Fig ) . These data demonstrated that the Δplb1-GFP mutant had a similar growth deficiency to our in vitro phenotype and with previous studies [21 , 23 , 24] . To confirm that PGE2 promotes cryptococcal infection in vivo we infected zebrafish larvae with Δplb1-GFP or H99-GFP and treated the larvae with exogenous PGE2 . In agreement with our in vitro findings ( Fig 1 ) , exogenous PGE2 increased the growth of both the parental H99 strain ( Fig 2B , p = 0 . 0137 , 1 . 35-fold increase vs . DMSO ) and the Δplb1-GFP mutant ( Fig 2E , p = 0 . 0001 , 2 . 15-fold increase vs . DMSO ) while PGD2 did not ( Fig 2C and 2F ) . Taken together these data show that PGE2 is sufficient to enhance the virulence of C . neoformans in vivo , furthermore our in vitro data suggest that this is a result of uncontrolled intracellular proliferation within macrophages ( Fig 1 ) . PGE2 can be enzymatically and non-enzymatically modified in cells to form a number of distinct metabolites . To distinguish the biological activity of PGE2 rather than its metabolites we used an analogue of PGE2 called 16 , 16-dimethyl PGE2 that cannot be dehydrogenated but otherwise has comparable activity to PGE2 [25] ) . We found that unlike PGE2 , 16 , 16-dimethyl PGE2 treatment did not increase the fungal burden of Δplb1-GFP ( Fig 3C p = 0 . 9782 ) or H99-GFP infected larvae ( Fig 3A p = 0 . 9954 ) . Therefore , the biological activity of PGE2 alone did not appear to promote cryptococcal pathogenesis suggesting that dehydrogenation of PGE2 was required . PGE2 and 16 , 16-dimethyl PGE2 both signal through PGE2 receptors ( EP1 , EP2 , EP3 and EP4 ) but 15-keto-PGE2 does not . In a murine model of pulmonary cryptococcosis the PGE2 receptors EP2 and EP4 have been identified as promoters of fungal virulence [20] . To confirm that PGE2 itself does not promote virulence in our model we treated zebrafish with antagonists against the EP2 and EP4 receptors . In support of our previous experiment we found that EP2 / EP4 inhibition had no effect on fungal burden in zebrafish ( Fig 3E ) . An abundant dehydrogenated form of PGE2 is 15-keto-PGE2 ( that has also been isolated from C . neoformans [18] ) and we tested if 15-keto-PGE2 was sufficient to rescue the growth defect of the Δplb1 mutant during infection . Therefore , we treated infected zebrafish larvae with exogenous 15-keto-PGE2 and found that this was sufficient to significantly increase the fungal burden of zebrafish larvae infected with both Δplb1-GFP ( Fig 3D , p = 0 . 0119 , 1 . 56-fold increase vs . DMSO ) and H99-GFP ( Fig 3B , p = 0 . 0048 , 1 . 36-fold increase vs . DMSO ) . To explore the Cryptococcus eicosanoid synthesis pathway further we used a second eicosanoid deficient C . neoformans mutant Δlac1 . Whereas Δplb1 is unable to produce any eicosanoid species Δlac1 is deficient only in PGE2 and 15-keto-PGE2 [18] . Using a Δlac1-GFP strain generated for this study we found that Δlac1-GFP also produced low fungal burden during in vivo zebrafish larvae infection ( S2B Fig ) . As with Δplb1-GFP , this defect could be rescued with the addition of exogenous PGE2 ( S2C Fig ) , but not with 16 , 16-dm-PGE2 ( S2D Fig ) or with 15-keto-PGE2 ( S2E Fig ) . PGE2 is known to affect haematopoietic stem cell homeostasis in zebrafish [26] . This could affect macrophage number and subsequently fungal burden . We have previously observed that a large depletion of macrophages can lead to increased fungal burden in zebrafish larvae [22] . We performed whole body macrophage counts on 2 dpf uninfected larvae treated with PGE2 or 15-keto-PGE2 2 days post treatment ( the same time points used in our infection assay ) . Following PGE2 and 15-keto PGE2 treatment macrophages were still observable throughout the larvae . For PGE2 treatment we saw on average a 15% reduction in macrophage number while 15-keto PGE2 did not cause any decrease ( S1C Fig ) . Due to the fact that fungal burden increased during both PGE2 and 15-keto PGE2 treatments it is highly unlikely that this reduction could account for the increases in burden seen . After determining that PGE2 promotes the growth of C . neoformans in vitro and in vivo via its metabolite 15-keto-PGE2 , we wanted to determine whether the source of these prostaglandins was the host or the fungus . Our in vitro data show that the C . neoformans strain Δplb1 has a growth deficiency in vitro and in vivo that can be rescued with the addition of PGE2 , because this phenotype is mediated by cryptococcal phospholipase B1 it indicates that pathogen-derived rather than host-derived prostaglandins are required . A previous study reports that C . neoformans infection can induce higher PGE2 levels in the lung during in vivo pulmonary infection of mice [20] , although it was not determined if the PGE2 was host or pathogen derived . Therefore , we tested the hypothesis that host prostaglandin synthesis was not required for cryptococcal virulence . To block host prostaglandin synthesis in vitro we inhibited host cyclooxygenase activity because it is essential for prostaglandin synthesis in vertebrates [27] . We treated H99 and Δplb1 infected J774 macrophages with aspirin at a concentration we determined was sufficient to block host PGE2 synthesis ( S3A Fig ) . Aspirin is a non-reversible inhibitor of cyclooxygenase-1 ( COX-1 ) and cyclooxygenase-2 ( COX-2 ) enzymes , we therefore included a condition where J774 cells were pretreated with aspirin only prior to infection and then at a condition where aspirin was present throughout infection . We found that aspirin treatment did not affect the intracellular proliferation of H99 or Δplb1 ( Fig 4A ) , suggesting that host cyclooxygenase activity is not required for the phospholipase B1 dependent virulence of C . neoformans during macrophage infection . To confirm this in vivo , we pharmacologically blocked zebrafish COX-1 and COX-2 . We used separate cyclooxygenase inhibitors with zebrafish larvae instead of aspirin because we found that aspirin treatment led to lethal developmental defects in zebrafish larvae ( unpublished observation ) . We infected 2 dpf zebrafish larvae with H99-GFP and Δplb1-GFP and treated with inhibitors for COX-1 ( NS-398 , 15 μM ) and COX-2 ( SC-560 , 15 μM ) . We found that both inhibitors decreased the fungal burden of H99-GFP , but not Δplb1-GFP infected zebrafish larvae ( Fig 4Bi and 4Bii , H99-GFP—NS-398 , p = 0 . 0002 , 1 . 85-fold decrease vs . DMSO . SC-560 p = <0 . 0001 , 3 . 14-fold decrease vs DMSO ) . These findings were different to what we had observed in vitro but because this phenotype was phospholipase B1 dependent we reasoned that these inhibitors could be having off target effects on C . neoformans . C . neoformans does not have a homolog to cyclooxygenase however other studies have tried to inhibit eicosanoid production in Cryptococcus using cyclooxygenase inhibitors but their efficacy and target remain uncertain [17 , 28] . To support our pharmacological evidence , we used a CRISPR/Cas9-mediated knockdown of the prostaglandin E2 synthase gene ( ptges ) [29] . We used a knockdown of tyrosinase ( tyr ) –a gene involved in the conversion of tyrosine into melanin as a control because tyr-/- crispants are easy to identify because they do not produce any pigment . We infected 2 dpf ptges-/- and tyr-/- zebrafish larvae with H99-GFP or Δplb1-GFP and measured the fungal burden at 3 dpi . We found that ptges-/- zebrafish infected with H99-GFP had a higher fungal burden at 3 dpi compared to tyr-/- zebrafish infected with H99-GFP whereas there was no difference between ptges-/- and tyr-/- zebrafish larvae infected with Δplb1-GFP ( Fig 4C ) . Thus , both pharmacological and genetic inhibitions of host prostaglandin synthesis were not determinants of C . neoformans growth . To further evidence that C . neoformans was the source of PGE2 during macrophage infection we used a co-infection assay which has previously been used to investigate the interaction of different C . gattii strains within the same macrophage [30] . We hypothesised that if C . neoformans derived prostaglandins promoted fungal growth , co-infection between H99 and Δplb1 would support mutant growth within macrophages because the parental strain H99 would produce growth promoting prostaglandins that are lacking in Δplb1 . To produce co-infection , J774 murine macrophages were infected with a 50:50 mixture of Δplb1 and H99-GFP [30] ( Fig 4Ei; as described previously for C . gattii [30] ) . This approach allowed us to differentiate between Δplb1 ( GFP negative ) and H99 ( GFP positive ) Cryptococcus strains within the same macrophage and to score their proliferation separately . The intracellular proliferation of Δplb1 was calculated by counting the change in number of GFP negative Δplb1 cells over an 18hr period from time-lapse movies of infected cells . We found that co-infected macrophages did not always contain an equal ratio of each strain at the start of the 18hr period so we scored the proliferation of Δplb1 for a range of initial burdens ( 1:2 , 1:3 and 1:4 ) . We found that Δplb1 proliferated better when accompanied by two H99-GFP yeast cells in the same macrophage ( Fig 4Eii , 1:2 p = 0 . 014 ) as opposed to when two Δplb1 yeast cells were accompanied by one H99-GFP yeast cell ( Fig 4Eii , 2:1 ) . We observed a similar effect for ratios of 1:3 and 1:4 , but these starting burden ratios are particularly rare , under powering our analysis ( S3 Fig ) . This effect was also recapitulated for J774 macrophages co-infected with Δlac1 and H99-GFP ( S2A Fig ) . These data indicate that a phospholipase B1 dependent factor found in H99 infected macrophages , but absent in Δplb1 and Δlac1 infected macrophages , is required for intracellular proliferation within macrophages . Our data indicate that host cells are not the source of virulence promoting prostaglandins , and that secreted factors produced by wild type—but not Δplb1 or Δlac1 Cryptococcus strains–promote fungal growth within macrophages . We wanted to test if there were detectable differences in PGE2 levels between infected and uninfected macrophages caused by cryptococcal prostaglandin synthesis . To do this we performed ELISA analysis to detect PGE2 concentrations in supernatants from C . neoformans infected J774 macrophages ( Fig 4Di ) . We found that J774 macrophages produced detectable levels of PGE2 ( mean concentration 4 . 10 ng/1x106 cells ) , however we did not see any significant difference between infected or uninfected macrophages ( infected with H99 , Δplb1 or Δplb1:PLB1 strains ) . To confirm our ELISA results , we performed LC MS/MS analysis of lysed J774 macrophages using a PGE2 standard for accurate quantification ( Fig 4Dii ) . The concentrations detected were similar to those measured by our ELISA ( mean concentration 6 . 35 ng/1X106 cells ) and also did not show any significant differences between conditions . Taken together these data suggest that any Cryptococcus-derived prostaglandins present during infection were likely to be contained within the macrophage in low , localized concentrations and that the host receptor targeted by these eicosanoids is therefore likely to be intracellular . We next wanted to determine how PGE2 / 15-keto-PGE2 promotes C . neoformans infection . We hypothesized that these prostaglandins were interfering with inhibition of fungal growth via a host receptor . Our experiments inhibiting EP2 / EP4 suggest that 15-keto-PGE2 must signal through a different receptor to PGE2 ( Fig 3E ) . 15-keto-PGE2 is a known agonist of the peroxisome proliferation associated receptor gamma ( PPAR-γ ) [31]; a transcription factor that controls expression of many inflammation related genes [32–34] . We first tested if PPAR-γ activation occurs within macrophages during C . neoformans infection by performing immunofluorescent staining for PPAR-γ in macrophages infected with H99 and Δplb1 . To quantify PPAR-γ activation we measured its nuclear translocation by comparing nuclear and cytoplasmic fluorescence intensity in infected cells . PPAR-γ is a cytosolic receptor that translocates to the nucleus upon activation , therefore cells where PPAR-γ is activated should have increased nuclear staining for PPAR-γ . We found that J774 macrophages infected with H99 had significantly higher levels of nuclear staining for PPAR-γ compared to Δplb1 infected and uninfected cells ( Fig 5Ai and S3C Fig ) . This confirmed that C . neoformans activates PPAR-γ and that this phenotype is phospholipase B1 dependent . To test the activation of PPAR-γ during infection we first wanted to confirm that exogenous 15-keto-PGE2 activates zebrafish PPAR-γ in vivo by using transgenic PPAR-γ reporter zebrafish larvae [35 , 36] . We treated these larvae at 2 dpf with 15-keto-PGE2 and Troglitazone ( TLT ) which is a specific agonist of PPAR-γ . TLT treatment was performed with a concentration ( 0 . 55 μM ) previously shown to strongly activate PPAR-γ in these zebrafish larvae [35] . We found that TLT treatment at 2 dpf strongly activated GFP reporter expression in the larvae . We employed a receptor competition assay as a sensitive measurement of binding by simultaneously treating zebrafish with with 15-keto-PGE2 and TLT . We observed a reduction in GFP expression compared to TLT treatment alone ( Fig 5Aii ) , demonstrating competition for the same receptor . This could mean that 15-keto-PGE2 was a partial agonist [37 , 38] or an antagonist to PPAR-γ in this experiment . Existing studies suggest 15-keto-PGE2 is an agonist to PPAR-γ [31] but to confirm this in our model we treated Δplb1-GFP infected zebrafish larvae with exogenous 15-keto-PGE2 as before but at the same time treated fish with the PPAR-γ antagonist GW9662 . We found that 15-keto-PGE2 treatment significantly improved the growth of Δplb1-GFP during infection but that inhibition of PPAR-γ was sufficient to reverse this effect ( Fig 5C ) . Therefore , we could demonstrate that that 15-keto-PGE2 was an agonist to PPAR-γ , and that PPAR-γ activation was sufficient to promote a permissive environment for C . neoformans growth during infection . To determine if PPAR-γ activation by C . neoformans facilitates growth within macrophages we treated H99 and Δplb1 infected J774 murine macrophages with the PPAR-γ antagonist GW9662 . GW9662 treatment significantly reduced the proliferation of H99 , but not Δplb1 ( Fig 5B , p = 0 . 026 , 1 . 22-fold decrease vs . DMSO ) . Further supporting that PPAR-γ activation is necessary for the successful intracellular parasitism of host macrophages by C . neoformans . Finally , to confirm that PPAR-γ activation alone promoted C . neoformans infection we treated 2dpf infected zebrafish larvae with TLT at the same concentration known to activate PPAR-γ in PPAR-γ reporter fish ( Fig 5Aii and [35] ) . We found that TLT treatment significantly increased the fungal burden of Δplb1-GFP ( Fig 5E and , p = 0 . 0089 , 1 . 68-fold increase vs . DMSO ) , H99-GFP ( Fig 5D , p = 0 . 0044 , 1 . 46-fold increase vs . DMSO ) and Δlac1-GFP ( Fig 5F , p = 0 . 01 , 1 . 94-fold increase vs . DMSO ) infected larvae similar to 15-keto-PGE2 treatment . Thus , we could show that host PPAR-γ activation was sufficient to promote cryptococcal growth during infection and was a consequence of fungal derived prostaglandins . We have shown for the first time that eicosanoids produced by C . neoformans promote fungal virulence both in vitro and in vivo . In this respect we have shown that the intracellular growth defects of two eicosanoid deficient C . neoformans strains Δplb1 and Δlac1 [21 , 23] can be rescued with the addition of exogenous PGE2 . Furthermore , our in vitro co-infection assay , in vitro infection assays with aspirin , and in vivo infection assays provide evidence that the source of this eicosanoid during infection is from the pathogen , rather than the host . Using an in vivo zebrafish larvae model of cryptococcosis we find that that PGE2 must be dehydrogenated into 15-keto-PGE2 before to influence fungal growth . Finally , we provide evidence that the mechanism of PGE2/15-keto-PGE2 mediated growth promotion during larval infection is via the activation of PPAR-γ [32 , 33 , 39–41] . In a previous study it was identified that the C . neoformans mutant Δplb1 ( that lacks the PLB1 gene coding for phospholipase B1 ) was deficient in replication and survival in macrophages [21] , a phenotype also observed by a number of studies using different in vitro infection assays [19 , 23] . In this study we demonstrate that supplementing Δplb1 with exogenous prostaglandin E2 during in vitro macrophages infection is sufficient to restore the mutant’s intracellular proliferation defect . A key goal of this study was to investigate how eicosanoids produced by C . neoformans modulate pathogenesis in vivo [22] . To facilitate in vivo measurement of fungal burden we created two GFP-tagged strains with constitutive GFP expression - Δplb1-GFP and Δlac1-GFP—to use alongside the GFP-tagged H99 parental strain previously produced [42] . These two mutants are the only C . neoformans mutants known to have a deficiency in eicosanoid synthesis . Δplb1 cannot produce any eicosanoid species suggesting phospholipase B1 is high in the eicosanoid synthesis pathway while Δlac1 has a specific defect in PGE2 suggesting it might be a prostaglandin E2 synthase enzyme . To our knowledge these are the first GFP expressing strains of Δplb1 and Δlac1 created . Characterisation of Δplb1-GFP and Δlac1-GFP in vivo revealed that both strains have significantly reduced fungal burdens compared to H99-GFP . This is the first report of Δlac1 in zebrafish but these observations do confirm a previous zebrafish study showing that non-fluorescent Δplb1 had attenuated infectious burden in zebrafish larvae [43] . To confirm that PGE2 is also required for cryptococcal growth in vivo we treated Δplb1-GFP and Δlac1-GFP infected zebrafish larvae with exogenous PGE2 to determine how it would affect fungal burden . In agreement with our in vitro findings we found that PGE2 significantly improved the growth of both of these strains within larvae . Interestingly we also found that PGE2 improved the growth of H99-GFP , perhaps representing a wider manipulation of host immunity during in vivo infection . In vertebrate cells , PGE2 is converted into 15-keto-PGE2 by the enzyme 15-prostaglandin dehydrogenase ( 15PGDH ) , furthermore it has been reported that C . neoformans has enzymatic activity analogous to 15PGDH [18] . To investigate how the dehydrogenation of PGE2 to 15-keto-PGE2 influenced fungal burden we treated infected larvae with 16 , 16-dm-PGE2 –a synthetic variant of PGE2 which is resistant to dehydrogenation [25] . Interestingly we found that 16 , 16-dm-PGE2 was unable to promote the growth of Δplb1-GFP , H99-GFP or Δlac1-GFP within infected larvae . These findings indicate that 15-keto-PGE2 , rather than PGE2 , promotes cryptococcal virulence . We subsequently treated infected larvae with exogenous 15-keto-PGE2 and confirmed that 15-keto-PGE2 treatment was sufficient to promote the growth of both Δplb1-GFP and H99-GFP , but not Δlac1-GFP ( discussed below ) , without the need for PGE2 . We therefore propose that PGE2 produced by C . neoformans during infection must be enzymatically dehydrogenated into 15-keto-PGE2 to promote cryptococcal virulence . These findings represent the identification of a new virulence factor ( 15-keto-PGE2 ) produced by C . neoformans , as well as the first-time identification of an eicosanoid other than PGE2 with a role in promoting cryptococcal growth . Furthermore , our findings suggest that previous studies which identify PGE2 as a promoter of cryptococcal virulence [19 , 20 , 44] may have observed additive effects from both PGE2 and 15-keto-PGE2 activity . Our experiments with the Δlac1-GFP strain reveals that this strain appears to respond in a similar way to PGE2 , 16 , 16-dm-PGE2 and troglitazone as Δplb1-GFP but appears to be unresponsive to 15-keto-PGE2 . At this time , we cannot fully explain this phenotype , Δlac1-GFP was generally less responsive to PGE2 and troglitazone treatments compared to Δplb1-GFP so it is possible that that higher concentrations of 15-keto-PGE2 would be needed to rescue its growth defect , however experimentation with higher concentrations of 15-keto-PGE2 led to significant host toxicity . The unresponsiveness of Δlac1 to eicosanoid treatment could be due to unrelated virulence defects caused by laccase deficiency . Cryptococcal laccase expression is required for the production of fungal melanin–a well characterized virulence factor produced by C . neoformans [45 , 46] . It is therefore likely that virulence defects unrelated to eicosanoid synthesis are responsible for the differences between the two mutant phenotypes . We have found that the phospholipase B1 dependent attenuation of Δplb1 can be rescued with the addition of exogenous PGE2 . This indicates that synthesis and secretion of PGE2 by C . neoformans is a virulence factor . Although our data indicated that C . neoformans was the source of PGE2 we wanted exclude the possibility that host-derived PGE2 was also contributing to virulence . To explore this possibility , we blocked host prostaglandin synthesis—we reasoned that if host PGE2 was not required that blocking its production would not affect the growth of C . neoformans . To do this we first treated J774 macrophages infected with H99-GFP and Δplb1-GFP with aspirin–a cyclooxygenase inhibitor that blocks both COX-1 and COX-2 activity–and found that this had no effect on intracellular proliferation . Subsequently we attempted to block cyclooxygenase activity in zebrafish but found aspirin was lethal at the zebrafish larvae’s currently stage of development , instead we used individual inhibitors specific for COX-1 ( NS-398 ) and COX-2 ( SC-560 ) . We found that each inhibitor decreases fungal burden of H99-GFP infected larvae but not Δplb1-GFP infected larvae . Due to the phospholipase B1 dependence of this phenotype we think that these inhibitors might be affecting eicosanoid production by the C . neoformans itself . The ability of broad COX inhibitors like aspirin/indomethacin to inhibit eicosanoid production by C . neoformans is controversial [17 , 28] however our study is the first to use selective COX-1 and COX-2 inhibitors on C . neoformans . This experiment remained inconclusive as to whether host PGE2 synthesis promotes virulence so to block PGE2 in zebrafish larvae without potential off target effects we used CRISPR Cas9 technology to knockdown expression of the prostaglandin E2 synthase gene ptges in zebrafish larvae . Ablating zebrafish ptges with this approach did not affect the fungal burden of Δplb1-GFP infected larvae but it did cause increased burden in H99-GFP infected zebrafish . This phenotype is interesting because it suggests host PGE2 might actually be inhibitory to cryptococcal virulence , furthermore this phenotype was phospholipase B1 dependent which suggests host-derived PGE2 might interact in some way with Cryptococcus-derived eicosanoids . This phenotype was not seen in vitro with aspirin so it is possible that the inhibitory effects of host-derived PGE2 influence a non-macrophage cell type in zebrafish larvae . To confirm our observations that C . neoformans was the source of PGE2 during infection we performed co-infection assays with H99 wild type cryptococci ( eicosanoid producing ) and Δplb1 ( eicosanoid deficient ) within the same macrophage and found that co-infection was sufficient to promote the intracellular growth of Δplb1 . We also observed similar interactions during Δlac1 co-infection ( a second eicosanoid deficient C . neoformans mutant ) . These observations agree with previous studies that suggest eicosanoids are virulence factors produced by C . neoformans during macrophage infection [19 , 28] . To identify if the secreted factor produced by C . neoformans was PGE2 , we measured the levels of PGE2 from Cryptococcus infected macrophages to see if there was an observable increase in this eicosanoid during infection . Although PGE2 was detected , we did not see any significant difference between infected and uninfected macrophages , an observation confirmed using two different detection techniques–ELISA and LC MS/MS . These data suggest that PGE2 produced by C . neoformans during macrophage infection is contained within the macrophage , likely in close proximity to the fungus and that the host receptor targeted by these eicosanoids is therefore likely to be intracellular . Our in vitro co-infection experiments indicate that C . neoformans secretes virulence enhancing eicosanoids during infection . The biological activity of 15-keto-PGE2 is far less studied than PGE2 . It is known that 15-keto-PGE2 cannot bind to prostaglandin E2 EP receptors , this means it can act as a negative regulator of PGE2 activity i . e . cells up-regulate 15PGDH activity to lower PGE2 levels [47] . Our findings however suggested that 15-keto-PGE2 did have a biological activity independent of PGE2 synthesis , possibly via a distinct eicosanoid receptor . It has been demonstrated that 15-keto-PGE2 can activate the intracellular eicosanoid receptor peroxisome proliferator associated receptor gamma ( PPAR- γ ) [31] . Activation of PPAR-γ by C . neoformans has not been described previously but it is compatible with what we know of cryptococcal pathogenesis . PPAR-γ is a nuclear receptor normally found within the cytosol . Upon ligand binding PPAR-γ forms a heterodimer with Retinoid X receptor ( RXR ) and translocates to the nucleus where it influences the expression of target genes which possess a peroxisome proliferation hormone response element ( PPRE ) [48] . If eicosanoids are produced by C . neoformans during intracellular infection , it is likely that they bind to an intracellular eicosanoid receptor . Additionally , activation of PPAR-γ within macrophages is known to promote the expression of anti-inflammatory genes which could make the macrophage more amenable to parasitism by the fungus . To investigate whether PPAR-γ activation within macrophages occurs during C . neoformans infection we performed immunofluorescent staining of H99 and Δplb1 infected J774 macrophages . We found that infection with C . neoformans led to increased nuclear localization of PPAR- γ indicating that the fungus was activating endogenous PPAR- γ during infection . We also found that macrophages infected with H99 had higher levels of PPAR- γ activation than Δplb1 infected macrophages . This strongly suggests that eicosanoids produced by C . neoformans are responsible for activating PPAR- γ . To confirm that 15-keto-PGE2 is an agonist to PPAR-γ we performed experiments with zebrafish larvae from PPAR-γ GFP reporter fish [35] and demonstrated that 15-keto-PGE2 binds to zebrafish PPAR- γ . To determine if 15-keto-PGE2 is an agonist to PPAR- γ we found that treating Δplb1-GFP infected zebrafish larvae with GW9662 at the same time as 15-keto-PGE2 blocked the virulence enhancing effects of the eicosanoid . These data indicate that 15-keto-PGE2 is a partial agonist to PPAR-γ ( in zebrafish at least ) . Partial agonists are weak agonists that bind to and activate receptors , but not at the same efficacy as a full agonist . Partial agonists to PPAR-γ have been reported previously , partial PPAR agonists bind to the ligand binding domain of PPAR-γ with a lower affinity than full PPAR agonists and as a result activate smaller subsets of PPAR-γ controlled genes [37 , 38 , 49–52] . We also found that activation of PPAR-γ alone was sufficient to mediate cryptococcal virulence . In this respect , we found that the in vitro intracellular proliferation of the wild type H99 cryptococcal strain within J774 macrophages could be suppressed using a PPAR-γ antagonist GW9662 . We could also block the rescuing effect of 15-keto-PGE2 on Δplb1-GFP during zebrafish infection using GW9662 . Finally we found that Cryptococcus infected zebrafish treated with troglitazone at a concentration that is known to activate PPAR-γ [35] had increased fungal burdens when infected with Δplb1-GFP , Δlac1-GFP and H99-GFP strains . Taken together these experiments provide convincing evidence that a novel cryptococcal virulence factor—15-keto-PGE2 –enhances the virulence of C . neoformans by activation of host PPAR-γ and that macrophages are one of the key targets of this eicosanoid during infection . In this study , we have shown for the first time that eicosanoids produced by C . neoformans can promote virulence in an in vivo host . Furthermore , we have provided evidence that this virulence occurs via eicosanoid mediated manipulation of host macrophages . We have identified that the eicosanoid responsible for these effects is 15-keto-PGE2 which is derived from the dehydrogenation of PGE2 produced by C . neoformans . We have subsequently demonstrated that 15-keto-PGE2 mediates its effects via activation of PPAR- γ , an intracellular eicosanoid receptor known to promote anti-inflammatory immune pathways within macrophages . We provide compelling evidence that eicosanoids produced by C . neoformans enhance virulence , identifies a novel virulence factor– 15-keto-PGE2 –and describes a novel mechanism of host manipulation by C . neoformans—activation of PPAR- γ . Most importantly this study provides a potential new therapeutic pathway for treatment of cryptococcal infection , as several eicosanoid modulating drugs are approved for patient treatment [53] . Animal work was performed following UK law: Animal ( Scientific Procedures ) Act 1986 , under Project License PPL 40/3574 and P1A4A7A5E . Ethical approval was granted by the University of Sheffield Local Ethical Review Panel . Experiments using the PPAR-γ reporter fish line [35] were conducted at the University of Toronto following approved animal protocols ( # 00000698 “A live zebrafish-based screening system for human nuclear receptor ligand and cofactor discovery” ) under a OMAFRA certificate . The following zebrafish strains were used for this study: Nacre wild type strain , Tg ( mpeg1:mCherryCAAX ) sh378 ) transgenic strain [22] and the double mutant casper , for PPARγ reporter experiments [35] , which lacks all melanophores and iridophores [54] . Zebrafish were maintained according to standard protocols . Adult fish were maintained on a 14:10 –hour light / dark cycle at 28 oC in UK Home Office approved facilities in the Bateson Centre aquaria at the University of Sheffield . The H99-GFP strain has been previously described [42] . The Δplb1-GFP and Δlac1-GFP stains was generated for this study by transforming existing deletion mutant strains [23 , 55] with a GFP expression construct ( see below for transformation protocol ) . All strains used are in the C . neoformans variety grubii H99 genetic background . Cryptococcus strains were grown for 18 hours at 28 oC , rotating horizontally at 20 rpm . Cryptococcus cultures were pelleted at 3300g for 1 minute , washed twice with PBS ( Oxoidm Basingstoke , UK ) and re-suspended in 1ml PBS . Washed cells were then counted with a haemocytometer and used as described below . C . neoformans strains Δplb1 and Δlac1 were biolistically transformed using the pAG32_GFP transformation construct as previously described for H99-GFP [42] . Stable transformants were identified by passaging positive GFP fluorescent colonies for at least 3 passages on YPD agar supplemented with 250 μg/ml Hygromycin B . CRISPR generation was performed as previously described [29] . Briefly gRNA spanning the ATG start codon of zebrafish ptges or tyr was injected along with Cas9 protein and tracrRNA into zebrafish embryos at the single cell stage . Crispant larvae were infected with C . neoformans as described above at 2 dpf . The genotype of each larvae was confirmed post assay–genomic DNA was extracted from each larvae and the ATG was PCR amplified with primers spanning the ATG site of ptges ( Forward primer gccaagtataatgaggaatggg , Reverse primer aatgtttggattaaacgcgact ) producing a 345-bp product . This product was digest with Mwol–wildtype digests produced bands at 184 , 109 and 52 bp while mutant digests produced bands at 293 and 52 bp ( S4 Fig ) . J774 macrophage ( J774 cells were obtained from the ATCC , American Type Culture Collection ) infection was performed as previously described [21] with the following alterations . J774 murine macrophage-like cells were cultured for a minimum of 4 passages in T75 tissue culture flasks at 37 oC 5% CO2 in DMEM ( High glucose , Sigma ) supplemented with 10% Fetal Bovine Calf Serum ( Invitrogen ) , 1% 10 , 000 units Penicillin / 10 mg/ml streptomycin and 1% 200 mM/L–glutamine , fully confluent cells were used for each experiment . Macrophages were counted by haemocytometer and diluted to a concentration of 1x105 cells per ml in DMEM supplemented with 1 μg/ml lipopolysaccharide ( LPS from E . coli , Sigma L2630 ) before being plated into 24 well microplates ( Greiner ) and incubated for 24 hours ( 37 oC 5% CO2 ) . Following 24-hour incubation , medium was removed and replaced with 1 ml DMEM supplemented with 2 nM prostaglandin E2 ( CAY14010 , 1mg/ml stock in 100% ethanol ) . Macrophage wells were then infected with 100 μl 1x106 yeast/ml Cryptococcus cells ( from overnight culture , washed . See above ) opsonized with anti-capsular IgG monoclonal antibody ( 18b7 , a kind gift from Arturo Casadevall ) . Cells were incubated for 2 hours ( 37 oC 5% CO2 ) and then washed with 37 oC PBS until extracellular yeast were removed . After washing , infected cells were treated with 1ml DMEM supplemented with PGE2 . To calculate IPR , replicate wells for each treatment/strain were counted at 0 and 18 hours . Each well was washed once with 1ml 37 oC PBS prior to counting to remove any Cryptococcus cells released by macrophage death or vomocytosis . Intra-macrophage Cryptococci were released by lysis with 200 μl dH2O for 20 minutes ( lysis confirmed under microscope ) . Lysate was removed to a clean microcentrifuge tube and an additional 200 μl was used to wash the well to make a total lysate volume of 400 μl . Cryptoccoccus cells within lysates were counted by haemocytometer . IPR was calculated by dividing the total number of counted yeast at 18hr by the total at 0hr . To assess the viability of C . neoformans cells recovered from macrophages we used our previously published colony forming unit ( CFU ) viability assay [21] . Lysates from C . neoformans infected J774 cells were prepared from cells at 0hr and 18hr time points . The concentration of C . neoformans cells in the lysate was calculated by haemocytomter counting , the lysates were then diluted to give an expected concentration of 2x103 yeast cells per ml . 100 μl of this diluted lysate was spread onto a YPD agar plate and incubated for 48 hr at 25 oC prior to colony counting . J774 cells were prepared and seeded at a concentration of 1x105 per ml as above in 24 well microplates and incubated for 24 hours ( 37 oC 5% CO2 ) , 45 minutes prior to infection J774 cells were activated with 150 ng/ml phorbol 12-myristate 13-acetate in DMSO added to 1 ml serum free DMEM . Following activation J774 cells were washed and infected with 100 μl / 1x106 yeast cells/ml 50:50 mix of Δplb1 ( non-fluorescent ) and H99-GFP ( e . g . 5x105 Δplb1 and 5x105 H99-GFP ) or Δlac1-GFP and H99 ( non-fluorescent ) . Infected cells were incubated for 2 hours ( 37 oC 5% CO2 ) to allow for phagocytosis of Cryptococcus and then washed multiple times with 37 oC PBS to remove unphagocytosed yeast , each well was observed between washes to ensure that macrophages were not being washed away . After washing 1 ml DMEM was added to each well . Co-infected cells were imaged over 20 hours using a Nikon TE2000 microscope fitted with a climate controlled incubation chamber ( 37 oC 5% CO2 ) using a Digital Sight DS-QiMC camera and a Plan APO Ph1 20x objective lens ( Nikon ) . GFP and bright field images were captured every 4 minutes for 20 hours . Co-infection movies were scored manually . For example co-infected macrophages that contained two Δplb1 ( non-fluorescent ) and one H99-GFP ( GFP positive ) yeast cells at 0 hr were tracked for 18 hours and before the burden of each strain within the macrophage was counted again . The IPR for Δplb1 within co-infected macrophages was calculated by dividing the number of Δplb1 cells within a macrophage at 18 hr by the number at 0 hr . J774 cells were cultured to confluency as discussed above and seeded onto sterile 13 mm glass coverslips at a density of 105 cells per ml without activation by phorbol 12-myristate 13-acetate . H99-GFP and Δplb1-GFP were opsonised with 18B7 for one hour . 18B7 was then removed by centrifugation and fungal cells were suspended in 1 ml of PBS with 1:200 FITC for one hour . Supernatant was removed again , cells were resuspended in PBS , and J774 were infected with 106 of either H99-GFP or Δplb1-GFP in serum free DMEM . After two hours of infection media was removed from J774 cells and the cells were washed three times with PBS . 10 μm TLT or DMSO was added to uninfected cells to act as controls . Cells were then left for 18 hours at 37 oC 5% CO2 . After 18 hours supernatants were removed and J774s were fixed with cold methanol for 5 minutes at -20 oC before washing with PBS three times , leaving PBS for five minutes at room temperature between washes . Coverslips were blocked with 5% sheep serum in 0 . 1% triton ( block solution ) for 20 minutes before being transferred into the primary PPAR-γ antibody ( 1:50 , Santa Cruz Biotechnology sc-7273 lot #B1417 ) with 1:10 human IgG in block solution for one hour . Coverslips were washed 3 times in PBS and incubated with 1:200 anti-mouse TRITC , 1:40 anti-human IgG , and 0 . 41 μl/ml DAPI in block solution for one hour . Coverslips were then washed three times with PBS , three times with water , and fixed to slides using MOWIOL . Slides were left in the dark overnight and imaged the following day . Imaging was performed on a Nikon Eclispe Ti microscope with a x60 DIC objective . Cells were imaged with filter sets for Cy3 ( PPAR-γ , 500ms exposure ) GFP ( Cryptococcus , 35 ms exposure ) and DAPI ( Nuclei , 5ms ) dyes in addition to DIC . The intensity of nuclear staining was analysed for at least 30 cells per coverslip , using ImageJ 2 . 0 . 0 a line ROI was drawn from the outside of cell , through the nucleus measuring the mean grey value along the line . For Cryptococcus infected conditions uninfected and infected cells were measured separately upon the same coverslip using the GFP channel to distinguish between infected and uninfected cells . Macrophages were seeded at 105 per ml into 24 well plates as described above . After two hours cells requiring aspirin were treated with 1 mM aspirin in DMSO in fresh DMEM . Cells were then incubated overnight for 18 hours at 37 oC 5% CO2 . H99-GFP and Δplb1-GFP were prepared at 106 cells per ml as described above , and opsonised with 18B7 for one hour . J774s were then infected with the fungal cells in fresh serum free DMEM for two hours before removing the supernatant , washing three times in PBS , and adding fresh serum free DMEM . Cells were imaged for 18 hours on a Nikon Eclispe Ti equipped with a climate controlled stage ( Temperature—37 oC , Atmosphere—5% CO2 / 95% air ) with a x20 Lambda Apo NA 0 . 75 phase contrast objective brightfield images were taken at an interval of 2 minutes , 50 ms exposure . Analysis was performed by manual counts of intracellular and extracellular cryptococci . Washed and counted Cryptococcus cells from overnight culture were pelleted at 3300g for 1 minute and re-suspended in 10% Polyvinylpyrrolidinone ( PVP ) , 0 . 5% Phenol Red in PBS to give the required inoculum in 1 nl . This injection fluid was loaded into glass capillaries shaped with a needle puller for microinjection . Zebrafish larvae were injected at 2-days post fertilisation; the embryos were anesthetised by immersion in 0 . 168 mg/ml tricaine in E3 before being transferred onto microscope slides coated with 3% methyl cellulose in E3 for injection . Prepared larvae were injected with two 0 . 5 nl boluses of injection fluid by compressed air into the yolk sac circulation valley . Following injection , larvae were removed from the glass slide and transferred to E3 to recover from anaesthetic and then transferred to fresh E3 to remove residual methyl cellulose . Successfully infected larvae ( displaying systemic infection throughout the body and no visible signs of damage resulting from injection ) were sorted using a fluorescent stereomicroscope . Infected larvae were maintained at 28 oC . All compounds were purchased from Cayman Chemical . Compounds were resuspended in DMSO and stored at -20 oC until used . Prostaglandin E2 ( CAY14010 , 10 mg/ml stock ) , Prostaglandin D2 ( CAY12010 , 10 mg/ml stock ) , 16 , 16-dimethyl-PGE2 ( CAY14750 , 10 mg/ml stock ) , 15-keto-PGE2 ( CAY14720 , 10 mg/ml stock ) , troglitazone ( CAY14720 , 10 mg/ml stock ) , GW9662 ( CAY70785 , 1 mg/ml stock ) , AH6809 ( CAY14050 , 1 mg/ml stock ) , GW627368X ( CAY10009162 , 10 mg/ml stock ) , NS-398 ( CAY70590 , 9 . 4 mg/ml stock ) , SC-560 ( CAY70340 , 5 . 3 mg/ml stock ) . Treatment with exogenous compounds during larval infected was performed by adding compounds ( or equivalent solvent ) to fish water ( E3 ) to achieve the desired concentration . Fish were immersed in compound supplemented E3 throughout the experiment from the time of injection . Individual infected zebrafish embryos were placed into single wells of a 96 well plate ( VWR ) with 200 ul or E3 ( unsupplemented E3 , or E3 supplemented with eicosanoids / drugs depending on the assay ) . Infected embryos were imaged at 0-days post infection ( dpi ) , 1 dpi , 2 dpi and 3 dpi in their 96 well plates using a Nikon Ti-E with a CFI Plan Achromat UW 2X N . A 0 . 06 objective lens . Images were captured with a Neo sCMOS ( Andor , Belfast , UK ) and NIS Elements ( Nikon , Richmond , UK ) . Images were exported from NIS Elements into Image J FIJI as monochrome tif files . Images were threshholded in FIJI using the ‘moments’ threshold preset and converted to binary images to remove all pixels in the image that did not correspond to the intensity of the fluorescently tagged C . neoformans . The outline of the embryo was traced using the ‘polygon’ ROI tool , avoiding autofluorescence from the yolk sac . The total number of pixels in the threshholded image were counted using the FIJI ‘analyse particles’ function , the ‘total area’ measurement from the ‘summary’ readout was used for the total number of GFP+ pixels in each embryo . PPARγ embryos [35 , 36] were collected from homozygous ligand trap fish ( F18 ) . Embryos were raised in a temperature-controlled water system under LD cycle at 28 . 5°C . in 0 . 5× E2 media in petri-dishes till 1 dpf . Any developmentally delayed ( dead or unfertilized ) embryos were removed . Chorions were removed enzymatically with Pronase ( 1 mg/ml ) and specimens were dispensed into 24 well plates ( 10 per well ) in 0 . 5 ×E2 media . For embryos 1% DMSO was used as a vehicle control . Chemicals were stored in DMSO , diluted appropriately and added individually to 400 ul of 0 . 5×E2 media with 0 . 05 U/ml penicillin and 50 ng/ml streptomycin and vortexed intensively for 1 min . 0 . 5×E2 media was removed from all wells with embryos and the 400 ul chemical solutions were administered to different wells to embryos . For drug treatment embryos were pre-incubated for 1 hour with compounds and then heat induced ( 28→37° C . ) for 30 min in a water bath . Embryos were incubated at 28°C for 18 h and then monitored using a fluorescent dissection scope ( SteREO Lumar . V12 Carl Zeiss ) at 2 dpf . For analyzing GFP fluorescent pattern , embryos were anesthetized with Tricaine ( Sigma , Cat . # A-5040 ) and mounted in 2% methyl cellulose . 2 dpf transgenic zebrafish larvae which have fluorescently tagged macrophages due to an mCherry fluorescent protein driven by the macrophage specific gene marker mpeg1 [56] Tg ( mpeg1:mCherryCAAX ) sh378 ) [22] were treated with 10 μM PGE2 , 10 μM 15-keto-PGE2 or an equivalent DMSO control for 2 days . Larvae were then anesthetized by immersion in 0 . 168 mg/ml tricaine in E3 and imaged using a Nikon Ti-E with a Nikon Plan APO 20x/ 0 . 75 DIC N2 objective lens , taking z stacks of the entire body with 15 μM z steps . Macrophage counts were made manually using ImageJ from maximum projections . J774 macrophages were seeded into 24 well plates at a concentration of 1x105 per well and incubated for 24 hours at 37 oC 5% CO2 . J774 cells were infected with C . neoformans as described above , at the same MOI 1:10 and incubated for 18 hours with 1ml serum free DMEM . At 18 hours post infection the supernatant was removed for ELISA analysis . For ELISA analysis with aspirin treatment , wells requiring aspirin had supernatants removed and replaced with fresh DMEM containing 1 mM aspirin in 1% DMSO . Cells were left for 24 hours total . At 24 hours all wells received fresh serum free media . Wells requiring aspirin for the duration received 1 mM aspirin in DMSO . Aspirin treated cells requiring arachidonic acid were treated with 30 μg per ml arachidonic acid in ethanol . Control wells received the following: either 1% DMSO , 30 μg per ml arachidonic acid , or ethanol . Cells were again left at 37 oC 5% CO2 for 18 hours . Supernatants were then removed and frozen at -80 oC until use . Supernatants were analysed as per the PGE2 EIA ELISA kit instructions ( Cayman Chemical ) . J774 macrophages were seeded into T25 tissue culture flasks at a concentration of 1 . 3x106 cells per flask and incubated for 24 hours at 37 oC 5% CO2 . J774 cells were infected with C . neoformans as described above , at the same MOI 1:10 and incubated for 18 hours with 2 ml serum free DMEM . At 18 hours post infection infected cells were scraped from the flask with a cell scraper into the existing supernatant and immediately snap frozen in ethanol / dry ice slurry . All samples were stored at -80 oC before analysis . Lipids and lipid standards were purchased from Cayman Chemical ( Ann Arbor , Michigan ) . Deuterated standard Prostaglandin E2-d4 ( PGE2-d4 ) , ≥98% deuterated form . HPLC grade solvents were from Thermo Fisher Scientific ( Hemel Hempstead , Hertfordshire UK ) . Lipids were extracted by adding a solvent mixture ( 1 mol/L acetic acid , isopropyl alcohol , hexane ( 2:20:30 , v/v/v ) ) to the sample at a ratio of 2 . 5–1 ml sample , vortexing , and then adding 2 . 5 ml of hexane [57] . Where quantitation was required , 2 ng PGE2-d4 , was added to samples before extraction , as internal standard . After vortexing and centrifugation , lipids were recovered in the upper hexane layer . The samples were then re-extracted by addition of an equal volume of hexane . The combined hexane layers were dried and analyzed for Prostaglandin E2 ( PGE2 ) using LC-MS/MS as below . Lipid extracts were separated by reverse-phase HPLC using a ZORBAX RRHD Eclipse Plus 95Å C18 , 2 . 1 x 150 mm , 1 . 8 μm column ( Agilent Technologies , Cheshire , UK ) , kept in a column oven maintained at 45°C . Lipids were eluted with a mobile phase consisting of A , water-B-acetic acid of 95:5:0 . 01 ( vol/vol/vol ) , and B , acetonitrile-methanol-acetic acid of 80:15:0 . 01 ( vol/vol/vol ) , in a gradient starting at 30% B . After 1 min that was ramped to 35% over 3 min , 67 . 5% over 8 . 5 min and to 100% over 5 min . This was subsequently maintained at 100% B for 3 . 5 min and then at 30% B for 1 . 5 min , with a flow rate of 0 . 5 ml/min . Products were monitored by LC/MS/MS in negative ion mode , on a 6500 Q-Trap ( Sciex , Cheshire , United Kingdom ) using parent-to-daughter transitions of m/z 351 . 2 → 271 . 2 ( PGE2 ) , and m/z 355 . 2 → 275 . 2 for PGE2-d4 . ESI-MS/MS conditions were: TEM 475°C , GS1 60 , GS2 60 , CUR 35 , IS -4500 V , dwell time 75 s , DP -60 V , EP -10 V , CE -25 V and CXP at -10 V . PGE2 was quantified using standard curves generated by varying PGE2 with a fixed amount of PGE2-d4 .	Cryptococcus neoformans is an opportunistic fungal pathogen that is responsible for significant numbers of deaths in the immunocompromised population worldwide . Here we address whether eicosanoids produced by C . neoformans manipulate host innate immune cells during infection . Cryptococcus neoformans produces several eicosanoids that are notable for their similarity to vertebrate eicosanoids , it is therefore possible that fungal-derived eicosanoids may provoke physiological effects in the host . Using a combination of in vitro and in vivo infection models we identify a specific eicosanoid species—prostaglandin E2 –that is required by C . neoformans for growth during infection . We subsequently show that prostaglandin E2 must be converted to 15-keto-prostaglandin E2 within the host before it has these effects . Furthermore , we find that prostaglandin E2/15-keto-prostaglandin E2 mediated virulence is via activation of host PPAR-γ –an intracellular eicosanoid receptor known to interact with 15-keto-PGE2 .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "blood", "cells", "cryptococcus", "neoformans", "medicine", "and", "health", "sciences", "neurochemistry", "fish", "cryptococcus", "immune", "cells", "pathology", "and", "laboratory", "medicine", "enzymes", "pathogens", "eicosanoids", "immunology", "microbiology", "verteb...	2019	15-keto-prostaglandin E2 activates host peroxisome proliferator-activated receptor gamma (PPAR-γ) to promote Cryptococcus neoformans growth during infection
Cholesteryl ester transfer protein ( CETP ) transports cholesteryl esters , triglycerides , and phospholipids between different lipoprotein fractions in blood plasma . The inhibition of CETP has been shown to be a sound strategy to prevent and treat the development of coronary heart disease . We employed molecular dynamics simulations to unravel the mechanisms associated with the CETP-mediated lipid exchange . To this end we used both atomistic and coarse-grained models whose results were consistent with each other . We found CETP to bind to the surface of high density lipoprotein ( HDL ) -like lipid droplets through its charged and tryptophan residues . Upon binding , CETP rapidly ( in about 10 ns ) induced the formation of a small hydrophobic patch to the phospholipid surface of the droplet , opening a route from the core of the lipid droplet to the binding pocket of CETP . This was followed by a conformational change of helix X of CETP to an open state , in which we found the accessibility of cholesteryl esters to the C-terminal tunnel opening of CETP to increase . Furthermore , in the absence of helix X , cholesteryl esters rapidly diffused into CETP through the C-terminal opening . The results provide compelling evidence that helix X acts as a lid which conducts lipid exchange by alternating the open and closed states . The findings have potential for the design of novel molecular agents to inhibit the activity of CETP . Cholesteryl ester transfer protein ( CETP ) is a 476-residue-long glycoprotein which promotes the transfer of cholesteryl esters ( CEs ) , triacylglycerols ( TGs ) and phospholipids ( PLs ) between the different lipoprotein fractions ( high density lipoprotein ( HDL ) , low density lipoprotein ( LDL ) , and very low density lipoprotein ( VLDL ) ) in human blood plasma . CETP is believed to mediate the transfer by a hetero-exchange mechanism in which CEs are carried from HDL to VLDL and LDL particles , and TGs are carried in the opposite direction from VLDL and LDL to HDL particles , resulting in CE depletion and TG enrichment of HDL [1] . Interestingly , CETP is structurally homologous to the phospholipid transfer protein ( PLTP ) , the lipopolysaccharide binding protein ( LBP ) , and the bactericidal/permeability-increasing protein ( BPI ) [1] . As all these proteins are able to bind phospholipids , similarity in their transportation mechanisms has been suggested . Importantly , however , CETP is the only protein able to transfer neutral lipids ( cholesteryl esters and triglycerides ) in human plasma [2] . The broad interest to understand CETP and its lipid trafficking properties stems from the fact that it has a potentially protective role in the development of cardiovascular diseases , in particular atherosclerosis , which are currently the main cause of death in Western countries , claiming ∼17 million lives a year . The role of CETP in the development of atherosclerosis became evident when it was found that CETP deficiency and the inhibition of CETP lower LDL and increase HDL levels in human plasma [3] . High HDL levels have been clinically found to be inversely correlated with the development of atherosclerosis , since HDL particles are considered crucial components in the transport of cholesterol from atherosclerotic plaques back into the systemic circulation . Unfortunately , the clinical trial with the first oral anti-atherogenic drug candidate with a CETP-inhibitory activity , torcetrapib , was unsuccessful because of its potentially lethal side effects [4] . Treatment with torcetrapib increased blood pressure and circulating aldosterone levels and also altered serum electrolyte levels . However , subsequent studies indicated that these adverse effects of torcetrapib were unrelated to the inhibition of CETP and are not necessarily shared by the other members of the class of CETP inhibitors . Indeed , a recent clinical trial showed that another CETP inhibitor , anacetrapib , effectively raises HDL and has an acceptable side-effect profile in patients with coronary heart disease or risk factors for coronary heart disease [5] . Importantly , a recent meta-analysis of 92 studies involving 113 , 833 participants concluded that the CETP genotypes that have lower CETP activity are associated with a decreased coronary risk [6] . Considering the central role of CETP in the development of coronary atherosclerosis and its complications , we face an outstanding challenge to better understand the mechanisms associated with CETP functions . Recently , Qiu et al . resolved the X-ray structure of CETP showing that it carries CE molecules inside a long hydrophobic tunnel , whose ends are plugged by phospholipids ( Figure 1 ) [7] . This kind of hydrophobic tunnel is unique among proteins , and it was speculated that CEs diffuse into and out from the tunnel through the two tunnel openings , which are closed by PLs during the transportation in aqueous surroundings . In addition , based on the X-ray structure it has been speculated that CETP is attached to lipoproteins via its concave surface where also the two hydrophobic tunnel openings reside [7] . Further , it has been proposed that the formation of CETP-lipoprotein complexes is modulated by pH , surface pressure , and the ionic interactions between CETP and phospholipids [8] , [9] . Fluorescence quenching has been used to demonstrate that the interaction between tryptophan residues of CETP and PLs could be important in the attachment [10] . Regarding the lipid exchange mechanism of CETP , helix X has been suggested to play a role in lipid loading and unloading by acting as a lid at the C-terminal tunnel opening , being in the open state when the exchange of lipids takes place , and in the closed state when CETP detaches from the lipoprotein surface to become surrounded by aqueous medium [5] . Various mutational studies further suggest that helix X is possibly crucial in the transfer of CEs and TGs but not in the transfer of PLs [11] , [12] . The above findings and suggestions are appealing and insightful , but call for better understanding of the structure-function relationship and of the dynamics that drive CE , TG and PL transfer . In essence , atomic and molecular scale insight into the lipid exchange between CETP and lipoproteins is limited , which largely stems from exceptional difficulties to experimentally probe the related transient processes in the nanometer scale . In the current study , our objective is to complement experiments through atomistic and coarse-grained molecular dynamics simulations to investigate the binding of CETP to a small lipid droplet and a planar lipid trilayer , and to determine the initial stages of the lipid exchange mechanism . By doing so , we can follow the lipid exchange in atomic detail , shed light on its mechanism , consider the effect of lipoprotein curvature , and unravel the dynamics of the related processes . These mechanisms and phenomena are considered over a multitude of time scales by bridging atomistic and coarse-grained simulations , which are shown to provide consistent results . The present study paves the way for future simulations to elucidate interactions of anacetrapib with CETP and CETP-lipoprotein complexes , with an objective to unlock its inhibitory mechanism . Given the significant role of CETP in cardiovascular diseases , the broad interest of the topic is hoped to attract substantial interest to extend the present work . We carried out three 100 ns atomistic simulations for fully hydrated systems containing CETP with different interior lipid compositions and a small pre-equilibrated HDL-sized lipid droplet composed of POPCs and CEs ( A1 , A2 , A3; Figure 1; see Materials and Methods ) . In addition to these spherical droplets , CETP was simulated with a pre-equilibrated planar POPC-CE trilayer system ( A4; Figure 1C ) to study the effect of less curved lipoprotein particles , like VLDL and LDL , on the conformation of CETP . Root mean square deviation ( RMSD ) profiles indicate that the structures do not deviate considerably from the X-ray structure ( Figure 2B ) . The radius of gyration fluctuated between 3 . 2 and 3 . 5 nm , and its profiles together with snapshots from simulation trajectories show that the conformation of CETP is able to bend to bind to surfaces with different curvatures ( Figure 2A ) . In the case of spherical A3 the curvature of CETP is clearly higher than in the planar A4 system , and it became apparent that the conformation of CETP is not able to rearrange sufficiently to fully match the planar surface . Nonetheless , our results imply that the structure of CETP is elastic and facilitates the binding of CETP to different lipoprotein surfaces with varying curvatures . Yet , due to its inherent curvature that closely matches the curvature of HDL , CETP prefers to bind to HDL-sized particles compared to larger VLDL-sized particles . Consequently , we propose that the free energy change associated with the binding of CETP to HDL is more favorable compared to the formation of a CETP-VLDL complex . Radial distribution functions and density profiles indicate that CETP does not penetrate deeper than to the level of POPC phosphate groups ( Figure 2C ) . Therefore , in all atomistic simulations the core CEs were observed not to interact directly with CETP , as instead they were found to reside only in the core . This suggests that the surface-core lipid ratio is important for the exchange of neutral lipids by CETP . During the simulation A2 , CETP-bound DOPCs did not diffuse into the lipid droplet . However , we found that during the simulation S1 the hydrophobic tunnel of CETP collapsed , which strongly suggests that the structure of CETP is not stable without interior lipids ( See Figure S2 and Text S2 ) . This finding is important regarding the lipid exchange process of CETP as it suggests that during the neutral lipid exchange , the hydrophobic cavity is not empty at any point . We return to this matter later . In atomistic lipid droplet simulations , we calculated the number of salt bridges that formed between CETP and POPCs as a function of time , in order to characterize the key charged residues involved in the attachment of CETP . The number of salt bridges that formed between the positively charged lysine residues of CETP and the negatively charged phosphate ( P ) groups of POPCs stabilized to a level of 12–20 ( Figure 3A ) . Salt bridging of lysines is much more efficient in A2 and A3 than in A1 ( 19–20 compared to 12 , see Figure 3A ) . The number of salt-bridges between arginines and P groups was on average two or three . Additionally , we calculated the number of salt bridges formed by the negatively charged Asp and Glu residues and found that Asp residues were able to form 6–8 and Glu residues 2–4 salt bridges with the positively charged choline groups . Amino acids that form most of the salt bridges are shown in Figure 3 , revealing that they are mainly located at the edge of the concave surface of CETP . In the spirit of the earlier Trp quenching study [10] , we inspected more carefully the behavior of Trps during binding . In all droplet simulations , Trp299 formed hydrogen bonds with POPCs ( Figure 3B ) . Trp264 stayed buried inside the structure of the protein and Trp162 was able to interact with the water molecules . In A1 and A2 , Trp105 and Trp106 were located facing the water phase , while in A3 the flap Ω5 interacted with POPCs by anchoring Trps 105 and 106 to the carbonyl region of POPC surface , highlighted in Figure 2 . In the trilayer simulation only two Trp residues ( 105 and 299 ) were able to interact with the POPC surface . Our results highlight the importance of electrostatic interactions between CETP and phospholipids in the formation of CETP-droplet complexes . The results provide compelling evidence that three Trp residues anchor CETP to lipid droplets , introducing additional stability to CETP-lipoprotein complexes where the curvature of CETP and a lipoprotein matches . Interpretation of atomistic simulations requires care due to the limited time and length scales that are feasible through atomistic studies . For example , the diffusion of lipids in HDLs is slow compared to the time scales we have simulated and , thus , claims regarding the principal binding site and penetration depth of CETP must be carefully considered . In order to add liability to our atomistic simulations , we also carried out coarse-grained simulations , covering time scales beyond 2 µs . CG simulations support and validate atomistic simulations by showing that the concave surface is the principal lipoprotein binding site of CETP . We did not observe any deviations from this conclusion during the three independent 2-microsecond simulations . Radial distribution functions shown in Figure 2 depict a similar distribution of molecules as in atomistic simulations . However , intriguingly we found that POPCs which were in contact with the concave surface of CETP migrated away from the tunnel openings , forming a small hydrophobic patch under the concave surface ( Figure 4A ) . In essence , CETP drives phospholipids to diffuse away from the slightly hydrophobic tunnel openings to its edges where most of the salt bridge-forming amino acids reside . We analyzed the spatial densities of the polar beads ( GL1 , GL2 , and NC3 , PO4 , using the descriptions of the Martini model ) of POPCs to clarify the patch formation more clearly . The spatial density map revealed the formation of a hydrophobic patch under the concave surface , and specifically in the region where the N- and C-terminal tunnel openings reside ( see Figure S1 and Text S1 in Supporting Information ( SI ) ) . The time associated with the formation of the hydrophobic patch is difficult to estimate accurately , since the process fluctuates depending on the dynamics of the CETP-droplet complex . In practice we found the patch to emerge in roughly 10–40 ns , and it increased to a size of about 1 nm×3 nm in 100–500 nanoseconds , depending on the system studied ( see Figure S1 ) . At longer times the patch fluctuated quite a lot but there was a trend showing a slow increase in size , suggesting that the total formation time may be of the order of microseconds . To gain further support for patch formation , as predicted by CG simulations , we repeated the analysis with two additional CG simulations where we used PME for electrostatics with the non-polarizable Martini water model , and PME with the polarizable Martini water model [13] . With the polarizable water model the solubility of charged species to apolar media should be better described compared to the standard Martini model . In both additional CG simulations , hydrophobic patch formation was observed too ( Figure S1 ) . Given that the patch emerged in CG simulations in tens of nanoseconds , and the atomistic simulations lasted for 100 ns , we returned to our atomistic simulation data to consider this aspect in atomic detail . We analyzed the spatial density profile of POPC head groups and observed similar hydrophobic patch formation . For example , in A3 we noticed that a small hydrophobic patch was formed under CETP already in ∼20 ns , and the patch slowly grew in size and number as in ∼80–100 ns there were two patches close to one another ( data not shown ) . The fact that also 100 ns atomistic simulations show the hydrophobic patch formation confirms that the CETP-lipoprotein interaction is strong specifically under the concave surface and promotes the formation of a path between the droplet core and CETP . The hydrophobic patch formation exposes the hydrophobic parts of the lipids to the concave surface where the hydrophobic tunnel openings are located . However , hardly any of the CEs were in contact with CETP , as can be seen from Figure 4 . Thus , we reduced the number of POPCs from 180 to 90 ( CG3-90POPC ) and simulated the system again for 2 µs in order to see if the lower surface pressure of a lipid droplet would promote the solubility of CEs to the surface lipid monolayer and the interaction between CETP and CEs . Indeed , the contacts between core CEs and CETP increased . Clearly , the concave surface of CETP has some affinity for CEs , over random thermal fluctuations , as the hydrophobic patch under CETP guides core CEs to the concave surface . However , the surface pressure must be low enough for CEs to localize to the surface monolayer and CETP to bind to the surface . This implies that the ratio of surface and core lipids ( surface pressure ) and the formation of the small hydrophobic patch under CETP are important factors modulating the core lipid transfer activity of CETP . Root mean square fluctuations ( RMSFs ) of the protein backbone were analyzed after 40 ns of atomistic lipid droplet simulations to find the regions , which wobble the most after CETP attached to the surface of the lipid droplet ( Figure 5 ) . This was done as follows . First , the RMSF of backbone atoms was computed by fitting the atomic positions to the reference structure ( average structure of CETP after its binding to the lipid droplet surface ) and then calculating the average distance deviation from the reference structure . The RMSFs of individual backbone atoms were then averaged per residue to determine the residual RMSF profile . Similar results were observed in all droplet simulations . The N- and C-terminal ends and loop regions ( marked by omegas ) of CETP showed high fluctuations , as expected . We also found that in the helix X region ( residues 460–476 ) the conformational fluctuations peaked near the residue 462 . This region has previously been proposed to be a potential hinge region of helix X with elevated B-factors [7] . In addition , it was found that the flaps Ω1 and Ω2 resulted in high fluctuations to the RMSF profile as was also proposed based on the B-factors of the X-ray structure of CETP [7] . In addition to the suggested high fluctuations , we found that another five regions of CETP were also highly fluctuating in each simulation . These regions were Ω3 ( residues 380–400 ) , Ω4 ( residues 40–50 ) , Ω5 ( residues 90–110 ) , Ω6 ( residues 150–170 ) , and Ω7 ( residues 230–260 ) . All regions are found in the loops , and hence high fluctuations can be expected . Previously it has been speculated that the hinge region could promote the needed flexibility to helix X that is important in the lipid exchange process [5] . To study further the role of helix X in lipid exchange , we did two additional atomistic simulations to probe its role in the lipid exchange process , see below . Earlier point and deletion mutations suggest that helix X is important in the transfer of core lipids , while it is not needed in phospholipid transfer [12] . Since we found that the hydrophobic patch was formed under the concave surface of CETP in both CG and atomistic simulations , we asked if the fully formed hydrophobic patch could induce changes to the conformation of helix X . To test this hypothesis , we did one additional atomistic simulation with 90 POPCs ( that is , starting from the system A3-90POPC ) where we expanded the hydrophobic patch under the concave structure by removing POPCs near the two tunnel openings of CETP , so that helix X was only able to interact with the hydrophobic parts of POPCs and CEs . Here , it is worth to mention that atomistic simulations are the only method of choice for this purpose , since this kind of conformational change can not take place in our CG simulations , where we used the elastic network model to keep the secondary structure of CETP stable [14] . We found that the conformation of helix X rearranged and became buried inside the hydrophobic cavity of CETP , where it interacted with CETP-bound CE ( Figure 6 ) . This conformational change generated a hydrophobic pathway from the droplet surface to the tunnel , increasing the accessibility of core CEs to the hydrophobic tunnel of CETP . To further assess the regulatory role of helix X , we created a deletion mutant of CETP , in which the residues 462–476 ( helix X ) were removed from the structure , and we simulated this structure for 80 ns . Deletion mutation simulation revealed that three CEs readily diffused into CETP when helix X was completely removed from the structure ( see the spatial density maps and the number of contacts plot in Figures 6B and 6C ) . This provides further support for the view that helix X acts as a lid at the C-terminal tunnel opening , and that its conformation regulates the accessibility of CEs to the hydrophobic tunnel . Previously , the role of electrostatic interactions in the formation of isolable CETP-lipoprotein complexes was demonstrated by Pattnaik et al . , who showed that ( in addition to CETP-HDL complexes ) CETP was able to form isolable complexes with LDL and VLDL particles when negative surface charge was increased by phospholipase A2 digestion or by acylation of phospholipid amino groups . They reached the conclusion that the phospholipid phosphate groups are the primary sites for the interaction of lipoproteins with CETP [8] . They also found that the formation of isolated CETP-HDL complexes was hindered by decreasing the pH , introducing positive divalent ions into the solution , or by digesting lipoproteins by phospholipase C . Moreover , Nishida et al . reported that the affinity of CETP for various lipoproteins is governed by a delicate balance of electrostatics and hydrophobic interactions [15] . The importance of electrostatic interaction in CETP binding has been shown also by several point mutation studies applied to the positively charged lysine residues of CETP [16] . Our results are in agreement with experiments , as in the present simulations most of the salt-bridges with the negatively charged phosphate groups of POPCs were formed by the Lys residues at the concave surface , when CETP fastened to the lipid surface . However , also Glu and Asp residues formed salt-bridges with the positively charged choline groups of POPCs , although the ratio of salt-bridges formed by the positively and negatively charged amino acids is approximately 1 . 8 , indicating that mostly the positively charged amino acids contribute to the formation of CETP-lipoprotein complexes . Another important factor playing a role in the binding of CETP is Trp residues located in the flaps Ω5 and Ω1 that were found to become buried into the lipid matrix . Most likely Trp residues add more stability to the CETP-lipoprotein complexes by anchoring CETP to a lipoprotein surface . Interestingly , Desmuraux et al . made mutations to the structurally similar flap Ω5 region ( Trp-91 , Phe-92 and Phe-93 ) of PLTP and showed that the phospholipid transfer activity of PLTP from liposomes to HDL particles decreased up to 60% [17] . This finding together with our results suggests that the flexible flap Ω5 region of CETP and Trp residues therein are crucial in the binding of CETP to HDL particles , playing an important role in the CETP-mediated lipid transfer . Penetration depth of CETP is an important factor in CETP-mediated lipid exchange , as it determines how efficiently the neutral core lipids are able to interact with CETP . Previous studies have shown that the exclusion pressure of CETP is lower than the exclusion pressure of other apolipoproteins , like apoA-I [9] , [10] . Moreover , it has been argued that the weaker penetration of CETP to the emulsion particles compared to apoA-I makes the activation energy of the attachment and detachment of CETP lower , rendering the transportation process more efficient [9] , [10] . Our atomistic and CG simulation results showed that CETP is not able to bury its amino acid residues deeper than to the level of the phosphate groups of POPCs . The above findings therefore imply that core lipids have to diffuse to or reside at the surface to enter CETP . Therefore , the amount of core lipids at the lipoprotein surface is an important factor modulating the activity of CETP , as has been suggested previously based on liposome studies [18] , and it can be promoted by defects as is outlined below . The number of surface-located neutral lipids can be regulated by the lipid and apolipoprotein composition of lipoprotein particles . Interestingly , we found that when CETP attaches to the surface by the aid of electrostatic interactions , the head groups of POPCs moved aside , providing access to the hydrophobic lipid region . In this manner , the two tunnel openings of the concave surface are exposed to the hydrophobic lipid matrix of the lipoprotein . The hydrophobic patch formation facilitates , by generating a defect to the surface monolayer , the diffusion of core lipids to the surface monolayer region located under CETP ( Figure 6B ) . Thus , the localization of neutral lipids at the surface monolayer itself is not crucial to allow CETP to exchange neutral lipids between lipoproteins but the neutral lipids can enter CETP through the formation of the hydrophobic patch . Consequently , we envision that the activity of CETP could be inhibited by nonpolar drugs that are transferred into the hydrophobic tunnel of CETP through the hydrophobic core of lipoproteins . Further , we observed that the concave surface interacted directly with CEs that diffused more readily to the hydrophobic tunnel openings when the surface-core lipid ratio was decreased . Finally , we found CEs to diffuse into the hydrophobic tunnel of CETP and interact with CETP-bound CE when the conformation of helix X was in the open state or completely removed . A previous mutational study argued that CETP is not able to transfer neutral lipids when helix X is removed from the structure [11] , [12] . However , CETP is able to transfer phospholipids without helix X . Our results showed that the conformation of helix X rearranges , and helix X moves inside the hydrophobic tunnel of CETP where it can interact with CETP-bound CE . Given this , we suggest that there are two important functional properties of helix X that make the neutral lipid exchange possible . First , helix X is able to facilitate the neutral lipid exchange by opening the hydrophobic pathway from a lipoprotein surface to the hydrophobic tunnel of CETP . Second , helix X promotes the diffusion of neutral lipids from the hydrophobic tunnel to lipoproteins by filling the volume of CETP-bound neutral lipid when it diffuses out from CETP . Afterwards , another neutral lipid from the lipoprotein core or inside CETP could take the place of helix X after which the C-terminal tunnel opening closes again . We propose that helix X is needed to prevent the structure of CETP from collapsing as was registered in the simulation A1 without the CETP-bound lipids . The above reasons would explain why helix X is important in the neutral lipid exchange , but not in the exchange of phospholipids . It is tempting to contemplate the possible roles of helix X in the inhibition of CETP . It has been reported that dalcetrapib , a novel CETP inhibitor , binds covalently to CETP by forming a disulphide bond with Cys-13 , which is located inside the hydrophobic tunnel of CETP [19] . In addition , it has been suggested that the disulphide bond formation is a necessary requirement for the dalcetrapib-mediated CETP inhibition . However , that is not the case with torcetrapib or anacetrapib ( another novel CETP inhibitors ) , both of which bind reversibly to CETP [19] . Yet , all inhibitors stabilize HDL-CETP complexes , which has been found to be the second major inhibitory mechanism of the neutral lipid transfer exerted by the synthetic CETP inhibitors [19] . Based on our simulations , we can hypothesize that helix X is locked to the open state when inhibitors are bound to CETP . The driving force for this could be the small size of an inhibitor that enforces helix X to be located inside the hydrophobic tunnel and , thus , prevents collapse of the lipid pocket . Another reason could be the more favorable interaction between helix X and the inhibitor inside the tunnel , which could conceivably force the conformation of helix X to the open state , or change the conformation distribution to favor the open state . Consequently , the detachment of CETP from the surface of HDL would be hindered since the helix X is not able to shield the hydrophobic tunnel opening of the lipid pocket when CETP is completely in the aqueous phase . In addition , the open state could prevent the binding of phospholipids to the C-terminal tunnel opening , which , based on the X-ray structure of CETP , is known to be occupied by phospholipids when CETP is not attached to a lipoprotein surface . A reduced ability of CETP to bind and transport phospholipids could further stabilize the HDL-CETP complex . In summary , we have provided a detailed atomistic picture regarding the initial steps in the lipid exchange mechanism of CETP and , furthermore , we have offered a plausible mechanism for the exchange of neutral lipids mediated by CETP . Overall , our work paves the way for additional future studies to elucidate interactions of the available promising CETP inhibiting drugs , such as anacetrapib and dalcetrapib , with CETP and CETP-lipoprotein complexes . Our findings for the factors that affect the lipid exchange process can also be exploited in the design of novel molecular agents capable of inhibiting the activity of CETP , one possible strategy being the design of nonpolar drugs which can be transferred into the hydrophobic tunnel of CETP . Together with recent simulation models for both HDL and LDL [20] , these ideas are a reasonable goal already at present . The coordinate file of CETP in the PDB format with an accession code 2OBD was acquired from the RCSB Protein Data Bank . In addition to the protein , the structure provides information of the lipids carried by CETP: there are two CEs located inside the long hydrophobic tunnel of CETP , and two dioleoylphosphatidylcholine ( DOPC ) lipids that cover the two endings of the hydrophobic tunnel . The charge state of CETP was chosen to represent the physiological pH that is 7 . 4 . A detailed explanation of the protein structure is given elsewhere [7] . For atomic-scale simulations , three different setups were constructed by combining pre-equilibrated lipid droplets consisting of 180 palmitoyloleoyl-PC ( POPC ) and 35 CE molecules [21] . In each system , CETP was placed approximately at a distance of 1 nm from the surface of the lipid droplet ( Figure 2 ) . In the first simulated system ( A1 ) , the two DOPCs and CEs were removed from CETP . In the second simulation ( A2 ) , both DOPCs and CEs were included , while in the third simulation ( A3 ) only CEs were kept inside CETP . We also simulated CETP with a planar trilayer system composed of 512 POPCs and 796 CEs ( A4 ) . The droplet systems were solvated with ∼180 , 000 water molecules at a salt concentration of 0 . 2 M including counter ions , while the trilayer system included ∼50 , 000 water molecules . Altogether , the systems included ∼500 , 000 atoms . Finally , three additional atomistic systems were constructed to characterize the role of helix X in lipid exchange in more detail ( see text ) . First , we studied the effect of the hydrophobic patch on the structure of helix X by removing half of the POPCs from A3 at 100 ns ( A3-90POPC ) . Second , we also considered CETP through its helix X deletion mutant to probe the regulatory role of the helix . Third , we used A3-90POPC as a basis and removed some of the surface lipids to model the complete formation of a hydrophobic patch under the concave surface of CETP . The context of these simulations to the studied processes will become clear in the discussion below . In addition to atomistic simulations , we carried out four coarse-grained ( CG ) simulations . First , the system A3 was directly coarse grained ( in the text , we refer to this simulation as CG3 ) by using a script that is available at the homepage of the Martini force field . Second , 90 POPCs were removed from CG3 , ending up in the system denoted as CG3-90POPC . The systems CG3 and CG3-90POPC were simulated under standard Martini model ( see below ) conditions with regard to electrostatics ( using truncation of electrostatic interactions ) and the water model that is non-polarizable . To clarify the influence of long-range interactions and the water model , we simulated two additional systems . That is , in the third case we focused on the system CG3 which was simulated with full electrostatics using the particle mesh Ewald ( PME ) method [22] . Finally , in the fourth coarse-grained model , we simulated the system CG3 using both PME and the polarizable Martini water model [13] . The GROMACS simulation package with version 4 . 0 was used in the simulations [23] . In atomistic studies , we used the Nose-Hoover thermostat [24] , [25] with a coupling constant of 1 . 0 ps to set the temperature to 330 K in which the particle core is certainly in liquid state . The pressure was set to 1 bar using the Parrinello-Rahman barostat [26] with isotropic pressure coupling and a coupling constant of 0 . 1 ps . The van der Waals interactions were chosen to have a cutoff at 1 . 0 nm . Electrostatic interactions were evaluated by the particle mesh Ewald technique with a real space cut-off of 1 . 0 nm [22] . Water molecules were described using the SPC water model . All non-water bonds were constrained using the LINCS algorithm [27] and the SETTLE algorithm [28] was used to constrain water molecules , allowing the use of a time step of 2 fs in the integration of equations of motion . Berger parameters [29] were used for lipids , while the GROMOS53A6 force-field [30] was employed for the protein . Combination rules were introduced for the interactions between lipids and the protein . The four leading atomistic systems ( A1–A4 ) were simulated for 100 ns , and the last two ones that focused on helix X for 80 ns . The total simulation time of atomistic simulations was 0 . 56 µs . CG simulations were also carried out with GROMACS , using the Martini force field with an extension to proteins [31] , [32] . The ElNeDyn elastic network model was used to keep the structure of CETP stable [14] . The Berendsen thermostat and barostat were used with time constants of 1 . 0 ps . Temperature was set to 320 K and isotropic pressure coupling was used with pressure set to 1 bar . Cut-off distance for electrostatic interactions was set to 1 . 2 nm . For Lennard-Jones interactions we used a cut-off of 1 . 2 nm , and Lennard-Jones interactions were shifted to zero from 0 . 9 nm . Time step was 25 fs . The simulation time of each CG system was beyond 2 µs , and the times reported here are given in units of the effective Martini time . All rendered figures were done by VMD [33] .	Coronary heart disease is a major cause of death in the Western societies . One of the most promising interventions to prevent and slow down the progress of coronary heart disease is the elevation of high density lipoprotein ( HDL ) levels in circulation . Animal models together with early clinical studies have shown that the inhibition of cholesteryl ester transfer protein ( CETP ) is a promising strategy to achieve higher HDL levels . However , drugs with acceptable side-effects for CETP-inhibition do not yet exist , although the next generation CETP inhibitor ( anacetrapib ) has great potential in this regard . In this study , our objective is to gain more detailed information regarding the interactions of CETP with lipoprotein particles . We show how the CETP-lipoprotein complex is formed and how lipid exchange between CETP and lipoprotein particles takes place . Our findings help to understand in a mechanistic way how CETP-mediated lipid exchange occurs and how it could be exploited in the design of new and more efficient molecular agents against coronary heart disease .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "biology", "computational", "biology", "biophysics" ]	2012	Lipid Exchange Mechanism of the Cholesteryl Ester Transfer Protein Clarified by Atomistic and Coarse-grained Simulations
Predicting organismal phenotypes from genotype data is important for plant and animal breeding , medicine , and evolutionary biology . Genomic-based phenotype prediction has been applied for single-nucleotide polymorphism ( SNP ) genotyping platforms , but not using complete genome sequences . Here , we report genomic prediction for starvation stress resistance and startle response in Drosophila melanogaster , using ∼2 . 5 million SNPs determined by sequencing the Drosophila Genetic Reference Panel population of inbred lines . We constructed a genomic relationship matrix from the SNP data and used it in a genomic best linear unbiased prediction ( GBLUP ) model . We assessed predictive ability as the correlation between predicted genetic values and observed phenotypes by cross-validation , and found a predictive ability of 0 . 239±0 . 008 ( 0 . 230±0 . 012 ) for starvation resistance ( startle response ) . The predictive ability of BayesB , a Bayesian method with internal SNP selection , was not greater than GBLUP . Selection of the 5% SNPs with either the highest absolute effect or variance explained did not improve predictive ability . Predictive ability decreased only when fewer than 150 , 000 SNPs were used to construct the genomic relationship matrix . We hypothesize that predictive power in this population stems from the SNP–based modeling of the subtle relationship structure caused by long-range linkage disequilibrium and not from population structure or SNPs in linkage disequilibrium with causal variants . We discuss the implications of these results for genomic prediction in other organisms . Most efforts to understand the genetic architecture of quantitative traits have focused on mapping the variants causing phenotypic variation in quantitative trait locus ( QTL ) mapping populations derived from crosses between lines genetically divergent for the trait , or in association mapping populations , with the goal of understanding the biological underpinnings of trait variation [1] . However , the ability to accurately predict quantitative trait phenotypes from information on genotypic variation in the absence of knowledge of causal variants will revolutionize evolutionary biology , medicine and human biology , and breeding of agriculturally important plant and animal species . The premise of personalized medicine is based on prediction of individual genetic risk to disease from genome-wide association studies [2] , [3] , and the ability to select individuals or lines in animal and plant breeding programs based on genotypic information circumvents the costly process of progeny testing and reduces the generation interval in applied breeding programs , leading to greater efficiency [4] , [5] . In classical animal and plant breeding , the genetic quality of individuals or lines is predicted from phenotypic values of selection candidates and their relatives . The widely used Best Linear Unbiased Prediction ( BLUP , [6] ) method models the covariance structures between individuals via the numerator relationship matrix , which is constructed from known pedigree information and thus reflects expected relationships between individuals ( i . e . the proportion of shared alleles of identical ancestral origin ) given the pedigree . The advent of high-throughput genotyping platforms for many agronomic species [7] enabled genotyping large numbers of individuals for dense panels of single nucleotide polymorphisms ( SNPs ) spanning the genome . The expected , pedigree-based numerator relationship matrix can then be replaced by a realized , genome-based relationship matrix ( often called the “genomic” relationship matrix , [8] ) . This approach is equivalent to a random regression approach in which all SNP genotypes are simultaneously accounted for as explanatory variables in a multiple regression model [9] . In animal and plant breeding , selection based on genome-based predictions of genetic values is expected to massively increase genetic progress [4] , [10] and has quickly found its way into widespread practical application ( see [4] , [5] for reviews ) . Genome based-prediction follows a different paradigm than genome wide association studies ( GWAS ) . GWAS identify single molecular variants associated with phenotypic variability using individual statistical tests for significance of each variant . Genome-based prediction uses the entire genomic variability captured by the available marker set to explain the observed phenotypic variation , and does not rely on selection of single loci based on significance tests . Standard prediction methods are thought to work for traits with a highly polygenic or even infinitesimal [11] genetic architecture , where the effect of a single variant is too small to be captured by a statistical test in a GWAS . There is strong empirical evidence that many quantitative traits have such a highly polygenic genetic architecture in farm animals [12] , agriculturally used plants [13] , model organisms and humans [14] , [15] . With the advent of next generation sequencing technologies , it is now feasible to implement genomic prediction based on complete genome sequences of higher organisms . While these techniques have only been applied to individuals or cohorts of limited size [16] to date , initiatives to sequence larger panels are under way [17] , [18] , and genotyping by whole genome resequencing will become a standard technology in the foreseeable future . The accuracy of prediction methods based on marker data depends on the heritability of the trait , its genetic architecture ( number of loci affecting trait variation , mode of inheritance , and distribution of allelic effects , [19] ) , the LD reflecting effective population size , the size of the genome , the marker density and the sample size used in the statistical analysis [20] . Various methods of prediction incorporating genomic information have been studied on real and simulated data , including Genomic Best Linear Unbiased Prediction ( GBLUP ) approaches with genomic relationship matrices [8] , Random Regression BLUP ( RRBLUP ) , Bayesian linear regression methods [10] , [21] or fully non-parametric approaches [22]–[25] . GBLUP approaches are based on a linear model for the phenotypic values , which encompasses a vector of random genetic values of individuals whose covariance structure is inferred from genomic data . The linear model underlying the RRBLUP approach includes a vector of random marker effects ( instead of a vector of genetic values ) which are assumed to be drawn from the same normal distribution and uncorrelated . The model primarily provides estimates of SNP effects , but estimated genetic values of individuals can be derived as linear combinations of the estimated SNP effects , yielding the same predictions of individual genotypic or phenotypic values as GBLUP . The BayesB method [10] , on the other hand , fits only a small fraction of the available markers to conform with the assumption that most loci are expected to have zero effect on the phenotype , and the remaining non-zero marker effects are drawn from normal distributions with random variances . It has been suggested [26] that differences between prediction methods will become more pronounced with the availability of full genome sequence data . According to a study with simulated data [26] , RRBLUP and equivalent GBLUP procedures do not take full advantage of high-density marker data if the number of causal SNPs is small , while approaches with an implicit feature selection such as BayesB might be more accurate . If , on the other hand , the number of causal loci is large , RRBLUP or GBLUP methods may yield accurate predictions because the assumption that every SNP has an effect is closer to reality . Implementing genomic prediction with full genome sequence data raises a number of questions . What is the most efficient way to incorporate the complete genomic information in prediction ? How much predictive ability is gained by using whole genome sequence data compared to high density SNP panels ? Is it possible to increase predictive ability by a pre-selection of SNPs or models with an internal feature selection ? How comparable are the results of genomic prediction and genome wide association ? Here , we address these questions empirically based on full genomic sequences of a population of Drosophila melanogaster inbred lines . The inbred lines have been sequenced , and constitute the Drosophila Genetics Reference Panel ( DGRP ) , a new community resource for genetic studies of complex traits [27] . We report the results of a full sequence based genomic prediction for two quantitative traits , starvation stress resistance and locomotor startle response , both of which display considerable genetic variation in natural populations and respond rapidly to artificial selection [28]–[30] . We used whole-genome sequences determined on the Illumina platform for DGRP-lines for starvation resistance ( startle response ) [27] . Our reference method is a GBLUP approach in which ∼2 . 5 million polymorphic SNPs are used to derive a genomic relationship matrix [8] . We evaluated predictive ability via cross-validation ( CV ) , and compared prediction within vs . across sexes , various SNP densities , and training set sizes . We assessed whether BayesB is superior over GBLUP given full genome sequence data [26] , and compared our genomic prediction results with those of GWAS conducted on the same DGRP lines [27] . To our knowledge , this is the first application of genomic prediction on empirical whole genome sequence in a substantial sample of a higher organism . However , this study , as well as all previous association studies , only assesses the effects of common SNPs , since the effects of rare alleles cannot be estimated due to the small sample of sequenced lines . The results illustrate both the potential of the approach and challenges to be addressed in the future . We constructed a genomic relationship matrix [8] from ∼2 . 5 million SNPs for which the minor allele was present in at least four of the DGRP lines [27] . A histogram of the off-diagonal elements of this matrix for DGRP lines used in the GBLUP analyses ( Figure 1 ) and a corresponding heatmap ( Figure 2 ) show that there were no large blocks of high genomic relationship among the lines . The average genomic relationship is close to zero , as expected , but there is considerable variance around this average ( Figure 1 ) , as indicated by two block of lines with average genomic relationships within each block of and ( Figure 2 ) . We performed genomic prediction for starvation stress resistance and locomotor startle response . The phenotypes used were the medians of many ( ) individually tested males and females for each line , or the average of the male and female medians ( Table S1 ) . We used several cross-validation ( CV ) procedures for each trait ( Table 1 ) . In the -fold CV , predictive ability was for starvation resistance and for startle response . In human studies the efficiency of a predictor is reported as the squared correlation rather than [31] , so that in terms of variance explained the estimates were for starvation resistance and for startle response . The observed accuracy depends on the size of the training set ( Figure 3 ) , with decreasing accuracies obtained with smaller training sets . Predictive abilities are roughly halved for both traits when using only instead of of the data to train the model . Maximum likelihood estimates of narrow-sense heritabilities based on the GBLUP model using the genomic relationship matrix were in all analyses ( Table S2 ) , reflecting the fact that phenotypes are averages over many replicates and thus residual variance is minimal . Hence , the phenotypes used represent the line genotypes with maximum accuracy , which is the ideal case for training the genomic model . Using male performance data to train the model and using the results to predict the female performance ( or vice versa ) does not affect the predictive ability for startle response , but substantially reduces the predictive ability for starvation resistance , reflecting a higher degree of genotype by sex interaction in this trait ( [27] , and see below ) . Prediction is more accurate in females than in males ( vs . ) for starvation resistance , while there is little difference for startle response . A series of -fold CVs for starvation resistance using different SNP densities showed that predictive ability remained almost constant if every SNP ( ∼150 , 000 SNPs ) was used to construct the genomic relationship matrix ( Figure 4 ) . The predictive ability began to deteriorate when fewer than SNPs were used , but only vanished completely when as few as ∼2 , 500 SNPs ( every SNP ) were used . The corresponding LD distribution for SNP neighbors for different SNP densities is shown in Figure 5 , illustrating the extreme short-range extent of LD in the D . melanogaster genome . The average LD between SNPs ( after imputation ) whose distance lay in the interval bp was for the autosomes and for the X-chromosome . Long-range LD between pairs of loci at the opposite ends of chromosome arms or across different chromosome arms was on average both for the autosomes and the X-chromosome . For starvation resistance , the influence of the minor allele frequency of the SNPs used on the predictive ability was assessed with a series of 5-fold CVs using SNP sets with different average minor allele frequency . We find that the variability of the predictive ability increases when the average minor allele frequency of the SNPs used to construct the genomic relationship matrix is decreased ( Figure S1 ) . In replicates of an additional 5-fold CV , in which we randomly chose SNPs to build the genomic relationship matrix , an average predictive ability of was obtained , which is in the range obtained when every SNP ( ∼77 , 817 SNPs ) was used ( , Figure 4 ) . Running replicates of a 5-fold CV using randomly chosen blocks of adjacent SNPs ( each block consisting of SNPs ) led to an average predictive ability of . To analyze whether the predictive ability is due to lines which are more highly related , we ran an additional 5-fold CV with replicates in which the two groups of higher overall relatedness ( Figure 2 ) were excluded . Here we found an average predictive ability of for starvation resistance , which is larger than the average predictive ability we obtained using all lines ( ) . For startle response , excluding the two groups led to a decrease in predictive ability ( in comparison to ) . The accuracy of genomic prediction is a function of a number of quantities , including the size of the training set and the effective population size [20] . has an effect on the number of independently segregating chromosome segments , , in a population ( the larger , the larger ) ; and the predictive ability of GBLUP is higher when the number of segments is small . By varying the size of the training set in a series of CVs , we can estimate by fitting a curve through the empirical accuracies obtained ( Figure 3 ) . We estimated for starvation resistance and for startle response . The coefficient of determination of the fitted curve was for starvation resistance ( startle response ) . The bias corrected empirical confidence intervals for the estimates obtained with bootstrapping [32] were for starvation resistance and for startle response . The effective population size in the Raleigh population ( from which the DGRP-lines were drawn ) was estimated to be ∼19 , 000 in , with a massive fluctuation between years [33] . Our estimates of correspond to independently segregating chromosome segments . In this formula is the length of the female genome in Morgans ( there is no recombination in male Drosophila ) . Since the sequenced animals resulted from generations of full sib mating following the original sampling from the Raleigh population , the DGRP lines are not expected to have the same as the original population and are consequently expected to have a different . We can use the curves fitted through the empirical accuracies ( Figure 3 ) , to predict the expected accuracy of prediction for an arbitrarily large size of the training set: If lines were available in the training set , the curve would predict accuracies of ∼0 . 58 for starvation resistance and startle response . This value was obtained by using and as well as and in the modified formula of [20] . We also estimated the effective population size based on LD directly . For a distance bin of Morgan we obtained average LD-values of for chromosome 2L ( 2R , 3L , 3R , X ) . These values correspond to an estimated effective population size of , approximately generations ago . The average estimated effective population size is , which is in the range of the estimates based on the observed accuracies . Genomic prediction might be improved if we only fit SNPs which are associated with variance in a trait , because we then concentrate on the biologically relevant genomic regions , and excluding SNPs which are not associated with the trait reduces statistical noise . We tested this hypothesis using the starvation resistance data . We identified the SNPs with the highest absolute estimated effect or the highest estimated genetic variance , respectively , in the training set of the respective of the folds in a -fold CV . We then used these subsets of selected SNPs to predict the phenotype in the remaining of the fold . Predictive ability was improved by over the reference scenario when using the SNPs with largest effects ( average predictive ability of in comparison to ) . Using the SNPs with greatest variance explained , predictive ability was improved by ( average predictive ability of ) . In both cases , the improvement is marginal and provides little support for the idea of SNP pre-selection . We also compared our GBLUP results to those from a method which does not assume that all SNP effects are drawn from the same normal distribution and carries out an internal feature selection . We ran replicates of a -fold CV for starvation resistance using BayesB [10] . In each round of the Markov Chain Monte Carlo based procedure ( see Methods ) , of the SNPs were assumed to have no effect and the effects of the remaining of the SNPs were drawn from normal distribution with random variances . In most folds of each single CV and for all replicates of CV , the observed predictive abilities differed only marginally between BayesB and GBLUP ( Figure 6 ) . The average predictive ability obtained with BayesB was which is not appreciably different from the result obtained with GBLUP ( ) . Although genomic prediction follows a different paradigm than genome-wide association studies , it is informative to compare significant SNP positions from the GWAS to areas of large estimated SNP effects resulting from the GBLUP model . Previously [27] , a GWAS of DGRP lines ( of which the material used here is a subset ) identified SNPs associated with starvation resistance and SNPs associated with startle response at a nominal p-value in the analyses of sex-averaged data . We estimated SNP effects using RRBLUP and compared them to the significant SNPs from the GWAS study ( Figure S2 , Figure S3 ) . There is excellent concordance of signals from both approaches in some regions ( e . g . the genome-wide largest SNP effects on chromosome 3L for starvation resistance and 2L for startle response ) , while concordance is poor in other regions , especially on the X chromosome . We further investigated whether the most significant SNPs detected in the GWAS are reflected by large SNP effects in the GBLUP study using a different approach . For each significant SNP position from the GWAS we took the neighboring SNPs ( on each side ) and calculated the sum of the absolute values of their estimated effects using the GBLUP model . To avoid an effect of different sample size , we used the most significant loci from the GWAS for both traits . We compared these sums to the sums of the absolute values of estimated SNP effects in sliding windows spanning the whole genome ( with each window containing neighboring SNPs ) . We observed a clear separation of the density functions of these sums for both startle response and starvation resistance ( Figure 7 ) . The density resulting from the sliding window approach reflects the overall distribution of the suggested statistic in the sample . For starvation resistance ( startle response ) a threshold value of , cf . Figure 7 , cuts off the upper of the respective distribution . Applying the same threshold with the density function reflecting the statistic for the significant GWAS positions , of the distribution exceeds the threshold , indicating that signals found in the GWAS are also associated with large estimates of the SNP effects in the genomic model . In addition to the line means we also analyzed individual records ( individual flies per line tested for starvation resistance and for startle response ) to assess whether the variance between lines can be fully explained by additive gene effects or if non-additive mechanisms have an impact . This was done by modeling the covariance structure between lines based on the additive and additiveadditive genomic relationship matrix and testing the goodness of fit of the respective models . Most applications of genomic prediction are for outbred populations , for which the additive genetic variance and corresponding narrow-sense heritability determine the extent to which phenotypes in the next generation can be predicted from information obtained on the current generation . However , the variance among DGRP lines is the total genetic variance , and is possibly inflated by additive by additive epistatic variance [34] . Therefore , we performed several analyses on measurements of individual flies to determine the nature of the total genetic variance , especially to what extent the presence of non-additive genetic variance might have affected predictive abilities . We fitted three different models to the individual phenotype data: Model 1 contained a random line effect , and lines were assumed to be unrelated . In Model 2 , a random additive line effect was added , whose covariance structure was modeled via the genomic relationship matrix . In Model 3 , an additional random additiveadditive epistatic effect was included , whose covariance structure was modeled via the Hadamard product . Since the between line variance relates to inbred lines , while the additive and additiveadditive variance component pertain to the non-inbred base population ( or a hypothetical random mating produced from the inbred lines ) , the variance between inbred lines in Model 1 is expected to be twice the additive genetic variance in Model 2 or 3 under a fully additive model . We estimated variance components for all three models pooled across sexes and separately for males and females ( Table S3 , Table S4 ) . We find little evidence for non-additive genetic variance for these traits . The estimate of from Model 2 is from Model 1 , and Model 2 gave a significantly better fit than Model 1 when applying the likelihood ratio test , again indicating that the observed between line variance is due to additive gene action . Inclusion of the component was not significant for either of the traits . We found significant sex by line interaction variance for starvation resistance , but not for startle response ( Tables S3 , S4 ) , which is in accordance with the findings of the genomic prediction across sexes ( Table 1 ) and previous analyses of these data [27] . We report the first ( to our knowledge ) application of genomic prediction to a real set of full genomic sequencing data in a eukaryotic organism . Although predictive abilities obtained with starvation resistance and startle behavior are only moderate to low , and although we limited our analysis to SNPs that are common due to the small sample size of lines , this study can be seen as a proof of concept for this approach . There are several reasons for the limited predictive ability obtained in this study . First , the training set is small , with a maximum of observations in the -fold CV , and the accuracy of genomic prediction is a function of the size of the training set [20] . Using the curves fitted through the empirical accuracies ( Figure 3 ) , we predict accuracies of for starvation resistance and startle response , if sequenced lines were available for the training set . The second important factor affecting accuracy of prediction is the number of independently segregating chromosome segments , [20] . In our study we obtained . This is larger than usually observed for Holstein cattle ( with and genome length Morgans [35] ) , but is smaller than the corresponding value in the human genome ( with Morgans , [36] ) . ( Note that in mammalian species , there is recombination in both sexes and [9] . ) Accuracy of genomic prediction is thought to come from two sources: ( i ) SNPs in useful LD with causal loci; and ( ii ) SNPs reflecting the relationship structure between the training set and the set to be predicted [37] . Due to the very fast decay of LD in the D . melanogaster genome , few SNPs are in useful LD with any causal polymorphism . Even if we define “useful LD” very conservatively as , then on average only a region of bp around a causal polymorphism was in useful LD on an autosome ( bp on the X chromosome ) . This means that on average ( ) SNPs were in useful LD with a causal autosomal ( X-linked ) polymorphism , as the average distance between neighboring SNPs was bp ( bp ) on an autosome ( X chromosome ) . If predictive ability was mainly driven by SNPs in LD with causal polymorphisms , reducing the SNP density should lead to a massive decay of predictive ability of the models , which was not observed . Little decrease in accuracy was seen , even if every SNP was used in the model , in which case hardly any SNP would be in useful LD with causal polymorphisms . The underlying mechanism therefore seems to depend on a sufficient number of SNPs being in low LD with causal polymorphisms , rather than few SNPs in close physical association and high LD . In the DGRP population , LD approaches a small but positive baseline level with increasing physical distance [27] , so that even with large physical distances a minimum level of LD is maintained , which was on average with being the sample size . The number of SNPs for maximal accuracy of genomic prediction with unrelated individuals has been estimated as [38] , corresponding to SNPs in the present study . For starvation resistance , we find that the empirical accuracy levels off when approximately every SNP is used , which is equivalent to or SNPs . Adding more SNPs beyond this value does not lead to any improvement in the genomic prediction of starvation resistance , but also does not reduce accuracy , which one might expect when using more SNPs than actually needed . While fitting large numbers of “superfluous” SNPs may be considered as noise in the RRBLUP model , these SNPs can also be seen to provide a better basis to estimate the realized relationship matrix in the GBLUP model , which leads to a higher accuracy of the estimated realized relationships . Since both models are fully equivalent [9] no penalty is expected in the prediction of genomic values . Since pedigree information for the founders of the inbred lines was not available , our estimates of heritability and genomic prediction are based on the actual degree of identity-by-descent sharing between relatives [39] . There is little pedigree structure in the DGRP lines , with the exception of two distinct blocks of higher relatedness , comprising and lines , respectively , with a genomic relationship within blocks of and . When these blocks were excluded from the data , predictive accuracy in a -fold CV increased ( decreased ) for starvation resistance ( startle response ) , suggesting that prediction in the DGRP population does not rely on distinct family structures . Given this together with the short-range extent of LD in the D . melanogaster genome and the robustness of the accuracy of genomic prediction with reduced marker density , we conclude that the observed accuracy of prediction for starvation resistance and startle response is primarily due to the long-range LD in the population , or equivalently , the subtle relationship structure as reflected by the genomic relationship matrix . We restricted our analyses to SNPs for which the minor allele was present in at least four DGRP lines ( a minor allele frequency of ) . We applied this threshold to avoid computational limitations , especially when applying the BayesB method; and for consistency with the GWAS in the DGRP [27] , which used the same filtering criterion . Thus , we did not utilize the million SNPs with minor allele frequencies less than this , nor did we take other forms of molecular variation into account . Structural variations such as transposable elements have been repeatedly reported to be associated with phenotypic variation [40] , therefore we must consider to what extent not including these variants in the models affected prediction accuracy . Given that we do not observe an increase in accuracy when increasing the number of SNPs from to million , we do not expect that increasing the marker density by adding more SNPs and other variants will have a significant effect on predictive ability . Additionally , SNPs with low minor allele frequencies were shown to be highly variable in predictive ability , so that the potential amount of information possibly added by the million low frequency SNPs is limited . However , accounting for all polymorphisms in the model means that some fraction of the genetic variants must causally affect the trait . Simulations [26] including the causal polymorphism in the model improves the predictive ability over models based only on neutral SNPs in LD with the causal variants . Further research is needed to understand these mechanisms in the context of genomic prediction based on empirical data . The accuracy of BayesB has outperformed that of GBLUP in several simulation studies [10] , [37] . Simulation results have suggested that GBLUP did not take full advantage of genome sequence data , suggesting that Bayesian methods are needed to obtain maximum accuracy [26] . The superiority of BayesB over GBLUP is expected to increase with marker density , and decrease when the size of the training data set is increased [38] . However , we did not find that BayesB yielded a significantly higher predictive ability than GBLUP in the replicates of -fold CV with starvation resistance implemented in the present study . We used a very high marker density and a small training set , and yet GBLUP performed as well as BayesB . These conclusions should be taken with caution , since the available size of the training set was extremely small in our study due to the limited availability of fully sequenced lines . In [20] , BayesB yielded a higher accuracy than GBLUP , when the number of simulated QTL was low; but GBLUP slightly outperformed BayesB , when the number of QTL became large , since the GBLUP model is equivalent to RRBLUP , in which all SNPs are assumed to have an effect drawn from the same normal distribution . Although this model may not seem biologically plausible , it performed as well as BayesB in the present study , consistent with several studies on real data from dairy cattle for different traits [4] , [41] . The finding that BayesB did not outperform GBLUP in the present study is consistent with a quasi-infinitesimal genetic architecture; and results indicate that starvation resistance and startle response are complex traits with a highly polygenic genetic architecture rather than being driven by a few major causal genes . This is in agreement with previous studies stating that starvation resistance and startle response can be considered to be model traits with a complex ( i . e . quasi-infinitesimal ) genetic background [28]–[30]; and it is also in line with the results from the GWAS [27] . One reasonable conclusion might be that there are so many causal polymorphisms , each with a small effect , that the effective chromosome segments are saturated with causal variants and the effects of segments follow a normal distribution . Under this circumstance , GBLUP is expected to perform as well as BayesB . However , these hypotheses clearly need further investigation . More systematic model comparisons based on the available data were not considered here due to the prohibitive computing time required for BayesB . Previously , gene centered multiple regression and partial least square ( PLS ) regression models were used to predict starvation resistance and startle response phenotypes from genotypic data [27] . In both cases only SNPs that had nominal significance levels of from the GWAS were used . The gene centered prediction models found that a few SNPs explained a large fraction of the genetic and phenotypic variance of the traits , while the PLS models found that the significant SNPs explained a high fraction of the phenotypic variance . The purpose of these studies was a comparison with human association studies , in which the faction of the variance explained by significant variants in the entire sample is commonly quoted . These approaches are fundamentally different from the BLUP approach used in this study . The BLUP approach includes random components and their covariance structure in the model , whereas regression models do not incorporate random terms except from the residuals; and the BLUP approach does not rely on a pre-selection of SNPs based on a GWAS . Most critically , we evaluated the robustness of the BLUP predictions using -fold cross-validation; whereas the previous analyses only tested the explanatory power of the most significant associated SNPs using the entire sample . Had we done the same analysis using GBLUP , we would be able to predict of the variance . The imperfect concordance of the positions of the most significant SNPs from the GWAS and the largest estimates of SNP effects from RRBLUP is a consequence of the different objectives of the two approaches . A sequence-based GWAS is conducted to identify causal polymorphisms and provide estimates of allelic effects and frequencies . Also , the GWAS suffers from estimating one effect at a time and so does not necessarily position the QTL accurately . The goal of RRBLUP is to predict the phenotype using all available SNP information simultaneously . Here , estimated SNP effects are a by-product and mapping causal variants is not the primary objective . Given that the number of SNP effects to estimate is much larger than the number of observations , effects are estimated using penalized multiple regression approaches , shrinking estimated effect sizes towards zero . In addition , the magnitude of estimated SNP effects from RRBLUP is a function of the marker density . The higher the marker density , the more SNPs will be in LD with a causal mutation; therefore , the true allele substitution effect of a causal polymorphism will be split up and assigned in parts to a series of SNPs in the respective haplotype block . This can mask both the effect size , because one large effect may come in many small pieces; and the mapping position , because any SNP in LD with the causal polymorphism may have a substantial estimated effect . Nevertheless , some of the largest SNP effects from RRBLUP are in the proximity of prominent SNPs identified in the GWAS , so that to some extent positional information can still be retrieved from the RRBLUP results . A methodology combining the strengths of both approaches – unbiased effect estimates and high positional resolution of GWAS with the simultaneous analysis of all SNPs , high predictive power and quality control via CV of genomic approaches – still needs to be developed . Results obtained in our study cannot be directly compared to predictive abilities in human studies due to the extremely small training set size ( in CV ) , and Drosophila has much larger and rapid decline of LD compared to humans . When genomic prediction in human studies was based on large training sets ( thousands ) , substantial SNP panels ( k ) and a highly heritable trait ( ) , predictive ability of genomic models was found to exceed what has been previously reported using a reduced number of markers pre-selected based on GWAS [31] and genomic prediction based on pre-selected SNPs was found to be of limited use in human studies of height [42] . In the near future individual whole genome sequences will become increasingly available for large numbers of individuals in many species [17] , [18] . Sequence-based predictions will therefore be relevant for prediction of risk disease and individualized medicine in humans , and for genome-based selection in farm animals and crops . The main findings of our study are: ( i ) genomic prediction can be efficiently implemented via GBLUP with full genome sequence data; ( ii ) there is little , if any , gain in predictive ability if the number of SNPs is increased above ( equivalent to in Holstein cattle and in humans ) ; and ( iii ) approaches based on external or internal ( BayesB ) selection of subsets of SNPs were not found to provide a substantial gain in accuracy of prediction compared to GBLUP . All findings must be seen against the background of the small sample size and the specific genetic constellation , with almost unrelated inbred lines and highly accurate phenotypes . Nevertheless , these results provide a realistic assessment of the potential benefits of sequenced-based prediction applied to non-model organisms and indicate avenues for future research . The full Drosophila Genetic Reference Panel ( DGRP ) [27] , a recently developed new community resource for genetic studies of complex traits , consists of D . melanogaster lines derived by generations of full sib mating from wild-caught females from the Raleigh , North Carolina population . Whole genome sequence data of DGRP lines ( Freeze 1 . 0 ) have been obtained using a combination of Illumina and next generation sequencing technology , which are available from the Baylor College of Medicine , http://www . hgsc . bcm . tmc . edu/project-species-i-DGRP_lines . hgsc . We used the Illumina sequences for DGRP lines in this study . SNPs were called from the raw sequence data as described previously [27] . We used SNPs with a coverage greater than 2X but less than 30X , for which the minor allele frequency was present in at least four lines , and for which SNPs were called in at least 60 lines . This series of filters gave a total of SNPs for this analysis; on 2L , on 2R , on 3L , on 3R and on the X chromosome . We did not consider the few SNPs on the very short chromosome 4 . In total there were missing SNP genotypes ( ) , which we imputed using Beagle Version 3 . 3 . 1 software [43] . Phenotypic measurements for starvation resistance were available for all DGRP lines , and for startle response on lines [27] . We used the average of the medians of measurements for each trait in males and females as the phenotypic value of the line , i . e . , where and are the medians of the measurements for female and male individuals of the line . We used medians because of the skewed distribution of traits; however , medians are highly correlated with line means . For starvation resistance ( startle response ) there were on average measurements for females , and measurements for males ( Table S1 ) . Measurements were taken in several replicates for each trait [27] . We used different cross-validation ( CV ) procedures [44]–[46] to assess the predictive ability of different methods . In one replicate of a CV , the lines are randomly divided into a training set , which is used for parameter estimation; and a validation set , for which genetic values are predicted . The CV procedures differ in the ratios of the numbers of lines belonging to the training and validation sets: In a -CV ( with integers and ) , the lines are randomly divided into groups . The groups build the training set , and the remaining groups build the validation set . For this classification , there are possibilities . For each of these possibilities ( “folds” ) , total genetic values for the lines of the validation set are predicted and the corresponding predictive ability is calculated . The predictive abilities are then averaged to obtain one average correlation per CV replicate . For example , one ( 3∶2 ) -CV , consists of CV folds , over which predictive abilities are averaged . A -CV is also called -fold CV . We used ( 4∶1 ) - , ( 3∶2 ) - , ( 2∶3 ) - and ( 1∶4 ) -CVs to analyze the effect of decreasing training set size . The CVs also differed in the constellations of phenotypic records used for the training and validation set . For example , the notation “ ( 4∶1 ) male – female” indicates that only the medians of male records were used in the training set , and that the predicted genetic values were correlated with the medians of female records of the validation set to obtain the predictive ability in a ( 4∶1 ) -CV . CVs were also run for different marker densities , using every -th SNP ( ) . Additionally , 5-fold CVs using only the SNPs with the largest absolute values of estimated effects ( obtained in the training set ) , or using only the SNPs with the largest SNP variances ( obtained in the training set ) were performed . The additive genetic variance marked by the SNP was calculated as with allele frequency and estimated SNP effect . In another series of 5-fold CVs we randomly chose SNPs to build the genomic relationship matrix or we randomly chose blocks of adjacent SNPs ( each block consisting of SNPs ) . In an additional 5-fold CV we excluded the lines in the two blocks of higher relatedness ( Figure 2 ) from the data . Each type of CV was replicated times , resulting in average predictive abilities . We also analyzed the influence of minor allele frequency on the predictive ability by another series of 5-fold CV . For this , we sorted all SNPs by their minor allele frequency and divided the sorted vector into blocks . For each block we ran replicates of a 5-fold CV using GBLUP and the corresponding SNPs . Predictive ability was measured in terms of correlation between predicted genetic values and observed phenotypic values . The corresponding accuracy , defined as the correlation between true and predicted genetic value , was obtained by dividing the observed predictive ability by the square root of the observed heritability [47] . The heritability was based on the GBLUP model ( see below ) . The underlying statistical model is ( 1 ) In this model , the component of the -vector is the phenotypic value of the line that is used for prediction , i . e . the average of the medians of the phenotypic measurements for males and females for this line . Moreover , is the overall mean; is assumed to be multivariate normal , with the genomic relationship matrix of all lines [8] and the additive genetic variance among lines . The matrix is an -incidence matrix , whose rows consist of unit vectors with one component being and all the others zero , indicating the respective positions of lines used for prediction in the -vector of genetic values of all lines . is the residual term , where is the residual variance . Following the approach of [8] , was defined aswhere is the -matrix of SNP genotype vectors for the lines with the SNPs coded as and the column of is , where is the frequency of the second allele at locus . Variance components were estimated via maximum likelihood ( ML ) using the R-package “RandomFields” , Version 2 . 0 . 46 ( http://CRAN . R-project . org/package=RandomFields ) , and its function “fitvario” . The BLUP approach to obtain the vector of genetic values is equivalent to solving the so-called Mixed Model Equations ( MME ) :A narrow-sense heritability based on the GBLUP model ( 1 ) was calculated as The GBLUP model ( 1 ) is equivalent to the following linear model in which all SNPs are assumed to have an effect drawn from the same normal distribution [9]:where and are as described above and is the vector of SNP effects with . Using this equivalence , the SNP effects can be predicted asTo estimate the SNP effects resulting from GBLUP for a single trait , we used all of the available lines , i . e . in model ( 1 ) contained the phenotypic values of all lines so that in the corresponding formulas . Note that only the inversion of a matrix of size equal to the number of sequenced lines is required . We used [48] as a measure of LD between a pair of loci . With two biallelic loci and with alleles and and frequencies and , we denote the frequencies of the genotypes and as and respectively . Then , We performed the LD analyses using the imputed SNP matrix of million SNPs for the lines . We calculated the distribution of LD between all pairs of neighboring SNPs for different marker densities , using every -th SNP ( ) . The extent of long-range LD was calculated for pairs of SNPs randomly sampled from the first and the last SNPs per chromosome arm . Moreover , the average LD was calculated between SNPs on different chromosome arms , by sampling pairs of SNPs for each combination of chromosome arms . We modified the formula [20] for the expected accuracy , of GBLUP given different population parameters ( see Text S1 for more details on the derivation in the case of D . melanogaster ) : ( 2 ) is the effective population size , is the size of the training set , is the length of the female genome in Morgans and is the narrow-sense heritability of the trait estimated from model ( 1 ) . The term describes the number of independently segregating genome segments [9] . We ran CVs with different numbers of lines ( for starvation resistance and for startle response ) in the training set ( replicates each ) . Average numbers of lines in the training set are reported , which are non-integer values for starvation resistance because in a -CV , division of lines into groups may give unequal numbers of lines in the different partitions . Given the corresponding average accuracies for the CV replicates , we estimated by fitting a curve to the points . To fit the curve , we chose such that the sum of the squared differences of the observed accuracies and the accuracies obtained by ( 2 ) was minimized:using and Morgan . We calculated the length of the female genome in Morgans by summing the lengths of the chromosomes in base-pairs ( ( , , , ) Mbp for chromosome 2L ( 2R , 3L , 3R , X ) , [49] ) and multiplying by the average recombination rates of females for the different chromosomes in Morgans per base-pair [50] . After performing bootstrapping ( replicates ) , the bias corrected empirical confidence intervals ( error in each tail ) for the estimates [32] , [51] were calculated aswhere is the -percentile of the bootstrap cumulative distribution function , is the -percentile of the standard normal distribution function , and . To estimate the effective population size based on LD , the following formula was used [52]:where is the number of lines and is the recombination rate in female individuals , cf . Text S1 for more details on this formula . The underlying model for the Markov Chain Monte Carlo based BayesB [10] method iswhere and are as defined previously and is the vector of normally distributed and independent SNP effects . The variance of the SNP effect , , is assigned an informative prior . The prior distribution of the genetic variances aims to resemble a situation where there are many loci with zero variance and only some loci with variance not equal to zero . Therefore , the prior distribution of the variance of a marker effect is a mixture of distributions which is given byNote that this implies that the unconditional distribution of each single marker effect is a mixture of a point mass at ( with probability ) and of a t-distribution with zero mean , degrees of freedom and scale parameter [21] , i . e . BayesB assigns the same unconditional prior distribution to each marker effect . In our studies , we used and the scale parameter was calibrated asWe chose , such that approximately markers were contributing to the additive genetic variance . For the residual variance , , the prior distribution was with andValues for and were chosen in the order of magnitude of the variance components of the GBLUP model ( 1 ) , which were estimated using all lines and “fitvario” . The BayesB procedure is described in detail in [10] . It consists of running a Gibbs chain , where additionally a Metropolis-Hastings algorithm ( iterations ) is used to sample from , where denotes the data corrected for the mean and all genetic effects other than the marker effect . Following graphical inspection , we ran BayesB with a chain length of iterations including a burn in of iterations that were discarded . To perform the BayesB approach , we used the software “GenSel” , Version 2 . 36 , by R . Fernando and D . Garrick ( cf . http://taurus . ansci . iastate . edu/Site/Welcome . html ) , which is implemented in C++ . BayesB is computationally very intensive . The analyses were run on a Mac Pro 2× 2 . 93 GHz 6-Core Intel Xeon with 64 GB RAM running Mac OS X Server 10 . 6 . 7 . One fold of a -fold CV for starvation resistance took approximately hours . A genome-wide association study ( GWAS ) revealed significant SNP positions for starvation resistance ( startle response ) [27] , where a SNP position was considered significant if at least one of the three p-values , obtained using only male , only female or sex-pooled phenotypic records , was . We considered the subset of SNPs for which p-values of SNP effects of pooled data were , to be more conservative and to be consistent with the previous analyses , leading to significant SNPs for starvation resistance ( startle response ) . We compared genomic regions for which GBLUP estimated large SNP effects to these significant SNP positions of the GWAS . To avoid an effect of different sample sizes , we chose the most significant SNPs from the GWAS analysis for each trait . For each of these SNPs , we chose the closest ( neighboring ) SNPs ( on each side ) and calculated the sums of absolute values of the corresponding SNP effects ( resulting from the GBLUP model ) . We compared the distribution of these sums to the distribution of the sums of the absolute values of estimated SNP effects in windows of neighboring SNPs covering the whole genome by plotting the corresponding density functions . To obtain the sums of the absolute values of estimated SNP effects covering the whole genome , the windows were overlapping , displaced by SNP positions . If the genomic regions for which GBLUP estimated large SNP effects coincide with the significant SNP positions of the GWAS , we expect the density functions to be separated . For each trait , we fitted three different models using individual trait records . The first model included a fixed sex effect , a random line effect , a random line-sex-interaction term and a random term accounting for the different replicates in which measurements of the traits were taken:In the second model , an additional random genetic effect was added for each line . The variance-covariance matrix of the vector of these genetic effects was assumed to be given by the genomic relationship matrix of [8]:In the third model , an additional random additiveadditive epistatic effect was included for each line . The variance-covariance matrix of the vector of these genetic effects was given by the Hadamard product [53] of the genomic relationship matrix of [8]:Other two-way epistatic interactions , like additivedominance or dominancedominance , should not exist in inbred lines , provided inbreeding is complete . Variance components and their standard errors were estimated using ASReml 2 . 0 [54] . The analyses were done pooled across sexes as well as separately for males and females . The analyses of separate sexes did not include the sex term , and the replicate ( sexline ) term was reduced to replicate ( line ) . The broad-sense heritability for Model 1 was calculated ascf . [28] . Narrow sense heritabilities for Models 2 and 3 were calculated asandThese heritabilities are based on individual trait records . Unless stated otherwise , all statistical analyses were performed using R software [55] . The R-package “ff” , Version 2 . 2-1 ( http://CRAN . R-project . org/package=ff ) , was used to handle the large amount of SNP data efficiently in terms of memory capacity .	The ability to accurately predict values of complex phenotypes from genotype data will revolutionize plant and animal breeding , personalized medicine , and evolutionary biology . To date , genomic prediction has utilized high-density single-nucleotide polymorphism ( SNP ) genotyping arrays , but the availability of sequence data opens new frontiers for genomic prediction methods . This article is the first application of genomic phenotype prediction using whole-genome sequence data in a substantial sample of a higher eukaryote . We use ∼2 . 5 million SNPs with minor allele frequency greater than 2 . 5% derived from genomic sequences of the “Drosophila Genetic Reference Panel” to predict phenotypes for two traits , starvation resistance and startle-induced locomotor behavior . We systematically address prediction within versus across sexes , genomic best linear unbiased prediction ( GBLUP ) versus a Bayesian approach , and the effect of SNP density . We find that ( i ) genomic prediction can be efficiently implemented using sequence data via GBLUP , ( ii ) there is little gain in predictive ability if the number of SNPs is increased above 150 , 000 , and ( iii ) neither implicit nor explicit marker selection substantially improves the predictive ability . Although the findings must be seen against the background of small sample sizes , the results illustrate both the potential of the approach and the challenges ahead .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "biotechnology", "genome-wide", "association", "studies", "animal", "genetics", "statistics", "applied", "mathematics", "population", "genetics", "animal", "models", "genome", "sequencing", "algorithms", "drosophila", "melanogaster", "model", "organisms", "mathematics", "ef...	2012	Using Whole-Genome Sequence Data to Predict Quantitative Trait Phenotypes in Drosophila melanogaster
Kaposi’s sarcoma-associated herpesvirus ( KSHV ) is the infectious cause of the highly vascularized tumor Kaposi’s sarcoma ( KS ) , which is characterized by proliferating spindle cells of endothelial origin , extensive neo-angiogenesis and inflammatory infiltrates . The KSHV K15 protein contributes to the angiogenic and invasive properties of KSHV-infected endothelial cells . Here , we asked whether K15 could also play a role in KSHV lytic replication . Deletion of the K15 gene from the viral genome or its depletion by siRNA lead to reduced virus reactivation , as evidenced by the decreased expression levels of KSHV lytic proteins RTA , K-bZIP , ORF 45 and K8 . 1 as well as reduced release of infectious virus . Similar results were found for a K1 deletion virus . Deleting either K15 or K1 from the viral genome also compromised the ability of KSHV to activate PLCγ1 , Erk1/2 and Akt1 . In infected primary lymphatic endothelial ( LEC-rKSHV ) cells , which have previously been shown to spontaneously display a viral lytic transcription pattern , transfection of siRNA against K15 , but not K1 , abolished viral lytic replication as well as KSHV-induced spindle cell formation . Using a newly generated monoclonal antibody to K15 , we found an abundant K15 protein expression in KS tumor biopsies obtained from HIV positive patients , emphasizing the physiological relevance of our findings . Finally , we used a dominant negative inhibitor of the K15-PLCγ1 interaction to establish proof of principle that pharmacological intervention with K15-dependent pathways may represent a novel approach to block KSHV reactivation and thereby its pathogenesis . Kaposi’s sarcoma-associated herpesvirus ( KSHV ) , also known as human herpesvirus –8 ( HHV-8 ) , causes Kaposi’s sarcoma ( KS ) [1] and two lymphoproliferative disorders: primary effusion lymphoma ( PEL ) [2] and the plasmablastic variant of multicentric Castleman’s disease ( MCD ) [3] . KS is the commonest neoplasm associated with KSHV infection and among the first clinical manifestations in untreated AIDS patients [4] . Histologically , it is characterized by KSHV latent nuclear antigen ( LANA ) positive proliferating spindle cells of endothelial origin , infiltrates of immune cells such as monocytes , plasma cells , and B and T cells , as well as extensive neo-angiogenesis with slit-like vascular spaces which allows the extravasation of erythrocytes to the surrounding tissue and results in the characteristic purplish appearance of the lesion [5] . In the later nodular stage of KS the majority of spindle cells , 90% , harbor the virus in its latent state , while a small proportion of these cells also display lytic replication , an important phenomenon in the pathogenesis of KS [6–8] . Cells supporting KSHV lytic replication are thought to contribute to the progression of KS through secretion of proangiogenic and proinflammatory factors , which can act both in an autocrine and paracrine manner and can release infectious virus that replenishes the pool of infected cells recruited to latency [5 , 7 , 9 , 10] . The environmental and physiological cues leading to KSHV lytic reactivation in infected individuals are not well defined . However , inflammation [11 , 12] , hypoxia [13–15] , oxidative stress [16] and co-infection with other viruses such as HIV-1[17–19] and other herpesviruses [20–22] have been shown , mostly in experimental settings , to trigger KSHV reactivation from latency through the release of inflammatory cytokines such as IFN-γ , hepatocyte growth factor/scatter factor and Oncostatin M [11 , 12 , 17 , 23 , 24] or the production of reactive oxygen species ( ROS ) [16 , 25 , 26] . These stimuli induce KSHV reactivation through the activation of specific intracellular signaling pathways and their downstream transcription factors such as AP-1 which can directly act on the promoters of several viral lytic genes including the master lytic switch protein RTA ( replication and transcription activator ) , which is sufficient to disrupt KSHV latency [27 , 28] . The activation of all three mitogen-activated protein kinase ( MAPK ) pathways , extracellular signal-regulated kinase ( ERK ) , c-Jun N-terminal kinase ( JNK ) and p38 , as well as their downstream transcription factors AP-1 and Ets-1 have been shown to occur as a result of ROS production from oxidative stress and inflammation , during co-infection with other viruses or treatment with chemical inducers such as 12-O-tetradecanoyl-phrobol-13-acetate ( TPA ) and to be crucial for KSHV reactivation from latency as well as viral infection and replication during primary infection of endothelial cells [25 , 29–35] . Increased expression of Pim-1 and Pim-3 kinases in response to cytokines or chemical inducers of KSHV reactivation can lead to the phosphorylation of the latency-associated nuclear antigen ( LANA ) on serine residues 205 and 206 which counteracts its repressive function on viral lytic transcription [36] . Blocking the activation of protein kinase C ( PKC ) isoforms by using GF 109203X or the PKCδ isoform specifically decreases TPA-induced KSHV reactivation [37]; while intracellular calcium mobilization elicits KSHV lytic replication through the activation of the Ca++ dependent calcineurin-NFAT pathway independently of PKC activation [38] . In addition to the ERK MAPK pathway , co-infection of BCBL-1 cells with either HIV-1 or HSV-1 activates the phosphatidylinositol 3-kinase ( PI3K ) -Akt pathway , inactivates the downstream GSK-3β through its phosphorylation and decreases the expression of the negative regulator PTEN which in turn resulted in KSHV reactivation from latency [34 , 35 , 39]; on the other hand , pharmacological inhibition of the PI3K-AKT pathway induced KSHV lytic reactivation in BC-3 cells [40] . Moreover , activation of the toll-like receptors TLR7/8 , crucial players in pathogen recognition and activation of the host innate immune response , using either its agonists or as a result of co-infection with vesicular stomatitis virus ( VSV ) have also been shown to be crucial for KSHV lytic gene transcription and replication [41] . The open reading frame ( ORF ) K15 , located at one end of the long unique coding region of the KSHV genome between ORF 75 and the terminal repeat region , contains eight alternatively spliced exons encoding multiply spliced variants of the K15 transcript [42–44] . These transcripts encode a K15 protein containing a variable number of membrane spanning domains with an attached short amino ( N ) - terminal and long carboxyl ( C ) -terminal cytoplasmic domain; the longest transcript ( containing all eight exons ) codes for a protein of 45 KDa apparent molecular weight with 12 predicted membrane-spanning domains [44 , 45] . So far , at least three highly divergent ( less than 33% amino acid sequence similarity ) , K15 alleles have been identified in different KSHV genomes , designated as predominant ( P ) , minor ( M ) and N [43–46] . Despite their low sequence similarity , all three K15 alleles feature several conserved putative signaling motifs such as a proline-rich Src homology 3 ( SH3 ) -binding site ( PPLP ) and two SH2-binding sites ( YASIL and YEEVL ) in their C-terminal cytoplasmic tail [43 , 44] . Upon ectopic expression , K15 recruits and constitutively activates the phospholipase C γ1 ( PLCγ1 ) protein by using a constitutively phosphorylated tyrosine residue at its second SH2-binding site ( Y481EEVL ) and one of two SH2 domains of PLCγ1 [42 , 45 , 47–49] . K15 also induces the activation of the two MAP-kinases Erk2 and JNK , calcineurin-NFAT as well as NF-κB pathways leading to the activation of the NFAT and AP-1 transcription factors and expression of proangiogenic and proinflammatory factors such as Dscr-1 , interleukin ( IL ) -6 , IL-8 , IL-1α/β , CCL20 , CCL2 , CXCl3 and Cox-2 as well as intracellular calcium ion ( Ca++ ) mobilization [42 , 45 , 47–50] . Depletion of K15 from the infected cell or its deletion from the viral genome reduces the invasiveness of KSHV infected endothelial cells and their ability to form capillary tubes in a PLCγ1 dependent manner [48 , 49] , linking its activation of PLCγ1-dependent signaling pathways to KSHV induced angiogenesis . K15 has also been shown to interact with the anti-apoptotic protein HAX1 and induce the expression of a number of anti-apoptotic genes including birc2 , brc3 , bf , A20 and bcl2a1 which can provide a survival advantage to the infected cell [47 , 51 , 52] . Given the importance of inflammatory and angiogenic factors as well as intracellular signaling pathways such as the MAPK pathways in KSHV lytic replication ( discussed above ) , we investigated whether K15 could also play a role in this regard . Our results revealed that the expression of the K15 protein in the infected cell is crucial for KSHV lytic replication and its reactivation from latency; K15 also contributes to the activation of PLCγ1 , Erk1/2 and Akt1 signaling during viral lytic replication . We establish K15 as one of the viral proteins responsible for KSHV induced spindle cell formation in stably infected lymphatic endothelial cells ( LECs ) and show its abundant expression at the protein level in KS biopsies . As a proof of principle , we used a dominant negative inhibitor to intervene with the K15-PLCγ1 interaction , K15-dependent signaling pathways and KSHV lytic replication , suggesting that the recruitment of PLCγ1 to K15 might represent a druggable target to block KSHV lytic gene expression at an early stage in its replication cycle and thereby KSHV pathogenesis . Previously , our group has established BJAB-rKSHV cells [39 , 53] , a B cell line stably infected with rKSHV . 219 , in which the KSHV lytic cycle can be activated by cross-linking the B-cell receptor ( BCR ) using an antibody to IgM . This culture system obviates the need for chemical inducers or RTA overexpression and produces a relatively high amount of infectious virus in contrast to PEL cell lines and other KSHV-infected adherent cell lines . To investigate whether K15 could play a role in KSHV lytic reactivation in this system , the K15 mRNA was depleted using an siRNA targeting exon 8 of the K15 mRNA , and the KSHV lytic cycle was induced by using a sub-optimal amount of anti-human IgM antibody ( 0 . 8 μg/ml ) . Viral lytic replication was then assessed by western blot analysis of the expression of KSHV lytic proteins RTA , ORF 45 and K8 . 1 as well as infectious virus release . As shown in Fig 1A , cross-linking the BCR with anti-IgM antibody induced the expression of the indicated KSHV lytic proteins in cells treated with the control , non-targeting/scrambled ( Scr ) siRNA . In contrast , knock down of K15 from these cells led to reduced expression of early lytic viral proteins including the master lytic switch protein RTA as well as ORF 45 , indicating that K15 expression is required at an early step during KSHV reactivation; the expression of the late lytic structural glycoprotein K8 . 1 was also inhibited in response to K15 depletion . An alternative explanation can also be that K15 is involved later during the lytic cycle , by inhibiting a positive feed-back loop during the activation of the lytic cycle . Further analysis of virus production from these cells revealed that knock down of K15 also strongly reduced the amount of infectious virus released after anti-IgM antibody treatment ( Fig 1B ) . We conclude from these results that K15 plays an important role during early viral lytic gene expression and productive replication in this system . To follow up our observation with another approach , we used a recombinant KSHV genome cloned in the Bacterial artificial chromosome clone 36 backbone ( KSHV-Bac36 ) [54] , which harbors a Hygromycin resistance gene as a selectable marker and expresses GFP as a marker of infection . The gene encoding the K15 protein ( nucleotide sequence between 135338 and 136900 ) was then deleted to generate a KSHV-Bac36ΔK15 construct , as described in [48] . The other KSHV non-structural signaling membrane protein K1 , encoded by the K1 gene at the opposite end of the viral genome , also activates a set of signaling pathways which can induce the expression of inflammatory and angiogenic factors similar to the K15 protein [55–58] . Recently , Zhang et al . reported that K1 is required for the activation of the KSHV lytic cycle [59] . Also Steinbrück et al . reported that the KSHV K1 and K15 proteins control a related set of cellular genes when substituted for the LMP2A gene in the EBV genome , an indication that their functions might overlap [52] . To compare the role of K1 with that of K15 during KSHV reactivation , we also produced a K1 deletion construct , KSHV-Bac36ΔK1 , lacking the nucleotide sequence between nt 104 and 970 . The integrity of the wild type as well as the two KSHV-Bac36 deletion mutant constructs was first validated by deep sequencing of the complete genome ( S1 Fig ) . The increased sequence coverage of the region between ORFs K4 . 2 and ORF 19 is due to the duplication of this region in the TR of this Bac [60] . Stably transfected , polyclonal HEK-293 cells were established for all three constructs ( Fig 2A ) . In these cells , the KSHV lytic cycle was induced by using a reactivation cocktail of SF-9 cell supernatant containing a baculovirus expressing the KSHV RTA , which from now on will be referred as only RTA , and sodium butyrate ( SB ) as described before [61] . Treatment of the HEK-293 cells stably harboring the wild type virus with the reactivation cocktail induced the expression of the KSHV immediate-early lytic genes K-bZIP and ORF 45 as well as of K1 and K15 ( Fig 2B ) . In agreement with our K15 knock down results in the Bjab-rKSHV system above ( Fig 1 ) and with a recent report using a K1 deletion virus [59] , the absence of either the K1 or K15 gene reduced the ability of KSHV to reactivate from latency , as evidenced by the decreased expression level of KSHV early lytic proteins K-bZIP and ORF45 as well as reduced levels of infectious virus released into the supernatant of these cultures ( Fig 2B and 2C ) . To extend our investigation to endothelial cells , which are thought to be targets of KSHV infection in vivo and relevant to KS pathogenesis , HuARLT2 cells ( an immortalized endothelial cell line ) [62] were stably infected with the wild type or either of the deletion mutant viruses as described in materials and methods ( Fig 2D ) , and the KSHV lytic cycle was induced using a cocktail of RTA and SB . In these endothelial cells , K15 expression is already detectable before lytic induction , as reported previously [48 , 49] and treatment with the reactivation cocktail further increased its expression as well as the expression of the two early lytic marker proteins K-bZIP and ORF 45 ( Fig 2E ) . Similar to the results described above ( Figs 1 and 2 ) , the expression of these two KSHV lytic proteins as well as virus production were both attenuated in reactivated HuARLT2 cells stably infected with the recombinant viruses lacking either the K1 or K15 gene in comparison to cells infected with the wild type virus ( Fig 2E and 2F ) . We have recently reported [63] an increase in the expression of low molecular weight isoforms of the KSHV LANA protein upon KSHV lytic reactivation . In agreement with this previous finding , induction of the lytic cycle increased the expression level of these low molecular weight LANA isoforms in the HuARLT2-KSHV-Bac36 Wt cells , but not when either the K1 or K15 gene is lacking from the viral genome ( Fig 2E ) . Regarding the role of K1 in KSHV reactivation , our results are consistent with the finding recently reported by Zhang et al . who used a KSHV-Bac16 construct carrying either five stop codons inside the K1 gene or a K1 deletion in HEK-293 or iSLK cells [59] . Together , these results establish a role for both the K1 and K15 proteins in KSHV lytic replication . Upon ectopic expression , K15 has been shown to activate the PLCγ1-calcineurin-NFAT , MAP-Kinase as well as the NF-κB pathways ( Fig 3A ) [45 , 47 , 48 , 50] , which are crucial both during KSHV primary infection as well as during its reactivation from latency ( see Introduction ) . As a first step in investigating the role of K15-dependent signaling pathways in KSHV lytic reactivation , we transfected HEK-293 cells with a K15 expression vector and measured the phosphorylation levels of PLCγ1 ( Tyr783 ) , Akt1 ( Ser473 ) and Erk1/2 ( Thr202/Tyr204 ) on western blots by using phospho-specific antibodies against the indicated residues . In line with experiments reported earlier [45] , in which co-transfection of Erk2 together with K15 had been shown to increase Erk2 kinase activity as measured by the increased phosphorylation of the myelin basic protein ( MBP ) in an in vitro kinase assay , we found that endogenous Erk1/2 protein is phosphorylated on Thr202 and Tyr204 residues in response to K15 expression ( Fig 3B ) . Consistent with previous results in endothelial cells from our group [48] , K15 expression in HEK-293 cells also induced the activation of the PLCγ1 pathway as evidenced by its increased phosphorylation on Tyr783; in contrast , we did not observe any activation of the PI3K-Akt pathway in these cells in response to K15 expression as shown by the lack of increased Akt1 phosphorylation on Ser473 ( Fig 3B ) . Next , we investigated the activation level of these signaling components in HEK-293 cells stably transfected with the KSHV-Bac36 wild type virus before and after the induction of the viral lytic cycle and showed that PLCγ1 , Akt1 and Erk2 are activated upon lytic replication as can be seen from their increased phosphorylation at the indicated residues ( Fig 3C ) . Deleting either K1 or K15 from the viral genome compromised the ability of KSHV to activate PLCγ1 , Akt1 and Erk1/2 upon lytic reactivation in HEK-293 cells ( Fig 3C ) . The fact that the lack of K15 in the infected cell abolished Akt1 phosphorylation in response to reactivation ( Fig 3C ) , even though K15 did not activate this pathway directly when expressed in isolation in these cells ( Fig 3B ) could be due to the decreased lytic replication which results in decreased expression of other KSHV lytic proteins such as vGPCR and K1 that are known to activate the PI3K-Akt pathway directly ( reviewed in [64] ) . In contrast , K15 activates phosphorylation of Erk1/2 and PLCγ1 directly ( Fig 3B ) , and the decreased levels of Erk1/2 and PLCγ1 phosphorylation in cells infected with the K1 or K15 deletion viruses could therefore be the direct result of the absence or reduced expression of K15 in the case of the K1 deletion virus . Consistent with the higher level of K15 expression in uninduced KSHV infected endothelial ( HuARLT2-rKSHV ) cells than in HEK-293 cells ( Fig 2B and 2E ) , there is already a detectable level of PLCγ1 ( Tyr783 ) phosphorylation in the uninduced KSHV-infected HuARLT2 cells compared to the uninfected control cells . PLCγ1 phosphorylation further increased upon induction of the lytic cycle , in line with the increased levels of K15 following lytic reactivation ( Fig 3D ) . Similar to the HEK-293 cells , deleting either K1 or K15 from the viral genome also compromised the ability of KSHV to activate PLCγ1 and Akt1 ( Fig 3D ) upon lytic reactivation in the HuARLT2 cells . However , unlike either PLCγ1 or Akt1 , the basal level of Erk1/2 phosphorylation on Thr202/Tyr204 is already abundant in the uninfected HuARLT2 cells ( Fig 3D ) , which could be due to the hTERT and SV40 large T antigen used for immortalization of these cells ( see Materials and methods ) . We also observed an increased Erk1/2 ( Thr202/Tyr204 ) phosphorylation without induction of the lytic cycle in these endothelial cells infected with viruses lacking either the K1 or K15 gene which decreased upon lytic reactivation ( Fig 3D ) . Taken together , our results so far indicate that the K1 and K15 proteins play an important role during KSHV lytic replication and its activation of cellular signaling pathways and that the absence of K1 or K15 affects the PI3K-Akt and MEK/Erk pathways in endothelial cells in a manner that would not have been predicted from the outcome of experiments involving overexpression of these proteins . Since the phenotypes of both K1 and K15 deletion mutant viruses were similar in regard to virus reactivation and signaling activation ( at least for the signaling components shown here ) , we wondered whether the two proteins might be present in similar cellular compartments . To investigate this in the context of viral infection ( in stably infected HuARLT2-rKSHV cells after lytic induction ) , an immunofluorescence assay ( IFA ) was performed using monoclonal antibodies to K1 [65] and K15 [49]; examples of cells expressing K15 ( shown in green ) alone or together with K1 ( shown in red ) are shown in Fig 4A . As can be seen in Fig 4A , K1 shows a uniform localization throughout the cytoplasm , while K15 is found in large vesicle-like structures more abundant in the perinuclear area and in the cell periphery . Cells co-expressing the two proteins show no clear co-localization ( Fig 4A ) . To further understand the intracellular localization of these two proteins , we analyzed their incorporation into lipid rafts . Lipid rafts are highly dynamic membrane microdomains formed by the association of glycosphingolipids with cholesterol that are involved in protein/lipid trafficking and cellular signal transduction [66 , 67] . Isolation of detergent-resistant membranes ( DRMs ) by lysis of cells in cold nonionic detergents such as Triton X-100 coupled with ultracentrifugation in a sucrose gradient , known as flotation assay , is a commonly used method to assess the composition of lipid rafts and associated signaling proteins . In this assay , DRMs/lipid rafts float to a position of lower density , while soluble proteins remain at the bottom of the gradient with the high-density sucrose . Using this biochemical assay , we showed previously that the K15 protein expressed from a transfected plasmid associates with lipid rafts in Cos7 cells [45] . Here , we assessed the association of both K1 and K15 in KSHV-infected cells with DRMs both before and after induction of the lytic cycle . In HEK-293 cells stably infected with KSHVBac36 virus , a portion of the K15 protein can be detected in fractions 3 and 4 of the sucrose gradient in the same manner as the lipid raft marker protein Caveolin , while a significant portion still remains at the bottom of the gradient in fractions 9 to 11 which contains solubilized proteins ( Fig 4B ) . We did not observe a differential localization of K15 to or away from lipid rafts in response to KSHV lytic reactivation in these cells ( Fig 4B ) . Unlike K15 , most of the K1 protein remained in the soluble fraction at the bottom of the sucrose gradient ( fractions 10 to 11 ) with a very small portion of it floating to fraction 4 after induction of the lytic cycle ( Fig 4B ) . Similar results were observed in the stably infected HuARLT2 cells ( Fig 4C ) , although K1 was hardly detectable in the DRM fractions of this experiment . Further analysis of the intracellular localization of K15 in stably infected HuARLT2-rKSHV cells by IFA staining using markers specific for different cytoplasmic compartments revealed the presence of K15 in vesicular structures in the perinuclear area . Here , K15 co-localized with a cis-Golgi network marker GM130 as well as a late endosome marker LAMP1 ( Fig 4D ) . Consistent with the flotation assay experiment , only a portion of the K15 protein co-localized with the lipid raft marker Caveolin at the cell periphery while there was no clear localization to clathrin positive vesicles ( Fig 4D ) . In order to get further mechanistic insight into the role of K15-induced signaling in KSHV reactivation , BJAB-rKSHV . 219 cells were treated with pharmacological inhibitors for the PLCγ ( U73122 ) , MAPK ( U0126 ) and PI3K/Akt ( Ly294002 ) pathways and the KSHV lytic cycle was induced by using anti-hIgM antibody . Treatment of these cells with 2 μM of U73122 decreased PLCγ1 phosphorylation in reactivated cultures; this was accompanied by reduced viral lytic gene expression and infectious virus release ( Fig 5A–5C ) . Similarly , consistent with the literature ( see Introduction ) , 12 . 5 μM of the MAPK inhibitor U0126 as well as 10 μM of the PI3K-Akt inhibitor Ly294002 efficiently inhibited KSHV lytic replication as shown both by the dramatic reduction in lytic gene expression as well as infectious virus production ( Fig 5D–5F ) . In accordance with the results of the these experiments , depleting PLCγ1 from the stably infected HuARLT2-rKSHV cells by siRNA affected the ability of the virus to reactivate from latency as shown by the reduced number of RFP positive cells as well as lytic gene expression ( S2 Fig ) . This observation confirms the involvement of PLCγ1 in KSHV reactivation in endothelial cells . Previously [49] , we have shown that the isolated cSH2 domain of PLCγ2 can bind to the K15 protein and inhibit the recruitment of PLCγ1 to K15 in a dominant negative manner , thereby inhibiting K15-dependent PLCγ1 signaling , K15-induced angiogenesis and invasiveness . To investigate whether it could inhibit KSHV reactivation as well , a plasmid expressing the isolated PLCγ2-cSH2 domain or its empty vector control was transfected into the HEK-293 cells carrying the KSHV Bac36 wild type virus genome and its effect on KSHV reactivation was assessed . Expression of the PLCγ2-cSH2 domain in these cells inhibited virus reactivation as shown by the decreased expression level of the KSHV lytic proteins K-bZIP , ORF45 and K15 itself ( Fig 6A ) , which was accompanied by reduced infectious virus release ( Fig 6B ) . Transfection of the isolated PLCγ2-cSH2 domain also inhibited the activation of PLCγ1 ( phosphorylation on Tyr783 ) , Akt1 ( Ser473 ) and Erk2 ( Thr202/Tyr204 ) after induction of the KSHV lytic cycle compared to KSHV-carrying cells transfected with the empty vector ( Fig 6C ) . To extend this observation to endothelial cells , stably infected HuARLT2-rKSHV cells were transduced with a lentiviral vector expressing the isolated PLCγ2-cSH2 domain . Similar to the results in the epithelial cells , expression of the isolated PLCγ2-cSH2 domain abolished virus reactivation as shown by the reduced number of RFP ( expressed under the control of the lytic PAN promoter ) expressing cells ( S3A Fig ) , and decreased K-bZIP and ORF 45 expression ( S3B Fig ) , as well as reduced infectious virus release ( S3C Fig ) compared to cells transduced with an empty vector control . The isolated PLCγ2-cSH2 domain was also cloned into a retroviral vector ( pSF91-IRES-GFP ) which , unlike the lentivirus vector , can express GFP to allow the monitoring of transduction efficiency . The stably infected HuAR2T-rKSHV cells were then transduced with this retroviral vector expressing the PLCγ2-cSH2 ( pSF91-PLCγ2-cSH2-IRES-GFP ) or with the empty vector as a control . Expression of the isolated PLCγ2-cSH2 domain from the retroviral vector also inhibited KSHV reactivation as shown by the reduced RFP ( Fig 6D lower panel ) and K-bZIP expression ( Fig 6E ) compared to cells transduced with the empty vector; equal transduction efficiency of both retroviral stocks was assessed based on GFP expression in uninfected HuAR2T cells that had been transduced with the pSF91-PLCγ2-cSH2-IRES-GFP or control retroviral vector in parallel ( Fig 6D upper panel ) . Chang and Ganem have recently described a unique KSHV transcription program with a widespread expression of both latent and lytic genes in stably infected lymphatic endothelial cells ( LECs ) , which differs from the traditional latency program observed in stably infected blood endothelial cells such as HUVECs [68] . The fact that these cells support lytic replication without the need for ectopic RTA expression or treatment with chemical inducers such as sodium butyrate may permit the investigation of the KSHV lytic cycle in LECs in a perhaps more physiological model . To investigate the role of K15 in this regard , LECs as well as HuARLT2 , an endothelial cell line derived by immortalizing HUVECs , were infected with the recombinant rKSHV . 219 virus and maintained under puromycin selection for two weeks to generate stably infected polyclonal cell populations , LEC-rKSHV and HuARLT2-rKSHV . KSHV infection of endothelial cells in culture is known to induce a change from their typical coble-stone morphology into an elongated spindle shape , a phenotypic change known as spindling [69] , which is similar to the appearance of KS spindle cells in KS lesions . Microscopic examination of the infected cells revealed that KSHV induced extensive spindling in LEC cells , which was already visible after 48 hours of infection ( see images taken 7 days after infection shown in Fig 7A ) , while HuAR2T-rKSHV cells did not show an obvious change in their morphology; the level of infection is shown by the expression of GFP in both cell populations ( Fig 7A ) . Additionally , the LEC-rKSHV cells are also expressing the lytic marker RFP a week after infection indicating that there is still ongoing lytic replication ( Fig 7A upper panel ) . In contrast , lytic gene expression in the HuARLT2-rKSHV cells has subsided one week after infection ( Fig 7A lower panel ) , but can be induced using RTA and SB ( S4 Fig ) . Consistent with the RFP expression , the KSHV lytic proteins K-bZIP and ORF 45 are expressed in the LEC-rKSHV cells after two weeks of infection as shown by western blot analysis ( Fig 7B upper panel ) ; in contrast , in HuARLT2-rKSHV cells , this can only be seen after lytic reactivation with RTA and SB ( Fig 7B lower panel ) . These findings are in line with those reported by Chang and Ganem [68] . Analysis of the expression of the K15 protein in the stably infected HuARLT2-rKSHV cells ( Fig 7B lower panel ) revealed that , unlike K-bZIP and ORF 45 , which are expressed only after lytic induction , K15 protein is already detectable in latently infected cells and its expression increases upon lytic reactivation in a similar fashion to what was observed in HuARLT2 cells infected with KSHV Bac36 virus ( see Fig 2E ) . Interestingly , a similar experiment in the stably infected LEC-rKSHV cells revealed an abundant K15 expression ( Fig 7B upper panel ) . Further , we performed immunofluorescence ( IF ) staining of endogenous K1 and K15 proteins as well as ORF 59 ( KSHV DNA polymerase processivity factor-8/PF-8 ) as a lytic marker in the stably infected LEC-rKSHV or HuARLT2-rKSHV cells with or without induction of the lytic cycle . The result ( Fig 7C ) showed that a small fraction of HuARLT2-rKSHV cells expresses the K1 or K15 proteins , 17% and 6 . 5% respectively , which increased to 28% and 10% respectively upon induction of the lytic cycle . In contrast , the majority of the LEC-rKSHV cells were expressing the K15 protein while the number of K1 expressing cells is negligible ( Fig 7C ) in agreement with the absence of detectable K1 mRNA expression in the previously reported KSHV tiling microarray study on KSHV-LECs [68] . Unlike the K1 and K15 proteins , ORF 59 is expressed only after induction of the lytic cycle in 6% of HuARLT2-rKSHV cells , while it is already present in 8% of LEC-rKSHV cells without any stimulation for lytic reactivation ( Fig 7C ) . Next , we co-stained the K15 and ORF 59 proteins by IFA in HuARLT2-rKSHV cells , in which the KSHV lytic cycle was induced using KSHV RTA and SB . Interestingly , ORF 59 staining was evident only in the K15 expressing cells , while the reverse is not true ( representative IF images of cells expressing K15 alone or together with ORF 59 are shown in Fig 7D upper and lower panel respectively ) , in agreement with the greater number of cells ( 10% ) expressing the K15 protein compared to the number of cells expressing ORF 59 ( 6% ) during lytic replication ( Fig 7C ) . Moreover , K15 expression in lytic ( ORF 59 co-expressing ) cells shows a dispersed and peripheral localization ( Fig 7D ) from its perinuclear abundance in vesicular structures in the non-lytic cells ( Figs 7D and 4D ) . Similar to the result in the induced HuARLT2-rKSHV cells , no detectable ORF 59 expression was observed in the absence of K15 expression in the stably infected LEC-rKSHV cells ( representative IF images are shown in Fig 7E ) , indicating that K15 expression in the infected cell might be a pre-requisite for KSHV lytic replication . We then asked whether the abundant expression of K15 in the stably infected LEC-rKSHV cells could contribute to the lytic gene expression program and ongoing lytic replication in these cells . To investigate this question , stably infected LEC-rKSHV cells were transfected with either a scrambled ( Scr ) control siRNA or two siRNAs against K15 or a single siRNA against K1; siRNA targeting the mRNA for the small capsid protein encoded by ORF 26 was included as a control for a late lytic gene which affects only virus production but not early lytic viral gene expression . First , we investigated the effect of K15 depletion on KSHV lytic replication by western blot analysis of the expression of K-bZIP and ORF 45 as well as LANA proteins . The result ( Fig 8A ) shows that silencing the K15 mRNA in LEC-rKSHV cells results in a pronounced decrease of K-bZIP and ORF 45 expression as well as the low molecular weight isoform of LANA , while the higher molecular weight isoforms of LANA were unaffected . Transfection of the siRNA against K1 did not reduce lytic gene expression in KSHV-infected LECs ( Fig 8A ) ; this is in line with the absence or low expression of K1 in these cells ( see Fig 7C ) . Likewise , as expected , silencing ORF 26 mRNA did not affect early lytic gene expression ( Fig 8A ) . As reported previously [68] , we also found that KSHV-infected lymphatic endothelial cells can , in addition to showing an extended lytic gene expression program , also release substantial titers of infectious KSHV ( Fig 8B ) . Silencing K15 mRNA in these cells abrogated infectious virus release to the level of knocking down the capsid protein ORF 26 , while transfection with siRNA against K1 did not affect virus production ( Fig 8B ) . Further , western blot analysis of the phosphorylation level of PLCγ1 , Akt1 and Erk1/2 in the LEC-rKSHV cells revealed that all three pathways are activated in the stably infected cells compared to the uninfected control cells ( Fig 8C , lanes 1 and 2 ) . Phosphorylation at the indicated residues of PLCγ1 , Akt1 and Erk1/2 was reduced in response to K15 silencing but not in cells transfected with siRNAs against either K1 or ORF 26 ( Fig 8C ) . Interestingly , in addition to inhibiting the lytic transcription in the stably infected LEC-rKSHV cells , K15 depletion also led to the loss of the pronounced spindle cell morphology induced by KSHV infection ( Fig 8D ) ; in contrast , we did not observe an obvious phenotypic change after transfecting siRNAs against either K1 or ORF 26 ( Fig 8D ) . In support of a role for K15 in spindle cell formation , we also show that ectopic expression of K15 alone in LECs can induce morphological change similar to KSHV infection ( Fig 9A ) and this is accompanied by the activation of PLCγ1 , Akt1 and Erk1/2 signaling pathways ( Fig 9B ) . This suggests a direct role of K15 in the activation of these pathways in the stably infected LEC-rKSHV cells in agreement with the reduced phosphorylation of PLCγ1 , Akt1 and Erk1/2 following K15 silencing in KSHV-infected LECs ( Fig 8C ) . We could also show that inhibiting the K15-PLCγ1 interaction by transducing the retroviral vector for the PLCγ2-cSH2 domain ( see above ) as a dominant negative inhibitor can reduce the phosphorylation of PLCγ1 , and , to a lower extent , Akt1 and Erk1/2 ( Fig 9C ) , viral lytic protein expression ( Fig 9D ) and infectious virus release ( Fig 9E ) in the stably infected LEC-rKSHV cells , consistent with our results obtained with the HEK-293-Bac36 and HuARLT2-rKSHV cells ( see above ) . Hosseinipour and colleagues have recently studied the transcription profile of KSHV in KS tissue biopsies from treatment-naïve HIV-positive patients by using a KSHV real-time quantitative PCR array with multiple primers per open reading frame [70] . In this study , the K15 mRNA was shown to be expressed as abundantly as the genes in the viral latency locus ( LANA , vCyc , vFLIP , Kaposin , and miRNAs ) in KS tissue biopsies with both a restricted and relaxed KSHV transcription profile [70] . However , K15 protein expression depends on post-transcriptional regulation exerted by KSHV ORF 57 [71] . Owing to the lack of suitable antibodies , it has not been possible so far to investigate whether the K15 protein is expressed in KS lesions . Here , the suitability of our rat anti-K15 monoclonal antibody ( clone 18E5 ) [49] for staining paraffin-embedded tissue blocks was first tested on paraffin embedded cell blocks prepared from HuARLT2 or HuARLT2-rKSHV cells representing K15 negative and positive samples respectively . As shown in Fig 10A , when used with a tyramide signal amplification protocol , this antibody detects a cytoplasmic K15 staining in a fraction of the cells in the HuARLT2-rKSHV cell block which is consistent with its expression profile in our IFA experiments . In contrast , there is no detectable signal in the HuARLT2 cell blocks indicating the specificity of our staining protocol . Next , we applied this staining protocol to 9 KS tissue biopsies obtained from HIV positive patients . In this experiment , we also stained for KSHV LANA as a marker of infection , the endothelial cell marker CD34 as well as Hematoxylin and Eosin ( H&E ) to visualize tissue structures . Fig 10B shows four examples in which the K15 protein is expressed at a varying level of abundance , mirroring the abundance of LANA positive cells , in these KS tissues . LANA immunostaining for the fourth case is not shown because of the small size of the paraffin block in this case . We also observed a similar K15 expression of varying abundance in the remaining 5 analyzed tissue samples; however , because of the small size and low quality of these tissue biopsies , we could not stain for LANA as well as CD34 and H& E and therefore did not include them in Fig 10B . These findings are consistent with the K15 mRNA expression profile in KS biopsies described recently [70] . Genotyping of the KSHV genome extracted from these KS tissue biopsies using primers specific for the K15P and M alleles revealed that 7 out of the 9 samples were carrying the K15P type allele while 2 samples contained the less common K15M allele ( Fig 10B ) . Interestingly , we were also able to detect the K15M protein using our monoclonal antibody in a KS biopsy that was positive for the K15M allele ( Fig 10B last panel ) . Despite the fact that the two K15 protein forms are very divergent in their amino acid sequence , they contain conserved sequence motifs ( see Introduction ) . Therefore , we analyzed the K15 epitope recognized by the monoclonal antibodies clone 10A6 ( used in western blot ) and 18E5 ( used in IFA and IHC staining ) by using a synthetic peptide array of 44 peptides derived from the K15 cytoplasmic tail sequence and overlapping with 3 amino acids . The last two peptides ( number 43 and 44 ) with the peptide sequence ATQPTDDLYEEVLFP and PTDDLYEEVLFPRN respectively reacted with both monoclonal antibodies ( S5A and S5B Fig ) . Although it was raised against a recombinant GST-tagged cytoplasmic domain of the K15P type protein , our monoclonal antibody recognizes one of these conserved motifs ( the protein sequence PTDDLYEEVLFP ) , including and flanking the SH2 domain-binding site YEEVL at the c-terminal end of the K15 cytoplasmic tail ( S5A and S5B Fig ) . To confirm that this antibody could also detect proteins encoded by the K15M type allele , we transfected a plasmid expressing either of the two K15 forms and performed IFA . Indeed , both K15P and M proteins were recognized by our monoclonal antibody clone 18E5 ( S5C Fig ) . The majority of KS endothelial spindle cell in KS tumors is latently infected with KSHV and there is substantial evidence that several latent KSHV genes , i . e . LANA [72 , 73] , vCyc [74 , 75] , vFLIP [76 , 77] and the KSHV miRNAs [78 , 79] may contribute to the increased survival and proliferation of infected endothelial cells . In addition , several early KSHV genes , such as vIL6 and vGPCR , as well as productive replication and seeding of KSHV into new cells are important for the pathogenesis of KS [7 , 80–83] . In keeping with this , a small fraction of KS spindle cells in advanced KS lesions express lytic viral proteins [6 , 8 , 84–86] , most KS lesions display a viral gene expression pattern that includes early lytic genes [70] and release of angiogenic and inflammatory factors from such type of cells during lytic replication have been suggested to create a conducive microenvironment for the development of KS [5 , 9 , 82 , 83] . In this study , we investigated the role of the K15 protein in the virus life cycle and in KSHV-induced pathogenesis . Having previously shown that K15 contributes to the increased invasiveness and angiogenic properties typical for KSHV-infected endothelial cells by recruiting and activating PLCγ1 [48 , 49] , we now provide evidence that K15 uses the same signaling pathway to support the activation of the early stages of the lytic cycle . We also provide proof of principle that the recruitment of PLCγ1 by K15 may represent a novel therapeutic target to silence KSHV reactivation ( Figs 6 and 9 ) . Furthermore , we demonstrate the expression of the K15 protein in a substantial proportion of KSHV-infected endothelial spindle cells in KS lesions ( Fig 10 ) as a prerequisite for defining K15 as a novel therapeutic target . Both by siRNA-mediated silencing as well as deleting the K15 gene from the viral genome we show , in a variety of KSHV infected cells , that the expression of K15 is crucial for reactivation from latency , early viral gene expression and virus production ( Figs 1 , 2 and 8 ) . K15 expression is required earlier during lytic replication as demonstrated by the reduced expression of immediate early lytic viral proteins including the master lytic switch protein RTA , ORF 45 and K-bZIP upon K15 knock down in stably infected cells or when K15 is deleted from the viral genome ( Figs 1A , 2B and 2E ) . The decreased expression of these immediate-early/ early viral proteins RTA , ORF 45 and K-bZIP in the absence of K15 can further affect the expression of late lytic proteins , such as the envelope glycoprotein K8 . 1 ( Fig 1A ) and therefore lead to reduced virus production ( Figs 1B , 2C and 2F ) . We also compared the role of K15 to that of K1 , which has previously been shown to be required for efficient viral reactivation [59 , 87 , 88] . Interestingly , while confirming these earlier reports for epithelial and blood vascular endothelial cells ( Fig 2 ) , we also found that K1 is only expressed at low levels , and may therefore not contribute to KSHV-reactivation , in infected lymphatic endothelial cells ( Figs 7 and 8 ) . KSHV infected lymphatic endothelial cells are spontaneously lytic [68] and their cellular transcriptome profile has been shown to resemble that of KS biopsies [89 , 90] . They may therefore reflect the in vivo situation in KS tumors better than KSHV-infected blood vascular endothelial cells , in which KSHV tends to be more latent ( Fig 7 ) . In keeping with this notion , the previously reported expression of K15 mRNA in many KS biopsies from AIDS patients [70] , as well as the presence of activating epigenetic marks on the K15 genomic locus [91] , we show here that K15 is spontaneously expressed in KSHV-infected blood vascular ( HuARLT2 ) and lymphatic endothelial cells ( Figs 2E and 7–9 ) , as well as in a substantial proportion of KS spindle cells in tumor biopsies ( Fig 10 ) . In cultured endothelial cells , the number of K15 positive cells vastly exceeds that of cells undergoing lytic viral DNA replication , as indicated by expression of the viral polymerase-associated factor encoded by ORF 59; in addition , ORF 59-positive cells are always positive for K15 , while the reverse is not the case ( Fig 7 ) . It appears therefore possible that K15 expression ‘primes’ endothelial cells for the onset of lytic replication but is not sufficient for this process to initiate . In addition to K1 and K15 , the viral G-protein coupled receptor ( vGPCR ) has been shown to positively regulate viral lytic replication [92 , 93] . All three viral membrane proteins exert their roles by activating intracellular signaling pathways , in particular PLCγ , Erk1/2 , Akt1 or NF-κB [64 , 87 , 88 , 92 , 93] . We show here that the absence of K15 impedes the activation of PLCγ1 , Erk1/2 and Akt1 ( Figs 3C , 3D and 8C ) , and that these pathways are also directly activated by K15 , at least in the lymphatic endothelial cells ( Figs 3B and 9B ) . We believe that in HEK-293 cells , in which Akt is not activated by K15 ( Fig 3B ) , the decreased expression of other viral proteins such as K1 and vGPCR account for the lack of Akt1 activation in the absence of K15 ( Fig 3C and 3D ) . In fact , our results suggest that deleting K15 is accompanied by reduced K1 expression upon viral lytic reactivation ( Fig 2 ) . The phenotypic similarity of the K1 and K15 deletion mutant viruses in regard to virus reactivation as well as intracellular signaling activation ( Figs 2 and 3 ) led us to investigate their intracellular localization . In the context of viral infection , we found both viral membrane proteins to be predominantly localized in intracellular vesicular structures rather than at the cell surface membrane ( Fig 4A and 4D ) . Overexpressed K15 is known to localize to DRMs ( lipid rafts ) [45] . We show here that , in KSHV infected cells , a portion of K15 is also associated with lipid rafts at the cell periphery while a substantial amount localizes into the perinuclear area in GM130 and LAMP1 positive vesicles ( Fig 4B–4D ) . Not only did K15 depletion affect viral replication in KSHV-infected lymphatic endothelial cells , it also reversed the virus-induced endothelial cell spindling ( Fig 8D ) , and K15 overexpression in these cells recapitulates KSHV infection with regard to spindle cell formation ( Fig 9A ) . Previous work [94 , 95] has shown the latent KSHV protein viral FLIP ( vFLIP ) to be crucial in this regard through its activation of the NF-κB pathway [68] . K15 can also mediate NF-κB activation [45 , 96 , 97] , suggesting that K15 induced LEC spindling may also be the result of NF-κB activation . Activation of the PLCγ1 and MAPK pathways ( Fig 9B ) which can also affect the expression of genes involved in cytoskeleton reorganization could also contribute to this phenotype . In view of our observation ( Figs 7 and 8 ) that , the absence of K15 also impedes the spontaneous lytic reactivation observed in KSHV-infected lymphatic endothelial cells , we explored if K15 might serve as a potential therapeutic target to interfere with KSHV early gene expression and lytic replication . We found that overexpression of the PLCγ2-cSH2 domain , which we have previously shown to interfere with the recruitment of PLCγ1 to K15 [49] can inhibit PLCγ1 , Erk1/2 and Akt1 phosphorylation , the expression of early viral proteins and virus production in KSHV-infected epithelial and endothelial cells ( Figs 6 and S3 ) , as well as in the spontaneously lytic lymphatic endothelial cells ( Fig 9 ) . These results suggest that , if a small molecule inhibitor of the K15-PLCγ1 interaction could be developed , it might be used to interfere with KSHV early gene expression and thereby its role in the pathogenesis of Kaposi’s Sarcoma . Written informed consent obtained from all patients covered the use of human biopsies , taken for diagnostic purposes , for this study and was approved by the Hannover Medical School Ethics Committee ( Approval Nr . 3381–2016 ) . Experiments were conducted in accordance with the Declaration of Helsinki . The list of primary and secondary antibodies used for western blot analysis is shown below . The rabbit anti-KSHV RTA polyclonal antibody [98] was a kind gift from David Lukac ( Rutgers new Jersey Medical School , Newark , New Jersey , USA ) . The production of the rat anti-KSHV K15 monoclonal antibody ( clone number 10A6 ) was described before [49] and the mouse anti-KSHV K1 monoclonal antibody ( Clone 3H4 ) [65] was kindly provided by Jae U . Jung ( University of Southern California , Los Angeles , California , USA ) . Primary antibodies rabbit anti-GAPDH ( #14C10 ) , rabbit anti-S-Tag ( #8476 ) , rabbit anti-PLCγ1 ( #2822 ) , rabbit anti-phospho PLCγ1 ( Tyr783; #2821 ) , mouse anti-phospho MAP-Kinase p44/42 ( Thr202/Tyr204; #91062 ) , rabbit anti-phospho Akt1 ( Ser473; #4058 ) , mouse anti-Akt1 ( #2967 ) and mouse anti-MAP-Kinas p44/42 ( #9107 ) were purchased from Cell Signaling Technology; mouse anti-Erk2 ( D-2 ) ( sc-1647 ) , mouse anti-KSHV ORF 45 ( sc-53883 ) and mouse anti-KSHV K-bZIP ( sc-69797 ) , were purchased from Santa Cruz Biotechnolocy Inc . ; mouse anti KSHV K8 . 1 ( A2B ) ( 13-212-100 ) , rat anti-KSHV ORF73 ( LNA-1;13-210-100 ) and mouse anti-KSHV ORF 59 ( 13-211-100 ) were purchased from Advanced Biotechnology Inc . ; mouse anti-HHV-8 ORF26 ( LS-C41403 ) was purchased from LifeSpan Biosciences Inc . ; mouse anti-β-actin ( A5441 ) was purchased from Sigma Aldrich; and rabbit anti-caveolin ( 610060 ) was purchased from BD Transduction Laboratories . HRP-conjugated secondary antibodies: goat anti-rabbit IgG ( P0448 ) , rabbit anti-mouse IgG ( P0260 ) and goat anti-mouse IgG ( P0447 ) were purchased from DAKO; goat anti-rat IgG ( #3050–05 ) was purchased from SouthernBiotech . The following primary and secondary antibodies were used for IFA . The production of the rat anti-KSHV K15 monoclonal antibody ( clone 18E5 ) was described before [49] . Here , the K15 epitope necessary for antibody binding for both clones used in western blot ( clone 10A6 ) and IFA ( clone 18E5 ) was mapped by using a microarray of 44 overlapping synthetic peptides from the K15 cytoplasmic tail sequence with 3 amino acid shifts as described before [99] . In brief , the peptides were synthesized using the SPOT method on a soluble cellulose matrix [100] . After dissolving , the cellulose-peptide conjugates were printed to glass slides via the SC2 process [101] , resulting in 8 arrays per slide . The microarrays were blocked with 2% ( w/v ) casein in TBS-T overnight at room temperature , followed by incubation with primary antibodies ( 10A6 and 18E5 ) in a dilution of 1:100 in blocking buffer overnight at 4°C . After three washes with TBS-T , staining of bound antibodies and control spots was carried out at room temperature via incubation with secondary antibodies ( Cy3-anti-rat IgG conjugate and Streptavidin-Cy5 ) at a dilution of 1:240 for 90 min . The slides were then washed three times with TBST and two times with dH2O for 5min each , dried in a stream of nitrogen and scanned with an Agilent microarray scanner ( Agilent Technologies , Inc . , Santa Clara , CA , USA ) . The mouse anti-KSHV K1 monoclonal antibody ( Clone 2H5 ) [65] was kindly provided by Jae U . Jung ( University of Southern California , Los Angeles , California , USA ) ; and the mouse anti-KSHV ORF 59 antibody ( 13-211-100 ) was purchased from Advanced Biotechnology Inc . The following fluorescently labelled secondary antibodies were used: FITC-conjugated donkey anti-rabbit IgG ( 711-095-152; Jackson Immuno Reasearch ) , Cy3-conjugated donkey anti-rat IgG ( 712-165-153; Jackson Immuno Reasearch ) , Cy3-conjugated donkey anti-mouse IgG ( 715-165-151; Jackson Immuno Reasearch ) , Streptavidin-Cy5 ( 016-170-084; Jackson Immuno Research ) , and Cy5-conjugated goat anti-mouse IgG ( 115-175-164; Jackson Immuno Research ) . HEK-293 ( ATCC , CRL-1573 ) , HEK-293T ( DSMZ No . : ACC 305 ) and HeLa CNX ( DSMZ No . : ACC 57 ) cells were maintained in Dulbecco’s modified Eagle medium ( DMEM; Gibco , lifetechnologies , Paisley , UK ) supplemented with 10% heat inactivated fetal bovine serum ( FBS; Hyclone , Cramlington , UK ) . BJAB-rKSHV . 219 cells are BJAB cells ( DSMZ No . : ACC 757 ) stably infected with rKSHV . 219 , a recombinant virus derived from JSC-1 cells that expresses the red fluorescent protein ( RFP ) from the KSHV lytic gene PAN promoter , the green fluorescent protein ( GFP ) from the cellular EF-1α promoter , and a puromycin resistance gene as a selectable marker [61] and have been described previously [53]; they were grown and maintained in RPMI 1640 medium ( Gibco , lifetechnologies , Paisley , UK ) supplemented with 20% heat inactivated FBS in the presence of 4 . 2 μg/ml puromycin ( A2856; Aplichem GmbH , Darmstadt , Germany ) . Primary juvenile foreskin lymphatic endothelial cells ( LECs; Lonza ) were provided by Päivi M . Öjala ( University of Helsinki , Helsinki , Finland ) and grown in EGM-2MV medium ( Lonza , Walkersville , MD , USA ) . HuARLT2 , a conditionally immortalized human endothelial cell line derived from HUVECs by expressing doxycycline-inducible transgenes simian virus 40 ( SV40 ) large T antigen ( TAg ) and human telomerase reverse transcriptase ( hTert ) [62] , was kindly provided by Dagmar Wirth ( HZI , Braunschweig , Germany ) and was grown and maintained in EGM-2MV medium in the presence of 1 μg/ml doxycycline ( D9891; Sigma-Aldrich , Saint Luis , Missouri , USA ) . HuARLT2-rKSHV . 219 , HuARLT2 cell line stably infected with rKSHV . 219 as described elsewhere in [94] , was grown and maintained in EGM-2MV medium in the presence of 1 μg/ml doxycycline and 5 μg/ml puromycin . rKSHV . 219 virus was produced from BJAB-rKSHV . 219 cells as described before [49 , 53] . Briefly , 6 X 105 BJAB-rKSHV . 219 cells/ml were inoculated and grown for four to five days in TECHNE spinner flasks ( Cole-Parmer GmbH , Wertheim , Germany ) with 60 rpm agitation in the presence of 2 . 5 μg/ml of anti-human IgM antibody ( #2020–01; SoutherBiotech ) to induce the KSHV lytic cycle . Culture supernatant was collected after low speed centrifugation to remove cells and debris , filtered through a 0 . 45 μm filter and virus particles were concentrated by ultracentrifugation at 15 , 000 rpm for 5 hours in a Type 19 rotor ( Beckman Coulter Inc . , Brea , CA , USA ) . The pellet containing virus was re-suspended in EBM-2MV medium . To determine the infectious virus titer , 2 . 7 x 104 HEK-293 cells were plated per well of a 96 well plate and infected by serial dilution of the virus preparation on the next day . Three days after infection , the number of GFP positive cells was counted and the infectious virus titer was calculated . For inhibitor treatment , 8 x 105 BJAB-rKSHV . 219 cells were plated per well of a 12 well plate and treated with 1μg per ml of anti-human IgM antibody plus the individual inhibitors for the PLCγ ( U73122 ) , MAPK ( U126 ) and PI3K/Akt ( Ly294002 ) pathways at their respective concentrations . Seventy two hours later , cells and culture supernatant were collected separately and the expression level of KSHV lytic proteins as well as the activation level of the respective signaling components was analyzed by western blot and infectious virus release was assessed by titrating the culture supernatant on HEK-293 cells . LECs ( between passage 2 and 5 ) were plated at 5 x 105 cells per well of a 6 well plate and infected with rKSHV . 219 virus at a multiplicity of infection ( MOI ) of 1 by spinoculation ( 450 x g for 30 minutes at 30°C ) in the presence of 10 μg/ml polybrene ( H9268; Sigma-Aldrich , Milwaukee , WI , USA ) . The culture medium was changed after 8 hours . Two days post-infection , the medium was changed to selection medium containing 0 . 25 μg/ml puromycin and stably infected cells ( LEC-rKSHV . 219 ) were maintained under selection for two weeks . For experiments comparing rKSHV . 219 infection in LECs and HuARLT2 cells , HuARLT2 cells were infected in parallel in a similar fashion and maintained with 5 μg/ml puromycin and 1 μg/ml doxycycline . SF-9 cells ( DSMZ No . : ACC 125 ) were grown and maintained in Grace’s insect medium ( Gibco , lifetechnologies , Paisley , UK ) supplemented with 10% FBS standard quality ( PAA Laboratories GmbH , Pasching , Austria ) and 100 units/ml of penicillin/streptomycin ( CytoGen GmbH , Sinn , Germany ) at 28°C . A recombinant baculovirus expressing the KSHV regulator of transcription activation ( RTA ) was kindly provided by J . Vieira ( University of Washington , Seattle WA , USA ) and propagated in SF-9 cells . Culture supernatant containing the baculovirus was used in combination with Sodium butyrate ( B5887; Sigma-Aldrich , Saint Louis , Mo , USA ) for induction of the KSHV lytic cycle in stably infected cells as described before [61] . For the investigation of K15-induced signaling in HEK-293 cells , 2 . 5 x 105 cells were plated per well of a 12 well plate . On the next day , 1μg of the K15 expression vector pFJ-K15 [42] , a kind gift from Jae U . Jung ( University of Southern California , Los Angeles , California , USA ) , or an empty vector control was transfected by using the Fugene 6 transfection reagent ( Promega , Madison , WI , USA ) following the manufacturer’s protocol . Two days after transfection , cells were washed and lysed; and western blot analysis was performed by using phospho-specific antibodies for the respective signaling components . A recombinant KSHV genome in the Bacterial artificial chromosome 36 backbone ( KSHV-Bac36 ) [54] was used to construct deletion mutants KSHV-Bac36ΔK1 and KSHV-Bac36ΔK15 by replacing the genes encoding either K1 ( nucleotide sequence between 104 and 970 ) or K15 ( nucleotide sequence between 135338 and 136900 ) with rpsL/neomycin cassette using Red/ET recombination system ( Gene Bridges GmbH , Heidelberg , Germany ) as described before [48 , 94] . The integrity of all three KSHV-Bac36 constructs was verified by whole genome deep sequencing . Briefly , purified Bac DNA was sheared by sonication . To avoid bias by over-amplification , library preparation was performed using the KAPA real-time library preparation kit ( KAPA Biosystems , Wilmington , MA , USA ) . Paired-end 300 bp reads were generated on a MiSeq sequencer by running the v3 chemistry ( Illumina , San Diego , CA , USA ) . A semi-confluent culture of HEK-293 cells was then transfected with 2 μg of Bac DNA per well of a 6 well plate per each construct ( KSHV-Bac36Wt , KSHV-Bac36ΔK1 or KSHV-Bac36ΔK15 ) using the Fugene 6 transfection reagent . Three days after transfection , cells were re-plated and maintained with a selection medium containing 150 μg/ml of Hygromycin-B ( P06-08100; PAN Biotech GmbH , Aidenbach , Germany ) until a fully selected polyclonal population for each construct ( HEK 293-KSHV-Bac36Wt , HEK 293-KSHV-Bac36ΔK1 or HEK 293-KSHV-Bac36ΔK15 ) was obtained . 2 . 5 X 105 stably transfected cells passaged more than 4 times were then plated per well of a 12 well plate . After 24 hours , the KSHV lytic cycle was induced using a reactivation cocktail containing 1mM Sodium butyrate ( SB ) and SF-9 cell supernatant containing KSHV RTA expressing baculovirus . Forty-eight hours later , the cell culture supernatant was collected and the titer of infectious virus was determined as before ( see section cells , viruses and infection ) , and cells were washed once with 1x PBS and subsequently lysed with lysis buffer for western blot analysis . For experiments involving the PLCγ2-cSH2 domain , 1 μg of the plasmid DNA or its empty vector control was reverse-transfected into 2 . 5 X 105 HEK 293-KSHV-Bac36Wt cells per well of a 12 well plate during plating , 24 hours after transfection cells were treated with reactivation cocktail as before and subsequent steps were performed similarly . The pTriEx-4 vector expressing the isolated PLCγ2-cSH2 domain was a kind gift from Matilda Katan ( University College London , London , UK ) . To establish a stably infected HuARLT2 cell line , recombinant viruses were produced from the stably transfected HEK-293-KSHV-BAC36Wt/ΔK1/ΔK15 cells . Briefly , a semi-confluent monolayer of the respective stably transfected cells was treated with a reactivation cocktail containing 1 . 25 mM Sodium butyrate and SF-9 cell supernatant containing baculovirus expressing KSHV RTA . Cell culture supernatant was collected after 72 hours , filtered through 0 . 45μm filter and concentrated by ultra-centrifugation at 15 , 000 rpm for 5 hours in a Type 19 rotor; the pellet containing the respective viruses was re-suspended in EBM-2MV medium . A semi-confluent monolayer of HuARLT2 cells was then infected with the wild type or either of the deletion mutant viruses by spinoculation ( 450 x g for 30 minutes at 30°C ) in the presence of 5μg/ml polybrene . Three days after infection , 100 μg/ml of Hygromycin-B was added and cells were grown and maintained until a fully selected , stably infected , polyclonal population was obtained for each virus ( HuARLT2-KSHV-Bac36Wt , HuARLT2-KSHV-Bac36ΔK1 or HuARLT2-KSHV-Bac36ΔK15 ) . 5 X 105 stably infected polyclonal cells passaged more than 4 times were plated per well of a 6 well plate and experiments were performed similar to the stably transfected HEK-293 cells described above . First , in order to generate the pSF91-PLCγ2-cSH2-IRES-GFP retroviral vector , the PLCγ2-cSH2 domain containing S- and His-tags at the N-terminus was amplified by PCR from the pTriEx-4-PLCγ2-cSH2 expression vector using Fwd: 5’-CCTGCGGCCGCATGGTACACCATCACCACC-3’ and Rev 5’-GTAGCGGCCGCTTATTTACTCGGGGTCACGGGG-3’ primers . The amplified segment was then inserted into the retroviral vector pSF91-IRES-GFP ( provided by Christopher Baum; Hannover Medical School , Hannover , Germany ) using the Not I restriction site . Retroviruses containing the pSF91-PLCγ2-cSH2-IRES-GFP , pSF91-sK15-IRES-GFP or the empty vector control pSF91-IRES-GFP were produced in HEK-293T cells after calcium-phosphate co-transfection of the respective vector constructs together with packaging plasmids pM57DAW ( gag/pol ) and pRD114 ( env ) as described before [48] . Lentiviral vector pRRL-PPT-SF-PLCγ2-cSH2 [49] or its empty vector pRRL-PPT-SF-GFP was similarly produced in HEK-293T cells by co-transfection with the helper plasmids pMDLGg/p , pRSV-REV and pMD-G . A semi-confluent monolayer of endothelial cells ( LEC , LEC-rKSHV . 219 , HuARLT2 or HuARTL2-rKSHV . 219 ) was then transduced with the indicated retroviral/lentiviral vectors by spinoculation ( 450 x g for 30 minutes at 30°C ) in the presence of 5μg/ml of polybrene . After 48 ( for K15 overexpression ) or 72 hours ( for PLCγ2-cSH2 ) , cell culture supernatant and cells were collected for virus titration and western blot analysis , respectively . Transduction efficiency was estimated by infecting KSHV-negative parental cells ( LEC , HuARLT2 ) and scoring the number of GFP positive cells . Transfection of small interfering RNA ( siRNA ) in BJAB-rKSHV . 219 and HuARLT2-rKSHV cells was performed by using the Neon transfection system ( Invitrogen , Carlsbad , CA , USA ) according to the manufacturer’s instruction . 150 pmol of each individual siRNA was microporated into 4 x 105 BJAB-rKSHV . 219 or 1 x 105 HuARLT2-rKSHV cells and the KSHV lytic cycle was induced the next day by using anti-human IgM antibody or a cocktail of KSHV RTA and Sodium butyrate . Cell culture supernatant and cells were separately collected after 48 hours of lytic induction and infectious virus titer as well as the expression level of KSHV lytic proteins was analyzed subsequently . LECs or LEC-rKSHV . 219 cells were transiently transfected with individual siRNA complexed with Lipofectamin RNAiMAX reagent ( Invitrogen , life technologies , Carlsbad , CA , USA ) as per the manufacturer’s recommendation . Briefly , 150 pmol of each individual siRNA and 6 μl of Lipofectamin RNAiMAX reagent were each diluted in 150 μl of low serum Opti-MEM ( Gibco , lifetechnologies , Paisley , UK ) media and combined . The siRNA: transfection reagent complex was then incubated for 20 minutes at room temperature and applied on to a semi-confluent monolayer of cells in a well of a 6 well plate . Cell culture supernatant and cells were collected 72 hours after transfection for virus titration and western blot analysis respectively . Control/ non-targeting scrambled ( Scr ) siRNA pool #2 D-001206-14-20 , siRNA against KSHV K15 targeting exon 8 #1 ( CAACCACCUUGGCAAUAAU ) , siRNA against KSHV K15 targeting exon 8 #2 ( CAGGCUUGGUCAUGGGUUA ) , siRNA against KSHV K1 ( GUCACUUGUGGUCAGCAUG ) and siRNA against KSHV ORF 26 ( CCAUUGUGCUCGAAUCCAA ) were all purchased from Dharmacon , Thermo Scientific . Cells were lysed with 1x SDS sample buffer ( 62 . 5 mM Tris-HCl pH 6 . 8 , 2% ( W/V ) SDS , 10% ( V/V ) Glycerol , 50 mM DTT , 0 . 01% ( V/V ) β-mercaptoethanol and 0 . 01% ( W/V ) bromophenol blue ) and centrifuged at 20 , 000 x g for 10 minutes at 4°C . Cleared total cellular lysate was then separated by SDS-PAGE ( Note: for the analysis of KSHV K15 protein , samples were not boiled prior to SDS-PAGE ) and transferred onto 0 . 45 μm nitrocellulose membranes ( Amersham , GE healthcare Europe GmbH , Freiburg , Germany ) . Membranes were then blocked with 5% ( W/V ) non-fat milk ( Carl Roth GmbH+Co . KG , Karlsruhe , Germany ) or IgG free albumin ( Carl Roth GmbH&Co . KG , Karlsruhe , Germany ) for 1 hour and probed with an appropriate primary antibody in blocking solution overnight , at 4°C . The next day , membranes were washed three times and incubated with the corresponding horseradish peroxidase ( HRP ) -conjugated secondary antibodies for an hour at room temperature ( RT ) . After a subsequent wash , signal was developed using a standard enhanced chemiluminescence ( ECL ) kit ( #34096; Thermo Scientific , Rockford , IL , USA ) . Primary and secondary antibodies used for western blot are listed above ( section antibodies ) . 7x 107 293-KSHV-BAC36Wt or HuARTL2-rKSHV cells , with or without induction of the KSHV lytic cycle as described above , were collected , washed once with 1xPBS and lysed in 1 ml of ice-cold TNE buffer ( 10mM Tris-HCl pH 7 . 4 , 150mM NaCl , 5mM EDTA ) containing 1% Triton X-100 and protease inhibitors and incubated on ice for 30 minutes . Cellular lysates were then homogenized by passing through a 200 μl pipet tip 10 times and centrifuged at 900 x g at 4°C for 10 minutes . Cleared supernatant was then mixed with 1 ml of 85% sucrose in TNE buffer in a 1:1 ratio and pipetted at the bottom of a 14 X 95 mm ultracentrifuge tube ( Beckman Coulter Inc . , Brea , CA , USA ) , overlaid with first 6 ml of 35% and then 3 . 5 ml of 5% sucrose solutions in TNE buffer and subjected to ultracentrifugation at 200 , 000 x g for 24 hours at 4°C in an SW40 rotor ( Beckman ) . Twelve , fractions of 1 ml were then collected starting from the top of the gradient and analyzed for the desired proteins by western blot . For overexpression experiments , Hela CNX cells were plated on sterile 20 x 20 mm cover slips ( 1 x 105 cells per well of a 6 well plate ) . On the next day , cells were transfected with 2 μg each of the respective plasmid DNA using the Fugene 6 transfection reagent . Forty-eight hours after transfection , cells were washed with 1x PBS , fixed with 4% paraformaldehyde ( PFA ) in PBS for 20 minutes at RT and quenched with 125 mM glycine for 10 minutes at RT . After 3 washes in 1xPBS cells were permeablized with 0 . 2% Triton X-100 for 10 minutes at RT , blocked with 10% FBS in PBS for an hour at 37°C and incubated with the respective primary antibody in blocking solution for an hour at 37°C . After another 3 washes , cells were incubated with the corresponding fluorescently labeled secondary antibody for an hour at 37°C . Plasmid constructs used include pFJ-K15P/M ( encoding either K15P or M ) . For staining endogenous proteins in infected cells , 2 x 105 HuARLT2-rKSHV cells were plated on cover slips as before and the KSHV lytic cycle was induced the next day using RTA and SB . Forty-eight hours after lytic induction , cells were washed in 1xPBS and fixed with 100% ice cold methanol at -20°C . Coverslips were then washed thoroughly with 1xPBS , blocked with 10% FBS in PBS for an hour at 37°C and stained with the respective primary antibodies and secondary antibodies as described above . For staining in LEC-rKSHV cells , samples were processed similar to HuARLT2-rKSHV cells but without induction of the lytic cycle . Images were acquired with a Leica TCS SP2 AOBS confocal microscope ( Leica Microsystems , Wetzlar , Germany ) or with a Zeiss Axio Observer . Z1 epi-fluorescent microscope ( Carl Zeiss Iberia , Madrid , Spain ) . To embed HuARLT2 and HuRLT2-rKSHV cells in paraffin we used a previously described protocol [102] . Briefly , cells were harvested , pelleted by centrifugation and fixed in 4% PFA for 20 minutes at room temperature . After 3 washes in 1XPBS the cell pellets were re-suspended in 500 μl of 1 . 5% ( W/V ) melted ( 50°C ) agarose basic ( Aplichem GmbH , Darmstadt , Germany ) in PBS and quickly transferred to a capped inverted 2ml Eppendorf tube . The agar-cell suspension was left on ice to solidify and gently extracted by opening the tube cup . The agar cylinder was then embedded in paraffin after processing as for histopathology specimens . The HuARLT2 and HuARLT2-rKSHV cell blocks were then used for testing the suitability of the rat anti-K15 mAb ( clone 18E5 ) for K15 staining in paraffin embedded tissue and as a negative and positive control during KS tissue staining . Paraffin embedded cell blocks and biopsy samples obtained from HIV positive KS patients were sectioned into 5μm thick slices and deparaffinized in Roti-Histol ( Carl Roth GmbH , Karlsruhe , Germany ) . After rehydration in descending ethanol dilution series , antigen retrieval was performed using a citrate-based antigen unmasking solution ( H-3300 , Vector Laboratories; 15 minutes , 100°C ) . Endogenous peroxidases were blocked with 3% H2O2/H2O for 10 minutes and staining was performed using the rat anti-K15 mAb ( clone 18E5 ) [49] as primary antibody followed by a biotinylated goat anti-rat IgG ( 1:250 , 111065033; Dianova ) secondary antibody . The signal of the rat anti-K15 mAb was amplified using the Tyramide Signal Amplification ( TSA ) system ( NEL702001KT; Perkin Elmer ) and nuclei were counterstained with 4' , 6-diamidino-2-phenylindole ( DAPI ) . Images were acquired using a Leica DM6000 microscope with Leica DFC350FX digital camera and contrast and brightness was adjusted afterwards in Adobe Photoshop CS4 .	Both the latent and lytic replication phases of the KSHV life cycle are thought to contribute to its persistence and pathogenesis . The non-structural signaling membrane protein K15 is involved in the angiogenic and invasive properties of KSHV-infected endothelial cells . Here we show that the K15 protein is required for virus replication , early viral gene expression and virus production through its activation of the cellular signaling pathways PLCγ1 and Erk 1/2 . K15 is abundantly expressed in KSHV-infected lymphatic endothelial cells ( LECs ) and contributes to KSHV-induced endothelial spindle cell formation . The abundant K15 protein expression observed in LECs is also observed in KS tumors . We also show that it may be possible to target K15 in order to intervene therapeutically with KSHV lytic replication and pathogenesis .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "phosphorylation", "transfection", "medicine", "and", "health", "sciences", "pathology", "and", "laboratory", "medicine", "biopsy", "gene", "regulation", "endothelial", "cells", "pathogens", "microbiology", "surgical", "and", "invasive", "medical", "procedures", "epitheli...	2017	The Kaposi's sarcoma-associated herpesvirus (KSHV) non-structural membrane protein K15 is required for viral lytic replication and may represent a therapeutic target
In a landmark paper , Nadeau and Taylor [18] formulated the random breakage model ( RBM ) of chromosome evolution that postulates that there are no rearrangement hotspots in the human genome . In the next two decades , numerous studies with progressively increasing levels of resolution made RBM the de facto theory of chromosome evolution . Despite the fact that RBM had prophetic prediction power , it was recently refuted by Pevzner and Tesler [4] , who introduced the fragile breakage model ( FBM ) , postulating that the human genome is a mosaic of solid regions ( with low propensity for rearrangements ) and fragile regions ( rearrangement hotspots ) . However , the rebuttal of RBM caused a controversy and led to a split among researchers studying genome evolution . In particular , it remains unclear whether some complex rearrangements ( e . g . , transpositions ) can create an appearance of rearrangement hotspots . We contribute to the ongoing debate by analyzing multi-break rearrangements that break a genome into multiple fragments and further glue them together in a new order . In particular , we demonstrate that ( 1 ) even if transpositions were a dominant force in mammalian evolution , the arguments in favor of FBM still stand , and ( 2 ) the “gene deletion” argument against FBM is flawed . In 1970 , Susumu Ohno came up with two fundamental models of chromosome evolution that were subject to many controversies in the last 35 years [1] . One of them ( the whole genome duplication model ) was first met with skepticism and only recently was proven to be correct [2 , 3] . The other , the random breakage model ( RBM ) , had a very different fate . It was embraced by biologists from the very beginning ( due to its prophetic prediction power ) and only recently was refuted by Pevzner and Tesler [4] using a theorem from [5] . However , the rebuttal of RBM caused a controversy and shortly after [4] was published Sankoff and Trinh [6 , 7] gave a rebuttal of the rebuttal of RBM . Rearrangements are genomic “earthquakes” that change the chromosomal architectures . The fundamental question in molecular evolution is whether there exist “chromosomal faults” ( rearrangement hotspots ) where rearrangements are happening over and over again . RBM postulates that rearrangements are “random , ” and thus there are no rearrangement hotspots in mammalian genomes . For the sake of completeness , we give a simple version of both the Pevzner-Tesler and Sankoff-Trinh arguments . Shortly after the human and mouse genomes were sequenced , Pevzner and Tesler [4] argued that if ( 1 ) the human–mouse synteny blocks are constructed correctly , and ( 2 ) chromosomal architectures mainly evolve by the “standard” rearrangement operations ( reversals , translocations , fissions , and fusions ) , then every evolutionary scenario for transforming the mouse genome into the human genome must have a very large number of breakpoint re-uses . This result implies that the same regions of the genome are being broken over and over again in the course evolution ( rearrangement hotspots ) , a contradiction to RBM ( note that high breakpoint re-use by itself does not invalidate RBM; however , a combination of high breakpoint re-use with scan statistics of the human–mouse breakpoint arrangements invalidates RBM; see Text S1 ) . Consequently , Pevzner and Tesler [4] suggested an alternative fragile breakage model ( FBM ) of chromosome evolution that was later supported by Murphy et al . [8] . Recent studies further argued for existence of fragile regions ( rearrangement hotspots ) in mammalian genomes [9–17] . These results are in conflict with the classical Nadeau and Taylor [18] analysis of RBM that implies that there are no rearrangement hotspots in the human genome . In the next two decades , numerous studies with progressively increasing levels of resolution made RBM the de facto theory of chromosome evolution . As a result , the Nadeau-Taylor analysis was until recently viewed as among the most significant results in “ . . . the history and development of the mouse as a research tool” [19] . The paper [4] challenged this view and was quickly followed by other studies questioning the RBM . For example , Kikuta et al . [16] recently wrote “ . . . the results in this study suggest that the Nadeau and Taylor hypothesis is not plausible for the explanation of synteny in general . ” Sankoff and Trinh [6 , 7] did not question the validity of combinatorial arguments against RBM in [4] , but instead argued that the synteny block generation algorithm is parameter-dependent and that question ( 1 ) above is more subtle than it may look at first glance . Sankoff and Trinh [6] emphasized how important it is to generate synteny blocks by constructing a series of random rearrangements that create an appearance of breakpoint re-use . They generated a series of random rearrangements according to RBM ( i . e . , no rearrangement hotspots ) , computed the resulting synteny blocks , applied the same arguments as in [4] , and came to the conclusion that the rearrangement hotspots exist . These hotspots , however , are clearly artifacts of synteny block generation rather than real hotspots , since the simulation in [6] followed RBM . Recently , Peng et al . [20] re-examined Sankoff and Trinh's arguments and demonstrated that Sankoff and Trinh fell victim to their inaccurate synteny block generation algorithm . Peng et al . [20] further demonstrated that if Sankoff and Trinh had fixed these problems and chosen realistic parameters , their arguments against [4] would disappear . Sankoff recently acknowledged the flaw in [6] ( see [21] ) , and it seems that condition ( 1 ) is not controversial anymore . However , Sankoff still appeared reluctant to acknowledge the validity of the Pevzner-Tesler rebuttal of RBM , this time arguing that condition ( 2 ) above may also be violated in mammalian evolution . This led to a split among researchers studying chromosome evolution: while most recent studies support the existence of rearrangement hotspots [9–14 , 16 , 17] , others feel that further analysis is needed to resolve the validity of RBM [22] . Indeed , since the mathematical theory used to refute RBM does not cover more complex rearrangement operations ( like transpositions ) , the arguments in [4] do not apply for the case when transpositions are frequent . In this paper , we develop a theory for analyzing complex rearrangements ( including transpositions ) and demonstrate that even if transpositions were a dominant evolutionary force , there are still rearrangement hotspots in mammalian evolution . This results in a rebuttal of the rebuttal [21] of the rebuttal [20] of the rebuttal [6 , 7] of the rebuttal [4] of RBM . The standard rearrangement operations ( i . e . , reversals , translocations , fusions , fissions ) can be modelled by making two breaks in a genome and gluing the resulting fragments in a new order . One can imagine a hypothetical k-break rearrangement operation that makes k breaks in a genome and further glues the resulting pieces in a new order . In particular , the human genome can be modelled as the mouse genome broken into ≈280 pieces that are glued together in the “mouse” order . Sankoff [21] is correct in stating that the rebuttal of RBM is not applicable if there was a significant presence of k-break rearrangements for large k ( in fact , it was acknowledged in [4] ) . However , rearrangements are rare evolutionary events and , starting from the classical Dobzhansky and Sturtevant studies of Drosophila , most biologists believe that k-break rearrangements are unlikely for k > 3 , and relatively rare for k = 3 ( at least in mammalian evolution ) . Indeed , biophysical limitations and selective constraints are already severe for k = 2 , let alone for k > 2 . However , 3-break rearrangements ( e . g . , transpositions ) undoubtedly happen in evolution , although it is still unclear how frequent they are in mammalian evolution . Also , in radiation biology , chromosome aberrations for k > 2 ( indicative of chromosome damage rather than evolutionary viable variations ) may be more common ( e . g . , complex rearrangements in irradiated human lymphocytes [23–26] ) . Thus , both the existing controversy about RBM and radiation/cancer biology call for studies of k-break rearrangements for k > 2 . We recently proved the duality theorem for the k-break distance between multichromosomal genomes , and showed how to compute it [27] . In this paper , we focus on the case k = 3 ( the most relevant case in evolutionary studies ) and show that even if 3-break rearrangements were frequent , the Pevzner-Tesler argument against RBM still stands . We further discuss the claim [7 , 21] that deletion of short synteny blocks may also create an appearance of high breakpoint re-use , an argument against FBM . We invalidate this argument by showing that deletion of short blocks does not lead to increase in breakpoint re-use under the realistic choice of parameters . We start our analysis with circular genomes ( i . e . , genomes consisting of circular chromosomes ) . We will find it convenient to represent a circular chromosome with genes x1 , . . . , xn as a cycle ( Figure 1 ) composed of n directed labeled edges ( corresponding to genes ) and n undirected unlabeled edges ( connecting adjacent genes ) . The directions of the edges correspond to signs ( strand ) of the genes . We label the tail and head of a directed edge xi as xit and xih , respectively . Vertex xit is called the obverse of vertex xih , and vice versa . Vertices in a chromosome connected by an undirected edge are called adjacent . We represent a genome as a graph consisting of disjoint cycles ( one for each chromosome ) . The edges in each cycle alternate between two colors: one color ( usually black or gray ) is reserved for undirected edges , and the other color ( traditionally called “obverse” and portrayed by dashed lines in Figure 1 ) is reserved for directed edges . We do not explicitly show the directions of obverse edges since they are defined by superscripts “t” and “h” ( Figure 1 ) . Let P be a genome represented as a collection of alternating black-obverse cycles ( a cycle is alternating if the colors of its edges alternate ) . For any two black edges ( u , v ) and ( x , y ) in the genome ( graph ) P , we define a 2-break rearrangement as replacement of these edges with either a pair of edges ( u , x ) , ( v , y ) , or a pair of edges ( u , y ) , ( v , x ) ( Figure 2 ) . 2-Breaks correspond to standard rearrangement operations of reversals ( Figure 2A ) , fissions ( Figure 2B ) , or fusions/translocations ( Figure 2C ) . This definition of elementary rearrangement operations follows the standard definitions of reversals , translocations , fissions , and fusions for the case of circular chromosomes . For circular chromosomes , fusions and translocations are not distinguishable; i . e . , every fusion of circular chromosomes can be viewed as a translocation and vice versa . The 2-break rearrangements can be generalized as follows . Given k black edges forming a matching ( i . e . , a vertex-disjoint set of edges ) on 2k vertices , define a k-break as replacement of these edges with a set of k black edges forming another matching on the same set of 2k vertices . Note that a 2-break is a particular case of a 3-break ( as well as of a k-break for k > 3 ) , in which case only two edges are replaced and the third one remains the same . Let P and Q be two signed genomes on the same set of genes . The breakpoint graph G ( P , Q ) is defined on the set of vertices V = {xt , xh \| x ∈ } with edges of three colors: obverse , black , and gray ( Figure 1 ) . Edges of each color form a matching on V: obverse matching ( pairs of obverse vertices ) , black matching ( adjacent vertices in P ) , and gray matching ( adjacent vertices in Q ) . Every pair of matchings forms a collection of alternating cycles in G ( P , Q ) called black-gray , black-obverse , and gray-obverse cycles , respectively . The chromosomes of the genome P ( respectively , Q ) can be read along black-obverse ( respectively , gray-obverse ) cycles . The black-gray cycles in the breakpoint graph play an important role in analyzing rearrangements [28] ( see Chapter 10 of [29] for background information on genome rearrangements ) . The k-break distance dk ( P , Q ) between circular genomes P and Q is defined as the minimum number of k-breaks required to transform one genome into the other . Every k-break in the genome P corresponds to a transformation of the breakpoint graph G ( P , Q ) . Since the breakpoint graph of two identical genomes is a collection of trivial black-gray cycles with one black and one gray edges ( the identity breakpoint graph ) , the problem of transforming the genome P into the genome Q by k-breaks can be formulated as the problem of transforming the breakpoint graph G ( P , Q ) into the identity breakpoint graph G ( Q , Q ) . Different from the genomic distance problem [5 , 30 , 31] ( for linear multichromosomal genomes ) , the 2-break distance problem for circular multichromosomal genomes has a trivial solution ( first given in [32] in a slightly different context ) . For the sake of completeness , we reproduce a proof from [33]: Theorem 1 . The 2-break distance between circular genomes P and Q is \|P\| − c ( P , Q ) where c ( P , Q ) is the number of black-gray cycles in G ( P , Q ) . Proof . It is easy to see that every nontrivial black-gray cycle in the breakpoint graph G ( P , Q ) can be split into two by a 2-break , implying that d2 ( P , Q ) ≤ \|P\| − c ( P , Q ) . Since every 2-break adds two new edges , it can create at most two new black-gray cycles . On the other hand , since every 2-break removes two old edges , it should remove at least one old black-gray cycle . Hence , no 2-break can increase the number of black-gray cycles by more than one , implying that d2 ( P , Q ) ≥ \|P\| − c ( P , Q ) . Therefore , d2 ( P , Q ) = \|P\| − c ( P , Q ) . Q . E . D . While 2-breaks correspond to standard rearrangements , 3-breaks add transposition-like operations ( transpositions and inverted transpositions ) as well as three-way fissions to the set of rearrangements ( Figure 3 ) . Different from standard rearrangements ( modeled as 2-breaks ) , transpositions introduce three breaks in the genome , making them notoriously difficult to analyze . Computing the minimum number of transpositions transforming one genome into another is called “sorting by transpositions . ” A number of researchers considered transpositions in conjunction with other rearrangement operations [34–40] . Despite many studies , the complexity of sorting by transpositions remains unknown [41–45] . Let codd ( P , Q ) be the number of black-gray cycles in the breakpoint graph G ( P , Q ) with an odd number of black edges ( odd cycles ) . Theorem 2 . The 3-break distance between circular genomes P and Q is ( \|P\| − codd ( P , Q ) ) / 2 . Proof . It is easy to see that as soon as there is a nontrivial black-gray odd cycle in the breakpoint graph G ( P , Q ) , it can be split into three odd cycles by a 3-break , thus increasing the number of odd cycles by two . On the other hand , if there exists a black-gray even cycle , it can be split into two odd cycles , thus again increasing the number of odd cycles by two . Therefore , there exists a series of ( \|P\| − codd ( P , Q ) ) / 2 3-breaks transforming G ( P , Q ) into the identity breakpoint graph , implying that d3 ( P , Q ) ≤ ( \|P\| − codd ( P , Q ) ) / 2 . On the other hand , since no 3-break can increase the number of black-gray cycles by more than two , we have d3 ( P , Q ) ≥ ( \|P\| − codd ( P , Q ) ) / 2 . Therefore , d3 ( P , Q ) = ( \|P\| − codd ( P , Q ) ) / 2 . Q . E . D . For the sake of completeness , below we formulate the duality theorem for the k-break distance for an arbitrary k from [27] . A subset of cycles in the breakpoint graph G ( P , Q ) is called breakable if the total number of black edges in these cycles equals 1 modulo ( k − 1 ) . Let sk ( P , Q ) be the maximum number of disjoint breakable subsets in G ( P , Q ) . For example , for k = 3 , every odd cycle forms a breakable subset and every breakable subset must contain at least one odd cycle , implying that s3 ( P , Q ) = codd ( P , Q ) . Theorem 3 . The k-break distance between circular genomes P and Q is . Sankoff summarized arguments against FBM in the following sentence [21]: Below we consider the “other processes” argument . The “noise in block construction” argument consists of two parts: synteny block generation and gene deletion . The flaw in the first argument was revealed in [20] . The second argument ( “gene deletion” ) is analyzed after the “other processes” argument . In this paper , we study transformations between the human genome H and the mouse genome M with 3-breaks , using the 281 synteny blocks from [46] and assume that all chromosomes are circular . While analyzing linear chromosomes would be more adequate than analyzing their circularized versions , it poses additional algorithmic challenges that remain beyond the scope of this paper . The related paper [47] addressed these challenges and demonstrated that switching from linear to circular chromosomes does not lead to significant changes in the multi-break distance . The breakpoint graph G ( H , M ) contains 35 black-gray cycles , including three odd black-gray cycles , implying that d2 ( H , M ) = 246 ( Theorem 1 ) and d3 ( H , M ) = 139 ( Theorem 2 ) . If each of 139 3-breaks on a shortest evolutionary path from H to M made three breaks , it would imply that there were 139 × 3 – 281 = 136 breakpoint re-uses ( for this particular evolutionary path ) , resulting in the breakpoint re-use rate 1 . 48 ( see Peng et al . [20] ) . While this is a high breakpoint re-use rate ( inconsistent with RBM and the scan statistics ) , this estimate relies on the assumption that each 3-break on the evolutionary path from H to M makes three breaks ( complete 3-breaks ) . In reality , some 3-breaks can make two breaks ( incomplete 3-breaks ) as 2-breaks are particular cases of 3-breaks , reducing the estimate for the number of breakpoint re-uses . Moreover , the minimum number of breakpoint re-uses may be achieved on a suboptimal evolutionary path from H to M . The rebuttal of RBM raises a question about finding a transformation of H into M by 3-breaks that makes the minimal number of individual breaks . The following theorem shows that there exists a series of 3-breaks that makes the minimum number of breaks while transforming P into Q . Theorem 4 . Any series of m k-breaks transforming a circular genome P into a circular genome Q makes at least m + d2 ( P , Q ) breaks . Moreover , there exists a series of d3 ( P , Q ) 3-breaks transforming P into Q that makes d3 ( P , Q ) + d2 ( P , Q ) breaks . Proof . For each k-break operation , let Δ ( cycles ) be the increase in the number of cycles and Δ ( breaks ) be the increase in the number of breaks . It is easy to see that Δ ( cycles ) ≤ Δ ( breaks ) − 1 . Summing up over a series of m k-breaks transforming P into Q , we have \|P\| − c ( P , Q ) ≤ b − m , where b is the total number of breaks made in the series . Therefore , b ≥ \|P\| − c ( P , Q ) + m = d2 ( P , Q ) + m . Consider a shortest series of complete 3-breaks transforming every odd black-gray cycle into a trivial cycle and every even black-gray cycle into trivial cycles and a single cycle with two black edges . This series consists of d3 ( P , Q ) − ceven ( P , Q ) 3-breaks and results in ceven ( P , Q ) cycles with two black edges that can be transformed into trivial cycles with a series of ceven ( P , Q ) incomplete 3-breaks ( i . e . , 2-breaks ) . The total number of 3-breaks in this transformation is d3 ( P , Q ) , and they make 3 ( d3 ( P , Q ) − ceven ( P , Q ) ) + 2ceven ( P , Q ) = 3d3 ( P , Q ) − ceven ( P , Q ) = d3 ( P , Q ) + d2 ( P , Q ) breaks overall . Q . E . D . Corollary 5 . Every transformation between the circularized human genome H and mouse genome M by 3-breaks requires at least 104 breakpoint re-uses ( implying that there exist rearrangement hotspots in the human genome ) . Proof . Any transformation of H into M requires at least d3 ( H , M ) + d2 ( H , M ) = 139 + 246 = 385 breaks . Since there are 281 breakpoints between the human and mouse genomes , it implies that there were at least 385 − 281 = 104 breakpoint re-uses on the evolutionary path from human to mouse , resulting in breakpoint re-use rate 1 . 37 . This is still higher than the expected breakpoint re-use rate of RBM as computed by scan statistics ( see [4] and simulations in the next section ) . It provides an argument against RBM not only for k = 2 but also for k = 3 and invalidates arguments from [21] in the case k = 3 ( see also [47] ) . Since k-breaks for k > 3 were never reported in previous evolutionary studies , it is unlikely that they significantly affect our conclusions . Q . E . D . Theorem 4 implies that any transformation of the human genome H into the mouse genome M with 2-breaks makes at least d2 ( H , M ) + d2 ( H , M ) = 246 + 246 = 492 breaks , while any transformation of H into M with 3-breaks makes at least d3 ( H , M ) + d2 ( H , M ) = 139 + 246 = 385 breaks . Below , we show how the number of breaks made in a series of 3-breaks depends on the number of complete 3-breaks in this series . Theorem 6 . For any series of m 3-breaks transforming a genome P into a genome Q with t complete 3-breaks , m ≥ max{d2 ( P , Q ) − t , d3 ( P , Q ) } . Moreover , there exists a series of max{d2 ( P , Q ) − t , d3 ( P , Q ) } 3-breaks transforming P into Q with at most t complete 3-breaks . Proof . Since k-break can increase the number of cycles in the breakpoint graph by at most k − 1 , a series with t complete 3-breaks and m − t incomplete 3-breaks ( i . e . , 2-breaks ) can increase the number of cycles by at most 2t + ( m − t ) = m + t . If it transforms the genome P into the genome Q , then m + t ≥ \|P\| − c ( P , Q ) = d2 ( P , Q ) . Therefore , m ≥ d2 ( P , Q ) − t . Consider a series of complete 3-breaks , transforming every black-gray cycle with q ≥ 3 black edges into two trivial cycles and a cycle with q − 2 black edges . Note that such a series may have at most d3 ( P , Q ) − ceven ( P , Q ) 3-breaks ( the longest possible series results in ceven ( P , Q ) cycles with two black edges and \|P\| − ceven ( P , Q ) trivial cycles ) . Since every such 3-break increases the number of cycles by two , a series of q = min {t , d3 ( P , Q ) − ceven ( P , Q ) } such 3-breaks results in c ( P , Q ) + 2q cycles . These cycles can be transformed into trivial cycles with a series of \|P\| − ( c ( P , Q ) + 2q ) = d2 ( P , Q ) − 2q 2-breaks . The total number of 3-breaks and 2-breaks in this transformation is q + d2 ( P , Q ) − 2q = d2 ( P , Q ) − min {t , d3 ( P , Q ) − ceven ( P , Q ) } = max {d2 ( P , Q ) − t , d3 ( P , Q ) } . Q . E . D . Theorems 4 and 6 imply: Corollary 7 . Any series of 3-breaks with t complete 3-breaks , transforming a genome P into a genome Q , makes at least d2 ( P , Q ) + max {d2 ( P , Q ) − t , d3 ( P , Q ) } breaks . In particular , any such series of 3-breaks with t ≤ d2 ( P , Q ) − d3 ( P , Q ) complete 3-breaks makes at least 2d2 ( P , Q ) − t breaks . Corollary 7 gives the lower bound for the breakpoint re-use rate as a function of the number of complete 3-breaks ( i . e . , transpositions and three-way fissions ) in a series of 3-breaks transforming one genome into the other . For the human genome H and mouse genome M , this lower bound is shown in Figure 4 . Corollaries 5 and 7 address only the case of circularized chromosomes and further analysis is needed to extend it to the case of linear chromosomes ( see [47] ) . Recently , Bergeron et al . [48] described another promising approach to analyzing both circular and linear chromosomes ( using double-cut-and-join operations proposed in [32] ) that also opens a possibility to obtain the breakpoint re-use estimates for linear genomes . However , the above estimate is based on the extreme assumption that certain 3-breaks ( transpositions and three-way fissions/fusions ) represent the dominant rearrangements while reversals and translocations are extremely rare ( contrary to the existing view ) . We emphasize that we do not share the point of view that genomes mainly evolve by transpositions and three-way fissions/fusions , and that we analyzed this assumption only to refute the arguments against FBM . A more realistic analysis of 3-breaks leads to a much higher estimate of the breakpoint re-use ( see Figure 4 ) . The papers [7 , 21] claim that deletion of some synteny blocks in [4] may create an appearance of breakpoint re-use even if there was no breakpoint re-use at all . Below , we show that this argument suffers from the same problem ( unrealistic parameter choice ) that was revealed in [20] . Sankoff and Trinh acknowledged the problem with unrealistic parameter choice in [6] in application to synteny block generation: Despite the importance of choosing realistic parameters , the paper [7] has no discussion of parameters that are relevant to the human–mouse analysis . Below , we study the deletion process , reproduce simulations in [7] , and show that if Sankoff and Trinh used realistic parameters they would confirm ( rather than refute ) FBM . Sankoff and Trinh [7] show that deletion of a large number of elements ( genes ) from a permutation produced by “random” rearrangements would produce a permutation with large breakpoint re-use ( Figure 5A ) . This is not surprising—the only question is what is the realistic number of deleted elements ( we use the term “deleted elements” instead of the term “deleted blocks” in [7] to avoid confusion with synteny blocks ) to match the reality of human–mouse comparison . If this number does not match the reality of human–mouse comparison , then the observation that the breakpoint re-use increases with element deletions turns into a purely mathematical statement that we are not debating and that is irrelevant to the conclusion in [4] about breakpoint re-use in mammalian evolution . For example , if only 20% of all elements are deleted , then Figure 5A ( reproduced from [7] ) supports rather than rejects FBM ( low breakpoint re-use at θ = 0 . 2 ) . However , if one deletes 50% of all elements , the breakpoint re-use becomes rather high , and the Sankoff-Trinh argument against [4] stands . This observation seems to imply that a long-standing debate must be easy to resolve—one should compute the number of deleted elements ( genes ? ) in the human genome and consult Figure 5A . Unfortunately , since it is unclear how one can estimate the number of deleted elements , Figure 5A cannot refute or validate FBM . The inability to connect Figure 5A with the realities of human–mouse genomic architectures is only part of the problem with the simulations in [7] . Another problem is the parameter choice; for example , it is not clear why the parameter n = 100 in Figure 5A is chosen , since the number of rearrangements between the human and mouse genomes clearly exceeds 100 . Moreover , most plots for n = 1 , 000 in Figure 5A ( particularly those with high breakpoint re-use ) produce synteny blocks that do not even fit RBM , which [7] is arguing for . Figure 6A shows that the distribution of synteny block sizes ( for n = 1 , 000 elements and m = 320 rearrangements , averaged over 100 simulations ) is quite different from the exponential distribution characteristic for RBM . In fact , ≈400 out of 640 resulting synteny blocks have ( minimal ) size 1 ( compare with Figure 1 , middle panel , in [4] that is used as an argument against RBM ) . One can argue that Sankoff and Trinh [7] are only interested in reversal distance of the resulting synteny block arrangements , and the sizes of the synteny blocks do not matter . While this argument is correct for θ = 0 , it becomes flawed for θ > 0 , since the results of the deletion process are highly dependent on the distribution of the synteny block sizes . Short synteny blocks ( of size 1 ) are “easy” to delete and the unrealistically high proportion of such blocks in the Sankoff-Trinh simulation makes the plot in Figure 5A look quite different from what one would expect if the simulations would follow RBM . This particular deficiency of the Sankoff-Trinh simulations is easy to fix: one should simply increase the granularity ( i . e . , increase n ) to better model RBM . Figure 5B shows the results of simulations with n = 25 , 000 ( rough estimate of the number of genes in mammalian genomes ) and m = 320 , while Figure 6B shows that the distribution of the sizes of the synteny blocks ( for these parameters ) fits the exponential curve and is consistent with RBM ) . If Sankoff and Trinh presented a ( more realistic ) plot in Figure 5B in their paper , they would likely confirm rather than refute FBM—indeed , one needs to delete more than 90% of genes ( elements ) to see significant breakpoint re-use . The sequenced mammalian genomes do not show any evidence of such extreme gene loss . However , although the plot in Figure 5B shows small breakpoint re-use ( for any realistic choice of parameters ) , we prefer not to use it as a counter-argument against the Sankoff-Trinh argument since ( similarly to [7] ) we do not know what is the best way to choose the parameters ( e . g . , the number of rearrangements ) matching the realities of the human–mouse analysis . This problem did not escape the attention of Pevzner and Tesler [4]; in fact , they implicitly constructed an analog of Figure 5 and even described the scan statistics to analyze it . The only difference is that instead of parameter θ ( the number of deleted blocks ) , they used different parameters σ ( the minimum size of a synteny block , all smaller ones are deleted ) and γ ( the total size of deleted synteny blocks ) . Although all these parameters seem to be interchangeable , there is a big difference between them when it comes to the real human–mouse comparison: σ and γ , different from θ , are easy to estimate . Indeed , the key conclusion of [4] is that the large synteny blocks ( >1 Mb ) cover almost the entire genome ( 95% ) , while breakpoint regions ( where elusive short synteny blocks hide ) cover only ≈5% of the human genome . We emphasize that synteny blocks in [4] are hardly controversial since all follow-up studies with different synteny block generation algorithms came up with roughly the same set of blocks . These blocks are further confirmed by a large number of genes ( in the same conserved order with few micro-rearrangements ) . Figure 7 describes a simulation similar to the Sankoff-Trinh simulations , but in σ rather that in θ coordinates ( all synteny blocks shorter than σ × GenomeLength are deleted ) . It is as good as Figure 5 for refuting FBM since the breakpoint re-use eventually increases when σ increases . However , one can see that breakpoint re-use is low at σ = 0 . 00033 ( corresponding to 1 Mb , the maximal size of deleted blocks in [4] ) and it is nowhere close to the observed human–mouse breakpoint re-use for any realistic values of parameter σ . In 100 , 000 simulations , the breakpoint re-use never reached the value 1 . 37 specified in Corollary 5 , indicating that reaching such high breakpoint re-use is highly unlikely in the RBM framework . We admit that since the choice of 1 Mb ( σ = 0 . 00033 ) as the threshold for the deleting short synteny blocks is somewhat arbitrary , one can argue that the breakpoint re-use becomes large when σ exceeds 0 . 00150 ( ≈5 Mb ) . Therefore , one can argue that if Pevzner and Tesler [4] had chosen 5 Mb as the threshold for synteny block deletion , they would fall into the trap described in [7] . Below we explain a flaw with this counter-argument . Indeed , in this case all synteny blocks shorter than 5 Mb would have to be deleted , and thus would have to be declared to be the breakpoint regions rather than the synteny blocks ( for σ = 0 . 00150 ) . It would result in a genome with an extremely high proportion of breakpoint regions ( as opposed to 5% reported in [4] for 1 Mb threshold ) . Application of scan statistics to such a genome would not reveal any surprising breakpoint clustering , and the conclusion that evolution follows RBM would be confirmed—therefore , in this case [4] would never be written ( let alone , published ) . This flawed “counter-argument” illustrates the key problem with [7]: it never took into account or even commented on the 5%–95% split between breakpoint regions and synteny blocks in the human and mouse genomes , the key argument against RBM . The rebuttal of RBM is based on both arguments ( breakpoint-re-use and 5%–95% split ) and [4] never claimed that breakpoint re-use alone invalidates RBM . Therefore , the rebuttal of [4] based solely on the breakpoint re-use argument ( as in [7] ) is flawed . We emphasize that Figures 4 and 6 represent rather similar simulations and differ mainly in the choice of parameters for representing the results of these simulations ( θ versus σ ) . There is no intrinsic advantage in choosing one simulation over another; the only difference is that one of these simulations ( θ ) is difficult to connect to the realities of the human–mouse analysis , while the other one ( σ ) has a clear interpretation . We also remark that for typical parameters , the Sankoff-Trinh “gene deletion” process is not dramatically different from the Pevzner-Tesler “synteny block deletion” process . For example , even if half of all genes are deleted ( θ = 0 . 5 ) , the Sankoff-Trinh simulation deletes ( on average ) 1/2i blocks of size i; i . e . , removes mainly short blocks as in [46] . Of course , there is no one-to-one correspondence between the Sankoff-Trinh and Pevzner-Tesler deletion processes: some blocks shorter than the threshold are retained , and some blocks larger than the threshold are deleted in the Sankoff-Trinh simulation . To better compare the Sankoff-Trinh gene deletion process with the synteny block deletion process , one may switch to parameter γ , the proportion of the total size of deleted blocks ( this parameter can be directly computed for the Sankoff-Trinh simulation ) . For γ = 0 . 05 corresponding to the 5% proportion of the breakpoint regions in human–mouse comparison , the breakpoint re-use is small ( ≈1 . 2 ) . The fact that it becomes as large as 1 . 9 for γ = 0 . 3 is irrelevant , since it does not reflect the reality of human–mouse comparison: indeed , we do not find that 30% of the human genome is formed by breakpoint regions that do not exhibit similarity with other mammalian genomes and have few orthologous genes . Again , if the human–mouse analysis in [4] revealed that the breakpoint regions account for a third of the genome , the paper [4] would never be written . Nadeau and Taylor [18] proposed RBM based on a single observation: the exponential distribution of human–mouse synteny block sizes . There is no doubt that jumping to this conclusion was not fully justified mathematically: there are many other models ( e . g . , FBM ) that lead to the same exponential distribution of the sizes of the “visible” synteny block . Apart from the 20-year old legacy , human and mouse genomes provide no evidence that would allow one to claim that RBM is correct and FBM is not; indeed , all statistical support for RBM immediately translates into statistical support for FBM . From this perspective , it is not clear how one can favor RBM over FBM without a single piece of evidence that holds for RBM but is violated for FBM . Pevzner and Tesler [4] presented the first evidence that RBM is in conflict with mammalian genomic architectures . Sankoff and Trinh [6 , 7] argued that the Pevzner-Tesler arguments against RBM are flawed . We acknowledge the important contribution of [6] in raising awareness that there are many subtle details and parameters in rearrangement analysis . At the same time , we emphasize that [6] did not present any arguments against FBM and did not connect their simulations with the realities of mammalian genomes . Perusal of the UCSC Genome Browser ( http://genome . ucsc . edu ) reveals large numbers of short adjacent regions corresponding to parts of several chromosomes [49] . For example , the antibody regions in mammalian genomes show signs of multiple recurrent rearrangements . However , until recently , it remained unclear whether these regions reflect genome rearrangements ( relevant to this paper ) , or duplications/assembly errors/alignment artifacts [50] . While previous studies attributed the fragile regions to high repeat density , high recombination rate , or pairs of tRNA genes , it remained unclear how to distinguish “true” short synteny blocks from computational artifacts [50] . When RBM was formalized in 1984 [18] , the short blocks in the human–mouse comparison were not available . By 2003 , many short blocks were found , but it was not possible to decide which of them ( if any ) were real synteny blocks and which represented algorithmic or statistical artifacts . Acknowledging that these newly found short blocks were unreliable , Pevzner and Tesler [4] did not use any of them to refute RBM . Instead , they proved that such short blocks exist ( without finding them ) and predicted that the distribution of the synteny block sizes looks like Figure 8A ( with an abnormally high bar corresponding to “hidden” short blocks ) . Recently , Ma et al . [22] finally revealed some short synteny blocks via the analysis of multiple mammalian genomes . Their distribution ( Figure 8B ) is remarkably similar to the distribution predicted by Pevzner and Tesler in 2003 [4] ( Figure 8A ) . The paper [4] has been cited in many biological papers , and we feel it is important to resolve the controversy that now confuses many researchers studying genome evolution . Since the rebuttal of RBM is based on a sophisticated theorem for computing rearrangement distances , few biologists can grasp all the details of both [4] and [6] . Fortunately , since both [4] and [6] use only computational arguments and simulations to refute/support RBM , this controversy ( different from some biological controversies ) is easy to resolve: one should simply check all computational arguments and simulations . In this paper , we developed algorithms for analyzing 3-breaks that generalize the standard rearrangements and make the analysis of rearrangements more transparent . We further analyzed the effects of transpositions ( and other 3-breaks ) on breakpoint re-use and came to the conclusion that even if transpositions and three-way fissions/fusions were dominant rearrangement operations , the arguments against RBM still hold . While one can still argue that rearrangements even more complex than 3-breaks ( e . g . , 4-breaks ) are common , this argument is not supported by existing biological knowledge . We also reproduced the simulations from [6] and came to the conclusion that the “block deletion” argument in [7] is flawed , similarly to the already refuted “synteny blocks” argument in [6] . If RBM is put to rest in favor of FBM , one has to answer the question of what makes certain regions break and others not break . Peng et al . [20] argued that long regulatory regions and inhomogeneity of gene distribution in mammalian genomes might be responsible for the breakpoint reuse phenomenon . The link between rearrangements and regulatory regions was explored in depth by Kikuta et al . [16] , who argued that long-range interactions between genes and their regulatory regions might explain solid and fragile regions in the genomes . However , revealing all factors responsible for genomic fragility and discovery of all fragile regions in the human genome remains an open problem . A computational approach based on comparison of gene orders was pioneered by David Sankoff [51 , 52] . Since some methods and notations used in this paper differ from the previous papers , below we briefly review the key concepts/methods that are relevant for this paper and put them in the context of previous studies . Initially , genome rearrangements were modeled by a combinatorial problem of sorting by reversals , as described below . The order of genes in two organisms is represented by permutations π = π1π2…πn and σ = σ1σ2…σn . A reversal ρ ( i , j ) of an interval [i , j] is the permutation The reversal ρ ( i , j ) has the effect of reversing the order of πiπι+1…πj and transforming π1…πi-1πi…πjπj+1…πn into π⋅ρ ( i , j ) = π1…πi-1πj…πiπj+1…πn . Given permutations π and σ , the reversal distance problem is to find a series of reversals ρ1 , ρ2 , … , ρt such that π·ρ1·ρ2 . . . ·ρt = σ and t is minimal . We call t the reversal distance between π and σ . Sorting π by reversals is the problem of finding the reversal distance d ( π ) between π and the identity permutation ( 12 . . . n ) . We extend a permutation π = π1π2…πn by adding π0 = 0 and πn+1 = n + 1 . We call a pair of elements ( πi , πi+1 ) , 0 ≤ i ≤ n , of π an adjacency if \|πi - πi+1\| =1 , and a breakpoint if \|πi − πi+1\| > 1 . It is easy to see that d ( π ) ≥ b ( π ) / 2 , where b ( π ) is the number of breakpoints in π . However , the estimate of reversal distance in terms of breakpoints is very inaccurate . Bafna and Pevzner [53] showed that another parameter ( size of a maximum cycle decomposition of the breakpoint graph ) estimates reversal distance with much greater accuracy . Originally , the breakpoint graph of a permutation π was defined as an edge-colored graph G ( π ) with n + 2 vertices {π0 , π1 , … , πn , πn+1} = {0 , 1 , . . . , n + 1} . We join vertices πi and πi+1 by a black edge for 0 ≤ i ≤ n . We join vertices πi and πj by a gray edge if πi − πj = 1 . A cycle in an edge-colored graph G is called alternating if the colors of every two consecutive edges of this cycle are distinct . It is easy to see that G ( π ) contains an alternating Eulerian cycle . Therefore , there exists a cycle decomposition of G ( π ) into edge-disjoint alternating cycles ( every edge in the graph belongs to exactly one cycle in the decomposition ) . We are interested in the decomposition of the breakpoint graph into a maximum number c ( π ) of edge-disjoint alternating cycles . Cycle decompositions play an important role in estimating reversal distance . Bafna and Pevzner [53] proved the bound d ( π ) ≥ n + 1 − c ( π ) , which is much tighter than the bound in terms of breakpoints d ( π ) ≥ b ( π ) / 2 . Finding a maximal cycle decomposition is a difficult problem . Fortunately , in the more biologically relevant case of signed permutations , this problem is trivial . Genes are directed fragments of DNA , and a sequence of n genes in a genome is represented by a signed permutation on {1 , . . . , n} with a “+” or “−” sign associated with every element of π . In the signed case , every reversal of fragment [i , j] changes both the order and the signs of the elements within that fragment . We are interested in the minimum number of reversals d ( π ) required to transform a signed permutation π into the identity signed permutation ( +1+2 . . . +n ) . The concept of a breakpoint graph extends naturally to signed permutations by mimicking every directed element by two undirected elements , which substitute for the tail and the head of the directed element [53] . For signed permutations , the bound d ( π ) ≥ n +1 − c ( π ) approximates the reversal distance extremely well . Hannenhalli and Pevzner [54] showed that where h ( π ) is the number of hurdles in π . In the model of the multichromosomal genomes we consider , every gene is represented by an integer whose sign ( “+” or “–” ) reflects the direction of the gene . A chromosome is defined as a sequence of genes , while a genome is defined as a set of chromosomes . Given two genomes π and Γ , we are interested in a most parsimonious scenario of evolution of Π into Γ ( i . e . , the shortest sequence of rearrangement events [defined below] required to transform Π into Γ ) . We assume that Π and Γ contain the same set of genes . Let Π be a multichromosomal genome . Every chromosome π in Π can be viewed either from left to right ( i . e . , as π = ( π1…πn ) ) or from right to left ( i . e . , as −π = ( −πn…−π1 ) ) , leading to two equivalent representations of the same chromosome ( i . e . , the directions of chromosomes are irrelevant ) . The four most common elementary rearrangement events in multichromosomal genomes are reversals , translocations , fusions , and fissions , defined below . Let π = π1…πn be a chromosome and 1 ≤ i ≤ j ≤ n . A reversal ρ ( π , i , j ) on a chromosome π rearranges the genes inside π = π1…πi-1πι…πjπj+1…πn and transforms π into π1…πi-1 − πj…−πiπj+1…πn . Let π = π1…πn and σ = σι…σm be two chromosomes and 1 ≤ i ≤ n +1 , 1 ≤ j ≤ m + 1 . A translocation ρ ( π , σ , i , j ) exchanges genes between chromosomes π and σ and transforms them into chromosomes π1…πi-1σj…σm and σι…σj-1πi…πn with ( i – 1 ) + ( m – j + 1 ) and ( j – 1 ) + ( n – i + 1 ) genes , respectively . We denote as Π·ρ the genome obtained from Π as a result of a rearrangement ( reversal or translocation ) ρ . Given genomes Π and Γ , the genomic sorting problem is to find a series of reversals and translocations ρ1 , ρ2 , … , ρt such that Π·ρ1·ρ2·…·ρt = Γ and t is minimal . We call t the genomic distance between Π and Γ . The genomic distance problem is the problem of finding the genomic distance d ( Π , Γ ) between Π and Γ . A translocation ρ ( π , σ , n + 1 , 1 ) concatenates the chromosomes π and σ , resulting in a chromosome π1…πnσ1…σm and an empty chromosome Ø . This special translocation , leading to a reduction in the number of ( nonempty ) chromosomes , is known in molecular biology as a fusion . The translocation ρ ( π , Ø , i , 1 ) for 1 < i < n “breaks” a chromosome π into two chromosomes: ( π1…πi-1 ) and ( πi…πn ) . This translocation , leading to an increase in the number of ( nonempty ) chromosomes , is known as a fission .	Rearrangements are genomic “earthquakes” that change the chromosomal architectures . The fundamental question in molecular evolution is whether there exist “chromosomal faults” ( rearrangement hotspots ) where rearrangements are happening over and over again . The random breakage model ( RBM ) postulates that rearrangements are “random , ” and thus there are no rearrangement hotspots in mammalian genomes . RBM was proposed by Susumo Ohno in 1970 and later was formalized by Nadeau and Taylor in 1984 . It was embraced by biologists from the very beginning due to its prophetic prediction power , and only in 2003 was refuted by Pevzner and Tesler , who suggested an alternative fragile breakage model ( FBM ) of chromosome evolution . However , the rebuttal of RBM caused a controversy , and in 2004 , Sankoff and Trinh gave a rebuttal of the rebuttal of RBM . This led to a split among researchers studying chromosome evolution: while most recent studies support the existence of rearrangement hotspots , others feel that further analysis is needed to resolve the validity of RBM . In this paper , we develop a theory for analyzing complex rearrangements ( including transpositions ) and demonstrate that even if transpositions were a dominant evolutionary force , there are still rearrangement hotspots in mammalian genomes .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "evolutionary", "biology", "homo", "(human)", "mus", "(mouse)", "computational", "biology" ]	2007	Are There Rearrangement Hotspots in the Human Genome?
Frameshift and nonsense mutations are common in tumors with microsatellite instability , and mRNAs from these mutated genes have premature termination codons ( PTCs ) . Abnormal mRNAs containing PTCs are normally degraded by the nonsense-mediated mRNA decay ( NMD ) system . However , PTCs located within 50–55 nucleotides of the last exon–exon junction are not recognized by NMD ( NMD-irrelevant ) , and some PTC-containing mRNAs can escape from the NMD system ( NMD-escape ) . We investigated protein expression from NMD-irrelevant and NMD-escape PTC-containing mRNAs by Western blotting and transfection assays . We demonstrated that transfection of NMD-irrelevant PTC-containing genomic DNA of MARCKS generates truncated protein . In contrast , NMD-escape PTC-containing versions of hMSH3 and TGFBR2 generate normal levels of mRNA , but do not generate detectable levels of protein . Transfection of NMD-escape mutant TGFBR2 genomic DNA failed to generate expression of truncated proteins , whereas transfection of wild-type TGFBR2 genomic DNA or mutant PTC-containing TGFBR2 cDNA generated expression of wild-type protein and truncated protein , respectively . Our findings suggest a novel mechanism of gene expression regulation for PTC-containing mRNAs in which the deleterious transcripts are regulated either by NMD or translational repression . A subset of colorectal carcinomas exhibit a molecular phenotype commonly referred to as high microsatellite instability ( MSI-H ) [1] . The microsatellite instability ( MSI ) pathway begins with the inactivation of one of a group of genes responsible for DNA nucleotide mismatch repair ( MMR ) , which leads to extensive mutations in both repetitive and non-repetitive DNA sequences [2–4] . The mechanism of tumorigenesis in MSI-H tumors is thought to involve frameshift mutations of microsatellite repeats within the coding regions of affected genes , and the inactivation of these genes is believed to contribute directly to tumor development and progression [5 , 6] . The frameshift mutations observed in the affected genes are expected to generate previously undescribed amino acid sequences in the C-terminal part of the respective proteins ( Figure S1 ) . If abnormal mRNAs and proteins are generated from the frameshift-mutated genes , tumor-specific antigen may be generated . High peritumoral lymphocytic infiltration and a relatively good prognosis have been reported in MSI-H tumors [7 , 8] . One of the important consequences of frameshift mutations is the formation of premature termination codons ( PTCs ) . In mammalian cells , mRNAs containing a PTC due to a nonsense mutation or a frameshift mutation are recognized and degraded by nonsense-mediated mRNA decay ( NMD ) , thus eliminating the production of the potentially deleterious truncated proteins [9 , 10] . NMD of mRNAs carrying PTCs is mediated through the recognition of the PTC by its position relative to the 3′-most last exon–exon junction . As a general rule , mammalian transcripts that contain a PTC more than 50–55 nucleotides ( nt ) upstream of the last exon–exon junction will be subjected to NMD [11 , 12] . Although PTC formation in frameshift mutation-derived mRNAs and their subsequent degradation through NMD is widely accepted , PTCs located within 50–55 nt or downstream of the last exon–exon junction are not recognized by NMD ( NMD-irrelevant ) , and some mRNAs with PTCs more than 50–55 nt upstream of their last exon–exon junction are not degraded by NMD ( NMD-escape ) [13 , 14] . In MSI-H tumors , several NMD-sensitive or NMD-escape PTC-containing mRNAs have been reported . A previous study compared the total mRNAs of affected genes from various cell lines [15] . However , this study did not differentiate the proportion of wild-type and mutant mRNAs , and did not confirm the mutant mRNAs through sequencing . This study also did not consider that the amount of mRNA from the affected genes might vary between cell lines . Moreover , the expression statuses and biological effects of the NMD-escape PTC-containing mRNAs are essentially unknown . In order to clarify the protein expression status of affected genes with frameshift mutations and the role of NMD in these mutated genes , we selected MSI-H cancers as a model system because these cancers have accumulated genes with frameshift mutations , and the mRNAs expected from these mutated genes contain PTCs . We analyzed the expression of 20 mutant mRNAs from 12 genes and evaluated their regulation along with the regulation of associated proteins . We demonstrate that some PTC-containing mRNAs escaped from NMD , but did not generate truncated proteins , indicating that PTC-containing transcripts can be regulated either by NMD or translational repression . To examine the effect of NMD on the affected genes with frameshift mutations in MSI-H tumors , we selected 12 genes from MSI-H tumors based on the reported frameshift mutation frequencies greater than 30% ( ABCF1 , ACVR2 , hMSH3 , hMSH6 , hRad50 , MARCKS , PRKWNK1 , RFC3 , SEC63 , TAF1B , TCF-4 , and TGFBR2 ) . We used genome sequencing of these 12 genes to identify frameshift mutation status . In these 12 genes , we identified 20 frameshift mutations that fell into three categories: 12 mutations were single base pair ( bp ) deletions , six were 2-bp deletions , and two were single bp insertions in coding mononucleotide repeats ( cMNR ) ( Table S1 ) . All 20 frameshift mutations of the 12 genes resulted in mRNAs containing a PTC ( Table 1 ) . We analyzed mRNA expression of the 12 genes by reverse transcriptase PCR ( RT-PCR ) in seven MMR-deficient ( LS174T , HCT-8 , SNU C2A , SNU C4 , DLD-1 , HCT116 , and LOVO ) and three MMR-proficient ( NCI-H508 , SW480 , and HT29 ) colorectal cancer cell lines . Primers were designed to contain at least one exon–exon junction region and to amplify the coding repeat sequences ( Table S2 ) . Of the 20 frameshift mutations in the genomic DNA , mutation-derived transcripts were detected from ten alleles representing six genes ( hMSH3 , TAF1B , TGFBR2 , ACVR2 , MARCKS , and TCF-4 ) , whereas ten alleles representing six genes ( ABCF1 , hMSH6 , hRad50 , PRKWNK1 , RFC3 , and SEC63 ) did not generate frameshift mutation-derived transcripts . No differences in expression of frameshift mutation-derived mRNA were observed between cell lines . Of the ten transcripts with frameshift mutations , five transcripts ( representing three genes: hMSH3 , TAF1B , and TGFBR2 ) had PTCs more than 50–55 nt upstream of the last exon–exon junction and were therefore expected to be degraded by NMD but instead escaped from NMD ( NMD-escape ) . On the other hand , the five remaining transcripts ( representing three genes: ACVR2 , MARCKS , and TCF-4 ) had PTCs within 50–55 nt upstream of the last exon–exon junction and were therefore expected to be irrelevant to NMD ( NMD-irrelevant ) . Accordingly , the 20 transcripts from 12 genes were classified as NMD-sensitive , NMD-escape , and NMD-irrelevant ( Table 1 ) . In order to confirm the effect of NMD on the NMD-sensitive and NMD-escape PTC-containing mRNAs , we used RT-PCR and a ribonuclease protection assay ( RPA ) to analyze the expression of the target gene mRNAs after treatment with puromycin , a translation inhibitor . In the five NMD-escape alleles that generated detectable frameshift mutation-derived mRNAs , no expression differences were found after puromycin treatment . In the ten NMD-sensitive alleles that produced no detectable frameshift mutation-derived mRNAs , mutant transcripts were detected after puromycin treatment ( Figure S2 ) . We analyzed the amount of two degraded NMD-sensitive transcripts , hRad50 and hMSH6 , by RPA and found a total loss of mutant transcripts , as evidenced by a 2-fold increase in hRad50 and hMSH6 products after puromycin treatment . In contrast , there was no loss of TGFBR2 mutant mRNA , an NMD-escape transcript , because the amount of product was unchanged after puromycin treatment ( Figure 1 ) . Next , we evaluated the effect of down-regulating UPF1 or UPF2 , which are key NMD factors , on the stability of the frameshift mutation-derived mRNAs , hRad50 and hMSH6 , using specific small interfering RNA ( siRNA ) . Upon the treatment of luciferase siRNA , expression of the mutation-derived hRad50 and hMSH6 mRNAs were not detected in the cell lines with hRad50 and hMSH6 mutations . In contrast , down-regulating UPF1 or UPF2 abundantly increased the frameshift mutation-derived mRNAs , as confirmed by RT-PCR , and sequence analysis . These findings indicate that frameshift mutation-derived mRNAs of hRad50 and hMSH6 are recognized and degraded by the NMD system ( Figure S3 ) . In order to determine if the truncated protein products from PTC-containing mRNAs can be detected , we first performed Western blotting analyses using antibodies directed against the N-terminus of hRad50 , hMSH6 , and hMSH3 . Truncated proteins were not detected for the NMD-sensitive ( hRad50 and hMSH6 ) genes , whereas wild-type proteins were detected in the cell lines containing the wild-type allele . These results support the previous finding that NMD-sensitive PTC-containing mRNAs are degraded by the NMD system . In the NMD-escape hMSH3 gene , we detected full-length hMSH3 proteins in cell lines with no mutations or with monoallelic mutations in these genes; however , we could not detect truncated hMSH3 proteins in any of the cell lines carrying frameshift mutations ( Figure 2 ) . The hMSH3 antibody detected the truncated proteins from the cell lines transfected with mutant hMSH3 cDNA , indicating that the antibodies used in our experiments specifically react with the N-terminal region of hMSH3 protein ( Figure S4 ) . Furthermore , we could not detect the truncated proteins of hMSH3 genes after treatment with the proteasome inhibitors MG132 or E64 ( unpublished data ) , which excludes the possibility of rapid degradation of mutated proteins . We interpreted the failure to detect truncated protein from the NMD-escape PTC-containing mRNAs as follows: ( 1 ) truncated proteins were generated , but at an amount not sufficient for detection by Western blotting , ( 2 ) truncated proteins were generated , but then rapidly degraded , or ( 3 ) truncated proteins were not generated from the mutant mRNA . To rule out an insufficient amount of endogenous truncated proteins , we constructed expression plasmids with NMD-escape PTC-containing genomic DNA or cDNA of TGFBR2: ( 1 ) wild-type cDNA of TGFBR2 ( K: TGFBR2 ( WT ) -cDNA ) , ( 2 ) PTC-containing mutant cDNA of TGFBR2 without downstream exons and introns ( L: TGFBR2 ( −1 ) -cDNA ) , ( 3 ) wild-type TGFBR2 genomic DNA ( M: TGFBR2 ( WT ) -splicing ) , ( 4 ) mutant TGFBR2 genomic DNA with a 1-bp deletion ( N: TGFBR2 ( −1 ) -splicing ) , and ( 5 ) mutant TGFBR2 genomic DNA with a 1-bp deletion and a PTC artificially located in the last exon ( O: TGFBR2 ( −1 ) -irrelevant ) ( Figure 3A ) . Among three NMD-escape PTC-containing mutated genes that we found in MSI-H tumors , we selected TGFBR2 . TAF1B and hMSH3 were excluded because of their large size and number of exons , which result in the failure or inefficient transfection of genomic DNA . In all of the constructs described above , the nucleotide sequences encoding FLAG peptide was introduced immediately downstream of the initiation codon , which allows for detection of the encoded proteins by Western blotting . These vectors were designed to differentiate the effect of spliced wild-type mRNA , spliced mutant NMD-escape mRNA , and spliced mutant NMD-irrelevant mRNA in terms of truncated protein expression . We observed abundant expression of PTC-containing TGFBR2 mRNA in cell lines transfected with TGFBR2 ( −1 ) -cDNA , TGFBR2 ( −1 ) -splicing , and TGFBR2 ( −1 ) -irrelevant ( Figure 3B ) . Cell lines transfected with TGFBR2 ( WT ) -splicing , TGFBR2 ( −1 ) -splicing , and TGFBR2 ( −1 ) -irrelevant showed accurate splicing , and all normal and mutant mRNA products were confirmed by sequence analysis ( unpublished data ) . A semi-quantitative RT-PCR analysis designed to detect exogenous TGFBR2 mRNA showed similar and abundant amounts of TGFBR2 mRNA expression in all of the cell lines transfected with the five different constructs ( unpublished data ) . In this analysis of protein expression using the anti-FLAG antibody , we demonstrated the expression of wild-type TGFBR2 protein in cell lines transfected with TGFBR2 ( WT ) -cDNA and TGFBR2 ( WT ) -splicing . We also demonstrated the expression of truncated TGFBR2 protein in cell lines transfected with TGFBR2 ( −1 ) -cDNA and TGFBR2 ( −1 ) -irrelevant . Intriguingly , we could not detect any TGFBR2 protein in cell lines transfected with TGFBR2 ( −1 ) -splicing , indicating a selective translational repression of NMD-escape mutant mRNA ( Figure 3C ) . In order to confirm that translational repression is responsible for the failure to detect truncated protein from PTC-containing TGFBR2 mRNA , we examined the mRNA distribution of TGFBR2 ( WT ) -splicing and TGFBR2 ( −1 ) -splicing by polysome analysis . In the cell line with the TGFBR2 ( WT ) -splicing vector , TGFBR2 ( WT ) -splicing mRNA was found in the polysome-containing fractions similar to endogenous GAPDH mRNA ( Figure 3D ) . However , in the cell line with the TGFBR2 ( −1 ) -splicing vector , a greater percentage of TGFBR2 ( −1 ) -splicing mRNA was found in the fractions that contained ribosomal subunits and monosomes , whereas endogenous GAPDH mRNA co-sedimented with polysomes ( Figure 3E ) . Furthermore , upon the treatment of puromycin , a greater percentage of TGFBR2 ( WT ) -splicing mRNA and endogenous GAPDH mRNA were shifted into ribosomal subunits and monosome-containing fractions ( Figure 3F ) . In order to rule out the possibility that the weak polysome association of TGFBR2 ( −1 ) -splicing mRNA is due to its shorter open reading frame as compared to the TGFBR2 ( WT ) -splicing mRNA , we repeated the same experiment using the TGFBR2 ( −1 ) -splicing vector with puromycin treatment . The results show that there is no significant difference in the cell line transfected with TGFBR2 ( −1 ) -splicing after puromycin treatment ( Figure 3E and 3G ) . These results indicate that ( 1 ) the sedimentation of TGFBR2 ( WT ) -splicing mRNA in heavy fractions was due to polysome association , and ( 2 ) the shift of TGFBR2 ( −1 ) -splicing mRNA into ribosomal subunits and monosome-containing fractions is due to translational repression , similar to TGFBR2 ( WT ) -splicing mRNA and TGFBR2 ( −1 ) -splicing mRNA treated with puromycin ( Figure 3E–3G ) . This novel mechanism , whereby PTC recognition itself triggers translational repression , is referred to as nonsense-mediated translational repression ( NMTR ) . We demonstrated the selective translational repression of the NMD-escape mutant TGFBR2 ( −1 ) -splicing mRNA after normal splicing , and this repression was not found in the NMD-irrelevant mutant TGFBR2 mRNA , which lacks a downstream sequence of the termination codon . Therefore , we examined other possible factors influencing the expression of the truncated protein by: ( 1 ) changing the 3′ UTR length ( the length between the PTC and poly ( A ) tail ) to check the possible effect of 3′ UTR length on translational repression [16] , ( 2 ) treating with a proteasome inhibitor ( MG132 ) in the cell lines transfected with TGFBR2 ( −1 ) -splicing and TGFBR2 ( −1 ) -irrelevant to rule out that the truncated proteins are generated but rapidly degraded by the proteasome , and ( 3 ) down-regulating key NMD factors , UPF1 and UPF2 , to evaluate whether NMD factors are involved in the translational repression of NMD-escape PTC-containing spliced TGFBR2 mutant mRNA . In order to change the 3′ UTR length , we constructed another TGFBR2 ( −1 ) -irrelevant with a full-length cDNA sequence spanning from the PTC to the 3′ end of TGFBR2 ( P: TGFBR2 ( −1 ) -irrelevant-F ) ( Figure 4A ) . When the genomic DNAs of TGFBR2 ( −1 ) -splicing , TGFBR2 ( −1 ) -irrelevant , and TGFBR2 ( −1 ) -irrelevant-F were transfected , normal splicing and a large amount of mutant mRNAs were present in all three cell lines transfected with the different genomic DNAs ( Figure 4C ) . However , no proteins were expressed in the cell lines transfected with TGFBR2 ( −1 ) -splicing , whereas a large amount of truncated proteins were expressed in the cell lines transfected with TGFBR2 ( −1 ) -irrelevant . In the cell lines transfected with TGFBR2 ( −1 ) -irrelevant-F , the amount of truncated proteins was reduced to about 35% of that of the cell lines transfected with TGFBR2 ( −1 ) -irrelevant . These findings indicate that the 3′ UTR length itself or specific cis-acting element ( s ) within the 3′ UTR seem to contribute to the translational inhibition of TGFBR2 ( −1 ) mRNA . However , more importantly , a splicing event downstream of the PTC may be involved in NMTR because truncated proteins are expressed in cells transfected with TGFBR2 ( −1 ) -irrelevant-F , but not in cells transfected with TGFBR2 ( −1 ) -splicing ( Figure 4D ) . We excluded the possibility that mutated proteins are generated , but then rapidly degraded by the proteasome , because cell lines transfected with TGFBR2 ( −1 ) -splicing and treated with MG132 , a proteasome inhibitor , demonstrated no truncated proteins . In contrast , the cell lines transfected with TGFBR2 ( −1 ) -irrelevant-F and treated with MG132 demonstrated similar amounts of truncated proteins compared to cell lines only transfected with TGFBR2 ( −1 ) -irrelevant-F ( Figure 4D ) . Finally , we evaluated whether key NMD factors are involved in NMTR . We expected that the most significant difference between PTC-containing TGFBR2 mRNA and PTC-containing TGFBR2 mRNA , which lacks an intron downstream of the PTC , would be the presence of exon junction complexes ( EJCs ) behind the PTC . An EJC recruits the NMD factors , UPF1 and UPF2 , which play a key role in mRNA quality control . We treated cells with UPF1 and UPF2 siRNA in order to elucidate whether these two NMD factors are involved in NMTR . The level of UPF1 was down-regulated to about 20% of normal , where normal is defined as the level in the presence of the nonspecific control , luciferase siRNA , whereas the level of UPF2 was down-regulated to about 10% of normal ( Figure 4B ) . Treatment of any of the siRNAs failed to produce truncated proteins in the cell lines transfected with TGFBR2 ( −1 ) -splicing , indicating that at least these two NMD factors do not play an important role in the NMTR of NMD-escape mutant TGFBR2 mRNA ( Figure 4D ) . We then examined whether NMD-irrelevant PTC-containing mRNA can generate truncated protein . We selected one NMD-irrelevant mRNA , mutant MARCKS , and performed a transfection assay using wild-type MARCKS genomic DNA ( MARCKS ( WT ) -splicing ) and mutant MARCKS genomic DNA with a 2-bp deletion ( MARCKS ( −2 ) -splicing ) ( Figure 5A ) . We found expression of wild-type and truncated protein in cell lines transfected with MARCKS ( WT ) -splicing and MARCKS ( −2 ) -splicing , respectively , by Western blotting with an anti-FLAG antibody ( Figure 5B ) . To verify that these protein products were identical to MARCKS , we performed Western blotting with the anti-MARCKS antibody and confirmed the expression of wild-type and truncated proteins ( Figure 5C ) . We also found that the truncated MARCKS protein is subject to active proteasome-mediated degradation; the amount of truncated MARCKS protein increased with time when cells were treated with MG132 , a proteasome inhibitor ( Figure 5B and 5C ) . In this study , we found that some PTC-containing mRNAs are not degraded by the NMD system , and their protein translations are repressed . We therefore suggest that PTC-containing mRNAs resulting from frameshift mutations can be classified into three groups: NMD-sensitive mRNAs , which are degraded by the NMD system; NMD-escape mRNAs , which are not degraded by the NMD system , but do experience repression of protein expression; and NMD-irrelevant mRNAs , which are not recognized by the NMD system , and generate truncated proteins . Our findings indicate that both NMD and NMTR , an additional surveillance mechanism for translational control , are involved in the recognition of PTC and suppression of truncated protein from PTC-containing genes that can be deleterious to cell function ( Figure 6 ) . NMD is a quality control-based surveillance mechanism that protects cells from the potentially dominant negative effects of truncated mutant proteins . The primary role of the NMD pathway is to eliminate nonsense transcripts that result from faulty transcription , alternative splicing , or somatic mutation [17 , 18] . This pathway selectively degrades mRNAs that prematurely terminate translation due to a frameshift or nonsense mutation . Although NMD is a quality control-based surveillance mechanism , avoidance of NMD by PTC-containing mRNAs has been reported for the mutated genes of many diseases . Moreover , about one third of the alternative transcripts in cells are expected to contain PTCs due to splicing errors and regulated unproductive splicing and translation ( RUST ) . Some of these PTC-containing mRNAs belong to the NMD-escape variety [14 , 15] . If translated , these NMD-escape mRNAs could produce truncated proteins that may critically interfere with cell viability . Among the PTC-containing mRNAs , some PTCs , which are called NMD-irrelevant mRNAs , are located within 50–55 nt or downstream of the last exon–exon junction and are not detected by NMD . Proteins generated from these types of PTC-containing mRNAs and their causal relationship to specific diseases have been well documented [19–21] . However , there are no reports describing translated proteins from mutation-derived NMD-escape mRNAs , although many NMD-escape PTC-containing mRNAs have been reported [22–24] . In this study , we demonstrated that NMD-escape TGFBR2 mRNA is subject to NMTR . Our transfection study of TGFBR2 constructs demonstrated that PTC-containing mRNAs from mutant TGFBR2 were abundant after transfection of mutant cDNA and mutant TGFBR2 genomic DNAs with a 1-bp deletion . However , truncated proteins were only detected in the cell lines transfected with mutant TGFBR2 cDNA , and no truncated proteins were detected in the cell lines transfected with mutant TGFBR2 genomic DNAs with a 1-bp deletion . In contrast , strong expression of TGFBR2 protein was observed in the cell lines transfected with wild-type TGFBR2 genomic DNA . Next , we confirmed using polysome analysis that the lack of truncated protein translated from PTC-containing TGFBR2 mRNA is due to translational repression , not instability of TGFBR2 mRNA . The major expected differences between the two PTC-containing TGFBR2 mutant mRNAs and mRNA from the cDNA of TGFBR2 or TGFBR2 genomic DNA are the deposition of the EJC and the possible recruitment of NMD factors to the mRNA during translation termination . We also confirmed the NMTR by demonstrating that mutant mRNA and truncated proteins were efficiently expressed in the cell line transfected with mutant TGFBR2 genomic DNA containing a PTC in the last exon without downstream introns ( Figures 3C and 4D ) . We therefore suspected that EJC proteins and/or NMD factors might play an important role in the NMTR . It is well known that NMD recognizes PTC and downstream splicing events that deposit an EJC at an exon–exon junction . The EJC is composed of proteins involved in splicing and the subsequent steps of mRNA transport and translation . EIF4A3 , RNPS1 , Y14 , and MAGOH are involved in EJC formation , and the EJC–mRNA complex is then exported to the cytoplasm together with nuclear cap binding proteins CBP80/20 and nuclear poly ( A ) binding protein 2 ( PABP2 ) [25–28] . The mRNA then recruits UPF2 and undergoes a so-called “pioneer” round of translation during mRNA export . NMD occurs when translation terminates more than 50–55 nt upstream of the last exon–exon junction . Transient SURF formation at the termination codon , which is composed of Smg1 , UPF1 , and translation termination factors eRF1–eRF3 , is thought to interact with the downstream EJC so as to trigger phosphorylation of UPF1 and thereby elicit NMD [26–30] . In this study , we demonstrated that key NMD factors , UPF1 and UPF2 , did not play an important role in the NMTR , which is evidenced by the fact that treatment of UPF1 and UPF2 siRNA did not produce truncated proteins in cell lines transfected with TGFBR2 ( −1 ) -splicing , even though both siRNAs drastically down-regulate endogenous UPF1 and UPF2 . Moreover , down-regulating Y14 or EIF4A3 , which are EJC components , using siRNA failed to restore translational repression of TGFBR2 ( −1 ) -splicing ( unpublished data ) . Our results indicate that NMTR is at work on the NMD-escape PTC-containing TGFBR2 mRNA by some unknown surveillance mechanism . The involvement of another messenger ribonucleoprotein particle ( mRNP ) complex in this translational repression is essentially unknown . Future studies should be focused on the role of translational repression of the various RNA binding proteins in the PTC-containing mRNP complex . Several recent reports have demonstrated the importance of termination codon context , especially 3′ UTR length , in PTC recognition [16] . Therefore , we tested two TGFBR2 ( −1 ) -irrelevant vectors with short and extended 3′ UTR lengths . If the PTC-containing , 3′ UTR-extended construct failed to produce truncated protein , then unlike NMD , splicing and EJCs may not be involved in the mechanism of NMTR . In this experiment , we demonstrated a 65% reduction of truncated protein in the cell lines transfected with TGFBR2 ( −1 ) -irrelevant vector with an extended 3′ UTR length . These findings indicate that 3′ UTR length is an important factor; however , other important factors are involved in NMTR . Because NMTR depends on 3′ UTR length or a putative cis-element residing in the 3′ UTR , it is , in part , reminiscent of EJC-independent NMD . A PTC within the penultimate exon of the β-globin or TPI gene elicits NMD depending on the position of PTC relative to the last EJC [12 , 31] . However , for a PTC within the penultimate exon that normally elicits NMD , deleting the last intron fails to eliminate NMD , indicating that the last exon has a so-called “failsafe” sequence that allows for PTC recognition and triggers NMD in the absence of a downstream EJC . Intriguingly , this element requires that splicing occur upstream of the PTC because the PTC-containing mRNA that is derived from an intronless TPI or β-globin gene is immune to NMD [12] . It remains to be clarified whether NMTR also requires a splicing event upstream of the PTC . Recently , another case for EJC-independent NMD has been reported in immunoglobulin-μ mRNA [16] . Similar to NMTR , EJC-independent NMD of this mRNA depends on the 3′ UTR length . However , the mode of PTC recognition looks quite different between NMTR and EJC-independent NMD of this mRNA , in the sense that NMTR does not require the NMD factors , UPF1 and UPF2 , as shown in our study . Important questions of whether mRNAs targeted by EJC-independent NMD are subject to NMTR should be addressed . Marked degradation of PTC-containing mRNAs and decreased protein synthesis from PTC-containing mRNAs by the NMD system have been reported in yeast [32] . Another important mRNA surveillance mechanism , nonsense-mediated altered splicing ( NAS ) , has been reported in mammalian cells [33] . NAS induces alternative splicing in the PTC-containing mRNA , thus avoiding the production of toxic mutant proteins . Although the exact mechanism of NAS had not been reported , UPF1 plays an important role in NAS [34] . Together with our findings that UPF1 did not play a significant role in NMTR , all of these findings indicate that ( 1 ) NMD , NAS , and NMTR play important roles in the inhibition of deleterious mutant protein production , and ( 2 ) unlike NMD and NAS , UPF1 and UPF2 do not a play key role in NMTR , suggesting novel factor ( s ) or pathways exist in the NMTR . In conclusion , we demonstrated three different molecular pathways of PTC-containing mRNAs . We propose a novel mechanism of gene expression regulation for PTC-containing mRNAs , in which the deleterious transcripts are regulated either by NMD or NMTR . Future studies of the NMD and NMTR control mechanism will enable us to better understand the reason for specific protein expression among the numerous mRNA isoforms , as well as the selective cellular control mechanism of protein expression . Ten cell lines were obtained from either the American Type Culture Collection ( ATCC; http://www . atcc . org ) or the Korean Cell Line Bank ( KCLB; http://cellbank . snu . ac . kr ) . Seven cell lines ( LS174T , HCT-8 , SNU C2A , SNU C4 , DLD-1 , HCT116 , and LOVO ) were MMR-deficient , and three ( NCI-H508 , SW480 , and HT29 ) were MMR-proficient in terms of their MSI status , as determined by previous studies [35 , 36] . We confirmed the presence of MSI using BAT26 and BAT25 markers . Cells were grown in RPMI supplemented with 10% FBS ( Life Technologies , Grand Island , New York , United States ) , penicillin , and streptomycin at 37 °C in 5% CO2 . Genomic DNA and cDNA preparation , analysis of MSI , and identification of target gene frameshift mutations were performed as described previously [37] . We used puromycin ( Sigma , St . Louis , Missouri , United States ) to inhibit the synthesis of proteins involved in NMD . Cells were grown to 80% confluence and then treated with 30-μg/ml puromycin . Six hours later , the cells were harvested , and total RNA was isolated using an RNeasy Mini kit ( QIAGEN , Valencia , California , United States ) according to the manufacturer's instructions . Twenty-one–nucleotide RNAs were chemically synthesized using a Silencer siRNA Construction kit ( Ambion , Austin , Texas , United States ) . Synthetic oligonucleotides were deprotected and gel-purified . HCT116 and SNU C2A cells growing in six-well dishes were transfected with 50 nM siRNA and oligofectamine ( Invitrogen , Carlsbad , California , United States ) according to the manufacturer's protocol . For RT-PCR analysis , total RNA was harvested 48 h after siRNA transfection . Targeted nucleotides , numbered relative to the start codon , were as follows: rent1/UPF1 , 1 , 879–1 , 901 ( 5′-AAGATGCAGTTCCGCTCCATTTT-3′ ) ; rent2/UPF2 , 1 , 423–1 , 445 ( 5′-AAGGCTTTTGTCCCAGCCATCTT-3′ ) ; and luciferase GL2 , 153–173 ( 5′-AACACGTACGCGGAATACTTCGA-3′ ) . The inhibition of UPF1 and UPF2 expression by siRNA targeting was evaluated by semi-quantitative RT-PCR or Western blotting . To study expression of TGFBR2 , we selected the secreted expression vector , pSecTag2B ( Invitrogen ) . The vector was cut by Hind III , and the FLAG oligonucleotide was inserted into the Hind III site to allow for specific immunodetection , thereby creating the pSecTag-FLAG vector . The complete coding sequence of full-length TGFBR2 begins with the first ATG at codon 1 and encodes a 568–amino acid protein , with a stop codon at 569 . Mutant TGFBR2 with a 1-bp deletion at ten adenosine repeats results in a premature stop at codon 162 ( Figure S1 ) . We constructed wild-type ( constructs K and M ) and truncated TGFBR2 expression vectors ( constructs L , N , O , and P ) . Wild-type TGFBR2 cDNA was obtained from a 293T cell line and mutant ( 1-bp deletion ) TGFBR2 cDNA was obtained from a HCT116 cell line by RT-PCR and then cloned using a T&A cloning kit ( RBC , Taipei , Taiwan ) . Construct K ( TGFBR2 ( WT ) -cDNA ) and construct L ( TGFBR2 ( −1 ) -cDNA ) were generated by inserting the Hind III fragment from the TGFBR2 cDNA into the Hind III site of pSecTag-FLAG . In order to analyze splicing and subsequent protein expression , we inserted wild-type and mutant TGFBR2 genomic DNA into the expression vectors . To generate constructs M ( TGFBR2 ( WT ) -splicing ) and N ( TGFBR2 ( −1 ) -splicing ) , exons 1 to 7 , except exon 3 , of TGFBR2 were obtained by PCR using 293T cell genomic DNA and cloned into the yT&A cloning vector ( RBC ) . Exon 3 of TGFBR2 was obtained from LOVO and 293T DNA by PCR , because a 1-bp deletion in the cMNR exists in exon 3 of TGFBR2 in LOVO . All primers for cloning were designed to contain more than 100 bp of intron sequence on each side of the exon boundary to ensure accurate splicing of the construct . These exon fragments were ligated to each other using a restriction enzyme site in the multiple cloning site ( MCS ) of the yT&A vector . In the case of exon 1 , deletion of the signal peptide was performed using Dpn1 for N-terminal FLAG tagging . The ligated genomic construct of TGFBR2 was inserted into the pSecTag-FLAG vector . To generate construct O ( TGFBR2 ( −1 ) -irrelevant ) , exon 4 to exon 7 of TGFBR2 ( −1 ) -splicing was deleted by cutting with BstX1 , and the remaining construct was self-ligated . To generate construct P ( TGFBR2 ( −1 ) -irrelevant-F ) , cDNA from exon 4 to exon 7 of TGFBR2 was inserted into the pSecTag-FLAG vector using EcoR1 and Pst1 restriction enzyme sites , and then genomic DNA from exon 1 to exon 3 was inserted into the same vector using the BamH1 restriction enzyme site . For the expression study of MARCKS , the expression vector pcDNA3 . 1 ( + ) ( Invitrogen ) was cut by Hind III , and the FLAG oligonucleotide was inserted at the Hind III site to allow for specific immunodetection , thereby creating the pcDNA-FLAG vector . In order to analyze splicing and subsequent protein expression , we inserted wild-type and mutant MARCKS genomic DNA into expression vectors . To generate constructs P ( MARCKS ( WT ) -splicing ) and Q ( MARCKS ( −2 ) -splicing ) , exon 1 of MARCKS was obtained through PCR using genomic DNA of 293T and cloned into the yT&A cloning vector ( RBC ) . Exon 2 of MARCKS was obtained from SNU C2A by PCR , since 2-bp monoallelic deletions in the cMNR exist in exon 2 of MARCKS in SNU C2A . These exon fragments were subcloned into the expression vector pcDNA-FLAG . Primer sequences are shown in Table S3 . HCT116 and HeLa cells ( 2 × 106 ) were transiently transfected using Lipofectamine 2000 ( Invitrogen ) with the specific construct plasmid and pSecTag-FLAG vector in a 60-mm plate . Cells were harvested 2 d later . Protein was purified from half of the cells using passive lysis buffer ( Promega , Madison , Wisconsin , United States ) , and total RNA was purified from the other half using TRIzol Reagent ( Invitrogen ) . Whole lysates from cell lines were prepared using passive lysis buffer ( Promega ) . Thirty micrograms of the total protein lysates were loaded into each lane , size-fractionated by SDS-PAGE , and then transferred to a PVDF membrane that was blocked with TBST containing 5% skim milk . Primary antibodies against GAPDH ( Trevigen , Gaitherburg , Maryland , United States ) , FLAG ( Sigma-Aldrich ) , hRad50 ( Ab13B3; Gene Tex ) , MARCKS ( Santa Cruz Biotechnology , Santa Cruz , California , United States ) , or hMSH6 and hMSH3 ( BD Bioscience , Mountain View , California , United States ) were incubated for 1 h at room temperature . After washing , membranes were incubated with HRP-conjugated secondary antibody ( Santa Cruz Biotechnology ) , washed , and then developed with ECL-Plus ( Amersham Pharmacia Biotech , Little Chalfont , United Kingdom ) . RNA samples were extracted from each cell line using TRIzol Reagent ( Invitrogen ) . RPA was carried out according to the instructions provided by the manufacturer ( RiboQuant RPA kit; BD Biosciences ) using 20 μg of total RNA per sample . The template antisense 32P-labeled RNA probes were specific for TGFBR2 ( unprotected , 271 nt; protected , 254 nt ) , hRad50 ( 371 , 353 ) , hMSH6 ( 332 , 312 ) , and GAPDH ( 134 , 120 ) mRNA . GAPDH was used as an internal control . The intensity of specific bands corresponding to individual riboprobes was determined by densitometry . All RPAs were evaluated at different exposures , and only bands that were within the linear range of the film were analyzed . Total RNA was extracted from HCT116 with TRIZOL Reagent ( Invitrogen ) . Full-length TGFBR2 cDNA was amplified by RT-PCR using 293T total RNA . Expression of the exogenous TGFBR2 construct mRNA was analyzed by Northern blot analysis using 10 μg of total RNA according to standard protocols . Cultures were supplemented with 100-μg/ml cycloheximide for 24 h post-transfection and incubated for 5 min at room temperature . HeLa cells were washed three times with ice-cold PBS containing 100-μg/ml cycloheximide . HeLa cells were collected by scrapping in PBS , transferred to Eppendorf tubes for additional washes , and then lysed in lysis buffer ( 15 mM Tris-Cl [pH 7 . 4] , 3 mM MgCl2 , 10 mM NaCl , 0 . 5% Triton X-100 , 100-μg/ml cycloheximide , 1-mg/ml heparin , and 200 U RNasin [Promega] ) . When indicated , puromycin ( 100 μg/ml ) was added to the cultures 2 h prior to harvesting , and cycloheximide was omitted from the gradient . For each construct , lysates from two 100-mm dishes were pooled into a microcentrifuge tube and incubated for 10 min on ice with occasional mixing . Nuclei and debris were removed by centrifugation at 12 , 000g for 2 min . Then , 1 ml of each cytoplasmic lysate was layered onto an 11-ml 10%–50% sucrose gradient and centrifuged at 4 °C in an SW40 rotor ( 39 , 000 rpm ) for 2 h . Sixteen fractions were collected from the top with concomitant measurement of absorbance at 254 nm , using a fraction collection system . RNA was extracted with TRIZOL Reagent and analyzed by RT-PCR . The GenBank ( http://www . ncbi . nlm . nih . gov/Genbank ) accession numbers for the gene products discussed in this paper are as follows: ABCF1 ( NM_001090 ) , ACVR2 ( NM_001616 ) , EIF4A3 ( NM_014740 ) , GAPDH ( NM_002046 ) , hMSH3 ( NM_002439 ) , hMSH6 ( NM_000179 ) , hRad50 ( NM_005732 ) , MARCKS ( NM_002356 ) , PRKWNK1 ( NM_018979 ) , RFC3 ( NM_002915 ) , SEC63 ( NM_007214 ) , TAF1B ( NM_005680 ) , TCF-4 ( NM_030756 ) , TGFBR2 ( NM_003242 ) , UPF1 ( NM_002911 ) , UPF2 ( NM_015542 ) , and Y14 ( NM_005105 ) .	A class of mutations found in many cancers introduces aberrant termination signals during the synthesis of mRNA . In mammalian cells , abnormal mRNAs containing premature termination codons ( PTCs ) are normally degraded by a process called nonsense-mediated mRNA decay ( NMD ) , thus avoiding potentially deleterious effects from abnormal protein production . However , some PTC-containing mRNAs are known to escape from NMD . By screening protein expression from genes with serious mutations in colon cancers , we confirmed that PTC-containing mRNAs of some genes escape from NMD . However , their abnormal proteins were not found in the tumor cells . To study the means by which these proteins were regulated , we transfected separate cell lines with NMD-escape mutant genomic DNA , wild-type genomic DNA , and mutant cDNA . We found that truncated proteins are not generated from the NMD-escape mutant genomic DNA , whereas wild-type protein and truncated protein were generated normally . These results indicate that the translation of PTC-containing mutant mRNAs is repressed in the cytoplasm .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods", "Supporting", "Information" ]	[ "oncology", "cell", "biology", "in", "vitro", "molecular", "biology", "genetics", "and", "genomics" ]	2007	Selective Translational Repression of Truncated Proteins from Frameshift Mutation-Derived mRNAs in Tumors
The protease GP63 is an important virulence factor of Leishmania parasites . We previously showed that GP63 reaches the perinuclear area of host macrophages and that it directly modifies nuclear translocation of the transcription factors NF-κB and AP-1 . Here we describe for the first time , using molecular biology and in-depth proteomic analyses , that GP63 alters the host macrophage nuclear envelope , and impacts on nuclear processes . Our results suggest that GP63 does not appear to use a classical nuclear localization signal common between Leishmania species for import , but degrades nucleoporins , and is responsible for nuclear transport alterations . In the nucleoplasm , GP63 activity accounts for the degradation and mislocalization of proteins involved amongst others in gene expression and in translation . Collectively , our data indicates that Leishmania infection strongly affects nuclear physiology , suggesting that targeting of nuclear physiology may be a strategy beneficial for virulent Leishmania parasites . The protozoan species Leishmania major ( L . major ) and L . mexicana are the causative agent of the cutaneous form of leishmaniasis , an ulcerative disease with ~1 million new cases reported worldwide annually . In mammalian hosts , Leishmania is an intracellular parasite that replicates predominantly within macrophages ( MΦ ) . To survive , Leishmania suppresses microbicidal and immune functions of the MΦ , caused by alterations of signaling pathways [1] . A crucial molecule for the subversion of host signaling is the leishmanial surface metalloprotease GP63 [2] . In recent years GP63 was found to be a prerequisite not only for the activation of host protein tyrosine phosphatases ( PTPs ) , such as SHP-1 , PTP1B and TCPTP [3] , but also for the alteration of transcription factor ( TF ) function through proteolytic cleavage [4] . Interestingly , GP63-mediated cleavage in the case of the TF AP-1 was shown to occur in the nucleus rather than the cytoplasm revealing that a parasite protease may enter the nucleus via an undefined mechanism [5] . In eukaryotic cells , the majority of nucleocytoplasmic exchange occurs through nuclear pore complexes ( NPCs ) . Those channels are composed of nucleoporins ( Nups ) and enable passive transport of smaller molecules . Larger proteins usually require a nuclear localization sequence ( NLS ) and carrier proteins to be transported through the NPC . Apart from nuclear transport , NPC and Nups have also been implicated in other critical cellular processes such as cell differentiation , genome organization , and gene expression [6–8] . To date , a number of obligate intracellular pathogens ( e . g . rhinovirus , poliovirus , HIV-1 ) have been shown to interfere with the nucleocytoplasmic transport machinery including the degradation of Nups and/or carrier proteins [9–11] . Leishmania parasites have been shown to affect host TF proteins and possibly their transport to the nucleus [4 , 5 , 12 , 13] . However , whether Leishmania parasites are able to further subvert host nuclear functions is unknown . Thus , the aim of this study was to elucidate the consequences of Leishmania infections on nuclear physiology with a special focus on the importance of GP63 . We here show that GP63 targets the nuclear envelope ( NE ) after Leishmania infection of murine MΦs . As a result GP63 degrades Nups of the NPC . Thereby , translocation of the leishmanial protease to the NE appears to be independent of a putative classical NLS . Proteomic analysis of MΦ nuclei after infection reveals an impact of GP63 on critical proteins involved in nuclear transport , nucleic acid metabolism , and essential mRNA processes . Comparison of the proteomic data sets of L . major and L . mexicana infections shows that both species affect predominantly chromatin remodeling and transcriptional/translational regulation , likely through GP63 . To our knowledge , this is the first in-depth proteomic analysis of macrophage nuclei after Leishmania infection revealing extensive alterations of nuclear physiology . We previously demonstrated that the Leishmania metalloprotease GP63 was able to cleave substrate proteins within the nucleus of host MΦs and can be localized in the perinuclear area of the nuclei [5] . In our experiments , we initially sought to confirm the previously observed distribution of GP63 within host cells after infection . In the conditions used , in accordance with the work of Contreras et al . , cells were typically infected by 1–2 parasites and images of host cells after infection indicate a preference of GP63 to localize in the perinuclear region ( S1 Fig ) . To investigate the nuclear localization of GP63 more accurately , we optimized a protocol to purify whole nuclei ( WN ) from MΦs and to fractionate the nuclear envelope ( NE ) from the nucleoplasm ( NP ) . Our western blot analysis confirmed that L . major GP63 is indeed present in nuclear fraction ( WN ) of infected MΦ . In-depth analysis of GP63 localization within the nucleus revealed that the protease predominantly remains at the NE ( Fig 1 ) . This was supported by gelatin zymography , which showed strong GP63 activity at the NE . These results coincided with the finding that two GP63 targets , the PTPs SHP-1 and TCPTP , were primarily localized at the NE putting protease and substrates in close proximity ( Fig 1 ) . Most nuclear proteins with a molecular mass of more than 40 kDa are transported into the nucleus through NPCs via the recognition and binding of a NLS by importins . The best-characterized transport signal is the classical NLS ( cNLS ) for nuclear protein import , which consists of either one ( monopartite ) or two ( bipartite ) stretches of basic amino acids [14] . We first searched for a cNLS in the GP63 sequence which is common to most Leishmania species . The PSORT II software did not identify a common cNLS [15] . However , we identified a motif very similar to a described consensus monopartite cNLS [16] . Mutants for this sequence ( GP63NLS ) were generated by site-directed mutagenesis . As the cNLS-like sequence was localized in the GP63 catalytic domain , we also generated a GP63 mutant for the active site ( GP63AS , E265D ) to discriminate between effects on the cNLS—and possibly transport—and effects on protease activity [17] ( Fig 2A ) . L . major GP63 knockout mutant parasites ( GP63-/- ) and L . major GP63-/- parasites complemented by transfection of the L . major GP63 gene 1 ( GP63 Rescued; GP63R ) [18] were used as controls for the new parasite strains ( Fig 2 ) . Examination of GP63 activity demonstrated that GP63 was not functional in GP63AS parasites while it was unaffected in GP63NLS parasites . Hence , we concluded that the mutation in the cNLS-like sequence did not alter GP63 proteolytic activity ( Fig 2A ) . In order to investigate whether GP63 utilized this putative cNLS-like sequence to target the NE enabling substrate cleavage events , we infected MΦs with different L . major species ( WT , GP63NLS , GP63AS , GP63-/- and GP63R ) or not ( Nil = non-infected LM1 ) , and monitored GP63 as well as GP63-mediated cleavage of TCPTP and c-Jun . Both substrate cleavage events were previously reported to take place in the nucleus or nuclear fractions , respectively [3 , 5] . The infection experiments showed that all GP63-expressing parasite strains but L . major GP63AS were able to cleave the substrates analyzed ( Fig 2B ) . GP63R and GP63NLS parasites had very similar GP63 expression and substrate cleavage patterns , probably due to the fact that only one GP63 gene copy was rescued [18] . The results indicate that the mutation in the putative cNLS-like sequence of GP63 does not impair the targeting of GP63 to the nucleus . However , this result does not exclude the possibility for the presence of a non-classical NLS in Leishmania GP63 , which may be common or specific to Leishmania species . As Leishmania GP63 is unlikely to use a cNLS-mediated nuclear import ( Fig 2 ) , we investigated other possibility for the protease to achieve nuclear localization . In this context , proteins without cNLS sequences have been shown to trigger nucleocytoplasmic transport through a direct interaction with components of the NPC like Nups [19] and pathogens like viruses have been shown to degrade the NPC and Nups as a way of altering nucleocytoplasmic transport [20] . A sequence screen of Nup proteins for putative GP63 cleavage sites ( using the ScanProsite tool ) , displayed hits for different family members Nup62 , Nup358 , Nup214 and Nup93 ( S1 Dataset ) . Western Blot analysis revealed a degradation of NPC Nup proteins after MΦs were infected either with L . major WT ( Fig 3A ) , or L . mexicana ( S1A Fig ) in a time-dependent manner . Both Leishmania species caused cleavage of Nup proteins after infection . Cleavage of Nup62 was also observable after infections with either L . mexicana or L . major using only low doses of infection ( S2B Fig ) . This emphasizes the likelihood of a Nup62 cleavage and consequently a NPC-degradation during in vivo parasite infections . Interestingly , some Nups were not or weakly affected by Leishmania infection , possibly due to a restricted accessibility within the NPC . Moreover , antibodies detecting Nup62 and Nup358 were able to recognize cleavage fragments that were in accordance with the sequence analysis ( S2C Fig , S1 Dataset ) . Indeed , after sequence analysis of Nup62 and Nup358 for GP63 cleavage site using the ScanProsite Tool ( S2C Fig ) , we found an exact and a potential site for GP63-dependent cleavage in the sequence of Nup62 that would result in fragments of ~ 50kDa and ~ 42kDa , as observed in the S2C Fig . In the case of Nup358 , six exact cleavage sites for GP63 ( and several additional potential ones—S2C Fig ) were found and those may explain the cleavage-dependent smears seen in S2C Fig . GP63 cleavage sites were also identified in Nup214 and Nup93 . To investigate whether NPC fragmentation was a result of GP63 activity , we compared infections of MΦs with L . major WT , GP63-/- , GP63R and L . mexicana parasites ( Fig 3B ) . We observed a substantial degradation of Nups in MΦs infected with L . mexicana , L . major WT and GP63R , but not with L . major GP63-/- , suggesting a pivotal role of GP63 for NPC degradation . This was further supported by the finding that L . major GP63AS was not able to target Nups . Proteolytic cleavage of Nups by GP63 was unaffected after infection of MΦs with L . major GP63NLS parasites ( Fig 3C ) . Confocal microscopy using a Nup62 specific antibody and intensity quantifications confirmed our previous findings ( Fig 3D and 3E ) . Cells infected with L . major WT and GP63R parasites showed a clear reduction of intensity for Nup62 at the NE and inside the nucleus , while cells infected with L . major GP63-/- displayed the same cellular Nup62 distribution as non-infected cells ( Fig 3D ) . Nuclear Nup62 intensity was quantified and results for non-infected cells and L . major WT infected cells were compared . This demonstrated that L . major infection was responsible for a significant reduction of the Nup62 signal at the NE and inside the nucleus ( Fig 3E , S1 Dataset ) . The usage of a large array of virulent Leishmania species and the non-virulent L . tarentolae [21] strain in infection experiments also indicated that the GP63-dependent alteration of the NPC , as shown by Nup62 cleavage , is a conserved phenomenon in virulent Leishmania species ( S2D Fig ) . Taken together , our data strongly suggests a GP63-dependent alteration of NPCs due to the cleavage of Nups after infection , which may affect the nuclear transport and offer GP63 access to nuclear targets . Both previous reports and the preceding results demonstrated that GP63 is able to interfere with the nuclear transport machinery , as well as with the activation of nuclear localized TFs and phosphatases [4 , 5] . To determine to which extent Leishmania and GP63 respectively modify nuclear physiology , whole nuclei ( WN ) and nucleoplasms ( NPs ) were purified from murine LM1 MΦs , after infection with L . major WT , L . major GP63-/- and L . mexicana ( Fig 4A ) . The analysis of the protein content of both purified WNs and NPs revealed global differences ( Fig 4B ) , confirming that Leishmania infections result in extensive alterations within the nuclei . The protein patterns obtained after L . major WT and L . mexicana were similar to each other , while the protein pattern after L . major GP63-/- infection resembled the results obtained for non-infected cells . These findings support the hypothesis that leishmanial GP63 protease activity plays a crucial role in alterations of nuclear proteins after infection . Both WNs and NPs were subjected to LC-MS/MS proteomic analysis and the proteomic data analyzed using the exponentially modified protein abundance index ( emPAI ) , which calculates a ratio of observed to observable peptides , based on factors like mass spectrometry analyses sensitivity , biochemical properties of proteins and published empirical data . The emPAI values are proposed to be linearly correlated to protein concentration [22 , 23] . With these criteria , proteomic analysis on WN identified a total of 932 different proteins: 726 in uninfected Nil samples , 684 in L . mexicana , 694 in L . major WT , and 762 in L . major GP63-/- samples ( S1 Dataset ) . Given our previous results , we focused our analysis of WN on proteins from the NE and/or involved in the nucleocytoplasmic transport ( Fig 4C ) . The proteomics data confirmed that protein levels of Nups ( including Nup62 , Nup93 , Nup214 and Nup358 , also called RanBP2 ) were largely decreased or below detection limit in infected cells compared to non-infected cells ( Fig 4C ) . Interestingly , LC-MS/MS proteomic analysis of WN highlighted the fact that other nucleoporins , such as Nup54 , Nup98 , or Nup107 , and proteins involved in nucleocytoplasmic transport ( RanGAP1 , SUMO1 , Importin alpha-2 , Importin beta-1 ) were also altered in dependency of GP63 ( Fig 4C ) . In addition , levels of proteins from the inner nuclear membrane ( SUN2 ) , the outer nuclear membrane ( Nesprin-1 and 2 ) , and from the lamina ( Lmna , Lmnb1 and 2 ) , which are all part of the LINC complex , connecting the nucleoskeleton to the cytoskeleton and involved in mechanotransduction of extracellular stimuli , were also decreased in the presence of GP63 [24] . We validated the LC-MS/MS results for different proteins identified in WN by western blot analysis , confirming the cleavage of Nup93 , RanGAP1 , and Importin beta-1 ( Kpnβ1 ) in the presence of GP63 ( Fig 4D , middle panel ) . RanGAP1 , a Ran GTPase activating protein , which has been implicated in importin-dependent nucleocytoplasmic transport , exists both free in the cytosol ( 70kDa ) and in a sumoylated form , attached to the NPC ( 90kDa ) via Nup358 . We discovered that GP63 decreases the protein levels of RanGAP1 at the NE ( Fig 4D , middle panel ) . Consequently , an accumulation of potentially sumoylated RanGAP1 was observable in TPLs after infection and only in the presence of GP63 ( Fig 4D , left panel ) , implying that GP63 may mediate the detachment of the sumoylated RanGAP1 from the NPC and Nup358 specifically [25] . In this regard , it remains unclear whether the detachment depends on Nup358 cleavage . Leishmania infections also resulted in diminished Kpnβ1 protein levels in the WN , but only with GP63 present ( Fig 4D , middle panel ) . However , an increase in Kpnβ1 abundance in the cytoplasm was not observable ( Fig 4D , left panel ) . Importin beta-1 facilitates the docking of the Importin alpha/NLS-containing protein ( cargo ) complex at the cytoplasmic side of the NPC . In the presence of nucleoside triphosphates and the small GTP binding protein Ran , the complex enters the NPC and the importin subunits dissociate . While Importin beta remains at the pore , the complex of Importin alpha/cargo protein is transported through the NP . RanGAP1 increases the rate at which Ran hydrolyzes GTP into GDP in the cytoplasm . A decrease of RanGAP1 and Kpnβ1 at the NE would both be consistent with an impairment of nuclear import . Moreover , the analysis of the Nups , RanGAP1 , and Kpnβ1 proteins sequences revealed the presence of putative GP63 cleavage sites ( S1 Dataset ) . In accordance to the data previously introduced , L . major GP63-/- induced changes of protein levels were largely negligible ( Fig 4C , 4D , and 4E ) . Thus , our proteomics data further substantiates our hypothesis that Leishmania is able to alter NPCs and reveals additional targets of the host nuclear transport machinery that are affected after parasite infection . For a further characterization of the impact of GP63 on nuclear physiology of host MΦs , nucleoplasms were extracted and submitted to proteomic analysis . Extraction purity and LC-MS/MS results were validated by western blot analysis ( Fig 4A and 4E ) , showing the cleavage of the importin Kpna3 in NP in the presence of GP63 only ( Fig 4E , middle panel ) . Besides the cleavage of proteins involved in nucleocytoplasmic transport we also observed GP63-dependent mislocalization of proteins such as Kif1B ( Fig 4E left panel and middle panel ) after infection by L . mexicana and L . major WT but not after infection by L . major GP63-/- ( Fig 4E ) . With the criteria mentioned previously , we identified a total of 996 different proteins by LC-MS/MS analysis: 761 proteins in NP of uninfected cells , 653 proteins after infection with L . major WT , 643 proteins after L . major GP63-/- infection and 756 proteins in the case of L . mexicana infection . We considered a difference in emPAI values significant if the change was at least 1 . 5 fold , as frequencies analysis demonstrated that the majority of the proteins were under that range ( Fig 5D and S4D ) . For an in-depth analysis of proteins affected by Leishmania infection we utilized the STRING-software to generate functional clusters of altered proteins , using gene ontology ( GO ) annotations . In two separate batches of analyses , we investigated proteins that exhibited changes in abundance in the NP in dependency of GP63 expression ( Figs 5 and 6 , S3 , and S2 Dataset ) and compared the differences of the nucleoplasmic protein content after infection with L . major WT and L . mexicana ( Figs 7 , S4 , S5 , and S3 Dataset ) . In order to confirm that GP63 was involved in host nuclear processes , we compared LC-MS/MS results obtained for nuclei of Nil , L . major WT and L . major GP63-/- samples ( Fig 5 ) . The emPAI values of all the proteins identified for L . major WT or L . major GP63-/- samples were displayed in regard to the proteins found in the control sample ( n = 820—Nil emPAI values are represented in ascending order ) ( Fig 5A and 5B ) . The majority of the proteins identified were detectable in all samples ( 544 ) . However , depending on the infection status the abundance of individual proteins often differed . Accordingly , our results demonstrate that most alterations after Leishmania infection were decreases in protein abundance and at least in part relied on GP63 . Furthermore , L . major WT samples featured more and stronger changes of protein levels than L . major GP63-/- samples ( Fig 5A and 5B ) . This is also illustrated through the observation that 37 proteins were exclusively detected in L . major WT samples , while only 11 proteins were unique to L . major GP63-/- samples ( Fig 5C ) . Taken together , 20 . 5% of proteins detected in non-infected conditions were absent in L . major WT samples , while 18 . 4% of proteins were absent in L . major GP63-/- ( Fig 5D ) . The importance of GP63 was substantiated by direct comparison of nucleoplasmic proteins levels after L . major WT and L . major GP63-/- infection of MΦs . L . major WT infection clearly induced the appearance of a substantial number of proteins and the extent of changes in protein abundance in L . major WT samples differed markedly ( Fig 5E ) . Thus , the presence of GP63 seems to be a prerequisite for a number of alterations within the NP . Infection with L . major strains ( WT and GP63-/- ) primarily resulted in a reduction of protein levels in the NP ( 339 ) . An increase in NP protein levels was less prevalent ( 121 ) ( Fig 6A ) . Among the proteins showing a GP63-dependent increase in abundance ( 65 ) , 37 proteins were not present at all in NP of uninfected cells ( S2 Dataset ) , again indicating a mislocalization of proteins due to alterations to the NE and the nucleocytoplasmic transport machinery . Using the STRING software and GO annotations , we identified several functional protein clusters which are involved in chromatin remodeling ( Actl6a , Smarca4 , Smarca5 , Smarcc2 , Hdac1 , Ruvbl1 ) , nuclear RNA splicing and export ( Sfrs4 , Hnrnpm , U2af1 , Ncbp1 , Nxf1 ) and co-translational modification ( Ddost , Mogs , Rpn1 , Rpn2 , Srp68 ) ( Fig 6B and 6C ) . Interestingly , proteins of the NE as well as proteins of the endoplasmic reticulum and the endomembrane system were found in NP after L . major WT infection , possibly due to the activity and protein degradation-mediated by GP63 ( Fig 6B ) . The comparison of L . major WT and L . major GP63-/- samples revealed a GP63-dependent decrease of nucleoplasmic protein levels ( 91 ) ( Fig 6A and 6D , S3 , and S2 Dataset ) . STRING software and GO annotations revealed that a number of NP-proteins , which exhibited a GP63-dependent decrease in abundance , were involved in ribosome assembly ( Rpl10 , Rpl36 , Rpl38 ) or subunits of the eukaryotic translation initiation factor ( Eif2s3x , Eif2s3y , Eif3f ) . Furthermore , ~40 different histones were identified ( Fig 6D ) . Although the comparative LC-MS/MS analysis of host nucleoplasms after L . major WT and L . major GP63-/- infection identified the protease GP63 as a key factor for the alterations of nuclear physiology , we identified 157 proteins with altered protein levels in both L . major WT and L . major GP63-/- samples ( Fig 6A ) . These changes comprised alterations of ATP-dependent mechanisms including enzymatic metabolic processes , as well as nucleic acid metabolism , chromosome organization ( helicase activity ) , gene expression ( TF activity / nucleolus , spliceosome ) and translation ( translation factor activity , ribosome ) ( S3 Fig , S2 Dataset ) . Therefore , it is likely that Leishmania has GP63-independent means to alter nuclear protein levels as well . In our experiments , we further aimed to elucidate the impact of different Leishmania species on the MΦ NP . Thus , we compared infections with L . major WT and L . mexicana parasites . We subjected the proteomic data obtained for the L . major WT and L . mexicana samples to the same comparative analyses as previously performed for the comparison of L . major WT and L . major GP63-/- samples ( S4 Fig ) . The emPAI values of all the proteins identified for L . major WT and L . mexicana samples were displayed in consideration of the proteins found in the control sample ( n = 920—Nil emPAI values in ascending order ) ( S4A and S4B Fig ) . We detected 552 proteins with changed protein levels , which were present in all infected samples . A closer examination of our data revealed that L . mexicana infection was able to trigger a variety of alterations , which were not present after L . major infection ( S4A and S4B Fig ) . Indeed , we found 111 proteins whose abundance was only changed after L . mexicana infection ( 111/157 proteins ) , with only two proteins uniquely changed in L . major WT samples ( 2/48 ) . ( S4C Fig ) . In addition , it is noteworthy that not only the number of altered proteins is higher in L . mexicana samples but recorded changes were more extensive ( S3D Fig and S4E ) . This finding possibly reflects the fact that L . mexicana is considered more virulent than L . major WT . Therefore , our comparative studies indicate that L . mexicana parasites alter nuclear physiology to a larger extent than L . major parasites . We then performed STRING and GO analyses of the proteins identified in both L . mexicana and L . major WT samples ( Figs 7 , S5 , and S3 Dataset ) . In L . mexicana samples 258 proteins were detected that either appeared or increased in abundance . This group comprised 85% of the proteins detected in L . major WT samples ( 73/86 ) ( Fig 7A ) . Consequently , we were able to identify similar clusters as in our previous analysis for L . major WT with new groups unique for L . mexicana infections ( Figs 6B , 6C , and 7B ) . Thus , the created STRING biological network revealed that both L . mexicana and L . major parasites act on chromatin remodeling and transcription regulation . For L . mexicana samples we identified more detailed clusters distinguishing mRNA , tRNA and rRNA processes ( Fig 7B ) . GO annotations also indicated that nucleic acid metabolic processes were still the predominant functions altered by both Leishmania species , but transport activity , signaling and response to stimuli were also strongly affected after L . mexicana infection ( S4 Fig ) . Newly identified clusters consisted of Nups and mitochondrial proteins ( Fig 7C ) further substantiating a Leishmania-dependent dysregulation of the nucleocytoplasmic transport machinery . Detailed GO annotations confirmed the identification of eight Nups after infection with L . mexicana possibly due to the effect of GP63 ( Figs 2 and 3 ) . The identification of proteins involved in mitochondrial processes within the nucleus after L . mexicana infection indicates that L . mexicana may also affect the integrity of other cellular organelles ( Figs 7D , S4 , and S3 Dataset ) . The groups of proteins with diminished protein levels after either L . major WT or L . mexicana infection were partly overlapping , too . 194 proteins were identified in both L . mexicana and L . major WT samples , while 119 were unique to L . mexicana , and 131 were unique to L . major WT samples ( Fig 7D ) . STRING and GO analyses highlighted the possibility of a parasite-dependent interference with proteins involved in rRNA maturation and ribosome assembly , as well as translation initiation . However it is interesting to note that L . major parasites seemed to act mostly on histones , while L . mexicana parasites impacted strongly on pre-mRNA splicing and consequently mRNA export and translation ( Figs 7D , S5 , and S3 Dataset ) . Together , our data and the observation that GP63 can affect the protein content of host nuclei may offer new explanations for a multitude of observations previously published , including the inhibition of cellular translation and the downregulation of host protective mechanisms , as well as effects on cell proliferation and nucleotide metabolism [5 , 26 , 27] . Although our study suggests GP63 is of pivotal importance for Leishmania-induced changes of nuclear physiology , our results did not exclude the possibility of a direct or indirect dependency on the activity of additional leishmanial proteins . To identify other Leishmania factors within the host cell nuclei , we subjected our proteomic samples to the Leishmania database . Overall , we detected 94 parasite proteins within host MΦ nuclei . Interestingly , beside GP63 , no other known Leishmania virulence factors were identified ( S1 Dataset ) , thus substantiating the critical role of GP63 for the alterations of nuclear physiology . Leishmania-mediated subversion of signaling pathways involving protozoan virulence factors has been a subject of high interest in recent years . Although primary indications of Leishmania-mediated effects on nuclear physiology existed , no further analysis was carried out to date . For the first time we here present an in-depth proteomic analysis—both quantitative and qualitative—of host MΦ nuclei after Leishmania infection with a special focus on the implications of the leishmanial protease GP63 . Herein , we provide evidence that the metalloprotease GP63 , a key virulence factor of Leishmania parasites , can localize to the NE of host MΦs . With the determinants for GP63 localization within MΦs unknown , we identified a putative cNLS-like sequence within the GP63 sequence of several Leishmania species . Classical nuclear import via an NLS sequence is one of the major routes to deliver proteins to the nucleus in higher eukaryotes . Nevertheless , mutation of the cNLS-like sequence did not result in an alteration of GP63-targeting to the NE . Although we have not ruled out the possibility that a non-classical NLS is present in the GP63 sequence , only few reports describe such sequences in protozoan trypanosomatids like Leishmania [28 , 29] . This could support our results that GP63 reaches the NE independently of the identified putative classical NLS sequence , and utilizes a different yet unknown pathway to target the nucleus . A possibility may be an import factor-independent mode of translocation . In this context , some proteins , like β-catenin , have been shown to enter the nucleus via the NPC by an NLS/importin-independent mechanism , where they recognize and bind Nups similar to importin family proteins . These proteins contain specific motifs called Armadillo or HEAT repeats that are recognized by FXFG-motifs of Nups [19 , 30 , 31] . A preliminary blast analysis revealed partial matches for Armadillo/HEAT repeats in the amino acid sequence of GP63 . In the future , site-directed mutagenesis studies of these motifs could further elucidate whether or not these sequences are determinants of GP63 localization within host cells . It is also possible that the substrate recognition process for Nups itself may play a role in the targeting of GP63 to host MΦ nuclei . In any case the localization of the protease seems to enable an alteration of both nuclear transport proteins and proteins within the nucleoplasm with potentially drastic consequences for key nuclear processes . Proteomic analysis of L . major and L . mexicana infected nuclei finally revealed the extent of parasite-dependent alterations of host nuclear physiology and specifically the impact of GP63 . Indeed , after both L . major and L . mexicana infection , we observed alterations in the localization and abundance of Nups from the NPC , as well as of different components of the NE , the nucleocytoskeleton , and the nucleocytoplasmic transport machinery . Thus , our proteomic data on WN and NP were in accordance with our results obtained by molecular biology methods . The observed alterations of the NE , the nucleocytoskeleton , and the nucleocytoplasmic transport machinery may very well represent a parasite survival strategy that is a consequence of GP63-passage through the NPC . Although corresponding modification of the nuclear transport machinery and nuclear physiology have not yet been described for any trypanosomatid parasites , our results correlate with studies demonstrating that viruses are able to disrupt the NE or alter the composition and function of the NPC for their own survival and propagation . For instance , the alteration of nucleocytoplasmic transport—including the mislocalization of proteins to the nucleus—has been shown in the case of picornavirus , rhinovirus and poliovirus infections . In these cases the changes have been linked to changes in nucleocytoplasmic transport , signal pathway alterations and an impaired immune response [32] . Interestingly , the mislocalization of proteins to the nucleus and defects in nucleocytoplasmic transport after rhinovirus and poliovirus infections were attributed to the degradation of Nups ( Nups 62 , 153 , 214 , and 358 ) , resembling the results we obtained for GP63-dependent cleavage of NPC-proteins [10 , 33–35] . Here we showed that GP63 seems to act in a similar fashion as proteases from different viruses that have been shown to degrade Nups , including Nup62 , to interfere with host nuclear transport and other host nuclear mechanisms . However , some viruses have been known to utilize interactions with host Nups to their own benefit . For instance , HIV-1 has been shown to specifically use the mislocalization of the host Nup62 to increase HIV-1 gene expression and infectivity [11 , 36] . During Herpes simplex virus infections , viral ICP27 directly interacts with the nuclear pore complex through Nup62 , inhibiting host nucleocytoplasmic transport pathways [37] . Consequently , in this context purified Leishmania GP63 could represent an interesting approach for a treatment interfering with these mechanisms of virulence . The comparison of our proteomic data for L . major WT and L . major GP63-/- infection provides further evidence that GP63 is indeed strongly involved in a wide variety of Leishmania-induced changes of nuclear physiology besides nucleocytoplasmic transport . This is illustrated by the similarities in our proteomic analysis and comparison of NPs after L . major WT , L . major GP63-/- and L . mexicana infection . Both virulent Leishmania species affect whole protein clusters that are of importance for or act in chromatin remodeling , RNA-related processes ( transcription , splicing , and export ) , nucleoskeleton and ribosome maturation . In this regard it is noteworthy that L . major and L . mexicana led to a diminished abundance of several subunits forming the eukaryotic initiation Translation Factor ( eiTF ) . This is most likely mediated by GP63 as depicted by the proteomic data obtained after L . major GP63-/- infection . The finding of GP63 interfering with various eiTF subunits corroborates the results of a previous study . There , L . major GP63 was shown to inhibit host protein translation through the cleavage of mTOR [26] . Thus , the direct GP63-mediated cleavage of eIF proteins may present a second parasitic strategy to impair cellular translation after infection . Our results also indicated that GP63 is involved in the alteration of proteins from the LINC complex ( Sun , Nesprins ) and the nuclear lamina ( lmna , lmnb1 , lmnb2 , lmnbr ) , both crucial for nucleus integrity . Moreover , the LINC complex has also been shown to be involved in the mechanotransduction of extracellular stimuli [24] , which may represent yet another novel process for Leishmania to evade host protective mechanisms . In addition , lamins and their associated proteins are also involved in other nuclear functions besides the maintenance of the shape and the mechanical strength of the nucleus: chromatin organization , DNA replication , transcription regulation , RNA processing and the linkage of the nucleus to all major cytoskeleton networks [38] . In any case the diversity of the nuclear targets of GP63 substantiates the impact of GP63s’ proteolytic activity on host MΦ function and nuclear physiology . Our results concerning L . major WT , L . major GP63-/- and L . mexicana infections do not exclude the possibility of additional proteins involved in the nuclear physiology alteration . To our knowledge , a nuclear localization within the host cell after infection has only been proposed for a hypothetical leishmanial protein from L . pifanoi and L . donovani . But both protein function and localization could not be confirmed to date [39] . Although our proteomic analysis of nuclei after infection identified various leishmanial proteins that gain access to the nucleus , none of them are parasite virulence factors . In conclusion we herein show through different methodologies that Leishmania GP63 can target the perinuclear region of host cells . At the nuclear envelope the parasite protease is able to alter host nucleoporins , the nucleocytoskeleton , and the nucleocytoplasmic transport through proteolytic cleavage . Moreover , our proteomic data sets greatly increase our understanding how parasites , specifically through GP63 , may impact on both host gene expression and translation and our data may explain observations of previous studies in this regard . Furthermore , our data reveals a parasite-mediated overall interference with key processes of nuclei , providing novel leads , as the basis for future functional studies , on how Leishmania can potentially subvert host functions to their own benefit . In addition to the cleavage of host transcription factors and the PTP-mediated signaling hijacking , our study strongly suggest that Leishmania infections are likely to cause the shutdown of a wide range of integral host cell nuclear and cytoplasmic functions possibly to ensure parasite survival and dampening of anti-leishmanial immune responses . Thus , GP63 arguably represents a possible approach to be included in future research for efficient treatments of Leishmania infections as its inhibition could negate the parasites ability to subvert various host-protective functions , not only the dysregulation of protein phosphorylation and inhibition of ROS and NOS production , but also intracellular transport mechanisms , gene expression and translation as presented herein . This is further emphasized by the finding that GP63 can confer resistance against antimicrobial peptides [40 , 41] . Our findings , and given the importance of GP63 for the subversion of host protective mechanisms early during infection , strengthen approaches that try to introduce either the DNA or the metalloprotease itself in vaccination studies . Generally those approaches have shown elevated success [42–44] . The immortalized murine bone marrow derived MΦ LM1 cell line ( generated in our lab [45] ) was maintained in culture at 37°C in 5% CO2 in Dulbecco’s Modified Eagle Medium ( DMEM ) supplemented with 10% heat inactivated FBS and antibiotics ( Penicillin 100 U/ml—Streptomycin 100 μg/ml ) . Leishmania promastigotes ( L . major WT , L . major GP63-/- , L . major GP63R , L . major GP63NLS mutant , L . major GP63AS mutant , L . mexicana , L . donovani infantum , L . donovani donovani , L . amazonensis , L . tarentolae ) were grown and maintained at 25°C in SDM-79 culture medium supplemented with 10% heat inactivated FBS . L . major GP63R , L . major GP63NLS and L . major GP63AS mutant parasites were selected with geniticin G418 ( Sigma-Aldrich , Oakville , ON , Canada , 50 ng/ml ) . MΦs were infected at a parasite to MΦ ratio of 20:1 with stationary phase promastigotes for the indicated times . To generate total parasite lysates , promastigotes were collected at day 7 by centrifugation , washed 3 times in PBS and lysed with cold buffer ( 40 mM Tris HCl , 275 mM NaCl , Glycerol 20% , Igepal 1% , 1X protease inhibitor cocktail ( PIC ) ; Roche , Mississauga , Ontario , Canada ) . To generate cell lysates after infection , LM1 cells were washed 3 times with PBS and submitted to either total protein lysis or nuclear separation . For total protein lysates , cells were lysed with cold lysis buffer ( 50 mM Tris pH 7 . 0 , 0 . 1 mM EDTA , 0 . 1 mM EGTA , 1% Igepal , 0 . 1% 2-Mercaptoethanol , 1X PIC ) . Proteins were quantified by Bradford assay ( Bio-Rad , Mississauga , Ontario , Canada ) . For nuclear separation , we adapted a protocol from Matunis et al . [46] . Washed and recovered LM1 cells were resuspended in cold buffer A ( 0 . 25 M sucrose , 25 mM KCl , 5 mM MgCl2 , 50 mM Tris pH7 . 5 , 1X PIC ) . After centrifugation ( 800g , 4°C , 20 min ) , the cytoplasmic fraction ( supernatant ) was discarded and the nuclei fraction ( pellet ) was resuspended in cold buffer A . Subsequently , the nuclei fraction was purified using a sucrose gradient ( 2M and 1 . 5M sucrose solutions ) by ultracentrifugation ( 31000 rpm , 4°C , 3 hrs ) . The isolated nuclei were resuspended in buffer A and counted on hemacytometer ( 1×106 nuclei = 15μg proteins ) . To extract NE , whole nuclei were pelleted ( 5 min , 2500 rpm , 4°C ) , lysed ( 500 μl of buffer containing 0 . 1 mM MgCl2 , 1 mM DTT , 5 μg/ml DNaseI , 5 μg/ml RNase A , 1X PIC ) and resuspended in 2 ml of extraction buffer pH 8 . 5 ( 10% sucrose , 20 mM triethanolamine pH 8 . 5 , 0 . 1 mM MgCl2 , 1 mM DTT , 1X PIC ) . Nuclei were underlayed with sucrose cushion solution ( 30% sucrose , 20 mM triethanolamine pH 7 . 5 , 0 . 1 mM MgCl2 , 1 mM DTT , 1X PIC ) and NEs were pelleted ( 4000g , 15 min , 4°C ) . The NE pellet was resuspended in 500μl of extraction buffer pH 7 . 5 ( 10% sucrose , 20 mM triethanolamine pH7 . 5 , 0 . 1 mM MgCl2 , 1 mM DTT , 1X PIC ) , followed by 50 μl of extraction buffer pH 7 . 5 containing 0 . 3 mg/ml Heparin . The NEs were underlayed with sucrose cushion solution , spun ( 30 min , 4000g , 4°C ) and resuspended in extraction buffer pH 7 . 5 . To extract NP , whole nuclei were pelleted ( 10 min , 4°C , 5000 rpm ) , resuspended in buffer C ( 20 mM HEPES pH 7 . 9 , 0 . 4 M NaCl , 1 mM EDTA , 1 mM EGTA ) and incubated for 20 min at 4°C . NP was collected by centrifugation ( 13000 rpm , 4°C , 15 min ) . NE and NP samples were dosed by Bradford assay . Protein extracts were treated as previously described [5] . Briefly , between 25 and 40μg of proteins were separated by SDS-PAGE ( 8% , 10% or 12% acrylamide ) , and transferred to PVDF membranes . Membranes were blocked in Tris Buffered Saline and tween 0 . 1% ( TBS-T ) containing 5% BSA for 1 hr and incubated either 2 hrs at room temperature or over-night at 4°C with primary antibody . After washing with TBS-T ( 2 times for 5min ) , membranes were incubated 1 hr with secondary anti-HRP-conjugated antibody ( GE Healthcare , Mississauga , ON , Canada ) . After washing with TBS-T ( 3 times for 5min ) , they were developed by chemoluminescence immunodetection with ECL reagents ( Thermo Fisher Scientific , Rockford , IL ) and autoradiography . List of antibodies used: GP63 monoclonal antibody clone #253 ( Button et al . 1991 ) ; Histone H2B ( Genscript corporation , A01174 ) ; KDEL ( ab12223 ) , SHP-1 ( ab3254 ) , NPC Mab414 ( ab24609 ) , Nup62 ( ab96134 ) , Nup358 ( RanBP2 , ab64276 ) , KpnB1 ( ab2811 ) , KpnA3 ( ab105348 ) from Abcam; PGK1 ( ProteinTech , 17811-1-AP ) ; Actin ( Sigma , A5316 ) ; TCPTP ( MediMabs , MM-0018-P ) ; NF-κB p65 ( SC-8008 ) , Nup93 ( SC-374400 ) , Nup214 ( SC-26055 ) , RanGAP1 ( SC-1862 ) , Kif1B ( SC-28540 ) from Santa Cruz; C-Jun ( Cell Signaling , 60A8 ) . Protease activity of GP63 was assayed as previously described by 10% SDS-PAGE incorporated with gelatin ( 1mg/ml ) [47] . With some modifications to the previous protocol , the gels were loaded with parasite extracts ( 10μg of proteins ) that were added to SDS-PAGE sample buffer ( 15 . 6mM Tris pH6 . 8 , 2% SDS , 10% glycerol , 0 . 05% Bromophenol Blue ) . Electrophoresis was performed at a constant current of 20mA at room temperature . After electrophoresis , SDS was removed by incubation with washing buffer ( 2 . 5% Triton X-100 in 50mM Tris pH 7 . 4 , 5mM CaCl2 , 1μM ZnCl2 ) for 1 hr on a rotating shaker at room temperature . Then , the gels were briefly rinsed twice with deionized water and incubated in a renaturation buffer containing 50mM Tris pH 7 . 4 , 5mM CaCl2 , 1μM ZnCl2 , over-night at 37°C . After incubation , gels were stained 30 min in 0 . 5% Coomassie Brilliant Blue R-250 in 30% ethanol and 10% acetic acid , and destained for several hours in a solution containing 30% ethanol and 10% acetic acid . Active GP63 was detected as clear bands on the gel . In-gel PTP assay was performed as previously described [3] . Briefly , poly ( Glu , Tyr ) substrate was tyrosine-phosphorylated by overnight ( O/N ) incubation with GST-FER protein kinase ( 10μg ) and 150 μCi [γ-32P]dATP . The substrate was then incorporated in a 10% SDS-polyacrylamide gel at a concentration of 2×105 cpm/ml . Mϕ protein extracts , prepared as described above , were denatured for SDS-PAGE and loaded onto the gel . After electrophoresis , the gel was incubated O/N in the fixative buffer A ( 50mM Tris-HCl pH 8 . 0 , 20% isopropanol ) , washed twice 30 min with buffer B ( 50mM Tris-HCl pH 8 . 0 , 0 . 3% β-ME ) , and followed by full protein denaturation in buffer B containing 6M guanidine hydrochloride and 1mM EDTA . Gels were washed twice 1 hr in buffer C ( 50mM Tris-HCl pH 7 . 4 , 1mM EDTA , 0 . 3% β-ME , and 0 . 04% Tween 20 ) and incubated for final renaturation O/N in Buffer C . Gels were dried and exposed to x-ray film . Active PTPs were detected as clear bands on the film . The day before the infection , cells were plated in 24 well-plates on poly-L-lysine coated and UV sterilized coverslips . After 2 hrs of infection , cells were fixed , stained with DAPI to visualize nuclei , and anti-Nup62 antibody to detect the nucleoporin , and the fluorescence was quantified ( 65 non-infected cells vs . 36 infected cells—S1 Dataset ) as previously described [48] . In parallel , as a control , Leishmania-infected macrophages were stained with anti-GP63 antibody to show the perinuclear localization of GP63 within the cells , as described previously [5] . LC-MS/MS analysis was carried out as described before [49] . Protein database searching was performed with Mascot 2 . 2 ( Matrix Science ) against the NCBI Mus musculus and Leishmania protein databases . The mass tolerances for precursor and fragment ions were set to 10 ppm and 0 . 6 Da , respectively . Trypsin was used as the enzyme allowing for up to 2 missed cleavages . Carbamidomethyl and oxidation of methionine were allowed as variable modifications . Duplicates of separately analyzed sets of MS/MS data were used for calculation of the exponentially modified Protein Abundance Index ( emPAI ) values using emPAICalc web server ( http://empai . iab . keio . ac . jp/ ) . Mascot output files were uploaded to emPAICalc server and hits with a minimum of 3 peptides and a minimum score of 20 were chosen as true hits for further analyses . Gene Ontology ( GO ) annotations of identified proteins were extracted and protein-protein interaction networks of the identified proteins were created using STRING database with specific parameters ( action view , high confidence—score 0 . 7 ) [50] .	Unicellular parasites of the genus Leishmania are the causative agent of leishmaniasis , a disease affecting 12 million people worldwide , mainly in tropical and subtropical regions of the developing world . They have evolved strategies to circumvent cellular defense mechanisms favouring their survival . This includes the cleavage and activation of proteins and the subsequent block of signals within the host cells . In this study we discovered that a Leishmania virulence factor , GP63 , is able to reach host cell nuclei and affect protein transport from and into the nucleus . Through the analysis of the protein content of nuclei after parasite infection we revealed that Leishmania , predominantly through the protein cleaving enzyme GP63 , can alter several processes within the nucleus , amongst others mechanisms associated with gene expression and nucleic acid metabolism . Thus , we here introduce a novel strategy of how Leishmania parasites may overcome host cell defense and ensure their own survival .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[]	2015	Impact of Leishmania Infection on Host Macrophage Nuclear Physiology and Nucleopore Complex Integrity
Survival at host temperature is a critical trait for pathogenic microbes of humans . Thermally dimorphic fungal pathogens , including Histoplasma capsulatum , are soil fungi that undergo dramatic changes in cell shape and virulence gene expression in response to host temperature . How these organisms link changes in temperature to both morphologic development and expression of virulence traits is unknown . Here we elucidate a temperature-responsive transcriptional network in H . capsulatum , which switches from a filamentous form in the environment to a pathogenic yeast form at body temperature . The circuit is driven by three highly conserved factors , Ryp1 , Ryp2 , and Ryp3 , that are required for yeast-phase growth at 37°C . Ryp factors belong to distinct families of proteins that control developmental transitions in fungi: Ryp1 is a member of the WOPR family of transcription factors , and Ryp2 and Ryp3 are both members of the Velvet family of proteins whose molecular function is unknown . Here we provide the first evidence that these WOPR and Velvet proteins interact , and that Velvet proteins associate with DNA to drive gene expression . Using genome-wide chromatin immunoprecipitation studies , we determine that Ryp1 , Ryp2 , and Ryp3 associate with a large common set of genomic loci that includes known virulence genes , indicating that the Ryp factors directly control genes required for pathogenicity in addition to their role in regulating cell morphology . We further dissect the Ryp regulatory circuit by determining that a fourth transcription factor , which we name Ryp4 , is required for yeast-phase growth and gene expression , associates with DNA , and displays interdependent regulation with Ryp1 , Ryp2 , and Ryp3 . Finally , we define cis-acting motifs that recruit the Ryp factors to their interwoven network of temperature-responsive target genes . Taken together , our results reveal a positive feedback circuit that directs a broad transcriptional switch between environmental and pathogenic states in response to temperature . Cells adapt to their environment by responding to specific environmental stimuli such as light , temperature , and nutrients . For microbial pathogens , mammalian body temperature can signal the induction of pathways required for host colonization and pathogenesis [1] . One such group of organisms is the thermally dimorphic fungal pathogens , which include Coccidioides , Paracoccidioides , Blastomyces , and Histoplasma species . These evolutionarily related fungi are notable among fungal pathogens in that they all cause disease in healthy individuals [2] . Each of these organisms grows in a mold form in the soil , forming long , connected filaments that produce vegetative spores [3] . When the soil is aerosolized , filamentous cells and spores can be inhaled by mammalian hosts and converted into a parasitic form within the host lung . Conversion entails a dramatic change in cell shape to a budding yeast form for the majority of these pathogens , as well as the transcriptional induction of virulence genes required to cause disease in the host [3] . For all thermally dimorphic fungi , host temperature is the key signal that triggers this developmental switch , but little is known about the coordinated induction of morphologic changes and virulence gene expression by temperature . Histoplasma capsulatum , which is endemic to the Ohio and Mississippi River Valleys of the United States , can cause life-threatening respiratory and/or systemic disease ( histoplasmosis ) [2] , [4] . It is estimated that up to 25 , 000 people develop life-threatening infections in endemic regions each year , with at least 10-fold more mild or asymptomatic infections [2] , [4] . Although the pathogen propagates as spores and in a filamentous form in the environment , H . capsulatum is found almost exclusively in the yeast form within mammalian hosts . Despite the prevalence of H . capsulatum and its threat to human health , we have a limited understanding of the transcriptional regulatory network that governs pathogenic yeast-phase growth . Previously , we identified three regulators , Ryp1 , Ryp2 , and Ryp3 , and showed that they are required for yeast-phase growth [5] , [6] . Whereas wild-type cells grow in the yeast form at 37°C , ryp1 , ryp2 , and ryp3 mutants grow constitutively in the filamentous form independent of temperature . In wild-type cells , RYP1 , RYP2 , and RYP3 transcripts and proteins accumulate preferentially at 37°C and each Ryp protein is required for the wild-type expression levels of the others [5] , [6] . RYP1 encodes a fungal-specific transcriptional regulator that is required for modifying the transcriptional program of H . capsulatum in response to temperature [5] . Ryp1 belongs to a conserved family of fungal proteins that regulate cellular differentiation in response to environmental signals . The best-studied member of this family of proteins is Wor1 , which was identified as a master transcriptional regulator that controls a morphological switch required for mating in Candida albicans [7]–[9] . In the model yeast Saccharomyces cerevisiae , the Ryp1 ortholog , Mit1 , is required for a morphologic switch that occurs under nutrient limitation [10] . Ryp1 orthologs in the plant pathogens Fusarium oxysporum ( Sge1 ) , Fusarium graminerium ( Fgp1 ) , and Botrytis cinerae ( Reg1 ) are required for full pathogenicity and conidiation [11]–[13] . All of these observations signify the importance of Ryp1 orthologs for transduction of environmental cues to regulate cell morphology and virulence . Furthermore , it was recently demonstrated that Wor1 contains a DNA-binding domain that is conserved throughout the WOPR ( Wor1 , Pac2 , Ryp1 ) family of proteins [14] , suggesting that these regulators respond to specific signals by triggering a transcriptional program . In contrast , Ryp2 and Ryp3 belong to the Velvet family of regulatory proteins [6] , whose molecular function is unknown . This family is typified by Velvet A ( VeA ) , which was initially characterized as a regulator of sexual spore production in Aspergillus nidulans [15] , [16] , but is now known to also regulate secondary metabolism and development in many fungi including Aspergillus species , Fusarium species , Neurospora crassa , and Acremonium chrysogenum ( reviewed in [17] ) . In H . capsulatum , the VeA ortholog Vea1 has a role in sexual development but is dispensable for yeast-phase growth [18] . Additionally , many fungi have multiple Velvet family proteins that collaborate to serve regulatory functions . For example , in A . nidulans , three Velvet family proteins ( the Ryp2 ortholog VosA , the Ryp3 ortholog VelB , and VeA itself ) act together to regulate asexual and sexual development and secondary metabolism [19] . Notably , since Velvet family proteins do not contain canonical DNA binding domains or other domains of known function , their mechanistic role in regulation of developmental processes is unclear . As noted above , both WOPR and Velvet family proteins are widely distributed among fungi , although the Hemiascomycetes , including Saccharomyces and Candida species , lack Velvet family proteins . Since both families of proteins are required for yeast-phase growth in H . capsulatum , we explored if and how these two distinct classes of fungal regulators work together to govern temperature-responsive traits by dissecting the Ryp regulatory network in H . capsulatum . To this end , we performed whole-genome transcriptional profiling and chromatin immunoprecipitation experiments to determine the shared and unique roles of Ryp1 , Ryp2 , and Ryp3 in regulating yeast-phase growth . We show that 96% of yeast-phase enriched transcripts are dependent on Ryp1 , Ryp2 , and Ryp3 for their enhanced expression in response to temperature , whereas 66% of filamentous-phase enriched transcripts require Ryp1 , Ryp2 , and Ryp3 to prevent their inappropriate expression at 37°C . We demonstrate that all three Ryp factors physically interact and associate with the upstream regions of a core set of target genes , including those required for yeast-phase growth and virulence . Additionally , we identify a fourth transcriptional regulator , Ryp4 , to be a component of the Ryp regulatory network required for temperature-responsive yeast-phase growth . Finally , the identification of two distinct cis-acting regulatory sequences that are bound and utilized by Ryp proteins provides the first evidence that highly conserved Velvet family proteins can directly bind to DNA and activate gene expression using a unique cis-acting element . Overall , our results provide a molecular understanding of how regulation of cell morphology and virulence gene expression is coordinated in response to temperature in H . capsulatum . Our previous studies showed that there are marked differences in the transcriptional profiles of wild-type yeast-form cells grown at 37°C and filamentous cells grown at room temperature [5] , [20] . Cells lacking RYP1 , RYP2 , or RYP3 grow constitutively as filaments independent of temperature [5] , [6] , and Ryp1 is required for the expression of the majority of the transcripts enriched during yeast-phase growth at 37°C [5] . Here we sought to understand whether Ryp2 and Ryp3 are also involved in regulating expression of genes required for yeast-phase growth . To this end , we performed whole-genome expression profiling experiments comparing the transcriptional profiles of multiple biological replicates of ryp1 , ryp2 , ryp3 mutants and wild-type strains grown at room temperature ( RT ) and 37°C . We identified 388 genes with significantly increased transcript levels and 376 genes with significantly decreased transcript levels in wild-type yeast cells grown at 37°C compared to wild-type filaments grown at RT ( Figure 1A and Table S1 ) . These gene sets were referred to as yeast-phase–specific ( YPS ) and filamentous-phase–specific ( FPS ) genes , respectively . Gene Ontology ( GO ) analysis showed that YPS genes were enriched in terms such as cell cycle regulation , chromosome segregation , and DNA recombination , all of which are characteristics of metabolically active , budding yeast cells ( Figure 1C ) . In contrast , FPS genes were enriched in terms such as cell wall , extracellular region , response to oxidative stress , hydrolase , and lipase activity , consistent with the idea that filamentous cells have a distinct cell wall structure and produce enzymes related to saprophytic activity ( Figure 1C ) . Additionally , YPS genes included the previously identified virulence genes of H . capsulatum: Genes encoding for calcium-binding protein 1 ( CBP1 ) [21] , yeast-phase specific protein 3 ( YPS3 ) [22] , super-oxide dismutase 3 ( SOD3 ) [23] , alpha- ( 1 , 4 ) -amylase ( AMY1 ) [24] , and heat-shock protein 90 ( HSP82 ) [25] were among the most differentially regulated YPS genes . Furthermore , YPS genes included M antigen/catalase B ( CATB ) , which has been shown to display yeast-specific expression in the H . capsulatum strain used here and encodes a secreted catalase that may help cope with oxidative stress in the host [26] . Transcript levels of these virulence genes were significantly down-regulated in ryp mutants , showing that Ryp factors are important for expression of virulence genes in H . capsulatum in response to temperature ( Figure 1B ) . Global comparison of the gene expression profile of wild-type cells to that of each ryp mutant revealed that the ryp mutants had a strongly diminished transcriptional response to temperature . The transcriptome of each ryp mutant grown at 37°C strongly resembled the transcriptome of wild-type cells grown at RT ( Figure 1D ) . Additionally , transcriptional profiles of the individual ryp mutants were strikingly similar to each other ( Figure 1A ) , indicating that they may act in the same temperature-responsive pathway . Transcript levels of the overwhelming majority of the YPS genes ( 96% ) were decreased ( >1 . 5-fold ) in ryp mutants at 37°C , indicating that Ryp proteins were required for their wild-type expression level . Similarly , the Ryp proteins were required to prevent inappropriate expression of the majority of the FPS genes at host temperature: Transcript levels of 66% of the FPS genes were increased ( >1 . 5-fold ) in ryp mutants compared to the wild-type strain at 37°C . These results showed that Ryp1 , Ryp2 , and Ryp3 are master regulators required for the appropriate temperature-responsive transcriptional program in H . capsulatum . As observed previously , our set of YPS genes included RYP1 , RYP2 , and RYP3 , and each of the RYPs depended on the other two for its temperature-regulated expression ( Figure S1B ) [6] . Since RYP transcript levels are low at RT under laboratory conditions , we expected that the Ryp proteins might play only a minor role in regulation of gene expression at RT . Consistent with this idea , the Ryp factors are not required for the normal transcriptional profile of filaments grown at RT ( Figure S1A ) . In sum , our transcriptional profiling experiments revealed that Ryp1 , Ryp2 , and Ryp3 are major regulators of yeast-phase growth at 37°C; are dispensable for filamentous-phase growth at RT; can either facilitate or repress transcript accumulation; and may act in the same pathway to regulate gene expression . Previous studies reported that Wor1 , an ortholog of Ryp1 , associates with DNA at hundreds of intergenic regions to regulate gene expression in C . albicans [14] . However , genome-wide DNA associations of Ryp1 have not been investigated in H . capsulatum . Additionally , Velvet family proteins do not contain a known DNA-binding domain and the ability of Ryp2 and Ryp3 orthologs to associate with DNA has not been explored . To establish the genome-wide association of Ryp factors with DNA and to distinguish between direct and indirect targets in the Ryp regulons , we performed chromatin immunoprecipitation-on-chip ( ChIP-chip ) using antibodies raised against Ryp1 , Ryp2 , and Ryp3 . Experiments were performed in either wild-type yeast cells grown at 37°C , or in the respective ryp mutant control grown at 37°C . We observed 361 ChIP events throughout the genome of wild-type cells ( Figure 2 , Figure S2A , and Table S2 ) . Most notably , there were a large number of targets ( 182 loci ) that associated with at least two Ryp factors , and 94 loci associated with all three Ryp factors , suggesting that Ryp1 , Ryp2 , and Ryp3 can act together to regulate gene expression . Interestingly , only Ryp1 had a large number of events ( 161 loci ) that were not shared with other Ryp factors , indicating that Ryp1 has a broader regulon than Ryp2 and Ryp3 ( Figure S2A ) . Further characterization of the ChIP events revealed that intergenic lengths corresponding to Ryp association events were significantly longer than the average intergenic length in the genome ( Figure S2B ) . A similar trend was noted previously for Wor1 association events in C . albicans , but its biological significance is unknown [27] . Notably , shared ChIP events that involved all three Ryp factors showed the most drastic shift in intergenic length distribution . To map the genomic regions defined by ChIP-chip events to specific genes , we used a validated gene set that was defined previously based on gene expression and sequence conservation [28] and identified genes that lie downstream of each ChIP event ( Table S2 ) . Our first notable observation was that Ryp1 , Ryp2 , and Ryp3 showed interdependent regulation . All three Ryp factors associated upstream of RYP1 and RYP2 , although none of the three factors associated upstream of RYP3 ( Figure 3A and Table S2 ) . To further explore the relationship between DNA association and gene expression , we overlaid gene expression data onto all ChIP-chip target genes ( Figure 2 ) . Targets that associated only with Ryp1 were not significantly enriched in YPS genes compared to the whole genome . In contrast , we observed significant enrichment of YPS genes for DNA-association events that were shared by Ryp1 , Ryp2 , and Ryp3 , as well as events that were shared only by Ryp1 and Ryp3 . This analysis revealed a correlation between shared association events and genes whose enhanced expression was induced by growth at host temperature . Strikingly , many of the known virulence genes ( CBP1 , SOD3 , and YPS3 ) were shared Ryp targets ( Figure 3B and Table S2 ) . These results indicate that the known core virulence genes are direct targets of the Ryp factors , and suggest that the remaining shared Ryp targets are interesting candidates for potential virulence factors . Notably , shared Ryp targets and all Ryp targets were enriched for predicted signal peptides ( p = 1 . 4e-06 and p = 2 . 0e-06 , respectively ) compared to the whole genome , which is of interest since secreted proteins produced by intracellular pathogens are often involved in virulence . In addition to virulence factors , the shared Ryp targets included a single transcription factor with a known DNA-binding domain and yeast-phase–specific expression . The corresponding gene , designated HISTO_DM . Contig933 . eannot . 1650 . final_new , encodes a Zn ( II ) 2Cys6 zinc binuclear cluster domain protein ( Figure 4A ) . As shown by microarray and qRT-PCR experiments , transcript levels of this gene were 6 . 5- to 35-fold higher at 37°C compared to RT , indicating that it displays enhanced expression in wild-type yeast-phase cells as compared to filaments ( Figure 4B and Table S1 ) . Moreover , Ryp1 , Ryp2 , and Ryp3 were required for expression of this gene at 37°C ( Figure 4B and Table S1 ) . We investigated the possibility that , similar to the RYP factors , HISTO_DM . Contig933 . eannot . 1650 . final_new is essential for growth in the pathogenic yeast form . Since targeted gene disruption in H . capsulatum is inefficient , we generated knockdown strains using RNA interference ( RNAi ) . We were able to deplete mRNA levels of HISTO_DM . Contig933 . eannot . 1650 . final_new by 72%–88% ( Figure 4C ) . Additionally , these RNAi strains were unable to grow in the yeast form and instead exhibited robust filamentous growth at 37°C , similar to the ryp mutants ( Figure 4D ) . Loss of RNAi plasmids resulted in reversion to yeast-phase growth at 37°C ( Figure S3A ) , indicating that the phenotype was dependent on the RNAi plasmids . Thus , we renamed this gene RYP4 ( Required for Yeast-Phase Growth ) . Although these morphologic studies revealed that ryp4 knockdown strains appeared drastically different from wild-type cells grown at 37°C , they were indistinguishable from wild-type filaments in appearance when grown at RT ( Figure 4D ) . BLASTP analysis indicated that the closest homolog of Ryp4 is FacB of A . nidulans ( E-value <1e-180 ) , and phylogenetic analysis ( see Materials and Methods ) revealed that Ryp4 is an ortholog of FacB ( Figure S3B and S3C ) . FacB is a transcriptional regulator of genes involved in acetate utilization in Aspergillus species and N . crassa [29]–[33] . Considering this conserved role in multiple fungal species , we investigated whether Ryp4 has a role in acetate utilization in H . capsulatum . However , unlike facB mutants in other organisms , ryp4 knockdown strains were able to grow in acetate as a major carbon source ( unpublished data ) , leading us to favor the hypothesis that Ryp4 has been rewired to regulate morphology in H . capsulatum . To assess whether Ryp4 is required for normal yeast-phase–specific gene expression in response to temperature , we performed whole-genome expression profiling experiments using wild-type cells carrying control vectors ( hereafter referred to as wild-type ) and ryp1 , ryp2 , ryp3 , and ryp4 knockdown strains grown either at RT or at 37°C . In contrast to the gene expression studies performed in Figure 1 using insertion mutants , here we used ryp1 , ryp2 , and ryp3 knockdown strains to match the ryp4 knockdown strain ( hereafter referred to as “mutant” ) . Statistical analyses revealed 441 YPS genes and 362 FPS genes in this dataset ( Table S3 ) . Similar to the aforementioned microarray dataset , regulation of the expression of the majority of YPS and FPS genes was dependent on Ryp1 , Ryp2 , Ryp3 , and Ryp4 ( Figure 4E and Table S3 ) . Notably , the transcriptional profile of the ryp4 mutant was very similar to that of ryp1 , ryp2 , and ryp3 mutants ( Figure 4E ) . The transcriptional profile of ryp4 mutants grown at 37°C was similar to that of wild-type filamentous cells grown at RT ( Figure S4A ) . Additionally , both microarray analysis and qRT-PCR experiments showed that Ryp4 was required for the expression of RYP1 , RYP2 , and RYP3 ( Figure S3D ) . In contrast to its critical role at 37°C , Ryp4 was not required for the normal transcriptional profile of filaments grown at RT ( Figure S4B ) . Overall , these results indicate that Ryp4 , like Ryp1 , Ryp2 , and Ryp3 , is critical for temperature-regulated gene expression at 37°C . To further explore the Ryp4 regulon , we performed ChIP-chip experiments in wild-type cells using antibodies raised against Ryp4 . We identified 61 Ryp4 association events that occur in cells grown at 37°C ( Table S2 ) . Interestingly , the majority ( 74% ) of these events were shared with other Ryp factors , indicating that Ryp4 acts in concert with other Ryp factors to regulate gene expression . Next , we identified genes downstream of each Ryp4 ChIP event ( Table S2 ) , and visualized these data together with Ryp1 , Ryp2 , and Ryp3 events ( Figure S5 ) . We found that about a quarter ( 26% ) of Ryp1 , Ryp2 , and Ryp3 shared targets were also occupied by Ryp4 ( Figure 4F ) . Furthermore , these common Ryp targets were even more significantly enriched for YPS genes than targets that were shared by Ryp1 , Ryp2 , and Ryp3 , but not Ryp4 ( Figure 4F ) . These YPS genes included core regulators of morphology ( RYP1 , RYP2 , and RYP4 ) and genes required for virulence ( CBP1 and SOD3 ) , further emphasizing the role of Ryp4 as a fundamental regulator of yeast-phase growth and an essential component of the temperature-responsive Ryp regulatory network ( Figures 3A , B and 4A ) . The ChIP data described above revealed that Ryp1 , Ryp2 , Ryp3 , and Ryp4 share a large number of overlapping targets , suggesting that these proteins may physically interact . To test this hypothesis , we performed co-immunoprecipitation ( co-IP ) experiments in wild-type cells grown at 37°C using each of the Ryp antibodies . Of all possible co-IP combinations , we were able to reliably and reproducibly detect Ryp1 , Ryp2 , and Ryp3 in Ryp2 and Ryp3 IPs , indicating that at least these three Ryp proteins physically interact ( Figure 5A ) . Ryp4 IPs did not yield reproducible interactions with other Ryp proteins . No Ryp protein signal was present in control IPs performed in the corresponding ryp mutant grown at 37°C ( unpublished data ) . In addition to these biochemical experiments , we used yeast two-hybrid experiments to assess Ryp1–Ryp2–Ryp3–Ryp4 interactions . We observed a reciprocal interaction between Ryp2 and Ryp3 , confirming the co-IP results ( Figure 5B ) . Our results also revealed that the Ryp2 N-terminus ( which contains the Velvet domain ) , but not the Ryp2 C-terminus , is required for interaction with Ryp3 ( Figure 5C ) . On the other hand , the Ryp2 C-terminus mediates Ryp2–Ryp2 interactions ( Figure 5B and 5C ) . No interactions were observed for Ryp4 ( unpublished data ) . Despite numerous attempts at transformation and analysis of multiple clones , we were unable to express Ryp1 bait or prey fusion proteins in the S . cerevisiae strains used for yeast-two-hybrid analysis . As a result , yeast-two-hybrid interactions with Ryp1 could not be assessed . Similarly , a previous study reported difficulty expressing yeast-two-hybrid constructs made with the Ryp1 ortholog from F . oxysporum [13] , suggesting that expression of these Ryp1 two-hybrid fusion proteins is toxic in standard laboratory strains of S . cerevisiae . To further explore the Ryp regulatory network , we analyzed each Ryp ChIP event set for the nonrandom occurrence of conserved cis-acting regulatory sequences ( hereafter referred to as DNA motifs ) . This de novo motif analysis was especially interesting for Ryp2 and Ryp3 , since there has been no prior evidence that Velvet family proteins associate with DNA , and thus , no DNA motifs have been identified for Velvet family proteins . Through the analysis of Ryp1 , Ryp2 , and Ryp3 ChIP events , we identified two distinct DNA motifs , which we named Motif A and Motif B . Specifically , different variants of Motifs A and B were identified through the analysis of Ryp ChIP events ( Table S4 , Figure S6A and S6B ) , and motifs that had the best predictive characteristics using receiver operating characteristic ( ROC ) plots ( as described in [34] ) were selected as Motif A and Motif B ( Figure 6A ) . Motif A is very similar to the Wor1 motif , which was previously identified using biochemical approaches [14] . Identification of Motif A by a completely independent approach validates our ChIP-chip data and our motif analysis pipeline . Motif B , which is quite distinct from Motif A , did not resemble any previously identified motifs according to searches of the motif databases JASPAR ( http://jaspar . genereg . net/ ) and YETFASCO ( http://yetfasco . ccbr . utoronto . ca/ ) . ROC plots using all ChIP events from each of the regulators demonstrated that Motif A and B were enriched in the entire ChIP set . Shuffled versions of each motif resulted in loss of specificity ( Figure 6B ) . Of note , motif-finding algorithms did not yield a meaningful result for Ryp4 , which is not surprising as the complex binding sites of zinc binuclear cluster transcriptional regulators can be difficult to predict [35] , [36] . Further analysis revealed that Motif A was associated with Ryp1 ChIP events , whereas Motif B was associated with Ryp2 and Ryp3 ChIP events: Motif A specificity was dependent only on the inclusion of Ryp1 ChIP events ( Figure S6C ) , whereas Motif B specificity was dependent only on the inclusion of Ryp2 and Ryp3 events ( Figure S6D ) . Furthermore , Motif A enrichment was similar in all Ryp1 targets regardless of whether they were shared targets with the other Ryp factors or whether they associated only with Ryp1 ( Figure S6E ) . In contrast , Motif B was enriched in Ryp1 targets that were shared with Ryp2 and Ryp3 , but had no specificity in targets that were unique to Ryp1 ( Figure S6E ) . These results corroborate the model that association of Ryp1 with DNA correlates with the presence of Motif A , and association of Ryp2 and Ryp3 with DNA correlates with the presence of Motif B . Motifs A and B are distributed throughout the H . capsulatum genome and are present in many of the Ryp targets ( Tables S5 ) . Specific examples are shown in Figures 3 and 4 where Motif A and B distribution is shown in the upstream regulatory regions of RYP1 , RYP2 , RYP4 , CBP1 , SOD3 , and CATB . Neither motif was found in the RYP3 upstream region ( Figure 3 ) , which is in agreement with our ChIP results that the Ryp factors do not associate upstream of RYP3 . Our motif analyses suggested that Ryp1 associates with DNA via Motif A , whereas Velvet family proteins ( Ryp2 and Ryp3 ) associate with DNA via Motif B . To test whether Ryp1 , Ryp2 , or Ryp3 can directly bind to Motif A or Motif B , we performed electrophoretic mobility shift assays ( EMSAs ) . The promoter region of the CBP1 gene , which associates with the Ryp proteins by ChIP-chip and contains both Motif A and Motif B ( Figure 3B ) , was used to design two distinct 60 bp probes encompassing either Motif A or Motif B ( Figure 3B ) . His-tagged versions of Ryp1 , Ryp2 , and Ryp3 were synthesized in coupled in vitro transcription and translation reactions and then subjected to purification . Additionally , it was already known that the N-terminus of C . albicans Wor1 , which is highly homologous to the N-terminus of Ryp1 , contains two globular domains that are sufficient to bind the Wor1 motif [14] . Therefore , we also synthesized and purified a His-tagged version of the N-terminus of Ryp1 ( Ryp1 ( N ) , 1–267 aa ) that harbors only the globular domains . EMSA revealed that full-length Ryp1 binds directly to Motif A , whereas control extracts contained no binding activity ( Figure 6C ) . In analogy to Wor1 , we also observed that Ryp1 ( N ) binds directly to Motif A ( Figure 6C ) . To explore the ability of Ryp2 and Ryp3 to bind Motif B , we performed mobility shift assays with either Velvet protein alone or with a combination of Ryp2 and Ryp3 . Whereas mobility shift assays performed with either Ryp2 or Ryp3 showed no binding activity , addition of both of these Velvet proteins resulted in binding of the Motif B probe ( Figure 6D ) . All observed band shifts were diminished upon addition of unlabeled competitor probe into the binding reactions . Notably , these experiments provide the first evidence that Velvet family proteins bind DNA directly . Since Ryp1 is sufficient to bind Motif A and Ryp2–Ryp3 is sufficient to bind Motif B , we next used an in vivo transcriptional activation assay [14] to explore whether each motif was sufficient to drive gene expression in a heterologous system when the appropriate Ryp proteins were expressed . We cloned a single copy of Motif A or B upstream of the UAS-less CYC1 promoter fused to the lacZ gene . We introduced these reporter plasmids along with constructs expressing the Ryp proteins , either individually or in combination , into S . cerevisiae and monitored β-galactosidase activity . S . cerevisiae provides an ideal heterologous expression system for our experiments , since the reporter strain we used lacks Ryp1 orthologs , and Velvet family proteins are not present in S . cerevisiae . In the strains that contained the Motif A reporter construct , β-galactosidase activity was detected only when Ryp1 was heterologously expressed ( Figure 7A ) . This activity was severely diminished when the conserved residues of Motif A were mutated . Co-expression of Ryp2 , Ryp3 , or Ryp4 individually with Ryp1 , or Ryp1 , Ryp2 , and Ryp3 together , did not lead to an increase in β-galactosidase activity ( Figure 7A ) . These results indicate that Ryp1 is necessary and sufficient to drive gene expression via Motif A . In contrast , in strains containing Motif B , β-galactosidase activity was detected only when Ryp2 and Ryp3 were expressed together , but not when Ryp2 or Ryp3 were expressed singly ( Figure 7B ) . This activity was also dependent on the conserved nucleotides of Motif B . Co-expression of Ryp1 or Ryp4 along with Ryp2 and Ryp3 did not enhance Ryp2-Ryp3–dependent β-galactosidase activity ( Figure 7B ) . Additionally , even though the Ryp2 N-terminus ( Ryp2 ( N ) ) , which contains the Velvet domain , is sufficient to interact with Ryp3 ( Figure 5C ) , co-expression of Ryp2 ( N ) and Ryp3 was not sufficient to drive gene expression using Motif B ( Figure 7C ) . For all these experiments , β-galactosidase activity was dependent on the presence of the appropriate motif ( Figure S7A ) , but independent of motif orientation ( Figure S7B and S7C ) . These results indicate that Ryp2 and Ryp3 together are necessary and sufficient to drive gene expression via Motif B and , taken together with the EMSA studies , provide the first evidence that Velvet proteins bind DNA via a conserved motif to direct transcriptional activation . Cells have derived complex regulatory networks to reprogram their transcriptional response upon changes in the environment . In thermally dimorphic fungi , regulation of cell morphology and virulence traits is coupled: they respond to host temperature by altering their cell shape and inducing virulence gene expression . This study is the first elucidation of the transcriptional circuitry underlying this dramatic change in any of these organisms . In this article , we have shown that an interlocking network of transcription factors regulate each other and common target genes to trigger a transcriptional program that is required for cell shape changes and virulence gene expression in response to host temperature in H . capsulatum . Each of the Ryp transcription factors described here , including the newly discovered Ryp4 , is absolutely required for the normal transcriptional profile of cells grown at 37°C . The structure of the corresponding transcription factor network is diagrammed in Figure 8 , which illustrates the connectivity between the four proteins required to program the switch from filamentous to yeast-form growth in response to temperature . Each Ryp factor is required for the expression of the others , and each associates upstream of the RYP1 , RYP2 , and RYP4 genes—but none of the factors , including Ryp3 itself , associates upstream of the RYP3 gene . Thus we propose that at 37°C , each of the factors acts in a positive-feedback loop to regulate itself and each of the others . The exception is Ryp3 , which regulates the other factors but is only indirectly regulated by them and does not directly regulate itself . Perhaps the interlocking nature of the network promotes a robust response by amplifying and stabilizing the signal generated by increased temperature . Additionally , the absolute requirement for activation of all four Ryp factors may provide specificity by requiring a sustained increase in temperature ( as experienced within a mammalian host ) to trigger the appropriate developmental program . Moreover , the requirement for multiple factors may allow unknown host signals other than temperature to influence the transition from filaments to yeast-form cells . The nature of the molecular signal that induces an initial increase in Ryp expression in response to temperature is unknown , but others have identified a histidine kinase , DRK1 , that is required for yeast-phase growth in H . capsulatum and the related thermal dimorph Blastomyces dermatitidis [37] . A possible relationship between DRK1 and regulation of Ryp factors has not been explored . Ryp factors are required both to promote transcription of genes with enhanced expression at 37°C and to prevent the inappropriate expression of filament-associated genes at this temperature . Transcriptional profiling of wild-type cells grown either at 37°C ( in the yeast form ) or at room temperature ( in the filamentous form ) yielded 764 genes ( about 8 . 5% of the genome ) that passed our criteria for differential expression in one of the two growth phases . Remarkably , 96% of YPS genes required each of the four Ryp proteins to achieve differential expression in response to temperature . The majority of this regulon is a consequence of indirect regulation by the Ryp factors , since only a fraction of YPS genes are downstream of Ryp association events . The most significant enrichment for YPS genes was observed for shared targets of multiple Ryp factors: about 21% of shared targets were YPS genes as opposed to 4% of the whole genome . Interestingly , Ryp association events that are unique to an individual Ryp factor , such as the large cohort of ChIP-chip events that associate only with Ryp1 and not the other three Ryp factors , showed no enrichment for YPS genes . Additionally , Ryp targets that were not YPS genes did not depend on Ryp factors for their expression under standard laboratory conditions . It is possible that Ryp factors are poised upstream of these genes to regulate their expression in response to environmental signals other than temperature , such as light intensity , nutrient availability or exposure to reactive radicals in the host . Recruitment of the Ryp transcription factors to their direct targets is driven , at least in part , by DNA sequence motifs that are sufficient to recruit either Ryp1 ( Motif A ) or the Ryp2–Ryp3 complex ( Motif B ) ( Figure 8 ) . Motif A , which is highly similar to the DNA motif that was defined for the Ryp1 ortholog Wor1 , is enriched in Ryp1 targets , whereas Motif B is enriched in Ryp2 and Ryp3 targets . Figure 8 illustrates the major categories of motif distribution for YPS gene regulatory regions that associate with Ryp factors . Interestingly , some YPS genes that contain only Motif A or only Motif B show association with Ryp1 , Ryp2 , and Ryp3 . Although there are several models that could explain this multifactorial association , we favor the idea that there are different subcomplexes of Ryp proteins in the cell . For example , we propose that Ryp1 associates with DNA directly via Motif A in the absence of Ryp2 and Ryp3 , thereby accounting for the many Ryp1 ChIP-chip events that are not shared by the other Ryp factors . Supporting this model is the finding that Ryp1 is sufficient to drive gene expression via Motif A in S . cerevisiae , and the biochemical experiments that show that the Ryp1 ortholog Wor1 associates directly with DNA via a motif that is highly similar to Motif A [14] . Alternatively , Ryp1 can be present in a complex with Ryp2 and Ryp3 , and then could be recruited to the DNA either directly , via Motif A , or indirectly , via interaction of Ryp2–Ryp3 with Motif B . This model also predicts that Ryp2–Ryp3 can be recruited to the DNA either directly or via interaction with Ryp1 . Interestingly , the existence of ChIP-chip targets that contain Motif A but associate only with Ryp1 and not with Ryp2–Ryp3 suggests that there is a genomic feature , such as chromosomal context for Motif A , that distinguishes the set of Ryp1-only targets from the shared targets . For example , perhaps chromosomal context causes the association of Ryp1 with Motif A in the shared targets to require nonspecific interactions of Ryp2–Ryp3 with the DNA . Interestingly , we identified a number of ChIP-chip targets that associate only with Ryp3 , and found that these targets are enriched for Motif B . Since binding of Ryp3 to Motif B , at least in vitro , requires the presence of Ryp2 , it is possible that these orphan events are actually shared with Ryp2 but fall below our level of detection for Ryp2 association . Alternatively , Ryp3 might be recruited to these targets via association with one of the other two Velvet domain proteins in H . capsulatum . Finally , the newly identified yeast-phase regulator Ryp4 seems to play no role in recognition of Motif A or Motif B , at least in vitro , and is incapable of driving gene expression via these motifs in a heterologous transcriptional activation assay . Additionally , we did not observe any physical interactions between Ryp4 and other Ryp factors . Taken together , these observations suggest that a third cis-regulatory element might recruit Ryp4 to its targets . In this study , we provide the first evidence that Velvet family proteins such as Ryp2 and Ryp3 can bind DNA directly . This family of proteins is well studied in environmental fungi , fungal plant pathogens , and an opportunistic fungal pathogen of humans; however , although physical interactions between multiple Velvet family proteins have been defined , the molecular mechanism of Velvet family proteins in general is unknown . In addition to showing binding of Ryp2 and Ryp3 to the DNA via mobility shift assays , we show that Motif B is sufficient to drive gene expression in the model yeast S . cerevisiae when Ryp2 and Ryp3 are co-expressed . The N-terminus of Ryp2 , which contains the Velvet domain , is required for Ryp2–Ryp3 interaction in the yeast two-hybrid assay , whereas the C-terminus of Ryp2 mediates multimerization of Ryp2 ( Figure 5C ) . In a previous study , the C-terminal region of the Ryp2 ortholog VosA was predicted to be a transcriptional activation domain [38] . Since co-expression of the Ryp2 N-terminus with Ryp3 is not sufficient to activate gene expression through Motif B ( Figure 7C ) , we conclude that the Ryp2 C-terminus is required for activation of gene expression in our S . cerevisiae transcriptional activation assays . Whether a Ryp2 homodimer or multimer has a role in gene expression in H . capsulatum or is important for the higher order formation of a Ryp complex remains to be investigated . One of the most interesting findings to arise from this network analysis is that the Ryp factors , which themselves are YPS genes that are required for morphology , directly regulate YPS genes that are required for pathogenesis ( e . g . , CBP1 , SOD3 , and YPS3 ) . A hallmark of thermally dimorphic fungal pathogens is that temperature regulates cell shape and other factors required for pathogenesis . In the case of H . capsulatum , yeast-phase morphology is thought to be critical for the lifestyle of this fungus in the host , especially since replication of the fungus within host immune cells is likely to be incompatible with filamentous growth . Thus , Ryp factors catalyze both the cell shape change and the increased expression of known virulence factors , resulting in a molecular link between temperature and the expression of traits required to cause disease in a mammalian host . A link between morphology and virulence has been explored for many of the major human fungal pathogens , perhaps most extensively for C . albicans ( reviewed in [39] ) . Unlike the thermally dimorphic fungi , which are found almost exclusively in a hyphal form in the soil and a yeast or spherule form in the host , C . albicans is present in multiple morphologies in the host , and filamentous cells are thought to play a major role in disease . We observed co-regulation of yeast-phase morphology and known virulence genes in H . capsulatum . Similarly , numerous studies in C . albicans have observed that hyphal formation is coordinated with the expression of genes required for adhesion , host tissue invasion , antifungal drug resistance , and oxidative stress response [39] , suggesting that a subset of fungal pathogens have evolved regulatory circuits that link virulence gene expression with morphologic changes . Notably , despite the importance of the yeast-to-hyphal transition in the infectivity of C . albicans , a systematic analysis of a large-scale C . albicans deletion library showed that morphology can be uncoupled from virulence . Although a large number of mutants were identified to have defects in both virulence and morphology , a sizeable number of mutants were defective for virulence despite maintaining normal morphologic responses in vitro , and some mutants displayed normal infectivity despite having substantial defects in morphogenesis [40] . In contrast , the ability to differentiate into yeast-form cells is likely to be essential for the intracellular parasitism that is characteristic of H . capsulatum infection . Therefore , it may be the case that yeast-phase morphology is categorically linked to virulence in the thermally dimorphic fungal pathogens . Finally , this work expands the rich history of elucidating fungal transcription circuits to understand regulatory networks in eukaryotic cells . Studies of fungal gene regulation continue to provide examples of novel DNA-binding domains , as was previously the case with the Ryp1 family of proteins [14] . In this study , we show association between DNA and two Velvet family proteins that lack known DNA-binding domains , now implicating this highly conserved protein family in direct gene regulation . The most interesting molecular implication of these data is that the Velvet domain could be a novel DNA binding domain . Additionally , this analysis reveals a fundamental example of transcriptional rewiring: although Ryp4 regulates acetate utilization genes in related fungi outside of the thermal dimorphs , in H . capsulatum its major role is to regulate the transcriptional response to temperature . As a result , the temperature-dependent circuit elucidated here demonstrates cooperation and regulation between three distinct families of transcription factors: WOPR , Velvet , and Zn ( II ) 2Cys6 . Expression of each of the Ryp proteins has been wired to be absolutely dependent on all of the others , so strains that lack any of the regulators fail to undergo a developmental program in response to temperature . This network structure is interesting , in part because of the regulatory implications described above , but also because the relationship between orthologs of these transcription factors is not static throughout the fungal kingdom . For example , Hemiascomycetes have retained WOPR family members but lack Velvet family proteins . These phylogenetic differences provide an opportunity to study how regulatory circuits evolve . Finally , the role of the Ryp proteins in the life cycle of other thermally dimorphic fungi has not been examined . Elucidating the nature of this circuit in these related fungal pathogens will provide insight on the evolution of this temperature-responsive circuit . H . capsulatum strains G217B ( ATCC26032 ) and G217Bura5 ( WU15 ) were kind gifts from Dr . William Goldman ( University of North Carolina , Chapel Hill ) . ryp1 , ryp2 , ryp3 T-DNA , and knockdown mutants were previously generated [5] , [6] . ryp4 knockdown mutants and wild-type strains with control RNAi plasmids were generated in this study . All plasmids and primers used in this study are listed in Table S6 and Table S7 , respectively . All plasmids were maintained in Escherichia coli DH5α strain and sequenced to ensure no mutation was introduced during the cloning process . Introduction of integrating and episomal RNAi plasmids into H . capsulatum G217Bura5 strain was done as previously described [6] . ryp4 knockdown strains with integrated RNAi constructs were used in all experiments , except for the plasmid loss experiments where episomal RNAi plasmids were used . Four independent ryp4 mutants generated with two different episomal RNAi plasmids ( pSB30 and pSB31 ) were grown in HMM broth supplemented with uracil ( 0 . 2 mg/ml ) at 37°C , 120 rpm with 5% CO2 . After 4 wk , an inoculating loop was used to transfer cells from each flask onto HMM+uracil agarose plates followed by incubation at 37°C with 5% CO2 . Individual yeast-phase colonies were streaked again and tested for uracil auxotrophy on HMM agarose plates . Representative images of the resulting strains are shown in Figure S3A . All images were obtained using a Zeiss Axiovert 200 microscope with 40× Phase objective . Total RNA from H . capsulatum strains was harvested using Trizol ( Life Technologies-Invitrogen ) following manufacturer's instructions . RNA was treated with RNase-free DNase set ( Qiagen ) and cleaned using the RNeasy mini kit ( Qiagen ) . 2 µg of total RNA was used in cDNA synthesis using Superscript RT II ( Life Technologies-Invitrogen ) following manufacturer's instructions . cDNAs were used to amplify RYP1 , 2 , 3 , 4 , and GADPH transcripts using OAS1057-58 , OAS1942-43 , OAS1944-45 , OAS3320-31 , and OAS1452-53 , respectively ( Table S7 ) . qRT-PCR reactions were performed in the Stratagene Mx3000P QPCR System ( Agilent Technologies ) and contained 300 nm of each primer and 0 . 8× FastStart Universal SYBR Green Master mix ( Roche ) . The transcript levels of RYP genes were normalized to the GAPDH transcript levels . Total RNA from H . capsulatum strains was harvested as previously described [5] . Reference RNA was generated by mixing total RNA from ryp mutants and wild-type control strains in three different ratios: For microarray experiments with ryp T-DNA mutants , equal amounts of total RNA from each ryp mutant and wild-type were combined . For microarray experiments with ryp1 , ryp2 , and ryp3 knockdown strains , total RNA from wild-type yeast , wild-type filaments , and ryp knockdown strains were mixed in a 1∶1∶2 ratio . For microarray experiments with ryp4 knockdown strains , total RNA from wild-type yeast , wild-type filaments , and ryp4 knockdown strains was mixed in 1∶1∶1 ratio . For each sample , 15 µg of total RNA was used to generate amino-allyl–labeled cDNA as described previously [42] with the following modifications: Reverse transcription reaction was performed with Superscript RT II ( Life Technologies-Invitrogen ) , and reactions were cleaned using QIAquick PCR clean-up kit ( Qiagen ) with phosphate wash buffer ( 5 mM KPO4 , pH 8 . 0 , 80% ethanol ) . cDNA from each ryp mutants and wild-type controls was labeled with Cy5 ( GE Healthcare Life Sciences-Amersham ) as described [42] and competitively hybridized against the Cy3-labeled pooled reference sample using a H . capsulatum whole-genome 70-mer oligonucleotide microarray . For the experiments with ryp T-DNA mutants , there were four to six replicates for each strain and condition , and for the experiments with ryp knockdown strains , there were three to 12 replicates for each strain and condition . Raw data for all hybridizations ( total of 88 ) performed are available through Gene Expression Omnibus ( GEO , www . ncbi . nlm . nih . gov/geo/ ) with accession numbers GSE46936 , GSE46937 , GSE46938 , and GSE46939 ( GEO superseries accession number GSE 47832 ) . Arrays were scanned on a GenePix 4000B scanner ( Molecular Devices ) to determine the intensity units in the 635- and 523-nm channels ( detecting Cy5 and Cy3 , respectively ) for each spot on the microarray . Data were analyzed by using GENEPIX PRO version 6 . 0 , NOMAD ( http://nomad2 . ucsf . edu/NOMAD/nomad-cgi/login . pl ) , and MultiExperiment Viewer 4 . 0 ( www . tm4 . org/mev . html ) [43] , [44] . After removal of values for flagged spots , background subtraction , and median normalization , the ratio of the median Cy5 intensity/median Cy3 intensity for these spots was used for further analysis . For the genes represented by multiple 70 mer probes , only values for the first probe were used for subsequent analysis . Normalized data was analyzed by Bayesian Analysis for Gene Expression Levels ( BAGEL ) to obtain relative expression levels for each spot in each condition or mutant [45] . Using the values obtained by BAGEL , we compared any two samples by dividing the value of the first sample by the value of the second sample . To determine the number and identity of genes that changed significantly in expression in any given comparison , we used >2 . 0-fold change in transcript levels and p value <0 . 01 as a cutoff criteria . Results of these analyses are given in Tables S1 and S3 . A directed graph was constructed from all “is-a” relationships in the 12/7/2011 version of the Gene Ontology ( go_daily-termdb . obo-xml . gz downloaded from the GO Consortium website [http://www . geneontology . org/] ) . GO-to-gene edges were added from the 12/7/2011 versions of the AspGD [http://www . aspgd . org/] [46] , CGD [http://www . candidagenome . org/] [47] , and SGD [http://www . yeastgenome . org/] [48] GO association files for A . nidulans ( An ) , A . fumigatus ( Af ) , C . albicans ( Ca ) , Candida glabrata ( Cg ) , and S . cerevisiae ( Sc ) . Gene-to-gene edges were added for An/Af/Ca/Cg/Sc genes to their H . capsulatum G217B ( HcG217B ) orthologs based on InParanoid mapping of each genome pair ( using default parameters with no outgroup ) . InterPro domains ( IPR ) from the Pfam , TIGRFAMS , SMART , PANTHER , and Gene3D databases from InterPro version 34 were mapped to HcG217B genes with InterProScan version 4 . 8 [49] . GO-to-IPR and IPR-to-gene edges were added to the graph by parsing the InterProScan results . An/Af/Ca/Cg/Sc genes with no ortholog in HcG217B were pruned from the graph , as were obsolete GO terms . The set of G217B genes deriving from each GO term was tabulated by traversing the graph in reverse topological sort order , assigning each parent node to the union of the G217B genes spanned by its children . For each query gene set , the GO association graph was pruned to the subset of GO terms spanning the query . GO terms in the resulting subgraph with a single GO term parent and child were considered uninformative relative to the query and were replaced by a direct edge from parent to child . We further removed any GO terms that were associated only with a single gene . The graph was further reduced to only GO terms present in the union of the AspGD/CGD/SGD GO slim sets ( 12/7/2011 version ) . For each GO term spanning k genes in the query , the probability of that term spanning at least k terms in a random gene set of the same size was calculated from the hypergeometric distribution as in [50] . The probability calculation was carried out using the phyper function in R [51] with graph operations implemented using NetworkX [52] . The top five terms are reported for each query with no adjustment for multiple hypothesis testing . Initial sets of Ryp4 homologs were identified using HMMer 3 . 0 and the Pfam hidden Markov models ( HMMs ) for the fungal Zn ( II ) 2Cys6 binuclear cluster domain ( PF00172 . 13 ) and the fungal-specific transcription factor domain ( PF04082 . 13 ) . Each HMM was searched against the BROAD predicted gene sets for Coccidioides immitis RS , Trichophyton verrucosum HKI 0517 , Paracoccidioides brasiliensis Pb03 , Blastomyces dermatitidis er-3 , Fusarium graminearum ph-1 , Neurospora crassa OR74A , Magnaporthe oryzae 70-15 , Botrytis cinerea B05 . 10 , and Stagnospora nodorum SN15; the SGD curated gene set for Saccharomyces cerevisiae S288C; the AspGD curated gene set for Aspergillus nidulans FGSC; and the GSC predicted gene set for Histoplasma capsulatum G217B using hmmscan . Hits were aligned to each HMM using hmmalign , and neighbor-joining trees were generated from the aligned domain sequences using CLUSTALW 2 . 1 . For PF00172 . 13 , Ryp4 was found in a monophyletic clade , sister to Sip4 , with orthologs from each target species except for B . cinerea . For PF04082 . 13 , Ryp4 was found in a monophyletic clade identical to the previous one except for the gain of B . cinerea BC1G_13551 and the loss of Sc Cat8 . TBLASTN of the Ryp4 protein sequence against the B . cinerea genome yielded a hit to the Zn ( II ) 2Cys6 domain ( E = 2e-29 ) about 170 bp upstream of BC1G_13551 , consistent with an incorrectly predicted 5′ for this gene . Based on these results , the union of pezizomycotina hits was annotated as orthologs of Ryp4 . The full-length protein sequences of the Ryp4 orthologs , as well as Cat8 and Sip4 , were aligned , and a bootstrapped ( n = 1 , 000 ) neighbor-joining tree was generated using CLUSTALW . ryp1 , ryp2 , ryp3 , and ryp4 mutants and wild-type ( G217B ) cells were grown at 37°C . Cells were fixed and collected as previously described [5] . Frozen cell pellets were ground using a mortar and a pestle in liquid nitrogen or in Retsch Mixer Mill MM 400 . Chromatin immunoprecipitation was performed as described [53] with the following modifications: 300 mg of ground samples were vortexed vigorously in lysis buffer [50 mM Hepes/KOH ( pH 7 . 5 ) , 140 mM NaCl , 1 mM EDTA , 1% Triton X-100 , 0 . 1% sodium deoxycholate supplemented with Halt protease and phosphatase inhibitors ( Pierce ) ] for 3 h with 0 . 5 mm glass beads to lyse the cells . Following cell lysis , lysates were sonicated in a Diagenode Bioruptor for 30 min ( 30 s on , 1 min off ) to shear DNA . Input DNA was collected and the rest of the sample was subjected to immunoprecipitation . For filamentous-phase samples ( ryp mutants ) , lysates from three identical tubes were combined prior to antibody incubation . Each sample was incubated overnight at 4°C with 20 µg of polyclonal antibodies against portion of Ryp1 ( ID:2877 , ELDKPFPPGEKKRA ) , Ryp2 ( ID:237 , QTNRDYPFYNGPDAKRPR ) , Ryp3 ( ID:338 , GIKIPIRKDGVKGPRGGQ ) , or Ryp4 ( ID:8 , PPPQQSLQGWSPEEW ) . On the next day , 40 µl of Protein-A Sepharose 4B Fast Flow ( Sigma-Aldrich ) beads were added to the lysate-antibody mixture and further incubated for 2 h . Subsequently , beads were collected by centrifugation and washed nine times . Protein–DNA complexes that are bound to the beads were collected by incubating beads at 65°C with elution buffer ( 50 mM Tris HCl pH 8 . 0 , 10 mM EDTA , 1% SDS ) . Protein–DNA crosslinks in input and ChIPed samples were reversed overnight at 65°C . Proteins were digested with proteinase K and DNA was purified with phenol∶chloroform extraction as described [53] . Both input and output DNA were amplified and labeled with fluorescent dyes ( Cy3 and Cy5 ) using strand displacement amplification following published protocols [53] . Labeled DNA samples were hybridized following Agilent's ChIP-on-chip protocol onto Agilent 2× 400K arrays comprised of 60 mer oligos tiling the entire H . capsulatum genome at a frequency of one probe per 50 bp . Oligos were selected based on a previously published method [54] . Slides were scanned with an Agilent scanner and raw ratios were obtained with the Agilent ChIP_107_Sep09 protocol . Background subtracted probe intensities were transformed to log2 ( red/green ) ratios ( M ) and log2 ( sqrt ( redgreen ) ) geometric mean intensities ( A ) , excluding probes with intensities below background . The M , A values were fit using the LOWESS algorithm [55] as implemented in R [51] , and the fit curve was subtracted from the transformed data to yield lowess normalized log2 ratio values . Normalized spot intensities are available through GEO with accession number GSE47341 ( GEO superseries accession number GSE 47832 ) . After normalization , data were analyzed using Mochiview [56] . Data from three biological replicates for each Ryp-IP were analyzed together to identify peaks . ChIP-chip experiments done with ryp mutants were used as a negative control in this analysis . Default parameters of Mochiview peak extraction were used with the exception of increasing total random samples to 100 , 000 and maximum random samples to 100 ( p value <0 . 001 ) . Median+Interquartile range ( IQR ) was used as a threshold to filter extracted peaks . Peaks that were eliminated based on the noisy enrichment values in ryp mutants were included back if there was >2-fold difference between wild-type and ryp mutant enrichment ratios . Additionally , peaks that were greater than a median value for a given Ryp event and had another Ryp event greater than median+IQR were also included in the finalized list of events ( Table S2 ) . ChIP enrichment ratios plotted in Figures 3 and 4 were generated using ChIP tracks that were a combination of three biological replicates for each Ryp-IP . Each ChIP event was mapped to specific genes using an H . capsulatum G217B strain validated gene set that was defined previously using gene expression and sequence conservation criteria [28] . Genes that have a ChIP event in their 5′ region within a 10 kb distance from the center of the peak were listed as target genes ( Table S2 ) . In addition to target genes found in the validated gene set , some additional target genes were included if they had detectable expression levels in the whole-genome transcriptional profiling experiments performed in this study . Signal peptide and transmembrane helix prediction algorithm Phobius 1 . 01 [57] was run with default parameters on the full G217B predicted protein set , excluding two sequences with predicted internal stops . Three nondisjoint prediction sets were derived from the Phobius output: genes predicted to have a signal sequence , genes predicted to have at least one transmembrane helix , and the intersection of the previous two sets . For each phobius prediction set spanning k genes in a given Ryp associated set , the probability of that term spanning at least k terms in a random gene set of the same size was calculated from the hypergeometric distribution as in [50]: P ( k≤X ) = 1−sum∧[k−1]_[i = 0] c ( M , i ) c ( N−M , n−i ) /c ( N , i ) , where c is the binomial coefficient function , N is the size of the full gene set , n is the size of the Ryp associated set , and M is the size of the phobius prediction set . The probability calculation was carried out using the phyper function in R [51] . Each set of Ryp ChIP events was randomly split into two sets using Mochiview . Each set was subjected to motif finding algorithms using Mochiview and Bioprospector [56] , [58] . For each set , the top five motifs identified using Mochiview and the top three motifs identified using Bioprospector were further analyzed for specificity in the corresponding randomly split set . Motif searches were carried out using MAST [59] as implemented in version 4 . 8 . 1 of the MEME suite [60] . The MAST -hit_list parameter was used to yield all nonoverlapping motif instances in the query sequences with no adjustment for query length or number of motif hits ( such that a reported p value reflects only the alignment of the motif instance to the query matrix ) . For ROC plots , the motif threshold ( -mt ) was set to 0 . 01 in order to explore a wide range of possible parameter values . For genome-wide searches , the motif threshold was set to the values determined from the ROC plots using false positive rate of 10% as a cutoff ( corresponding to p values of 2 . 08e-04 for Motif A and 9 . 18e-05 for Motif B ) . Motif locations identified in the genome are given in Table S5 . For expression of the Ryp1-N-terminus , Ryp1 , Ryp2 , and Ryp3 proteins for EMSA experiments , we subjected pSB122 , pSB124 , pSB128 , and pSB130 , respectively , to TNT Coupled Wheat Germ Extract systems ( Promega ) following the manufacturer's instructions . For purification of Ryp proteins , eight 50-µl reactions were pooled and diluted 10-fold in binding buffer ( 10 mM Tris-HCl , pH 8 . 0 , 50 mM KCl , 5% glycerol , 1% TritonX-100 , and 20 mM imidazole ) supplemented with HALT phosphatase and protease inhibitors ( Pierce ) . For each sample , 100 µl of Ni-NTA agarose beads ( Qiagen ) was washed in binding buffer and incubated with extracts for 1 h at 4°C . Following the incubation , beads were washed once with binding buffer , and once with binding buffer with no detergents . His-tagged proteins were recovered with elution buffer ( 10 mM Tris-HCl , pH 8 . 0 , 50 mM KCl , 5% glycerol , 250 mM imidazole ) . The presence of His-tagged Ryp proteins was confirmed by SDS-PAGE analysis and Western blotting . Wheat germ extract with no DNA template was subjected to a similar purification process and use as a control in mobility shift assays . 5′-IRDye800-labeled Motif A and Motif B probes were prepared by annealing 5′-IRDye800-CBP1-MotifA-Fwd and CBP1-MotifA-Rev , and 5′-IRDye800-CBP1-MotifB-Fwd and CBP1-MotifB-Rev , respectively , in 10 mM Tris-HCl , pH 7 . 9 , 50 mM NaCl , 10 mM MgCl2 , and 1 mM DTT . Nonlabeled competitor probes were prepared similarly with nonlabeled oligonucleotides ( Table S7 ) . Two µg of each purified protein ( Ryp1-N-terminus , Ryp1 , Ryp2 , Ryp3 , or control extract ) and 1 nM of labeled probes were mixed in binding buffer ( 10 mM Tris-HCl , pH 8 . 0 , 50 mM KCl , 5% glycerol , 1 mM EDTA , 0 . 5 mM DTT , 100 ug/ml BSA , and 25 ug/ml poly ( dI:dC ) ) and incubated for 30 min at room temperature . Reactions were separated on 6% DNA retardation gels ( Invitrogen ) in 0 . 5× TBE buffer . Mobility shifts were visualized and analyzed using the ODYSSEY imaging system ( LI-COR Biosciences ) . Wild-type ( G217B ) cells grown to late log phase at 37°C were harvested by filtration , and the pellet was frozen in liquid nitrogen . Whole cell extracts were made by cryogrinding the pellet in Retsch Mixer Mill MM 400 . Co-immunoprecipitation experiments were performed using the Dynabeads Co-immunoprecipitation kit from Invitrogen following the manufacturer's instructions . Briefly , 100 ug of polyclonal α-Ryp2 ( ID:387 , SQSAGHMQSPSQVPPAWG ) or α-Ryp3 ( ID:356 , SHGSKGQDGEGEDWENEG ) antibodies were covalently linked to 5 mg of magnetic beads using Dynabeads Antibody Coupling kit . To prepare cell lysate , 2 g of ground samples were mixed with lysis buffer , vortexed , and spun down . Then , supernatant was incubated with 5 mg of antibody-coupled magnetic beads for 8 h at 4°C . After multiple washes , protein complexes that are bound to antibody-coupled beads were eluted in low pH buffer provided by the manufacturer . IPs with ryp2 and ryp3 mutants grown at 37°C and no antibody control were performed similarly . All fractions were separated by SDS-PAGE , visualized by silver staining , and analyzed by Western blotting using standard procedures . Polyclonal α-Ryp1 ( ID:3873 , ASSYQPGPPASMSWNTAATG ) , α-Ryp2 ( ID:387 , SQSAGHMQSPSQVPPAWG ) , or α-Ryp3 ( ID:356 , SHGSKGQDGEGEDWENEG ) antibodies were used to detect Ryp proteins . β-galactosidase assays were performed as previously described [61] . At least three independent isolates of each S . cerevisiae strain were grown to late log ( for the yeast-two-hybrid assay ) or stationary ( for the in vivo transcriptional activation assay ) phase . Each isolate was assayed in quadruplicate , and the results of representative isolates are shown in Figures 5 , 7 , and S7 .	Microbial pathogens of humans display the ability to thrive at host temperature . So-called “thermally dimorphic” fungal pathogens , which include Histoplasma capsulatum , are a class of soil fungi that upon being inhaled into the human lung , undergo dramatic changes in cell shape and virulence gene expression in response to host temperature . The ability of these pathogens to cause disease is exquisitely coupled to temperature response . Here we elucidate the regulatory network that governs the ability of H . capsulatum to switch from a filamentous form in the soil environment to a pathogenic yeast form at body temperature . The circuit is driven by three transcription regulators ( Ryp1 , Ryp2 , and Ryp3 ) that control yeast-phase growth . We show that these factors , which include two highly conserved proteins of the Velvet family of unknown function , bind to specific regulatory DNA elements and directly regulate expression of virulence genes . We identify and characterize Ryp4 , a fourth regulator of this pathway , and define DNA motifs that recruit these transcription factors to their temperature-responsive target genes . Our results provide a molecular understanding of how changes in cell shape are linked to expression of virulence genes in thermally dimorphic fungi .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "signal", "transduction", "fungi", "mycology", "signaling", "in", "cellular", "processes", "molecular", "cell", "biology", "microbial", "physiology", "transcriptional", "signaling", "cellular", "stress", "responses", "medical", "microbiology", "gene", "expression", "micro...	2013	A Temperature-Responsive Network Links Cell Shape and Virulence Traits in a Primary Fungal Pathogen
Crimean-Congo hemorrhagic fever virus ( CCHFV ) is a tick-borne zoonotic agent that is maintained in nature in an enzootic vertebrate-tick-vertebrate cycle . Hyalomma genus ticks have been implicated as the main CCHFV vector and are key in maintaining silent endemic foci . However , what contributes to their central role in CCHFV ecology is unclear . To assess the significance of host preferences of ticks in CCHFV ecology , we performed comparative analyses of hosts exploited by 133 species of ticks; these species represent 5 genera with reported geographical distribution over the range of CCHFV . We found that the composition of vertebrate hosts on which Hyalomma spp . feed is different than for other tick genera . Immatures of the genus Hyalomma feed preferentially on species of the orders Rodentia , Lagomorpha , and the class Aves , while adults concentrate mainly on the family Bovidae . With the exception of Aves , these hosts include the majority of the vertebrates consistently reported to be viremic upon CCHFV infection . While other tick genera also feed on these hosts , Hyalomma spp . almost completely concentrate their populations on them . Hyalomma spp . feed on less phylogenetically diverse hosts than any other tick genus , implying that this network of hosts has a low resilience . Indeed , removing the most prominent hosts quickly collapsed the network of parasitic interactions . These results support the intermittent activity of CCHFV foci: likely , populations of infected Hyalomma spp . ticks exceed the threshold of contact with humans only when these critical hosts reach adequate population density , accounting for the sporadic occurence of clinical tick-transmitted cases . Our data describe the association of vertebrate host preferences with the role of Hyalomma spp . ticks in maintaining endemic CCHFV foci , and highlight the importance of host-tick dynamics in pathogen ecology . Crimean-Congo hemorrhagic fever ( CCHF ) is a tick-borne zoonotic disease seen exclusively in humans that can progress from mild , non-specific signs to a severe and fatal hemorrhagic disease . The etiologic agent , Crimean-Congo hemorrhagic fever virus ( CCHFV; family Nairoviridae , genus Orthonairovirus ) , is transmitted to humans predominantly via tick bites , but may also be transmitted nosocomially or by handling tissues from viremic animals ( e . g . , in abattoirs ) . As non-human vertebrate hosts do not develop clinical signs [1] , maintenance in nature is largely silent . Recent reviews summarize the current knowledge about serology in animals [2] , routes of transmission [3] , and the tick species unambiguously involved in CCHFV circulation in natural and permanent foci [4] . Ticks are vectors and reservoirs for CCHFV; vertebrates act as a bridge , transmitting the virus to new generations of ticks . Infected vertebrates develop a short viremia [1] , and virus is transmitted to ticks feeding on viremic hosts , or through co-feeding with infected ticks ( demonstrated by [5 , 6] ) , which release the virus into the feeding cavity where other uninfected ticks feed . Most reports support ticks of the genus Hyalomma as the main CCHFV vector , but laboratory and field studies allude to other tick species that may also be responsible for virus circulation [4] . To date , interactions between CCHFV and the tick are not well known , including hypothetical molecular factors that could regulate infection and viral dissemination through a variety of physiological and anatomical barriers in the tick [7] . While molecular interactions are an obvious research target to explain the prominence of Hyalomma spp . in CCHFV maintenance and transmission , other non-molecular relationships may also be involved or even predominate . Some ticks of other genera may act as efficient CCHFV vectors under adequate laboratory conditions ( e . g . , as reviewed in [4] ) , raising the question about the importance of intimate molecular relationships between tick and virus versus the simple ecological interactions of ticks and key hosts in supporting silent CCHFV foci . In other words , the dynamics of CCHFV transmission may be driven by purely ecological factors and not depend on molecular compatibility . Most tick species are not restricted in range by their hosts; rather , climate is the main driver of their distribution patterns [8] . With the exception of some monoxenic species , ticks regularly feed on a wide range of hosts . Permanent foci of some tick-transmitted pathogens are restricted to the range of key reservoirs or vectors . For example , Borrelia burgdorferi s . l . is intimately linked with the tick genus Ixodes; Babesia bovis is transmitted exclusively by boophilid ticks; and tick-borne encephalitis virus is restricted to rodents reservoirs [9–11] . Likewise , CCHFV could conceivably circulate only in areas that support a delicate equilibrium of abundance and composition of appropriate hosts for the ticks . We hypothesized that a key factor for maintaining CCHFV foci is a precise combination of host species that feed the ticks , thereby amplifying infection in the tick population . A network-based analysis , comprising relationships among ticks and hosts connected by pairwise relations , enables the reconstruction of associations between ticks and key vertebrates in the circulation of a pathogen . This kind of ecological modeling is made possible by an extensive toolbox developed for network research ( see , for example , [12–14] ) . The structural properties of tick-vertebrate networks reveal new insights into the associations linking ticks and hosts that are key for supporting permanent CCHFV foci . Here we aimed to compare the combinations of hosts parasitized by ticks colonizing the reported range of CCHFV , focusing on explicit relationships between ticks and the hosts reported to support viremia . We explicitly tested the phylogenetic diversity and the centrality ( i . e . , the relative importance in the network of connections ) of the groups of hosts used by these ticks , the resilience of the networks to the removal of hosts , and the existence of clusters of tick-host interactions . We used these findings to elaborate on the specific relationships of Hyalomma spp . with their hosts and determine if these relationships differ from those of other ticks . We also attempted to identify the key factors shaping the circulation of CCHFV mainly by Hyalomma spp . ticks , and ascertain how these specific combinations drive unstable foci of the virus . A species-by-species analysis of the tick-host relationships is not possible , because i ) some ticks species are underreported ( e . g . , prevalence , host preference ) in the literature and therefore an evident bias in the number of hosts is expected; and ii ) the immatures of some ticks are difficult to identify , leading to the reporting of host species that support improperly identified ticks . Following the same reasoning , the analysis of the relationship between ticks and specific host species is not possible; some vertebrates may be very poorly surveyed ( because they are rare , difficult to trap , or protected , etc . ) , which would undoubtedly bias the holistic approach . We therefore used the data on families of hosts for each species of tick , as reported by [15] . The geographical range refers to the complete Palearctic and Afrotropical regions , which are the territories in which CCHFV circulation has been described . Then , data were summarized at the level of tick genera . An estimation of the relative importance of each family of vertebrates was developed in the context of a network of tick-host relationships , at the level of tick genera and life stages ( larvae , nymphs , or adults ) , similar to network approaches commonly performed in other scientific fields [16–18] . A network is a construct that reflects organisms ( nodes ) that interact in any way ( links ) . In our approach , nodes are ticks and vertebrates , and links display the reported finding of a given tick and life stage on a vertebrate of a given family . It is thus a directed network , since ticks have been recorded on hosts . The basic index of a network is the weighted degree ( WD ) , defined as the weighted number of times a group of hosts is recorded for the complete set of tick genera [19] . We calculated the betweenness centrality ( BNC , [20] ) , an index that explains how important a node is in linking several other nodes of either ticks or vertebrates . The ecological significance of the index in our application is immediate: BNC is higher for families of vertebrates that are predominantly used as hosts by several genera of ticks . Separate calculations of BNC for each genus and life stage of ticks give a comparitive overview of the relative importance of the hosts . Clusters of the network were calculated using the algorithm of Neumann [21] . A cluster is a group of nodes that interact more among themselves than with other nodes in the network . Clusters have importance in this context because they reflect groups of vertebrates among which ticks interact more commonly , thus displaying an ecological relationship . Network calculations were done for the complete dataset ( to capture the structure of the complete network of interactions ) , as well as separately for every genus and stage of ticks ( to understand the ecological relationships of every tick genus independently of the rest ) . For each genus and stage of ticks , we calculated the genetic richness of the exploited hosts using Faith’s phylogenetic distance ( PD ) . This metric is based on the sum of distances of the branches that link any pair of families in the phylogenetic tree [22] , and is used to determine if the different genera and stages of ticks parasitize phylogenetically narrow or wide host ranges . PD is an adequate estimate in this context , and supersedes simple measures of host variability based solely on the number of different taxa that serve as tick hosts . We first obtained the phylogenetic tree of the families of hosts , as available in the Open Tree of Life ( OTL , https://tree . opentreeoflife . org ) . The OTL is a repository of phylogenetic trees and produces synthetic trees for a broad range of organisms . It can be accessed through its API to obtain portions of the complete phylogenetic tree stored in the repository . We used a script in the R programming environment to query OTL for the phylogenetic pattern of the families of hosts used by the ticks examined in this study . The resulting tree ( see S1 Fig ) contained data on the phylogenetic relationships of 92 families of vertebrates and was suitable for obtaining estimates of the relative branching of the vertebrates utilized by each genus of tick , but had no calibrated date because it was a synthetic tree . The remaining 40 families of hosts had no information in OTL , and the available information in GenBank was too fragmented to be combined with the already built tree . We evaluated the resilience of the network of tick-vertebrate relationships separately for each tick genus and life stage to understand how random or directed attacks could affect its stability . Resilience of a host-parasite network is an important feature emanating from the network approach , and can be evaluated by removing the hosts either randomly or based on their BNC order in the network . Resilience is measured in terms of the probability of network collapse; the removal of key hosts may lead the network to break down without further links of the parasites to the remaining hosts . We built and obtained indexes of the network structure on the R programming environment [23] using the igraph [24] , bipartite [25] , and picante [26] packages . The resilience of the networks of each genus and stage of ticks after recursive removal of host nodes was evaluated with the package NetSwan for R [27] . Visualization of the networks was done in Gephi v0 . 91 [28] . To analyze the relationships between ticks and hosts with evidence of a potential role in CCHFV circulation , we compiled a list of tick species distributed over the geographical range of CCHFV . This range includes the complete Afrotropical region , the Mediterranean Palaearctic region , and most of Central Asia from the Turkish steppes to India . We focused on 5 tick genera: Amblyomma , Dermacentor , Hyalomma , Ixodes , and Rhipicephalus , which contain species that have been implicated in CCHFV transmission through either studies of natural foci or in the laboratory . All species in the genera were included in the analysis , but not every species included has been reliably linked to CCHFV transmission . Our dataset contained 22 species of genus Amblyomma , 2 of Dermacentor , 18 of Hyalomma , 44 of Ixodes , and 47 of Rhipicephalus , with a total of 1591 pairs of reported associations between the 133 tick species and 132 families of hosts . Phylogenetic calculations were performed on 92 vertebrate families . The network construct had a total of 147 nodes ( genera and stages of ticks , families of hosts ) and 553 links . The complete list of ticks and hosts is included as S1 Table . Values of BNC for each family of hosts ( converted to the range 0–100 to improve comparisons ) are included in S1 Data . The network construct provides a representation of the tick-host relationships , enabling research on the principles behind complex interactions . Fig 1 displays the network and its clusters , explicitly describing the sets of nodes that interact more among themselves than with others ( see also S2 Fig ) . Up to 5 groups or clusters of organisms can be detected , denoting dominant interactions between sets of ticks and vertebrates . Interestingly , each cluster was formed by the 3 life stages of the same genus of ticks , except for Dermacentor and the adults of genus Rhipicephalus . The adults of genus Dermacentor appeared in the same cluster as adults of genus Hyalomma , and the immatures clustered with immatures of the genus Ixodes . The adults of Rhipicephalus formed their own cluster of interacting organisms . The overview of the relationships between ticks and vertebrates ( Fig 1 ) shows 54 families of hosts ( 36 . 73% of total ) that are only slightly important , in terms of BNC , for supporting the network of ticks; these families appear at the periphery of the network . Most Amblyomma species use hosts of the classes Reptilia and Aves , and , in a large proportion , the family Bovidae . Patterns of diversity for Amblyomma spp . are high , with 36 , 35 , and 37 vertebrate families used by larvae , nymphs , and adults , respectively . A similar pattern is seen for Hyalomma spp . : the larvae , nymphs , and adults use 41 , 39 , and 37 vertebrate families . Most of the vertebrates families used as hosts by Hyalomma spp . are small endotherms , including members of the orders Rodentia , Lagomorpha , and Artiodactyla , and the class Aves . Leporidae and Muridae had BNC values of 1222 and 583 for larvae of the genus Hyalomma , and 1332 and 581 for nymphs , demonstrating the importance of these vertebrate families in supporting ticks of this genus . In contrast , Muridae had a BNC value of only 76 for larvae of the genus Amblyomma , and neither Muridae nor Leporidae was parasitized by Amblyomma nymphs . Adults of both Hyalomma and Amblyomma spp . feed on ungulates ( family Bovidae , BNC = 460 for Hyalomma and 1544 for Amblyomma ) . Interestingly , adults of the genus Hyalomma also utilized Leporidae ( BNC = 248 ) . The pattern was completely different for Ixodes spp . , the ticks that feed on the widest variety of vertebrates , including 51 families exploited by larvae , 59 by nymphs , and 56 by adults . Every stage of this tick genus was widely and loosely distributed over a wider variety of hosts than other tick genera . Rhipicephalus spp . were highly eclectic in host preference , and the 3 life stages of this genus do not cluster in a discrete group , preferring Bovidae , Canidae , and a variety of Carnivora hosts during the larval , nymphal , and adult stages , respectively . Results for the genus Dermacentor , with only 2 species included in analysis , showed that immatures mainly parasitize Muridae , while adults concentrated on Suidae . To evaluate whether ticks of each genus were restricted to a wide or narrow host range according to PD of vertebrate families , we aimed to capture the PD of the hosts supporting every genus and stage of the ticks examined . For example , a tick genus may use several vertebrate families that are phylogenetically very related and thus uses a narrow range of hosts , or utilize a few vertebrate host families that are phylogenetically distant , thus covering a broad range of phylogenetic diversity . Analysis resulted in a tree containing data on the phylogenetic relationships of 92 vertebrate families ( S1 Fig ) . We found that the PD of the different genera and stages of ticks varied highly ( summarized in Table 1 ) . With the exception of the genus Dermacentor , the genus Hyalomma showed the lowest PD , even though Hyalomma spp . can parasitizes a number of host families similar to other tick genera . Although the low PD for genus Dermacentor was notable , only 2 species of this genus were included in the study , compared to 18 species of Hyalomma ticks . From the results in Table 1 and Fig 1 , we summarized both BNC and PD for each host family , tick genus , and tick life stage ( Figs 2 and 3 ) . Hyalomma was i ) the genus with lowest phylogenetic diversity of hosts during all 3 developmental stages; and ii ) the only genus the immatures of which concentrate on Leporidae and Muridae hosts while the adults fed mainly on Bovidae . This was demonstrated by high BNC values of these vertebrate families for Hyalomma spp . , which was not seen for other tick genera . Most important in this context , adult Hyalomma spp . ticks have also been found associated with Leporidae hosts . Additionally , a few Aves species have a relatively high importance as hosts for Hyalomma spp . , and are involved mainly in circulating immature ticks . Many complex systems display a surprising degree of error tolerance . However , networks with prominent hubs have low resilience and are extremely vulnerable to attacks ( that is , to the selection and removal of a few nodes that play a vital role in maintaining the connectivity of the network ) [29] . We aimed to capture the behavior of the networks of each tick genus and stage after recursive removal of host nodes . Resilience is an important feature of the ecological networks in which some organisms ( ticks ) depend on the presence of others ( hosts ) that may be key for the circulation of the parasite . After removing each node , the complete network was recalculated and its connectivity was re-evaluated . The percent of connectivity loss was the key measure of the resilience of the network to attack ( random attack , or removing in order of decreasing BNC or decreasing WD ) . These calculations could not be done for the genus Dermacentor , because the connectivity dropped unrealistically after the removal of only a few nodes due to the limited number of species studied . The networks of every stage and genus of tick analyzed were very resilient to random removal of hosts , all of them resulting in a loss of ~75% of connectivity after the removal of 50% of the nodes according to decreasing BNC of hosts ( Figs 4 and 5 ) . Random removal of host nodes promoted higher loss of connectivity in every network . However , lowest resilience of the networks was obtained when they were subjected to removal of hosts according to their WD; Hyalomma larvae and nymph results were deeply affected by removing as few as 2% of hosts with the highest WD . Removing Bovidae , Muridae , and Leporidae , host families that have been reported to develop consistent viremia upon CCHFV infection , resulted in almost complete collapse of the Hyalomma spp . larvae and nymph networks . Removing Bovidae , Leporidae , and Suidae promoted a ~50% loss of connectivity for Hyalomma spp . adults . Although immatures of the genus Amblyomma were also affected by the removal of hosts according to their WD , these ticks were most affected by removing lizards and amphibians , which are not known to be involved in the CCHFV lifecycle . The ecological significance of these findings is that Amblyomma spp . larvae and Hyalomma spp . immatures depend on key vertebrate families as hosts . Notably , the hosts on which Hyalomma spp . mostly depend are of pivotal importance for CCHFV circulation . The main vectors of CCHFV are considered to be ticks of the genus Hyalomma . However , viral transmission has been confirmed under laboratory conditions in ticks of other genera co-occurring with Hyalomma . What differentiates CCHFV vector capacity of Hyalomma spp . from that of other ticks , including those capable of virus transmission , is not clear . We examined ecological factors to investigate whether special characteristics of the communities of hosts used by each tick genus could have a role in CCHFV epidemiology and distribution . The main aim was to describe the ecological relationships among ticks and vertebrates , and to discern if distinct interactions could capture the prominent role of the tick genus Hyalomma in CCHFV circulation . CCHFV is well known to circulate through the 3 stages of the tick developmental cycle . The virus persists in ticks through the developmental stages by transstadial survival , and is maintained in new tick generations by transovarial passage [5 , 30] . Brief viremia in vertebrate tick hosts is the bridge by which the virus accesses other ticks . CCHFV can also infect ticks by transmission among co-feeding ticks , a process by which uninfected ticks feeding in close proximity with infected ticks on non-viremic hosts become infected [31] . Field surveys systematically report clumped distributions of ticks on vertebrates [32]: a few hosts carry large numbers of ticks aggregated in close proximity , while most of the hosts carry few or no ticks . This is of particular interest for the co-feeding mechanism , since the ticks concentrate highly on small mammals , which develop longer viremia and serve as hosts for immature ticks ( reviewed by [1] ) . Since these groups of hosts are important for CCHFV transmission , the preferences for them would concentrate most of the tick populations on key carriers for viral circulation , and increase the probability of CCHFV transmission by co-feeding . Here , based on network analyses of CCHFV associated tick-host relationships , we found that every genus of ticks examined , except Dermacentor , had its own set of preferred hosts; the exception was probably because only 2 Dermacentor species were included in the dataset . Also , our results suggest that Rhipicephalus adults prefer a set of hosts completely different from those exploited by immatures of the same genus . Most importantly , we identified unequivocal host relationships of Hyalomma genus ticks . Immatures of this tick feed on only a few members of Rodentia , Lagomorpha , and Aves , while adults are tightly associated with large ungulates , with lower but still prominent preferences for Lagomorpha and Suidae . Rodents , lagomorphs , and ungulates , but not birds ( with the exception of ostriches [33] ) , have been demonstrated to develop viremia for a variable , but brief , period of time [1 , 34 , 35] . Values of centrality of these hosts show that they are of major importance for Hyalomma spp . Other genera of ticks may use similar groups of hosts , but are also widely distributed over many other host families . In other words , the immatures of the genus Hyalomma tend to concentrate and over-aggregate on vertebrates that have been shown to be important in CCHFV transmission , even if viremia in these hosts is transient . Other genera of ticks feed on these same hosts , but also on a wide array of alternative hosts that are not known to circulate the virus . This feature has been called the dilution effect for other tick-transmitted pathogens , like B . burgdorferi [36 , 37] . While the effect seems not to be universal [38 , 39] , an adequate balance of carrier and non-carrier hosts that ticks can use would attenuate the prevalence of a pathogen in ticks . We explicitly propose that the particular associations of Hyalomma spp . with their hosts are responsible for the prominent role of this tick genus in CCHFV circulation . It must to be noted that , depending on abundance of various host species in a given region , ticks of other genera could theoretically circulate the virus also , as demonstrated in laboratory protocols [4] . This scenario of immature Hyalomma ticks over-aggregating on some key vertebrates while adults infest large ungulates ( which may be also viremic ) has been reported as the main driver of CCHF epidemics [40] . Abandoned agricultural areas become populated by large patches of natural flora , facilitating shelter for rodents , birds , wild suids , and wild ungulates . We hypothesize that the overpopulation of these key hosts increases the abundance of Hyalomma spp . ticks , which could fuel CCHFV prevalence rates in the vectors in a feedback mechanism . As more ticks become infected , the probability of infecting reservoirs and naïve ticks and of transmittng the virus to humans increases . Further analyses suggest that Hyalomma spp . ticks are associated with a phylogenetically narrow spectrum of hosts , accounting for the low resilience of the network of hosts for this tick genus . The lowest PD value was obtained for genus Hyalomma together with Dermacentor , but the result for the latter was considered biased , since only 2 Dermacentor species were included in this study . Other genera of ticks , including Ixodes , Amblyomma , and Rhipicephalus , had significantly higher values of host PD , suggesting that these ticks feed on a much wider range of hosts . In other words , the lack of significant host preference for the immatures of these tick genera would result in greater stochasticity in tick abundance , affecting the circulation of the virus . CCHFV epidemiology is characterized by silent persistence of virus foci with intermittent epidemics . When abundance of key hosts is low , Hyalomma tick populations could also remain low , enough to further circulate CCHFV ( probably only by transovarial passage [31] ) , but well below the threshold ( R0 ) necessary to break the barrier of contact with humans . Small changes in the composition of vertebrate hosts could slighlty increase the value of R0 , leading to the few CCHF cases reported annually in endemic countries [41] . Expansion of key host populations would lead to CCHF epidemics . It is necessary to stress that the only data about CCHFV distribution come from the detection of human clinical cases . Therefore , no information exists about tick densities or serology in hosts for areas where the virus circulates at levels below the epidemic threshold . When the key hosts for CCHFV circulation are absent , immature Hyalomma spp . ticks would use other hosts that are not viremic , deeply affecting the prevalence of the virus in these vectors . The approach of this study is purely ecological and is based on the tenets of the network theory , which has deep roots in social behavior [42] , links among computers [43] , or mutualism between plants and pollinators [44 , 45] . An unbalanced representation of the tick-host interactions could constrain the results of this development , since poorly collected species could introduce a bias in the total number of records . We , however , evaluated the strength of associations between partners using a purposely inclusive systematic division of ticks ( genera ) and hosts ( families ) to prevent the noise generated by undersampled species , together with robust markers representing the relationships in directed networks [46] . This approach guarantees a minimum bias in indexes of the network but not in PD estimations . Furthermore , this approach reduces potential for biogeographical bias based on co-distribution of both tick and hosts in cluster analysis , and supports that the reported cluster formation is derived from an acual preference towards particular vertebrates . A species-by-species analysis of tick-host relationships is not possible here , as detailed earlier . Our broad approach does not account for CCHFV strains ( e . g . , AP92 , lineage Europe 2 ) that , in addition to circulation by Hyalomma spp . ticks [47] , have been suggested to be circulated predominantly by other species such as those of the genus Rhipicephalus [48] . It should be noted , however , that no proof of the vectoral capacity of Rhipicephalus spp . ticks for strain AP92 has ever been obtained under adequate laboratory conditions [2] . A broad approach also prevents the ability to delineate diverse viral lineages circulated by different species of Hyalomma in our analyses . However , these relationships , which are likely due to overlapping geographical ranges of both ticks and viral strains , do not affect the observations detailed here . Analyses are based on inclusion of all tick reports , irrespective of Hyalomma species , and serves as a broad investigation on what differentiates Hyalomma ticks ecologically from other genera of potential CCHFV tick vectors . While our approach is validated by field data supporting the importance of a critical combination of hosts that coexist during the life cycle of Hyalomma ticks , we must take into account the complete lack of data regarding the intimate molecular relationships of CCHFV with the tick . The tick gut represents the first barrier against pathogens , and the gut cell membrane is the key to dissemination of the virus into the body of the vector . The need to conduct these studies under high biocontainment has precluded the understanding of basic mechanisms that CCHFV uses to enter the tick gut and to disseminate to salivary glands for further circulation . While the ecological hypothesis that we outlined here provides a meanignful interpretation of CCHFV dynamics in the field , the relationships between the molecular machinery of the virus and the tick as an environment must be understood . Our current knowledge on CCHFV distribution has been gathered from reported human clinical cases , which provide a fragmented picture of the much wider geographical range of the virus . Surveying the virus in wild hosts and questing ticks , together with an extensive record of tick-host relationships , is urgently needed to update exisiting data about this potentially lethal agent . Viral foci must also be associated with adequate definitions of the environmnetal niche to make sense of the elusive behaviour of the so-called silent foci . These studies , together with a deeper knowledge of the molecular mechanisms shaping the virus-vector interactions , are esential to identify the routes of CCHFV circulation and the exposed populations , and to outline adequate preventive mesaures .	Crimean-Congo hemorrhagic fever virus ( CCHFV ) , a cause of severe hemorragic disease in humans , is maintained in nature in a tick-vertebrate-tick enzootic cycle characterized by silent persistence of endemic foci with intermittent epidemics . Most studies support ticks of the genus Hyalomma as the main CCHFV vectors , but some laboratory reports and field studies also allude to other tick species that may be responsible for virus circulation . Here we converted the tick-host interactions of 133 species of ticks with reported geographical distribution over the range of CCHFV into a network of relationships . By a series of network analyses , we found that immatures of the genus Hyalomma are unique in their hosts preferences among the examined tick genera . Immatures of the genus Hyalomma concentrate on rodent and rabbit hosts , which most efficiently support CCHFV maintenance and transmission by ticks in nature . Based on these data , we formulate the hypothesis that the ecological relationships between Hyalomma spp . and their hosts form a delicate equilibrum that differentiates human epidemics from periods of silent CCHFV maintenance .	[ "Abstract", "Introduction", "Materials", "and", "methods", "Results", "Discussion" ]	[ "taxonomy", "invertebrates", "medicine", "and", "health", "sciences", "ixodes", "pathology", "and", "laboratory", "medicine", "vertebrates", "animals", "developmental", "biology", "phylogenetics", "data", "management", "nymphs", "network", "analysis", "ticks", "infectious...	2018	Host preferences support the prominent role of Hyalomma ticks in the ecology of Crimean-Congo hemorrhagic fever
Polymorphic inversions contribute to adaptation and phenotypic variation . However , large multi-centric association studies of inversions remain challenging . We present scoreInvHap , a method to genotype inversions from SNP data for genome-wide association studies ( GWASs ) , overcoming important limitations of current methods and outperforming them in accuracy and applicability . scoreInvHap calls individual inversion-genotypes from a similarity score to the SNPs of experimentally validated references . It can be used on different sources of SNP data , including those with low SNP coverage such as exome sequencing , and is easily adaptable to genotype new inversions , either in humans or in other species . We present 20 human inversions that can be reliably and easily genotyped with scoreInvHap to discover their role in complex human traits , and illustrate a first genome-wide association study of experimentally-validated human inversions . scoreInvHap is implemented in R and it is freely available from Bioconductor . Frequent polymorphic inversions contribute to adaptation and phenotypic variation [1 , 2] . However , their global contribution to complex traits remains unknown because there is no specific high-throughput technology to genotype inversions in large cohorts . Previous methods have successfully used SNP data to detect the presence of polymorphic inversions by linkage differences at the breakpoints [3–5] as well as to infer inversion genotypes from the mapping of inversion status to haplotype groups , when the breakpoints are known [6–8] . While inversion calling can be performed by the congruence of different SNP signals [8] , only a limited amount of experimentally-validated inversion genotypes have been available for assessing reliable inferences in large cohorts . As such , large association studies that infer inversion genotypes from SNP data have been limited to three human inversions [9–11] . Those inversions that have been successfully genotyped at large scale are either tagged by SNPs ( e . g . inv-17q21 . 31 ) or their genotypes fully explain the clustering of individuals in the first principal components ( PCs ) of the SNPs within their breakpoints , such as inv-8p23 . 1 [6] or inv-16p11 . 2 [10] . In the later cases , the subject clusters correspond to different haplotype-genotypes ( e . g . A/A , A/B , B/B ) of divergent haplotype groups ( A and B ) , supported by the suppression of recombination between inversion states ( inverted: I , non-inverted: N ) [7] . Few individuals can be then selected for costly experimental genotyping , with methods like FISH , to help labeling the clusters according to the inversion-genotypes ( I/I = A/A , I/N = A/B , N/N = B/B ) . The genotypes of the rest of the subjects are then inferred by haplotype-genotype cluster membership [6] . This unsupervised inference , with posterior experimental labeling of the clusters , has allowed the genotyping of inversions in large cohorts [6–8] . However , this approach is still very limited because individual inferences are based on the analysis of entire population samples , making them computationally inefficient [9] and forcing the reanalysis of the whole dataset when new individuals are included . In addition , it has been observed that some inversions exhibit multiple clusters that exceed the three inversion genotypes and therefore their labeling is unclear [10 , 12] . Current methods do not address the needs required for the meta-analyses of inversion association studies that include efficient and reliable genotyping in large population samples and inversion-genotype harmonization across different sources of SNP data . To tackle these problems , we developed scoreInvHap , a novel inversion-genotype classifier that enables the inclusion of inversions in regular GWASs . scoreInvHap compares how similar the SNPs of a new individual are to those in reference haplotype-genotypes , previously linked to reported experimental inversion-genotypes . Our current implementation enables the efficient and reliable genotyping of 20 human inversions in large cohorts . We studied the performance of the method on the inversion calling of inv-8p23 . 1 and inv-17q21 . 31 against two other methods ( invClust and PFIDO ) in a wide range of data types: whole genome sequencing , four SNP microarray studies and two exome datasets . We also evaluated the performance of scoreInvHap in inv-7p11 . 2 and inv-Xq13 . 2 [13] , showing that scoreInvHap can confidently call inversions with multiple haplotypes . We illustrate how scoreInvHap can be used to replicate previous associations of inv-8p23 . 1 and inv-17q21 . 31 with autism and schizophrenia , and to perform a genome-wide association study of 15 inversions on breast cancer . scoreInvHap can generate reliable and scalable inferences for 20 human inversions , whose experimentally validated inversion-genotypes are highly concordant with the haplotype-genotypes of the European individuals of the 1000 Genomes Project [14] ( S1 Text , S1 Dataset ) . These inversions can be genotyped with scoreInvHap in any GWAS of European individuals , showing high prediction accuracy on experimental genotypes not used in the classifier ( Table 1 , S1 Table ) . Six of these inversions cannot be called with previous methods as they support more than two inversion-haplotypes , revealed by the presence of more than three clusters in the first PCs of the SNPs within the inverted region . We demonstrated that haplotype-inversion labeling for these inversions is recovered at higher PC dimensions , where subject clusters are reliably mapped to numerous haplotype-genotype groups ( Fig 1 ) . Using a coalescent simulator for inversions [15] , we observed that the existence of more than two haplotype groups is a common feature ( S1 Fig ) . We studied the performance of scoreInvHap on the inversion calling against invClust [8] and PFIDO [6] . First , we assessed the methods’ accuracies to predict experimental genotypes in the European subjects of the 1000 Genomes project . We found that invClust and PFIDO had low accuracy for inversions with more than 2 haplotypes ( Table 1 ) or when the first MDS component did not completely capture the inversion genotype groups ( S1 Text ) . Second , we tested how sample size affected the calling accuracy of the methods for inversions inv-8p23 . 1 and inv-17q21 . 31 . We resampled varying number of individuals from 1000 Genomes Project and observed that scoreInvHap had high accuracy even when using only one individual reference per haplotype-genotype , whereas invClust and PFIDO required at least 20 and 30 subjects for accurate classification ( S2 Fig ) . We also run the three methods on the African individuals of 1000 Genome project for inversion 8p23 . 1 and observed that scoreInvHap and invClust had lower accuracy ( 85% ) , while PFIDO was unable to return a classification . These results can be explained by a low concordance between haplotype-genotypes and experimental inversion-genotypes . Nonetheless , we could not completely rule-out experimental error that penalized the methods’ accuracy ( S3 Fig ) . Third , we compared the three methods on the inversion calling of inv-8p23 . 1 and inv-17q21 . 31 ( Table 2 , S1 Table ) among studies with different sources of SNP data: four SNP microarray studies and two exome sequencing datasets ( S2 Table ) . The four SNP microarray studies came from trio studies , so we could evaluate the transmission errors . Although the three methods returned similar inversion frequencies ( S4 and S5 Figs ) , we observed that scoreInvHap and invClust had very low transmission errors while PFIDO underperformed ( Table 2 ) . We did not find any substantial differences among the accuracies of the methods in imputed data ( S6 and S7 Figs , Table 2 ) . Inversion calling in the UK10K exome data allowed us to demonstrate the suitability of the method under low SNP coverage . We observed that scoreInvHap returned consistent inversion frequencies with those observed for the Europeans of the 1000 Genomes , while PFIDO’s frequencies were significantly different and invClust failed to identify the inv-8p23 . 1 genotype clusters ( Fig 2 , S8 Fig ) . Finally , we compared the runtime of the three methods on one of the trio datasets ( SSC 1Mv3 ) and found that the parallel version of scoreInvHap was the fastest method ( S3 Table ) . We then demonstrated that the method is efficient in calling genotypes of inversions with multiple haplotype groups . We specifically studied the performance of scoreInvHap to call inversion genotypes of inv-7p11 . 2 and inv-Xq13 . 2 , the largest inversions with multiple haplotype-genotypes ( Table 1 , S1 Table ) . We observed that scoreInvHap classification matched true inversion-genotypes under low SNP densities ( 10% of original SNP coverage ) , for both inversions ( S9 Fig ) . Testing the performance in SNP array data , we observed consistent inversion frequencies with those found for the Europeans of the 1000 Genomes project ( S4 Table , S10 and S11 Figs ) and found low transmission errors ( S5 Table ) . We applied scoreInvHap to validate initial association analyses of inv-17q21 . 31 and inv-8p23 . 1 with autism ( cases/controls = 604/5 , 529 ) and schizophrenia ( cases/controls = 1 , 308/5 , 528 ) , using the exome data of UK10K studies [16] . Note that scoreInvHap is the only method that allows testing associations with inv-8p23 . 1 since inversion calling from such a low coverage data could not be performed with other methods ( Fig 2 ) . We tested the associations under three inheritance models , adjusting by genome-wide PCs ( Fig 3A ) . We replicated a significant association between schizophrenia and inv-8p23 . 1 ( additive OR = 0 . 91 , P = 4 . 9×10−2 ) and inv-17q21 . 31 ( additive OR = 0 . 84 , P = 1 . 4×10−3 ) . However , we did not replicate the association with autism ( Fig 3A ) where we could not rule-out remaining differences in genetic ancestry between the studies nor the lack of power for a study with 604 cases and 4358 controls to detect OR~1 . 12 ( power = 0 . 466 ) , as computed with Genetic Association Study ( GAS ) Power Calculator [17] . Finally , to illustrate a first genome-wide association study of experimentally-validated human inversions , we tested the association between breast cancer and 15 inversions of Table 1 that could be reliably called in a GWAS study of 1 , 061 cases and 1 , 033 controls [18 , 19] . We did not detect any significant association adjusting for genome-wide PCs , age and multiple comparisons ( Fig 3B ) . However , we did observe associations at a nominal significance level for inversions at 7p11 . 2 ( additive OR = 1 . 14 , P = 4 . 2×10−2 ) , 6p21 . 33 ( recessive OR = 1 . 36 , P = 1 . 8×10−2 ) and 6q23 . 1 ( recessive OR = 4 . 30 , P = 3 . 5×10−2 ) , which should be further investigated in larger association studies . These applications demonstrate that scoreInvHap is a robust genotyping tool of inversions , easy to use on already available GWAS data . We developed scoreInvHap , a new bioinformatics tool to call inversions from SNP data . Its main advantage is the quick call of inversion genotypes from SNP data at the individual level with consistent genotype labeling . As a consequence , inversion genotyping is readily harmonized . Another important advantage is that the method allows the calling of inversion-genotypes using different sets of SNPs . As a result , inversions can be called on datasets with lower SNP coverage than the dataset used for the references as well as to call inversion-genotypes on individuals with missing SNP genotypes . Previous bioinformatics methods relied on applying a dimensionality reduction technique to SNP data followed by clustering the individuals . Although these methods have been used to associate chromosomal inversions to phenotypic traits [20] , they have some limitations . First , these methods partition a population sample into inversion-genotypes but require external information for labeling the inverted-homozygous group , challenging the harmonization of inversion calling in multi-centric studies . Second , the methods are computationally intensive and are inefficient for calling inversion genotypes in large cohorts [9] . Finally , they require a minimum number of individuals to compute accurate calls , so the whole dataset needs to be recalled to include inferences in new individuals . In contrast , the link between haplotypes and inversion status is previous to scoreInvHap classification . Consequently , scoreInvHap classification is readily comparable across different studies and genotyping techniques ( from SNP array to exome data ) , allowing the harmonization of inversion calling in large meta-analyses . As the method classifies each individual separately , further gains in computational efficiency can be obtained from processing large datasets by batches allowing the genotyping of multiple inversions to be included in association studies . scoreInvHap is the only method designed to genotype inversions with multiple haplotypes , whose abundance in the human genome is likely underestimated . We found inversions with multiple haplotypes on simulations under neutrality and in the inversions reported in invFEST and 1000 Genomes . This result suggests that the less common presence of only two haplotypes in inversions inv-8p23 . 1 and inv-17q21 . 31 could be due to the reported selection process that occurred in these regions [6 , 11 , 21] . Inversions supporting three or four haplotypes have already been described in the literature [10 , 12] . For inversions inv-7p11 . 2 and inv-Xq13 . 2 , Aguado and colleagues generated inversion-haplotype trees [13] . Based on the major branches of these trees , they observed that both inversions support four possible haplotype groups , where inv-7p11 . 2 supports two standard and two inverted haplotypes and inv-Xq13 . 2 supports three standard and one inverted haplotypes . The tetrahedron structure that we observed for the first three MDS components of these inversions clearly matched the phylogeny of the haplotypes . Sanders and colleagues described more than 100 polymorphic inversions based on a single cell sequencing method [22] . Most of these inversions have not been previously detected with bioinformatics methods designed for inversions with two haplotypes . Therefore , an assessment of the haplotype complexity of these inversions for inference in large association studies is warranted . Further research is also needed for establishing the frequency of complex haplotype patterns in inversions and for elucidating the mechanisms involved in the formation of divergent haplotype groups , supported by the presence of an inversion polymorphism . scoreInvHap , nonetheless , also has limitations . In particular , its performance depends on the representativeness of the reference inversion-genotypes . For inversions inv-8p23 . 1 and inv-17q21 . 31 and European samples , we captured the haplotypic variability of the inversions using only one reference per inversion genotype . However , scoreInvHap needs to increase the number of experimental references for inversion calling in population samples with higher within haplotype variability , such as inv-8p23 . 1 in Africans . Further studies are needed to determine the accuracy of the method in inversions with larger genetic variability or populations with admixture . The inversion genotyping by scoreInvHap , like other SNP based methods , is indirect: it does not detect the change of DNA orientation but relies on the haplotype structures generated by inversions . Therefore , it has some clear limitations against experimental methods to detect inversions , such as iPCR [13] , next generation sequencing or single strand sequencing [22] . In particular , scoreInvHap cannot detect small , recent or de novo inversions , as these inversions do not support different haplotype groups . In addition , scoreInvHap will produce wrong classifications for recurrent inversions , where the same haplotype can be found in standard and inverted chromosomes . Despite these limitations , scoreInvHap has the advantage of working with stringent conditions of SNP coverage and sample sizes . All inversions in Table 1 can be genotyped with scoreInvHap using SNP data in common formats , like PLINK , snpMatrix or vcf . Performing the genotyping of new inversions in large studies , in human and other species , can be achieved by creating their classifiers within scoreInvHap . To build a new classifier , the first step is to demonstrate that a reference sample of individuals can be clustered into haplotype-genotypes . The second step is to show that haplotype-genotypes are unambiguously labeled by experimentally inversion-genotypes . Finally , the reference haplotype-genotypes can be included in the program for genotyping the inversion in new individuals . We showed how scoreInvHap inferences can be used to perform association studies , but additional analyses are needed to understand how the inversion affects the phenotype . One option is that the positional change caused by the inversion affects the regulation of nearby genes , leading to phenotypic differences between individuals . Another option is that the inversion captures the allele ( or a combination of alleles ) that are the causal variants . Structural variants , such as deletions , copy number alterations or complex re-arrangements , can also be captured by the inversion status and produce the phenotype . Only further analyses can elucidate the mechanism linking an inversion to a phenotype . In summary , scoreInvHap can reliably perform inversion calling for large multi-centric studies with SNP genotype data . The method has been implemented for the calling of 20 human inversions which can be immediately included in any GWAS , to forward our understanding of the role of inversions in complex traits . The method is easily extended to other inversions , in humans and other species , as soon as experimental inversion genotypes become available . Inversions suppress recombination within the inverted segment when heterozygous . Therefore , for an ancient non-recurrent inversion , two divergent haplotype groups emerge for each inversion status [7] . Haplotype groups that map to a single inversion status are defined as inversion-haplotypes . In this model , standard and inverted homozygous can be considered as subpopulations where chromosomes belong to the same haplotype group while individuals that are heterozygous for the inversion belong to a 1:1 mixture of standard and inverted chromosomes . This mixture can be seen in the first Multi Dimensional Scaling ( MDS ) components of the SNPs within the inverted region . In the simplest cases ( i . e . inv-17q21 . 31 and inv-8p23 . 1 ) , two clear haplotype groups ( A and B ) emerge for each inversion status ( N and I ) , resulting into three differentiable clusters , or haplotype-genotypes , on the first MDS component ( Fig 1A ) . Heterozygous haplotype-genotype individuals ( AB ) are visualized equidistant to the homozygous haplotype-genotype groups ( AA/BB ) . Therefore , a univocal map , given by experimental validations , between inversion status and haplotype groups can be established ( A = N , B = I ) . However , in other inversions , more than two haplotype groups have been observed . Inversion at 16p11 . 2 shows , for instance , a pattern consistent with two haplotype groups ( A , C ) in the standard configuration and one haplotype group in the inverted allele ( B ) [10] . In the first MDS components of the SNPs in the region , one can see that heterozygous haplotype-genotype clusters ( i . e . AC ) are equidistant to their respective homozygous haplotype-genotype clusters ( AA and CC ) , forming a triangular 6-cluster pattern ( Fig 1B ) . Experimental validation is therefore needed to correctly assign an inversion status to each haplotype group ( A , C = N , B = I ) . More complex scenarios are also possible , where four haplotype groups are observed in the region , supporting ten clusters in the first three MDS components consistent to all possible haplotype-genotypes . Experimental inversion-genotypes are then needed to identify the inversion status to which the haplotype-genotypes map . We studied 59 inversions reported in the 503 European individuals of the 1000 Genomes projects . For each inversion , we perform an MDS analysis for all SNPs within the inverted region and studied whether the clustering conformed to a model where haplotype-genotypes could be unambiguously defined . We , therefore , selected the inversions that followed any of the patterns illustrated in Fig 1 , increasing the number of MDS components , from 1 to 3 , until one of the patterns was clearly identified . This heuristic procedure is described in S1 Text and can be used as a guideline to extend scoreInvHap to new inversions . Each cluster of individuals was then identified as a reference group for a given haplotype-genotype to which new individuals are compared for inferring their own haplotype-genotypes . Consequently , at least one reference individual is needed for each haplotype-genotype . The haplotype-genotypes were then mapped to experimental inversion-genotypes to determine their inversion status . At this stage , we measured the degree of concordance between the haplotype and inversion genotypes by their percentage of agreement across individuals , accounting for the cases where more than one haplotype group was found in a single inverted status . We developed scoreInvHap for inversions that could be consistently mapped to haplotypes . Therefore , scoreInvHap is suitable for those inversions for which the clustering pattern presents no haplotype sharing between the inverted and standard status , and where individuals can be reliably classified into haplotype-genotype groups . We considered that both conditions were fulfilled when clusters followed at least one of the inversion-haplotype mappings in Fig 1 . scoreInvHap computes a similarity score between a subject’s SNP genotypes in the inverted region and the haplotype-genotypes that map to experimentally validated inversion-genotypes . Note that the mapping is at the level of SNP and haplotype genotypes and not on individual chromosomes . As such , no phasing is needed for the inferences . scoreInvHap then classifies a new individual into the reference haplotype-genotypes for which their link to inversion-genotypes has been established . The classification is based on similarity scores between the SNP genotypes of the individual and the SNPs in each haplotype-genotype reference group . To compute the score , we first build the classifier from the frequency of each SNP i in each reference haplotype-genotype k made of Mk reference individuals , fki ( xi ) =nk ( x ) Mk where fki is the frequency of the i-th SNP genotype x = {0 , 1 , 2} in the haplotype-genotype reference group k . The frequency is the ratio between the number of reference individuals ( nk ) in k with SNP genotype xi and Mk . The score of a subject S , with L ( L ⊆ N ) SNP genotypes in the inverted segment ( s1 , …sL ) , s = {0 , 1 , 2} , in the haplotype-genotype reference group k is defined as Hk=∑i=1Lfki ( s ) ·ρi2∑i=1Lρi2 where ρi2 is the maximum linkage disequilibrium between the SNP i and the haplotype groups in the reference individuals . For inversions with two haplotypes , ρi2 corresponds to the linkage disequilibrium R2 between SNPi and the inversion-genotypes . For inversions with three haplotypes ( A , B and C ) , we compute the R2 between SNPi and each haplotype-genotype . For instance for haplotype A the three haplotype-genotype are given by RR: {BB , BC , CC} , RH: {AB , AC} and HH: {AA} . We use these three haplotype-genotypes to compute the R2 between the haplotype group A and SNPi in the reference individuals . ρi2 is then , the highest R2 across A , B and C . The inferred haplotype-genotype of the individual S is , therefore , the genotype for which the score is maximum , that is arg ( max{H1 , …HJ} ) where J is the total amount of haplotype-genotypes; that is , 3 haplotype-genotypes for 2 haplotype groups , 6 for 3 groups , 10 for 4 , and so on ( Fig 1 ) . The inversion-genotype for the individual follows from the link between haplotype-genotypes and experimental inversion-genotypes in the reference individuals . For imputed data , the score is computed as Hk=∑i=1L∑si=0 , 1 , 2Pi ( t ) ·fki ( si ) ·ρi2∑i=1Sρi2 , where Pi ( t ) is the probability that the individual S has genotype t . We implemented scoreInvHap in an R package that supports snpMatrix or VCF formats , two standard Bioconductor classes for SNP data . The stable version is available in Bioconductor ( https://bioconductor . org/packages/release/bioc/html/scoreInvHap . html ) while the development version can be installed from the GitHub repository ( https://github . com/isglobal-brge/scoreInvHap/ ) . scoreInvHap requires the SNP genotypes of an individual in the inversion region . The allele frequencies of SNPs in each genotype reference and the ρ2 between the SNPs and the validated inversion genotypes are built in the classifier and included in the package for the 20 inversions described in Table 1 . We have also developed imputeInversion , a wrapper to impute SNP array data to use scoreInvHap . This tool is available from the GitHub repository ( https://github . com/isglobal-brge/imputeInversion ) . We used the SNP data ( MAF > 5% ) of 503 European individuals of 1000 Genomes phase 3 [23] . To test the performance of scoreInvHap , invClust and PFIDO under different conditions , we re-sampled the original dataset under different scenarios . Four scenarios run 200 times each , with different SNPs coverage ( 10% , 20% , 50% and 75% ) for the three methods and six scenarios with different number of individuals ( 5 , 10 , 15 , 20 , 25 and 30 ) for invClust and PFIDO . To evaluate scoreInvHap performance under different number of individuals , subsets for the references varied from 1 to 5 individuals per haplotype-genotype group . Full scoreInvHap performance was tested with a leave-one-out classification approach , classifying one individual with experimental inversion-genotype and using the remaining individuals as references . We analyzed autism cohorts from the Autism Genome Project ( AGP ) [24] and the Simon Simplex Collection ( SSC ) [25] . SSC contained data from three different arrays: Illumina 1Mv1 , Illumina 1Mv3 Duo and Illumina HumanOmni 2 . 5 . We considered each array as a different dataset . To include European subjects only , we run a Principal Component Analysis ( PCA ) using 128 SNP markers for ancestry [26] including all autism cohorts and HapMap3 individuals [27] . We generated a confidence ellipse of 0 . 99999 around European HapMap subjects and we discarded all individuals outside the ellipse . We discarded 111 subjects of AGP . We obtained exome data from the UK10K Neurodevelopment datasets . We analyzed two datasets to compare scoreInvHap to clustering methods: one of schizophrenia cases ( UK10K_NEURO_ABERDEEN ) and another of autism cases ( UK10K_NEURO_ASD_GALLAGHER ) . Both datasets are deposited in the European Genome-phenome Archive ( EGA ) under study accession codes EGAD00001000433 and EGAD00001000436 . To select European individuals , we performed a genome-wide PCA of the merge between the UK10K neurodevelopment datasets and two control GWAS datasets: British Birth Cohort ( BBC ) and National Blood Service ( NBD ) . We discarded subjects outside the central PCA cluster , likewise AGP . We generated four different inversions using invertFREGENE [15] . We used default values of recombination ( 1 . 25×10−7 ) and mutation rates ( 2 . 3×10−7 ) . In all simulations , the entire simulated region was 2Mb while the inversion comprised 800Kb . Stop frequency was set at 0 . 4 for the first three inversions and to 0 . 2 for the forth . We run scoreInvHap in SNP arrays , imputed data and exome data using the inversion-genotype references included in the package . We discarded SNPs with call rate lower than 0 . 9 . We ran invClust using the first two multidimensional scaling components of the SNPs in the inverted regions . We ran PFIDO with the default values of SNPs and subject call rate filtering ( 0 . 9 ) . We forced the model to return 3 groups and set all the other parameters to default . We tested the associations between autism spectrum disorder and schizophrenia , and inversions inv-8p23 . 1 and inv17q21 . 31 in ten UK exome studies of the UK10K project ( S6 Table ) . We used subjects from Welcome Trust Case Control Consortium 2 as controls . This dataset consists of two cohorts ( National Blood Service ( NBS ) Cohort and 1 , 958 British Birth Cohort ) genotyped with Illumina 1 . 2M . We only included individuals classified as Europeans by peddy [28]: 5 , 529 controls , 604 autism cases and 1 , 308 schizophrenia cases . To run peddy , we created two datasets: the first one was the merger between controls and autism cohorts and the other was the merger between controls and schizophrenia . In both cases , we included the 68 , 689 SNPs that were common between the SNP arrays and the exome data . We applied scoreInvHap on each dataset . As cases and controls cohorts belong to different studies , we tested whether the differences in inversion frequencies were not statistically significant ( chi-squared test ) between the two control cohorts , and among the ten cases cohorts . We used SNPassoc for association testing between disease status and inversion genotypes in the joint dataset across all cohorts , adjusting for the joint genome-wide PCs . We tested the association between inversions and breast cancer on the Cancer Markers of Susceptibility ( CGEMS ) study [18 , 19] , available in dbGaP ( dbGaP Study Accession: phs000147 . v3 . p1 ) . We only included individuals classified as European with a probability higher than 0 . 9 inferred by peddy [28]: 1 , 061 cases , 1 , 033 controls . We imputed the chromosomes containing inversions in Table 1 with Michigan Imputation Server [29] . We selected HRC r1 . 1 2016 as reference panel and SHAPEIT as phasing algorithm . We removed SNPs with an imputation R2 smaller than 0 . 4 . We called inversion genotypes with scoreInvHap in the 15 inversions having at least 4 SNPs with high quality imputation . We used SNPassoc for association testing between disease status and inversion genotypes , adjusting for age and the joint genome-wide PCs .	Chromosomal inversions are structural variants consisting on an orientation change of a chromosome segment . Inversions have been linked to some phenotypic differences between individuals and to genetic divergence . However , their overall contribution to complex diseases is largely underdetermined as there are no high-throughput methods to call inversion-genotypes in large cohort studies . Here , we propose a new method , scoreInvHap , to call individual inversion genotypes from their haplotype similarity . We show that scoreInvHap has a high performance when analyzing heterogeneous sources of SNP data . Our current implementation contains 20 human inversions that can be readily genotyped in existing GWAS datasets . We exemplify the utility of scoreInvHap by running the first-genome wide association of experimentally validated inversions and a multi-centric inversion association study . All in all , scoreInvHap can substantially contribute to increase our knowledge of the role of chromosomal inversions in complex diseases by re-analyzing data from existing genetic association studies .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "genome-wide", "association", "studies", "medicine", "and", "health", "sciences", "pervasive", "developmental", "disorders", "variant", "genotypes", "geographical", "locations", "autism", "social", "sciences", "developmental", "psychology", "neuroscience", "genetic", "mappi...	2019	scoreInvHap: Inversion genotyping for genome-wide association studies
Cortical fast-spiking ( FS ) interneurons display highly variable electrophysiological properties . Their spike responses to step currents occur almost immediately following the step onset or after a substantial delay , during which subthreshold oscillations are frequently observed . Their firing patterns include high-frequency tonic firing and rhythmic or irregular bursting ( stuttering ) . What is the origin of this variability ? In the present paper , we hypothesize that it emerges naturally if one assumes a continuous distribution of properties in a small set of active channels . To test this hypothesis , we construct a minimal , single-compartment conductance-based model of FS cells that includes transient Na+ , delayed-rectifier K+ , and slowly inactivating d-type K+ conductances . The model is analyzed using nonlinear dynamical system theory . For small Na+ window current , the neuron exhibits high-frequency tonic firing . At current threshold , the spike response is almost instantaneous for small d-current conductance , gd , and it is delayed for larger gd . As gd further increases , the neuron stutters . Noise substantially reduces the delay duration and induces subthreshold oscillations . In contrast , when the Na+ window current is large , the neuron always fires tonically . Near threshold , the firing rates are low , and the delay to firing is only weakly sensitive to noise; subthreshold oscillations are not observed . We propose that the variability in the response of cortical FS neurons is a consequence of heterogeneities in their gd and in the strength of their Na+ window current . We predict the existence of two types of firing patterns in FS neurons , differing in the sensitivity of the delay duration to noise , in the minimal firing rate of the tonic discharge , and in the existence of subthreshold oscillations . We report experimental results from intracellular recordings supporting this prediction . Among inhibitory neurons in the neocortex , the “fast-spiking” ( FS ) compose the most prominent type . These neurons are characterized by brief action potentials with a width smaller than 0 . 5 ms followed by a deep monophasic afterhyperpolarization ( AHP ) [1 , 2] . Delayed rectifier currents of the types Kv3 . 1–Kv3 . 2 are responsible for these characteristics of FS action potentials [3 , 4] . The firing patterns of FS cells in response to a step of injected current are highly variable . Depending on the neuron and on the amplitude of the current pulse , FS cells fire action potentials immediately after the onset of the current step or after a prolonged delay , which can be on the order of several hundreds of milliseconds [2] . Interestingly , voltage-dependent subthreshold oscillations in the gamma range have been reported during the delay period . They typically occur in a narrow voltage range just negative to threshold [4 , 5] . The steady-state firing pattern is reached after an adapting [6] , non-adapting , or accelerating [2] transient . This steady state can be tonic or bursting . In the latter case , the neuron fires rhythmic irregular bursts of action potentials; this activity pattern is named “stuttering” [7] . The goal of the present modeling study is to address the origin of this variability . Bifurcation theory , which classifies how the behavior of dynamical systems changes as their parameters vary , reveals that in strongly nonlinear systems , qualitatively different dynamical regimes can emerge as a result of a continuous variation of some parameters [8] . Hence , heterogeneities in biophysical parameters of neurons can induce distinct classes of firing patterns even if their distributions are smooth . In the present paper , we propose that a great deal of FS electrophysiological variability is a consequence of heterogeneities in the maximal conductance of a slowly-inactivating d-type K+ current [9 , 10] , known to be present in FS cells [11 , 12] , and in the strength of the Na+ window current . This window current , governed by the overlap between the activation and inactivation curves of the Na+ current , affects the ability of the neuron to fire at low rates [13] . To assess this proposal , we consider a minimal , conductance-based neuronal model that incorporates these two ionic currents and a fast delayed rectifier K+ current . We investigate this model using techniques from nonlinear dynamical system theory . We find that as the conductance of the d-current and the overlap of the activation and the inactivation curves of the Na+ current are varied within a range compatible with experimental data [14] , the model neuron displays a “”variability of firing patterns similar to those observed experimentally in FS cells . Our study leads to several predictions that can be tested experimentally . In particular , we predict that FS cells that fire tonically at low rates do not exhibit subthreshold oscillations during the delay period . We present experimental results consistent with this prediction . In all of this section , we assume that the system is noiseless . We will change the Na+ window current by varying the half-maximum potential , θm , of the activation curve of INa . As shown in Figure 1 , the amplitude of the window current decreases as θm is more depolarized . The strength of the d-current is controlled by its maximal conductance gd . The amplitude of the current step injected into the neuron is denoted by Iapp . For very large window current , i . e . , small θm ( θm < −31 . 4 mV for gd = 0 and θm < −32 . 9 mV for gd = 2 mS/cm2 ) , the neuron is spontaneously active . In contrast , for very small window current , i . e . , sufficiently large θm ( θm > −15 . 2 mV for gd = 0 and θm > −16 . 4 mV for gd = 2 mS/cm2 ) , the neuron remains quiescent for all amplitudes of the step current . In the intermediate range of θm , the neuron is quiescent for Iapp smaller than a threshold Ith , while it fires spikes for Iapp > Ith . Then , depending on θm , two qualitatively different behaviors of the neuron occur . In this section , we clarify the mechanisms underlying the different firing patterns described above and the dependency of the phase diagram on the Na+ window current . The time constant of the inactivation variable b of the d-current is τb = 150 ms ( see the Model section in Materials and Methods ) . Hence , b varies much more slowly than all the other dynamical variables of the model . The full dynamical system describing the neurons can subsequently be separated into fast ( variables V , h , n , and a ) and slow ( variable b ) subsystems . This allows us to dissect the dynamics of our model using the “fast–slow method” [13 , 16–18] . The first step in this method is to study how the attractors of the dynamics of the fast subsystem depend on the value of b , taken as a time-independent parameter . In a second step , one derives the dynamics of the full system taking into account the slow variations of b . The minimal firing rates in the tonic discharge following the delay period is broadly distributed among FS neurons [22–24] . The theoretical results described above established correlations between the minimal firing rate of a neuron and the existence of subthreshold oscillations during the delay period . Neurons that fire at high rates display subthreshold oscillations during that period [22 , 23] . In contrast , neurons that can fire at low rates lack those oscillations . We report here experimental results supporting this prediction . The responses to steps of depolarizing current pulses were recorded intracellularly in FS neurons ( n = 20 , average ratio of the spike's rising phase dV/dt to falling phase dV/dt = 1 . 17 ± 0 . 29 ) , as described in the section “Whole cell recordings and analysis” in Materials and Methods . Eight neurons responded to the depolarization onset almost instantaneously with a tonic non-adapting spike train . The other 60% of the neurons ( n = 12 ) displayed a prolonged delay period , which was preceded by a transient firing of 1–3 spikes in some cases ( Figure 11A and B ) , similar to the simulations of our model for low Na+ window current ( Figure 7C ) . One cell from this group of 12 cells fired irregularly ( Figure 11C ) . Eight out of those 12 cells had properties as reported in previous experimental studies of FS neurons ( see also [22 , 25] ) . These eight cells exhibited a delay period followed by a tonic , high-frequency regular discharge ( Figure 11A ) or stuttering discharge with large , instantaneous intra-burst firing rate ( Fig 11B ) . The minimal average firing frequencies in this group of neurons ranged from 25 to 220 Hz ( average 81 ± 59 Hz ) and the voltage threshold to action potential was −31 . 6 ± 5 . 0 mV . All the computed trial-average power spectra of the membrane fluctuations during the delay period ( n = 4 ) displayed a peak within the gamma range ( 20–100 Hz ) . Two examples are shown in Figure 12A and 12B . These features are similar to those exhibited by our model neuron for small window current ( Figure 9A and 9B ) . Different properties were found in the remaining three neurons ( out of these 12 cells ) that displayed delay to firing . First , their minimal firing frequencies were under 10 Hz ( average 6 . 8 ± 2 . 6 Hz ) ( Figure 11D ) , substantially lower than in the other group of eight neurons . Their spike threshold was −39 . 6 ± 6 . 4 mV , a significantly more hyperpolarized value than in the other group ( p = 0 . 01 , student t-test ) . These neurons did not exhibit stuttering behavior . Finally , spectral analysis of the fluctuations during the delay period failed to reveal subthreshold oscillations ( Figure 12C ) . These properties are as predicted in our model for large Na+ window current ( Figure 9C ) . The minimal model of FS neurons studied in our work displays four types of behavior in response to a current step , depending on the Na+ window current and on the strength of the conductance of the K+ current Id . 1 . When the Na+ window current and the conductance gd are small , the neuron exhibits tonic , high-frequency firing that follows the current step onset almost immediately , even if the step amplitude is just above firing threshold . 2 . When the Na+ window current is small and the conductance of the d-current is of intermediate strength , delayed high-frequency tonic firing occurs for just suprathreshold step amplitudes . The delay duration decreases as the step amplitude increases and abruptly jumps to zero at some critical value . Noise dramatically reduces this duration . Noise also induces subthreshold oscillations during the delay and can also induce stuttering . 3 . For small Na+ window current but large values of the d-current conductance , the response to just suprathreshold input is delayed stuttering with high-frequency firing within the bursts . As the current step amplitude increases , the response becomes tonic firing , first delayed and subsequently and abruptly non-delayed . Other properties are as in 2 . 4 . For large value of the Na+ window current , the neuron responds with delayed tonic firing for small amplitude of the current step and non-delayed tonic firing when the amplitude of the current step is large . In contrast to what happens in 2 , the average firing rate is low near firing threshold; noise very weakly affects the delay duration and does not induce subthreshold oscillations during the delay period . A large spectrum of K+ currents with different activation and inactivation properties and kinetics has been reported in FS neurons [7 , 11 , 26] . Delayed rectifier K+ channels from the Kv3 . 1–Kv3 . 2 types are responsible for the fast spike repolarization and strong AHP of these neurons [3 , 4 , 27] . Slowly inactivating K+ channels from the Kv1 . 1 , Kv1 . 2 , and Kv1 . 6 types have also been found in FS cells [12]; blockade of these currents with DTx-I eliminates delays to firing present in the control situation . The d-current incorporated in our model can be thought of as representing these slow channels . We are aware of a single experimental study where the activation and inactivation properties of the Na+ channels in FS neurons were measured [14] . One conclusion of that study is that the overlap between the activation and inactivation curves of this current is small . Clearly this does not mean that the Na+ window current has no effect , since this will depend on the maximum conductance of this channel , a value which is not known . As a matter of fact , in our model , the inactivation mid-point potential , and the gain of the activation and inactivation functions are in accordance with the data provided by Martina and Jonas [14] . We take the same value of the Na+ conductance gNa as in [3 , 28] . With this value , we find variability of firing patterns while varying the activation mid-point potential in a range compatible with the data provided by Martina and Jonas [14] . Throughout this article , we use the half-activation curve of INa , θm , to quantify the strength of the Na+ window current . Effects of modifying the window current by depolarizing or hyperpolarizing θh are similar to the effects of hyperpolarizing or depolarizing θm , respectively ( Figure S2 ) . We did not include a persistent Na+ current in our minimal model because this current was not found in FS neurons [14] . However , if added to our model , this current would have an effect similar to increasing the Na+ window current ( unpublished data ) . Synchronization properties of neuronal networks are tightly related to single neuron properties [29–33] . For instance , increasing the strength of the Na+ window current transforms the bifurcation of the rest state from a Hopf type to an SN type . This change may switch an inhibitory-coupled network of FS neurons from an asynchronized state to a synchronized state [34] . Furthermore , Skinner et al . showed that networks of FS cells that possess both sufficiently strong Na+ window current ( or persistent Na+ current ) and Id , and that are coupled by both inhibitory and electrical coupling [35] , may exhibit collective bursting oscillatory behavior [36] . Hence , the variability in single-cell properties , presented in this article , is very relevant to the network's behavior . The firing patterns exhibited by our model include “classical” non-delayed tonic firing , delayed tonic firing , and delayed stuttering . These three patterns of firing are consistent with those described in recent experimental studies of FS cells ( e . g . , [7 , 22] ) as well as in the experimental results reported in the present study . Increasing Iapp in the model eventually causes the disappearance of the delay . This is also consistent with experimental observations [2] . A large variability is observed between FS neurons in their minimal firing rates in response to steady current . Whereas many FS cells have high steady-state minimal firing frequencies on the order of tens of Hz or more [2 , 23] , the minimum firing rate of other FS neurons can be as low as 20 Hz [22] or even less than 10 Hz [37] . Especially , FS neurons with neurogliaform morphology can fire at low rates [24 , 38] . Although in our experimental data we classified neurons as FS based on their spike width and repolarization rate [1] , and did not examine their morphology , our experimental results are consistent with such variability in the minimal firing rate . Further experimental work is needed to verify whether only FS neurons with neurogliaform morphology can fire at low rates . Relying on our modeling study , we propose that heterogeneities in the Na+ window current contribute strongly to variability in the minimal firing rate . In our model , a delay in action potential firing is induced by a slow crossing of a bifurcation driven by the slowly inactivating d-current . Depending on the window current of the Na+ current , this bifurcation can be of Hopf or SN types . As a consequence , the properties of the neuron during the delay period depend also on the Na+ window current . In particular , subthreshold oscillations are found during that period ( and also during the quiescent periods in stuttering patterns ) . This is consistent with the observation of subthreshold oscillations at frequencies in the gamma-range ( 20–100 Hz ) in FS neurons in cortex [22] during delay or interburst periods . Moreover , in our model , these oscillations exist only when the INa window current is small ( Figure 9A and 9B ) . Subsequently , we predict that subthreshold oscillations in an FS neuron are more likely to be observed in neurons with a large minimal firing frequency . Our experimental results ( Figures 11 and 12 ) are consistent with this prediction . Finally , noise can induce irregular stuttering in our FS model ( Figure 10 ) . Similar patterns were found in previous experiments ( Figure 1C in [23] ) , as well as in the experiments reported here ( Figure 11C ) . There are several physiological observations that the model does not replicate . The AHP and the spike amplitude are larger in the model ( Figures 2 and 7 ) than in FS neurons recorded in slice experiments ( Figure 11 ) . Changing the reversal potential of the K+ current reduces the AHP of the model neuron to some extent . It may also happen that in FS cells , the spike-generating area is distant from the soma , and therefore action potentials recorded in the soma are filtered by cable properties . This effect , which cannot be included in our single compartment model , may contribute to the reduction of the AHP . Another limitation of our model is that it can account neither for the substantial accommodation observed in some FS cells ( AC cells in Figure 5 in [7] ) nor for the burst of action potentials that precedes tonic firing sometimes observed in FS neurons ( “b” cells in Figure 5 in [7] ) . Although our model may exhibit some adaptation , it is a result of the strong AHP , which causes Id to inactivate during firing , but this inactivation is only weak . However , accommodation and initial bursting can probably be accounted for if one incorporates additional slowly activating K+ currents into the model . Finally , stuttering FS cells that do not exhibit an initial delay in response to the injection of a step current are observed experimentally [6] . Such a behavior is not present in the phase diagram of Figure 2A . However , it can occur in the framework of our model if the reversal potential of the leak current is taken to be more depolarized ( e . g . , VL = −60 mV in Figure S1B ) than the reference parameter set VL = −70 mV ( Figure 1 ) . To our knowledge , our work is the first to propose a minimal conductance–based model incorporating ionic channels known to exist in FS cortical interneurons and which accounts in a comprehensive way for the variability of firing patterns these cells display . The analysis we have made of this model builds on previous theoretical works . The role of the window INa in achieving low firing rates was considered in [13 , 18] . The fact that slowly inactivated K+ currents can induce delay to action potential firing and bursting was also described in [13 , 39] . The stuttering pattern displayed by our model is an example of “elliptic bursting” [18 , 20 , 40] ( also named “SubHopf/Fold cycle” [19] ) . In addition , the present paper relates the appearance of subthreshold oscillations and the dependence of the delay duration on the levels of noise and applied current with the bifurcation structure of the fast subsystem . Marder and colleagues proposed that for a specific pattern of activity , one can find parameter subspaces within which the model displays qualitatively , and even quantitatively , similar behavior [41 , 42] . They also proposed that neuronal function can be stabilized by homeostatic mechanisms ensuring that the neuron always remains in those subspaces [43] . Clearly , the bifurcation point cannot exist in such subspaces . There are , however , directions in parameter space along which the qualitative behavior of the neuron varies via bifurcations of the dynamics , as shown in our paper . These bifurcations can underlie the variability observed in the electrophysiological properties of FS cells . We predict that for FS cells that exhibit delay before firing , the delay duration , tdelay , decreases with the amplitude of the current step , Iapp , and disappears at a non-zero value as Iapp is elevated ( Figure 4A ) . When the neuron displays stuttering in response to just suprathreshold oscillations , we predict that elevating Iapp will first increase the average number of spikes during the stuttering state ( Figure 2D ) , and then will transform the cell into a tonic firing cell ( Figure 2A ) . A depolarizing pre-pulse shortens tdelay and even eliminates it if the pre-pulse is large enough , but does not affect the stuttering behavior ( Figure S1B ) . Similarly , a hyperpolarizing pre-pulse increases tdelay . These predictions can be tested by current-clamp experiments . Our theoretical work and the experimental results presented here suggest the existence in FS cells of two types of responses to step current pulses . They differ in the minimal firing frequencies , the properties of the membrane potential fluctuations during the delay period , and the sensitivity of the delay duration to noise . More specifically , we predict that the minimal firing rate , the sensitivity of tdelay to noise , and the presence of subthreshold oscillations of the membrane potential during the delay period are negatively correlated to the strength of the Na+ window current . Furthermore , we predict that FS neurons that can fire at low firing rates cannot stutter , and that increasing gd artificially via dynamic clamp may convert a tonic-delay response into stuttering . These predictions can be tested in a detailed population study of electrophysiological properties of FS neurons . The modeling results presented in this paper can be applied to understand the effect of some neuromodulators on the firing patterns of FS cells . Dopamine attenuates the d-type K+ current Id in a subgroup of FS neurons [44] . Consistent with the results of our modeling study , dopamine also transforms the firing pattern of an FS cell from a tonic-delay type to the tonic-no delay type [44] ( see also [45] ) . Na+ currents are affected by metabotropic glutamate receptor subtype 1 ( mGluR1 ) . As shown by [46] , it shifts θh to more hyperpolarized potentials in pyramidal neurons , decreasing their Na+ window current . It also facilitates the persistent Na+ current INaP by shifting its activation curve leftward [46] . Similarly , serotonin makes θh more negative in pyramidal cells , and also reduces the maximal conductance of INaP [47] . If one assumes that these modulators have similar effects on Na+ currents in FS cells , our results suggest that they may modify qualitatively the firing patterns of these neurons . Our model of FS cells is based on that of [3 , 28] , with several modifications based on voltage clamp data . The current balance equation is where V is the membrane potential of the neuron , C = 1μF/cm2 is the membrane capacitance , and the parameters of the leak current are gL = 0 . 25 mS/cm2 and VL = −70 mV . The external current injected into the neuron is denoted by Iapp . The Na+ current INa is given by: where the gating variables , h and m , follow: The parameters are: gNa = 112 . 5 mS/cm2 , VNa = 50 mV , σm = 11 . 5 mV , θh = −58 . 3 mV , σh = −6 . 7 mV , θth = −60 mV , σth = −12 mV [14] . In this work , we study the effect of the strength of the Na+ window current , controlled by the parameter θm , on the dynamics of the neuron . The delayed rectifier K+ current IKdr is of the Kv3 . 1–Kv3 . 2 type . It is responsible for the brief duration of the spike , about 0 . 5 ms [2 , 48] , and for the high firing frequency [3 , 49] . It is given by: with: All the parameters of the delayed rectifier current are fixed: gKdr = 225 mS/cm2 , VK = −90 mV , θn = −12 . 4 mV , σn = 6 . 8 mV , θtn = −27 mV , σth = −15 mV [50] . The K+ current Id incorporated in the model [10 , 11] has fast activation and slow inactivation . It is defined by: Throughout the paper , all the parameters of the d-current but gd are fixed: θa = −50 mV , σa = 20 mV , τa = 2 ms , θb = −70 mV , σb = −6 mV , τb = 150 ms [51 , 52] . The parameter gd is varied to study the effect of the strength of this current . Finally , to study the effect of noise in the external input on the firing pattern of the neuron , we add an additional external input , Inoise , of the form: where ξ ( t ) is a Gaussian white noise with an average 0 and a unit variance , and D has the units of μA2 × ms/cm4 . Numerical methods . Simulations were performed using the fourth-order Runge-Kutta method with a time step of 0 . 01 ms implemented as a C program or within the software package XPPAUT [53] , which was used also for computing bifurcation diagrams . Delay . The delay duration tdelay is defined to be the time from the onset of current injection , or , if the neuron fires transient 1–3 spikes , from the last transient spike to the first spike of the sustained firing . We define that the neuron shows a delay if tdelay is at least twice as large as the inter-spike interval during steady-state spiking tISI , or if it is larger than both 100 ms and 1 . 2 tISI . Fourier spectrum . Discrete Fourier transforms of subthreshold oscillations were calculated numerically over a time window of TFT ending TBS = 5 ms before the first spike of the steady-state firing . The absolute values of the Fourier components were averaged over nR repetitions of the same stimulus . Parameters for Figure 9 are: TFT = 120 ms , nR = 20 . Parameters for Figure 12 are: TFT = 90 ms , nR = 5 ( A ) , TFT = 120 ms , nR = 13 ( B ) , TFT = 120 ms , nR = 11 ( C ) . The bifurcations of the fast subsystem when b varies depend on the shape of the function VFP ( b ) , where VFP is the value of the membrane potential of the neuron at the fixed point of the dynamics for fixed b . Equivalently , one can relate the bifurcations to the shape of the curve b = b ( VFP ) , in the b-VFP plane ( the “b-VFP curve” ) , which is defined by ( see Equations 5 , 6 , 11 , and 15 ) : where and The denominator in Equation 21 is positive and increases with VFP . The numerator therefore is positive in the relevant range of VFP for which b > 0 . The functions IL ( VFP ) and IKdr ( VFP ) are increasing with VFP . Only the function INa ( VFP ) may decrease with VFP . Therefore , for a small Na+ window current , b decreases monotonously with VFP . This happens for instance for θm = −24 mV ( Figure 5A ) . In contrast , if the overlap between the activation and the inactivation curves of the Na+ current is sufficiently large [13] , the term −INa ( VFP ) in Equation 21 can contribute substantially to make the function b ( VFP ) be non-monotonous . This happens for instance for θm = −28 mV ( Figure 5D ) . We estimate tdelay , the duration of the delay to firing of action potentials , using the “fast–slow method , ” and derive the dependence of tdelay on Iapp near the current threshold Ith in the noiseless case . During the delay period , the fast subsystem is at its fixed point , and b decreases slowly . We use this fact to compute the scaling of the divergence of tdelay with Iapp−Ith for Iapp ≳ Ith for the two bifurcation scenarios . Hopf bifurcation of the fast subsystem . The evolution of b is given by Equation 17 , and it becomes very slow when b approaches b∞ ( V ) . For large τb , V follows the curve VFP ( b ) during the delay period ( Figure 5A ) . We denote by bx̃ the solution of the equation , and define Vx̃ ≡VFP ( bx̃ ) Note that the fixed point of the fast subsystem for b = bx̃ is unstable if the neuron fires following the delay . When Iapp is near Ith , VFP ( b ) is approximated by Vx̃ [13] , and therefore Equation 17 for the evolution of b is approximated by Using Equation 22 for computing tdelay is justified because tdelay is determined mainly by the slow dynamics of b when it is near bx̃ . Before the current step is applied at time t = 0 , the system is at rest with b = brest . The subsystem converges immediately to its fixed point on the slow time scale , and the solution to Equation 22 is According to Equation 3 . 13a from [20] , tdelay is determined by the equation Near a Hopf bifurcation , where bHopf is the value of b at the Hopf bifurcation and α is a constant . Substituting Equation 23 in Equation 24 and using Equation 25 , we obtain Near the threshold current Ith , tdelay >> τb , and Equation 26 becomes Generically , when Iapp−Ith is small , bx̃ depends only weakly on Iapp and bHopf − bx̃ depends linearly on Iapp−Ith . Therefore , tdelay scales as ( Iapp−Ith ) −1 . Saddle-node bifurcation of the fast subsystem . The dynamics are very slow near the SN bifurcation , occurring at ( bSN , VSN ) . Neglecting the changes in VFP during the evolution , Equation 23 becomes where bISN = b∞ ( VSN ) . Namely , Generically , when Iapp−Ith is small , bSN−bISN depends linearly on Iapp−Ith . Therefore , tdelay scales as −log ( Iapp−Ith ) . We calculate the dependence of tdelay on the noise variance , D , for weak noise and weak window INa . In this case , the delay ends because the fast subsystem is destabilized via a Hopf bifurcation . According to Theorem 4 . 1 in [40] , assuming that the variance of the noise , D , is neither too large nor too small , tdelay is determined by the equation where Ax̃ and Bx̃ are constants . As above , tdelay is determined mainly by the slow evolution near bx̃ , and therefore one can use the approximation VFP ( b ) ≈ Vx̃ . Substituting Equations 23 and 25 in Equation 30 , we obtain Solving this equation for tdelay , in the limit tdelay >> τb , we obtain where and . We consider the case that , for a certain value of gd , gd = gd1 , and a step current with an amplitude Iapp , there is a solution to the equation F ( b ) = b . We prove here that for a large enough gd , a solution to this equation does not exist for this value of Iapp . The function F ( b ) is defined only when the limit cycle exists , namely only for b ≤ bSNP . Since the current Id depends on gd and b only through the product gdb ( Equation 15 ) , bSNP for any other value of gd is This means that bSNP ( gd ) is very small for large gd . We continue by noticing that: ( 1 ) F ( b ) is the time-average of the function b∞ ( V ( t ) ) over LC ( b ) ( Equation 2 ) ; ( 2 ) b∞ ( V ) is a positive , decreasing function of V ( Equation 19 ) ; and ( 3 ) V ≤ VNa . Therefore , F ( b ) ≥ F ( b∞ ( VNa ) ) . From the fact that bSNP ( gd ) decreases with gd ( Equation 33 ) , one finds that , for large enough gd , F ( b∞ ( VNa ) ) > bSNP ( gd ) . Since F ( b ) is defined only for b ≤ bSNP ( gd ) , we obtain that F ( b ) > b , and there is no solution to the equation F ( b ) = b if gd is large enough . Therefore , the full system cannot exhibit a tonic firing state . If the rest state of the neuron is unstable , the neuron stutters . In practice , F ( b ) is much larger than F ( b∞ ( VNa ) ) because the membrane potential spends a large fraction of its period in subthreshold values , and therefore gd should not be extremely large to prevent a solution of the equation F ( b ) = b . Mice ( CD1 , 21–28 d old ) were deeply anaesthetized with pentobarbital , decapitated , and their brains quickly removed into cold ( 5 °C ) physiological solution . Coronal cortical slices ( 400 μm thick ) were cut with a vibratome ( Campden Instruments , http://www . campdeninstruments . com ) and then transferred to a holding chamber where they were kept at room temperature for at least 1 h before recording , continuously bubbled with 95% O2 , 5% CO2 . Recording was done in a chamber mounted on an upright microscope equipped with IR/DIC optics ( Nikon physiostation EC-600 ) , where they were held at 32–34 °C and constantly perfused . The normal bathing solution contained ( in mM ) : 124 NaCl , 3 . 5 KCl , 2 MgSO4 , 1 . 25 NaHPO4 , 2 CaCl2 , 26 NaHCO3 and 10 dextrose , and was saturated with 95% O2 , 5% CO2 ( pH 7 . 4 ) . Whole-cell recordings were made from neurons in the barrel field . Patch recording micropipettes ( 4–6 MΩ ) were filled with a solution containing ( in mM ) 125 K gluconate , 5 NaCl , 2 MgCl2 , 10 EGTA , 10 HEPES , and 2 Na2-ATP , pH 7 . 2 , 280 mOsm ) . Voltages were recorded with a patch clamp amplifier ( AxoPatch 2B , Axon Instruments , http://www . axon . com ) , and digitally sampled at 10 kHz . Data acquisition and analysis were performed with Labview ( National Instruments , http://www . ni . com ) . Series resistance was typically <15 MΩ . During all recordings , 50 μM DL-2-amino-5-phosphopentanoic acid ( AP5 , Sigma , http://wwwsigmaaldrich . com ) and 6 , 7-dinitroquinoxaline-2 , 3-dione ( DNQX; 20 μM , Sigma ) were present in the bath to block excitatory transmission . Identification of FS neurons . Non-pyramidal neurons were targeted by their soma and proximal dendrites image under the IR/DIC microscope . Among those , FS neurons were identified according to their electrophysiological properties . A neuron was classified as an FS neuron if: 1 ) it fired brief spikes with fast , deep , monophasic AHPs [1]; and 2 ) the ratio of the spike's rising phase dV/dt to falling phase dV/dt was smaller than 2 . Previous studies revealed that the morphological correlate of FS neurons can be either “basket” [7] or “neurogliaform” [24 , 54 , 55] . Most FS neurons express parvalbumin [56] , but others express somatostatin [57] . Thus , this type of interneuron may be heterogeneous in terms of its morphology or chemical content , but we did not use these criteria for our classification .	About 25% of the neurons in the mammalian neocortex are inhibitory , namely reduce the activity of neurons they contact . These inhibitory neurons exhibit diversity of morphological , chemical , and biophysical properties , and their classification has recently been the focus of much debate . Even neurons belonging to a single class of “fast-spiking” ( FS ) display a large variety of firing patterns in response to standard square current pulses . Previous works proposed that this class is in fact a discrete set of neuronal subtypes with biophysical properties differing in a discontinuous way . In this work , we propose an alternative theory , according to which the biophysical properties of FS neurons are continuously distributed , but distinct firing patterns emerge due to highly nonlinear dynamics of these neurons . We ascertain this theory by exploring with mathematical techniques a biophysically based model of FS neurons . We demonstrate that variable firing responses of cortical FS neurons can be accounted for if one assumes heterogeneity in the strength of some of the ionic conductances underlying neuronal activity . Our theory predicts the existence of two main firing patterns of FS neurons . This prediction is verified by direct recordings in cortical slices .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "neuroscience", "cortical", "fast-spiking", "neurons" ]	2007	Mechanisms of Firing Patterns in Fast-Spiking Cortical Interneurons
For many genes , proper gene expression requires coordinated and dynamic interactions between multiple regulatory elements , each of which can either promote or silence transcription . In Drosophila , the complexity of the regulatory landscape is further complicated by the tight physical pairing of homologous chromosomes , which can permit regulatory elements to interact in trans , a phenomenon known as transvection . To better understand how gene expression can be programmed through cis- and trans-regulatory interactions , we analyzed transvection effects for a collection of alleles of the eyes absent ( eya ) gene . We find that trans-activation of a promoter by the eya eye-specific enhancers is broadly supported in many allelic backgrounds , and that the availability of an enhancer to act in trans can be predicted based on the molecular lesion of an eya allele . Furthermore , by manipulating promoter availability in cis and in trans , we demonstrate that the eye-specific enhancers of eya show plasticity in their promoter preference between two different transcriptional start sites , which depends on promoter competition between the two potential targets . Finally , we show that certain alleles of eya demonstrate pairing-sensitive silencing resulting from trans-interactions between Polycomb Response Elements ( PREs ) , and genetic and genomic data support a general role for PcG proteins in mediating transcriptional silencing at eya . Overall , our data highlight how eya gene regulation relies upon a complex but plastic interplay between multiple enhancers , promoters , and PREs . The eukaryotic genome is rich in regulatory elements whose combined inputs lead to proper execution of programmed patterns of gene expression . Regulatory elements that promote gene expression include promoters , where RNA polymerases begin transcription of genes , and enhancers , which bind to transcriptional activator proteins and are thought to physically interact with promoters via looping , thereby recruiting or activating RNA polymerases [1 , 2] . Conversely , other DNA elements play roles in preventing transcription locally , including Polycomb Response Elements ( PREs ) , which bind to complexes of proteins known as the Polycomb Group ( PcG ) and can ultimately create a silenced chromatin domain via the histone mark H3K27me3 [3 , 4] . While our ability to identify these types of regulatory elements has grown with increasing accuracy via the refinement of sophisticated genomic approaches , our understanding of how specific elements interact with one another across diverse tissues remains incomplete . In Drosophila , specificity of interactions between regulatory sequences is further complicated by the phenomenon of somatic homolog pairing , where homologous chromosomes are held in close proximity in virtually all somatic cells of the organism [5] . A growing body of data supports that somatic homolog pairing permits regulatory elements on one homolog to interact with those on the homologous chromosome , a phenomenon coined transvection by its discoverer , Ed Lewis [6] . The term transvection encompasses several types of pairing-dependent genetic interactions , including those that positively impact gene expression , as is the case when an enhancer on one chromosome acts in trans to activate transcription from a promoter on the homologous chromosome , or that negatively impact gene expression , as observed in some cases when PREs interact in trans , which is thought to increase the efficacy of PcG proteins bound to the PRE in silencing transcription ( Fig 1 ) [5 , 7] . The study of transvection typically relies on specific mutant backgrounds and/or transgenic organisms with defined constructs placed at equivalent positions on homologous chromosomes . For example , enhancer action in trans has been studied extensively for the yellow gene of Drosophila , which is required for pigmentation of the adult cuticle [8 , 9] . The yellow gene has a simple structure , with a single promoter and transcription start site ( TSS ) and several well-defined tissue-specific enhancers , and is rich in classical alleles that impact gene expression . Notably , enhancer action in trans at yellow appears tightly regulated; intragenic complementation via transvection is observed only between two types of alleles , those in which enhancers are deleted or otherwise prevented from interacting with the yellow promoter in cis ( “Class A” alleles ) , and those in which the promoter region is compromised by deletion , mutation , or nearby transposon insertion ( “Class B” alleles ) [9] ( Fig 1A ) . In contrast , alleles of yellow that have an intact promoter but carry mutations in the coding region of the gene , known as “Class C” alleles , fail to complement Class A alleles despite carrying functional enhancers that would otherwise be available to act in trans ( Fig 1A ) . The failure of the yellow enhancers of Class C alleles to act in trans has been interpreted to be due to their preference for a promoter in cis; it is only when the cis-promoter is somehow compromised , as in Class B alleles , that the yellow enhancers are released to act in trans , suggesting a hierarchical regulation of potential promoter targets for the yellow enhancers . Several other examples of enhancer action in trans show evidence that enhancers prefer to act on a promoter in cis relative to a promoter in trans [10–14] . However , in these cases , activation of a trans-promoter is attenuated , but not eliminated , in the presence of a promoter in cis . A simple interpretation is that cis-preference is a global phenomenon that is relevant to many enhancers in Drosophila , and that yellow represents an extreme case where cis-preference is strong enough to reduce trans-activation to undetectable levels . However , aside from analyses at yellow , there has yet to be further characterization of transvection in Drosophila that uses a diverse collection of alleles analogous to the Class A , Class B , and Class C alleles of yellow . The eyes absent ( eya ) gene encodes an evolutionarily conserved transcriptional co-activator with protein phosphatase activity that is a component of the Retinal Determination Network ( RDN ) of transcriptional regulators required for normal eye development in Drosophila [15 , 16] . Eye-specific loss of eya function can cause an “eyeless” phenotype in adult flies , whereas ectopic expression of eya can lead to development of eye tissue elsewhere in the body [16 , 17] . Eya functions in part via the formation of a complex with the DNA-binding Sine oculis ( So ) protein , thereby acting as a bipartite transcription factor that regulates RDN gene expression and coordinates other downstream processes of eye differentiation [18] . The eya gene structure contains two major transcriptional start sites , with the first exon of the eya-B ( also known as Type I ) transcript encoded roughly 10kb upstream of that of eya-A ( Type II ) ( Fig 2A ) . Each of the alternate first exons splices to common second through fifth exons that together encode the majority of the protein . Presumed null mutations in the coding region result in embryonic lethality due to the requirement of eya activity in diverse tissues at early stages of development , likely via transcription from the eya-A promoter [19–23] . In contrast , retrotransposon insertions into Exon 1B result in homozygous viable flies entirely lacking compound eyes and ocelli , suggesting that the transcript initiated from the eya-B promoter is required for development of eye structures [24] . Upstream of the eya-B promoter is approximately 9 kb of non-coding sequence that carries several enhancers for distinct eye tissues [24–26] , and a DNA fragment carrying this 9 kb fragment can fully recapitulate the wild-type eya expression pattern in third instar larval eye discs [25] . Transvection has been previously demonstrated for alleles of eya that are analogous to the Class A and Class B alleles of yellow [27] . Specifically , trans-heterozygous flies carrying eya2 , an enhancer deletion , and eya4 , an insertion in the 5’ UTR of Exon 1B , result in pairing-dependent rescue of eye development , likely via enhancer action in trans . Furthermore , several other alleles of eya are structurally similar to Class C alleles of yellow , with missense and nonsense mutations in the eya coding region , and several show some degree of complementation with eya Class A alleles that could be consistent with transvection [19 , 20] . Here we demonstrate that Class A alleles of eya complement all Class B and Class C alleles tested , demonstrating that enhancer action in trans is less strictly regulated at eya relative to yellow . However , complementation by Class B is consistently stronger than that by Class C across several genetic backgrounds , suggesting that the availability of enhancers to act in trans differs depending on the molecular lesion in eya . Furthermore , by manipulating promoter availability in cis and in trans , we show that the eye-specific enhancers of eya show preference for the eya-B promoter relative to the eya-A promoter , and that the preference depends on competition between the two promoter targets . Finally , genetic interactions between eya Class A alleles demonstrate pairing-sensitive silencing , and genetic and genomic data support a role for PcG proteins in mediating eya transcriptional silencing . Prior analyses have demonstrated that , whereas expression during embryogenesis is specific to the eya-A promoter , both the eya-A and eya-B promoters are active in the developing eye disc [20 , 26] . Furthermore , qRT-PCR analysis supports that the eye-specific enhancers upstream of the eya-B promoter activate transcription of both transcript types [26] . To better understand transcript-specific expression in the developing eye , we performed in situ hybridization on wild type third instar larval eye discs using probes specific to the first exons of either the eya-A or eya-B transcript . Analysis of the eya-B isoform showed robust expression in progenitor cells anterior to the morphogenetic furrow and in differentiating cells immediately posterior the furrow , with lower levels of expression observed in more mature ommatidial clusters toward the posterior of the disc ( Fig 5A ) . Expression is also seen in the developing ocelli , consistent with a requirement for eya in ocelli development [16] . The pattern of staining for the eya-A transcript appears similar to that of the eya-B transcript , with highest expression seen immediately anterior and posterior to the morphogenetic furrow ( Fig 5B ) . However , the signal for the eya-A transcript is barely detectable above background fluorescence , suggesting that the eye-specific enhancers of eya act predominantly on the eya-B promoter in the developing eye , and only act weakly on the eya-A promoter . The alleles eya3 and eya4 are caused by retrotransposon insertions in the first exon of eya-B ( Exon 1B ) , and do not directly impact the eya-A transcript [24] . Furthermore , in situ hybridization demonstrates that the retroelement in eya4 is transcribed in an eya-like pattern , reflecting that the eye enhancers are functional in this allele [24] . Two plausible models could account for the lack of eye development in eya3 and eya4 flies: first , assuming that the eya-B transcript is rendered non-functional by the retroelement insertions , production of functional mRNA solely from the eya-A promoter could be insufficient to generate eye tissue . Alternatively , it could be that retrotransposon insertion into Exon 1B decreases or prevents communication between eye-specific enhancers and the eya-A promoter , resulting in loss of eya-A transcription and leaving only the non-functional eya-B transcript . To assess these models and thereby gain a better understanding of transcript usage in the developing eye , we employed isoform-specific in situ hybridization in discs that display enhancer action in trans . In eya4/eya2 discs , we observe robust expression of eya-B , likely reflecting strong activation of the retrotransposon-carrying Exon 1B in cis to the functional enhancers of the eya4 chromosome in addition to trans-activation of the functional Exon 1B on the eya2 chromosome ( Fig 5C ) . However , levels of the eya-A transcript appear strongly decreased in eya4/eya2 discs ( no detectable signal in 6/7 discs scored ) ( Fig 5D ) , suggesting that the retrotransposon insertion into Exon 1B significantly decreases transcription from the eya-A promoter , and that the eya-B promoter is the preferred target in cis and in trans to the eye-specific enhancers . In contrast to the small enhancer deletion of the eya2 allele , the eya1 deletion removes enhancer sequences and the eya-B promoter [26]; thus , in eya4/eya1 discs , functional eya-B transcript cannot be generated from either chromosome . In contrast to our observations in eya4/eya2 discs , we easily detect RNA signal for the eya-A transcript in eya4/eya1 discs ( 7/7 discs scored ) , consistent with a model wherein the eya-A promoter is trans-activated by the functional enhancers of the eya4 chromosome in this background ( Fig 5E and 5F ) . To further support a difference in promoter usage in eya4/eya2 vs . eya4/eya1 discs , we employed isoform-specific quantitative RT-PCR on eye-antennal discs from these genotypes and compared the levels of eya-A transcripts relative to those of eya-B . Notably , eya-A transcript levels in eya4/eya2 discs drop to 56% ( 95% CI 52 . 6%-59 . 9% , n = 3 biological replicates ) when compared to transcripts from eya4/eya1 discs , further supporting decreased expression of eya-A relative to eya-B in eya4/eya2 discs . In sum , our data demonstrate that the eye-specific eya enhancers show a preference for the eya-B promoter , and support that the loss of eye development in eya3 and eya4 flies involves a reduction in activation of the eya-A transcript in addition to the insertional disruption of the eya-B transcript . Furthermore , our data suggest that , in the absence of a functional eya-B promoter in trans , the enhancers can switch their specificity to trans-activate the eya-A promoter in order to produce eye tissue . To further assess the requirement for the eya-B transcript in eye development , we used CRISPR-Cas9 to completely remove Exon 1B and its associated core promoter from the genome ( Fig 6A ) . Synthetic guide RNAs designed to flank Exon 1B and under the control of the U6 promoter were injected into embryos carrying a source of Cas9 , and the progeny of the resulting flies were screened via PCR for the expected deletion , resulting in four independent mutants lacking exon 1B ( see Materials and Methods ) . All four mutant alleles are viable and fertile as homozygotes , consistent with the proposed eye-specific role for Exon 1B . Surprisingly , all mutants lacking Exon 1B develop near-wild type eyes as either homozygotes or in combination with Df ( 2L ) eya ( Fig 6B and 6C ) . To address eye-specific promoter usage in these mutants , we performed isoform-specific in situ hybridization on third instar larval eye-antennal discs that were homozygous for the Exon 1B deletion . We observed no signal above background for the eya-B probe , confirming that the induced deletion prevents transcription of these sequences ( Fig 6D and 6E ) . Remarkably , staining for the eya-A transcript shows robust signal in a pattern similar to that previously observed for eya-B , demonstrating that the loss of the eya-B promoter results in elevated activation of eya-A transcription ( Fig 6F and 6G ) . To further support this observation , we used quantitative RT-PCR to measure levels of eya-A and eya-B expression in wild type and exon 1B-deleted mutant eye-antennal discs ( Fig 6H ) . Consistent with our in situ data , we observed a 3-fold increase in expression of eya-A , and a complete loss of eya-B , in the mutant discs . Thus , our data support a model wherein wild type eye development relies primarily on activation of the eya-B promoter; in the absence of this promoter and its associated first exon , enhancers shift their specificity to the more distal promoter associated with the eya-A transcript , with near complete compensatory expression to support eye development . The Class A allele eyacs is homozygous viable and hypomorphic , with eyacs/eyacs flies showing a reduced adult eye phenotype with variable expressivity ( Fig 7A and 7B ) . Sequence analysis showed that eyacs carries a 115 bp deletion from -806 to -691 relative to the TSS of the B transcript , which is nested within the enhancer deleted by the eya2 allele ( -896 to -577 ) ( Fig 2A ) . To assess whether eyacs can support enhancer action in trans , we created flies with eyacs on one homolog and various Class B or Class C alleles on the other homolog as we had previously done with the Class A alleles eya1 and eya2 . Flies carrying trans-heterozygous combinations of eyacs and Class B or Class C alleles show greater numbers of ommatidia than eyacs homozygotes , comfirming increased expression of eya ( Fig 7A and 7B ) . As observed for other Class A alleles , the strength of transvection is higher when eyacs is in trans to Class B alleles relative to Class C alleles , and a transvection-disrupting rearrangement of the chromosome carrying eya4 [27] shows reduced complementation relative to a structurally wild type chromosome carrying eya4 , supporting that the observed complementation between eyacs and Class B and Class C alleles are pairing-dependent ( Fig 7A and 7B ) . Thus , the hypomorphic Class A allele eyacs can participate in enhancer action in trans . In establishing eyacs as a Class A allele , we were surprised to find that the eye phenotypes of flies carrying eyacs trans-heterozygous with the other Class A alleles are more severe than those of eyacs homozygotes , with eyacs/eya2 having a more severe phenotype than eyacs/eya1 ( Fig 8A–8E ) . Furthermore , the eye phenotype of flies carrying eyacs trans-heterozygous with Df ( 2L ) eya , a large deficiency spanning the entire eya locus , does not show an increased severity relative to eyacs homozygotes , but instead shows a more moderate phenotype ( Fig 8A and 8E ) . Thus , eyacs shows repressive trans-interactions with the small deletions carried by other Class A alleles , but not with a large deletion . To determine whether repressive trans-interactions involving eyacs are pairing-dependent , we created trans-heterozygotes between eyacs and the rearranged eya2 allele ETD2 . 2 ( Fig 8D and 8E ) . Notably , the disruption of pairing between eya alleles caused by the rearrangement carried by ETD2 . 2 restored partial eye development in these flies , indicating that the repression of eya by other Class A alleles is pairing-sensitive . Previous analyses of pairing-sensitive silencing have revealed a central role for PcG genes [7] . Specifically , known cases of pairing-sensitive silencing are caused by pairing-dependent interactions between PREs on homologous chromosomes , which is thought to augment the recruitment and/or silencing capacity of PcG complexes relative to those at unpaired PREs . To address whether PcG genes may function in pairing-sensitive silencing at the eya locus , we first assessed ChIP-seq binding profiles of key PcG proteins and the histone H3K27me3 mark associated with Polycomb repressive domains ( Fig 8F ) . Indeed , the compartmental domain occupied by eya is rich in H3K27me3 in third instar larval disc tissue and in cultured Drosophila cells [31–33] , and carries several putative PREs indicated by peaks of PcG proteins Ph , Psc , Pc , and E ( z ) [33] . Two putative PREs are located upstream of the eye-specific eya-B promoter and partially overlap the known eye-specific enhancers [26] ( Fig 8F and 8G ) . Notably , the eya2 deletion is predicted to leave all putative PREs intact , whereas the eya1 deletion removes the two putative PREs that are upstream of the eya-B promoter ( Fig 8G ) . To provide genetic evidence for a role for PcG proteins in pairing-dependent silencing at eya , we used two approaches to assess eye development in backgrounds with reduced expression of two key PcG components , E ( z ) and Pc . First , we scored the number of ommatidia in developed adult eyes ( excluding eyeless flies ) for eyacs/eya1 trans-heterozygous flies with and without reduced PcG gene dosage . In this assay , mutations in both E ( z ) and Pc act as dominant suppressors of pairing-sensitive silencing ( Fig 8H–8J ) , resulting in greater numbers of ommatidia relative to eyacs/eya1 flies with wild type PcG dosage . Secondly , we compared percentages of total adult flies that were completely eyeless; in an eyacs/eya2 background , nearly all ( 98 . 8 ± 1 . 4% ) adults completely lack eyes , but loss of a single functional copy of E ( z ) suppressed the number of eyeless flies to 75 . 9 ± 2 . 5% ( p = 0 . 008 , Mann-Whitney U test ) ( Fig 8H ) . A similar effect is observed in an eyacs/eya1 background ( p = 0 . 02 ) , but not when a functional copy of Pc was removed from either eyacs/eya1 or eyacs/eya2 flies ( Fig 8H ) , perhaps indicating a greater sensitivity for E ( z ) relative to Pc function at an early stage of eye specification . In sum , our genetic data and genomic analysis support a role for PcG genes in pairing-sensitive silencing at eya . Based on our data , we favor a model wherein the eya eye-specific enhancers play dual roles , activating transcription of the eya-B and eya-A promoters and opposing the silencing activity of PREs in the domain occupied by eya ( Fig 8L ) . In flies carrying the eyacs deletion , one or both of these activities is partially compromised , resulting in reduced eye growth; when eyacs is placed in trans to the eya2 allele , enhancer activity is further suppressed while pairing between homologous PREs strengthens their silencing capacity , resulting in complete loss of eye tissue . However , when eyacs is placed in trans with eya1 or larger deletions , or with a rearranged eya2 allele , pairing of some homologous PREs is lost , resulting in re-establishment of partial eye development . In sum , our data indicate an important role for PcG genes in regulating a critical eye determining gene . The presence of multiple eya promoters likely reflects an ancient promoter duplication event , which , like gene duplications , can lead to varying degrees of functional redundancy or sub-functionalization between the alternate TSS [42] . Promoter duplication appears to be widespread in the Drosophila genome; genome-wide mapping of TSS shows that approximately 27% of mapped genes can initiate transcription via two or more promoters , with an average number of 1 . 4 promoters per gene across all mapped TSS [43] . Comparison to genomes of other Drosophila species suggests that a promoter duplication at eya was a relatively recent event , with evidence of two TSS in the D . melanogaster , D . simulans , and D . ananassae genomes , but not in those of D . pseudoobscura or D . virilis [44 , 45] . Prior analyses at eya suggest that the eya-A and eya-B promoters have undergone some degree of sub-functionalization in D . melanogaster , with the eya-B promoter being active primarily within the developing eye disc and the eya-A promoter being more broadly expressed across multiple tissues [20] . Our data support the prior finding that the eya-B promoter is the preferred target of the eye-specific eya enhancers , which are primarily located just upstream of the eya-B promoter , but roughly 10 kb from the eya-A promoter [26] . These observations suggest a simple model wherein specificity can be dictated by relative position; for some enhancers , activity may be highest on nearby promoters , with less activity on more distal promoters . However , our data support that this model is largely driven by promoter competition at the eya locus such that , in the absence of the eya-B promoter , the eya-A promoter becomes a “preferred” target and is highly active . Interestingly , ChIP-seq analysis using antibodies to the insulator protein su ( Hw ) in embryos suggests the presence of an insulator element in the eya-B first intron that would be predicted to disrupt communication between the eye-specific enhancers and the eya-A promoter [46] . Based on our observations , it is unlikely that this candidate insulator is active in the developing eye disc . Our data regarding promoter competition and enhancer-promoter proximity is consistent with prior observations where a nearby promoter is preferred to one that is more distal [47–50] , and may therefore be generalizable to many enhancers . As a potential caveat in interpreting our data , two other DNA fragments that map close to the eya-A promoter support some degree of transgene expression in the late developing eye disc [26] . These candidate enhancers are not themselves sufficient to rescue eye phenotypes , and are therefore of unknown functional relevance in vivo , but we cannot exclude the possibility that their activity changes in some way upon deletion of the eya-B promoter such that they play a role in the upregulation of eya-A transcription observed in this background . Finally , we note that the current genome annotation supports evidence for a third eya TSS , defining an eya-C transcript that initiates further downstream from the eya-A TSS and is predicted to produce a truncated protein product [45 , 51–52] . The biological relevance of this potential promoter and its relationship to the eya-A and eya-B transcripts is as yet unclear . Our observation of pairing sensitive silencing of eya suggests a direct role for PcG genes in regulating eya expression . In support of this hypothesis , genomic data shows that eya is embedded in a domain of H3K27me3 in cultured cells , embryos , and third instar disc tissues , and distinct peaks of PcG proteins that are characteristic of PREs are found throughout the eya locus [31–33 , 53 , 54] . Furthermore , reduction in dosage of key PcG proteins , E ( z ) and Pc , suppresses pairing sensitive silencing of eya , providing genetic evidence for a role for PcG proteins in directly regulating eya expression . Recently , Erceg et al [55] characterized hundreds of sequences with overlapping PcG binding and enhancer activity , and showed that these fragments can act as enhancers in some cell types and as silencing PREs in others . Notably , one of the candidate PREs upstream of eya overlaps a previously characterized enhancer [26] , and the mutations that uncover pairing-sensitive silencing affect sequences in this region . Our data support a model wherein the PRE activity of this region is active in cells outside of the developing eye , silencing eya , whereas the PRE silencing activity is overcome in primordial eye cells in order for eye development to proceed . According to this model , H3K27 methylation would be reduced or suppressed by the activity of the eya enhancers in primordial eye cells ( Fig 8L ) , although we are unable to observe this directly using existing ChIP-seq data derived from mixed larval tissues . Given that several candidate PREs are found across the locus , it is as yet unclear how these different sequences may cooperate and/or interact to determine the transcriptional state of eya in a given tissue . Interestingly , several other genes of the Retinal Determination Network ( RDN ) are also characterized by domains of H3K27me3 and localized regions of PcG binding in multiple cell types [31–33 , 53 , 54] . Furthermore , in addition to eya , RDN genes toy and dac have been identified as having overlapping PRE and enhancer sequences , and a neuronal enhancer from the ey gene was shown to have enhancer activity in some tissues and PRE activity in others [55] , suggesting that direct regulation by PcG proteins could be a common feature of RDN genes . According to this model , PcG proteins would maintain RDN genes in an inactive state in non-eye tissues , whereas activation of RDN gene transcription in the developing eye would rely on the coordinated removal of repressive chromatin marks and simultaneous activation of transcription . Consistent with this hypothesis , ChIP-seq analysis shows that binding of the PcG proteins Pho and Ph at the TSS of the RDN genes so and toy is higher in haltere tissues ( where the RDN genes are inactive ) relative to binding in eye tissue , consistent with reduced binding of PcG proteins at PREs of RDN genes in cells with active expression [56] . However , investigations of roles for PcG proteins in the developing eye are complicated by the widespread pleiotropic effects on gene expression caused by PcG mutations combined with the deeply intertwined regulatory network that determines eye cell fates . For example , clonal loss of E ( z ) and Pc in cells anterior to the morphogenetic furrow can lead to reduced expression of eya and dachshund ( dac ) , but this is likely due to misexpression of teashirt , which can act as a negative regulator of eya and dac [57 , 58] . Similarly , seminal work by Zhu et al . [59] demonstrated a role for PcG proteins in maintaining eye cell fates in the developing eye via the repression of genes that would signal an alternative wing tissue fate . Furthermore , biochemical studies show that Eya protein is a binding partner for Combgap ( Cg ) , a sequence-specific DNA-binding protein that can recruit PcG complexes to PREs , although genetic analyses show that Cg may act in opposition to other PcG complexes in the developing eye [15 , 60 , 61] . Ultimately , a multifaceted approach involving targeted mutations of individual response elements , combined with transgenic strategies , in backgrounds with altered availability of PcG gene products will likely be required to unravel precise roles for PcG in regulating genes in the developing Drosophila eye . Stocks carrying alleles eya2 , eya3 , eya4 , eyaE1 , eyaE4 , eyaD1 , eyacs , Df ( 2L ) eya , ETD2 . 2 ( a chromosome carrying eya2 and the transvection-disrupting inversion In ( 2LR ) 29C;41 ) , and ETD4 . 3 ( an eya4 background carrying a transvection-disrupting cyclical translocation with new order T ( 2; 3; 4 ) 30A; 101; 98D ) were obtained from Nancy Bonini ( Department of Biology , University of Pennsylvania , PA ) . Stocks carrying eyaD3 , eyaD6 , and eyaD7 were obtained from Justin Kumar ( Department of Biology , Indiana University , IN ) . Stocks carrying eya137 . 39 , eya117 . 36 , and eya7 . 42 were provided by Jennifer Jemc Mierisch ( Department of Biology , Loyola University , Chicago , IL ) . A stock carrying eya54C2 was obtained from Denise Montell ( Department of Molecular , Cellular , and Developmental Biology , UC Santa Barbara , Santa Barbara , CA ) . Stocks carrying eya1 , eyaEY13242 , and eyacliftIID were obtained from the Bloomington Drosophila Stock Center ( Indiana University , IN ) . Stocks carrying Cap-H20019 and Cap-H25163 were provided by Giovanni Bosco ( Geisel School of Medicine , Dartmouth College , NH ) . Stocks carrying Pc1 and E ( z ) S1 ( also known as E ( z ) 60 ) were obtained from Judy Kassis ( NIH ) . All flies were maintained at 25°C in standard 25 mm-diameter vials containing cornmeal , yeast , sugar , and agar medium with p-hydroxybenzoic acid methyl ester to prevent mold [10] . To assess adult eye development , crosses were established between 1–4 males and 2–5 virgin females of the selected genotypes . Progeny flies were collected 1 to 5 days post-eclosion and frozen for preservation . Fly eyes were imaged using a Canon EOS Rebel Tli digital camera mounted on a Leica MZ7 . 5 stereomicroscope . For each eye , the number of ommatidia was scored manually from the digital images . Mean count data and standard deviations for crosses examining enhancer action in trans are presented in S3 Table as well as the main text figures . Statistical comparisons were made using Graphpad Prism or R . Backgrounds carrying PcG mutations occasionally showed suppression of pairing sensitive silencing in late-eclosing flies from vials that were overcrowded , which is similar to observations of sex comb phenotypes induced by other PcG mutations [62] . We did not observe evidence of changes in severity of phenotype for other allelic combinations of eya according to eclosion time or crowding . Nevertheless , we avoided overcrowding and did not score flies beyond day 5 of eclosion . Strategies and analysis for the identification of molecular lesions in alleles of eya are detailed in S1 File . To create transcript-specific RNA probes , exon 1 of the eya-A transcript and exon 1 of the eya-B transcript were each amplified from genomic DNA using primer pairs eyaISA1F/eyaISA1R and eyaISB1F/eyaISB1R , respectively ( Primer sequences are provided in S1 Table ) . For both primer pairs , the reverse primer included a 5’ extension carrying the promoter for T7 RNA polymerase . PCR products were purified using a PCR Purification kit ( Qiagen ) , and 1 μg of each PCR product was used as a template to create digoxygenin-labeled RNA probes using a Dig RNA Labelling Kit ( Roche ) . The products of the reaction were ethanol precipitated and resuspended in 250 μl of 50% formamide/50% TE with 0 . 1% Tween-20 . For in situ hybridization , eye-antennal discs were dissected in PBS , transferred to a 1 . 5ml microcentrifuge tube , fixed in 4% formaldehyde in PBS on ice for 20 minutes , then fixed further in 4% formaldehyde/PBS with 0 . 6% Triton X-100 at room temperature for 20 minutes . After washing in PBS + 0 . 6% Triton X-100 ( 3 x 5 min ) , discs were rinsed with 50% PBS/50% Hybridization Buffer ( HB , 50% formamide , 2X SSC , 1X Denhardt’s , 250 μg/ml tRNA , 250 μg/ml salmon sperm DNA , 50 μg/ml heparin sulfate , 5% dextran sulfate , 0 . 1% Tween-20 ) , then pre-hybridized for 1 hour in 500 μl HB at 52°C . Discs were then incubated overnight in HB with a 1:100 dilution of digoxygenin-labeled probe at 52°C with agitation , followed by four changes of wash solution ( 50% formamide/2xSSC/0 . 1% Tween-20 ) over the next 24 hours at 52°C . Discs were rinsed with PBT ( PBS + 0 . 1% Triton X-100 ) , then incubated in PBT for 30 minutes at room temperature . Next , anti-digoxygenin antibody conjugated to horseradish peroxidase ( HRP ) ( Abcam ) was added at a dilution of 1:500 , and discs were incubated overnight at 4°C . After four 20-minute washes in PBT at room temperature , discs were developed with a TSA-Plus Cy3 detection kit ( Perkin Elmer NEL744E001KT ) , washed in PBT ( 3 x 5 minutes ) and mounted in Fluoromount G ( Electron Microscopy Services ) . Discs were visualized on either a Zeiss Axio Imager . A2 fluorescence microscope with an AxioCam MRm camera and Zen software , or a Leica SP8 confocal microscope with LASX software . Assessment of eya mRNA levels via quantitative RT-PCR was carried out as previously described [10] . Briefly , for each sample , 20 imaginal discs were dissected from wandering third instar larvae and frozen at -80°C . Tissue homogenization , genomic DNA elimination , and RNA purification were carried out using an RNeasy plus kit ( Qiagen ) according to the manufacturer’s protocol . PCR was performed on a StepOne Real-Time PCR system ( Applied Biosystems ) using cDNA diluted 1:5 into SYBR green PCR Mastermix ( Applied Biosystems ) . Primers were designed to specifically amplify the first exon of either the B transcript ( primers eyaRTF1 and eyaRTR1 ) or the A transcript ( primers eyaRT_AF1 and eyaRT_AR1 ) . For discs from CRISPR-edited flies , primers RP49-58F and RP49-175R were used to amplify the housekeeping rp49 cDNA as an internal reference [10] . For discs wherein eye development depended on transvection , RP49 does not present a suitable internal control due to the varying levels of eye tissue relative to the remaining tissues in the disc; in these experiments , the eya-B transcript was used as an internal reference for eya-A to provide a relative measure of eya-A:eya-B transcription . Relative levels of transcript were calculated via the ΔΔCt method using StepOne software . To generate a deletion of the eye-specific eya exon 1B , primers were designed to create guide RNAs complementary to sequences 64–84 bp upstream of nucleotide +1 of exon 1B and 537–557 bp downstream of the last nucleotide of exon 1B , spanning roughly 1 . 1 kb of genomic DNA in total ( S1 Table ) . Guide RNAs were cloned into the plasmid pU6-BbsI-chiRNA as previously described [63] , and a mixture of two plasmids carrying upstream- and downstream-targeting guide RNAs ( 250 ng/μl each ) was injected into embryos expressing Cas9 under the control of the Actin5C promoter [64] by BestGene , Inc . From 200 injected embryos , 96 G0 adults eclosed and were crossed to flies carrying the second chromosome balancer CyO . 89 fertile G0 adults were subsequently tested for evidence of an exon 1B deletion via PCR using primers eyaCRISPR34check_F and eyaCRISPR34check_R , which flank the region to be deleted via non-homologous end joining of double strand breaks; 68 ( 76 . 4% ) of these PCRs produced a single 2 . 5 kb PCR fragment consistent with unmodified wt DNA , whereas 21 ( 23 . 6% ) of the PCRs produced additional smaller fragments indicative of putative deletions . Of the 21 G0 flies carrying candidate deletions , four were found to transmit the deletion through the germline , and isogenic stocks were established from three of these . Sequencing of PCR fragments generated from each stock confirmed that each carries a deletion of roughly 1 . 1 kb spanning the distance between the two guide RNAs and including exon 1B and its promoter .	Gene regulation requires interactions between regions of DNA known as regulatory elements , which , in combination , determine where and when a gene will be active or silenced . Some genes use just a few regulatory elements , whereas others rely on highly complex interactions between many different elements that are poorly understood . While we typically imagine regulatory elements interacting with one another along the length of a single chromosome , in a curious phenomenon called transvection , elements can communicate between two different chromosomes that are held in close proximity . Here , we use the study of transvection to better understand how different regulatory elements contribute to the expression of eyes absent ( eya ) , a gene required for proper eye development in Drosophila . Our data show that a class of elements that initiate eya gene expression , called promoters , will compete with one another for activation by eya’s enhancers , a second class of regulatory element , with the promoter that is closest to the enhancers being the favored target for activation . Furthermore , our study of transvection uncovers an important role for a silencing element , called a PRE , in opposing eya gene expression . Overall , our study sheds new light on how different elements combine to produce patterned expression of eya .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "invertebrates", "medicine", "and", "health", "sciences", "gene", "regulation", "alleles", "animals", "invertebrate", "genomics", "dna", "transcription", "animal", "models", "drosophila", "melanogaster", "model", "organisms", "experimental", "organism", "systems", "eyes",...	2019	Two modes of transvection at the eyes absent gene of Drosophila demonstrate plasticity in transcriptional regulatory interactions in cis and in trans
A previous survey for clinical cases of Buruli ulcer ( BU ) in the Mapé Basin of Cameroon suggested that , compared to older age groups , very young children may be less exposed to Mycobacterium ulcerans . Here we determined serum IgG titres against the 18 kDa small heat shock protein ( shsp ) of M . ulcerans in 875 individuals living in the BU endemic river basins of the Mapé in Cameroon and the Densu in Ghana . While none of the sera collected from children below the age of four contained significant amounts of 18 kDa shsp specific antibodies , the majority of sera had high IgG titres against the Plasmodium falciparum merozoite surface protein 1 ( MSP-1 ) . These data suggest that exposure to M . ulcerans increases at an age which coincides with the children moving further away from their homes and having more intense environmental contact , including exposure to water bodies at the periphery of their villages . It has been established that the chronic necrotizing skin disease BU is caused by the emerging pathogen Mycobacterium ulcerans , however the mode ( s ) of transmission and environmental reservoirs are still unknown . Comparative genetic studies have revealed that M . ulcerans has diverged from the fish pathogen M . marinum . Through the acquisition of a plasmid , M . ulcerans has gained the ability to produce a cytotoxic and immunosuppressive macrolide toxin , referred to as mycolactone [1] , [2] . In addition to M . ulcerans strains isolated from human lesions , which belong either to the classical or to the ancestral lineage [3] , other mycolactone-producing mycobacteria ( MPM ) have been identified as fish and frog pathogens and given diverse species names [4]–[7] . However , recent comparative genomic analyses have shown that all MPM are genetically closely related and can be divided into three principal ecovars of M . ulcerans [8] . Extensive pseudogene formation and genome downsizing of the human M . ulcerans pathogen are indicative for an adaptation to a more stable ecological niche . In African endemic settings both the physical environment and organisms such as amoeba , insects , fish and frogs have been proposed as possible environmental reservoirs of the pathogen [9] . Accordingly , direct inoculation of bacteria into the skin from an environmental reservoir , but also bites from insects , such as mosquitos or water bugs have been suggested as route of infection . While possums have been identified as an animal reservoir in BU endemic areas of Southern Australia [10] , no mammalian reservoir has so far been detected in Africa . The distribution pattern of lesions is not indicative for a particular route of infection [11] and a genetic fingerprinting study of M . ulcerans isolates has revealed a highly focal transmission pattern , which excludes certain modes of transmission [12] . While it has long been generalized that in African BU endemic areas children below the age of 15 are most affected by the disease [13] , population age-stratified data from our previous survey for BU in the Mapé Basin of Cameroon showed that children less than five years old were underrepresented among cases [11] . One explanation for this observation may be a lower degree of exposure of very young children to M . ulcerans . Sero-epidemiological studies in Ghana have shown that screening blood sera of local populations for the presence of IgG specific for the 18 kDa shsp of M . ulcerans represents a tool to monitor exposure of populations to M . ulcerans [14] . However , in these investigations study participants were older than five years of age . Since a proportion of study participants of all age groups tested positive , it is still not known at which age immune responses against M . ulcerans start to emerge and hence where and at which age exposure to the pathogen begins . In the present sero-epidemiological study the potential association between age and exposure to M . ulcerans was investigated by determining anti-18 kDa shsp IgG titres in 875 individuals from BU endemic sites in the Densu River Basin of Ghana and the Mapé Basin of Cameroon . In these cross-sectional surveys we included more than 100 children less than five years old allowing us to estimate the age of sero-conversion , which may provide another cornerstone in the search for the mode of M . ulcerans transmission . Ethical clearance for the collection and testing of human blood samples from Ghana and Cameroon was obtained from the institutional review board of the Noguchi Memorial Institute for Medical Research ( Federal-wide Assurance number FWA00001824 ) and the Cameroon National Ethics Committee ( N°172/CNE/SE/201 ) as well as the Ethics Committee of Basel ( EKBB , reference no . 53/11 ) . Written informed consent was obtained from all individuals involved in the study . Parents or guardians provided written consent on behalf of children . We investigated the association between age and exposure to M . ulcerans by determining serum antibody titres against the 18 kDa shsp in individuals living in two different BU endemic areas . In Cameroon , serum samples were collected from inhabitants of the village of Mbandji 2 . This village is located in the Bankim Rural Health Area of the Bankim Health District , where we conducted a cross-sectional house-by-house survey for BU in early 2010 , including the collection of data on the population age structure . These data and the subsequent identification of BU cases until June 2012 were published in our previous study [11] . In the present study we provide updated information based on a continued monitoring of new BU cases in this area until May 2013 . The age-specific incidence rates were calculated using the ages of the BU cases identified between March 2010 and May 2013 and the population age distribution as collected in the house-by-house survey in the Bankim Health District . Sera were collected in January 2011 from all inhabitants of Mbandji 2 , who agreed to participate ( 395 individuals with a nearly equal gender distribution ) . Re-sampling of 80 blood donors from Mbandji 2 was carried out one year after the first blood collection to analyze stability of anti-18 kDa shsp serum IgG levels over time . The second study site comprised villages within the Obom sub-district of the Ga-South district in Ghana . This sub-district is one of the major BU endemic communities along the Densu River Basin . The villages from which the sera were collected , have active transmission on-going as they have continuously reported cases for the past five years . Study participants included 96 laboratory confirmed BU patients ( 57 females and 39 males ) as well as 4 age- , sex- , and home village-matched controls for each patient ( 384 control individuals ) . Demographic data as well as history of known previous mycobacterial infections were recorded for all participants at both sites . While the majority of individuals had no history of mycobacterial infections , eight study participants from Mbandji 2 reported to having had tuberculosis ( 2 ) , leprosy ( 1 ) or BU ( 5 ) . All control participants recruited in Ghana had no history of mycobacterial infection . The age distribution of study participants from Cameroon and Ghana is shown in Figure 1A and 1B , respectively . Blood sera from the 875 individuals were tested for the presence of anti-18 kDa shsp antibodies in an ELISA format . In addition , 96 sera from children <5 living in Mbandji 2 were tested by Western Blot analysis for the presence of antibodies against this protein , as well as against a Plasmodium falciparum MSP-1 protein domain in order to assess the exposure and immune responses of child study participants to this mosquito transmitted parasite . 96-well Nunc-Immuno Maxisorp plates ( Thermo Scientific ) were coated with 0 . 25 µg recombinant M . ulcerans 18 kDa shsp per well in 100 µl phosphate-buffered saline ( PBS ) and incubated over night at 4°C . Plates were washed four times with washing buffer ( dH2O , 2 . 5% Tween 20 ) before being incubated with blocking buffer 1 ( 5% skim milk in PBS ) for 2 hours at room temperature ( RT ) . After washing as described above , 50 µl of 1∶100 diluted human blood sera in blocking buffer 2 ( 1% skim milk in PBS ) was added to each well and incubated for 2 hours at RT . Following a further washing step , 50 µl of 1∶8000 diluted goat anti-human IgG ( γ-chain specific ) antibodies coupled to horseradish peroxidase ( HRP , SouthernBiotech ) in blocking buffer 2 was added to each well and incubated for 1 . 5 hours at RT . Plates were washed and 50 µl TMB Microwell Peroxidase Substrate ( KPL ) was added per well . The reaction was stopped after 5 minutes using 0 . 16 M sulfuric acid . The absorbance was measured at 450 nm in a Tecan Sunrise microplate reader . 15 µg of recombinant M . ulcerans 18 kDa shsp or 5 µg of a Plasmodium falciparum MSP-1 protein domain ( amino acids 34-469 of strain K1 ) were separated on NuPAGE Novex 4–12% Bis-Tris ZOOM Gels with 1 . 0 mm IPG well ( Invitrogen ) using NuPAGE MES SDS Running Buffer ( Invitrogen ) under reducing conditions . After electrophoresis the proteins were transferred onto nitrocellulose membranes using an iBlot Gel Transfer Device ( Invitrogen ) . Membranes were blocked with blocking buffer 3 ( 5% skim milk in PBS containing 0 . 1% Tween 20 ) and cut into thin strips . Membrane strips were then incubated with human blood sera at a 1∶1000 dilution in blocking buffer 3 for 2 hours at RT . Strips were repeatedly washed with 0 . 3 M PBS containing 1% Tween 20 and after that incubated with 1∶20'000 diluted goat anti-human IgG ( γ-chain specific ) antibodies coupled to HRP ( SouthernBiotech ) for 1 hour at RT . After a second washing step , bands were visualized by chemiluminescence using ECL Western Blotting substrate ( Pierce ) . ELISA results were analyzed using GraphPad Prism version 6 . 0 ( GraphPad Software , San Diego California USA ) and R version 3 . 0 . 1 [15] . The distribution of antibody titres and the differences between two successive antibody titres are presented as box plots . These comprise a line for the median , edges for the 25th and 75th percentiles and traditional Tukey whiskers showing 1 . 5 times the interquartile distance . Dots on the graph represent individual points that lie outside that range . We compared changes in OD between age categories in the Cameroon dataset using the Kruskal-Wallis test . Levene's test for homogeneity of variances was used to compare the degree of variation by age category . We compared the OD values for the Ghana matched cases and controls using conditional logistic regression . The overall bias and variation between the first and second Ghanaian serum samples was estimated using the Bland-Altman method [16] . The age-specific BU incidence rates for the population in the Mapé Basin were calculated using 76 BU cases identified between March 2010 and May 2013 . Based on these cases , a low incidence rate of BU was detected for children less than 4 years of age ( Figure 2A ) . The age-distribution of IgG titres against the M . ulcerans 18 kDa shsp for a cross-sectional survey of 395 individuals from the village Mbandji 2 is shown in Figure 2B . While high antibody titres were detected in individuals of all age groups over 4 years , none of the children younger than 4 years showed an ELISA IgG titre above the background , which was determined by Western Blot analysis as OD < 0 . 35 . Analysis of the sera sampled from children less than 7 years old by Western Blot analysis showed no specific bands representing IgG antibodies against the 18 kDa shsp for sera from children <4 years of age ( Figure 3 ) . In contrast , Western Blot positive sera were found in all tested age groups >4 years old . Since very weak IgG titres were recorded for some of the sera from 4 year olds , sero-conversion may start in some children around this age . IgG titres against a recombinant fragment of MSP-1 were determined by Western Blot analysis . In contrast to the lack of antibody responses against the 18 kDa shsp in children <4 years old , serum IgG responses against a P . falciparum malaria parasite MSP-1 domain were detected in all age groups tested . Strong staining of the MSP-1 band was observed for the majority of sera collected from children between one and seven years of age as well as for one of the infants ( Figure 4 ) . One year after the first serum collection in Mbandji 2 , 80 of the 395 study participants were re-sampled . While only minimal changes in antibody titres against the 18 kDa shsp were recorded overall , more individuals had a decreased than an increased serum IgG level after one year ( Figure 5A ) . Increases in OD tended to be small and confined to the older children and young adults ( Figure 5B ) . The most distinct changes , characterized by a marked decrease of antibody titres between the two surveys , occurred in young adults . There was a significant association between age group and the absolute change in OD ( Kruskal-Wallis test p = 0 . 01 ) and borderline evidence of an association between the variation in changes in OD and age group ( Levene's test for homogeneity of variances , p = 0 . 08 ) . M . ulcerans 18 kDa shsp specific IgG titres were also determined in sera from 96 BU patients and 384 healthy matched control individuals living in a second BU endemic site in West Africa , the Densu River Valley in Ghana . Each serum sample was tested twice , once in each of two independent experiments ( Figure S1 ) . Negligible overall bias between experiments was observed with the mean difference ( OD1-OD2 ) of 0 . 024 . There was also a reasonably small variation in the individual differences with the 95% limits of agreement from −0 . 0796 to 0 . 1278 . There was no evidence of a difference in the ELISA OD values between the cases and controls ( p = 0 . 99 ) ( Figure 6A ) . While sero-responders were identified in all age groups of individuals more than 6 years old , none of the sera from children younger than 5 years exhibited a distinct anti-18 kDa shsp IgG titre ( Figure 6B ) . Western Blot analysis of sera from 2-year-old children confirmed the absence of anti-M . ulcerans 18 kDa shsp IgG in these samples ( Figure 6C ) . Results of representative subsets of sera which tested negative , moderately positive or highly positive by ELISA were reconfirmed by Western Blot analysis , showing good agreement between ELISA OD values and Western Blot band intensities ( Figure S2 ) . A high degree of antigenic cross-reactivity among mycobacterial species complicates investigations on M . ulcerans-specific humoral immune responses . However , the immunodominant 18 kDa shsp [17] , which is overexpressed in M . ulcerans [18] , represents a suitable serological marker for exposure to M . ulcerans [14] . Diverse outcomes of infection with other mycobacteria , such as M . tuberculosis and M . leprae have been associated with both host and pathogen factors . While only one study has investigated a possible association between BU and host genetics [19] , various behavioural factors that may lead to increased risk to develop the disease have been reported , with poor wound care , failure to wear protective clothing , and living or working near water bodies being the most common risk factors identified [20] . While the generalization persists that children <15 years old are most affected by the disease [13] , our recent survey for BU in the Mapé Basin [11] and continued monitoring of new BU cases in this region have revealed that the risk of BU is as high in individuals above the age of 50 as in young teenagers and that very young children below the age of four are underrepresented among cases when adjusting for the population age distribution . Data of our previous sero-epidemiological investigations revealed that the proportion of individuals from a BU endemic area showing serum IgG titres against the 18 kDa shsp of M . ulcerans is comparable for all age groups >5 years [14] . Results of the present study , including for the first time a substantial number of serum samples from children <5 years of age , showed that children of this age group have not yet sero-converted . Hence , young children appear to be considerably less exposed to M . ulcerans . This reduced exposure may be explained by the smaller movement radius away from the house of these very young children . Although , these small children do leave the house , they usually do so being carried by a caregiver and are therefore not in direct contact with the environment , at more distant places from their homes . No significant difference could be observed when comparing anti-M . ulcerans 18 kDa shsp antibody titres between BU patients and controls . This may be related to the immune-suppressive effect of mycolactone and concurs with the lack of a serological response in experimentally infected mice ( unpublished data ) . The results of a case-control study carried out in a BU endemic region of south-eastern Australia indicated reduced odds of having BU for individuals who frequently used insect repellent and increased odds for those who were bitten by mosquitoes [21] . In African BU endemic settings , the highly focal transmission of M . ulcerans haplotypes [12] , [22] , [23] , as well as the distribution pattern of BU lesions on the body [11] , speak against an exclusive role of mosquito vectors in transmission . Here we observed in children <5 years frequent sero-conversion for the MSP-1 antigen of the mosquito-transmitted malaria parasites in the absence of an IgG response against the M . ulcerans 18 kDa shsp . The age distribution of BU cases and the relatively abrupt changes in this risk of contracting BU with age do not speak for transmission of BU by mosquito species commonly found within the small movement radius of very young children . Within the framework of our analyses , blood was collected for a second time from a limited number of participants one year after the first sample . Results of this pilot study showed that anti-18 kDa shsp IgG titres were relatively stable in older adults . Future studies of the age-related changes in behaviour of three to six year old children , monitoring of their movement radius and water contact patterns in combination with larger longitudinal serological and environmental studies may have the potential to shed further light onto the mode of transmission and relevant environmental reservoirs of M . ulcerans .	Although M . ulcerans , the causative agent of Buruli ulcer ( BU ) , was identified in 1948 , its transmission pathways and environmental reservoirs remain poorly understood . The occurrence of M . ulcerans infections in endemic countries in West and Central Africa is highly focal and associated with stagnant and slow flowing water bodies . BU is often described as a disease mainly affecting children <15 years of age . However , taking the population age distribution into account , our recent longitudinal survey for BU in the Mapé Dam Region of Cameroon revealed that clinical cases of BU among children <5 years are relatively rare . In accordance with these findings , data of the present sero-epidemiological study indicate that children <4 years old are less exposed to M . ulcerans than older children . Sero-conversion is associated with age , which may be due to age-related changes in behavioural factors , such as a wider movement radius of older children , including more frequent contact with water bodies at the periphery of their villages .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "infectious", "diseases", "medicine", "and", "health", "sciences", "emerging", "infectious", "diseases", "medical", "microbiology", "microbial", "pathogens", "biology", "and", "life", "sciences", "immunology", "microbiology", "bacterial", "pathogens", "immune", "response"...	2014	Late Onset of the Serological Response against the 18 kDa Small Heat Shock Protein of Mycobacterium ulcerans in Children
Neurocysticercosis ( NCC ) is a major cause of epilepsy in regions where pigs are free-ranging and hygiene is poor . Pork production is expected to increase in the next decade in sub-Saharan Africa , hence NCC will likely become more prevalent . In this study , people with epilepsy ( PWE , n = 212 ) were followed up 28 . 6 months after diagnosis of epilepsy . CT scans were performed , and serum and cerebrospinal fluid ( CSF ) of selected PWE were analysed . We compared the demographic data , clinical characteristics , and associated risk factors of PWE with and without NCC . PWE with NCC ( n = 35 ) were more likely to be older at first seizure ( 24 . 3 vs . 16 . 3 years , p = 0 . 097 ) , consumed more pork ( 97 . 1% vs . 73 . 6% , p = 0 . 001 ) , and were more often a member of the Iraqw tribe ( 94 . 3% vs . 67 . 8% , p = 0 . 005 ) than PWE without NCC ( n = 177 ) . PWE and NCC who were compliant with anti-epileptic medications had a significantly higher reduction of seizures ( 98 . 6% vs . 89 . 2% , p = 0 . 046 ) . Other characteristics such as gender , seizure frequency , compliance , past medical history , close contact with pigs , use of latrines and family history of seizures did not differ significantly between the two groups . The number of NCC lesions and active NCC lesions were significantly associated with a positive antibody result . The electroimmunotransfer blot , developed by the Centers for Disease Control and Prevention , was more sensitive than a commercial western blot , especially in PWE and cerebral calcifications . This is the first study to systematically compare the clinical characteristics of PWE due to NCC or other causes and to explore the utility of two different antibody tests for diagnosis of NCC in sub-Saharan Africa . Epilepsy is one of the most common neurological disorders worldwide . More than 80% of people with epilepsy ( PWE ) live in low income countries [1] and more than 75% of them are not treated sufficiently [2] . The prevalence of epilepsy is two to ten times higher and the incidence rate twice that of high income countries [3] . In northern Tanzania , where our study took place , recent results indicate a prevalence of 11 . 2 per 1000 [4] . One important step towards reducing the burden of epilepsy is to assess its prevalence , causes and risk factors in resource poor countries . According to the International League Against Epilepsy ( ILAE ) , neurocysticercosis ( NCC ) is a growing problem in tropical countries and increasingly recognized as a leading cause of epilepsy [5]–[8] . A recent meta-analysis including only studies from Africa revealed a highly significant association between cysticercosis and epilepsy , suggesting that NCC is a major cause of epilepsy in Africa [9] . In endemic countries , NCC is the cause of epilepsy in more than a quarter of PWE [10] . Worldwide , NCC is the most common parasitic disease of the nervous system [1] . Cysticercosis occurs when humans become infected with Taenia solium eggs and develop the larval stage . Typical clinical manifestations of NCC are epileptic seizures , which are caused by the cysticerci themselves and by the host's immune response [11] . Pig farming has increased considerably in East and South Africa , especially in rural , low income , smallholder communities [12] . Thus cysticercosis may represent a growing problem , especially in rural areas like our study area , where proper meat inspection does not exist and lack of health education and hygiene is common [11] . The costs of cysticercosis are considerable . Carabin et al . calculated a loss of 34 . 2 million US-Dollars caused by cysticercosis for the Eastern Cape Province in South Africa with 7 million inhabitants in 2004 [13] . Diagnostic accuracy is the key to identify people with NCC and to treat them according to established guidelines . In most low income countries where NCC is rampant , neuroimaging is not readily available . Diagnosis is established based on medical history , physical examination and , if present , laboratory testing . In our study , we present clinical details of PWE with and without NCC , results of neuroimaging and of two different antibody tests for cysticercosis in an area of northern Tanzania , which only recently has been shown to be highly endemic for cysticercosis [14] , [15] . The project was ethically cleared by The National Institute for Medical Research ( NIMR ) , Tanzania . Patients were informed about the risks and benefits of participating in the study , computed tomography ( CT ) examination and necessary treatments . Written informed consent was obtained from all participants . Patients received free anti-epileptic and , if necessary , anthelmintic treatment according to national guidelines during and after the study . The shipment of samples to the Centers for Disease Control and Prevention , in Atlanta , GA ( CDC ) , was in accordance with institutional review boards and material transfer agreements between Haydom Lutheran Hospital ( HLH ) , CDC , NIMR and the Medical University Innsbruck . The study was conducted at HLH , which is situated in a remote rural area of northern Tanzania . The total immediate catchment area of the hospital comprises 316 , 168 people [16] . In 2002 , one of the authors ( ASW ) established a clinic for PWE at HLH , which nowadays cares for more than 400 PWE . The implementation of CT scanning in 2005 further improved diagnostic capacity for people with epileptic seizures/epilepsy . The main tribes in the study area are Iraqw , Datoga and Bantu tribes . The Iraqw and Bantu are agro-pastoralists , while the Datoga traditionally have been nomadic pastoralists , though many are now less nomadic and more agricultural . In a former project between August 2002 and October 2004 , our research group identified and characterised 346 people with epileptic seizures [17] . Demographic and clinical characteristics , psychosocial as well as sociocultural aspects and the pattern of injuries were evaluated . Epilepsy was defined as two or more afebrile seizures unrelated to acute metabolic disorders or to withdrawal of drugs or alcohol [18] and grouped according to a classification developed for resource poor countries [19] , which is based on the ILAE classification for epileptic seizures [20] . The diagnosis was based on seizure semiology , past and present medical history , other associated risk factors and physical examination only . Imaging or EEG was not available at time of diagnosis of epilepsy . Of these 346 people with epileptic seizures 212 participated in the CT study . Due to ethical reasons children younger than ten years at time of diagnosis of epilepsy and pregnant women were excluded . For more details on inclusion and exclusion criteria see Figure 1 . All participating PWE were interviewed by a doctor in training and two local nurses using a follow up protocol , which had been validated in a previous study [17] . Additional questions related to NCC were added . Compliance with treatment was assessed by checking regular attendance at HLH and , if attending regularly , calculating the number of tablets that should have remained since the last appointment and counter-checking it with the actual number of tablets . All 212 PWE underwent CT examination . The CT machine was a Toshiba Auklet Single Slice Spiral CT . The thickness of slices was 5 mm in the posterior cranial fossa and 10 mm for the rest . Intravenous contrast medium was applied in all patients . The pictures were saved digitally and sent to Innsbruck , where a neuroradiologist reviewed all scans . A detailed description of the results on neuroimaging has recently been published [15] . CT based diagnosis of NCC was divided into three groups: definite NCC lesions , lesions highly suggestive of , and those compatible with NCC . Definite NCC lesions were cystic lesions showing the scolex . Any cystic lesion without a visible scolex , single or multiple ring or nodular enhancing lesions and parenchymal brain calcifications were categorized as lesions highly suggestive of NCC [21] , [22] . Any pathology that might be caused by NCC such as hydrocephalus or enhancement of the leptomeninges was considered compatible [21] , [22] , single calcifications in parenchymal brain were also considered compatible . Active NCC was defined as any cystic lesions or lesions with ring enhancement . Parenchymal calcifications were classified as inactive [23] , [24] . Due to financial restrictions only part of the collected serum samples ( 20 of PWE with either highly suggestive or definite NCC lesions and 20 of PWE without NCC lesions on CT scan ) and all 11 CSF samples of PWE with highly suggestive or definite NCC lesions on CT scan were analysed with a commercially available western blot ( CWB ) for cysticercosis ( LDBio , Lyon , France ) [25]–[27] at the Department for Medical Parasitology , Institute of Specific Prophylaxis and Tropical Medicine , Medical University Vienna , Austria . The antigen used in this test was prepared from T . solium larvae ( cysticerci ) . At a later stage of the project all collected serum samples ( 28 of PWE with either highly suggestive or definite NCC lesions , 7 of PWE with compatible NCC lesions and 46 of PWE without NCC lesions on CT scan ) and CSF samples ( 11 CSF samples of PWE with highly suggestive or definite NCC lesions on CT scan ) were tested at the CDC using the CDC-developed enzyme-linked electroimmunotransfer blot ( CDC EITB ) . The CDC EITB was performed as described previously [28] . The criteria proposed by Del Brutto et al . were used for diagnosis of clinical NCC . A set of defined absolute , major and minor criteria , based on clinical signs and symptoms , neuroimaging , detection of antibodies and epidemiological considerations , were used to establish degrees of certainty for diagnosis , which are classified as definitive and probable NCC [21] , [22] . In our study , we compared PWE with definitive or probable NCC to those without NCC . A positive serological result in both tests ( CWB and CDC EITB ) was only considered once as a criterion for the diagnosis of NCC . All data were entered in a SPSS-database . Statistical analysis was performed with the same program . Since all numeric variables had a non-parametric distribution , differences between PWE with and without NCC were tested for significance with Mann-Whitney-U test . Categorical data were tested with Fisher's exact test . McNemar's test was applied to data from samples that were analysed with both antibody tests in order to find significant differences . A p-value lower than 0 . 05 was regarded as significant . Insufficient answers were left as missing data in the database and were excluded from statistical analysis . Hence the number of people within a specific category might differ from the total number of examined PWE ( n = 212 ) . The reduction of seizure frequency was calculated by dividing the seizure frequency after treatment in 2006 by the seizure frequency before treatment . People with no seizures in 2006 had a reduction of 100 percent . People with an increased number of seizures had zero percent reduction . The median duration of epilepsy from the day of diagnosis until the day of the follow up for this study was 28 . 6 months . The range was between 19 . 3 and 47 . 7 months . In our cohort of 212 PWE , 35 ( 16 . 5% ) satisfied the diagnostic criteria for either probable or definitive NCC . The main demographic findings are shown in Table 1 . People with active NCC lesions ( n = 6 ) with an average age of 47 . 7 years ( SD 23 . 4 ) were significantly older than people with inactive NCC ( n = 25 ) with an average age of 25 . 2 years ( ( SD 15 . 5 ) ; Mann-Whitney-U , p = 0 . 015 ) . Also , the mean age of PWE with active NCC at first seizure ( 41 . 2 years ( SD 23 . 0 ) ) was significantly higher when compared to PWE with inactive NCC ( 16 . 8 years ( SD 16 . 5 ) ; Mann-Whitney-U , p = 0 . 007 ) . The seizure frequencies , compliance with anti-epileptic drug ( AED ) and reduction of seizure frequency after treatment in PWE with and without NCC are listed in Table 1 . There was a trend towards generalised epilepsies without focal neurological signs that started outside the age range of idiopathic epilepsies ( implying that there may be structural brain damage ) and generalised epilepsies with clear focal signs being commoner in the NCC group . On the other hand , idiopathic epilepsies , which are usually generalised seizures starting within a specific age range , and epilepsies associated with brain damage were more frequent in PWE without NCC . The compliance with AED was slightly better in the group without NCC , although this did not reach significance . The administration of the type of AED was similar in both groups . Among compliant PWE taking their AED regularly , the group of PWE with NCC had a significantly higher reduction of seizures ( Table 1 ) . The percentage of people without seizures since their last visit was higher among compliant PWE ( compliant: 54 . 1% , 59/109; non-compliant: 40 . 6% , 41/101; Fisher's exact test p = 0 . 054 ) . Dividing compliant and non-compliant PWE according to the presence of NCC a higher percentage of people without seizures was observed among compliant PWE with NCC compared to those without NCC ( compliant: 78 . 6% , 11/14 PWE with NCC , 50 . 5% , 48/95 PWE without NCC , Fisher's exact test p = 0 . 082; non-compliant: 42 . 1% , 8/19 PWE with NCC , 40 . 2% , 33/82 PWE without NCC , p = 1 . 000 ) , although the result did not reach significance . In terms of past psychiatric history , family history of seizures and educational level no significant differences were found . For more details see Table 1 . Number of people in household as an indicator of crowding , number of pigs at home , close contact with pigs and use of latrines as possible risk factors for NCC were not significantly associated with NCC . However , the percentage of people who consumed pork was significantly higher in the group of people with NCC ( Table 2 ) . All samples which were positive using the CWB were also positive using the CDC EITB . Additionally , seven samples that were negative in the CWB were positive in the CDC EITB , which was statistically significant ( McNemar , p = 0 . 016 ) . Five of these seven had multiple calcified lesions highly suggestive of NCC , one had multiple calcified lesions and one hypodense lesion and one had a normal CT . The percentage of a positive CDC EITB result in the group of PWE with NCC calcifications ( = highly suggestive of NCC ) was much higher ( 52 . 2% , 12/23 ) than the percentage of the same group tested with the CWB ( 13 . 3% , 2/15; “sensitivity” , if CT was used as a gold standard ) . Even with the small number of samples that was analysed in both labs ( n = 15 ) , this approached significance ( negative in both tests: 8 , positive in both tests: 2 , CWB negative , CDC EITB positive: 5; McNemar , p = 0 . 063 ) . Concerning people with a normal CT , all samples ( 20/20 ) were negative with the CWB , whereas 91 . 3% ( 42/46 ) were negative with the CDC EITB ( “specificity” , if CT was used as gold standard ) . The results of CSF samples were identical in CWB and CDC EITB . Using the results of the CDC EITB as the serological gold standard , as suggested by Del Brutto et al . [21] , [22] , 17 people were classified with definitive NCC , versus 7 , if the data obtained from the CWB had been used ( Table 3 ) . Similarly , the percentage of PWE with probable and definitive NCC was 16 . 5% ( 35/212 ) using the CDC EITB results , compared to 13 . 7% ( 29/212 ) if the CWB was used . The results of both antibody tests together with the number of total NCC lesions on CT , including cysts and calcifications , are listed in Table 4 . The number of NCC lesions was significantly associated with a positive serum antibody result ( Mann-Whitney-U , CWB: p = 0 . 006 , CDC EITB: p<0 . 001 ) . People with more CT lesions had a higher chance of a positive antibody result . Probably due to the low number of samples , there was no significant association in CSF ( Mann-Whitney-U , p = 0 . 537 ) . Active NCC lesions were significantly associated with a positive antibody result in both tests for serum and CSF ( Fisher's exact test , Serum CWB: p = 0 . 002; Serum CDC EITB: p = 0 . 022; CSF both tests: p = 0 . 015 , see Table 4 ) . The prevalence of cysticercosis antibodies detected by the CDC EITB among people with single lesions was 22 . 2% ( 2/9 , “sensitivity” , if CT was used as gold standard ) . CWB was not performed in people with single lesions ( Table 4 ) . People with at least two lesions , when analysed with CDC EITB and CWB had antibodies in 65 . 4% ( 17/26 ) and 30% ( 6/20 ) , respectively ( “sensitivity” , if CT was used as a gold standard ) . Looking at individuals with a positive CSF result , all corresponding serum samples were positive in the CDC EITB and four out of five were positive in the CWB . To our knowledge this is the first study that systematically compares demographic data and clinical characteristics of PWE with and without NCC in sub-Saharan Africa . A differentiation between these two groups is important not only because of the possible treatment of active NCC , but also because of prevention and disease control programs for NCC . There are several limitations of our study . The diagnosis of epilepsy was made up to three years prior to the performance of the CT scan and the collection of specimens . Hence the number of people with active NCC might be underestimated . Due to lack of EEG the diagnosis of epilepsy is based on clinical examination and thorough interviews of patients and relatives . Epileptic seizures that appear generalized may have a short focal start that clinically goes unnoticed . Simple partial seizures may not be diagnosed , because patients do not report to the hospital . As to serology , it would have been preferable , if cysticercosis serology had been performed in all PWE . This unfortunately was not possible in our study . Demographic data about PWE in low income countries is limited . Singhvi et al . described the patterns of patients ( n = 100 ) with intractable epilepsy at a tertiary center in India . All patients received at least two AEDs . In their study , the mean age was 23 . 2 years with a mean duration of seizures of 11 . 4 years and their mean seizure frequency was 12 . 4 seizures per month [29] . In our study , the mean age was higher with 28 . 2 years and the average age at first seizure was 17 . 7 years . The older age of our study population could be explained by the exclusion of children younger than ten years at time of diagnosis . The mean seizure frequency in our drug naïve PWE with 7 . 7 seizures per month was lower compared to the study of Singhvi et al . The mean age of PWE with NCC in our study ( 32 . 5 years ) is comparable to the mean age ( 28 . 6 years ) of people with NCC in a large retrospective study of people with NCC treated in Houston , Texas , that included 202 individuals mainly presenting with seizures [30] . In our study , there was a trend that PWE with NCC were older and that they had their first seizure later compared to PWE without NCC . This may suggest that NCC should be considered as an underlying cause of epilepsy especially among PWE with a late onset of seizures . Other studies reported that NCC is the cause of late onset epilepsy in up to 50% [31]–[34] . However , a recent review revealed similar percentages of NCC associated epilepsy in the age group below 20 years and that of 20–54 years of age [10] . These data suggest that NCC has also to be considered as a cause of epilepsy in young PWE . In our study , PWE with active NCC were significantly older than PWE with inactive NCC , which might be explained by a different immune response to the infestation with age . Porcine cysticercosis was first detected in our study area in the late 1980s and previously pig husbandry was not common [12] . If infestations caused more calcifications in younger people , the lower number of calcifications among older people could also be explained by the absence of NCC before the introduction of pig husbandry . The gender in our study population of PWE was equally distributed , which is in accordance with another study in Tanzania [35] and with the conclusion of reviews about the prevalence of epilepsy in Asia [36] and sub-Saharan Africa [1] , which did not find a significant difference regarding gender . In our study slightly more males had NCC , a result similar to the one reported by Nicoletti et al . , who diagnosed 21 males and 14 females with NCC among 124 PWE [34] . It is remarkable that , although PWE with NCC were less compliant , they had a lower seizure frequency . The reduction of seizure frequency in compliant PWE was significantly higher in PWE with NCC compared to those without NCC . 78 . 6% of compliant PWE with NCC had no seizures , whereas this was only the case in 50 . 5% of compliant PWE without NCC . These observations not only suggest that the frequency of seizures due to NCC seems to decline faster over time , but also that anti-epileptic treatment with carbamazepine or phenobarbitone , which is available in most low income countries , is very efficient in controlling seizures due NCC . Del Brutto et al . , who investigated 203 PWE with NCC , reported a similar seizure control rate of 83% [37] . The reason for the higher percentage of NCC among people of the Iraqw tribe might be that according to our observation pig farming is more common in that tribe compared to other tribes . This difference might disappear in future , because tribes are intermingling and pig husbandry seems to become abundant in all tribes . The past psychiatric history and family history of seizures seem to be similar in PWE with and without NCC . In our study , the only significant risk factor associated with NCC was pork consumption . In this context , we want to stress that cysticercosis is a faecal-oral infestation and pork consumption is not a prerequisite for cysticercosis . The consumption of undercooked pork or the handling of infected pig meat could lead to a tapeworm carrier in the household . Lescano et al . showed that there is a significant cysticercosis seroprevalence gradient around a tapeworm carrier and that the risk for cysticercosis is high , if there is a tapeworm carrier in the household [38] . Other factors related to social status such as number of people per household or educational level and also pig contact or number of pigs per household were not significantly associated with NCC . The use of latrines , which has been identified as a risk factor for porcine cysticercosis [39] , was almost omnipresent in our study population . The number of people who did not use a latrine ( 6/209 ) was too low to identify significant differences . It is difficult to define sensitivity and specificity of serological tests for cysticercosis , because the gold standard for the diagnosis of NCC is not well defined . Probably the best available single examination to diagnose NCC in low income countries is a CT scan . Taking the CT result as a gold standard , the comparison between the two antibody tests for cysticercosis revealed a difference mainly in terms of sensitivity . Especially among people with lesions highly suggestive of NCC , which are mainly multiple calcified lesions , the CDC EITB had a higher sensitivity than the CWB . Antibodies were detected in 52 . 2% and 13 . 3% with the CDC EITB and the CWB , respectively . However , sensitivity was low compared to other studies that reported a sensitivity of 60% to 89% in sera with only calcified cysts [40] , [41] . Rajshekhar et al . , who studied NCC and epilepsy in India , showed a positive EITB result in 26 . 1% ( 12/46 ) of PWE with NCC lesions on CT scan [7] . This relatively low number could be explained by the high percentage of people with only one or two calcifications ( 81 . 3% ) . In Peru , Montano et al . detected antibodies in 46 . 7% ( 7/15 ) of PWE with NCC lesions on CT scan . The type of lesions however was not specified [6] . In active lesions , where the CDC EITB detected antibodies in all six samples and the CWB in five of six , the number of samples was too low to show significant differences . Our results confirm the high sensitivity of the CDC EITB in people with more than one viable NCC cysts [28] , [40] . As expected the total number of active and inactive NCC lesions was positively correlated with a positive antibody result in both tests . It seems that a higher burden of disease leads more often to a detectable antibody response . When looking at PWE with normal CT , it seems that the specificity of the CDC EITB ( 91 . 3% ) is lower compared to the CWB ( 100% ) . However , it has to be considered that cysticercosis may be present in other organs such as eye and subcutaneous/muscular tissue or that calcifications smaller than 10 mm may be missed because of the thickness of CT slices . The specificity of the EITB is very high , near 100% , with only a few anecdotal reports [42] , [43] of false positive results since the time the test was introduced in 1989 . Therefore , it seems probable that the four PWE with a positive CDC EITB and normal CT scan may have had some exposure to T . solium larvae in the past or may harbour cysts in other organs than the brain . The analysis of CSF with CDC EITB did not give additional information regarding the diagnosis of NCC , because all CSF positive cases had also positive serum samples . Our results are in accordance with the study of Proaño-Narvaez et al , in which analysis of serum and CSF samples with a EITB for cysticercosis were equally sensitive and specific [44] . We conclude that an analysis of CSF may only be indicated , if a quantitative test is used in order to calculate a specific T . solium antibody index of a serum/CSF pair . In summary , our study compares demographic and clinical characteristics of PWE with and without NCC in rural northern Tanzania , showing that PWE with NCC tend to be older with a later onset of seizures compared to those without NCC . Seizure frequency in compliant PWE with NCC , using AEDs available on site , seems to respond better than in compliant PWE without NCC . The only risk factor for NCC that could be identified is consumption of pork . In addition , the sensitivity and specificity of a commercially available western blot and the CDC EITB was tested in PWE with NCC showing a higher sensitivity of the latter , especially in PWE with calcifications .	Neurocysticercosis , a preventable and treatable disease , is one of the main causes of epilepsy in low income countries . In these countries , the diagnosis of epilepsy is often based on clinical presentation and interviews as neuroimaging is rarely available . It is crucial to distinguish people with epilepsy due to neurocysticercosis from other people with epilepsy by clinical symptoms and/or serology , because the former warrants a specific approach both in terms of diagnosis and treatment . The authors compared the demographic and clinical data of the two groups and found that people with epilepsy due to neurocysticercosis are older , more likely to consume pork , and respond better to anti-epileptic treatment . Additionally , the authors compared two antibody tests for cysticercosis with computed tomography images , which showed a higher sensitivity of the CDC electroimmunotransfer blot compared to a commercial western blot . The number of neurocysticercosis lesions was significantly associated with a positive antibody result in both tests . In summary , this research describes clinical characteristics of people with epilepsy and neurocysticercosis and assesses the usefulness of two immunoblots in those patients . This has implications not only for the diagnosis of neurocysticercosis in low income countries , but also for future epidemiological research .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "medicine", "public", "health", "and", "epidemiology", "neurocysticercosis", "infectious", "disease", "epidemiology", "test", "evaluation", "neglected", "tropical", "diseases", "population", "biology", "infectious", "diseases", "epilepsy", "epidemiology", "biology", "diagno...	2011	A Cross-Sectional Study of People with Epilepsy and Neurocysticercosis in Tanzania: Clinical Characteristics and Diagnostic Approaches
Percutaneous treatment ( PT ) emerged in the mid-1980s as an alternative to surgery for selected cases of abdominal cystic echinococcosis ( CE ) . Despite its efficacy and widespread use , the puncture of echinococcal cysts is still far from being universally accepted . One of the main reasons for this reluctance is the perceived risk of anaphylaxis linked to PTs . To quantify the risk of anaphylactic reactions and lethal anaphylaxis with PT , we systematically searched MEDLINE for publications on PT of CE and reviewed the PT-related complications . After including 124 publications published between 1980 and 2010 , we collected a total number of 5943 PT procedures on 5517 hepatic and non-hepatic echinococcal cysts . Overall , two cases of lethal anaphylaxis and 99 reversible anaphylactic reactions were reported . Lethal anaphylaxis occurred in 0 . 03% of PT procedures , corresponding to 0 . 04% of treated cysts , while reversible allergic reactions complicated 1 . 7% of PTs , corresponding to 1 . 8% of treated echinococcal cysts . Analysis of the literature shows that lethal anaphylaxis related to percutaneous treatment of CE is an extremely rare event and is observed no more frequently than drug-related anaphylactic side effects . Human cystic echinococcosis ( CE ) , caused by the larval stage of the cestode Echinococcus granulosus , is a cosmopolitan parasitic zoonosis , affecting mainly the liver ( ∼70% ) and the lung ( ∼20% ) of the human intermediate host . Clinical symptoms depend on the location , number , and size of the cysts . Until anthelminthic chemotherapy became available ( mebendazole in the 1970s and albendazole in the early 1980s ) , surgery was the only treatment choice . The spectrum of therapeutic options was further extended in the mid-1980s when the increasing availability of modern imaging techniques , namely ultrasound , allowed the introduction of image-guided percutaneous treatment ( PT ) methods . Over the years , various PTs have been developed , based on the classic PAIR ( Puncture of the cyst , Aspiration of the cyst fluid , Injection of a scolicidal agent , and Re-aspiration of the cyst content ) procedure [1] , [2] with minor variations of the essential steps [3]–[5] . Different catheterization techniques allowing aspiration of the solid content of cysts have also been developed for those cyst stages that are often unresponsive to PAIR [6] , [7] . Despite the wide use of PTs in the last two and a half decades , the fear of anaphylactic shock and dissemination due to the spillage of cystic fluid is still quoted by physicians favoring surgery for the treatment of CE [8] , [9] . However , anaphylactic reactions in CE occur not only as a side effect of PT , but also of surgical treatment [10]–[17] , result of accidental trauma [18]–[20] and even spontaneously [21]–[24] . To our knowledge there are no updated figures on the frequency of anaphylactic reactions , anaphylactic shock or lethal anaphylaxis following PTs of echinococcal cysts . To quantify the risk of allergic reactions and lethal anaphylaxis related to PT of echinococcal cysts we systematically reviewed the published literature . We performed a PubMed ( MEDLINE ) search of the literature using the key words “echinococcal cysts” , “hydatid cysts” , “cystic echinococcosis” , “hydatidosis” , “PAIR” , and “percutaneous treatment” and reviewed the available references published between January 1980 and December 2009 for eligible publications ( Figure 1 ) . The inclusion criteria were as follows: All publications on PT of E . granulosus cysts with information about the number of treated cysts , the number of PT procedures and the occurrence of lethal complications were included . When the original article was not obtainable but the abstract containing the requested information was , the publication was included in the analysis . In some cases repeated PTs of the same echinococcal cysts were performed during the course of treatment . In these instances , we collected the total number of treated echinococcal cysts and the corresponding total number of PT procedures . To avoid multiple counting ( duplication ) of identical procedures and cases , follow-up publications on identical procedures and cases were traced and excluded ( references S1 ) . In human CE , the liver is the organ most frequently affected . Therefore , we divided the collected cases and PT procedures according to the anatomical location into “hepatic cysts” ( Table S1 ) and “extra-hepatic cysts” ( Table S2 ) . When the exact anatomical locations of the cysts were not specified , the data was collected separately ( Table 1 ) . Information about the reported complications was collected accordingly , differentiated into “lethal complications” and “reversible complications” and summarized ( Tables 2 , 3 , 4 , 5 , 6 , 7 ) . It was impossible to retrospectively grade the severity of the reversible anaphylactic reactions due to the lack of a standardized definition of the events . If the authors labelled subjective severity levels of the observed anaphylactic reactions ( e . g . “severe” , “moderate” , “mild” , “minor” ) we collected , summarized and listed them accordingly . In addition to the evaluation and quantification of anaphylactic reactions , we also collected and summarized other PT related complications , to allow a representative overview of all PT relevant complications . One hundred-twenty-four publications met our inclusion criteria , with a total number of 5943 PT procedures performed for the diagnosis or treatment of 5517 echinococcal cysts . Ninety-two publications that did not meet the inclusion criteria were excluded from the analysis . Four publications were follow-up publications on identical cases or case series and therefore excluded from analysis . In all but three of the publications included , detailed information about the observed reversible complications were available . In one additional publication , the observed complications were specified but not quantified . These four publications were labeled in the tables accordingly ( Tables S1 , S3 , 5 , 6 , 7 ) . For 863 cysts , information concerning the organ location involved was available , but exact number of cysts for each organ was not . For 17 cysts , information about the location was not available . The publications covering these 880 cysts were labeled in the tables accordingly ( Table 1 ) . A detailed analysis of the observed complications in reference to size , stage , and exact location within the affected organs was impossible due to lack of details in the original publications . Overall , five lethal and 777 reversible complications were collected ( Tables 2 , 6 ) . Of the five lethal complications , three were related to the PT procedure , while two fatalities occurred due to PT “unrelated causes” . Of the three PT related fatalities , two lethal anaphylactic shocks and one fatality “associated with the use of the method” were reported . Unfortunately , detailed information about the two fatalities due to “unrelated causes” [25] , [26] and the fatality reported as “associated with the use of the method” [27] were not obtainable . There were five fatal cases reported in 5943 performed PT procedures . This occurred while treating 5517 echinococcal cysts resulting in an overall fatality rate of 0 . 08% ( 5 in 5943 ) and 0 . 09% ( 5 in 5517 ) respectively ( Table 2 ) . The overall fatality rate due to lethal anaphylaxis is 0 . 03% ( 2 in 5943 ) and 0 . 04% ( 2 in 5517 ) respectively ( Table 2 ) . Reversible complications were reported in 345 out of 3440 PT procedures ( 10% ) for the treatment of 3232 liver echinococcal cysts ( Table 3 ) , in nine out of 175 PT procedures ( 5% ) for the treatment of 142 extra-hepatic echinococcal cysts ( Table 5 ) and in 423 out of 2328 PT procedures ( 18% ) for the treatment of 2143 echinococcal cysts of unspecified anatomical location ( Table 4 ) . In summary , 777 reversible complications were observed in 5943 PT procedures for the treatment of 5517 echinococcal cysts . Therefore , reversible complications were observed in 13% of all PT procedures , corresponding to 14% of all treated echinococcal cysts ( Table 6 ) . The reversible complications fall into three categories: anaphylactic , potentially anaphylactic , and non-anaphylactic: In total , 99 reversible anaphylactic reactions were reported in 5943 PT procedures for the treatment of 5517 echinococcal cysts . Therefore , reversible allergic reactions complicated 1 . 7% of all PT procedures , corresponding to 1 . 8% of all treated echinococcal cysts ( Table 7 ) . The potentially anaphylactic complications include “fever” , “hypotensive reaction” , “vaso-vagal-reaction” , and “nausea and vomiting” . In total , 128 potentially anaphylactic reactions were reported during 5943 PT procedures ( 2 . 1% ) for the treatment of 5517 ( 2 . 3% ) echinococcal cysts ( Table 6 ) . Non-anaphylactic complications – ranging from frequently observed “biliary fistulas” to very rare events such as “active arterial hemorrhage” , “intracystic bleeding” or “gallbladder hemorrhage” – were reported in 550 cases during 5943 PT procedures ( 9 . 3% ) for the treatment of 5517 ( 10% ) echinococcal cysts . Allergic reactions and anaphylaxis are IgE-mediated immediate hypersensitivity reactions that occur when antigen-specific IgE , bound to Fc receptors on mast cells and basophils , are cross linked by the antigen , activating the cells to rapidly release a variety of mediators such as histamine , enzymes and lipid mediators [28] . While anaphylactic reactions and allergic symptoms are usually observed in cases of treatment-related rupture of echinococcal cysts , they may also occur spontaneously . The symptoms vary from mild urticaria to anaphylactic shock [29] . The presence of specific IgE in serum of patients is a well known feature of CE with levels varying according to cyst number , location , morphology , disease severity , and cyst stage [30] , [31] . Despite 75% of CE patients having detectable levels of specific IgE and histamine release by circulating basophils in response to E . granulosus , antigens can be detected in 100% of patients [32] . Consequently , allergic reactions are rare and unpredictable . So far , the predictive value of IgE titers ( or of IgG4 titers , considered “anti-anaphylactic” isotypes ) for the development of allergic reactions has not been investigated . Echinococcus allergens have mainly been studied with the aim of improving the performance of diagnostic tests . Three conserved proteins have been identified ( EgEF-1β/δ , EA21 and Eg2HSP70 ) , by screening of an E . granulosus cDNA library with IgE from patients with and without cutaneous allergic manifestations showing significantly different IgE-binding reactivity between groups [33] , [34] , [35] . Nevertheless , the identification of such reactivity by a patient's IgE as a predictive factor for the development of anaphylaxis has never been investigated . Another appealing , still unexplored possibility , is the use of these allergens for desensitization therapy . The control of CE-related allergic reactions relies on the administration of vasoactive agents ( e . g . epinephrine ) and corticosteroids . Although a study reported less severe hemodynamic alterations in surgical patients pre-treated with histamine H1 plus H2 receptor blockers [36] , the usefulness of any pre-operative treatment for the prevention of anaphylaxis has never been demonstrated . The pathogenesis of anaphylactic reactions in CE is still unclear but commonly explained by the disruption of the integrity of the cyst wall with spillage and translocation of allergenic cyst contents into the host's circulation . Despite this , rupture of echinococcal cysts does not always or necessarily lead to anaphylactic reactions . In a series of 24 patients with proven rupture of echinococcal cysts ( 12 patients with liver cysts and 12 patients with lung cysts ) only four patients ( 16 . 7% ) had symptoms or a history of allergic reactions [37] . The same observation has been made during surgery of echinococcal cysts , were apparent spillage of cyst fluid – despite all precautions taken – is reported to occur in 5% to 10% of cases , without this necessarily leading to anaphylaxis [38] . In our review , we found an incidence of three anaphylactic fatalities per 10 , 000 PT procedures ( 0 . 03% ) . To put this figure in perspective , one may consider other conditions where treatment entails the risks of lethal anaphylaxis: Overall , we found a frequency of 1 . 67 reversible anaphylactic reactions per 100 PT procedures of echinococcal cysts ( 1 . 67% ) ( Table 7 ) . The majority of these reversible anaphylactic reactions were allergic skin reactions ( urticaria , rash , pruritus ) , reported in 1 . 1 per 100 PT procedures ( 1 . 1% ) ( Table 7 ) . Again , to put these figures in perspective , we consulted the literature on drug-related allergic skin reactions: in a large surveillance program on drug-induced allergic cutaneous reactions – including 15 , 238 consecutive inpatients – Bigby et al . found antibiotics to be associated with the highest risk . Among the 51 drugs studied , allergic cutaneous reactions were observed in 1 . 8% to 5% of all treated patients ( penicillin G: 1 . 8% , erythromycin: 2% , semisynthetic penicillins: 2 . 1% , cephalosporins: 2 . 1% , ampicillin: 3 . 3% , trimethoprim-sulfamethoxazole: 3 . 4% , amoxicillin: 5% ) . In the same study , 2 . 2% of patients receiving blood products presented allergic cutaneous reactions [43] . One problem with allergic reactions from the puncture or surgery of echinococcal cysts is that the exact pathophysiological cause and correlation with consecutive symptoms remains unclear . Some peri-interventional complications reported as “fever” ( 111 cases ) , “hypotensive reaction” ( 15 cases ) , “vaso-vagal-reaction” ( 1 case ) , and “nausea and vomiting” ( 1 case ) ( Table 6 ) might represent allergic reactions . If this were to be the case , the risk of reversible allergic reactions might be as high as 3 . 8 per 100 PT procedures ( 3 . 8% ) . Even though some of these cases might represent anaphylactic reactions , it can be assumed that most of the “fever” events ( 111 of the 128 potentially anaphylactic reactions ) are due to infections , as post-interventional “cavity infections” and “abscesses” account for 60 of the total 550 non-allergic reversible complications ( Table 6 ) . Additionally , the concept of anaphylaxis awaits a stricter definition , as there is no consensus on exactly how to define it along with considerable disagreement about its prevalence , diagnosis and management [44] , [45] . The retrospective evaluation of publications on PT related complications is certainly limited by a number of factors such as non-uniform definitions of anaphylactic events , the merging of data from different kind of studies – covering different PT methods in different settings and dealing with a different composition of clinical cases – and the denominator issue . Due to the retrospective nature of our review and because we can only analyze published data , a publication bias can also be at work . It can be argued that severe events ( e . g . severe anaphylaxis ) might be more likely be be published . But one could counter that events assumed to be common ( especially the often quoted PT related anaphylaxis ) might not as readily be published . Nevertheless , we consider the analysis of the existing published literature a justified approach as no other source of more accurate data is currently available . Future work in this area is needed to investigate the pathophysiology of anaphylactic reactions in CE and to prospectively study the potential relationship between clinical variables such as location , number , size , stage of the cyst , and risk of anaphylactic reactions . While large , well-designed clinical trials are needed to develop treatment algorithms stratified by cyst stage and available level of health care resources , the analysis of the available literature shows that the traditional fear of lethal anaphylaxis and allergic reactions in PT of echinococcal cysts has been exaggerated by the critics of PT . Provided adequate stand-by resuscitation measures are available , each time an echinococcal cyst is punctured , fear of anaphylactic shock is no longer justified as an argument to avoid this therapeutic option . A necessary evolution in the clinical management of CE will be the comparative evaluation of different PT and surgical methods in certain situations . While surgery legitimately maintains a central role in complicated cysts ( rupture , biliary fistulas , compression of vital structures , bacterial superinfection , haemorrhage ) , cysts at high risk of rupture , or large cysts with many daughter vesicles , that are not suitable for percutaneous treatment approaches , PT has shown to be a safe and effective alternative for many patients with suitable cysts . What is needed now are evidence-based criteria to allocate the patient to the most appropriate treatment option according to the specific situation .	The risk of anaphylactic shock is the objection most often raised by opponents of percutaneous treatments for cystic echinococcosis , but there are no updated figures on the actual occurrence of anaphylaxis as a complication of this treatment . To assess the number of lethal and non-lethal anaphylactic reactions following percutaneous aspiration of echinococcal cysts , we systematically reviewed the literature published from 1980–2010 . The analysis of the available literature shows that the risk of severe anaphylactic reactions resulting from percutaneous treatment of echinococcal cysts has been widely exaggerated and the actual risk may be lower than that following administration of certain antibiotics . Provided adequate stand-by resuscitation measures are available , each time an echinococcal cyst is punctured , fear of anaphylactic shock is no longer justified as an argument to avoid this therapeutic option .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "medicine", "echinococcosis", "infectious", "diseases", "neglected", "tropical", "diseases" ]	2011	Justified Concern or Exaggerated Fear: The Risk of Anaphylaxis in Percutaneous Treatment of Cystic Echinococcosis—A Systematic Literature Review
The systemic immune response of Drosophila is known to be induced both by septic injury and by oral infection with certain bacteria , and is characterized by the secretion of antimicrobial peptides ( AMPs ) into the haemolymph . To investigate other possible routes of bacterial infection , we deposited Erwinia carotovora ( Ecc15 ) on various sites of the cuticle and monitored the immune response via expression of the AMP gene Diptericin . A strong response was observed to deposition on the genital plate of males ( up to 20% of a septic injury response ) , but not females . We show that the principal response to genital infection is systemic , but that some AMPs , particularly Defensin , are induced locally in the genital tract . At late time points we detected bacteria in the haemolymph of immune deficient RelishE20 flies , indicating that the genital plate can be a route of entry for pathogens , and that the immune response protects flies against the progression of genital infection . The protective role of the immune response is further illustrated by our observation that RelishE20 flies exhibit significant lethality in response to genital Ecc15 infections . We next show that a systemic immune response can be induced by deposition of the bacterial elicitor peptidoglycan ( PGN ) , or its terminal monomer tracheal cytotoxin ( TCT ) , on the genital plate . This immune response is downregulated by PGRP-LB and Pirk , known regulators of the Imd pathway , and can be suppressed by the overexpression of PGRP-LB in the haemolymph compartment . Finally , we provide strong evidence that TCT can activate a systemic response by crossing epithelia , by showing that radiolabelled TCT deposited on the genital plate can subsequently be detected in the haemolymph . Genital infection is thus an intriguing new model for studying the systemic immune response to local epithelial infections and a potential route of entry for naturally occurring pathogens of Drosophila . Antimicrobial peptides ( AMPs ) are key components of innate immunity . They are ubiquitous throughout the animal and plant kingdoms , reflecting the importance of these molecules in host defence . In Drosophila over 20 AMPs , divided into seven classes , have been described [1] . These Insect AMPs are thought to be active against microbial membranes and exhibit specificity of activity against Gram-negative bacteria ( e . g . Diptericin ) , Gram-positive bacteria ( e . g . Defensin ) or fungi ( e . g . Drosomycin ) . Drosophila AMPs are induced in the fat body , an analogue of the mammalian liver , in response to both septic injury and oral infection with certain bacteria . This response is referred to as the systemic response because it leads to secretion of AMPs into the haemolymph , which bathes all tissues . The systemic response has been well characterised and is regulated at the transcriptional level by the Toll and Imd pathways [2] , [3] . The Toll pathway is induced by both Gram-positive bacteria and fungi , via circulating pattern recognition receptors , and leads to the activation of NF-κB proteins ( Dif and Dorsal ) , controlling the production of AMPs active against fungi . In contrast , the Imd pathway mainly responds to Gram-negative bacterial infection and controls antibacterial peptide genes via the activation of the NF-κB-like protein Relish . PGRP-LC , one of several Peptidoglycan Recognition Proteins in Drosophila , acts as a membrane-bound pattern recognition receptor that activates the Imd pathway upon sensing of DAP-type Peptidoglycan ( PGN ) [4] , [5] , [6] . PGN is a cross-linked polymer which is an essential component of all bacterial cell walls . DAP-type PGN , which contains a diaminopimelic acid ( DAP ) moiety , is limited to Gram-negative bacteria and some Gram-positive bacilli , while the Lys-type PGN of most Gram-positive bacteria contains Lysine instead of DAP . Drosophila AMP genes are also expressed in barrier epithelia such as the epidermis , reproductive system , respiratory tract and digestive tract , which are exposed to environmental microorganisms and/or indigenous flora [7] , [8] , [9] . This AMP synthesis is referred to as the local immune response , as opposed to the systemic response . It is important , in epithelia , to distinguish between constitutive and inducible AMP gene expression . Some AMP genes are expressed constitutively , in specific tissues , in the absence of infection . This is the case for Drosomycin in salivary glands and the female spermatheca [7] , for Drosocin in the ovaries [10] and for Cecropin in the male ejaculatory duct [11] . This constitutive expression is not regulated by NF-κB pathways but by various tissue-specific transcription factors such as the homeobox-containing protein Caudal [12] . By contrast , the inducible local AMP gene expression is triggered upon infection by Gram-negative bacteria and is mediated by the Imd pathway [8] , [13] , [14] . For example , Drosomycin and Diptericin are induced in both trachea and gut in response to local infections by bacteria such as Erwinia carotovora . Like the systemic response , the local immune response is based on the recognition of Gram-negative PGN by PGRP-LC [15] . In addition to the immune response to infection , AMP expression has also been shown to be stimulated by mating , in females [16] . Surprisingly , this induction appears to be independent of any microbial elicitor . Indeed , experimental data support the idea that accessory gland proteins , including Sex Peptide , present in the sperm of males , are sufficient to induce a local AMP gene expression in the female genitalia . This immune response may limit the propagation of potential infectious agents just after copulation . This may be particularly important following female genital wounding , which has been shown to occur during copulation in D . melanogaster [17] . These recent observations have shown that the genes encoding AMPs are expressed in more sophisticated and differentiated patterns than were previously anticipated [9] . The expression of AMP genes appears to be specifically regulated both locally and systemically . In addition to some constitutive expression , they can be induced upon infection and upon mating . The starting point of this study was to investigate additional possible routes of infection and analyse whether Insects could sense the presence of bacteria through their cuticle , leading to expression of AMP genes . Here , we show that deposition of bacteria on the genital plate of males is sufficient to activate both systemic and local immune responses . Our results strongly suggest that the translocation of small fragments of PGN into the haemolymph triggers a protective systemic response upon genital infection . We wondered if flies are able to sense the presence of potentially infectious microorganisms in contact with their external cuticle . Entomopathogenic fungi have been shown to be able to cross the cuticle barrier and thereby activate an immune response [18] , but to date bacteria have only been shown to activate an immune response by direct introduction into the haemolymph ( septic injury ) or by oral infection . To address this question , we deposited Gram-negative bacteria on various sites of the cuticle of adult males and monitored the systemic immune response 6h later , by the expression of the AMP gene Diptericin in whole flies . Deposition of a droplet of the Gram-negative bacteria Erwinia carotovora carotovora 15 ( Ecc15 , OD600 = 200 ) on the head , thorax or abdomen , with a paintbrush , results in modest induction of Diptericin corresponding to less than 6% of the expression level observed after septic injury . In contrast , we observed that deposition of bacteria on the genital plate ( see Figure 1A ) reproducibly results in a substantial level of Diptericin expression , corresponding to 10 to 20% of the level obtained after septic injury ( Figure 2A , B ) . For all further experiments , genital infection ( GI ) was performed by applying 20–50nL of bacterial pellet using a 200µl pipette tip ( see Figure 1B ) depositing a droplet without touching the genital plate itself . A kinetic analysis reveals that Diptericin expression is already detectable at 1 . 5h post GI , reaches its maximum between 6 and 18h and decreases thereafter ( Figure 2B ) . RT-qPCR analysis also indicates that GI triggers the expression of all the families of AMPs known to be induced in the systemic immune response ( Figure 2C and Figure S1 ) . Of all the AMP genes , Defensin exhibited the highest level of expression upon GI , reaching levels comparable to those observed in response to a systemic infection . Importantly , Diptericin expression in females , following an equivalent deposition of bacteria on the abdominal terminalia , was very low , with a maximum of 5% of the level observed after septic injury , indicating that the observed immune reactivity is specific to the male genitalia ( data not shown ) . Since mating can induce wounding of the genital plate ( at least in females [17] ) , we compared the response of mated to unmated males and found that the response was similar or slightly higher in unmated males ( Figure S2 ) . All subsequent experiments were performed with males from stocks , which were likely mated . The genital plate is distinct from the other cuticle sites on which bacteria were deposited in that it includes openings to both the gut and genital tract . It should be noted however that other deposition sites included spiracles , opening into the trachea , and the proboscis , opening into the foregut . Thus the presence of openings into epithelial organs is not in itself sufficient to allow a systemic response to bacterial deposition on the cuticle . Systemic expression of AMP genes is regulated by both the Toll and Imd pathways , which are activated by different classes of microbes in Drosophila . We investigated the range of bacteria to which a response to GI occurs and observed that AMP genes are induced only upon GI by Gram-negative but not by Gram-positive bacteria ( Figure 2D and S1 ) . This suggests that the Imd pathway mediates the response to GI and , indeed , Diptericin gene expression upon GI was abolished in RelishE20 mutant flies that lack a functional Imd pathway ( Figure 2E ) . Our results indicate that male flies react strongly to the presence of bacteria deposited on the genital plate . The genital plate is a complex cuticular structure that provides many sites where bacteria could accumulate ( see Figure 1A ) . Furthermore , since it includes both the anal and genital openings , it could provide a route of entry for potential pathogens into the digestive and/or genital tracts . To test if the genital plate is indeed a site of entry for bacteria into organs and into the haemolymph , we investigated the fate of living bacteria during GI . We monitored the fate of GFP expressing Ecc15 ( Ecc15-GFP ) deposited on the genital plate and observed that , after 9h , GFP was still present at the infection site in 90% of the flies . In contrast , only 50% of the flies still show GFP signal when bacteria were deposited on the side of the abdomen ( data not shown ) . This suggests that the genitalia offer a better niche for the persistence of bacteria than the cuticle in general . We further observed that , after 24h , some flies which had had bacteria deposited on their abdomen exhibited bacteria on the genitalia , suggesting that bacteria washed from the body are able to infect the genitalia . Dissection of the genital tract of genitally infected flies revealed small quantities of fluorescent bacteria within the ejaculatory bulb and duct although the majority appears to remain on the external genitalia ( data not shown ) . We did not detect any difference between Ecc15-GFP persistence on the outside of the genitalia in RelishE20 and wild-type flies . However , by dissecting flies 24h after GI , we observed fluorescent bacteria in the body cavity of 19% of RelishE20 flies ( 5/27 ) , while no bacteria were found inside wild-type flies ( n = 20 ) . In order to verify the presence of live Ecc15-GFP we extracted haemolymph from genitally infected flies and plated the extracts , under Rifampicin selection for Ecc15 . Live Ecc15-GFP were present in the haemolymph of 20% of RelishE20 flies ( 4/20 ) but none were present in the haemolymph of wild-type flies ( n = 20 ) . Although a systemic immune response is already detected at 1 . 5h after GI , no bacteria were detected in the haemolymph at this time point , in either wild-type or RelishE20 flies , by direct observation of fluorescence . We cannot exclude however that small numbers of bacteria , below our limit of detection , were present . Consistent with the presence of bacteria in their body cavity , we observed a weak but significant mortality after GI with Ecc15 in RelishE20 flies ( Figure 3A ) . We also observed mortality in both wild-type and , more strongly , RelishE20 flies after GI with a strain of Pseudomonas aeruginosa ( CFBP2466 ) , further suggesting that the genital plate provides a favourable route of infection in Drosophila ( Figure 3B ) . These experiments demonstrate a role of the Imd pathway in controlling the progression of genital infections , preventing bacteria from entering the haemolymph . Two bacterial strains , Ecc15 in larvae and Pseudomonas entomophila in both larvae and adults , have been shown to be able to activate both local and systemic responses upon oral infection [13] , [19] . To determine which tissues express AMPs upon GI with Ecc15 , we used RT-qPCR to compare the expression levels of Diptericin in whole flies to those in dissected genital tracts , fat bodies ( carcasses ) and guts , 6h after infection . Figure 4A shows that Diptericin is most strongly expressed , upon GI , in the fat body . Using a Diptericin-lacZ reporter gene , we observed a strong fat body expression in about 20% of the males after GI ( Figure 4B ) . Thus the Diptericin expression observed in whole flies is mostly due to a strong systemic induction in only 20% of the infected individuals . By contrast , Diptericin expression was induced in the fat body of 100% of flies subjected to septic injury . In order to visualise the local immune response , we studied Diptericin-lacZ expression in the gut and genital tract . In contrast to oral infections , relatively little induction was observed in the gut ( Figure 4C ) . In the genital tract , we monitored Diptericin-lacZ expression upon GI in Galn1 flies lacking the endogenous ß-galactosidase gene , as endogenous ß-galactosidase is expressed in the ejaculatory bulb . We observed a weak induction of the transgene in the seminal vesicle , the ejaculatory duct and ejaculatory bulb ( Figure 4D ) . By qRT-PCR we observed that this induction is dependant on the Imd pathway ( Figure S3 ) . We extended this analysis to other AMPs , using qRT-PCR and AMP-GFP reporter genes . Interestingly , we observed that GI specifically triggers a very strong induction of Defensin in the genital tract of males ( Figure 4A ) and that this induction is dependent upon the Imd pathway ( Figure S3 ) . Observation of flies expressing the Defensin-GFP reporter gene reveals that this expression is localised to the thin part of the ejaculatory duct ( Figure 4E ) . We also confirmed previous observations that a Cecropin A-GFP reporter gene is strongly expressed in the ejaculatory duct in the absence of infection [8] , [11] ( data not shown ) , and expression analysis of the endogenous gene showed a slight induction upon GI ( Figure 4A ) . We conclude that deposition of bacteria on the genital plate is sufficient to activate both a systemic and a local immune response in Drosophila males . Most of the Diptericin expression observed in these flies corresponds to fat body expression . This indicates the existence of a link between genitalia and the fat body immune response . The observation that deposition of bacteria on the genital plate induces a systemic immune response could be explained either by the invasion of haemolymph by bacteria or by the existence of an immune signal between genitalia and the fat body . Recently , it has been proposed that PGN translocation through the gut could be a mechanism that activates the systemic immune response upon oral bacterial infection with P . entomophila or Ecc15 [15] . Our observation that systemic Diptericin expression was detectable before bacteria were observed in the haemolymph of even immune deficient flies , suggests that a similar mechanism explains the systemic response to GI . To distinguish between direct sensing of bacteria in the haemolymph and remote sensing of bacteria present in the genital tract and/or hindgut , we compared the systemic immune response , as monitored by the level of Diptericin expression , upon GI with living or dead bacteria . Figure 5A shows that dead bacteria activate a systemic immune response with a similar , or higher , efficiency to live bacteria . This demonstrates that the induction of a systemic response is not mediated by the crossing of the epithelial barriers by live bacteria but rather results from the release of bacterial elicitors . Monomeric or polymeric DAP-type PGNs of Gram-negative bacteria are the most potent inducers of the Imd pathway . This prompted us to analyze the effect of PGN deposition on the genital plate . As shown in Figure 5B , deposition of highly purified PGN from Ecc15 activates a systemic immune response to the same , or greater , extent as live bacteria . Comparing the responses to PGN from different bacteria we found that DAP-type PGNs of Gram-negative Ecc15 or P . entomophila induce a strong response , whilst PGNs from Gram-positive bacteria , either DAP-type from B . subtilis or Lys-type from M . luteus , activate little or no response ( Figure 5C ) . The GlcNAc-MurNAc ( anhydro ) -L-Ala-γ-D-Glu-meso-DAP-D-Ala monomer , also known as tracheal cytotoxin ( TCT ) , was previously identified as the minimum PGN motif capable of efficiently inducing the Imd pathway [5] , [6] . TCT provides an ideal “signature” of Gram-negative bacteria since this muropeptide is continuously released during cell growth and division as a result of PGN recycling [20] . We observed that GI with TCT strongly induces AMP genes ( Figures 2B and C ) . It has been proposed that TCT released by cleavage of the PGN polymer can cross epithelia to activate an immune response . If this were the case , then TCT should induce a stronger systemic immune response than polymeric PGN . To compare the immune response to TCT or polymeric PGN , we deposited around 30nL of HPLC-purified TCT or PGN on the genital plate at ( monomer equivalent ) concentrations respectively of 1mmol . L−1 and 5mmol . L−1 . Injecting serial dilutions of TCT or PGN , starting at these concentrations , the immune response to PGN is always the same or higher than that to TCT ( Figure S4 ) . By contrast , the response to genital deposition of 1mmol . L−1 TCT was particularly strong , three times the response to genital deposition of 5mmol . L−1 PGN at the 6h time point ( Figure 5B ) . These results show that the systemic activation of AMP genes in response to GI can be mediated through the sensing of TCT or small fragments of PGN and does not require the presence of live bacteria . They further suggest that this sensing takes place in the haemolymph . Recent studies in Drosophila have revealed that multiple levels of regulation limit Imd pathway activity and prevent excessive or prolonged immune activation [3] . A key role in bacterial tolerance of the gut has been attributed to the amidase PGRPs , PGRP-SC1 and PGRP-LB , as they are proposed to scavenge PGN released by gut microbes [15] , [21] . An additional negative regulator , Pirk ( also named Pims or Rudra ) , has been recently identified as an Imd immune responsive factor which removes the receptor PGRP-LC from the membrane , thereby shutting down Imd signalling [22] , [23] , [24] . We investigated the roles of PGRP-LB , Pirk and PGRP-LC in the immune response upon GI . As shown in Figure 6A , Diptericin was induced , upon deposition of TCT on the genitalia , at a higher level in flies subjected to ubiquitous PGRP-LB RNAi ( with a da-Gal4 driver ) than in control flies . Furthermore , ubiquitous overexpression of PGRP-LB suppresses Diptericin induction by TCT . Specific overexpression of PGRP-LB in the fat body and haemocytes ( with a c564-Gal4 driver ) also suppressed the induction of Diptericin by TCT ( Figure 6B ) . This result demonstrates that the presence of the amidase PGRP-LB in the haemolymph compartment blocks the systemic immune response to GI , presumably by degrading fragments of PGN entering the haemolymph . Interestingly , RNAi of PGRP-LB in the fat body ( driven by c564-Gal4 ) has a less reproducible effect than ubiquitous RNAi of PGRP-LB ( Figure 6A , B ) , suggesting that PGRP-LB normally plays a more important role outside the haemolymph , in limiting the availability of PGN fragments capable of entering the haemolymph . However , the lack of a Gal4 driver expressed specifically in the male genital tract prevented us from testing whether it is in this tissue that PGRP-LB is required to limit the systemic response to GI . In keeping with the model of TCT entering the haemolymph , we observed that selective depletion of PGRP-LC in the fat body blocks the systemic immune response to TCT ( Figure 6B ) . Similarly , PGRP-LCE12 mutant flies show no systemic response to GI with Ecc15 ( Figure 6C ) . Finally , we observed a higher immune response to GI in pirk deficient flies , the level of Diptericin expression being 2 . 5 fold higher than in wild-type flies ( Figure 6C ) . We conclude that the activity of the Imd pathway is required in the haemolymph for a systemic response to genital infection and that this is blocked by the presence of a PGN degrading enzyme . As was shown for the gut immune response , the immune response to GI is tightly regulated , being limited by the Imd pathway modulators PGRP-LB and Pirk . To strengthen the model that translocation of TCT into the haemolymph is responsible for the systemic immune response to genital deposition of PGN or TCT , we analysed whether traces of TCT could be detected in the haemolymph following its deposition on the genital plate . To this end , we deposited [14C]-radiolabelled TCT on the genital plate of males and tested for the presence of radioactivity in the haemolymph . Around 30nL of TCT was deposited on the genitalia of 100 flies and haemolymph samples were carefully collected 2h later ( shortly after the systemic immune response is first detected ) and deposited on a filter paper . Figure 5D shows that a radioactive signal could be detected in the haemolymph and Figure 5E indicates that about 1/1000th of the radioactivity deposited on the genitalia was recovered in the haemolymph 2h later . This is consistent with the observation that injection of 1µmol . L−1 TCT provokes a similar level of systemic immune response to genital deposition of 1mmol . L−1 TCT ( Figure S4 ) . RT-qPCR analysis revealed that Diptericin was expressed to similar levels in flies challenged with radioactive or non-radioactive TCT ( data not shown ) . Since the [14C]-TCT used was radiolabelled at the level of the DAP , this experiment shows that [14C]-TCT or fragments of TCT containing the DAP are indeed translocated from the genital plate to the haemolymph . In this paper , we show that the deposition of bacteria on genitalia is sufficient to trigger both local and systemic expression of AMP genes in Drosophila males . Of all the sites where bacteria were deposited , the immune response was the greatest in the case of the genitalia . This demonstrates an unexpected immune reactivity to the presence of bacteria on the genitalia , that probably relates to the possibility of pathogen entry via this route . Exploiting this new mode of infection , we also provide strong evidence that a systemic immune response can be activated at a distance by the presence of bacteria on the genitalia , through the diffusion of small fragments of PGN . Thus our study not only provides information on the mechanisms used to prevent microbial infection of the genital tract , but also describes a novel mode of infection , via the genitalia , which offers new opportunities for the study of the long-range activation of Drosophila immunity . Although the integument constitutes a formidable barrier with the outside world , there are weak points in the surface that parasites and pathogens might be expected to target , such as the gut , trachea and reproductive organs [25] . Thus the genital plate , with openings to the gut and genital tract , seems a likely route of pathogen entry [26] . The convolutions of the cuticular structures of the genital plate may facilitate the persistence of bacteria at this site , reducing the efficiency of cleaning by leg brushing , a typical grooming behaviour of the fly . This is corroborated by our observation that bacteria deposited on this structure persist longer than those deposited elsewhere on the cuticle . We observed that deposition of Gram-negative bacteria on the genital plate is sufficient to induce AMP gene expression locally in both the genital tract and hindgut . Due to the proximity between anal and genital openings , it was not possible to distinguish whether GI results in infection of the genital tract , anus or both . However , since we observe that deposition of bacteria triggers stronger expression of AMP genes in the genital tract than in the gut , the genital tract could well be the main site of infection . Our study reveals for the first time that Defensin , an antibacterial peptide with activity against Gram-positive bacteria , is strongly induced in the genital tract upon GI with a Gram-negative bacterium , under control of the Imd pathway . Expression of the analogous vertebrate Defensins has also been reported in the reproductive tract and seminal fluid of male Rats , Mice and Humans , where it is proposed to play a role in the protection of germ cells [27] . Whereas we observed no significant immune response to bacterial deposition on the genitalia of female Drosophila , mating has been shown to stimulate the expression of AMP genes in females , as a result of accessory gland proteins present in the male seminal fluid [16] . These results support the theory that mating is a significant cause of infection for Drosophila in the wild [26] . In Insects generally , a wide range of infections have been shown to be transmitted during copulation [28] and potential pathogenic bacteria have been reported on male sexual organs [29] . In Drosophila , transfer of the entomopathogenic bacterium Serratia marcescens from contaminated males to females during courtship and mating has been observed in an experimental setting [30] . Our observations showing a high immune response to the presence of bacteria on the male genitalia suggest that the genital tract provides an entry route for bacteria in both sexes . It is intriguing that the two sexes seem to show immune responses to different genital stimuli . Immune differences between the sexes are commonly observed in Insects [31] , [32] and it has been suggested that in some cases this is an efficiency measure since infections are more reliably coupled to mating in females than males [26] . This is taken to the extreme in the bedbug , Cimex lectularius , where the female has a specific immune organ developed in response to traumatic insemination [26] . Although traumatic insemination has not been reported in Drosophila melanogaster , the male is known to inflict genital wounds during mating [17] , thus infection may be reliably coupled to mating , leading to the evolution of an immune response to copulation . That males show a direct response to bacteria on the genitalia , rather than to copulation , could be explained by a reduced risk of infection associated with mating in males and an advantage to maintaining clean genitalia , thereby avoiding transmission of bacteria to mated females . To date , two methods , septic injury and oral ingestion of certain bacterial strains , have been shown to activate systemic expression of AMP genes in Drosophila in response to bacterial infection . In this study , we report a third mode of activation of the systemic immune response by showing that deposition of bacteria on the male genital plate results in an immune response in the fat body , without the presence of bacteria in the haemolymph . This shows that the genitalia allow the passage of ‘early-warning’ signals that recruit and activate haemolymph-based effector systems ( see below ) . As has been observed for oral infections [8] , [13] , the immune response to GI is mediated by the Imd pathway and activated by DAP-type PGN of Gram-negative bacteria . The observation that DAP-type PGN of the Gram-positive B . subtilis is a less potent inducer might be explained by the absence of TCT and/or the high proportion of amidated DAP in Bacillus PGN . Interestingly , systemic Diptericin expression was observed in only 20% of GI treated males , all of these exhibiting a strong fat body expression . This suggests the existence of a threshold , with an all or nothing response . This clearly contrasts with the uniform immune response following septic injury . This threshold response is probably linked to the existence of multiple negative feedback controls such as PGRP-LB and Pirk , that regulate the Imd pathway and prevent its activation . We can further speculate that the threshold response depends upon the ability of bacteria to colonise the genital tract . An important barrier role of the immune response to genital infection is demonstrated by our observations that the haemocoel of RelishE20 , but not wild-type , flies can be colonised by genitally deposited Ecc15 and that RelishE20 flies exhibit significantly higher lethality than wild-type flies upon GI with a strain of the opportunistic bacterium P . aeruginosa . Although much progress has been made in our understanding of the regulation of innate immunity , it remains to be determined how microbe-derived molecules activate the immune system under physiological conditions . In other systems small muropeptide fragments of PGN , notably TCT , have been shown to act as diffusible signalling molecules . For example , TCT released by the symbiont Vibrio fisheri is involved in the differentiation of the light organ of the squid Euprymna scolopes [33] . Small muropeptides were also identified as products of the intestinal flora that induce the genesis of lymphoid follicles in the gut of Mice , through activation of the Nod1 pathway [34] . Although the intracellular receptors Nod1 and Nod2 are known to be activated by small PGN fragments , how these ligands reach the intracellular compartment is unclear . Some progress has been made recently , with the identification of a transmembrane receptor , Pept1 , as a possible transporter of the Nod2 ligand , muramyl dipeptide [35] . Alternately , another recent study reports that the clathrin and dynamin dependent endocytosis pathway is a key component in the activation of the Nod2 pathway [36] . We have previously suggested a model whereby long-range activation of the systemic immune response in Drosophila is mediated by the translocation of small PGN fragments from the gut lumen to the haemolymph [15] . This view was supported by the observation that ingestion of monomeric PGN can stimulate a strong systemic immune response in PGRP-LB RNAi flies , which have reduced amidase activity and are unable to degrade PGN to its non-immunogenic form . Our present study of GI further supports this model of long-range activation of the immune system by PGN translocation . Deposition of PGN or TCT on the genitalia is sufficient to induce a systemic immune response and , in agreement with a model that involves transport of PGN , we observed that TCT activates a stronger systemic immune response than polymeric PGN . That the systemic immune response to GI requires PGRP-LC expression in the fat body , and is suppressed by amidase activity in the haemolymph , suggests that this response is mediated by PGN fragments present in the haemolymph . Finally , GI with radiolabelled TCT has allowed us to show for the first time that TCT ( or a fragment thereof ) can be found in the haemolymph shortly after genital deposition and that its presence correlates with the activation of an immune response . It seems likely that the detected molecules are intact TCT , since fragments of TCT are insufficient for immune activation via PGRP-LC [5] . These results demonstrate that TCT is indeed translocated into the haemolymph , although do not exclude that other signalling mechanisms contribute to the systemic immune response to local infections . Although TCT has been shown to be a diffusible signalling molecule in other systems , few studies have addressed the extent to which the gut and genital epithelial barriers are permeable in Drosophila . Interestingly , some of the male-derived Accessory Gland Proteins , transferred to the female reproductive tract during mating , access presumptive target tissues bathed by the haemolymph [37] . Further , some have been shown to directly traverse the female reproductive tract and enter the haemolymph by crossing the ventral intima of the posterior vaginal wall [38] . In spite of this apparent permeability , PGN applied to the genitalia of females does not activate an immune response , underlining the difference between the genital tracts of male and female Drosophila . It remains possible that PGN fragments are translocated into the haemolymph by endocytosis or by specific transporters as has been suggested for ligands of the Nod2 pathway . Future work should decipher the mechanism , passive or active , of PGN translocation as well as the site of translocation . By contrast to oral infection , which depends upon the complex physiological regulation of feeding , GI is easy to control , since bacteria or PGN are applied directly to the genitalia . Thus GI provides a simple and reproducible mode of infection for further study of the long-range activation of the systemic immune response . OregonR flies were used as wild-type controls unless otherwise indicated . RelishE20 ( e+ , RelE20 ) , PGRP-LCE12 and pirk ( pimsEY00723 ) are described elsewhere [23] , [39] , [40] . Galn1 is a null mutation in the ß-galactosidase gene of Drosophila [41] . The Gal4 lines used in this study specifically express GAL4 constitutively ( da-Gal4 ) or in adult fat body and haemocytes ( c564-Gal4 ) [42] , [43] . Defensin-GFP , Cecropin A-GFP and Diptericin-lacZ strains were described previously [44] . The UAS-PGRP-LB-IR ( insertion R1 and R3 ) and UAS-PGRP-LC-IR1 inverted repeat RNAi stocks were generated by Ryu Ueda and are described elsewhere [15] . A full-length cDNA of PGRP-LB ( using the CG31217_cDNA gold LD43740 from DGRC ) was placed downstream of the UAS sequence using the pUASt vector . The UAS-PGRP-LB-YFP transgene was obtained by a fusion of YFP to the C-terminus of PGRP-LB inserted in the pDONR221 Gateway entry clone ( Invitrogen ) and finally subcloned in the pTWG transgenesis vector . RNAi and overexpression experiments were controlled by crossing the Gal4 drivers to w1118 , the strain in which the UAS construct insertions were generated . F1 progeny carrying both the UAS construct and the Gal4 driver were transferred to 29°C at late pupal stage for optimal efficiency of the UAS/Gal4 system . Drosophila stocks and crosses were maintained at 25°C in yeasted tubes containing corn-meal fly medium [44] . All bacteria were stored as frozen stocks ( 15% DMSO ) and cultured on LB-Agar plates and in LB medium . Ecc15 and Ecc15-GFP strains are Rifampycin resistant and were described previously [13] . They were grown overnight at 29°C ( without selection ) and used as pellets of OD600 = 200 and 250 respectively [44] . The Pseudomonas aeruginosa strain 2466 ( Collection Française de Bactéries Phytopathogènes , INRA , Angers , France ) was used for survival experiments . It was grown at 37°C overnight and used as pellets of OD600 = 150–200 . Additional bacteria , Bacillus subtilis , Pseudomonas entomophila [19] and Micrococcus luteus , were grown at 29°C and used as pellets of OD600 = 200 . Pellets were not washed prior to use . Dead bacteria were produced by heating for 5 minutes at 95oC . Septic injuries were performed by pricking adults in the thorax with a thin needle dipped into a concentrated bacterial pellet [44] . For Figure 2A , bacterial depositions were performed using a paintbrush dipped into a bacterial pellet . For all other experiments , GI was performed by touching a 200µL pipette-tip containing 10µL of bacterial pellet or 5µL of PGN/TCT to the tip of the abdomen , thereby depositing a small droplet ( 20–50nL ) covering the whole genital plate ( see Figure 1B ) . Infected males were subsequently maintained at 29°C in tubes without yeast in the absence of females . In the case of the experiment using [14C]-TCT , each fly was maintained in a separate vial to prevent cross-contamination . For each bacterial strain , four independent survival experiments were performed with at least 20 flies per genotype , in replicates of 10–20 flies . Survival was scored every 24h . AMP and rpL32 mRNA quantification by RT-qPCR was performed as described [44] . All expression data are given as a ratio of the expression level of the invariant mRNA rpL32 . Each experiment was performed with approximately 20 flies for each genotype . For GFP observation , flies were dissected in PBS and either directly observed under a Leica MZ16F dissecting microscope , or mounted in PBS for imaging with a Zeiss Axioimager Z1 . β-galactosidase was visualised by X-gal staining , as previously described [44] , followed by mounting in a 50∶50 mix of ethanol and glycerol . Images were captured with a Leica DFC300FX camera and Leica Application Suite or with an Axiocam MRn camera and Axiovision respectively . All PGN was prepared as described in Leulier et al , 2003 [4] . Each PGN was used at a monomer equivalent concentration of 5mmol . L−1 . For production of TCT , E . coli PGN was purified from the BW25113 Δlpp::CmR strain that does not express the Braun lipoprotein and digested by SltY lytic transglycosylase [5] . Radiolabelled E . coli PGN was obtained by incorporation of meso-[14C] DAP ( 11 . 6 kBq . nmol−1 ) into the FB8 lysA::kan strain grown in M63 minimal medium supplemented with 0 . 2% glucose and 100µg . ml−1 of lysine , threonine and methionine [45] . Radiolabelled [14C]-TCT , produced by digestion of the resulting material with SltY , was purified by HPLC and its specific activity was estimated as 2kBq . nmol−1 . Both TCT and [14C]-TCT were used at a concentration of 1mmol . L−1 . Haemolymph was extracted manually using a Nanodrop microinjector ( Nanoject™ ) [44] . A glass needle containing approximately 100nL of protease inhibitors ( 1× Complete Mini , Roche , in PBS ) was used to prick the flies in the dorsal thorax , inject 25nL of the inhibitor and immediately extract as much haemolymph as possible ( through the same wound ) . For the experiment to detect [14C]-TCT , the dorsal thorax was washed , prior to haemolymph extraction , with a filter paper dipped in water to avoid contamination by [14C]-TCT potentially present on the cuticle . For each fly , the extracted haemolymph ( and remaining protease inhibitor ) was deposited on a filter paper , accumulating the haemolymph from 100 flies on the same paper . For the experiment to detect the presence of bacteria in the haemolymph , contamination between flies was avoided by washing the needle between each extraction first with one aliquot of protease inhibitor , followed by ethanol and a second , clean , aliquot of inhibitor , before refilling with a third aliquot of inhibitor . Haemolymph extracts or 1µL drops of a dilution series of the [14C]-TCT solution were deposited on Whatman filter paper , which was then exposed to a Phosphor Imager Screen ( GE Healthcare ) for 2 weeks at room temperature . Images were generated with a Typhoon Trio Phosphor Imager ( GE Healthcare ) and quantified with ImageQuant TL . PGN , TCT and water were injected using a Nanodrop microinjector ( Nanoject™ ) [44] . Flies were pricked in the thorax with a glass needle containing the elicitor solution and 13nL was injected . The Flybase ( http://www . flybase . org ) accession numbers for genes mentioned in this study are: Attacin A ( CG10146 ) , Cecropin A ( CG1365 ) , Defensin ( CG1385 ) , Diptericin ( CG12763 ) , Drosocin ( CG10816 ) , Drosomycin ( CG10816 ) , Gal ( CG9092 ) , Metchnikowin ( CG8175 ) , PGRP-LB ( CG14704 ) , PGRP-LC ( CG4432 ) , Pirk ( CG15678 ) and Relish ( CG11992 ) .	Innate immunity is the first line of antimicrobial defence for vertebrates and the only immune response present in invertebrates such as the fruitfly Drosophila , which provides a powerful model system to study innate immunity . Interestingly , local infections of epithelia like the gut and , in our study , the genital tract , result not only in a local immune response , but in an immune response of the whole body . The latter seems to protect Drosophila against the potential spread of local infections . We have investigated the immune response to bacteria placed on the genitalia , at the entrance to both the genital tract and hindgut . This could be a natural entry route of pathogens , possibly linked to sexually transmitted infections . We observe a strong immune response to Gram-negative bacteria , mediated by the immune responsive Imd signalling pathway . This response depends on peptidoglycan , a crucial component of the bacterial cell wall , as pure peptidoglycan placed on the genitalia is sufficient to trigger a whole body immune response . Finally , we present strong evidence that peptidoglycan fragments within the genital tract or hindgut can cross these epithelia , enter the body cavity and thus induce a system wide immune response to a local infection .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "immunology/innate", "immunity", "genetics", "and", "genomics/genetics", "of", "the", "immune", "system" ]	2009	Long-Range Activation of Systemic Immunity through Peptidoglycan Diffusion in Drosophila
Mosquito sampling during Japanese encephalitis virus ( JEV ) -associated studies , particularly in India , has usually been conducted via aspirators or light traps to catch mosquitoes around cattle , which are dead-end hosts for JEV . High numbers of Culex tritaeniorhynchus , relative to other species , have often been caught during these studies . Less frequently , studies have involved sampling outdoor resting mosquitoes . We aimed to compare the relative abundance of mosquito species between these two previously used mosquito sampling methods . From September to December 2013 entomological surveys were undertaken in eight villages in a Japanese encephalitis ( JE ) endemic area of Bangladesh . Light traps were used to collect active mosquitoes in households , and resting boxes and a Bina Pani Das hop cage were used near oviposition sites to collect resting mosquitoes . Numbers of humans and domestic animals present in households where light traps were set were recorded . In five villages Cx . tritaeniorhynchus was more likely to be selected from light trap samples near hosts than resting collection samples near oviposition sites , according to log odds ratio tests . The opposite was true for Cx . pseudovishnui and Armigeres subalbatus , which can also transmit JEV . Culex tritaeniorhynchus constituted 59% of the mosquitoes sampled from households with cattle , 28% from households without cattle and 17% in resting collections . In contrast Cx . pseudovishnui constituted 5 . 4% of the sample from households with cattle , 16% from households with no cattle and 27% from resting collections , while Ar . subalbatus constituted 0 . 15% , 0 . 38% , and 8 . 4% of these samples respectively . These observations may be due to differences in timing of biting activity , host preference and host-seeking strategy rather than differences in population density . We suggest that future studies aiming to implicate vector species in transmission of JEV should consider focusing catches around hosts able to transmit JEV . Japanese encephalitis virus ( JEV ) is one of the most important causes of viral encephalitis in Asia with an estimated 67 , 900 cases annually [1] . The transmission cycle of JEV was initially studied in Japan during the 1950’s [2–7] . The most important mosquito vector—Culex tritaeniorhynchus—and reservoir hosts—pigs and ardeid birds—first implicated in Japan are generally thought to drive transmission across Asia [8–11] . The high numbers of Cx . tritaeniorhynchus caught during JEV-associated field studies have reinforced current theory about the primary role of this mosquito species in JEV transmission [3 , 12–21] . In particular , Cx . tritaeniorhynchus has often constituted the majority of the mosquitoes sampled in regions of India reporting Japanese encephalitis ( JE ) cases [15–17 , 20–36] . For this reason , in addition to observations of this species feeding predominantly on mammalian hosts , including pigs , and the number of viral isolates obtained from this species , Cx . tritaeniorhynchus is thought to be the primary vector in India [17 , 37] . JEV has however , been detected in 15 other mosquito species that are found in this region [37] . Although detection of virus in a mosquito species does not indicate that it is a vector , it does demonstrate that the species has fed upon a viremic host and its potential as a vector should be further considered . Studies in India reporting high numbers of Cx . tritaeniorhynchus relative to other species have often been undertaken around cattle ( dead-end hosts for JEV [38] ) and , less frequently , pig sties and human dwellings [16 , 17 , 20 , 21 , 24–26 , 28–31 , 33 , 35] . Justification for sampling near cattle was made by reference to the ability of this method to catch large numbers of JEV vector species in China [24 , 39] . Outdoor resting collection methods have also been developed to collect Cx . tritaeniorhynchus in addition to other species , although these alternative methods have not been used as frequently in studies aiming to implicate vector species in JEV transmission [40] . In addition , few studies in South Asia have compared these methods with respect to the relative abundance of mosquito species caught . Mosquito sampling methods can be biased toward certain species [41–45] . However , differences in the observed abundance of Cx . tritaeniorhynchus relative to other potential vector species between sampling methods in South Asia has been little studied . Bias in estimates of mosquito species relative abundance could have important implications for the process of identifying the host and mosquito species that drive transmission , given that the two are linked by the host-feeding preferences of the mosquito [46] . In regions of India experiencing JE outbreaks , between 85–98% of Cx . tritaeniorhynchus bloodmeals have been found to be from cattle , even when resting mosquitoes were sampled away from potential hosts [47–49] . The preference of Cx . tritaeniorhynchus for feeding on cattle has also been demonstrated under experimental conditions [50] . Thus , if other species exhibit different host preferences , using collection methods near cattle may lead to an overrepresentation of Cx . tritaeniorhynchus within mosquito samples . We therefore aimed to compare the relative abundance of mosquitoes between two previously used mosquito sampling methods in a region of Bangladesh where JE cases have been reported [51] . More specifically , we aimed to: 1 ) quantify the difference in species composition of mosquitoes captured in resting collections adjacent to oviposition sites ( method 1 ) and those captured at light traps placed in households near humans and domestic animals ( method 2 ) ; and 2 ) quantify the association between the numbers of humans and domestic animals , including cattle , in a household and the relative abundance of common mosquito species captured at light traps . Hospital-based surveillance between 2003–2005 , and 2007–2008 and related JE burden studies indicated the Rajshahi Medical College Hospital catchment area in northwest Bangladesh as having the highest rates of human JE incidence in Bangladesh [51 , 52] . Based on this finding , eight villages within this catchment area ( Naogaon , Chapai Nawabganj and Rajshahi Districts within Rajshshi Division ) were randomly selected from a census database using a random number generator ( S1 Fig ) . These villages were surveyed during the JE transmission season [51] , between September and December 2013 . In each of the eight villages , village boundaries were established upon arrival by consulting with the local chairperson . Resting collections were made at only seven of the eight villages , due to time constraints . Collections were made for one to four days in each village , depending on the time available ( S1 Table ) . Resting collections were conducted in shaded areas of vegetation adjacent to mosquito oviposition sites ( e . g . ponds , ditches and rice fields ) . Ten to 20 resting boxes , constructed from 1m long metal frames covered with black refuse bags , were placed near various habitats in the chosen sites during the first day at each village and checked each morning between 8 . 30 and 12 . 00 . Resting collections were also made using a Bina Pani Das ( BPD ) hop cage , which was constructed according to Das [40] . Using the hop cage , approximately every meter along a transect , vegetation was disturbed for 30 seconds as described by Das [40] . The number of transects and their length varied according to the shape and size of the area available for sampling , but an average of three transects of 20m in length were surveyed in each village . The action of disturbing the vegetation resulted in resting mosquitoes flying up into the cage , from which they could then be collected after each 30 second sampling period . The length of each transect and number of samples were recorded . Two surveyors checked resting boxes and used the hop cage concurrently , collecting mosquitoes by hand-held battery powered aspirators . From six to 12 Centers for Disease Control and Prevention miniature light traps were set per night in each village depending upon the time available ( S1 Table ) . Light was the only attractant used with traps , which were set from dusk to dawn . Traps were set on the first night in households along the main village road so that each light trap was approximately evenly distributed from another and so that they covered the extent of the village . Traps were moved to different households upon subsequent days of trapping within a village . GPS locations in Google Earth were used to select households for the placement of the next night’s light traps to ensure traps were approximately evenly distributed . At the time of survey , numbers of all domesticated animals and humans living in each household where light traps were set were obtained by interview with a household member . Locations where animals were kept at night varied between selected households . Locations of light traps inside household areas , including an indoor room where humans slept , animal shed , indoor areas where both humans and animals were present , and outdoor courtyards where animals were kept were therefore alternated between households to enable approximately equal sampling effort in each area . Mosquitoes were collected from light traps each morning after resting collections were made . All mosquitoes were killed using chloroform in an open well ventilated space , and immediately separated from other insects at a local field station after every morning collection . Trained entomologists further separated female mosquitoes from males and preserved them in tubes containing silica gel and cotton wool . As there is no specific key to the mosquitoes of Bangladesh , a number of taxonomic keys [53–60] from other countries in Asia were used that collectively included the species listed by Ahmed [61] . Females , where possible , were identified to species level and numbers of each species per catch ( e . g . per light trap , per group of boxes , or per BPD hop cage transect ) recorded . Three of 123 light trap catches were estimated by identifying approximately one third of the catch , due to high numbers collected ( >5000 female mosquitoes each ) . All data are available in a GitHub repository along with source code for the analyses: https://github . com/PulliamLab-UFL/mosquito-Rajshahi . In addition to the eight villages where household host data were collected , light traps were also set in an additional two villages for which no household host data were collected . The data from these two villages were therefore not included in data analyses but are included in the repository . We use the term ‘relative abundance’ to describe numbers of mosquitoes caught by each sampling method in order to acknowledge that we do not know how counts per collection scales with actual population density . Common species were defined as those constituting at least 5% of the total mosquito collection by either method . For both sampling methods , the proportion of the total catch constituting each species and approximate 95% confidence intervals ( 1 . 96*√p ( 1-p ) /N , where p is species proportion and N is the total number of mosquitoes ) were calculated . Estimates of species richness were compared between sampling methods , taking into consideration the number of individual female mosquitoes collected . To do this , we considered the total number of species caught as a function of the number of female mosquitoes caught by sampling method 2- the method by which the most species were obtained . This relationship was assessed via linear regression , with log ( x + 1 ) number of female mosquitoes as the explanatory variable and species richness as the response . Regression coefficients were then used to compare species numbers across sampling methods based on the aggregated data for each method by calculating ( a ) the species richness expected from method 2 in a sample of 575 individuals ( equivalent to the number of females obtained by method 1 ) , and ( b ) the number of mosquitoes that would have been required using sampling method 2 to catch 24 species of mosquito ( equivalent to the total number of species collected via method 1 ) . Hill’s diversity numbers [62 , 63] were calculated for method 1 ( separately for the BPD hop cage and resting boxes ) and method 2 to compare sampled species diversity between the two methods . Potential differences in species composition between samples from each method were assessed . The probability of selecting each common species from a collection using method 2 compared to a collection using method 1 were calculated for each village where these species were collected by both methods , using log odds ratios , standard error ( SE ) and 95% confidence intervals for the log odds . A confidence interval that did not include zero indicated that a species was significantly more likely to be selected from a collection using method 2 than a collection using method 1 ( if greater than zero ) or the opposite ( if less than zero ) . The log ( x+1 ) transformed count data for each common species in light traps was analyzed using multiple linear regression , as a function of the number of cattle , goats , birds and humans in households where light traps were set . All explanatory variables , including interaction terms , were initially included in the model , and then a step-wise elimination of non-significant terms according to the F-test was undertaken to achieve the minimal adequate model [64] . Using the minimal adequate model , the predict function in the R stats package [65] was used to plot the predicted relationship between individual significant explanatory variables and the number of female mosquitoes . Given the substantial effect of cattle relative to other hosts upon common species in our analysis , arithmetic means and standard error of the means ( s . e . m . ) of numbers of each mosquito species per light trap sample were compared for households with cattle and with no cattle . All 36 species identified during the study were observed in the mosquito collection using method 2 , compared with 24 species using method 1 , which yielded 575 female mosquitoes . Using results from linear regression ( Fig 1 ) , we estimate that 18 species would have been expected in a sample of 575 individual mosquitoes ( the total captured by method 1 ) using method 2 . To collect 24 species by method 2 , at least 3500 mosquitoes would be required ( Fig 1 ) . Indeed , despite more mosquitoes being collected by method 2 , Hill’s diversity numbers H1 and H2 were both higher for method 1 ( Table 2 ) , indicating a greater diversity of species was sampled by method 1 than by method 2 . Culex tritaeniorhynchus was observed to be common by both methods . The collection from method 1 contained five other common species- Cx . pseudovishnui , Cx . gelidus , Ar . subalbatus , Cx . vishnui and Ar . kesseli ( Table 1 ) . Three other common species were observed in collections from method 2- An . peditaeniatus , Cx . gelidus , and Cx . pseudovishnui , with the rest of the observed species each constituting less than 5% of the total sample . With respect to the two resting collection methods , 153 mosquitoes were collected using resting boxes and 422 using the BPD hop cage . H1 and H2 were similar between these methods ( Table 2 ) . Species collected by hop cage but not resting box included Ae . lineatopennis , An . barbirostris , An . nigerrimus , Ar . kuchingensis , Cq . crassipes , Cx . whitmorei , Ma . annunlifera . Species collected by resting box but not hop cage included An . vagus , Cx . infula and Ma . indiana . In five of seven villages , Cx . tritaeniorhynchus was more likely to be selected from a collection using method 2 than from a collection using method 1 ( Fig 2 ) . There was no significant difference between sampling methods for this species in the 2nd and 8th villages visited . In the 8th village there were no significant differences between sampling methods for Cx . pseudovishnui , Ar . subalbatus and Ar . kesseli . These three species were however more represented in the collection using method 1 than in the collection using method 2 for all other villages ( Fig 2 ) . There were no significant differences between sampling methods for Cx . gelidus and Cx . vishnui in the majority of villages ( Fig 2 ) . Households had varying combinations of cattle , goats , ducks , chickens , pigeons and , rarely , geese or sheep . Pigs were only present in one of the eight villages . There was no association between the number of common mosquito species caught by light trap and the number of humans or goats in a household . Households with higher numbers of cattle on average had higher numbers of all four common species in light traps ( Fig 3 , S3 Table ) . The numbers of cattle in a household ranged from zero to 12 . For each additional cow present in a household there was an approximate two-fold increase in the number of Cx . tritaeniorhynchus and approximately 1 . 5 to 1 . 7-fold increase of other common species ( S3 Table ) . In households with no cattle , higher numbers of common mosquito species were more likely when there were higher numbers of birds in a household ( Fig 4 ) . The number of birds in a household ranged from zero to 120 . In the absence of cattle , there was approximately a 1 . 3-fold increase in the numbers of each of the four common species for each additional 10 birds in a household ( S3 Table ) . Though numbers of both birds and cattle influenced the relative abundance of all common mosquito species , the influence of cattle was higher for Cx . tritaeniorhynchus than the influence of birds relative to other mosquito species . The association between cattle numbers and abundance of individual mosquito species had implications for the mean relative abundance ( Table 3 ) and species composition ( Fig 5 ) by sampling method . Entomological studies of potential JEV vectors in India have frequently collected mosquitoes from around cattle sheds at dusk and many report Cx . tritaeniorhynchus to constitute at least 50% of the sample [20–27 , 29–31 , 33 , 35 , 36 , 66] . This finding has been used as evidence supporting the theory that Cx . tritaeniorhynchus is the primary vector of JEV in this country [20–27 , 29–31 , 33 , 35–37 , 66] . In our study , while Cx . tritaeniorhynchus constituted the majority of the sample when using light traps in households with cattle , in households without cattle Cx . tritaeniorhynchus did not constitute the majority of the sample , nor from resting collections near oviposition sites . Our findings are consistent with a study from India where Cx . tritaeniorhynchus was two to forty times more abundant in cattle sheds at dusk than other species whereas during daytime resting collections other species , including Cx . quinquefasciatus , Cx . pseudovishnui and Anopheles spp , were more abundant [66] . Studies aiming to implicate vector species in transmission of JEV in areas where the transmission ecology differs substantially from that first described in Japan ( accompanying article ) , should consider focusing catches around hosts able to transmit JEV instead of collections near dead-end hosts , such as cattle . Other species , including Cx . gelidus , Cx . quinquefasciatus , Ar . subalbatus and anophelines have been observed to be infected with JEV in the wild [22 , 67 , 68] . Species in addition to Cx . tritaeniorhynchus , including Ar . subalbatus , Cx . quinquefasciatus and Cx . pseudovishnui , are capable of transmitting JEV under experimental conditions [67 , 69] . The relative contributions of other competent mosquito species to JEV transmission should be further investigated . Differences in the relative abundance of competent species in collections from cattle using aspirators or light traps at dusk may be a result of alternative host-seeking strategies between species rather than actual differences in population density . Light traps in particular are biased because they capture phototactic species active after dark . Culex pseudovishnui , Ar . subalbatus and Ar . kesseli were significantly better represented in daytime resting collections near oviposition sites than light trap collections near hosts . During our surveys in Bangladesh , Ar . subalbatus was frequently observed during daylight hours . Similarly , Das et al . [70] , which was one of few studies in India to undertake human-bait collections during daylight hours , found Ar . subalbatus to constitute 76% of the sample whilst the Cx . vishnui subgroup including Cx . tritaeniorhynchus constituted less than 5% . It is likely that light traps are not effective for catching this species because host-seeking females are most active before light traps become effective . This may also be the case for Cx . pseudovishnui , as Bhattacharyya et al . [71] reported the initial peak in biting for this species to occur at 19 . 00 compared with 21 . 00 for Cx . tritaeniorhynchus . The study was undertaken between May and June , when sunset times ranged from 17 . 45 to 18 . 10 . The initial peak biting time for Cx . pseudovishnui was approximately one hour after sunset and for Cx . tritaeniorhynchus three hours after sunset . It is therefore possible that it may not be dark enough one hour after sunset for light traps to be fully effective for attracting Cx . pseudovishnui . Indoor resting collections were not conducted during the current study and we may therefore have under-sampled species which tend to rest indoors including Cx . quinquefasciatus , An . vagus and An . subpictus [32] . Thus , these species may constitute a greater proportion of the true mosquito community in the study area than our methods account for . We also acknowledge that the current study represents a snapshot of mosquito relative abundance , with fieldwork being conducted for only three months of the year ( Sep—Dec ) and in a relatively small number of villages . Due to the political situation during the survey period we were unable to spend an equal amount of time in each village , resulting in variation in the amount of time for resting collections and number of light traps set in each village . While this sample is sufficient to demonstrate that there are important differences between sampling methods we cannot therefore account for potential variation in mosquito community composition in space and time . Studies aiming to implicate vector species in JEV transmission should carefully consider potential method biases when estimating relative abundance . Initial studies in Japan [72] used traps baited with host species known to be able to transmit JEV , and those that were present in high density , to sample the mosquito community feeding on hosts of relevance to transmission rather than dead-end hosts such as cattle . Because cattle do not produce viremia sufficient to infect mosquitoes , they have the potential to dilute transmission , especially when they are present in high density relative to other hosts , as in regions of India and Bangladesh . This may in turn reduce the role of Cx . tritaeniorhynchus as a vector species given its preference for cattle [50] . Given that Cx . tritaeniorhynchus is not as abundant in samples using method 1 , is an indicator that other mosquito species may be just as , or more abundant and should not be discounted as potential vectors . Where possible , supplementing resting collections with traps baited with hosts able to transmit JEV would assist in implicating mosquito species in transmission . During our study , substantially more mosquitoes were required to be collected by light trap near hosts than by resting collection near oviposition sites in order to observe an equivalent number of species . As the most time consuming aspect of JEV- associated entomological fieldwork is often the identification of mosquitoes to species level , we suggest resting collection near oviposition sites to be more efficient for initially establishing the mosquito species present in an area because fewer individual mosquitoes would need to be identified in order to obtain a sample of the species present . This approach also requires relatively little equipment and does not require use of animal baits or electricity . Furthermore , because resting mosquitoes are often blood-fed , the same samples could be used for estimating the proportion of bloodmeals taken from different host species . The total number of JEV isolates from Cx . tritaeniorhynchus , often obtained from sampling near cattle at dusk , has additionally been used as evidence that Cx . tritaeniorhynchus is the primary vector [37]; however , the number of JEV isolates from any mosquito species is a product of both the abundance of infected mosquitoes of that species and the probability that an individual of that species will be caught by the particular trapping method used . As Cx . tritaeniorhynchus is collected in large numbers compared with other species near cattle at dusk with light traps or aspirators , it may be expected that the highest number of viral isolates would also be obtained from this species using this method , regardless of their actual importance in transmitting JEV to humans . In this study , Cx . tritaeniorhynchus was approximately 11 times more abundant in light traps at households with cattle than Cx . pseudovishnui . This apparent difference in abundance may result from differences in the true abundance of the species , differences in trap efficiency , or both . Given the difference in the number of individuals caught , however , the prevalence of JEV in Cx . pseudovishnui would have to be at least 11 times greater than that for Cx . tritaeniorhynchus in order to be likely to yield a higher number of viral isolates from this species using collections near cattle . We advise that maximum likelihood estimation of virus prevalence in mosquito populations be considered alongside total number of viral isolates , to more accurately characterize the potential role of mosquito species . Maximum likelihood estimates of prevalence would be more useful than total numbers , for implicating species important in maintaining JEV transmission , but additional studies are also required to identify species which present the highest risk of transmission to humans . The numbers of infectious bites on humans is related not only to the prevalence of virus in mosquitoes and the number of mosquitoes relative to hosts , but also mosquito host feeding patterns . Even if Cx . tritaeniorhynchus is the most important vector for maintaining JEV transmission , bloodmeal studies have shown Cx . tritaeniorhynchus to feed between 2 and 4% on humans [48 , 73] . In addition , human landing catches in India resulted in a sample of 2836 Ar . subalbatus , 579 Cx . quinquefasciatus , 41 Cx . vishnui , 29 Cx . pseudovishnui and no Cx . tritaeniorhynchus , in addition to species of Anopheles and Mansonia [70] . The species most important for human transmission in South Asia remain to be identified . In addition to relative abundance , greater focus should be placed upon mosquito species feeding patterns , fine-scale variation in host and mosquito species community composition , and competence of hosts and vectors for viral replication in regions experiencing JE outbreaks where the transmission context is different from that first described in Japan ( accompanying article ) .	The relative numbers of individuals of each mosquito species in an area are important to estimate when identifying species that contribute the most to vector-borne pathogen transmission . However , methods to sample mosquitoes and enumerate the number of individuals collected often vary in their catch efficacy between species . For example , species that take a bloodmeal during daylight hours are less likely to be caught using a light trap than a species that feeds predominantly at night . Similarly , sampling near a mammalian host will more likely collect mosquitoes with a preference for mammals than those with a preference for birds . In this study we compare sampling methods for assessing the relative abundance of mosquito species that may be involved in Japanese encephalitis virus ( JEV ) transmission . Collections near cattle- a species unable to transmit JEV- have been influential in implicating Cx . tritaeniorhynchus as the primary vector of JEV in South Asia , due to the high number of individuals of this species caught relative to other species . Indeed , this mosquito constituted the majority of the mosquitoes collected by light traps in households with cattle in this study . However , other species were more common when sampling households without cattle or resting mosquitoes near oviposition sites . We propose that methods used to sample mosquitoes in studies aiming to implicate species in JEV transmission in South Asia be reconsidered given that there are other mosquito species that are able to transmit JEV , and these species may be underrepresented when sampling using light traps near cattle .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[]	2016	Sampling Design Influences the Observed Dominance of Culex tritaeniorhynchus: Considerations for Future Studies of Japanese Encephalitis Virus Transmission
Members of the family Coronaviridae have the largest genomes of all RNA viruses , typically in the region of 30 kilobases . Several coronaviruses , such as Severe acute respiratory syndrome-related coronavirus ( SARS-CoV ) and Middle East respiratory syndrome-related coronavirus ( MERS-CoV ) , are of medical importance , with high mortality rates and , in the case of SARS-CoV , significant pandemic potential . Other coronaviruses , such as Porcine epidemic diarrhea virus and Avian coronavirus , are important livestock pathogens . Ribosome profiling is a technique which exploits the capacity of the translating ribosome to protect around 30 nucleotides of mRNA from ribonuclease digestion . Ribosome-protected mRNA fragments are purified , subjected to deep sequencing and mapped back to the transcriptome to give a global “snap-shot” of translation . Parallel RNA sequencing allows normalization by transcript abundance . Here we apply ribosome profiling to cells infected with Murine coronavirus , mouse hepatitis virus , strain A59 ( MHV-A59 ) , a model coronavirus in the same genus as SARS-CoV and MERS-CoV . The data obtained allowed us to study the kinetics of virus transcription and translation with exquisite precision . We studied the timecourse of positive and negative-sense genomic and subgenomic viral RNA production and the relative translation efficiencies of the different virus ORFs . Virus mRNAs were not found to be translated more efficiently than host mRNAs; rather , virus translation dominates host translation at later time points due to high levels of virus transcripts . Triplet phasing of the profiling data allowed precise determination of translated reading frames and revealed several translated short open reading frames upstream of , or embedded within , known virus protein-coding regions . Ribosome pause sites were identified in the virus replicase polyprotein pp1a ORF and investigated experimentally . Contrary to expectations , ribosomes were not found to pause at the ribosomal frameshift site . To our knowledge this is the first application of ribosome profiling to an RNA virus . Members of the family Coronaviridae have the largest genomes of all RNA viruses , typically in the region of 30 kilobases ( kb ) . Several coronaviruses , including SARS-CoV and MERS-CoV , are of medical importance , with high mortality rates and , in the case of SARS-CoV , significant pandemic potential . Other coronaviruses , such as Porcine epidemic diarrhea virus and Avian coronavirus , are important livestock pathogens . Coronavirus infections are frequent in bats and other mammals [1] and interactions between humans and non-human animal populations presents a constant risk of new zoonotic outbreaks [2] . Recent findings also indicate an evolutionary origin of the established human coronavirus species , Human coronavirus 229E in hipposiderid bats [3] . The family Coronaviridae is divided into the subfamilies Coronavirinae and Torovirinae . Torovirinae includes the genera Bafinivirus and Torovirus , infecting fish and mammals respectively , while Coronavirinae includes the genera Alphacoronavirus , Betacoronavirus , Gammacoronavirus and Deltacoronavirus , commonly infecting mammals and birds . SARS-CoV and MERS-CoV are members of the genus Betacoronavirus . Therefore , a useful model for these two viruses , especially with regard to their structure and replication , is Murine coronavirus , a betacoronavirus that is commonly referred to as mouse hepatitis virus ( MHV ) . Like all coronaviruses , MHV has a monopartite , positive-sense , single-stranded RNA genome ( gRNA ) ( Fig 1A ) . The 5′ two thirds of the genome contains two long open reading frames ( ORFs ) , ORF1a and ORF1b , which encode the replicative proteins . These ORFs are expressed as two polyproteins pp1a and pp1ab , where pp1ab is a “transframe” fusion of the ORF1a and ORF1b products , produced via −1 programmed ribosomal frameshifting ( −1 PRF ) [4 , 5] . Polyproteins pp1a and pp1ab are proteolytically cleaved by virus-encoded proteases , PLP1 and PLP2 ( in nsp3 ) and 3CL ( nsp5 ) to produce the non-structural proteins nsp1 to nsp16 . The 3′ third of the genome contains ORFs that encode structural proteins and accessory proteins . These ORFs are translated from a series of subgenomic mRNAs ( mRNAs 2 to 7 ) produced during virus infection . Each subgenomic mRNA is identical to a 3′-coterminal region of the virus genome with the exception of a 65 nucleotide ( nt ) leader sequence at the 5′ end that is identical to the 5′ end of the gRNA . These leader sequences are added ( as a reverse complement ) during synthesis of subgenomic negative-sense templates that then give rise to the positive-sense mRNAs . The process of discontinuous transcription during negative-strand RNA synthesis takes place via polymerase “jumping” at specific “transcription regulatory sequences” ( TRSs ) on the gRNA template . In MHV , mRNAs 2 to 7 encode , respectively , proteins 2 and haemagglutinin-esterase ( HE ) , spike ( S ) , protein 4 , protein 5 and envelope ( E ) , membrane ( M ) , nucleocapsid ( N ) and internal protein ( I ) , with mRNAs 5 and 7 being functionally bicistronic ( Fig 1A ) [6] . In the laboratory-adapted strain ( MHV-A59 ) employed in the present study , however , expression of HE and protein 4 is defective . Ribosome profiling is an emerging methodology that facilitates global mapping of the positions of translating ribosomes on the transcriptome , defining at the codon level the extent to which individual mRNAs species are engaged in protein synthesis [7–9] . The technique exploits the knowledge that translating ribosomes can protect from RNase digestion a defined fragment of mRNA of around 28–30 nt in length [10] . In ribosome profiling , often referred to as RiboSeq , cells are lysed under conditions optimised to minimise further ribosome movement ( addition of translation inhibitors , rapid freezing ) , the lysate is treated with ribonuclease ( often RNase 1 ) to degrade regions of mRNAs that are not physically protected , and the ribosomes harvested on sucrose gradients or through a sucrose cushion . The ribosome pellet is de-proteinized , the ribosome-protected fragments ( RPFs ) harvested by elution from a polyacrylamide gel , ligated to adapters , subjected to RT-PCR , deep sequenced and mapped back to the genome . This analysis reveals the location and abundance of ribosomes on mRNAs with up to single-nucleotide precision . The corresponding transcriptome is also determined from the same lysate: total RNA is harvested , fragmented , cloned and sequenced to generate a parallel RNA sequencing ( RNASeq ) library . Ribosome profiling has been applied to a variety of cellular organisms to address a range of questions in translational control and global gene expression [9 , 11–16] . Also , it has been employed in the study of the replication of large DNA viruses; namely , human cytomegalovirus [17–18] , Kaposi's sarcoma-associated herpesvirus [19] , herpes simplex virus 1 [20] , vaccinia virus [21] , and bacteriophage lambda [22] , providing insights into the temporal regulation of gene expression in these viruses and identifying numerous previously unrecognized translated ORFs , including novel protein-coding ORFs and short regulatory uORFs . In this paper , we describe the first analysis of RNA virus replication and gene expression by ribosome profiling ( and parallel RNASeq ) , using MHV as a model system . The data obtained allowed us to determine the time course of virus positive and negative-sense RNA production , as well as the translation of each of the virus genes , the expression of short and/or previously unannotated ORFs , and the efficiency of −1 PRF . We also investigated early time points of infections at high multiplicity to visualise the translation of input genomes . The profiling data also revealed examples of prominent ribosomal pausing within the coding regions for nsp3 and nsp6 . Nsp3 ribosomal pausing was confirmed in in vitro translation experiments . Surprisingly , we found that ribosomes do not pause appreciably during −1 PRF , arguing against a requirement for pausing in frameshifting . This study also provides insights into the challenges associated with the profiling of RNA viruses and suggests strategies that may prove beneficial in future studies . To study the kinetics of virus RNA and protein synthesis in a single cycle of virus replication , we performed two independent biological repeats ( repeats 1 and 2 ) of an MHV infection time course in which murine 17 clone 1 cells ( 17Cl-1 ) were infected with recombinant MHV-A59 at a multiplicity of infection ( MOI ) of 10 and cells harvested at 1 , 2 . 5 , 5 and 8 h post-infection ( p . i . ) , with mock-infected cells harvested at 1 and 8 h . For all time points , two dishes were prepared and , immediately prior to harvesting , cells were treated with cycloheximide ( CHX ) alone , or harringtonine ( HAR ) then CHX ( as detailed in Materials and Methods ) , for analysis of elongating ( CHX ) and initiating ( HAR ) ribosomes , respectively . Subsequently , RiboSeq ( CHX ) , RiboSeq ( HAR ) and RNASeq ( CHX only ) libraries were prepared for each time point , deep sequenced and reads mapped to host and virus sequences ( see Materials and Methods ) . The composition of each library is summarised in S1A Table and S1 Fig . As an example of the data provided by our experimental strategy , Fig 1B shows the density of RiboSeq CHX and RNASeq reads mapping to the virus genome at 5 h p . i . In general , there is a 5′ to 3′ increasing gradient in total ribosome density , with the N ORF being expressed at the highest level , the M , 5 , S and 2 ORFs at intermediate levels , and ORFs 1a and 1b at the lowest levels . As expected , very little ribosome density was observed within the defective genes HE and 4 . The step reduction in RiboSeq density between ORF1a and 1b reveals the proportion of ribosomes that terminate at the ORF1a stop codon instead of frameshifting into ORF1b . In contrast , RNASeq density is essentially constant across ORFs 1a and 1b , and then steadily increases 5′ to 3′ , reflecting the cumulative density summed over the genomic RNA and 3′-coterminal subgenomic transcripts . Extra RNASeq density in the 5′ UTR reflects the 5′ leader sequence that is present on all subgenomic transcripts as well as the genomic RNA . RiboSeq density was also observed in the 5′ leader , although not corresponding to known coding regions ( see below ) . Negative-sense virus RNA is present at much lower amounts than positive-sense RNA , but follows roughly the same expression patterns , including high density in the ( anti ) -leader region , consistent with discontinuous transcription occurring during negative-sense RNA synthesis [23] . Low levels of negative-sense RiboSeq reads were also observed but these had length distributions that did not match typical RPF length distributions ( see below ) . Thus , these are unlikely to derive from ribosomes loading onto negative-sense RNAs ( e . g . non-specifically onto uncapped , possibly degraded virus-derived RNAs ) . Instead , they may derive from low amounts of RNA non-specifically co-sedimenting with ribosomes ( see below ) . Since the RiboSeq analysis represents the product of transcript abundance and translation efficiency , we also plotted the RiboSeq/RNASeq ratio along the genome ( Fig 1C ) . This ratio was substantially lower in ORF1a and ORF1b than in the 3′ coding ORFs ( except the defective genes HE and 4 ) , which may indicate that a substantial proportion of genomic RNA is not being translated ( e . g . sequestered in replication-transcription complexes [RTCs] or destined for packaging ) or that genomic RNA intrinsically has a relatively low translation efficiency . Note , however , that this simple calculation ignores the fact that RNASeq density is present for all ORFs on a transcript whereas RiboSeq density is only present for the translatable ORFs ( normally the 5′ proximal ORF ) . This discrepancy is accounted for in the more detailed analysis of translation efficiencies below . Fig 2 shows enlarged views of the virus transcript 5′ UTR and 3′ ORFs with linear scales optimized separately for each region . This analysis shows that there is significant variability in the RNASeq read depth within a transcript , which we ascribe to biases such as fragmentation bias , PCR bias and ligation bias . Similarly , variability in the RiboSeq data within a CDS may be partly due to nuclease bias , PCR bias and ligation bias but also reflects real variations in ribosome progressivity . The depth of RNASeq reads in the 5′ UTR is similar to that of the N ORF , reflecting that the major contribution to 5′ leader sequence comes from mRNA7 . Peaks in the RiboSeq HAR data highlight the canonical translation initiation sites of the 2 , S , 5 , M and N ORFs . In the same dataset , the ORF1a/1ab initiation peak is dwarfed by RPFs in the 5′ leader ( presumably mostly coming from mRNA7; see below ) . It should be noted that HAR arrests ribosomes at initiation , but not during elongation thus allowing elongating ribosomes to run-off . However , in these samples it is apparent that elongating ribosomes have not yet cleared the S ORF . We considered it important to assess the quality of the datasets that were obtained by our experimental strategy . For RPFs derived from non-organellar ribosomes of eukaryotic organisms , mapping of the 5′ end positions to coding sequences ( CDSs ) characteristically reflects the triplet periodicity ( herein referred to as “phasing” ) of translational decoding [7] . Good phasing within datasets is beneficial in assigning ORFs with confidence , particularly if such ORFs are very short or overlap . The extent of phasing can vary between protocols and libraries due , presumably , to variation in the efficiency of RNase I ( or other nuclease ) trimming or other factors . S2 Fig ( repeat 1 ) and S3 Fig ( repeat 2 ) show , for each library , histograms of the codon positions to which the 5′ ends of host mRNA reads map for different read lengths . The RiboSeq libraries show excellent phasing with the majority of RPF 5′ ends mapping to the first codon position . Conversely , and as expected , the 5′ ends of RNASeq reads had a nearly uniform distribution between the three possible codon positions . The RiboSeq read length distributions were typically sharply peaked at around 29 nt consistent with other analyses [8] , while those of RNASeq were much broader , consistent with a length distribution set by the size of the gel slice excised during purification of fragmented RNA in the RNASeq protocol ( approximately 28–34 nt ) . S4 Fig shows the distribution of host mRNA RPF 5′ ends relative to initiation and termination codons , summed over all host mRNAs in each of the RiboSeq libraries . For all samples , a discrete peak in RPF abundance was observed just upstream of the initiation site . As noted previously , the peak is probably largely a result of drug treatment—either HAR which specifically arrests initiating ribosomes , or CHX which arrests elongating ribosomes but allows ribosomes to continue to accumulate at initiation sites [8] . This peak corresponds to the 5′ ends of RPFs derived from initiating ribosomes with the AUG codon in the ribosomal P-site , and allows calibration of the offset between the RPF 5′ end and RPF P-site position , which , for these libraries , is normally 12 nt ( e . g . S5 Fig ) . For many samples , a discrete peak was also observed 15 nt upstream of the stop codon , corresponding to ribosomes pausing during termination ( with the stop codon in the ribosomal A-site ) . The presence of this peak appears to be subject to minor variation in sample preparation as it was not consistent between repeats ( cf . repeat 1 and repeat 2 , RiboSeq CHX mock 1 h in S4 Fig ) . In contrast to [24] , we believe that the clear spike four codons downstream of the initiation peak is an artefact of ligation bias ( and potentially also other biases ) : every read mapping to this position begins with 5′-AUG ( thus compounding any ligation preferences ) , whereas reads that map to the initiation peak have different 5′ and 3′ ends in different mRNAs ( thus averaging out any ligation preferences ) . For 30-nt reads , a trough was also apparent four codons upstream of the termination peak ( S5 Fig ) ; this corresponds to reads that all end in UAG-3′ , UAA-3′ or UGA-3′ , and again is likely to be an artefact of ligation bias . Peaks at the start and stop codons were also apparent for RNASeq data , corresponding to reads with 5′ ends aligning to the A of AUG and the middle nucleotide of the stop codon , respectively ( S6 Fig , right ) ; the latter is not visible in RiboSeq data due to low RiboSeq density in the 3′ UTR . A peak 12 nt upstream of the AUG ( more noticeable in repeat 1 samples , S6 Fig , left ) together with a very low level of phasing within the CDS ( S6 Fig ) likely represents a low level of contamination of RNASeq samples by RiboSeq samples , although the latter could potentially also be a result of codon usage bias , e . g . a preference for RNY codons [25] , compounded with ligation biases . Averaged over all host mRNAs , very few RPFs were observed in 3′ UTRs while a larger but still low level of RPFs were observed in 5′ UTRs ( S4 Fig and S5 Fig ) . The latter may largely derive from translation of uORFs in various locations and phases with respect to the main ORF of each mRNA [8] . We also observed a remarkable perturbation in host cell translation at late time points ( S4 Fig , lower panels—RiboSeq CHX , compare 5 and 8 h p . i . with 1 and 8 h mocks ) that was not mirrored in RNASeq data ( S6 Fig ) and could be a consequence of cell stress [26–28] . This phenomenon and other host cell responses to virus infection will be discussed in future work . We also addressed the issue of possible contamination during sample preparation as we expected that RNASeq and RiboSeq analysis of virus-infected cells may present some specific challenges . For example , late in infection , virus RNA can be produced at very high levels and extreme care is required to minimise cross contamination between late and early time-point libraries . Indeed , a comparison of read length distributions of host-derived RNA and virus-derived RNA revealed contamination of this type in some of our libraries , despite great care in processing experimental samples ( S7 Fig and S8 Fig ) . For example , in the first biological repeat ( S7 Fig ) , the virus and host length distributions in the 5 h p . i . RiboSeq CHX sample were almost identical . However , for the 1 and 2 . 5 h p . i . RiboSeq CHX samples , virus and host length distributions were dissimilar to each other but instead the virus length distribution resembled the RiboSeq CHX 5 h p . i . length distribution , suggesting contamination of virus RPFs from the later time-point sample into the earlier time points . The absolute amount of contamination was very low and would have little effect on host mRNA analyses; however , relative to the amount of virus RNA at 1 and 2 . 5 h p . i . , it was significant . Contamination was also apparent for the 1 and 2 . 5 h p . i . RiboSeq HAR samples and the 1 h p . i . RNASeq sample . Similarly , the mock-infected controls each contained ~1000–2000 virus reads ( cf . ~2–22 million at late time points of infection ) ( S1A Table ) . In the second biological repeat , the mock samples were evidently less contaminated , containing from only 0 to 55 virus reads each ( S1A Table ) . Nevertheless , traces of contamination were still apparent in the 1 and 2 . 5 h p . i . RiboSeq CHX and 1 h p . i . RiboSeq HAR samples ( S8 Fig ) . A different type of contamination was observed for the 8 h p . i . RiboSeq CHX sample in repeat 2 . Here , the host read length distribution was broad compared to the virus read length distribution , and the host mRNA phasing was poor ( S8 Fig and S3 Fig , respectively ) . This suggests that this sample is contaminated with RNASeq material from a sample containing little or no virus RNA , thus affecting the host mRNA length distribution but not the virus RNA length distribution . In subsequent discussions of the MHV profiling data , any samples suffering from contamination have been excluded , or subjected to appropriate caveats . Another potential source of “contamination” in our experimental strategy is the problem of non-ribosomal ribonucleoprotein ( RNP ) complexes . For example , certain virus proteins have RNA binding properties and can associate with viral and , potentially , cellular RNA . These RNP complexes may co-sediment with ribosomes and lead to contamination of RiboSeq libraries . Such contamination may be revealed by unusually high read density in host mRNA 3′ UTRs ( which normally have very low RPF occupancy ) and differences in read length distributions [29] . S9 Fig and S10 Fig show length distributions for all libraries for reads mapping within 10 to 100 codons upstream ( green; CDS ) or downstream ( orange; 3′ UTR ) of CDS termination codons . In all RiboSeq libraries , the 3′ UTR read density was extremely low compared to the CDS read density ( left plot of each pair ) . ( It should be noted however that , as HAR enriches for initiating ribosomes , the above analysis is not well-suited to HAR samples . ) For comparison , the RNASeq library 3′ UTR read density was typically ~80% of the CDS read density ( that it is not 100% likely reflects the presence of transcripts with 3′ UTRs that are shorter than the annotated 3′ UTRs ) . Since the analysis is based on mapping to NCBI RefSeq mRNAs , a low level of 3′ UTR occupancy derives from genuine RPFs derived from coding exons in one splice form that have alternative mappings to the 3′ UTR in another splice form . Further , low levels of post-termination unrecycled 80S ribosomes may enter the 3′ UTR [30–32] . Thus , for mock infections , the 3′ UTR RiboSeq read length distributions largely matched those of the CDSs ( S9 Fig and S10 Fig , 1 and 8 h mock CHX ) , albeit with some differences ( e . g . a high-end tail ) arising from unknown sources of contamination potentially including host protein:mRNA RNPs . While such contamination is expected to be present throughout the mRNA , it is more apparent in the 3′ UTR due to the much lower density of bona fide RPFs in this region . For infected samples , the host mRNA 3′ UTR density for CHX samples was similar in magnitude ( 0 . 5–1 . 2% ) to that of the mocks ( 0 . 7–0 . 9% ) , except for the 8 h p . i . time points where the 3′ UTR density was 2 . 9–6 . 3% of the CDS density ( S9 Fig and S10 Fig ) . Consistent with the probable RNASeq contamination discussed above , the length profile of the 8 h p . i . CHX repeat 2 sample was broad for both the CDS and 3′ UTR regions . On the other hand , the length profile of the 8 h p . i . CHX repeat 1 sample was not qualitatively different from that of the mocks , suggesting that the increase in 3′ UTR occupancy might not simply be explained by virus-induced RNPs , but rather , or as well , by an increase in bona fide RPFs in the 3′ UTRs . A mechanism for the latter could be overloading of the host cell ribosome recycling factors ( ABCE1 and any cofactors ) , allowing an increase in post-termination unrecycled 80S ribosomes entering the 3′ UTR [31] . If a proportion of late time-point contamination results from virus proteins interacting with mRNA to form RNPs , it may be significantly higher for virus RNA than for host mRNA , as virus proteins are likely to interact selectively with virus RNA; for example , through specific binding signals or via compartmentalization within the cell . Excess contamination in the virus RPF fraction may be gauged by comparing length distributions of reads mapping to virus positive-sense RNA with length distributions of reads mapping to host mRNA CDSs . Reassuringly , in all cases , the virus positive-sense RiboSeq reads showed a similar or even tighter length distribution at late time points than the host RiboSeq reads ( S7 Fig and S8 Fig; 5 h . p . i and 8 h p . i . , CHX and HAR ) . In contrast , as mentioned above , the small quantity of negative-sense virus reads in the RiboSeq samples had very different length distributions ( S7 Fig and S8 Fig ) indicating that they are unlikely to be true RPFs; such reads comprised <4% of virus reads for all RiboSeq samples , and <0 . 05% for the two 5 h p . i . CHX repeats . Fig 3A shows a time course of the total amount of virus RNA expressed as a fraction of total virus RNA plus host mRNA , for both RiboSeq CHX and RNASeq samples . Samples with contamination ( see above ) could only be used to give upper bounds ( grey symbols ) . Total virus translation as a fraction of total cellular translation increased 700 to 20 , 000-fold from 1 to 5 h p . i . , while virus positive-sense RNA increased 80 to 200-fold over the same time period . In repeat 2 , virus translation and RNA appeared to have reached a maximum by 5 h p . i . , while infection progressed a little slower in repeat 1 . From 1 h p . i . to 2 . 5 h p . i . , the positive-sense RNA fraction remained roughly constant ( presumably reflecting the input RNA ) while the negative-sense RNA fraction grew from essentially negligible amounts to ~0 . 1% of total virus RNA and host mRNA ( Fig 3A ) . At late time points , virus negative-sense RNA ceased to increase , whilst positive-sense virus RNA showed significant increases ( Fig 3B ) . At the later time points , virus translation had reached ~50–75% of total cell translation and positive-sense virus RNA had reached ~80–90% of total virus RNA plus host mRNA . At the same time , negative-sense virus RNA represented ~0 . 3% of total virus RNA and host mRNA ( Fig 3B ) . These findings are consistent with previous analysis of virus RNA synthesis in MHV-A59-infected cells [33] . Virus infection and the kinetics of viral protein expression over the time course were confirmed by western blot with antisera to the N , S and nsp9 proteins ( Fig 3C ) . We also calculated the levels of transcription and translation for each virus ORF throughout the time course ( Fig 4A ) . Note , again , that the data only provide upper bounds for the contaminated samples ( as indicated in Fig 3 ) . The particularly contaminated repeat 1 RiboSeq 1 h p . i . data are omitted from Fig 4A , while the upper bounds provided by the cleaner repeat 2 are included as they are likely to be more accurate . To calculate translation efficiencies , it is necessary to determine the amount of each virus transcript but , in the case of coronaviruses , raw RNASeq densities represent the cumulative sum of genomic RNA and all subgenomic transcripts . For example , for the N ORF , RNASeq density includes contributions from mRNAs 2 to 7 and gRNA . Thus , to calculate the amount of mRNA7 , we subtracted the positive-sense RNASeq density in the region of mRNA6 upstream of the mRNA7 TRS from the density in the mRNA7 region . We then followed a similar procedure for all other mRNAs . The same analysis was also applied to the negative-sense virus RNAs and these “decumulated” values are plotted in the right-hand panels of Fig 4A . Due to the low production of mRNA4 relative to mRNA3 , the amount of mRNA4 could not be estimated in this way . We also omitted the 1 h p . i . time point due to the low levels of virus reads ( S1A Table ) . Translation efficiencies were calculated by dividing the raw RiboSeq densities for each ORF by the decumulated RNASeq densities for the corresponding mRNA . Note also that initiation and termination peaks were excluded from the RiboSeq density calculations ( see Methods ) . The 3′ ORFs 2 , S , 5 , E , M and N are all translated at comparable efficiencies ( Fig 4B ) . The translation efficiency of E was at the lower end , presumably due to it not being the 5′ proximal ORF on its transcript ( mRNA5 ) [34] . The translation efficiency of N was also at the lower end . The translation efficiency of ORF1a/1ab was , in comparison to the 3′ ORFs , very low . As mentioned above , this could be due to a proportion of gRNA being present in an untranslatable pool , perhaps as RTCs or RNPs destined for packaging , but may also indicate a real restraint on ORF1a/1ab translatablity ( see below ) . The gRNA translation efficiency calculated in this way was low even at 2 . 5 h p . i . ( repeat 2 , ORF1a translational efficiency ~0 . 11 ) . On the assumption that gRNA will not be directed to a packaging pathway at early time points , this suggests that incoming and early synthesis gRNA is largely involved in RNA synthesis , or is , indeed , inherently poorly translated . It should be noted that technically these calculations do not measure translational efficiency absolutely , as ribosome occupancy may also be affected by translational speed ( though , when averaging over ORFs , this effect is thought to be generally quite slight; [8] ) . Further , as new transcripts enter the translation pool , there may not have been time to establish steady state ribosome loading on any particular transcript , while , at late time points , translational efficiencies may be below their optimal values due to saturation of the host cell protein synthesis machinery . Transcript abundances can be calculated from the decumulated RNASeq densities ( as above ) or , independently , from the relative abundances of RNASeq reads spanning each leader/body junction . Such “chimeric” reads ( where the 5′ part maps to the leader sequence , and the 3′ part maps just downstream of a body TRS ) were not included in the initial mapping to the virus genome ( Fig 1B ) , but were identified subsequently ( see Materials and Methods ) . Fig 5 compares mRNA abundances estimated using these two methods . The “TRS method” has the advantage that it avoids the potential inaccuracies introduced by decumulation but may be more subject to fragmentation , ligation and PCR biases due to the relatively short window in which to calculate a mean RNASeq density . Nonetheless there is a good correlation between the two estimates ( R2 = 0 . 99 ) . In MHV , the consensus for canonical TRSs is UCUAAAC with minor exceptions being UCUAUAC for mRNA2 and UCCAAAC for mRNA6 [35–38] . A variable number of tandem copies ( two in MHV-A59 ) of UCUAA are present at the leader junction site , while an imperfect copy of UCUAA precedes the canonical UCUAAAC at several body junction sites ( S2 Table ) . Heterogeneity in the number of copies of the pentanucleotide has previously been observed to occur in mRNA6 for MHV-A59 , and both mRNA6 and mRNA7 for MHV-JHM , and this is presumably due to heterogeneity in the site of re-annealing following a polymerase jump [39] . Consistent with this , we also observed significant usage of a junction site 5 nt upstream of the canonical site for mRNA6 ( 13–17% of mRNA6 transcripts ) ( S3 Table ) . We also observed this phenomenon for mRNA7 ( 0 . 5–0 . 8% of mRNA6 transcripts ) . The greater usage for mRNA6 is likely due to it having an imperfect pentanucleotide UCCAA at the canonical junction site but a perfect pentanucleotide UCUAA 5 nt upstream; in contrast , other mRNAs have a better pentanucleotide match at the canonical site than at the site 5 nt upstream ( S2 Table ) . For mRNAs showing such heterogeneity , the summed values were used for Fig 5 . For mRNA7 , where the upstream pentanucleotide is CCUAA instead of UCUAA , we observed that the first nucleotide could be templated either by the body sequence ( i . e . 'C'; ~40% ) or by the leader sequence ( i . e . 'U'; ~60% ) ( S3 Table , nt 29653 sequences ) . We also observed many non-canonical leader/body chimeric sequences ( S3 Table ) , though even the most frequent were present at ≤20% the level of leader/body chimeric reads for mRNA2 ( the least abundant canonical mRNA ) . The coronavirus polymerase is known to engage in promiscuous jumping [39–41] and there is no reason to suppose that the additional transcript species generated this way are functionally relevant . Two of the most abundant ( genomic coordinates 41 and 34 in S3 Table ) involved apparent backward jumping by the polymerase ( although inter-template jumping is another possibility ) . The sequences at non-canonical junctions often partly resembled canonical TRSs ( e . g . UCUAAAa at nt 41 , UCUcAAC at nt 34 , cCUAcuu at nt 22483 , UCcAAgc at nt 27106 and UgUAAua at nt 28847; canonical TRS nucleotides in upper case ) . In cases where the nucleotides at +1 to +2 in the genome sequence differed from UC , the RNASeq read generally contained nucleotides templated by the genome sequence rather than the UC in the leader sequence ( e . g . CC instead of UC for the nt 22483 junction ) , although there were exceptions ( e . g . UC instead of AA for the nt 22582 junction ) ( S3 Table ) . This latter site , AAUAAGC , aligns with a TRS previously identified for an HE mRNA in the JHM strain of MHV [38] . The sequence in MHV-JHM is AAUAAAC , differing from the MHV-A59 sequence by a G to A substitution . An HE mRNA has not been observed for MHV-A59 and this is likely due to the greater deviation from the consensus TRS , UCUAAAC , in this strain [38 , 42] . Although we observe some usage of this site in our sequencing , the levels are extremely low—just 3 . 6–4 . 6% those of mRNA2 ( the least abundant canonical mRNA ) . Fig 6 compares the translation efficiencies at 5 h p . i . of virus and host CDSs . The former are as described previously in Fig 4B . The latter are calculated on a per gene ( rather than per transcript ) basis , using RNASeq and RiboSeq reads contained entirely within annotated CDS regions ( i . e . excluding 5′ and 3′ UTRs and also RPFs accumulating at or near to initiation or termination sites ) , and , like the virus values , are expressed relative to the mean levels for the cell ( due to normalization by library size ) . The analysis shows that the virus translation efficiencies fall within the general range of those of host genes ( except for ORF1a/1b which have particularly low translation efficiencies; see above ) indicating that virus transcripts are not preferentially translated during virus infection . Instead , massive production of virus proteins ( in particular the N protein ) is achieved through high levels of transcription . To study virus RNA synthesis and translation during the earliest stages of infection , we did high MOI ( ~200 ) infections to maximize the number of virus reads in the libraries . The composition of the high MOI libraries is summarized in S1B Table and S12A Fig . Fig 7A shows the distributions of RiboSeq and RNASeq reads on the virus genome at 1 h p . i . ( where 0 h p . i . , is the time at which the inoculating virus is removed ) . A 5′ to 3′ decreasing gradient in RPF density is visible within ORF1a in the RiboSeq CHX density profile , while very few RPFs were found within ORF1b , indicating that , at 1 h p . i . , ribosomes have only had time to translate part of the 4470-codon ORF1a . This does not indicate the translate rate per se , as time is also required for uncoating , recruitment of ribosomes , translation of a uORF on the gRNA ( see below ) , and potential delays with initiation and reinitiation ( see also below ) . In the RiboSeq HAR data , a clear trough in RPF density is visible after the ORF1a initiation peak , followed by higher density further downstream in ORF1a . The trough reflects run-off of elongating ribosomes in the three minutes between addition of HAR ( which inhibits new initiation events ) and CHX ( which freezes the ribosomes ) . Taking the width of the trough as ~750 codons , this gives an elongation rate of 4 . 2 amino acids s−1 , similar to that determined previously in mouse embryonic stem cells ( 5 . 6 amino acids s−1 ) [8] . Despite the very high MOI , virus RNA levels were low except , unexpectedly , in the N region where the mean RNASeq density was ~27 times that in the ORF1a region . To test whether this might be due to contamination from late time-point samples , we compared the length distribution of reads in the N region with the length distribution of reads mapping to host mRNAs for the same sample ( Fig 7B , right panel; red and green lines , respectively ) . The two distributions were very similar . In contrast , the length distributions of virus-derived reads from the 5 and 8 h p . i . RNASeq time points ( from repeat 2 which was co-processed with the high MOI libraries ) were different in shape ( Fig 7B , right panel; grey lines ) . While it is impossible to definitively rule out contamination in this way , the analysis suggests that the RNASeq density in the N region at 1 h p . i . is not a result of contamination . Since , for mRNA7 , negative-sense RNA is present at >0 . 1% of positive-sense RNA at 2 . 5 , 5 and 8 h p . i . ( Fig 4 ) , the absence of appreciable levels of negative-sense reads mapping to the N region in the high MOI 1 h p . i . sample ( 3 negative-sense compared with 48 , 429 positive-sense reads; 0 . 006% ) also argues strongly against the positive-sense reads being inter-library contaminants . The near-complete absence of negative-sense reads also argues against this phenomenon being a result of early synthesis . Moreover , the absence of equivalent RNASeq density in the leader region ( cf . Fig 2 ) argues against the density in the N region deriving from bona fide mRNA7 transcripts . Closer inspection revealed a number of a relatively abundant chimeric reads suggesting a mosaic structure of rearranged N-ORF sequences , reminiscent of defective interfering ( DI ) RNAs [43 , 44] . However , since coronavirus DI RNAs are expected to include parts of the 5′ end of the genome and a packaging signal , and only arise after multiple high-MOI passages , we believe the N ORF transcripts we have identified must represent a novel class of packaged transcripts . An alternative , albeit unlikely , explanation is that the excess 3′ density may be a result of selective degradation ( either natural or artifactual ) of ~96% of the input gRNA . Relative to RNA levels , very few RPFs mapped to the N ORF region and we were unable to ascertain whether or not they resulted from contamination from other samples as , in contrast to RPFs from ORF1a , their length distribution only partly matched the length distribution of host RPFs ( Fig 7B , left panel , red line ) . Using these RPF counts , the N ORF translation efficiency ( normalized to total virus RNA and host mRNA ) was calculated to be only 0 . 0005 , compared to values in the range 1 . 1 to 1 . 7 at the 2 . 5 , 5 and 8 h p . i . timepoints , indicating that the early timepoint N ORF RNA revealed by RNASeq is not , or only barely , translated . The −1 PRF signal that facilitates expression of MHV pp1ab comprises two elements , a heptanucleotide slippery sequence ( U_UUA_AAC ) , identical in all known coronaviruses , and an RNA pseudoknot structure located a few nucleotides downstream [5 , 45 , 46] ( Fig 8A ) . During translation of the gRNA , elongating ribosomes either terminate at the ORF1a stop codon , yielding pp1a , or frameshift on the slippery sequence to translate ORF1b , yielding pp1ab . Frameshifting likely provides a fixed ratio of translation products for assembly into a macromolecular complex [47 , 48] . Studies of frameshifting using reporter constructs expressed in transfected cells or through in vitro translation of synthetic mRNAs have indicated that the efficiency of the process in coronaviruses is generally in the region of 20–45% [4 , 5 , 49–51] However , the actual efficiency in the context of virus infection has not been previously determined . Simplistically , one can calculate this value by dividing the RiboSeq density in ORF1b by the density in ORF1a . However , in principle , RiboSeq density represents the quotient of expression level and translational speed so the above calculation assumes that , on average , translation speed is the same in ORFs 1a and 1b and that translation is steady state . Such a calculation is , therefore , invalid immediately after infection ( as ribosomes begin to translate ORF1a of the input gRNA but have not yet reached ORF1b; Fig 7 ) and may also be compromised if newly synthesised gRNA entering the translation pool represents a significant fraction of the gRNA undergoing translation . Thus , we measured the frameshifting efficiency at 5 h p . i . , calculating values of 48% for repeat 1 , and 70% for repeat 2 ( Fig 8B ) . The former value ( 48% ) is similar to previous in vitro measurements of the MHV frameshifting efficiency ( 40% ) [5] . As the infection appeared to be more advanced in repeat 2 ( Fig 3 ) , it is possible that the higher level measured ( 70% ) is a consequence of depletion of the host cell protein synthesis resources , e . g . exhaustion of initiation factors ( including free ribosomes ) could decrease the density of ribosomes in ORF1a as elongating ribosomes run off , and a partial exhaustion of elongation factors could slow the establishment of a new steady state . We also measured the frameshifting efficiency by means of transfected reporter constructs . We began by cloning a 100-bp fragment including the MHV frameshift signal ( Fig 8A ) into a dual luciferase frameshift reporter plasmid ( pDluc; [52 , 53] ) between the Renilla ( Rluc ) and firefly ( Fluc ) luciferase genes . In this plasmid , frameshifting permits expression of Fluc as a fusion with Rluc ( analogous to the expression of MHV pp1ab ) , while failure to frameshift results in expression of Rluc alone . Frameshifting efficiencies were calculated from the ratio of Fluc activity to Rluc activity , normalized by a control construct in which an extra C residue was inserted immediately downstream of the slippery sequence to place Rluc and Fluc in the same reading frame . The well-studied coronavirus frameshifting signal from Avian coronavirus , infectious bronchitis virus ( IBV ) served as a positive control , alongside a lower efficiency control ( the gag/pol −1 PRF signal of HIV isolate HXB2 ) [54 , 55] . The MHV frameshifting efficiency was found to be 38% in 17Cl-1 and 45% in BHK-21 cells , and similar in both instances to that of IBV ( Fig 8C ) . These data suggest that frameshifting in coronaviruses is not specifically modulated by virus infection , with the difference seen in the more advanced infection of repeat 2 likely due to the non-specific effects mentioned above . The relevance of ribosomal pausing to the mechanism of −1 PRF has long been a subject of debate [56–58] . Frameshift signal-associated pauses have been documented in a number of in vitro assays [59–64] , but there is , as yet , little evidence for a causal relationship , and pausing has not been examined in infected cells . We therefore looked to see whether there was an accumulation of RPFs at the MHV frameshift site . In the initial RiboSeq time courses we failed to see significant pausing at the frameshift site . However , reasoning that the frameshift-stimulatory pseudoknot beginning 6 nt 3′ of the slippery heptanucleotide U_UUA_AAC would be partly inside the mRNA entrance channel at the onset of frameshifting , and might , due to its compact structure be somewhat resistant to RNase 1 digestion , we considered the possibility that frameshift-associated pauses may generate longer RPFs , which would be excluded from the profiling analysis as a result of gel size selection ( 28–34 nt ) . Thus we prepared new libraries from the 5 h p . i . repeat 2 RiboSeq CHX samples ( see S1C Table for composition ) using a larger gel slice ( nominally 28–80 nt ) . However , even in these samples we failed to see noticeable pausing at the frameshift site ( S11 Fig ) . Although we failed to identify significant pausing at the frameshift site , there were other sites at which RPFs accumulated to a much higher level than at neighbouring sites . We frequently observed such accumulations at initiation sites ( possibly an artifact of CHX treatment; [8] ) ( Fig 1 and Fig 2 ) , but also at internal sites within ORFs . Besides ribosome pausing , fluctuations in RPF density may occur as a result of nuclease , ligation , and PCR biases . The latter two occur also for RNASeq , whilst in RNASeq nuclease bias is replaced by fragmentation bias . Following [65] , we compared the distributions of variability in RiboSeq and RNASeq densities within ORF1a , which revealed that RiboSeq densities were more variable than RNASeq densities ( Fig 9A ) , with the extra variability presumed to be a result of fluctuations in ribosome progressivity . We focused on two of the highest RiboSeq density peaks in the ORF1a region ( blue arrows in Fig 9B ) . RPFs at the second of the two pause sites , located in the nsp6 region , have 5′ ends that map almost exclusively to nt 11366 which , unusually , corresponds to the second codon position ( Fig 9C , right ) . The 3′ end positions of these RPFs were , as is normal , more variable , with the most abundant 3′ ends mapping to nt 11393–11394 for repeat 1 and nt 11394–11395 for repeat 2 , giving read lengths of 28–29 and 29–30 nt which are within the typical range for the respective samples ( S2 Fig and S3 Fig ) . For these samples , RiboSeq CHX 5 h p . i . repeats 1 and 2 , 64% and 66% , respectively , of host mRNA RPFs have 5′ ends mapping to codon position 1 , with only 7% and 8% mapping to codon position 2 . The reason for the deviation at the pause site is unknown but may be a result of “tension” within the mRNA or perturbation of the ribosome conformation [66] . Due to the unusual codon position of the 5′ end , it was not possible to definitively predict the P-site position of ribosomes at this pause site , but it is more likely to be at nt 11377 to 11379 ( AAA codon ) than at nt 11380 to 11382 ( CAG codon ) as the lengths of the most abundant reads ( 28–29 and 29–30 nt in repeats 1 and 2 , respectively ) are more consistent with reads being 1 nt shorter than normal at the 5′ end , rather than 2 nt longer . The nascent peptide sequence ( i . e . peptide sequence within the ribosome exit tunnel that could potentially affect pausing ) here is…IKHKHLYLTMYIMPVLCTLFYTNYLVVYKQ ( P-site amino acid underlined ) . An additional smaller peak was apparent 30 nt upstream ( in repeat 1 ) and potentially corresponds to a following ribosome stacking behind a proportion of paused ribosomes . RPFs at the first of the two pause sites , located in the nsp3 region , have 5′ ends that map to nt 4704 in repeat 1 but nt 4699 in repeat 2 ( Fig 9C , left ) . However , this 5 nt difference is made up in the length of the reads ( top three read lengths 28 , 27 and 29 nt in repeat 1 , but 33 , 34 and 32 nt in repeat 2 ) so that the 3′ ends of the RPFs map to similar positions in both repeats . Incidentally , this difference in 5′ end position between the two repeats makes it highly unlikely that the peak is an artefact of ligation bias . In general the nuclease trimming seems to be less stringent in repeat 2 than in repeat 1 ( host mRNA RPF lengths peak at 29–30 nt in repeat 1 but at 30 nt in repeat 2; S2 Fig and S3 Fig ) . The pause site read length in repeat 2 is unusually long , indicating that the extra ~5 nt at the 5′ end are , for some reason , partially protected , resisting trimming in repeat 2 but not in the more stringently trimmed repeat 1 . The nature of this protection ( scrunching of extra mRNA into the mRNA exit channel , formation of an RNase-resistant RNA structure 5′-adjacent to the ribosome , conformational changes in the ribosome , or an additional ribosome/mRNA-associated protein factor ) and whether and how it is linked to pausing remain undetermined . The nascent peptide sequence is…EKCQVTSVAGTKALSLQLAKNLCRDVKFVT ( P-site amino acid underlined ) . The nascent peptides at both pause sites lack the E- or P-site prolines or A-site GAA codons that are commonly associated with pausing in ribosome profiling meta-analyses [8] , though many diverse nascent peptides are also known to perturb ribosome progressivity [67 , 68] . As an alternative possibility , we analysed the RNA downstream of the pause sites for evidence of stable RNA structures that could induce pausing , but nothing was apparent . An alternative explanation is that these pauses are induced by trans-acting factors , e . g . RNA binding proteins , or chaperones of the nascent peptide . Coronaviruses induce substantial membrane rearrangements in the infected cell , including formation of a reticulovesicular network composed of two types of membrane modifications , double-membrane vesicles ( DMVs ) and convoluted membranes ( CM ) . The reticulovesicular network is contiguous with the endoplasmic reticulum ( ER ) and is the site of virus RNA synthesis [69 , 70] . Nsp3 , nsp4 and nsp6 are integral membrane proteins whose topology has been determined in vitro for SARS-CoV and MHV [71–74] , and , in the case of SARS-CoV , have been shown to be necessary and sufficient for double-membrane vesicle formation [75] . Nsp3 is the largest protein encoded in MHV ORF1a and contains multiple domains , including two small ubiquitin-like domains ( Ubl1 and Ubl2 ) , two papain-like cysteine proteinase domains ( PLP1 and PLP2 ) , a poly-ADP-ribose-binding activity ( ADRP domain ) , the newly determined domain preceding Ubl2 and PLP2 ( DPUP; [76] ) , the nucleic-acid binding domain ( NAB ) , the betacoronavirus marker ( G2M ) , a transmembrane domain ( TM ) and domain Y ( Fig 10A ) . The apparent ribosome pause occurs within the sequence DVKFVTNAC ( P-site at pause underlined; Fig 10A ) which is located between the ADRP and the DPUP domains . We investigated the nsp3 pause in vitro in rabbit reticulocyte lysate ( RRL ) translations using edeine assays [60] . Initially , a cDNA fragment comprising the first 1 , 125 residues of nsp3 ( nsp3; excluding the NAB , G2M , TM and Y domains ) was cloned into pcDNA3 . 1 and synthetic mRNAs translated in RRL for 3 min prior to addition of the translation initiation inhibitor edeine . Incubation was continued , samples withdrawn at the indicated times post-edeine addition and translation products separated on a 10% SDS-PAGE gel ( Fig 10B ) . To accurately mark the position of the predicted pause product , a control mRNA in which a UAA stop codon had been introduced at the pausing A-site was also translated . As seen in Fig 10B ( marked by a red asterisk ) a distinct translational pause was observed during translation of nsp3 , migrating at the same position as the “pause control” and accumulating and then diminishing as translation proceeded . To more closely define the stalling sequence , a region encoding 46 amino acids of nsp3 including the putative pausing peptide was cloned into the influenza PB1 reporter gene in transcription vector pPS0 ( Fig 10C ) [60] and edeine assays performed as above . Once again , a clear ribosomal pause was evident ( Fig 10C , red asterisk ) . The nsp3 sequence within pPS0 includes five upstream positively charged amino acids ( four Lys and one Arg ) and one aromatic residue ( Phe ) that could potentially contribute to pausing [77] . These residues were mutated to alanine sequentially and incrementally ( pPS0-nsp3 mutants ) , such that in Mut 1 , Lys-Phe adjacent to the pausing site was changed to Ala-Ala , Mut 2 had these changes plus Arg to Ala , and so on , as shown in Fig 10D . Edeine assays were performed and a single time point ( 20 min ) from each mutant analysed by SDS-PAGE . As seen in Fig 10D , pausing was obviated in Mut 3 , Mut 4 and Mut 5 , indicating that the residue substituted by alanine Mut 3 is likely to be a major contributor to the ribosomal pause . A complete time-course of pPS0-nsp3 Mut3 confirmed the lack of pausing ( Fig 10E ) . Fig 11A displays RiboSeq and RNASeq densities for the 5′ region of the genome at 5 h p . i . The leader sequence , 5′ of the leader TRS ( orange , Fig 11A ) , is present on all mRNAs so reads mapping to this region may derive from any mRNA , although most are expected to derive from the highly abundant mRNA7 . The plot excludes “chimeric” reads ( i . e . reads that span a TRS transcriptional discontinuity ) , so the RNASeq density drops close to the TRS site and the same is also expected to happen for RiboSeq . Probing of initiation sites through harringtonine treatment revealed unexpectedly that a substantial number of reads accumulate at or near the 5′ end of the leader , despite an absence of AUG codons . These 5′-proximal reads have a tight length distribution characteristic of true RPFs ( Fig 11B; left panel ) so are likely to be bona fide RPFs rather than some form of contamination . The 5′ portion of the leader contains a number of potential near-cognate non-AUG initiation codons , but most of the harringtonine reads do not obviously map to these . For example , the most abundant RPF position corresponds to a GCG codon ( genome coordinates 16–18 ) ; initiation at this point would generate a 12 amino acid peptide , but it should be noted that GCG is not a recognised non-AUG initiation codon . Elongation profiling with cycloheximide revealed a similar pattern of reads in the 5′ part of the leader but also a larger peak on a UUG codon close to the 3′ end of the leader sequence ( Fig 11A ) . UUG is a known , albeit quite inefficient non-AUG initiation codon [8 , 78] and , in this case , it is also in a poor initiation context ( cucUUGuag; in mammals contexts with an A at −3 , or a G at −3 and a G at +4 , may be regarded as “strong”; [79] ) , so only a very small proportion of ribosomes would be expected to initiate here . Consistent with this , the HAR peak is very small compared to that seen at the N initiation codon ( 1 . 4%; Fig 11C ) ( though similar in magnitude to initiation peaks at the uORF and ORF1a on the genomic RNA; Fig 11A ) . Interestingly , the difference between the UUG peak and the N initiation peak was much less for the CHX samples ( 69%; Fig 11C ) . The reasons for this are unknown , but may be related to the UUG codon being immediately followed by a termination codon , with the peak potentially being derived from both initiation and termination pauses ( UUG in P-site , UAG in A-site ) . We note also that , on mRNA7 , the UUG codon is 31 nt upstream of the N initiation codon , so that initiation at N might lead to stacking of ribosomes on the UUG codon , potentially increasing initiation on this ostensibly very weak start codon . Downstream of the leader TRS but upstream of ORF1a , is a single , short AUG-initiated uORF that is present in many coronaviruses and believed to play a role as a regulator of genomic RNA translation , virus replication and pathogenesis [80] . Upstream ORFs are present in ~40% of mammalian mRNAs and have been shown generally to cause repression of translation of the downstream ( main ) ORF [81 , 82] . We observed RPFs mapping specifically and in-frame to the uORF , confirming that it is translated . Indeed , at 5 h p . i . it appeared to be translated as efficiently as ORF1a ( Fig 12A ) despite its poor initiation context ( uccAUGc; cf . auaAUGg for ORF1a ) suggesting that it inhibits ribosomal access to ORF1a . This effect appeared less pronounced at early time points , suggesting a potential role for temporal regulation of replication protein synthesis ( Fig 12A , bottom panel ) . Interestingly , we observed the greatest density of RPFs on the second codon ( proline ) rather than the first codon ( methionine ) of the uORF , both for HAR and CHX-treated samples . Prolines are often associated with ribosome pausing due to their restrained geometry in the decoding centre and/or ribosome exit tunnel [8 , 83 , 84] . To see if N-terminal Met-Pro was associated with ribosomal pausing on other mRNAs , we compared mean ribosome profiles for host mRNAs with CDSs beginning with AUG-CCN with mean ribosome profiles for generic host mRNAs and found that , particularly under conditions of virus infection , ribosomes tend to pause more at the second codon in the former ( Fig 12B ) , although the ratio of ribosome occupancy between the AUG and CCN averaged over host mRNAs was less extreme than is the case for the virus uORF . It should be noted that , although presence of the uORF is conserved in 17 of 18 NCBI betacoronavirus RefSeqs , CCN occurs as the second codon in only six of these . Fig 13 shows histograms of RiboSeq ( CHX and HAR ) and RNASeq reads that map near to the 5′ ends of the HE , 4 and 5 ORFs . Again , “chimeric” leader/body reads spanning transcriptional discontinuities at the TRS sites are excluded from these plots . In the laboratory-adapted strain MHV-A59 , the HE ORF is disrupted by a premature termination codon ( red diamond , Fig 13A ) [85] , and , furthermore , the TRS upstream of HE in MHV-A59 is defective ( open green box , Fig 13A ) [38] , leading to only very low levels of HE mRNAs ( see above ) . Although ribosomes were observed to initiate at the authentic HE AUG codon , upstream of the premature termination codon ( Fig 13A , HAR ) , very little RiboSeq density was observed downstream of the premature termination codon . Translation of the annotated HE ORF ( i . e . the long 3′ fragment; grey , Fig 1A and Fig 13A ) was negligible , consistent with the presence of numerous AUG codons in other reading frames downstream of the “authentic” HE start codon , which would be expected to inhibit ribosomal access to the 3′ fragment of HE . The low level of initiation noted at the “authentic” HE start codon is likely explained by the very low levels of HE mRNA production inferred from the observation of a few RNASeq reads crossing the HE leader/body transcriptional discontinuity ( see above and S3 Table ) , since leaky scanning on mRNA2 is unlikely to allow access to HE due to the large number of intervening AUG codons . Similarly , in MHV-A59 the natural ORF4 coding sequence is split by a frameshift mutation into a short 5′ ORF4a ( pale yellow , Fig 13B ) and a longer 3′ ORF4b ( grey , Fig 1A and Fig 13B ) [86] . Again , we observed ribosomes initiating at the ORF4a AUG codon ( Fig 13B , HAR , the blue peak is in the ORF4a frame ) , but very little RiboSeq density in the annotated ORF 4b . Ribosome access via leaky scanning to ORF4b would be inhibited not only by the ORF4a AUG but also by an additional out-of-frame AUG codon ( Fig 13B ) . Upstream of ORF4a , but downstream of the mRNA TRS junction , a low level of initiation appeared to occur on an AUU codon ( RiboSeq , HAR , orange peak ) . Ribosomes initiating here would translate a 15-codon ORF resulting in the peptide MYSILIATWPRKRQS ( assuming the initiating codon AUU is decoded as Met ) . A similar ORF is present in other strains of MHV . Upstream of ORF5 , we identified an alternative initiation site at a CUG codon ( Fig 13C , HAR , blue peak ) which may have some bearing on the mechanism of expression of the E ORF , which lies downstream of ORF5 on the bicistronic mRNA5 ( Fig 13C ) . The CUG codon in question is in the same reading frame as the upstream ORF4 and initiation here would result in translation of the last 13 codons of ORF4 with peptide sequence MVVHILLRHCPGI ( assuming the initiating codon CUG is decoded as Met ) . The CUG is downstream of the mRNA5 TRS and appears to be utilized only on this mRNA as the RiboSeq density on the upstream part of the defective ORF4 ( see above ) is negligible . The level of initiation at the CUG was comparable to that at the ORF5 AUG ( Fig 13C ) and translation of this short ORF might be utilized to shunt a proportion of ribosomes past the ORF5 AUG codon . We also observed utilization of an AUG codon just downstream of the ORF5 AUG codon ( Fig 13C , HAR , orange peak , six-codon ORF , peptide sequence MDLACE ) . Access to this AUG is likely facilitated by the poor initiation context of the ORF5 AUG ( cauAUGa ) . After translating a very short ORF ( e . g . <30 codons ) , the small subunit of the ribosome can remain associated with the message , resume scanning , and reinitiate translation at a downstream AUG codon [87] . After translation of a short ORF , the 40S subunit of the ribosome is not immediately competent to reinitiate , but becomes competent after scanning for some distance . Thus , after translating the short CUG-initiated ORF , it is possible that the post-termination 40S subunits can scan past the five AUG codons present within the first 44 nt of ORF5 ( green +s , Fig 13C ) , before becoming initiation competent and able to reinitiate translation at the next available AUG codon , which is the initiation codon for the E ORF some 290 nt downstream ( Fig 13C ) ( see also [88] ) . The presence of an upstream CUG-initiated short ORF is preserved in other strains of MHV , though most ( other than MHV-A59 ) also have a separate AUG-initiated ( albeit in a weak initiation context ) short ORF that could be used to shunt even more ribosomes past the ORF5 initiation codon . These viruses also preserve a conserved absence of AUG codons ( in any reading frame ) throughout ORF5 except for the 5′-most 44 nt ( where there are from one to five AUG codons , depending on species and strain ) and the very 3′ end where the E ORF AUG is situated [88] . In contrast , related viruses such as Betacoronavirus 1 ( including bovine coronavirus and equine coronavirus ) have AUG codons spaced throughout ORF5 , but produce a separate mRNA for E protein expression so that bicistronic expression from the same mRNA as ORF5 is not required [89 , 90] . It should be noted , however , that expression of E ( but not protein 5 ) can occur from artificial reporters in which an additional ORF is added upstream of ORF5 , and therefore appears to involve internal ribosome entry [34 , 91] . It is possible that multiple strategies are used to enhance E expression . Alternatively , presence of the CUG-initiated uORF could simply be to downregulate production of protein 5 . A long internal ORF ( I ) is present within the N ORF of MHV and many other coronaviruses , encoding a largely hydrophobic polypeptide that is thought to confer a minor growth advantage to the virus [92 , 93] . As shown in Fig 14A , however , HAR profiling did not reveal an initiation spike for the I protein of MHV-A59 , suggesting that it might not be expressed . However , western blotting of infected-cell lysates using anti-N and anti-I sera revealed unambiguous expression of N and I from 5 h p . i . ( Fig 14B ) . To further confirm expression of the I protein , the N coding sequence was cloned into pcDNA . 3 and the mRNA translated in RRL ( Fig 14C ) and immunoprecipitated ( Fig 14D ) with anti-N and anti-I sera , and , as a negative control , anti-S serum . As shown , both N protein ( 50 kDa ) and I protein ( 23 kDa ) were expressed from the synthetic N mRNA , with I produced at a level of about 2% of N . Although we were unable to obtain strong evidence for the expression of I from the profiling data , a comparison of the phasing of RPFs ( a ) in the region where the I ORF overlaps the N ORF , and ( b ) in the region of the N ORF downstream of the I termination codon , revealed in the former a slight excess of RPFs with 5′ ends mapping to the second position of N-frame codons ( blue in Fig 14E; upper panels ) . The excess is consistent with 6–12% of ribosomes translating the +1 ( i . e . I ) reading frame in this region . It is possible that 5′ leader sequence present in mRNA7 ( e . g . the UUG-initiated uORF ) , but absent from the pcDNA . 3-transcribed mRNA , promote access to the I ORF . To ensure that the phasing difference was not due to a single RPF peak ( as individual peaks can sometimes map to a non-standard phase; cf . Fig 9C ) , mean phasing was also determined in a 55-codon sliding window and , consistent with the previous result , the proportion of RPFs mapping to the second position of N-frame codons ( blue in Fig 14E; lower panels ) was found to decrease abruptly around the I ORF stop codon . A caveat to note is that , in MHV-A59 , an upstream AUG ( bold ) is present in the I frame followed by a stop codon ( asterisk ) prior to the “designated” I AUG codon ( underlined;…MPVAEAPL*TALVMESSRRP; both AUGs are in a strong context ) . In some related virus sequences , the stop codon is replaced with a sense codon such that I is probably initiated from the upstream AUG . Thus , MHV-A59 may be somewhat defective with regards to I expression . We have used ribosome profiling to investigate virus gene expression kinetics , relative translational efficiencies , ribosomal frameshifting , ribosome pausing , and uORF translation in cells infected with MHV , a representative of the betacoronavirus genus of the coronavirus family of RNA viruses . These studies provide the highest resolution data on coronavirus translation to date . Using parallel RNASeq data , we examined the kinetics of virus replication and transcription , the relative abundances of different transcripts , and the degree of promiscuous polymerase jumping . We explored a number of data quality issues that can arise when applying ribosome profiling to the study of RNA viruses that replicate to high titres in cell culture and describe ways to bioinformatically assess and quantify potential contamination . Despite identifying low levels of different types of contamination , we were able to use impartial tests to avoid drawing incorrect conclusions from our data . Viruses present particular challenges in profiling experiments . One of these is library contamination , which in this study may have been derived from two sources . The first was low-level contamination of one sample by another , a problem that is compounded by the high levels of virus RNA synthesised in infected cells at late time points . We took precautions to avoid this source of contamination , including the use of designated work spaces , buffers and equipment , and avoiding parallel processing of early and late time points where possible . Potentially , contamination may also have been introduced through the multiplex adaptor sequences . In the relatively small number of published studies on virus ribosome profiling , data from mock-infected samples and tests for contamination are often not reported , so the level of contamination suffered by others is uncertain . A second potential source of contamination could derive from RNPs comprising virus or host mRNA complexed with virus or stress-induced host RNA binding proteins . Such RNPs might co-sediment with ribosomes during the sucrose cushion centrifugation step and contaminate RiboSeq libraries . Although we were mindful of the possibility of such contamination , we found little evidence for it occurring as a result of MHV infection . An increased 3′ UTR RiboSeq ( CHX ) density was not apparent until 8 h p . i . ( when the plateau of virus production has been reached and virtually all cells are involved in extensive syncytium formation ) and , even then , the read length distributions were similar to those of mock-infected cells; suggesting that the increased 3′ UTR occupancy was as much due to bona fide RPFs as contaminating RNPs . The former could be due to depletion of ribosome recycling factors resulting in increased amounts of unrecycled post-termination ribosomes accessing the 3′ UTR [31] . The high level of phasing in our RiboSeq data ( S4 Fig ) allowed us to carefully assess contamination issues , and our observations reinforce the essentiality of basic data quality checks ( e . g . S4–S10 Figs ) in profiling studies . Despite these challenges , the profiling and RNASeq analysis of MHV infection still showed itself to be a powerful tool to investigate specific aspects of MHV replication at high resolution . The kinetics of virus transcription in MHV-infected cells as observed through RNASeq were consistent with previous studies [33 , 94–96] . Up to 2 . 5 h p . i . , there was little amplification of positive-sense RNA , whilst negative-sense RNA levels rose from undetectable to about 0 . 1% of total virus RNA and host mRNA . Subsequently , positive-sense RNA levels increased rapidly—with the accumulation of negative-sense RNA plateauing at about 5 h p . i . —such that , at late time points , the former comprised 80–90% of total virus RNA plus host mRNA , while the latter comprised only ~0 . 3% . Despite differences in abundance , the patterns of expression of positive and negative-sense RNAs were similar , including high densities in the leader region , consistent with discontinuous transcription occurring during negative-strand synthesis [23] . The measurement of decumulated RNASeq densities and the analysis of specific RNASeq leader/body chimeric reads at TRSs determined the relative abundance of mRNAs at 5 h p . i . to be mRNA7 > mRNA6 > mRNA1/mRNA5/mRNA3 > mRNA4/mRNA2 . An earlier study of MHV-A59 transcription using [32P] pulse labelling in the presence of actinomycin D provided a similar but slightly different order ( mRNA7 > mRNA6 > mRNA5 > mRNA1/mRNA3/mRNA4 > mRNA2 ) [97] although it should be noted that , while mRNAs 7 and 6 are nearly always the most abundant subgenomic transcripts , the relative abundances of the other transcripts can vary greatly between different isolates , strains and mutants of MHV [42 , 97] . The translation of virus proteins was detectable at a very early stage of infection . Indeed , using a high MOI infection , we were able to visualize input gRNA translation at 1 h p . i . , a stage when the majority of ribosomes had not yet reached the pp1b ORF . Using RiboSeq ( HAR ) data at this time point , we were able to estimate a translation rate of 4 . 2 amino acids s−1 , consistent with previous estimates for mammalian systems [8] . During the course of infection , we found that virus mRNAs 2–7 were translated with generally similar efficiencies and , importantly , were not preferentially translated relative to host mRNAs . Rather , the synthesis of large quantities of virus proteins , especially N , is achieved through high levels of transcription ( note that , due to library normalization , the quotient of RiboSeq and RNASeq does not inform on global virus-induced host shut-off , which is likely to be occurring at late time points of infection [98] ) . The virus genomic RNA , however , appears to be poorly translated , as judged by the quotient of RiboSeq and RNASeq . During infection , much of the gRNA pool may , of course , be unavailable for translation . At earlier time points , it is , perhaps , sequestered in replication-transcription complexes; whereas at later time points , it may also be involved in packaging complexes . At 2 . 5 h p . i . , when gRNA is unlikely to be a substrate for packaging , its translational efficiency was still low , but at this point in the replication cycle , the formation of replication-transcription complexes would preclude the massive amplification of viral RNA that takes place between 2 and 6 h p . i . [33] . It may also be the case that the pp1a and pp1b ORFs on the gRNA are inherently poorly translatable , e . g . due to translation of the uORF ( see below ) inhibiting ribosomal access to ORF1a . We also observed significant amounts of RNASeq reads mapping to the N ORF region at 1 h p . i . , a time point at which negative-sense RNASeq reads were essentially absent . This suggests that the N ORF RNA is not newly synthesised . Further , the absence of similar amounts of RNASeq density in the leader region , together with a very low translation efficiency , suggest that the N ORF RNA does not correspond to bona fide mRNA7 . There has been considerable debate regarding the presence of subgenomic RNAs in coronavirus particles [99 , 100] but recent analyses [101] suggest that there is a very selective incorporation of MHV gRNA into virus particles and , although immunopurified virus particles may contain detectable amounts of mRNA7 , it is minimal . The N ORF RNA observed in our study may represent a part of a defective viral genome with some structural similarity to DI-like RNAs . An alternative possibility , namely that the RNASeq density corresponding to the N ORF may arise by selective degradation of the genomic RNA , is not without precedent in other virus infections [102] . However , it seems very unlikely to occur to ~96% of the input gRNA prior to replication complex formation . Further studies are needed to determine the source of this RNA and whether or not it has any biological relevance . Our data indicate that in MHV-infected cells , in addition to the “standard” coding sequences , ribosomes access and translate a number of short ORFs . In general , translation of upstream short ORFs ( uORFs ) is thought to regulate translation of downstream protein-coding ORFs , with the peptide product of the uORF only rarely being functional in itself [82] . The AUG-initiated uORF of the gRNA has been characterised previously and may play a role in attenuation of translation of ORFs 1a and 1b , with a beneficial but non-essential role in coronavirus replication in cell culture [80 , 103] . We found that translation of this uORF occurred at a level similar to that of ORF1a , reflecting its upstream position but poorer initiation context . Interestingly , ribosomes on this uORF paused predominantly at the second codon ( proline ) , probably as a consequence of the restrained geometry of this amino acid in the decoding centre [83] . Other translated uORFs included a UUG-initiated 1-codon ORF in the leader sequence , an AUU-initiated 15-codon ORF upstream of ORF4a , and a CUG-initiated 13-codon ORF upstream of ORF5 . The function , if any , of the first two is unknown , but we speculate that the latter uORF may play a role in expression of the E protein , which is encoded downstream of ORF5 on mRNA5 . E is a small , hydrophobic viroporin that plays multiple roles during infection , including a role in virion morphogenesis [104] . As the second cistron on mRNA5 , it is not clear how the E AUG is accessed for translation initiation . Previous evidence indicates that E can be expressed via internal ribosome entry [91] , although the experiments that led to this conclusion did not test for the production of alternative transcripts that might allow E expression in the system used . We now hypothesize , however , that E could also be expressed via a form of leaky scanning , where , after translating the short uORF on mRNA5 , the small subunit of the ribosome remains associated with the mRNA , resumes scanning , and re-initiates at the AUG of the E ORF . Intervening AUGs within the 5′ 44 nt of ORF5 could be bypassed , as the scanning 40S subunit may not have had time to reacquire the relevant initiation factors [87] . We were also able to confirm expression of the previously characterized internal ( I ) ORF embedded within the N gene [93] through western blotting , while analysis of profiling data ( taking advantage of the phasing quality to gauge translation levels in different frames ) was consistent with translation of I at a level not more than 12% of N protein expression . The mechanism of I expression is uncertain , but leaky scanning of ribosomes that fail to initiate at the N AUG is a possibility and the low level of I expression is consistent with such a mechanism . Note that failure to detect I ORF initiation ( and weak detection of E ORF initiation ) may indicate a shortcoming of the ribosomal profiling technique in the detection of initiation codons accessed by non-standard mechanisms . Coronavirus −1 PRF signals have been useful models for studies of ribosomal frameshifting in vitro , both from the perspective of structure-function relationships of RNA pseudoknots , and also because they stimulate efficient frameshifting [58] . From the profiling analysis presented here , we now know that frameshifting in the context of MHV infection is also extremely efficient , with around half of the ribosomes that translate ORF1a continuing on to translate ORF1b . We find little evidence that −1 PRF is modulated by MHV infection , with similar efficiencies observed both in infected cells and in transfected cells expressing a frameshift-reporter mRNA . Intriguingly , there is no evidence that ribosomes pause upon encountering the MHV frameshift-promoting pseudoknot . Several published in vitro studies have shown that RNA pseudoknots ( and certain other RNA structures ) can pause ribosomes [57 , 59–61] and recent kinetic studies have revealed that the translocation step of protein synthesis is significantly slowed by 3′ frameshift-stimulatory RNA structures [62–64] . Whilst the in vitro systems used to study pausing and frameshifting kinetics could be inappropriate , it may be that profiling is insufficiently sensitive to register what may , in vivo , be pseudoknot-induced ribosomal pauses of short duration . Relevant to this , despite the burgeoning literature on ribosomal profiling , only relatively few studies have addressed whether RiboSeq pauses can be generally correlated with intra-mRNA structure [105–107] . Until this is better understood , the significance of these observations remains to be determined . In this study , we did identify a number of strong ribosomal pauses , however , and confirmed the occurrence of pausing within nsp3 in an in vitro translation assay . The nsp3 pausing site is located in the linker region between two modular domains of the protein , i . e . ADRP [108] and the recently identified DPUP [76] , and we hypothesize that the pause may occur after synthesis of the first domain in order to allow it to fold properly before synthesis of the second domain . Ribosomal pausing as a way to optimize protein folding has been reported increasingly in recent years [109–111] . We show that replacing four residues ( Lys , Arg , Lys , Phe ) in the nascent peptide sequence ( within 10 aa upstream of the pausing P-site ) is sufficient to largely abrogate pausing , indicating that the pause is nascent peptide mediated and depends , at least in part , on positively charged residues acting within the ribosome exit tunnel , consistent with other ribosome profiling data where positively charged residues have been linked to ribosome retardation [77] . Our analysis of MHV by ribosomal profiling is the first such investigation for an RNA virus . Together with RNASeq , the datasets provide a high-resolution examination of MHV replication and gene expression and provides a basis for the subsequent analyses of virus-host responses ( manuscript in preparation ) . We anticipate that the information will also be valuable to researchers with an interest in translation and virology , not least due to the excellent phasing in the RiboSeq datasets and the good coverage of reads on virus and cellular mRNAs . Murine 17 clone 1 ( 17Cl-1 ) [112] and BHK-21 [C-13] ( ATCC CCL-10 ) cells were maintained in Dulbecco’s modification of Eagle’s medium supplemented with 10% ( vol/vol ) fetal calf serum ( FCS ) . Recombinant MHV strain A59 ( MHV-A59 ) was derived as previously described [113] . 17Cl-1 cells ( 107 ) were plated in 10 cm dishes and , upon reaching 70–80% confluence , were infected with MHV-A59 at a multiplicity of infection ( MOI ) of 10 PFU/cell ( or 200 PFU/cell in the “High MOI” experiment ) in Hank’s balanced salt solution ( HBSS ) containing 50 μg/ml DEAE-dextran and 0 . 2% bovine serum albumin ( BSA ) . After 45 min at 37°C , the inoculum was removed and the cells were incubated in DMEM containing 10% FCS , 100 U/ml penicillin and 100 μg/ml streptomycin at 37°C until harvest . At the appropriate time point , cells were treated with CHX ( Sigma-Aldrich; to 100 μg/ml; 2 min ) , or HAR ( LKT laboratories; 2 μg/ml , 3 min ) then CHX ( to 100 μg/ml; 2 min ) . Cells were rinsed with 5 ml of ice-cold PBS , the dishes were submerged in a reservoir of liquid nitrogen for 10 s and then transferred to dry ice and 400 μl of lysis buffer [20 mM Tris-HCl pH 7 . 5 , 150 mM NaCl , 5 mM MgCl2 , 1 mM DTT , 1% Triton X-100 , 100 μg/ml cycloheximide and 25 U/ml TURBO DNase ( Life Technologies ) ] dripped onto the cells . The cells were scraped extensively to ensure lysis , collected and triturated with a 26-G needle ten times . Lysates were clarified by centrifugation for 20 min at 13 , 000 g at 4°C , the supernatants recovered and stored in liquid nitrogen . Cell lysates were subjected to RiboSeq and RNASeq . The methodologies employed were based on the original protocols of Ingolia and colleagues [7 , 114] , except ribosomal RNA contamination was removed by treatment with duplex-specific nuclease ( DSN ) and library amplicons were constructed using a small RNA cloning strategy [115] adapted to Illumina smallRNA v2 to allow multiplexing . The methods used were as described [16] , with minor modifications for the analysis of ribosomal pausing at the MHV −1 PRF signal , namely a broader range of RPFs , migrating between 28 and 80 nt , were harvested prior to amplicon construction , and longer PCR amplicons of ~150–206 bp were gel purified . Amplicon libraries were deep sequenced using an Illumina HiSeq 2000 platform ( repeat 1 samples at the Wellcome Trust Centre for Human Genetics—Oxford Genomics Centre; repeat 2 , MOI 200 , and long read samples at the Beijing Genomics Institute ) . Adaptor sequences were trimmed using the FASTX-Toolkit and reads shorter than 25 nt were discarded . Trimmed reads were mapped first to Mus musculus rRNA ( GenBank accession numbers NR_003278 , NR_003279 , NR_003280 , NR_030686 , NR_046233 and GU372691 ) , followed by the MHV genome ( GenBank accession number AY700211 . 1 ) and subsequently Mus musculus mRNA , ncRNA and genomic DNA databases . In order to select good-quality samples of host mRNA-derived RPFs for analyzing RPF length , framing , and position-on-transcript distributions , the mouse mRNA database comprised NCBI RefSeq mRNAs . The non-coding RNA and genomic DNA databases comprised the Ensembl Mus_musculus . NCBIM37 . 64 . ncrna . fa and release-64 DNA chromosome files , respectively . Reads that map to the gDNA , but none of the RNA databases , are expected to derive from unannotated transcripts as the sequencing protocol is RNA-specific . Reads were mapped using bowtie version 1 [116] with parameters -v 2 —best ( i . e . maximum 2 mismatches , report best match ) . The order of mapping was tested to check that virus-derived reads were not lost accidentally due to mis-mapping to host RNA , or vice versa; a slight reduction ( ~0 . 05% ) in virus-derived reads was observed only on mapping to the entire host genome ( gDNA ) and thus mapping to virus RNA and host mRNA was considered to be specific . For host mRNA mapping , no specific consideration was given to the presence of multiple isoforms within the RefSeq database; reads that could be mapped to multiple transcripts were assigned at random to one transcript . Except where specifically stated , virus reads that mapped discontinuously to the MHV genome ( due to transcriptional discontinuities at TRS sites ) were excluded from the analyses . Host mRNA RiboSeq and RNASeq phasing distributions ( S2 Fig and S3 Fig ) were derived from reads mapping to the “interior” regions of annotated coding ORFs; specifically , the 5′ end of the read had to map between the first nucleotide of the initiation codon and 30 nt 5′ of the last nucleotide of the termination codon , thus , in general , excluding RPFs of initiating or terminating ribosomes . Histograms of 5′ end positions of host mRNA reads relative to initiation and termination codons ( S4 Fig , S5 Fig , S6 Fig ) were derived from reads mapping to RefSeq mRNAs with annotated CDSs ≥450 nt in length and annotated 5′ and 3′ UTRs ≥60 nt in length . All figures are based on total numbers of mapped reads , rather than weighted sums for highly expressed mRNAs [7] , because virus-induced shut-off of host cell translation at late time points reduces the efficacy of the latter approach for our data . Read length distributions ( S7 Fig and S8 Fig ) are based on total mapped reads ( to positive-sense host mRNA , or to positive or negative-sense MHV genome , as indicated ) without restriction to annotated coding regions . To compare read densities between CDSs and 3′ UTRs ( S9 Fig and S10 Fig ) , we used reads whose 5′ end offset by +12 nt ( i . e . estimated P-site positions for RPFs ) mapped within the regions from 30 nt to 300 nt upstream of stop codons ( CDSs ) , or from 30 nt to 300 nt downstream of stop codons ( 3′ UTRs ) . This analysis was restricted to mRNAs with annotated coding ORFs ≥450 nt in length and annotated 3′ UTRs ≥300 nt in length . The presence of transcript isoforms with 3′ UTRs shorter than the annotated ( ≥300 nt ) 3′ UTRs leads to a modest underestimation of the actual 3′ UTR density . For Fig 12B , RefSeq mRNAs with annotated CDSs ≥300 nucleotides in length and annotated 5′ UTRs ≥30 nt in length ( with no restriction on annotated 3′ UTR length ) were used , as only the 5′ end of CDSs was analysed , and the more relaxed thresholds increased the sample size [important for the more restricted set of CDSs beginning with AUG-CCN ( Met-Pro ) ; of 29600 NCBI RefSeq mRNA accessions , 1558 have CDSs beginning with AUG-CCN] . Transcripts with ≥20 RPFs with 5′ ends mapping between −30 and +15 relative to the annotated initiation codon were used , and histograms of 5′ end positions for individual transcripts were down-weighted by the number of RPFs mapping to this region before summing over the different transcripts ( i . e . a weighted sum of “highly expressed” mRNAs , [7] ) . Fig 12B is based on sums over 3620 and 203 transcripts for generic CDSs and CDSs beginning with AUG-CCN , respectively . Plots showing reads mapped to the MHV genome ( Figs 1 , 2 , 7 , 9 , 11 , 12A , 13 and 14 and S11 Fig ) show histograms of the positions to which the 5′ ends of reads map , with a +12 nt offset to indicate ( for RPFs ) the approximate P-site . ( More precisely , the +12 nt offset means that RPFs whose 5′ end aligns to the first position of a codon are mapped to the first nucleotide of the P-site codon , and RPFs whose 5′ end aligns to the third position of a codon are mapped to the last nucleotide of the codon preceding the P-site codon . ) In contrast , plots showing reads summed over large numbers of host mRNAs ( Fig 12B and S4 Fig , S5 Fig , S6 Fig ) show histograms of the positions to which the 5′ ends of reads map , without the +12 nt offset . This is because the host mRNA plots are used for calibration whereas the virus plots are used to illustrate specific features of virus gene expression . To normalize for different library sizes , while taking into account global shut-off of host gene expression in response to virus infection , counts expressed as reads per million mapped reads ( RPM ) or reads per kb per million mapped reads ( RPKM ) use the sum of total virus RNA ( positive and negative-sense ) plus total host mRNA ( reads that map to NCBI mRNA RefSeqs ) as the denominator . The same library normalization factors were also used for Fig 3 , S4 Fig and S6 Fig . To calculate the expression of individual virus ORFs ( Fig 4 , RiboSeq ) , we counted RPFs whose 5′ end mapped between the first nucleotide of the initiation codon and 30 nt 5′ of the termination codon , thus excluding RPFs of ribosomes paused during initiation or termination ( or nearby ) . The corresponding sequence length was used to calculate counts per kb . We used a similar procedure to calculate RNASeq densities for each inter-TRS region ( Fig 4 , RNASeq ) , with the inter-TRS regions ( prior to the 30 nt 3′ buffer ) being 72 to 21748 ( mRNA1 ) , 21754 to 23923 ( mRNA2 ) , 23929 to 27936 ( mRNA3 ) , 27942 to 28319 ( mRNA4 ) , 28325 to 28959 ( mRNA5 ) , 28965 to 29656 ( mRNA6 ) , and 29662 to 31335 ( mRNA7 ) . Frameshifting efficiencies ( Fig 8 ) were calculated using reads whose 5′ end offset by +12 nt ( i . e . estimated P-site positions for RPFs ) mapped within the regions 361 to 13452 ( for ORF1a ) and 13774 to 21596 ( for ORF1b ) . These coordinates leave a 150 nt buffer after the ORF1a initiation codon ( nt 211 ) , before the frameshift site ( nt 13602 ) , after the ORF1a termination codon ( nt 13623 ) and before the ORF1b termination codon ( nt 21746 ) , respectively . Read counts were divided by region lengths to obtain read densities . Phasing distributions in the N and I ORFs ( Fig 14E ) were calculated with respect to the N reading frame , using reads whose 5′ end offset by +12 nt ( i . e . estimated P-site positions ) mapped within the regions 29736 to 30353 ( for the I/N overlap ) and 30360 to 31031 ( for N downstream of I ) . For comparison , the coordinates of the N and I ORFs are 29670 to 31034 ( N ) and 29734 to 30357 ( I ) . For the analysis of RiboSeq and RNASeq count variability within ORF1a ( Fig 9A ) , counts were first smoothed with a 3-nt running mean filter and then the fold-change relative to mean was calculated using reads whose 5′ end mapped between nt 211 ( the start of ORF1a ) and nt 13572 ( 30 nt 5′ of the frameshift site ) . For the above analyses , virus reads with discontiguous mappings to the MHV genome ( i . e . reads spanning sites of discontinuous transcription—generally at the TRS sites ) were excluded . To identify such reads we re-mapped raw trimmed reads to host rRNA , virus genome , host mRNA , ncRNA and gDNA databases , this time permitting zero mismatches . We then pooled the remaining unmapped reads with the reads that mapped to the virus genome and , for each library , searched this set of reads for the query sequence UUUAAAUCUAA ( AY700211 . 1 nt 55 to 65; 5′-adjacent to the leader TRS ) . Reads were selected that had at least 17 nt 3′ of the query sequence and classified according to whether nucleotides +3 to +17 after the query sequence were compatible with mRNA1 , 2 , 3 , 4 , 5 , 6 or 7 , or were derived from non-canonical chimeric sequences . These criteria were motivated by previous data indicating that , in leader/body chimeras , nucleotides up to and including UUUAAAUCUAA are templated by the leader , nucleotides at +1 and +2 may be templated by leader or genome , and nucleotides at +3 and above are templated by the genome sequence [41] . Counts were normalized to reads per million mapped reads as described above . A possible source of error here is that different libraries have different RNASeq read length distributions ( due to variation in the gel-slice boundaries ) ; libraries with longer reads will have proportionally more reads found to span leader/body discontinuities due to the requirement of at least 17 nt 3′ of the 11-nt query sequence for selection . For this reason , inter-library comparisons are avoided . To calculate host translational efficiencies , after removing reads mapping to rRNA with bowtie1 as above , remaining reads were mapped to the mouse genome ( UCSC , assembly mm10 ) using TopHat ( parameters: —no-novel-juncs —bowtie1 —prefilter-multihits —max-multihits 500 , with —transcriptome-index defined using the genes . gtf file from the UCSC mm10 annotation available from the tophat website ) [117] . Reads entirely contained within annotated CDSs were enumerated with htseq-count ( parameters: -t CDS -m intersection-strict -i gene_id -s yes ) [118] , reporting read counts per gene rather than per transcript . Read counts were normalized for library size as above , and for CDS length according to the sum of all coding exon fragment lengths for a given gene ID in the genes . gtf file . This will tend to result in an overestimate of CDS lengths since many transcripts ( alternative splice forms and/or alternative transcription initiation sites ) will lack some coding exons . While this is likely to have only a modest effect on RiboSeq/RNASeq translation efficiencies ( Fig 6 , y-axis ) it will tend to result in underestimates for RNASeq RPKM values ( Fig 6 , x-axis ) . The sequencing data have been deposited in the ArrayExpress database ( http://www . ebi . ac . uk/arrayexpress ) under the accession number E-MTAB-4111 . The MHV frameshift signal , and the N , nsp3 and nsp6 protein coding sequences were amplified using specific oligonucleotides ( S4 Table ) and cDNA derived from 17 Cl-1 cells infected with MHV-A59 at an MOI 10 and harvested at 8 h p . i . For assessing frameshifting efficiencies in transfected tissue culture cells , the dual-luciferase reporter vector pDluc was employed ( kind gift from Dr M . Howard , University of Utah; [53] ) . DNA fragments of 100 bp spanning the MHV frameshift signal and flanked by XhoI and BglII restriction sites were derived by PCR amplification and ligated into appropriately cleaved pDluc vector . An in-frame control ( mimicking 100% frameshifting efficiency ) was also constructed . pDluc-IBV and pDluc-HXB2 have been described elsewhere [54 , 55] . BamH1-XhoI-digested PCR fragments were cloned into pcDNA 3 . 1 ( + ) ( Life Technologies ) previously digested with BamH1-XhoI . In pPS0 plasmids , PCR reactions were carried out using the pcDNA . 3 nsp3 plasmid as a template and cloned into a digested XhoI/PvuII-pPS0 plasmid . pPS0-nsp3 mutants were subjected to site-directed mutagenesis . For all pcDNA . 3 and pPS0 constructs , a “pause control” was also generated in which a UAA stop codon was introduced to generate a protein whose size corresponded to that produced by the predicted ribosomal pause . All sequences were confirmed by dideoxy sequencing . 17 Cl-1 and BHK-21 cells were seeded in dishes of a 24-well plate and grown for 16 h until 80% confluence was reached . Plasmids were transfected using a commercial liposome method ( TransIT-LT1 , Mirus ) . Transfection mixtures [containing plasmid DNA , serum-free medium ( Opti-MEM; Gibco-BRL ) and liposomes] were set up as recommended by the manufacturer and added dropwise to the tissue culture cell growth medium . Cells were harvested 24 h post transfection ( h . p . t . ) and reporter gene expression was determined using a dual-luciferase assay system kit ( Promega ) . Frameshifting efficiencies were calculated by dividing the Fluc/Rluc ratios of the test samples by the Fluc/Rluc ratio of the in-frame controls . Proteins were separated by 10% , 12% or 15% SDS-PAGE depending on the molecular weight of the protein of interest and transferred to nitrocellulose membranes . These were blocked for 30–60 min with 5% powdered milk ( Marvel ) in PBST [137 mM NaCl , 2 . 7 mM KCl , 10 mM Na2HPO4 , 1 . 5 mM KH2PO4 ( pH 6 . 7 ) , and 0 . 1% Tween 20] and probed with mouse monoclonal antibodies raised against nsp9 ( AM08450PU-N , Acris Antibodies , Inc , 1:500 in Marvel-PBST ) , N ( 1:1 , 000 ) , S ( 1:500 ) , GAPDH ( G8795 , Sigma-Aldrich , 1:20 , 000 ) or a polyclonal rabbit anti-I ( 1:1 , 000 , a kind gift of Prof . P . S . Masters , Wadsworth Center , New York State Department of Health ) . Membranes were incubated in the dark with an IRDye-conjugated secondary antibody in PBST [IRDye 800CW Donkey Anti-Mouse IgG ( H+L ) , IRDye 800CW Donkey Anti-Rabbit IgG ( H+L ) and IRDye 680RD Goat Anti-Mouse IgM ( μ chain specific ) ] . Blots were scanned and bands quantified using an Odyssey Infrared Imaging System ( Licor ) . pcDNA . 3 and pPS0 plasmids were linearized with XhoI and AvaII respectively and capped run-off transcripts generated using T7 RNA polymerase and SP6 RNA polymerase respectively as described previously [119] . RNAs were recovered by a single extraction with phenol-chloroform ( 1:1 vol/vol ) followed by ethanol precipitation . Remaining unincorporated nucleotides were removed by gel filtration through a NucAway spin column ( Ambion ) . The eluate was concentrated by ethanol precipitation , the mRNA resuspended in water , checked for integrity by agarose gel electrophoresis and quantified by spectrophotometry . RNAs were translated in nuclease-treated rabbit reticulocyte lysate ( RRL ) ( Promega ) programmed with ~50 μg/ml template mRNA . A typical reaction mixture had a volume of 10 μl and was composed of 90% ( vol/vol ) RRL , 20 μM amino acids ( lacking methionine ) , and 0 . 2 MBq [35S]-methionine . Reaction mixtures were incubated for 30 min at 26°C and stopped by the addition of an equal volume of 10 mM EDTA , 100 μg/ml RNase A followed by incubation at room temperature for 15 min . In ribosomal pausing assays , conditions were the same except that the reaction mixture had a volume of 40 μl and the translational inhibitor edeine was added 3 min after the start of the reaction in order to obtain synchronous initiation ( final concentration , 5 μM ) . Aliquots of 1 . 5 μl were withdrawn from the translation reaction mixture at specified intervals and mixed with an equal volume of EDTA/RNase A mixture , as above . In immunoprecipitations , 10 μl of RRL was mixed with either mouse anti-N , anti-S or rabbit anti-I for 30 min at 4°C prior to binding to protein A-Sepharose CL-4B ( Pharmacia Biotech AB , Uppsala , Sweden ) and subsequent washing . Samples were prepared for SDS-PAGE by addition of 10 volumes of 2X Laemmli’s sample buffer and boiling for 4 min . Proteins were resolved on 10% or 15% SDS-PAGE gels . 14C-labelled molecular weight standards ( MW ) were from Amersham International ( United Kingdom ) . Dried gels were exposed to a Carestream Kodak Biomax MR film ( Sigma-Aldrich ) and scanned .	Ribosome profiling is emerging as a powerful technique to monitor translation in living cells at sub-codon resolution . It has particular applicability to virology , with the capacity to identify viral mRNAs that are being translated during infection and to provide new insights into virus gene expression , regulation and host-virus interactions . In this work , we carried out the first ribosome profiling analysis of an RNA virus , using as a model system the murine coronavirus strain MHV-A59 , a betacoronavirus in the same genus as the medically important SARS-CoV and MERS-CoV . Parallel ribosome profiling and RNA sequencing of infected-cell time points was performed during the course of MHV replication in mouse tissue culture cells and used to determine virus gene expression kinetics and the relative translational efficiencies of virus and host mRNAs . The sensitivity and precision of the approach permitted us to uncover several unanticipated features of coronavirus translation , giving insights into ribosomal frameshifting , ribosome pausing , and the utilisation of short , potentially regulatory , upstream open reading frames . We also identified some challenges associated with the technique that are of general relevance to the ribosome profiling technique and developed bioinformatic strategies to address these .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "nucleic", "acid", "synthesis", "messenger", "rna", "negative-sense", "rna", "viruses", "viruses", "rna", "viruses", "genome", "analysis", "rna", "synthesis", "cellular", "structures", "and", "organelles", "chemical", "synthesis", "research", "and", "analysis", "metho...	2016	High-Resolution Analysis of Coronavirus Gene Expression by RNA Sequencing and Ribosome Profiling
Conservation outcomes are principally achieved through the protection of intact habitat or the restoration of degraded habitat . Restoration is generally considered a lower priority action than protection because protection is thought to provide superior outcomes , at lower costs , without the time delay required for restoration . Yet while it is broadly accepted that protected intact habitat safeguards more biodiversity and generates greater ecosystem services per unit area than restored habitat , conservation lacks a theory that can coherently compare the relative outcomes of the two actions . We use a dynamic landscape model to integrate these two actions into a unified conservation theory of protection and restoration . Using nonlinear benefit functions , we show that both actions are crucial components of a conservation strategy that seeks to optimise either biodiversity conservation or ecosystem services provision . In contrast to conservation orthodoxy , in some circumstances , restoration should be strongly preferred to protection . The relative priority of protection and restoration depends on their costs and also on the different time lags that are inherent to both protection and restoration . We derive a simple and easy-to-interpret heuristic that integrates these factors into a single equation that applies equally to biodiversity conservation and ecosystem service objectives . We use two examples to illustrate the theory: bird conservation in tropical rainforests and coastal defence provided by mangrove forests . Habitat conservation is central to biodiversity conservation . Habitat can be conserved by either protecting it if it remains intact or by restoring it once it has been degraded . Conservation organisations often pursue both restoration and protection simultaneously , and management guidelines advocate the use of both actions [1 , 2] . However , the orthodox position is that managers should “protect first , restore second” where possible , and the priority of protection has been argued in the scientific literature [3–7] and management guidelines internationally [8–10] , for both biodiversity conservation and ecosystem service provision . This prioritisation of investment in protection over restoration is justified with reference to the relative costs , expected benefits , and timescales of the two actions [3–12] . While restoration can improve a site’s ecological condition , restored habitat will often take decades to regain the majority of its biodiversity and ecosystem attributes [3 , 13] . Despite this prevailing wisdom , recent experiments have revealed that the disparity between restoration and protection is smaller than expected , and strongly context dependent . The majority of many ecosystem features , particularly certain ecosystem services , can be provided by restored habitat [3] , sometimes within a surprisingly short timeframe [14 , 15] . Moreover , ongoing technological advances [16] mean that the cost of this restoration can be low enough to generate a net social benefit [11 , 17] . At the same time , conservation theory has highlighted disadvantages to habitat protection that parallel problems identified for restoration . Protected areas also suffer from poor implementation and management and do not guarantee the conservation of intact habitat , species assemblages , or ecosystem services [18 , 19] . Dynamic landscape models further illustrate how the benefits of protection are also subject to time delays , since protection does not create new habitat , but only reduces the likelihood of future habitat loss [20] . Moreover , in highly degraded landscapes that leave many species with non-viable populations ( i . e . , extinction debts [21] ) , habitat protection will only have a secondary effect on biodiversity loss rates . Restoration is the only in situ conservation intervention that can actively reduce an extinction debt [22 , 23] . The relative priority of protection and restoration can only be coherently assessed by a conservation resource allocation theory that incorporates both actions and that quantifies both their costs and benefits in a comparable manner . In particular , this theory must be temporally explicit , since both the benefits of the two actions accrue at different rates . In this paper , we incorporate restoration and protection into a shared dynamic landscape model [20] that explicitly includes restoration rates , dynamic species loss ( i . e . , extinction debts ) , and ecosystem service provision ( see Materials and Methods ) . To contrast the performance of habitat protection and restoration , we apply this unified theory to two divergent examples: ecosystem service provision in the Coral Triangle and biodiversity conservation in the Atlantic forests . Intact mangrove ecosystems help defend coastal communities from storm surges and floods [24 , 25] , events that are increasing in frequency and intensity as the climate changes [26] . In Southeast Asia , the impact of frequent extreme weather events on coastal communities is pronounced due to vulnerable infrastructure and high rates of mangrove deforestation [27] . We focused on the northern tip of Borneo ( Sabah , Malaysia ) , where over 38 , 000 people live on the coast . Approximately 78% of mangrove forests remain from an original 535 km2 , after more than 30 years of losing 0 . 5%–0 . 8% annually to other uses ( e . g . , urban development and aquaculture ) [27] . The Sabah Forestry Department is charged with both the protection and restoration of mangrove forests , two actions that gained popularity across Southeast Asia after the 2004 Indian Ocean tsunami , but they lack a decision framework for deciding which action is best to implement [28] . Barbier et al . [25] calculated that coastal defence benefits accrue as a nonlinear function of the area of surrounding intact mangrove habitat: E ( t ) = 1 − exp ( − k ( P ( t ) + F ( t ) ) ) Eq ( 1 ) where P ( t ) is the area of protected mangrove forest , F ( t ) is the area of unprotected but intact mangrove forest , and k = 2 . 1x10–3 ( S1 Text ) . Eq ( 1 ) measures ecosystem service provision by the proportional reduction in damage caused by a lower wave height in areas sheltered by mangrove forests . In this example , over the course of a 30-year project starting in 2006 , managers must choose to share an annual conservation budget equivalent to US$15 million between the protection of intact mangrove forests , the restoration ( and subsequent protection ) of degraded mangrove habitat , or a combination of both actions . Their aim is to maximize the total coastal defence provided to communities by mangrove forests over the project lifetime: maxu ( t ) ∫ t=0 T e −rt [ 1 − e − k ( P ( t ) + F ( t ) ) ] dt Eq ( 2 ) where r is the economic discount rate . We assume that restored mangrove forests do not contribute to coastal defence until restoration is complete ( see S1 Text for parameterisation details ) . To maximise coastal defence , the model predicts that northern Sabah managers should prioritise the restoration of cleared or degraded mangrove habitat over the protection of intact forests ( Fig . 1A and Fig . 2A ) . The optimal allocation schedule suggests that restoration should be an absolute priority ( i . e . , all available funds should be directed towards restoration ) , even though the optimisation method allows managers to allocate part of their funds to both actions ( e . g . , 95% to restoration and 5% to protection ) . Restoration results in a smaller amount of protected forest than protection ( Fig . 1B ) , both because restoration is significantly more expensive , and because restored habitat only becomes intact and protected after a substantial time lag . Nevertheless , restoration provides the local community with more coastal defence because it results in less degraded land and more intact forest ( protected and unprotected; Fig . 2A ) . The optimal allocation therefore assigns a higher priority to restoration . Paraguay contains some of the last remnants of South America’s high-latitude tropical rainforests , which remained predominantly intact into the 1970s [29] . Given the extraordinary endemic bird species richness ( 148 species ) and high levels of habitat loss ( >90% ) in this ecoregion , the small number of extinctions to date indicates the presence of a substantial extinction debt [30] . Managers want to minimise the number of bird extinctions by either restoring and then protecting cleared or degraded habitat , by protecting the last remaining stands of intact rainforest , or by a combination of the two [31] . The species–area relationship relates the equilibrium species richness to the area of intact habitat , regardless of its protection status: S * = α ( P + F ) z Eq ( 3 ) where α represents regional species richness and z is a constant . We make the conservative assumption that restored habitat does not mitigate the extinction debt until restoration is complete , and we estimate the recovery rate using long-term surveys of species richness following forest clearance ( although the complete recovery of population abundances will take longer than the return of species; S1 Text ) . Species extinctions often lag significantly behind habitat loss , but debt “relaxation” can be averted if degraded habitat is quickly restored . The rate of species extinction is modelled as proportional to the size of the species debt , calculated as the difference between extant species richness and the number of species that would be supported by the current habitat distribution at equilibrium [32]: dS ( t ) dt = θ [ S ( t ) − S * ] = θ [ S ( t ) − α ( P ( t ) + F ( t ) ) z ] , Eq ( 4 ) where θ is the rate of extinction debt relaxation . The managers’ objective is to minimize the total number of extinctions during a T year conservation project: minu ( t ) ∫ t=0 T [ S ( t ) − α ( P ( t ) + F ( t ) ) z ] dt Eq ( 5 ) We calculated a retrospective optimal management strategy for protection and restoration between 1970–2013 ( Fig . 3; see S1 Text for parameterisation details ) , assuming an ongoing annual budget equivalent to US$100 million ( 2014 ) . The results show that , in the Atlantic Forests , restoration and protection would have achieved broadly comparable outcomes ( Fig . 2B ) . However , to optimally reduce species extinctions over the time period , managers should have pursued habitat protection for the first 20 years and then switched their efforts entirely towards restoration ( Fig . 3A ) . It is not optimal to fund both actions simultaneously , at any point in time . Rates of habitat degradation in Paraguay’s Atlantic Forests were so rapid that neither action would have had a large impact on the size of the extinction debt . Nevertheless , an initial focus on protection would have quickly reduced the amount of habitat that was unprotected , and that therefore could be degraded . Once this was achieved , a switch to restoration would have allowed managers to address the extinction debt directly , by converting degraded land back into intact habitat . We identified the best allocation schedules using optimal control theory , but the relative benefits of protection and reservation can often be better understood and implemented using myopic heuristics [20] . Both biodiversity and ecosystem service objectives are advanced by an increase in the total amount of intact habitat ( protected and unprotected ) . The act of protection increases intact habitat indirectly by reducing the amount of unprotected habitat available for degradation , while restoration converts degraded land to protected intact habitat , after a time delay . A linearisation of the state equations ( Eq 7; Materials and Methods ) indicates that protection should be a higher conservation priority if half the ratio of the restoration and habitat loss rates is less than the ratio of their costs ( see S1 Text for derivation ) : g 2δ < c R c P Eq ( 6 ) Roughly speaking , restoration is favoured if its relatively higher costs ( cR/cP ) are outweighed by its more rapid reduction of degraded land relative to protection ( g/2δ ) . In agreement with our optimal control solutions ( Fig . 1 and Fig . 3 ) , Eq ( 6 ) predicts that restoration should be strongly favoured over protection in mangrove forests ( g/2δ = 13 . 1; cR/cP = 3 . 0 ) . In contrast , protection should be considered a slightly higher priority than restoration in the Atlantic Forests ( g/2δ = 1 . 1; cR/cP = 1 . 6; see S1 Text for parameter values ) . This simple rule of thumb holds true for both ecosystem service provision and biodiversity conservation objectives . The right-hand side of this condition is familiar—relative costs of action are consistently identified as important elements of prioritisation [33–35] . However , the left-hand side highlights another important factor in conservation planning: the timescales over which interventions yield benefits . An interpretation of this temporal ratio is that restoration will become a higher priority when habitat quickly regains its pristine qualities; protection should be preferred when the restoration rate is slow or when the rate of land degradation is rapid . This rule can be readily modified to include the probability that restoration and protection will be unsuccessful; the results of this modification indicate that including failure rates are equivalent to increases in the costs of the respective action ( S1 Text ) . For example , if the probability of restoration being successful was only 50% , then that is equivalent to a doubling of the cost of restoration . Constrained by limited budgets , conservation organisations and governments must always choose between restoration and protection . Our framework provides a coherent framework that can help resolve longstanding uncertainties about the relative priority of these two fundamental actions . The results and implications are much more complex than simply “protect first , restore second . ” In both of our examples , the long-term objectives will be best achieved if priority is given to restoration at some point in the project , despite its higher cost , and despite a substantial delay before restored habitat can contribute to project objectives . As well as demonstrating the benefits of restoration , our results indicate that the optimal solution is always to spend all available resources on either restoration or protection , never both . While the option of splitting the budget between the two actions was available to our optimisation method , it was never optimal to fund both actions simultaneously . This type of either-or solution—known as bang-bang control—is often sensitive to model assumptions about homogeneity , uncertainty , and linearity . Our model makes all of these assumptions . The model in Eq ( 7 ) considers only the overall landscape scale and contains no fine-resolution ecological or economic variation , with all land incurring the same restoration or protection costs . In reality , conservation costs ( and benefits ) vary dramatically between locations and projects [36] . We assume that the rates of land loss and restoration are known and deterministic , but land degradation is a stochastic process , and depends on highly uncertain factors , as do both restoration success and protected area performance . Finally , the management control terms in Eq ( 7 ) are linear in the control variable u ( t ) ( e . g . , uB/cP ) , while in reality , expenditure generally achieves diminishing marginal returns [37] . Relaxing any of these assumptions will tend to smooth the abrupt switch between restoration and protection—that is , will create a more continuous transition period during which both actions are funded simultaneously . Our objective functions ( Eqs 2 and 5 ) aim to maximise unidimensional descriptions of biodiversity conservation ( the number of bird species ) and ecosystem service provision ( the amount of coastal protection ) . For individual ecosystem services like coastal protection or carbon sequestration , this is a reasonable formulation of conservation objectives . However , many conservation agencies carry out projects that aim to deliver multiple benefits simultaneously , and almost all conservation actions will inevitably provide benefits to more than one ecosystem service or measure of biodiversity . It is possible , for example , to imagine a conservation project that pursues both our stated objectives—coastal protection and species conservation—simultaneously . While it is possible to aggregate the provision of multiple benefits into unidimensional quantities by calculating the monetary value of different ecosystem services [25] or by assigning relative weightings to different species [38] , this is not always appropriate or desirable . The process of aggregation will also be further complicated by the divergent values of multiple stakeholders [39] . In such situations , optimisation may not be as useful as an exploration of how restoration and protection affect trade-offs between objectives and conflict between stakeholders . Our broad , qualitative conclusion—that protection should not always be prioritised over restoration—is robust to the parameterisation of our two examples and to the structure of the restoration process ( S1 Text ) . However , both restoration and protection are more complicated and nuanced than any of the abstracted models we apply here . Recent meta-analyses of terrestrial and aquatic restoration projects show that even successful restoration projects are unable to recover reference-level biodiversity and ecosystem services ( they achieved an average of 80%–86% of reference sites , although technological improvements continue to improve these outcomes [3 , 16] ) . Moreover , these benefits accrue to different ecosystem features at varying timescales , and may take decades to be realised , particularly for the restoration of biodiversity [40 , 41] . Alternate versions of the landscape model can incorporate incomplete restoration ( one alternative formulation is given in the S1 Text ) , but at the cost of increased complexity and greater information requirements . Protection is also rarely perfectly effective , both because effectively managed protected areas cannot halt all degrading activities [18] and because many protected areas are poorly managed [42] . Furthermore , decisions to restore or protect are influenced by a variety of important factors not considered here , including the feasibility of actions in a given place , which is influenced by operational ( e . g . , technical success ) , legal ( e . g . , land tenure ) , political ( e . g . , political will ) , and social constraints ( e . g . , the willingness of landowners ) . In particular , our nonspatial model omits the constraints placed on managers by spatial conservation objectives . For example , if managers want to connect particular areas or reduce fragmentation in a landscape , restoration will be the only suitable action . Despite these omissions , we believe that simple , general theory can still provide useful insights into a problem . By providing a unified , dynamic framework within which to compare their long-term outcomes , our theory provides evidence and rationale for pursuing restoration alongside protection—even in preference to protection—under the right circumstances . An explicit theoretical framework also helps to highlight relationships that determine the priority of the two actions: the relative costs of restoration and protection , and the rate at which restored habitat approaches the benefits of intact habitat relative to the habitat conversion rate . Our two examples illustrate this general theory and demonstrate the unexpected potential for restoration to parallel or precede protection . However , the approach and simple rule of thumb should be seen as informative , not prescriptive . Our framework is not intended to inform precisely where protection and restoration should occur at a fine scale within a landscape , but to offer insights into the allocation of a limited conservation budget at a meso-scale ( e . g . , bioregional or catchment conservation initiatives ) . The model is nonspatial , assuming that the conservation state of the landscape can be described by the aggregate proportions of the land in different states ( e . g . , the proportion protected , degraded , etc . ) . Actual decisions about precisely where to act are inherently spatial and should be informed using spatial zoning tools that optimize for multiple conservation actions [43] . The decision about whether to restore or protect is underpinned by a basic dynamic landscape model . Each small area of land at time t is classified as being in one of four states: intact and unprotected , F ( t ) ; intact and protected , P ( t ) ; degraded or cleared , C ( t ) ; or undergoing restoration , R ( t ) . We describe the total amount of land in each state as a proportion of the landscape , and the model ensures that F ( t ) + P ( t ) + C ( t ) + R ( t ) = 1 at all times . In our model , transitions of land between the four states are driven by two actions and two processes . The two actions , protection and restoration , are entirely determined by the managers . At each point in time across a T-year project , managers can allocate a varying proportion ( 0≤u ( t ) ≤1 ) of a fixed annual budget ( B ) to protection , and the remainder 1—u ( t ) to restoration . We note that this choice of u ( t ) allows managers to simultaneously fund both actions—that is , to allocate a proportion of their resources to protection , and the remainder to restoration ( e . g . , if u ( t ) = 0 . 1 , managers spend 10% of their budget on protection and 90% on restoration ) . Managers cannot directly alter the two processes . The first process is land degradation , in which we assume that unprotected land is being degraded at proportional rate δ [20 , 44 , 45] . That is , landscapes with large amounts of intact , unprotected habitat will experience large absolute rates of habitat loss . Managers can therefore reduce habitat loss by decreasing the amount of unprotected habitat through protection , but they cannot directly affect the loss rate δ . We note that alternative models of land degradation ( e . g . , constant rates ) could also be used . The second process is restoration . We assume that restoration actions , once undertaken , do not create intact habitat immediately . Once managers spend resources by purchasing degraded land and undertaking restoration actions , the land undergoing restoration , R ( t ) , only regains its intact qualities at a proportional rate g . We note that this is a continuous model of restoration , rather than a time-lag model , but show in the S1 Text that this simplification does not qualitatively alter our conclusions . By combining these actions and processes , the rate of change of each land state becomes: dF ( t ) dt =−δF− u ( t ) B c P dP ( t ) dt = u ( t ) B c P +gR dC ( t ) dt =δF− ( 1−u ( t ) ) B c R dR ( t ) dt = ( 1−u ( t ) ) B c R −gR Eq ( 7 ) The variables cP and cR denote the costs of protection and restoration respectively . Because habitat must be purchased before it is restored , cR is generally larger than cP . However , if land has been abandoned , or if the primary value of land comes from the intact habitat itself ( e . g . , timber ) , then purchasing degraded land for restoration may be cheaper than purchasing intact habitat for protection . The protection or restoration of a land parcel both require ongoing management , and we therefore assume that these two cost parameters measure the endowed cost of undertaking both actions until the end of the project . That is , cP is equal to the initial purchase price of the land , plus the time discounted cost of all future actions required to effectively manage the protected area . Likewise , cR is the sum of the land purchase price , the time discounted cost of the initial intensive restoration actions , and then the time discounted cost of ensuring that the restoration and subsequent protection are effective . The entirety of these endowed costs must be paid in the year that a given parcel of land is restored or protected . The dynamics described by Eq ( 7 ) are illustrated in Fig . 4 . Identifying the optimal management schedule is equivalent to determining the specific control function u ( t ) that maximises the objective function . There are clearly a very large number of candidate control functions , but fortunately the optimal function can be identified by applying Pontryagin’s maximum principle [46 , 47] to the system dynamics in Eq ( 3 ) , and each problem’s objective function ( Eq 1–2 ) . The details of this analysis are shown in the S1 Text . We note that it is not essential to understand the specifics of this optimisation method since alternative methods could equally be used to solve for the optimal solution , notably stochastic dynamic programming [20] .	Most species go extinct because humans have cleared their habitat . Habitat loss can also cause people to lose some of the services provided by ecosystems , such as the removal of carbon dioxide from the atmosphere or the protection of coastal communities from storm damage . There are two broad strategies for stopping and reversing habitat loss: we can either protect habitat that is currently intact , or we can restore habitat that has already been cleared . Superficially , we might imagine that , as with human health , “prevention is better than cure , ” and that therefore habitat protection should be given priority over habitat restoration . However , there is currently no scientific theory to justify this belief . Here , we used an ecosystem model and dynamic optimization tools from mathematics to show that habitat restoration ( such as tree planting ) can , surprisingly , be more cost-effective than habitat protection ( such as designating a national park ) for two case studies . We discovered that the best decision depends on the relative costs of the two actions , the rate at which habitat is being lost , and the time lag between restored habitat being as useful as intact habitat for securing species and ecosystem services .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods", "Dynamic", "Conservation", "Landscape", "Model" ]	[]	2015	Optimal Conservation Outcomes Require Both Restoration and Protection
Standard statistical approaches for prioritization of variants for functional testing in fine-mapping studies either use marginal association statistics or estimate posterior probabilities for variants to be causal under simplifying assumptions . Here , we present a probabilistic framework that integrates association strength with functional genomic annotation data to improve accuracy in selecting plausible causal variants for functional validation . A key feature of our approach is that it empirically estimates the contribution of each functional annotation to the trait of interest directly from summary association statistics while allowing for multiple causal variants at any risk locus . We devise efficient algorithms that estimate the parameters of our model across all risk loci to further increase performance . Using simulations starting from the 1000 Genomes data , we find that our framework consistently outperforms the current state-of-the-art fine-mapping methods , reducing the number of variants that need to be selected to capture 90% of the causal variants from an average of 13 . 3 to 10 . 4 SNPs per locus ( as compared to the next-best performing strategy ) . Furthermore , we introduce a cost-to-benefit optimization framework for determining the number of variants to be followed up in functional assays and assess its performance using real and simulation data . We validate our findings using a large scale meta-analysis of four blood lipids traits and find that the relative probability for causality is increased for variants in exons and transcription start sites and decreased in repressed genomic regions at the risk loci of these traits . Using these highly predictive , trait-specific functional annotations , we estimate causality probabilities across all traits and variants , reducing the size of the 90% confidence set from an average of 17 . 5 to 13 . 5 variants per locus in this data . Recent breakthroughs in high throughput genotyping technologies have ushered in the era of genome-wide association studies ( GWAS ) that have reproducibly identified thousands of genetic variants associated to many diseases and complex traits [1] . GWAS leverage the linkage disequilibrium ( LD ) patterns among genetic markers for probing genetic variation beyond the typed variants . Thus , it is often the case that the associated variant is not itself biologically causal , but rather , a proxy as a result of LD . Identification of causal variants underlying risk loci is performed within fine-mapping studies [2] , [3] , [4] through sequencing ( or array typing and imputation ) followed by variant prioritization using marginal association statistics or posterior probabilities [5] , [6] , [7] . Using these measures , a set of top candidate variants is selected for testing in functional experiments to validate biological causality . Many statistical approaches have been introduced for fine-mapping ranging from a simple ranking of marginal association statistics to Bayesian approaches that integrate elaborate priors [5] , [8] , [9] , [10] , [11] , . Due to the fact that fine-mapping can be casted as a variable selection problem , both LASSO-like procedures that estimate empirical probabilities of inclusion for SNPs based on sub-sampling [13] , as well as Bayesian approaches that perform joint multipoint inference to compute posterior inclusion probabilities [14] have been proposed . The inclusion probabilities provided by these methods offer a natural way to prioritize variants in fine-mapping . However , although neither of the two variable selection approaches assume a fixed number of causal variants , they both require individual level data which is often not readily available . Ranking of SNPs for follow-up analysis can also be performed based on correlation-adjusted t-scores that explicitly take into account the correlation structure among variants , thus requiring individual level data [12] as well . Recent works [5] , [8] , [9] have proposed to estimate posterior probabilities and credible sets for variants to be causal under the simplifying assumption of single causal variant per locus . A key advantage of such approaches is that they only require marginal association statistics which are readily available for large-scale data sets . Large-scale initiatives such as The Encyclopedia of DNA Elements ( ENCODE ) [18] have ascribed functional importance to more than 80% of the human genome and have provided a genome-wide catalogue of regulatory regions . This functional annotation data can be used jointly with the standard association signal to gain insights into the genetic basis of common traits . Indeed , variants associated with certain ENCODE genomic functional annotations such as DNase I Hypersensitive Sites , transcription factor binding sites and expression quantitative loci are enriched among GWAS hits [19] , [20] , [21] , [22] , [23] , with recent work demonstrating that it is possible to integrate such data with the GWAS association signal to identify novel risk loci [10] . However , existing integrative frameworks typically either assume a single causal variant per risk locus [10] that is likely to be incorrect at many risk loci [10] , [24] , [2] , [7] , [25] , [26] , [27] , [28] or do not make use of functional data [29] , [30] . Although ENCODE functional annotation data are clearly beneficial for fine-mapping [22] , a rigorous statistical framework for integrating the different types of information for the purpose of prioritizing plausible causal variants is currently lacking . In this work we introduce PAINTOR ( Probabilistic Annotation INTegratOR ) , a framework to combine external functional annotations ( sets of variants that localize within certain genomic features , e . g . enhancers , repressors ) with genetic association data ( the strength of association between genetic variants and the phenotype ) to improve the prioritization of causal variants in fine-mapping studies . As compared to existing approaches that only rely on the strength of association between genotype and phenotype [31] , [5] , [6] , our framework combines two orthogonal lines of evidence to estimate variant-specific probabilities for causality: functional relevance and genotype-phenotype association . These probabilities can then be used for prioritization of variants for functional validation studies to determine biological causality . More specifically , we incorporate the external functional annotation data through an Empirical Bayes prior [32] with parameters inferred from targeted fine-mapping data , obviating the need to make assumptions on which tissue-specific annotation is relevant to the trait of interest . Finally , budgetary constraints will invariably restrict the number of potential variants that can be validated in functional studies . We address this issue by proposing a cost-to-benefit optimization framework to guide the design of experimental follow-up studies . We use extensive simulations starting from the 1000 Genomes data to show that our approach improves resolution of statistical fine-mapping and is superior to existing frameworks . In our simulations of a trait with a heritability of across 100 risk loci , one needs to test in functional assays an average of 12 . 3 SNPs per locus to identify 90% of all causal variants if using our approach . In addition , if causal variants are preferentially enriched within certain genomic regions [19] , [21] , [10] , [23] , PAINTOR further reduces the average number of SNPs per locus needed to capture 90% of the causal variants to 10 . 4 . We show in simulations that the enrichment estimates provided by PAINTOR are largely unbiased , a fact that we can subsequently use to search for the annotations most phenotypically relevant . We then demonstrate an application of our approach using data from a large-scale meta-analysis study of blood lipid phenotypes ( triglycerides ( TG ) , total cholesterol ( TC ) , high density lipoprotein ( HDL ) , low density lipoprotein ( LDL ) [33] ) and find that causal variants at risk loci are preferentially enriched within coding regions and significantly depleted from repressed regions . In real data , PAINTOR is able to reduce the size of the 90% confidence set from an average 17 . 5 to 13 . 5 SNPs per locus , a reduction consistent to simulation results . We provide software implementing our framework freely available to the research community at http://bogdan . bioinformatics . ucla . edu/software/paintor/ . To illustrate PAINTOR , consider the case of two risk loci that are fine-mapped through sequencing to elucidate the causal variant ( s ) driving the phenotype ( Figure 1 ) . The observed association statistics at all SNPs at these loci are a function of the causal variants , their effect size and the locus-specific LD structure . We use a multivariate normal approximation to connect the LD structure of a fine-mapping locus to the association statistics ( e . g . association z-scores ) which allows for the possibility of modeling multiple causal variants – an important feature since the number of causals variants per locus is typically unknown a priori . We integrate functional annotation data through an Empirical Bayes prior [32] such that the prior probability of a variant to be causal is governed by its membership to functional classes ( see Methods ) . We perform maximum likelihood estimation over all fine-mapping loci using a variant of the Expectation Maximization algorithm to infer the parameters of the model , followed by estimation of the probabilities for each variant to be causal ( see Methods ) . Intuitively , PAINTOR up-weights variants residing in certain functional annotations ( e . g . transcription start sites ) while down-weighting variants within annotations less relevant to the trait ( e . g . intergenic ) . The weight associated to each functional annotation is inferred from the data itself without making any ad-hoc assumptions on which tissue-specific annotations are relevant to the trait of interest . The main output of PAINTOR is a probability for each variant to be causal that can be used for selection of SNPs to be tested for biological causality in functional assays . Numerous approaches for fine-mapping have been proposed , ranging from methods that require individual genotype data to methods that take as input summary association data and integrate functional annotations ( see Table S1 ) . We used simulations to compare PAINTOR to previously proposed methods . It is generally the case that in fine-mapping studies several risk loci are simultaneously sequenced ( or densely genotyped ) and a set of plausible causal SNPs is selected for follow-up in functional assays . We therefore simulated fine-mapping data sets across one hundred 10 KB risk loci that collectively explained 25% of the phenotypic variance in N = 10 , 000 individuals . We created three synthetic “functional annotations” that roughly correspond to coding exons ( 2 . 2% of all variants ) , transcription start sites ( 2 . 2% of all variants ) , and DNase Hypersensitivity Sites ( 30 . 7% of all variants ) and enriched them with causal variants at 9 . 5 , 5 . 7 and 3 . 7-fold to approximately match what we observed in real data ( see below ) . Each simulation resulted in approximately 64 loci that harbor at least one causal variant with 34 harboring a single causal variant and the remaining harboring multiple causal variants ( see Methods ) . We compared all approaches across only loci with at least one causal variant . We find that prioritizing variants using PAINTOR posterior probabilities achieves superior accuracy over existing methodologies ( see Figure 2 and Table 1 ) . Our approach identifies more causal variants at all selection thresholds , and is a consequence of PAINTOR's ability to model multiple causal variants while incorporating functional priors . For example , in order to find ( 50% , 90% ) of all causal variants one needs to select an average of ( 1 . 3 , 10 . 4 ) SNPs per locus if using PAINTOR . In contrast , ranking SNPs using frameworks that assume a single causal variant , such as Maller et al . [5] and fgwas [10] , require ( 2 . 7 , 25 . 4 ) and ( 2 . 0 , 21 . 5 ) SNPs per locus , respectively . In general , we observe an increase in performance for methods that incorporate functional data and allow for multiple causal variants at a risk locus ( see Tables 1 and 2 ) . Despite having access to individual level data , variable selection strategies [13] , [14] were less accurate than PAINTOR in our simulations ( see Figure 2 and Table 1 ) . Ranking SNPs based on correlation-adjusted t-scores [12] was superior to existing methodologies , however , still failed to achieve the same level of accuracy of PAINTOR , requiring an average of ( 2 . 0 , 13 . 3 ) SNPs per locus to find ( 50% , 90% ) of all causal variants . Across all methodologies , the relative performance holds irrespective of whether SNPs are prioritized across all fine-mapping loci or within each locus independently ( generally the latter strategy is sub-optimal ( see Table 1 ) ) . Finally , we note that iterative conditioning , a method typically used to detect multiple independent signals , performs worse than the prioritization strategies described here ( see Figure S1 ) [7] . Interestingly , as the number of SNPs selected for follow-up increases , the naive approach of selecting based on association p-value alone attains high accuracy , most likely due to the much smaller set of assumptions as compared to other methods . Having established that PAINTOR increases fine-mapping accuracy over existing methods in simulations , we next explored the gain in performance attributable to having access to functional annotation data . We find that prioritizing variants using PAINTOR with functional data increases accuracy at all significance thresholds . For example , in order to find ( 50% , 90% ) of all causal variants one needs to select an average of ( 1 . 3 , 10 . 4 ) SNPs per locus if integrating functional data as opposed to ( 1 . 7 , 12 . 3 ) if excluding annotation data . We note that our approach that does not empirically estimate the prior , but uses the known prior information does not lead to superior performance over PAINTOR in these simulations ( see Table 1 ) reflecting the fact that the prior probability for each SNP is accurately estimated . Furthermore , as the size of the fine-mapping locus is increased , PAINTOR continues to outperform simpler approaches . In particular , to resolve 90% of the causal variants for loci ( 10 Kb , 25 Kb , 50 Kb ) in size , one needs to select ( 27 . 4 , 52 . 3 , 110 . 7 ) SNPS per locus if ranking on posterior probabilities assuming a single causal variant as opposed to ( 11 . 4 , 16 . 0 , 24 . 1 ) SNPs per locus if ranking using PAINTOR ( see Table 3 ) . We next sought to determine at what types of loci is functional prior data providing the biggest increase in accuracy . Loci where the association signal is strong ( i . e . loci where the p-value at the causal variants are in the top quartile across all loci with at least one causal variant ) do not gain much from integration of functional annotation data , with the number of SNPs required to find 90% of the causal variants decreasing by only 6 . 5% . On the other hand , at loci where the association signal is weak ( i . e . loci where the p-value at the causal variants are in the bottom quartile ) we observe a 21 . 4% decrease in the total number of SNPs to be followed-up to find 90% of all causal variants ( see Table S2 ) . This suggests that as the causal status for a SNP becomes increasingly ambiguous on the basis of association data alone ( e . g . small effect size ) , the importance of incorporating additional sources of information is magnified . It is not guaranteed that the true causal variant will be present in the fine-mapping data set due to technical reasons ( e . g . capture sequencing technology or imputation accuracy ) . To explore this scenario , we simulated fine-mapping data sets at a locus 100 Kb in size after which we masked the true causal ( s ) from the data ( see Methods ) . To measure fine-mapping performance when causal variant is absent from the data , we looked at the distance in base-pairs between variants in the top N SNPs to the true masked causal SNP . As expected , we observed a decrease in performance when causal variants are absent from the fine-mapping dataset ( e . g . the average median distance to the true causal variant in the set of top 5 SNPs increases by 6% when the causal variant is masked , see Table 4 ) . The rather small nominal decrease in localization distance suggests that accurate localization may be attained even in the absence of the causal variant . Alternatively , we can recast the observed improvement in causal variant localization when incorporating functional annotations as a decrease in size of the set of SNPs to account for a fixed amount of posterior probability mass . We extend existing work for single-locus fine-mapping [5] , [8] , [9] , [7] to define an ρ-level causal set as the set of top SNPs ( rank-ordered based on probabilities ) across all fine-mapping loci that consume an fraction of the total posterior probability mass . We observe a reduction in the number of SNPs within the 90% , 95% and 99% confidence sets when using functional annotations as compared to no functional data ( see Table 2 ) . In addition , although PAINTOR with annotation yields fewer SNPs with high probability than the PAINTOR with no annotation ( 232 . 8 vs 265 . 2 at a threshold of ) , having access to annotation yields more simulated causals with high posterior probability ( 78 . 6 vs 73 . 8 at a threshold of ) ( see Table S3 ) . A vast resource for functional annotations is the ENCODE project [18] , which has ascribed regulatory biological function to a large fraction of the human genome and has shown that regulatory DNA regions are highly cell-specific . Coupling this insight with the fact that for most complex diseases the relevant tissues are unknown , stresses the importance of carefully selecting cell-specific annotations for any specific trait [22] . A byproduct of our framework is the estimation of enrichment of causal variants within functional annotations ( i . e . the ratio of prior probability of causality for SNPs within annotation versus those outside the annotation ) . Therefore , we can use PAINTOR to infer which functional annotations show significant effect on the probability of causality and use only those annotations to estimate probability of causality . To assess how accurately PAINTOR can recapitulate functional enrichment , we simulated fine-mapping studies over 100 loci with a synthetic functional annotation ( see Methods ) and either enriched or depleted causal variants within this annotation . We also compared our approach to fgwas [10] as it too is capable of inferring enrichment from summary data . Figure 3 demonstrates that both PAINTOR and fgwas are able to provide unbiased estimates of enrichment . However , we find that PAINTOR is more efficient than fgwas , and has a smaller variance attached to those estimates . We note that as causal variants become increasingly depleted from functional categories , fgwas tends to fail to converge ( e . g . fgwas fails in nearly 21% of cases for simulations with 8-fold depletions ) . Finally , we assessed PAINTOR and fgwas for more realistic annotation data ( i . e . contiguous segments in the genome ) and find that both methods attain very similar results ( see Figure S2 ) . Although PAINTOR ( and previous methods ) provide a quantification of the probability of each variant to be causal that can be used to rank variants based on their plausible causality , it remains unclear how to choose the number of variants to test in functional assays . The optimum number is constrained by the budget of the study and by an implicit cost to benefit ratio for selecting the optimal number of SNPs to be followed up . We propose a framework that assumes that every causal variant identified adds a benefit ( ) while every selected variant is tested at a cost ( ) ; therefore , the utility function we propose to maximize is , where is the total number of true causal variants from the total number of selected SNPs ( ) . We note that the ratio is the critical parameter of the utility function . Using the results from simulations with functional annotation enrichment described above , we assessed the capacity of the proposed utility function in selecting the number of SNPs for follow-up under various values for the ratio ( Figure 4 ) . For example , at a ratio ( the benefit of finding a causal outweighs 10 times the cost of testing 1 SNP ) , the utility is maximized by selecting approximately 3 . 5 SNPs per locus for validation resulting in 72 . 6% of causal variants successfully identified ( see Figure S3 ) . Selection of a set of variants for follow-up is usually performed based on a threshold on posterior probability or based on credible sets that account for a given amount ( e . g . ) of the probability of capturing all causal variants [5] , [8] . We assessed these two strategies for selecting variants for functional testing within the context of our benefit-to-cost framework . We find that a posterior probability threshold of ( 0 . 9 , 0 . 5 , 0 . 1 ) roughly corresponds to optimizing benefit-to-cost-ratios of . These results suggest that a simple translation of the arbitrary thresholds on posterior probabilities into cost-to-benefit optimum is attainable . In a similar fashion , we can assess credible sets within our cost-to-benefit framework . For example , the 90% credible set yields an average of 393 SNPs which is approximately 88% of the optimum for a benefit-to-cost of . To validate our approach , we applied PAINTOR to association summary data from a large meta-analysis of four lipid traits . Our goal was to build a model that incorporated all the independent sources of available information ( i . e . association signals alongside carefully selected functional annotations ) to produce a prioritization of plausible causal SNPs for these phenotypes . We used the GWAS hits reported by Teslovich et al . [33] under the assumption that these regions contain causal variants and therefore well-suited to fine-map using PAINTOR . We first ran our method on 450 cell-type-specific annotations ( see Methods ) and fit the model to each annotation independently on both the original and densely imputed data sets for all four traits . Consistent to previous works , we observe that imputation consistently enhances the signal of enrichment [34] , [23] , [10]; for example , for HDL , the relative probability for causality for coding exons increases from 7 . 4 to 12 . 4 from using the original data to 1KG-imputed data ( see Table S4 ) . This effect is most likely due to the availability of more variants through imputation thus being able to localize the association signal to genomic annotation more accurately . Across the four traits in general , we see consistent signal of increased relative probability for causality within transcribed regions ( e . g . exons and transcription start sites ( TSS ) ) and a depletion of causal variants in repressed regions; for example , for TG , the coding exons show a log2 relative probability for causality of 3 . 4 while the repressed regions show an log2 relative probability of −1 . 6 . Having identified functional annotations that are relevant to the four traits of interest ( see Table 5 ) , we devised trait-specific PAINTOR models that included the top marginal annotations in conjunction with the association statistics to estimate the probability of causality for all SNPs from the risk loci on the densely imputed data sets ( see Methods ) . Table 6 shows the HDL SNPs that attain a posterior PAINTOR probability greater than 0 . 9 ( results for the other traits are displayed in the Tables S5 , S6 , S7 ) . Unsurprisingly , the majority of these top SNPs localize in functional elements and attain a high marginal association statistic . We observe an abundance of liver associated cell types , DNase Hypersensitivity Sites , and genic elements annotated to these top SNPs . Notably , PAINTOR identifies four non-synonymous variants ( rs7607980 , rs1260326 , rs5110 , rs13107325 ) , two of which were not reported in the initial Teslovich et al . findings . Overall by incorporating functional annotations we see a marked improvement in fine-mapping resolution across all four traits as indicated by a reduction in the 90% confidence sets relative to PAINTOR models with no annotations of 19 . 0% , 34 . 9% , 50 . 6% , and 24 . 2% for HDL , LDL , TC , and TG , respectively ( Table 7 ) . This corresponds to approximately an average reduction of 17 . 5 to 13 . 5 SNPs per locus across the four traits . Recent efforts by large consortia such the ENCODE have provided a genomic map of regulatory regions and have shown that GWAS associated variants are preferentially enriched within these regions . In this work , we propose a principled approach to unifying these genomic features with the standard association signal to improve the localization accuracy in fine-mapping studies . Our method relies on empirical data to select trait-specific genomic annotations , thus removing the need for ad-hoc selection of relevant functional annotations a priori . Through simulated and real data results , we have shown that our integrative framework is able to reduce the number of variants that need to be investigated to identify causal variants that alter risk of disease . Our method shares similarities to recent integrative approaches proposed in the context of GWAS [10] . Although conceptually both approaches integrate functional and association signal , the two methodologies are fundamentally distinct in their aims . Whereas [10] seeks to identify novel risk loci by leveraging functional information , we instead propose our method as way to refine signal at known GWAS loci . This fundamental distinction leads to different statistical models and optimization procedures allowing for superior accuracy for refining association signal through fine-mapping . In addition our method addresses a limitation of [10] by allowing for the possibility of multiple causal variants at a risk locus . Several hierarchical Bayesian methods have been developed that combine prior information with genomic association data to help prioritize variants in various contexts [29] , [30] . The main contribution of our approach is that we explicitly account for LD between SNPs which we can learn from external reference panels such as the 1000 Genomes . Additionally , because we do not take a fully Bayesian approach [30] ( i . e . integrate over the entire hyper-parameter space ) , we are able to devise computationally efficient algorithms that allow our method to search over the ever-increasing number of functional annotations ( e . g . ENCODE ) to identify the most informative subset while retaining the ability to model multiple causal variants . We have shown that PAINTOR can unbiasedly estimate enrichment of causal variants in different functional elements on the basis of summary association data alone . This may prove to be particularly important as access to individual genotype data is more cumbersome than summary-level statistics . The unbiased nature of the estimation procedure may provide clues to the genetic basis of common traits . For example our results suggest that although coding variants are more likely to be causal than regulatory variants , the majority of the genetic variation contributing to the trait at these risk loci may lie within regulatory as opposed to coding regions due to the larger number of variants residing in regulatory regions . This is consistent with recent work that concluded that variants in regulatory regions show a higher contribution to traits than coding variants , however , such an analysis required individual level data [23] . One interesting implication of our results is that while higher-order functional data is very useful for gleaning insight into to the genetic architecture of human diseases genome-wide [20] , [23] , a critical component of accuracy in a fine-mapping study is the sample size ( see Table S8 ) . Consequently , the success of a fine-mapping experiment may hinge on first obtaining an adequate sample size and then augmenting that sample size with functional data . These findings are largely in-line with what was previously reported in the context of GWAS [10] . In this work we have applied our framework to known risk loci identified in GWAS in the search for plausible causal variants . As future work , our approach could be extended to risk loci that do not pass a genome-wide stringency , potentially leading to discovery of novel risk loci . Additionally , risk loci for related traits that are known to share a genetic basis could potentially be combined , leading to an increase in power to identify variants that contribute to both traits . Finally , we anticipate that the approximations of the non-centrality parameters could be handled in a more principled fashion using a Bayesian approach that integrates a prior distribution of effect sizes . We leave a thorough investigation of these directions as future work . A standard approach to model the strength of association of genotype to phenotype is through the Z-score . For a continuous phenotype , the trait values are marginally regressed on each SNP and the corresponding Z-score is taken to be the Wald statistic ( i . e ) , which is distributed under the null . For case-control designs , the Z-score can also be obtained through the standard test statistic for two proportions ( assuming equal sample sizes of : , where denotes the frequency of the SNP in the cases ( controls ) and . We define a fine-mapping locus as a contiguous region of the genome flanking a GWAS “hit” on both sides . Let be a vector of Z-scores from the locus of length . In addition , let be the corresponding LD matrix of pairwise correlation coefficients for locus that can be derived directly from individual level data if available , or approximated using an appropriate reference panel such as the 1000 Genomes . We obtain annotations from external repositories ( e . g . ENCODE [18] ) and for each SNP , create a -length binary annotation vector , where if the SNP at the locus is part of annotation . For example , one such annotation could be all coding sites and the annotation vector will contain a 1 only if the SNP is located within coding region . We note that and serves to represent the “baseline” annotation whose corresponding coefficient can be interpreted as the baseline prior odds for causality of any SNP within the set of fine-mapping loci . Let be the effect size of the annotation on the probability of a SNP being causal and the non-centrality parameter , , be the standardized effect size of SNP at locus . Finally , let be an indicator vector of causality where if SNP at locus is causal and 0 otherwise . Now , we can define the likelihood of the data relative to these terms as: ( 1 ) where the sum is taken across all causal indicator vector sets . We note that in order to keep the enumeration of the causal vector sets combinatorially tractable , we restrict the total number of potential causal variants at each locus to three or less in practice ( see Figure S4 for assessment of run time versus number of causal variants considered ) . We define the annotation effect on the causal probability through a standard logistic model: ( 2 ) and relate the causal set of SNPs to the observed association Z-scores under a standard multivariate normal assumption [35] , [36] , [37] as: ( 3 ) where denotes the elemental pairwise multiplication between two vectors . A summary of model parameters can be found in Table 8 . In order to compute the probability of causality , we must first fit the data to our model . We accomplish this through a maximum likelihood estimation over . The formulation of our approach lends itself to the standard Expectation Maximization ( EM ) algorithm . The E-step of the EM involves computing at each locus independently , the posterior probability of each using an application of Bayes Theorem: ( 4 ) To obtain the posterior probability , , for each SNPi , j we marginalize across all such that . ( 5 ) Despite the fact that posterior probabilities are calculated independently at each locus , we can set up the objective function to aggregate the results and borrow information across loci to compute estimates of . In doing so , we prevent over fitting of the data to any one locus , offering more robust estimates of the model parameters leading , in turn , to more accurate posterior probabilities . We define our Q function for the M step as follows thereby partitioning the likelihood , decoupling the estimation of from the . We simplify to obtainwhich is a concave function whose gradient is simply We optimized this function using the NLopt C++ package's implementation of the limited-memory BFGS algorithm [38] , a quasi-Newton method that only requires the objective and the gradient as input [39] . As stated previously , we fix the non-centrality parameters , , and only optimize over due to the fact that our model would be over-specified otherwise . Specifically , we set the non-centrality parameters at each SNP to the observed Z-score if the absolute Z-score is greater than 3 . 7 ( corresponding to a p-value of 10e-4 ) or the sign of the observed Z-score times 3 . 7 otherwise . Simulation results show that our strategy yields high accuracy to detect causal variants among several simulated approaches to approximate ( Figure S5 ) . Starting from the 1000 Genomes ( 1 KG ) European samples , we used HAPGEN [40] to simulate fine-mapping data sets over 10 Kb loci . We filtered monomorphic/rare SNPs ( MAF 0 . 01 ) and normalized genotypes to be mean-centered with unit variance . For each simulation we randomly chose one hundred 10 Kb loci and randomly assigned SNPs to binary annotations at a pre-specified proportion . We drew causal status for each SNP according to the logistic model above and varied to induce a desired prior probability for causality for SNPs part of the “functional” annotation , while maintaining an approximately fixed number of causals – typically one per locus in expectation . For example , to induce an 8-fold causal enrichment in a synthetic “functional” annotation that contained 1/3 of the SNPs , the ( ) values were set to be ( 4 . 62 , −2 . 15 ) . We note that the random assignment of causal status would lead to loci with either zero ( 36 ) , one ( 34 ) , or multiple causal ( 30 ) variants on the average . Once we established the causal SNPs , we used a linear model to simulate continuous phenotypes such that the causal SNPs aggregated to explain a fixed proportion of the phenotypic variance ( ) . This phenotypic variance was partitioned equally amongst all the causal SNPs ( qualitatively similar results were obtained when phenotypic variance was unevenly partitioned among causal variants ( see Figure S6 ) ) . In particular , the individual's phenotype was drawn according to , where is the total number of causal variants , is the effect size of the causal SNP , is number of copies of the risk allele ( randomly assigned as reference or alternate ) for individual m , and . Finally , we calculated association Z-scores ( ) at each SNP by taking the Wald statistic from the regression of the on , where Y is a vector of phenotypes for M individuals and is the vector of corresponding genotypes for the SNP at the locus . For simulations that required loci greater than 10 KB , we instead drew Z-scores from a Multivariate Normal distribution with covariance equal to LD based on the European 1 KG and non-centrality parameters at causal sites drawn from a Normal distribution with mean 5 and standard deviation 0 . 2 . When measuring performance of our simulations , we examine the proportion of causal SNPs identified as a function of the average number of SNPs per locus selected for follow-up restricted to loci that contain at least one causal variant ( we show in Figure S7 that using Positive Predictive Value as a metric of accuracy attains qualitatively similar results ) . We compared our approach to a several of existing methods that can be used for fine-mapping[5] , [10] , [12] , [13] , [14] . To compute Maller et al . [5] posterior probabilities , we first calculated Bayes factors with the R package , BayesFactor , using the default settings . We converted the resultant Bayes factors into posterior probabilities of association using the following formula: . We show in Figure S8 and Supplementary Note S1 that posterior probabilities approximated directly from the Z-scores give virtually indistinguishable results . We downloaded fgwas [10] version 0 . 3 . 4 from GitHub and ran the software using the -fine flag . Due to the fact that we fit linear models to obtain Wald statistics for each SNP , we were able to provide standard errors for the estimates of the prior variance . We segmented our input based on the individual loci as instructed in user manual , but provided a single file that contained all the fine-mapping SNPs so that fgwas could compute enrichment . The Guan and Stephens [14] method is implemented in the software piMass which we ran using the flags and MCMC parameters given in the user manual as defaults ( burn-in = 1000 , samples = 100 , 000 , step-size = 10 ) . We used the posterior inclusion probabilities ( PIPs ) that had undergone Rao-Blackwellisation for prioritization due to the fact these had superior performance over naive PIPs . The R package implementing LLARRMA [13] was run using the default settings . Zuber et al . was implemented in the R package , care , which we also applied to the data using the default settings . We prioritized variants using the square of the CAT scores as described in [12] . We note that with exception fgwas , all the aforementioned methods were applied to each locus independently . Conditional analysis is a common procedure used to tease out secondary signals at associated loci [41] . For a single locus , we iteratively condition on the SNP most strongly associated with the simulated phenotype . We accomplish this in a step-wise fashion through marginal regression of the phenotype onto each SNP and subsequently conditioning on the one that is most significantly associated . At each iteration a new SNP will enter the regression model as a covariate until all the causal SNPs have been selected . The order in which the SNPs enter the model provides a natural ranking thus enabling us to compare iterative conditioning to other methods that rank SNPs probabilistically . As expected , we show in Figure S1 that conditioning is suboptimal for fine-mapping . We explored whether integration of the location of tissue-specific regulatory and coding DNA regions could increase resolution in statistical fine-mapping . The ENCODE [18] project provided a wealth of genomic landmarks that were systematically integrated to segment the genome into seven major classes: transcription start site and predicted promotor region ( TSS ) ( 1 . 2% ) , predicted promotor flanking region ( PF ) ( 0 . 7% ) predicted enhancer ( E ) ( 1 . 8% ) predicted weak enhancer ( WE ) ( 2 . 5% ) , CTCF-enriched element ( CTCF ) ( 0 . 1% ) predicted transcribed region ( T ) ( 19 . 3% ) and finally , predicted repressed or low-activity region ( R ) ( 69 . 6% ) . We examined these genomic segmentations for the six primary ENCODE cell lines: gm12878 ( lymphoblastoid ) , h1hesc ( embryonic stem cells ) , helas3 ( cervical cancer ) , hepg2 ( liver carcinoma ) , huvec ( umbilical vein endothelial cells ) , and k562 ( chronic myelogenous leukemia ) . In addition , we also explored 403 broadly defined ( peak 1 Kb ) DNase I Hypersensitivity Sites spanning numerous tissues and cell lines . Of these 403 DHS I maps , 349 came from Maurano et al . [19] , 73 DHS I annotations from Thurman et al . [42] , with the remaining DHS annotation being an overall DHS map derived from UCSC genome browser . These annotations have been used recently in the context of GWAS [10] . Due to the fact that we fit our model using maximum likelihood , a natural way to ascribe statistical confidence to the inferred parameters is to use a likelihood ratio test . For example , to calculate the significance of a single annotation , we can compare a fitted null model to a model that contains the annotation under consideration using the following test statistic: . We demonstrate in simulations that under the null , this test statistic follows approximately its theoretical distribution ( see Figure S9 ) . In addition to a point estimate for the enrichment of functional annotation , it would be useful to derive an estimate of the variance . Unfortunately , the complex structure of the likelihood makes it difficult to derive an analytically tractable parameter covariance estimator . However , since we assume fine-mapping loci to be independent , we propose to use bootstrapping ( i . e . re-sampling entire loci with replacement ) and subsequently re-fitting the model ( see Methods ) . We confirm that such a strategy does indeed reproduce a correct estimate of the parameter variance in simulations . We find that the mean bootstrap standard deviation largely mirrors the “true” standard deviation of the parameter estimates ( see Figure S10 ) . As a result , a confidence interval based on the bootstrap standard deviation will attain desirable coverage properties due to the fact that estimation of the model parameters is unbiased . The budget of a fine mapping follow-up study constrains the total number of causal variants to be further examined for evidence of causality . This motivates approaches that , in addition to providing a prioritization of SNPs , also identify an optimal number of SNPs to be tested . We introduce here a benefit-to-cost framework for selecting the optimal number of SNPs for follow-up . Our framework assumes that every causal variant identified adds a benefit ( B ) while every selected variant is tested at a cost ( C ) ; therefore , the utility function we propose to maximize is U = BNc - CNt , where Nc is the total number of true causal variants identified from the total number of selected SNPs . A key parameter of the utility framework is the ratio of of benefit to utility . Publicly available GWAS summary data across four blood lipids phenotypes was downloaded from public access websites [43] . Data was part of a meta-analysis conducted in individuals of European ancestry that examined four plasma lipid traits ( number of significant loci ) : LDL cholesterol ( 14 loci ) , HDL cholesterol ( 37 loci ) , trigylcerides ( 23 loci ) , and total cholesterol ( 24 loci ) . From the original 2 . 0 M SNPs , we imputed an additional 5 . 3 million Z-scores using ImpG-Summary [34] . For each significant GWAS hit reported by Teslovich et al . , we centered a 100 KB window on the lead SNP and estimated LD from the European reference panel of the 1 KG . We chose a conservative window of 50 Kb on either side of the GWAS hit , as it has been previously shown that within European populations , average LD decays after 25 KB [44] . These loci contained an average of 718 SNPs in the 1000 genomes reference panel , of which we were able to on average accurately impute 261 ( see Table S9 ) . This resulted in 2837 ( 10778 ) , 1231 ( 3903 ) , 1693 ( 5504 ) , 1615 ( 5513 ) SNPs ( with 1 KG imputation ) to which we fit our model for HDL , LDL , TC , and TG respectively . In addition , we created the corresponding pool of functional annotations described above for every SNP in a window . We analyzed the dataset using PAINTOR in two phases . In the first phase we fit our model for each annotation independently to ascertain the functional annotations most phenotypically relevant . We did this for all four lipid traits for both the original and densely imputed data sets . After running PAINTOR marginally on each annotation , we selected the the top five most significant annotations for the final model ( denoted with a * in Table: 5 ) . We note that in the case of experimental replicates ( i . e . same tissue and class ) , we only report the top replicate .	Genome-wide association studies ( GWAS ) have successfully identified numerous regions in the genome that harbor genetic variants that increase risk for various complex traits and diseases . However , it is generally the case that GWAS risk variants are not themselves causally affecting the trait , but rather , are correlated to the true causal variant through linkage disequilibrium ( LD ) . Plausible causal variants are identified in fine-mapping studies through targeted sequencing followed by prioritization of variants for functional validation . In this work , we propose methods that leverage two sources of independent information , the association strength and genomic functional location , to prioritize causal variants . We demonstrate in simulations and empirical data that our approach reduces the number of SNPs that need to be selected for follow-up to identify the true causal variants at GWAS risk loci .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "genetic", "polymorphism", "gene", "expression", "genetics", "gene", "regulation", "biology", "and", "life", "sciences", "population", "genetics", "locus", "control", "region", "evolutionary", "biology", "alleles", "genetic", "loci" ]	2014	Integrating Functional Data to Prioritize Causal Variants in Statistical Fine-Mapping Studies
Gliomas are a highly heterogeneous group of brain tumours that are refractory to treatment , highly invasive and pro-angiogenic . Glioblastoma patients have an average survival time of less than 15 months . Understanding the molecular basis of different grades of glioma , from well differentiated , low-grade tumours to high-grade tumours , is a key step in defining new therapeutic targets . Here we use a data-driven approach to learn the structure of gene regulatory networks from observational data and use the resulting models to formulate hypothesis on the molecular determinants of glioma stage . Remarkably , integration of available knowledge with functional genomics datasets representing clinical and pre-clinical studies reveals important properties within the regulatory circuits controlling low and high-grade glioma . Our analyses first show that low and high-grade gliomas are characterised by a switch in activity of two subsets of Rho GTPases . The first one is involved in maintaining normal glial cell function , while the second is linked to the establishment of multiple hallmarks of cancer . Next , the development and application of a novel data integration methodology reveals novel functions of RND3 in controlling glioma cell migration , invasion , proliferation , angiogenesis and clinical outcome . Gliomas are brain tumours originating from the glial cells and neural stem cells that surround and support neurons [1] . They are classified on the basis of their clinical and histopathological characteristics in four grades with progressively more severe features . Grade I and II gliomas ( astrocytomas , oligodendrogliomas and oligoastrocytomas ) are considered relatively benign , well-differentiated tumours and have 5 year survival rates of 59 . 9% [2] . Amongst patients diagnosed with low-grade gliomas , approximately 70% progress to grade IV glioblastoma multiforme ( GBM ) within 5–10 years of diagnosis [3] . De novo GBM constitute the majority of grade IV glioma and are powerful inducers of angiogenesis , highly proliferative and invasive . They are largely resistant to treatment and have poor prognosis with two years survival rates as low as 3 . 3% [4] . A number of studies have identified key genomic alterations in GBM able to induce transformation in non-tumorigenic cells such as mutations within EGFR [5] [6] and PDGFRA [7] . A molecular classification for GBM has been proposed subdividing the tumours based on their molecular profile into 4 groups . This includes the classical type ( EGFR amplification , CDKN2A deletion ) , proneural type ( PDGFRA amplification , PTEN deletion ) , the mesenchymal ( NF1 deletion ) and the neural type [8] . However , angiogenic and invasive phenotypes are observed across the different groups , making this classification unsatisfactory . For example , EGFR amplification reminiscent of the classical type has been shown to drive invasive growth [9] . Amplification and overexpression of EGFR leads to activation of Ras GTPase and Akt signalling pathways controlling cell growth , differentiation and survival of tumour cells [10] [11] . The complexity of the factors involved in the biology of gliomas makes it difficult to develop a comprehensive model underlying GBM progression . Here we address this important challenge by integrating functional genomics datasets representing existing knowledge , clinical studies and in vivo and in vitro glioma models . We first show that network modules derived from a comprehensive integration of protein interaction databases and defined by a high density of genes differentially expressed between low and high-grade gliomas are consistent with the hypothesis that Rho GTPases may be part of a key regulatory mechanism controlling hallmarks of high-grade glioma . A key feature of GBM is invasion of tumour cells into the surrounding brain tissue and members of the Rho GTPase family known to control actin cytoskeleton dynamics and cell migration have been implicated in the survival and invasion of tumour cells [12] [13] , [14] . In addition , RhoA expression correlates to tumour grade in astrocytomas [15] . We reconstruct stage specific gene co-expression networks and analyse the connectivity profile of Rho GTPases . This reveals that regulatory Rho GTPases separates in two groups , one active in low and the other in high-grade gliomas . The functional profile of the putative targets of these two sub-sets predicts the functional differences observed between low and high-grade glioma . Further characterisation of a high-grade glioma transcriptional network highlights a pivotal role of the Rho GTPase RND3 ( also known as RhoE , Entrez: 390 ) in controlling tumour proliferation , migration and invasion . Ultimately , the clinical relevance of this regulatory network is proving that copy number variation in the RND3 gene is predictive of clinical outcome . Our study is based on a complex data analysis workflow which includes several complementary reverse engineering techniques to address the important challenge of generating and validating hypotheses outlining the main factors underlying the control and maintenance of glioma stage . The strategy we followed , which is summarised in Fig 1 , is based on several cycles of data acquisition , computational analysis , hypothesis generation and experimental validation . The workflow consisted of five distinct but interconnected steps . Step 1- Integration of protein-protein interaction networks with gene expression data derived from a human clinical study: This represents low and high-grade gliomas and were used to modularise a large network of known human protein-protein interactions ( PPi ) . The analysis of these modules identified a unique sub-network of regulatory factors , which represented a number of Rho GTPases . Step 2- Revealing the linkage between subsets of Rho GTPases and glioma grade: Genome scale gene expression data were used to construct co-expression networks centred on the regulatory Rho GTPases identified in step 1 . This revealed a switch in activity of Rho GTPases between grade II and grade IV glioma . Step 3- Reverse engineering of regulatory networks from an in vivo model: We used the experimental glioma model in the chick chorioallantoic membrane [16] which revealed regulation of Rho GTPases in the angiogenic and invasive phase of glioma development . Network analysis revealed RND3 as the most connected factor for high-grade glioma . Step 4- Validation step: The predictions from the network analysis were validated through functional assays ( apoptosis , proliferation , migration and angiogenesis in vitro and in vivo ) . Step 5- Development of data integration strategy: We developed a novel , multi-level data integration pipeline in order to shed light on the possible mechanisms of RND3 control of tumour function . This revealed a link between RND3 and DNA replication factors such as MCM3 , which we validate in vitro . In order to gain insight in the mechanisms underlying the pathophysiology of human glioma , we first developed an interaction network representing genes differentially expressed between grade II astrocytoma and grade IV glioblastoma . We then applied a modularization procedure to identify sub-networks of highly interconnected proteins , thus capturing important biological networks potentially representative of the differences between low and high-grade glioma . The procedure identified four main modules , with a highly statistically significant functional enrichment profile ( Fig 2 ) . The entire list of significantly enriched Gene Ontology terms can be found in S1 Dataset . Interestingly , the three largest modules ( M2-4 ) mainly comprised genes up regulated in grade IV gliomas and were enriched in typical effector functions in cancer . More precisely , the largest of the effector modules ( M2 ) was enriched in genes linked to apoptosis , blood vessel development and inflammatory response and included the oncogenes JUN , FOS and BCL3 ( for reviews of known oncogenes see [17] , [18] ) . The second largest module ( M3 ) included genes involved in cell adhesion , extracellular matrix , blood vessel development , adherens junctions and integrin complex and also included the oncogenes ERBB2 , MET and EGF1 . Effector module M4 was predominantly enriched with proliferation related functions such as cell cycle , DNA repair and DNA replication and the oncogenes MDM2 , CDK6 , FOXM1 and BIRC5 . We noticed that the smallest module ( M1 ) comprised mainly proteins linked to GTPase signalling ( 41/54 ) , suggesting that this class of proteins may represent more important regulators of tumour effector functions than previously anticipated . By examining the functional enrichment profile of this cluster we discovered that the only enriched family of GTPases was the Rho family with 5 members ( CDC42 , RHOJ , RAC2 , RHOC , and RND3 ) ( False Discovery Rate < 1 . 93−6 ) . This observation is consistent with the pro-tumour function of CDC42 , RHOG , RAC1 and RHOA in glioma [14] [13] [19] [20] [15] [21] [22] . However , our model suggests a broader role of GTPases in glioma than previously thought . We therefore tested this hypothesis by inferring the structure of grade II ( Fig 3A ) and grade IV ( Fig 3B ) glioma transcriptional networks in the neighbourhood of all Rho GTPases . We discovered that genes encoding for Rho GTPases were separated in two groups when characterized by tumour grade-specific connectivity profiles ( Fig 3D ) . Seven Rho GTPases ( RND1 , RND3 , RAC3 , RHOA , RAC2 , RAC1 and RHOC ) showed a significantly greater connectivity in grade IV glioma and seven ( CHP , RHOD , RHOF , RHOB , RHOQ , RND2 , RHOBTB3 ) showed greater connectivity in grade II glioma . In order to validate this differential connectivity we used expression data from grade II and grade IV glioma within the Cancer Genome Atlas database . 16 Rho GTPases could be matched between the original data and TCGA grade II and grade IV datasets , of which 13 ( 81% ) showed the same trend as the original analysis ( S2 Dataset ) . Remarkably , Rho GTPases with a higher number of connections in grade IV tumours were correlated to genes in the grade IV networks with a functional profile that included many of the effector functions associated with high-grade glioma ( immune response , regulation of apoptosis , regulation of cell proliferation , response to cytokine stimulus , inflammatory response , cell adhesion ) ( Fig 3C ) . On the other hand , in the grade II networks targets of Rho GTPases with a higher proportion of connections in grade II glioma showed a functional profile consistent with glial cells ( putative target genes for this group of regulators were enriched with clathrin coated vesicle membrane , dendrite , axonegensis , regulation of exocytosis , neuron development and regulation of synaptic transmission functional terms ) ( Fig 3C ) . The full list of Gene Ontology terms from the analysis of the grade II and grade IV networks can be found in S3 Dataset . The functional profiles of the targets of the two groups of Rho GTPases also contained a subset of similar functional terms such as synapse , cell projection , neurotransmitter transport , vesicle membrane , cell junction . This suggests that in grade IV gliomas the link between Rho GTPases and normal glial function is only partially interrupted . Interestingly , we found that although most ( 15/19 ) of the Rho GTPases were differentially expressed between grade II and grade IV gliomas , there was not a clear trend in the direction of change ( S2 Fig ) . Additionally , none of the Rho GTPases were differentially expressed between grade II and grade III gliomas ( S2 Fig ) . Overall , this supports the hypothesis that distinct subsets of Rho GTPases are potentially important players in grade IV gliomas . In order to validate the networks developed from the clinical study , we implanted a grade IV glioma derived cell line ( U87MG ) in the chicken egg chorioallantoic membrane ( CAM ) and followed the transcriptional profile of tumour cells for the first 5 days post implantation with 10 equally spaced time points . In this well established in vivo model , tumours implanted on the CAM form avascular , solid tumours which stimulate angiogenesis and become vascularised within 48 hours [16] . We found 2 , 999 unique genes differentially expressed during 5 days of tumour growth . We used these genes to build a high level map representing the dynamics of transcriptional changes in the developing tumour . A gene clustering procedure identified 14 distinct clusters which were stratified according to their expression profiles ( S3 Fig ) ( for gene lists of each cluster and the Gene Ontology analysis see S4 Dataset ) . Functional profiling of the early transcriptional response post-implantation ( up to 12hrs ) revealed an increase in expression of extracellular matrix components . The intermediate transcriptional response ( 13 to 24hrs ) was characterised by a wave of transcriptional repression related to the inflammatory response , regulation of apoptosis , lipid biosynthetic process and lysosomes functions . These included the extracellular matrix remodelling genes , BMP2 and BMP6 , chemokine signalling genes CCR1 and CCL20 and the inflammatory cytokine IL1A ( S3 Fig ) . Consistent with the development of a fully vascularized tumour at 48 hours post-implantation we saw a dramatic increase in the expression of genes related to the cell cycle , cell adhesion , blood vessel development , and cell migration in the time window between 37 and 48 hours . Four Rho GTPases were up-regulated in this time window . Two of these ( RHOC and RND3 ) were among the Rho GTPases with a strong grade IV specific connectivity profile in multiple datasets ( Fig 3D ) and the remaining two ( RHOBTB1 and RHOBTB2 ) were non-specific . Since GTPases and their potential transcriptional targets were all modulated at this specific time window we hypothesized that if a cause and effect relationship exists it should be within the time frame and resolution of our sampling ( 12 hours ) . We therefore reverse engineered a static ARACNE mutual information network representing the neighbourhood of RND3 , RHOC , RHOBTB1 and RHOBTB2 during the tumour implantation time course ( S4A Fig ) . Remarkably , we found that the resulting network shared many properties with the grade IV network derived from the clinical study ( Fig 3 ) . First of all , the GTPases which we predicted to be grade IV specific had a markedly larger number of connections . Secondly , we discovered that RND3 was the most connected gene ( S4B Fig ) at a high level of stringency ( p < 10−8 ) . Furthermore , we could verify a high degree of functional overlap between genes connected to RND3 in the static CAM network and the grade IV network inferred from the clinical samples ( Fig 3E; S1 Table ) , making this gene an ideal candidate for further analysis . We then sought to experimentally define the transcriptional response linked to RND3 and compare this with its predicted targets in the inferred grade IV networks . We therefore used RNA interference to knock-down expression of RND3 by siRNA in U87 cells ( siRND3 ) in vitro and performed an expression profiling analysis . Differential expression analysis revealed 2 , 606 genes up-regulated and 2 , 099 genes down-regulated compared to non-silencing controls ( siControl/siRND3 ) . A Gene Ontology analysis of these genes was consistent with the predictions made from the reverse engineered networks ( Fig 3E; S1 Table ) . The full Gene Ontology analysis can be found in S5 Dataset . Functions related to tumour development including inflammatory response , regulation of apoptosis , cell migration and cell cycle were differentially regulated in directions consistent with the correlation patterns within primary glioma and U87 implantation networks . Additionally , we noticed that RND3 knockdown resulted in the transcriptional up-regulation of genes related to DNA repair , histone modification , RNA splicing and transcription factor binding , and down-regulation of genes in cell communication , amino acid phosphorylation , cell growth and response to oestrogen functional terms ( S5 Fig ) . We also noticed that the transcription of key genes involved in inflammation , proliferation , angiogenesis and extracellular matrix remodelling ( including MMP2 , HIF1A , IL1B , IL1A , IL1R1 , MMP2 , VEGFA ) , processes vital to glioblastoma development , were down-regulated by RND3 knock-down ( S6 Fig ) . RND3 and other Rho GTPases are transcriptionally regulated in high-grade glioma ( see S2 Fig ) . However , the expression of RND3 protein in different glioma grades is unknown . We tested the expression of RND3 which on the basis of the numbers of inferred network connections we predicted to have differential activity across glioma grades . Remarkably , western blot analysis of grade II , III and IV human glioma samples confirmed that protein expression correlated with the transcriptional network connectivity . RND3 protein was significantly up-regulated in grade IV gliomas compared to both grade II and grade III ( Fig 4A and 4B ) . We confirmed the difference in RND3 expression in tumour cells by immunohistochemistry analysis of grade II , III and IV gliomas ( Fig 4C ) . RND3 was found in the cytoplasm of tumor cells but also in the nucleus . We then set to characterise RND3 expression in relation to glioblastoma sub-types and genetic mutations in key disease genes . We discovered that RND3 is up-regulated up to 2-fold in GBM of the mesenchymal subtype with respect to the others ( S13A Fig ) . Consistent with this observation , other markers of mesenchymal subtype ( MET , TLR4 , RELB , TNF receptor and CD44 ) were significantly down-regulated following RND3 knock down in glioblastoma cells ( S13B Fig ) . We then tested whether individual genetic markers ( copy number variations and SNPs ) may be able to explain the expression of RND3 . Interestingly , we were able to explain up to 5% of variance in RND expression with individual copy number variations and individual gene expression measurements in the EGFR and CDKN2A genes . In addition , the expression of NF1 was also able to explain a small part of RND3 expression ( S13C Fig ) . We also used a random forest regression approach to find combinations of genetic markers that could explain RND3 expression . The resulting models were able to explain 14% ( model with only genetic mutations ) and 21% ( Model including both genetic mutations and gene expression ) of the variance in RND3 expression ( S13D and S13E Fig ) . The observed transcriptional changes suggest alterations in cell proliferation , migration and cell cycle . In order to test whether the transcriptional signature truly reflects physiological changes we first used RNA interference and a panel of in vitro assays to test proliferation , invasion , migration and cell cycle . The results obtained were fully consistent with our predictions , suggesting that knock-down of RND3 induced an anti-tumour phenotype in U87 cells , which express high levels of RND3 ( S7A Fig ) . Proliferation , migration and invasion were all significantly reduced in U87 RND3-depleted cells ( Fig 5A–5C ) . Bromodeoxyuridine ( BrdU ) labelling of RND3-depleted cells revealed significantly reduced numbers of cycling cells ( Fig 5D; siRND3 U87 18 . 8±0 . 55% cells; control 32 . 3±0 . 32% cells ) . We then performed an over-expression experiment to test whether increased levels of RND3 may have the expected effect on proliferation and migration . Firstly , the expression of RND3 was determined in several cell lines from different glioma grades . As shown in patient samples , a grade II cell line , 1321N1 , expressed low levels of RND3 ( S7A Fig ) . Two grade IV cell lines were assessed , U87 and T98G cells . T98G cells expressed a much lower amount of RND3 than U87 cells ( S7A Fig ) . A lentiviral expression of myc-RND3 was then used to over-express RND3 in both low-expressing cell ( S7B Fig ) . Both cell line morphologies were changed by myc-RND3 overexpression and this was validated by RND3-GFP over-expression ( S7C Fig ) . F-actin staining showed formation of a large lamellipodia at the front of the cells and a decrease of cell volume . This overexpression impacted both cell lines in a manner fully consistent with the RNA interference assays . Both proliferation and migration ( Fig 5E–5H ) were significantly increased in both 1321N1 and T98G cells . These results confirmed the importance of RND3 in cell aggressiveness behaviours . We then validated the hypothesis that RND3 expression is linked to apoptosis . RND3-depleted cells showed an increase in cell death evidenced by a dramatic increase in both condensed nuclei ( Fig 5I ) and cleavage of caspase 3 ( Fig 5J ) . Use of the small molecule inhibitor Y-27632 to inhibit ROCK activity did not significantly affect cell death in RND3-depleted cells suggesting that the effects of depleting RND3 were not mediated via ROCK1 , which interacts with RND3 and is essential for its canonical function controlling cytoskeleton remodelling [23] . Encouraged by the in vitro analysis we performed an in vivo implantation of U87 siRND3 cells in the chicken egg CAM . Phenotypic characterization of the resulting tumours was again fully consistent with the results of the in vitro analysis . Visual inspection showed that tumours appeared dramatically reduced in size 48 hours after implantation ( Fig 5K and 5L ) . This was accompanied by a substantial reduction in proliferating cells shown by a 62% reduction in Ki-67 expression ( Fig 5M–5O ) . Further quantification of the tumour surface area in a horizontal section confirmed an average 40% reduction in tumour expansion ( Fig 5P; siRND3 U87 1 . 8x106 pixels ±1 . 7x105; control U87 2 . 9x106 pixels ±2 . 3x105 ) . We also quantified tumour thickness after immunostaining which demonstrated a reduction of 50% ( Fig 5Q; siRND3 U87 188 . 8μm ± 23 . 2; control U87 382 . 3μm±37 . 3 ) . Consistent with the reduced expression of VEGFA mRNA in siRND3 transfected U87 cells , blood vessel density in siRND3 tumours was significantly lower compared to control tumours ( Fig 5R; siRND3 oligo B U87 10 . 6±2 . 79; control U87 23 . 8±2 . 13 ) , suggesting a disrupted angiogenesis process . Additional histological analysis of tumours revealed reduced Vimentin expression in line with the reduced tumour mass ( S8 Fig ) . Pericyte coverage of blood vessels within the tumour , detectable by Desmin staining , was increased in siRND3 tumours ( S8 Fig ) . Taken together , this supports a critical role of RND3 in tumour expansion by modulating angiogenesis , cell migration , invasion , apoptosis and cell cycle dynamics . Some of the pro-tumour effects of RND3 ( migration and invasion ) are likely to be a direct consequence of its known role in cytoskeleton remodelling [24] . However , the mechanisms behind control of cell proliferation and apoptosis are less easy to interpret , particularly considering that these are ROCK1 independent ( Fig 5I and 5J ) . Therefore , in order to further explore RND3-dependent growth and survival , we designed an unbiased , open-ended approach based on the de-novo identification of RND3 interacting proteins followed by computational analysis . Co-immunoprecipitation of RND3 flag tagged protein and subsequent mass spectrometry analysis identified 205 putative interactors of RND3 . Functional profiling of the interactors revealed primarily nuclear associated proteins involved in regulation of translation , the nuclear lumen , intracellular transport and cell division ( Table 1 ) . In order to identify a specific mechanism that may explain RND3 tumour growth we developed a novel network modularization approach designed to identify a sub-network of the human interactome enriched with proteins interacting with RND3 and at the same time highly correlated in the CAM expression profiling time course . The algorithm ( see S1 Text for details of the development and validation of the algorithm ) identified a significant sub-network ( p<10−16 ) containing 49 genes ( Fig 6A ) . This network included genes known to be involved in Rho-mediated cytoskeletal remodelling ( ROCK1 , Vimentin , Moesin , Radixin ) as well as components of NFKB signalling ( IKBKAP , NFKBIA ) , apoptosis ( Caspase 3 , PSME3 ) and , interestingly , 4 nuclear proteins involved in DNA licensing ( MCM3 , MCM4 , MCM5 , MCM7 ) as well as the important cell cycle regulator CDC2 . During the cell cycle RND3 expression has been shown to increase during G1 followed by a rapid decrease at S phase [25] . Taken together , this raised the hypothesis that RND3 may be associated to nuclear proteins and that this may be part of the mechanism regulating cell cycle . Consistent with recent reports [26] [27] and the immunohistochemistry analysis ( Fig 4C ) we demonstrated by confocal imaging of U87 cells that RND3 can localise to the nucleus ( Fig 6B ) and that it is detectable by western blot in nuclear fractions ( S11 Fig ) . Complementing evidence of the nuclear occupancy of RND3 , we performed Fluorescence recovery after photo-bleaching ( FRAP ) experiments ( S12 Fig ) , showing motile RND3-GFP is able to partially recover into bleached regions in both the nuclear and cytoplasmic compartments . We hypothesised that RND3 might be controlling the cellular localisation of the nuclear DNA licensing factors and consistent with this we verified that MCM3 could be co-immunoprecipitated with RND3 ( Fig 6C ) in U87 cells . A previous report has suggested that high levels of MCM3 protein in the nucleus may result in cell cycle arrest [28] . We observed that depletion of RND3 leads to nuclear accumulation of MCM3 ( Fig 6D–6F ) . Taken together , this suggests a novel role for RND3 in controlling cell cycle by modulating the localisation of the DNA licencing protein MCM3 . Having shown that RND3 plays a role in the development of GBM , we asked whether its expression could be influenced by genetic mutations such as copy number variation of the RND3 locus , and if both expression and CNV may be predictive of patient survival . We first approached this question using data available within the REMBRANT database of functional genomics data [29] . We focused on GBM patients and found that patients with an increased RND3 copy number showed significantly lower probability of survival compared to patients with a normal or reduced RND3 copy number ( Fig 7A ) . Using the same database we then could verify that RND3 gene expression was also predictive of survival and that RND3 CNV and expression were positively correlated ( Fig 7B ) . Next , in order to validate this initial finding , we performed survival analysis using the Cancer Genome Atlas ( TCGA ) database of GBM expression and copy number . Consistent with the survival analysis performed using REMBRANDT , both RND3 expression and CNV were predictive of survival ( S9 Fig ) . Taken together these finding suggest that genetic mutations can trigger increased expression of RND3 and that this correlates with clinical outcome . An important property that emerged from our inferred networks is the existence of two sets of Rho GTPases , which may target functionally different molecular networks and explain the more aggressive nature of high-grade Gliomas . Overall , this hypothesis is consistent with the literature . Among the seven Rho GTPases that are linked to grade II networks we could find little evidence in the literature for a mechanistic linkage with tumour transformation or progression . On the contrary , we could find evidence of their involvement in normal tissue functions such as migration of neuronal shape change during brain development ( RND2 ) [30] , regulation of actin cytoskeleton involved in vesicle trafficking ( RHOD , RHOF ) [31] [32] , sub-cellular trafficking of growth factor receptors in normal cells ( RHOB ) [33] [34] , and protein degradation ( RHOBTB3 , CHP ) [35] [36] . Among these , only RHOB has been firmly linked to high-grade glioma albeit with contrasting results . Repression of RHOB has been shown to increase motility and invasion in glioblastoma cells [37] . However , it has recently been shown to support glioblastoma tumorigenesis [38] . On the other hand , 6/7 of the Rho GTPases linked to high-grade glioma in our model has previously been shown to play a role in GBM . This includes effects on cell invasion ( RAC1 [39] , RAC2 [40] , RAC3 [39] , RHOA [41] , RHOC [42] ) , focal adhesion formation ( RHOA [43] ) , stemness of glioma precursor cells ( RAC1 [44] ) , cell proliferation ( RAC1 [45] , RND3 [46] ) and cell cycle ( RND3 [25] ) . Our work shows that down-regulation or overexpression of RND3 supports its role as a pro-tumour gene in glioblastoma controlling proliferation , migration and invasion using different glioma cell lines . Inactivation of RND3 also alters survival at both the level of cell cycle regulation and induction of apoptosis . Importantly , its expression appears to be regulated by genetic mutations such as CNV rather than as a secondary event down-stream of other cancer signalling pathways . Our work identifying RND3 as a pro-tumour gene is consistent with data from endometroid adenocarcinoma cells where RND3 is described as a p53 inducible pro-tumour gene , promoting proliferation and survival of cells following DNA damage [47] . However , its role in other types of cancer may be different . RND3 is under-expressed in prostate cancer and induces apoptosis and cell cycle arrest [48] . In addition , a previous study in glioma positioned RND3 as an anti-tumour gene , decreasing proliferation and inducing apoptosis in U87 cells [46] . However , these data were based on the overexpression of RND3 in a cell type that already expresses high levels of RND3 ( S7A Fig ) , leading to a non-physiological situation . It is also possible that RND3 has a bell-shaped activity profile , where inhibition is seen when it is absent or low and when it is highly overexpressed such as after transfection into cells . The strength of our approach , in contrast to many of these previous studies , relies in the fact that a systems biology approach that combines clinical and experimental datasets has been used in our study . This provides us with an array of concording results that clearly point to a pro-tumour role of RND3 in glioblastoma . In this article , we have shown that by combining advanced network biology approaches with the right experimental models , we are able to reveal novel regulatory circuits controlling multiple hallmarks of cancer . However , these findings should stimulate several research directions . Firstly , while we now better understand the role and clinical relevance of RND3 in human Glioma , we still do not have a mechanistic model explaining the switch between networks of normal and abnormal regulatory Rho GTPases , which we hypothesise drives the establishment of a progressively more severe cancer phenotype , presumably at the expenses of normal glial function . In this context , we have identified MCM3 , a DNA licencing factor , as a new interacting partner . This interaction may participate , in addition to its known modulatory activity on RhoA activity , to the biological effects triggered by RND3 in glioma cells . The specific role of the RND3-MCM3 interaction is at present not established . One may speculate that it is involved in nuclear-cytoplasmic shuttling . Establishing a complete mechanistic model will involve extensive experimental analysis of the role of each regulator and , eventually , the development of a mathematical model to simulate glioma progression . Ultimately , the integration of molecular , phenotypic and clinical endpoints within a computational model will provide a new set of investigative and predictive tools to support clinical decision-making . Human tumour samples are provided by Rolf Bjerkvig ( University Bergen , Norway ) . He has ethical permission to store biopsy specimens from human patients in a biobank , as well as corresponding xenografts in animals . Regional ethical approval number: 013 . 09 . The tissue collected is anonymized at the department of Neurosurgery at Haukeland University Hospital . Animal experiments: the animal experiments are conducted in the Animalerie Mutualisée , Bordeaux , France . The number of authorisation is B33-522-22 and was obtained February 28 2012 . The approval for experimentation has been obtained from the University of Bordeaux ethical committee ( approval number R-45GRETA-F1-10 ) . Differentially expressed genes between grade II ( n = 45 ) and grade IV ( n = 81 ) gliomas were identified using an existing microarray study by Sun et al [49] . Raw data were downloaded from the GEO database ( accession GSE4290 ) and normalized using the Robust Multiarray Average algorithm ( RMA ) [50] . Statistical significance was determined using a t-test followed by correction for multiple comparisons using the Benjamini-Hochberg method to estimate the false discovery rate ( FDR ) [51] . Genes with a log2 fold change greater than 1 . 5 and an FDR < 1% were selected . This transcriptional signature was compared to a transcriptional signature from a similar study comparing grade II ( n = 50 ) and grade IV ( n = 24 ) glioma ( accession GSE52009 ) for validation purposes . Both transcriptional signatures showed a very high level of similarity ( FDR < 1% ) when compared using Gene Set Enrichment Analysis [52] ( S1 Fig ) . In order to achieve this goal we first subset the human protein interactome by selecting proteins encoded by genes differentially expressed between grade II and grade IV glioma samples . The interactions between this subset of proteins were identified using the Michigan Molecular Interaction ( MiMI ) protein interactions databases [53] , which merged and integrated a number of protein interactions databases such as BioGRID [54] and HPRD [55] . This protein-protein interaction ( PPI ) network consisted of 1423 nodes and 18 , 681 edges . Hubs were identified within the network by calculating the percolation score [56] for each node , in order to identify potentially important proteins . There were 93 hub nodes with a percolation score greater than 2 standard deviations above the mean . The complete glioma stage network was constructed by selecting the first neighbours of these network hubs , resulting in 682 nodes and 2472 edges . The network was modularized using the GLaY algorithm for community detection [57] . U87 glioblastoma cells were maintained in DMEM , 10% FBS , antibiotics and L-glutamine . U87 cell pellets were deposited on the chicken egg CAM at developmental day 10 as previously described [16] . For the transcriptomics time-course , tumours were dissected from the CAM and snap frozen every 12 hours following implantation at developmental day 10 until developmental day 15 ( 10 time points ) . To assess the effect of RND3 silencing on tumour growth on the CAM , U87 cells were transfected with siRNA for RND3 or non-silencing control siRNA before implantation . Tumours were imaged every day and grown for 5 days after deposition on the CAM . The tumour was then dissected and mounted in OCT and processed for measurement and immunohistochemistry . Low and high-grade glioma transcriptional networks were generated from grade II and grade IV samples from the Sun et al dataset ( GSE4290 ) [49] and Cancer Genome Atlas datasets ( TCGA ) ( http://cancergenome . nih . gov/ ) . Normalised , batch corrected TCGA datasets were downloaded from the MD Anderson MBatch ( http://bioinformatics . mdanderson . org/tcgambatch/ ) website . Mutual information networks were inferred using the ARACNE method [58] with the Rho GTPases set as hub genes . Significant gene-gene interactions were defined using a p-value cut-off of 1x10-5 . A dynamic model of gene expression in the developing tumour during the first 5 days following U87 deposition on the CAM has been inferred from our microarray data using a custom bioinformatics pipeline based on time-delay correlation . In order to reduce the complexity of the transcriptional response we generated 14 gene clusters with distinct expression profiles using the HOPACH algorithm [59] . The clusters were categorized into rapid ( 1–12hr ) , intermediate ( 13–24hr ) or delayed-responders ( 37–48hr ) to implantation ( no clusters fell into the 25–36hr category ) according to the time point at which the expression profile was altered by 50% of the dynamic range ( S3 Fig ) . In order to increase the number of data points available for correlation analysis we first identified the median expression profile of each cluster and then applied a polynomial interpolation algorithm to generate 100 data points . The correlation matrix between the interpolated expression profiles was calculated using Spearman’s Ranking Coefficient with the addition of a time-delay procedure . Interpolated expression profiles were considered correlated with a time-delay if a shift of one expression profile by 13–24 hours ( 1–2 time points ) improved the correlation value . The time-delay procedure identified the maximum correlation value within this time-delay window . Clusters with a highly stringent correlation value of >0 . 9 were considered significantly correlated . The static transcriptional networks of genes highly correlated to expression of Rho GTPases in the CAM tumour implantation were identified using the ARACNE method [58] . A range of high stringency p value thresholds ( p < 10−6 , p < 10−7 , p < 10−8 ) were applied in order to identify the most highly connected GTPases independent of statistical confidence . To knockdown expression of RND3 U87 cells were transiently transfected with two custom siRNA oligos ( 9 pmol ) supplied by Dharmacon ( oligo A 5’-AUAGUAGAGCUCUCCAAUCA-3’ or oligo B 5’-CAAACAGAUUGGAGCAGCU-3’ ) using Lipofectamine RNAiMAX ( Invitrogen ) as described elsewhere [60] . Control , non-silencing , oligos were purchased from Qiagen . Knock-down was confirmed to last up to 96 hours by western-blot analysis ( S10 Fig ) . Protein were separated by SDS-PAGE and then transferred to nitrocellulose membranes using previously described methods [61] . Western blotting was performed using antibodies against cleaved Caspase 3 ( Cell Signalling Technology ) , RND3 ( Millipore ) , MCM3 ( Abcam ) , Lamin A/C ( Santa Cruz Biotechnology ) , α-tubulin ( Sigma ) and Flag ( Sigma M2 ) . The western blot bands were quantified using ImageJ Software . RNA was extracted from U87 cells and snap frozen tumours using RNeasy columns ( Qiagen , UK ) . RNA purity was assessed using a NanoDrop spectrophotometer and each sample had a 260/280 ratio of 1 . 8–2 . 1 . RNA was reverse transcribed and the cDNA was labelled with fluorescent Cy3 dye using the Agilent Low-input Quick Amp Kit ( Agilent , UK ) . cRNA was purified using RNeasy columns ( Qiagen , UK ) and hybridised overnight to Agilent Human 8x60k Whole Genome or Agilent Chicken V1 Whole Genome microarrays according to the manufacturer’s protocol . Microarrays were scanned using an Agilent SureScan microarray scanner and processed using Agilent Feature Extraction software . Data was normalised using quantile normalisation . In order to remove probes from the analysis that could potentially hybridise to both chicken and human cRNA we performed a separate microarray analysis . We created separate pools of RNA from CAM and U87 cells to create chicken and human reference samples . The chicken and human reference RNA was then hybridised to both human and chicken whole genome microarrays . RNA extraction , generation of fluorescently labelled cRNA , microarray hybridisation and scanning protocols were identical to those used for the implanted tumour tissue . After subtraction of the background signal , the relative contribution of the human cRNA to the total fluorescence observed for each probe on the chicken array was calculated , and vice versa . Any probe for which the cross-hybridisation of cRNA from the other species resulted in a fluorescent signal > 64 or the relative contribution to the total signal was greater than 15% was removed . 5 , 272 probes were removed from the chicken dataset and 9 , 128 probes were removed from the human dataset . Genes differentially expressed during the time course of U87 implantation on the CAM were detected using a two-step method . First , genes with a minimum fold change of 1 in log2 scale were selected , then noisy genes were removed using the BETR [62] algorithm with α = 0 . 001 . Genes differentially expressed in U87 cells in response to siRND3 treatment were detected using the SAM method [63] . Genes with a false discovery rate of 5% or lower were deemed significant . We developed a procedure derived from the work of Dittrich et al [64] , which uses the prize-collecting Steiner tree framework to identify network modules in protein-protein interaction data . However , our novel procedure is able to integrate three independent sources of data including known protein-protein interactions , differential gene expression and correlation structure . This was applied to identify networks of RND3-interactors that are enriched with co-expressed genes linked to glioma grade and therefore potentially important for tumour progression . The development of the algorithm and application to simulated data is described in S1 Text . For proliferation , migration and invasion assays U87 ( 5x104 cells/well ) cells were plated in 96-well plates . For siRNA proliferation experiments , siRND3 or siControl transfections were performed 24h before plating . 1x105 or 5x105 T98G or 1321N1 cells respectively were plated for proliferation and migration assays . Each cell line was infected with control GFP-Turbo or myc-RND3 lentiviral plasmids to a MOI of 10 . IncuCyte technology ( Essen Bioscience ) was used to generate measurements of cell proliferation , migration and invasion over time . Growth curves ( proliferation ) were built from confluence measurements acquired during round-the-clock kinetic imaging . For invasion and migration assays cells were plated in 96-well ImageLock plates ( Essen Bioscience ) . Wells were pre-coated for 6h with 50μg/ml of reduced matrigel ( BD Biosciences ) . For invasion assay 150μg/ml of reduced matrigel was added on each well . At 90–100% of confluence the plates were scratched with a 96-Well WoundMaker ( Essen Bioscience ) . Migration/invasion was detected by IncuCyte scanning one image per well , every two hours for 18 hours . The time-course of cell migration/invasion was quantified using percentage of scar recovery ( cells migrated/invaded into the wound ) at 2 h time intervals . Proliferation was measured by Bromodeoxyuridine ( BrdU ) labelling of U87 cells using the BrdU Labelling and Detection Kit I ( Roche ) according to the manufacturer’s instructions . To induce apoptosis , U87 cells were treated with 5μM ROCK inhibitor Y-27632 and/or 50 μM cisplatin ( cis-diammineplatinum ( II ) dichloride ) for 16 hours and apoptosis assessed by detection of cleaved caspase 3 by western blot or by staining cells with DAPI ( 4' , 6-diamidino-2-phenylindole , Invitrogen ) and counting the number of cells with condensed nuclei , as previously described [60] . Tumour grafts grown on chicken egg CAM were excised , fixed with 4% paraformaldehyde for 5 minutes and processed for cryo-sectioning . Ten micrometre sections were placed on Super Frost slides and immunohistochemistry was performed directly after fixation of the tissue on the slide with 4% paraformaldehyde . For immunohistochemistry , we used the following primary antibodies: anti-human Vimentin ( 1:400; Santa Cruz ) , anti-human Ki-67 ( 1:200; Santa Cruz ) and anti-Desmin ( 1:100; clone D33 from DAKO ) . Corresponding fluorescent secondary antibodies were from Molecular Probes ( 1:1 , 000 , Invitrogen ) . Chick blood vessels were visualized by using fluorescein-coupled Sambucus nigra lectin-FITC ( SNA-1 lectin , 1:100 , Vector Laboratories ) . Cell nuclei were visualized by DAPI ( Invitrogen ) . Fluorescent labelling was viewed by confocal microscopy ( Nikon ) . Quantification of staining was performed using ImageJ software . Paraffin-embedded formalin-fixed glioma tissue sections were deparaffinized and heated at 99°C for 20 min in 10 mM citrate buffer at pH 6 . 0 or incubated with proteinase K diluted in 0 . 05 M Tris—Cl , pH 7 . 5 at 37°C for 10 min . The sections were incubated with the following primary antibodies: anti-RND3 ( Abcam ab79999 , 1/100 ) . Primary antibodies were incubated overnight at 4°C . Detection was performed using a biotinylated secondary antibody ( Vector Laboratories ) amplified with Vectastain ABC Reagent ( Vector ) . Sections were developed using 3′3-diaminobenzidine ( DAB , DAKO ) , following the manufacturer’s instructions . The immunohistochemical stainings were analyzed and pictures were taken with a Nikon light microscope ( Nikon Eclipse E600 , Melville , NY , USA ) using Nikon imaging software ( Nikon NIS Elements v 4 . 11 ) . Imaging of MCM3 localisation was performed on U87 cells transfected with siRND3 or non-silencing control oligos . PFA ( 4% ) fixed cells were stained for MCM3 ( 1:200 , Abcam ) and fluorescent labelling was viewed using a Nikon Eclipse Ti system . HEK293T or U87 cells were transfected with pCMV-Flag-RhoE ( RND3 ) or pCMV-Flag as a control as previously described [23] . To discover putative RND3 binding proteins cell lysates were immunoprecipitated using anti-Flag conjugated beads ( Sigma ) . After SDS-PAGE and staining with coomassie blue the gel was cut into 7 equal fragments and subjected to in-gel trypsin digestion ( along with matching gel slices from empty vector control ) and analysed by mass spectrometry . Co-immunoprecipitation of Flag-RND3 with Mcm3 was performed on U87 cells . Cells were washed with ice-cold serum-free medium and lysed on ice in buffer containing 20 mM Tris-HCl ( pH 7 . 4 ) , 150 mM NaCl , 1 mM EGTA ( pH 8 . 0 ) , 1 mM EDTA ( pH 8 . 0 ) , 2 . 5 mM pyrophosphate , 1 mM β-glycerophosphate , 1% Triton X-100 containing freshly added protease , and phosphatase inhibitor cocktail tablets ( Roche ) . Lysates were clarified by centrifugation at 4°C , and the protein concentrations were determined by using Bio-Rad protein assay reagent ( Bio-Rad Laboratories ) . For immunoprecipitation analyses , aliquots of cellular lysates were incubated with 2 μg of monoclonal anti-Flag ( Sigma M2 ) for 1 h at 4°C . Immunocomplexes were collected on protein G-Sepharose beads ( Sigma ) . The beads were washed three times with lysis buffer then boiled for 5 min in Laemmli sample buffer . HEK293T cells were used for mass spectrometry analysis . UltiMate 3000 HPLC series ( Dionex , Sunnyvale , CA USA ) was used for peptide concentration and separation . Samples were trapped on uPrecolumn Cartridge , Acclaim PepMap 100 C18 , 5 um , 100A 300μm i . d . x 5mm ( Dionex , Sunnyvale , CA USA ) and separated in Nano SeriesTM Standard Columns 75 μm i . d . x 15 cm , packed with C18 PepMap100 , 3 μm , 100Å ( Dionex , Sunnyvale , CA USA ) . The gradient used was from 3 . 2% to 44% solvent B ( 0 . 1% formic acid in acetonitrile ) for 30 min . Peptides were eluted directly ( ~ 300 nL min-1 ) via a Triversa Nanomate nanospray source ( Advion Biosciences , NY ) into a LTQ Orbitrap Velos ETD mass spectrometer ( ThermoFisher Scientific , Germany ) . The data-dependent scanning acquisition in positive ion mode was controlled by Xcalibur 2 . 7 software . The mass spectrometer alternated between a full FT-MS scan ( m/z 380–1 , 600 ) and subsequent collision-induced dissociation ( CID ) MS/MS scans of the 7 most abundant ions . Survey scans were acquired in the Orbitrap with a resolution of 30 , 000 at m/z 400 and automatic gain control ( AGC ) 1x106 . Precursor ions were isolated and subjected to CID in the linear ion trap with AGC 1x105 . Collision activation for the experiment was performed in the linear trap using helium gas at normalized collision energy to precursor m/z of 35% and activation Q 0 . 25 . The width of the precursor isolation window was 2 m/z and only multiply-charged precursor ions were selected for MS/MS . MS/MS scans were searched against NCBI database using Mascot algorithm in Proteome Discoverer 1 . 1 software ( Thermo Fisher Scientific ) . Variable modifications were deamidation ( N and Q ) , oxidation ( M ) and phosphorylation ( S , T and Y ) . The precursor mass tolerance was 10 ppm and the MS/MS mass tolerance was 0 . 8Da . Two missed cleavage was allowed and were accepted as a real hit protein with at least two high confidence peptides . To prepare nuclear and cytoplasmic fractions trypsinised cells ( 3x106 ) were washed in ice cold PBS and incubated in 500μl RSB ( 10mM Tris pH 7 . 4 , 5mM MgCl2 , 10mM KCl ) containing 0 . 5% v/v NP40 and protease inhibitors for 5 min on ice before being centrifuged at 500g for 5 minutes . 150μl of the supernatant ( cytoplasmic fraction ) was removed and 30μl of 6x protein sample buffer added . The pelleted nuclei were resuspended in 1ml of RSB and centrifuged for 5 minutes at 500g . This was repeated and the pellet re-suspended in 150μl of protein sample buffer . To confirm clear separation of nuclear and cytoplasmic fractions lysates were analysed by western blot with Lamin A/C ( nuclear marker ) and α-tubulin ( cytoplasmic marker ) . Single cell imaging of RND3 localisation was performed using U87 cells transfected with RND3-GFP ( Addgene #23229 ) alone ( with DAPI stain ) or RND3-GFP and H2B-mcherry ( Addgene #21044 ) plasmids . Transfection involved 24h incubation with a mixture of TransIT-LT-1 transfection reagent ( Mirus BIO ) in a ratio of 3:1 with plasmid DNA . Confocal imaging was carried out on a Zeiss LSM510 microscope using either 20x Fluar 0 . 8 NA or 63x Planapochromat 1 . 4 NA objectives at a temperature of 37°C , 5% CO2 and humidified atmosphere . Z-stack images were taken at sequential 1um depth slices . Fluorescence recovery after photo-bleaching ( FRAP ) experiments involved similarly-sized regions of the cytoplasm and nuclear compartments exposed to 20 iterations of a 488nm argon-ion laser set to 100% power . FRAP and Z-stack live cell imaging was carried out at the Centre for Cell Imaging , IIB , University of Liverpool , UK . To investigate if expression of RND3 is dependent on GBM subtype [8] , genes with known signatures and genetic mutations influencing the subtype were compared with RND3 . The eight genes included: NF1 , PDGFRA , IDH1 , EGFR , TP53 , FIP1L1 ( FIP1L1 can become a fusion protein with PDGFRA ) , PTEN and CDKN2A . Normalized gene expression , scaled CNV and somatic mutation data free from batch effects was downloaded from the MD Anderson Bioinformatics Cancer Genome Atlas MBatch resource ( bioinformatics . mdanderson . org/tcgambatch ) . A regression analysis was performed using either univariate regression or the ensemble classification and regression algorithm Random Forest ( RF ) , where RND3 was used as the dependent variable and the 8 gene CNV and expression data as the predictor variables . The most influential predictor variables and the % variability of RND3 expression explained by the RF models are reported . The effects of alterations in expression and copy number variation of Rho GTPases on glioblastoma patient survival was assessed using the REMBRANDT [29] and TCGA databases . The online tools available on the REMBRANDT website ( www . caintegrator . nci . nih . gov/rembrandt ) were used to generate Kaplan-Maier survival curves and log rank p values . Similarly , the TCGA data was analysed by first calculating ranks for each patient according to the expression or standardised CNV of RND3 , and then finding the optimum partitioning of patients that maximises the significance of the Cox regression model . This was then used to generate Kaplan-Maier survival curves and log rank p values . The following datasets are deposited within Gene Expression Omnibus: Implantation of U87 cells on the chicken CAM ( GSE43674 ) , RND3 silencing in U87 cells ( GSE43812 ) .	Gliomas are aggressive brain tumours that are invasive , heterogeneous , refractory to treatment and show poor survival rates . Surgical resection and chemotherapy can increase patient survival but ultimately the disease is fatal . Multiple grades of glioma exist , with lower grades associated to better prognosis . While the majority of high-grade gliomas occur de novo , it is common that low-grade gliomas progress to the more aggressive form known as glioblastoma . In this article , we have shown that by combining advanced network biology approaches with the right experimental models , we are able to reveal novel regulatory circuits controlling multiple hallmarks of glioma . Through analysis of multiple network models representing protein-protein interaction or gene co-expression data we have revealed a switch in the role of regulatory Rho GTPases between low and high-grade gliomas . Amongst these , we show that RND3 is up-regulated in glioblastomas and is a key regulator of tumour proliferation , migration and invasion . We confirm that expression and genomic copy number of RND3 are predictive of clinical outcome , suggesting that changes in the activity of this particular Rho GTPase could be an early event associated to transformation and tumour expansion .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[]	2015	Inference of Low and High-Grade Glioma Gene Regulatory Networks Delineates the Role of Rnd3 in Establishing Multiple Hallmarks of Cancer
Coq6 is an enzyme involved in the biosynthesis of coenzyme Q , a polyisoprenylated benzoquinone lipid essential to the function of the mitochondrial respiratory chain . In the yeast Saccharomyces cerevisiae , this putative flavin-dependent monooxygenase is proposed to hydroxylate the benzene ring of coenzyme Q ( ubiquinone ) precursor at position C5 . We show here through biochemical studies that Coq6 is a flavoprotein using FAD as a cofactor . Homology models of the Coq6-FAD complex are constructed and studied through molecular dynamics and substrate docking calculations of 3-hexaprenyl-4-hydroxyphenol ( 4-HP6 ) , a bulky hydrophobic model substrate . We identify a putative access channel for Coq6 in a wild type model and propose in silico mutations positioned at its entrance capable of partially ( G248R and L382E single mutations ) or completely ( a G248R-L382E double-mutation ) blocking access to the channel for the substrate . Further in vivo assays support the computational predictions , thus explaining the decreased activities or inactivation of the mutated enzymes . This work provides the first detailed structural information of an important and highly conserved enzyme of ubiquinone biosynthesis . Ubiquinone ( coenzyme Q or Q ) is an essential redox-active lipophilic molecule widely distributed from bacteria to mammals . [1 , 2 , 3] It is a key component of the ATP-producing mitochondrial aerobic respiratory chain in which it shuttles electrons from Complexes I and II to Complex III . Q also functions as a membrane-soluble antioxidant as well as a cofactor for the proton transport function of uncoupling proteins , and recently was further shown to act as a membrane stabilizing agent improving osmotic-stress tolerance in the bacterium Escherichia coli . [4] A fully-substituted benzoquinone ring mediates Q redox properties . This aromatic head is linked to a polyprenyl tail whose length varies between living organisms and which confers lipid solubility and ensures anchorage to cell membranes . [5] In Saccharomyces cerevisiae , it is formed by six isopentenyl units ( ubiquinone is therefore designated as Q6 ) , by 8 units in E . coli ( Q8 ) and by 10 in humans ( Q10 ) . Q10 is well-known as a nutritional complement and a lipid-soluble antioxidant , [6] and is also used clinically to treat patients with coenzyme Q10 deficiencies . [7] Ubiquinone biosynthesis corresponds to a highly-conserved pathway involving a large number of genes which are involved in the modification of a six-carbon ring derived from 4-hydroxybenzoate ( 4HB ) into a quinone via attachment of the polyprenyl chain followed by hydroxylations , O-methylations and a C-methylation . [8] Despite seminal genetic and biochemical investigations on S . cerevisiae and E . coli , [9 , 10] this pathway is far from being completely understood . Indeed , only a small number of the biosynthetic proteins ( referred to as the Coq/Ubi families ) have been isolated and characterized biochemically and structurally . Q6 biosynthesis in S . cerevisae requires at least 12 such Coq proteins , namely Coq1p-Coq9p , Yah1p and Arh1p , [10 , 11] and the recently-identified Coq11p . [12] Several , if not all , of these proteins form a multi-protein complex associated with the mitochondrial inner membrane—the CoQ synthome—where the presence of every member is mandatory to properly produce Q6 . [13 , 14] Such functional and structural interdependence of yeast Coq proteins complicates the elucidation of the whole synthesis mechanism , with single Coq gene knockouts accumulating only early pathway intermediates . In this context we recently initiated a long-term project aimed at providing new insights into the structure and function of Coq proteins in yeast . Coq6p from S . cerevisiae is of particular interest . It was initially proposed to be a flavin-dependent hydroxylase , but it was unclear whether it was involved in C1 or C5 hydroxylation . [10 , 15] However , genetic and biochemical studies in our group unambiguously established that it was specifically responsible for the C5 hydroxylation during Q6 synthesis in yeast , with the unexpected assistance of Yah1 and Arh1 . [16 , 17] Primary Q10 deficiency is a rare recessive disorder associated with mitochondrial dysfunction , encephalomyopathy , ataxia , and cerebellar atrophy . [18 , 19] In addition , a recent clinical study of 13 patients identified 6 mutations on the homologous human COQ6 gene that cause steroid-resistant nephrotic syndrome with sensorineural deafness as a likely consequence of Q10 deficiency . [20] We report here the first biochemical characterization of yeast Coq6p which establishes that Coq6p is a flavoprotein using FAD as a cofactor . In the absence of any crystallographic three-dimensional structure of this enzyme , we constructed several models using homology modeling followed by molecular dynamics simulations and analysis of the FAD-Coq6p model complexes . This allowed the identification of potential channels for access of lipophilic substrates to the active site . Using substrate docking calculations , a number of key residues were proposed for mutagenesis in order to block access to the substrate . Functional in vivo assays of these site-directed mutants are in agreement with our theoretical predictions , supporting the hypothesis that one particular Coq6p channel is involved in substrate binding . This work provides a molecular basis for further studies towards a deeper understanding of structure-function relationships with respect to Coq6p mutant dysfunctions . Coq6p was overexpressed in the presence of GroES and GroEL chaperones , and purified as an N-terminal MBP-tagged protein ( named Coq6p-MBP ) with a molecular mass of 96 kDa ( Fig 1A ) . Gel filtration experiments indicated that the protein formed tetramers and high molecular weight oligomers that could be further separated . The yellow color of the tetrameric Coq6p-MBP and its UV-visible spectrum with characteristic absorption bands at 370 nm and 446 nm ( Fig 1B ) show that it contains a flavin cofactor . The flavin extracted from Coq6p-MBP was identified as flavin adenine dinucleotide ( FAD ) based on its co-elution on HPLC with commercial standard ( S2 Fig ) . FAD occupancy was estimated at 85% per monomer in the tetrameric Coq6p-MBP by UV-vis spectroscopic analysis . FAD occupancy was estimated at 85% per monomer in the tetrameric Coq6p-MBP by UV-vis spectroscopic analysis and Bradford assay . Because the large size of the MBP tag ( 45 kDa ) may impede protein function , removal of this tag from Coq6p-MBP was attempted using enzymatic cleavage by Factor Xa . However it resulted in protein aggregation of cleaved Coq6p . The reduction of the FAD cofactor was attempted with different reductants ( NADH and NADPH ) in anaerobic conditions and monitored by UV-visible spectroscopy . The FAD cofactor of Coq6p-MBP could be reduced neither by NADH nor NADPH . No NADH oxidase activity of Coq6p-MBP was detected in aerobic conditions . We examined the ability of Coq6p-MBP to function as a flavin monooxygenase that utilizes dioxygen to catalyze the hydroxylation of various water-soluble aromatic compounds as models of the two potential substrates , 3-hexaprenyl-4-hydroxyphenol ( 4HP6 ) and 3-hexaprenyl-4-hydroxybenzoate ( 4-HB6 ) . However , none of the tested molecules ( 4-hydroxybenzoic acid , 3-methyl 4-hydroxybenzoic acid , 2-methyl 4-hydroxyphenol , 2-methylphenol ) were found to be substrates . Several hypotheses can explain this lack of activity: the presence of the MBP tag can affect the enzymatic activity of Coq6p-MBP; potential protein partners may be required for Coq6p to be enzymatically active; the positioning of Coq6p in the lipid membrane may also affect its biological activity , folding and binding with its substrate . [63] The modeling strategy is divided into four successive sections , whose results are presented and analyzed separately . First , we created homology models of Coq6p . We used two distinct approaches: one where template selection , template-target alignment , and model refinement steps are done with maximal user intervention using MODELLER[32–34] and the other using two highly ranked servers , ROBETTA and I-TASSER , [64 , 65] that fully automate these processes . At this stage , the FAD co-factor was introduced after model construction , as detailed in Methods . Second , the three Coq6p-FAD models were subjected to molecular dynamics simulations . The resulting trajectories were analyzed on the basis of substrate accessibility to the catalytic site by preliminary docking calculations on a model substrate , 4-HP6 . At the end of this process , one of the models was selected for further analysis . Third , detailed substrate docking in the wild type enzyme was performed . Representative enzyme conformations were extracted using a receptor-based scoring function designed to recognize catalytic site conformations likely to be compatible with catalysis . Finally , mutations were rationally designed on the basis of in silico results and experimentally tested by in vivo assays . The search for proteins of known structures presenting sequence homology with Coq6p consistently identified FAD-dependent monooxygenases . Such proteins constitute appropriate templates for homology modeling and their crystal structures are identified as follows in the PDB database: 4N9X , 4K22 , [24] 2X3N , [27] and 1PBE . [40] Sequence alignment of Coq6p with these templates reveals low sequence identities ( 28 . 32% , 27 . 30% , 20 . 30% , 18 . 51% , respectively , see S3 Fig ) . In spite of regions with unsolved coordinates in several of the templates , these templates exhibit similar overall tertiary structures with a Rossmann-like β/α/β fold forming an FAD-binding domain and a large beta sheet forming a substrate binding domain . Structural comparison shows that the core secondary structure elements are overlapping ( S4 Fig ) . If one assumes that such structural features also translate to Coq6p and that they are correlated to Coq6p function , then an approach based on homology modeling is relevant . In order to resolve alignment ambiguities among these low homology sequences , the pairwise alignments of Coq6p with 4K22 , [24] 4N9X , 2X3N , [27] and 1PBE[40] were manually curated by cross-referencing them against a multiple sequence alignment of 119 Coq6p homologues ( S5 Fig ) as well as by considering the 3D structure of each template . The resulting target-template pairwise alignments were then collated to build a structurally annotated alignment for Coq6p shown in Fig 2 . Compared to the templates , it is apparent that Coq6p contains a long additional sequence designated the Coq6 family insert ( 293–343 ) , with the insert of Coq6p S . cerevisiae being among the longest in the Coq6 family ( S5 Fig ) . Target-template alignments for each of the three modeling protocols are presented in S6 Fig . ROBETTA automatically chose a single template , 1PBE . I-TASSER proposed models from 4N9X followed by 2X3N; we rather chose the 2X3N-derived model on the basis that it is co-crystallized with FAD , ensuring a reliable initial conformation of the FAD binding domain , unlike 4N9X . Based on structure-function knowledge , the alignment passed to MODELLER combined 3 templates: 2X3N for most of Coq6p N-terminal region ( residues 1–426 ) , 4N9X for the C-terminal region not present in 2X3N ( residues 427–479 ) , and 4K224 for a small 6-residue segment of the beta-sheet domain ( residues 352–357 ) . In this manual homology modeling step , we superposed both helices 12 of 2X3N and 4N9X to determine the relative orientation between the two templates , selecting P426 as the final residue of the 2X3N template and S427 as the starting residue of the 4N9X template . Regarding the small template sequence of 4K22 , we similarly superposed the beta-strand of 4K22 and 2X3N templates to allow the modeling of this region in Coq6p_MODELLER . This small region of 4K22 was selected as the template one ( rather than 2X3N ) in order to allow the FAD to undergo in and out movements during subsequent MD runs , which do not occur using 2X3N for this region . Fig 3 highlights structural differences between the three homology models , Coq6p_I-TASSER , Coq6p_ROBETTA and the Coq6p_MODELLER model ( based on a manually curated alignment of rationally selected templates ) , prior to MD . It is interesting that in all monooxygenase structures from the PDB that are bound to FAD , the FAD binding pocket is always constituted of a small ribityl binding loop ( the GDAxH loop ) structured as a α-helix , suggesting that a catalytically competent Coq6p structure must also have its ribityl binding-loop in a α-helix . All three models are derived from templates with a GDAxH loop compatible with FAD binding ( see S1 Text ) , with the si loop ( for which the templates are incompletely resolved ) consistently reconstructed as a helix . Two major structural elements discriminate the models: the C-terminus and the 51-residue Coq6p-family insert . The former strongly depends on template choice , with the ROBETTA and the MODELLER models using 1PBE and 4N9X respectively , while the I-TASSER model does not rely on any template in this region , since it is not present in 2X3N . The Coq6 family insert is not present in any of the templates; therefore its final conformation in the models solely results from methodological differences for modeling ab initio regions . It is however consistently predicted to have a helix-turn-helix secondary structure . The insert is systematically localized to the exterior face of the beta-sheet domain , with no residues proximal to the active site . Since all flavin-dependent monooxygenases with known structure show catalytic activity without a similar structural element , the Coq6p insert is unlikely to be required for catalysis . However , its conservation among the Coq6 family ( S5 Fig ) and its high proportion of polar and charged residues suggest that it may be essential for integration of this enzyme into the obligate CoQ biosynthesis complex through ionic protein-protein interactions . Overall , the three homology models provide us with three significantly different starting models for further MD runs . Taking the Coq6p_MODELLER model as a reference , the Coq6p_ROBETTA and Coq6p_ITASSER models exhibit RMSDs of 6 , 53 Å and 6 , 39 Å respectively , which reflect the different structures of the insertion and of the C-ter region in the three models ( see Fig 3 ) . Additional RMSDs are given in S2 Table . All three Coq6p-FAD homology models were subjected to the same molecular dynamics ( MD ) protocol . The aim of this process is first to investigate the structural stability of the models and perform conformational sampling of each model , which is important for subsequent substrate docking . All models appeared structurally-stable through 50 ns MD runs , all having converged to a conformational plateau in which RMSD fluctuations are as expected for proteins of this size simulated in physiological conditions ( S7 Fig ) , all having stable secondary structures ( S8–S10 Figs ) . This suggests that approximations and assumptions made during the modeling process ( related to the ab initio modeling procedure for sections with no templates , or the manual addition of FAD post Coq6p model generation ) did not prevent any of the models to be close to a low-energy state . Conformations of the three Coq6p models after MD are illustrated in S11 Fig . It is noted that the regions identified as most important for Coq6p function , such as the FAD-binding domain helices and the β−sheet domain , appear more stable structurally than the rest of the protein and especially the Coq6p insert and C-terminal region ( as observed from their RMSDs in S7 Fig ) . The FAD binding mode as exemplified in Fig 4 was stable in the three homology models over the course of 50 ns MD calculations . As strongly suggested by the position of FAD in both 1PBE and Coq6p models , the most-plausible location of the Coq6p catalytic site is a buried volume immediately facing the isoalloxazine ring , about 14 Å away from the protein surface . Focusing on this specific region ( Fig 5 ) , key residues were identified in both the 1PBE crystal structure , in which they contact the substrate 4-HB ( S212 , R214 , P293 and T294 , Fig 5A ) , and in Coq6p ( T261 , M255 , P381 and L382 , Fig 5B ) . These residues were used to build the receptor-based scoring-function with the exception of M255 to identify catalytically plausible poses . Three of these residues ( T261 , P381 and L382 ) are very conserved among Coq6 eukaryotic sequences ( S5 Fig ) . They provide the same arrangement of 1 H-bond donor ( PHBH: S212/Oγ; Coq6p: T261/Oγ ) and 2 H-bond acceptors ( P293 , T294/O; P381-L382/O ) . Coq6p M255 cannot provide similar interactions to pHBH R214 ( H-bonding to the 4-HB carboxyl group ) although it is located at the same position , and hence was not included in the scoring-function . The 4-HB position in pHBH also gives us a diagnostic distance ( 4 . 32 Å ) between the C4X FAD-atom which bears the peroxo group of the reactive FAD-OOH intermediate and the substrate target carbon to be hydroxylated . [40] The FAD C4X atom position in Coq6p relative to T261 , P381 and L382 is also well predicted when compared to that in 1PBE ( Fig 5B ) . As preliminary investigations , the active site in pHBH was investigated using MD simulations starting from the crystal structure of the pHBH-FAD complex from 1PBE with the substrate removed . The arrangement of the identified key residues selected above could be recurrently found through the 50 ns MD , using the receptor-based scoring-function . When re-docking 4-HB in the pHBH-FAD substrate free complex , [66] the calculations reproduced the crystal pose of 4-HB in 1PBE . Translated to Coq6p , these calculations imply that enzyme conformations that are able to bind the substrate directly can be found by exploring substrate free conformational space through MD . The three Coq6p models were then analyzed for the presence of cavities for substrate binding and access to the catalytic site by computation of accessible void regions . Three distinct tunnels connecting the Coq6p surface to the putative catalytic site were identified on the pre-MD homology models ( Fig 6 ) . Two of these tunnels ( 1 and 2 , in purple and blue , respectively ) exit the enzyme from the re face , the other ( 3 , in red ) from the si face of the isoalloxazine ring . Assuming that substrate binding does not involve significant induced fit effects , it should traverse at least one of the three channels . For each of these , a pair of residues corresponding to the bottlenecks was identified , and the corresponding distance monitored through the MD simulations ( S12 Fig ) . MD snapshots were extracted according to the maximal values of this bottleneck distance and served as targets for automated substrate docking , using 3-hexaprenyl-4-hydroxyphenol ( 4-HP6 ) as a Q6 model substrate . The choice of the model substrate as 4-HP6 was motivated as follows . In earlier studies , 3-hexaprenyl-4-hydroxybenzoate ( 4- HB6 ) has been proposed to be the substrate of Coq6p on the basis that the decarboxylation reaction of the eukaryotic Q biosynthesis pathway could occur after the C5-hydroxylation step catalyzed by Coq6p . [67] However , we have recently reported that cells lacking an active Coq6p enzyme accumulate 3-hexaprenyl-4-hydroxyphenol ( 4-HP6 ) , [17] showing that the C1-decarboxylation and C1 hydroxylation reactions can occur prior to the C5-hydroxylation catalyzed by Coq6p ( S13 Fig ) . Yet clear experimental evidence are missing to strongly support either 4-HB6 or 4-HP6 as substrate of Coq6p since no in vitro assay is available and the decarboxylase enzyme is not identified . [13] Moreover , in pHBH , [40] the carboxyl group of the substrate 4-hydroxybenzoate is hydrogen bonded with R214 in the active site , but Coq6p lacks such a residue ( 5B Fig ) , suggesting that a benzoate group may not be present on the aromatic head of the Coq6p substrate . Overall , these features prompted us towards 4-HP6 as a model substrate rather than 4-HB6 , at variance with the known substrate of PHBH . Still , docking studies were performed also with 4-HB6 as a model substrate . While changing functional groups of the aromatic head from a hydroxyl to a carboxyl did not change the further conclusions in this study , the ensemble of 4-HB6 substrate docked poses is more diverse ( S14 Fig ) while in both cases the same most stable binding mode is found . This difference in calculation convergence is most probably related to the lack of recurrent stabilizing interactions between the carboxyl group of 4-HB6 and the residues of the Coq6p active site , which we assign to the absence of a PHBH R214 homolog that would facilitate the recognition of a carboxylate substrate . We therefore present here our investigation of substrate docking with 4-HP6 , while our conclusions regarding the substrate access channel are similar with both models . It was then found that only one specific tunnel in one model ( the re face tunnel 1 in the Coq6p_MODELLER model , Fig 6C , purple volume ) produced plausible binding modes of 4-HP6 , i . e . with the substrate aromatic head placed in the cavity of the putative catalytic site ( Fig 7 ) . Using the receptor-based scoring function , MD conformations from this model were further selected so as to resemble the 1PBE enzyme substrate-bound conformation as much as possible ( see Methods ) . In a second round of calculations , the substrate was docked into these Coq6p conformations . This resulted in a systematic convergence to a single binding mode with the aromatic head close to the FAD cofactor and the hydroxyls oriented as to bind the two following residues ( Fig 7 ) : the H-bond donor , T261 , and the H-bond acceptor , P381 , a highly conserved residue among Coq6 enzymes ( S5 Fig ) . The distance between the substrate C5 atom and FAD C4X atom is consistently is found at 4 . 7 Å ( Fig 7 ) , very similar to the corresponding 4 . 32 Å FAD-4-HB distance in the pHBH-FAD-4-HB complex from 1PBE . [40] The isoprenyl chain can adopt a variety of conformations traversing the channel and reaching the surface , and exhibits a number of contacts with hybrophobic residues such as P249 , L253 , L382 , L438 and F439 . Overall , these sequential series of calculations allowed us to identify the re face tunnel 1 as the putative substrate access channel in Coq6p . Evolutionary conservation of residues lining the tunnel system of Coq6p was determined by submitting the Coq6p sequence to the ConSurf server and mapping the residue conservation scores onto the Coq6p homology model ( Fig 8A ) . Interestingly , it identified the re face tunnel 1 leading to the active site as the most conserved tunnel . Furthermore , the walls of tunnel 1 are lined almost exclusively with hydrophobic and uncharged side chains , F244 , P249 , S265 , P381 , L382 and F439 , ( Fig 8B ) , which is appropriate for the channeling of the isoprenyl chain . For comparison , the 4N9X experimental structure ( an UbiI homologue ) was also submitted to establish a map of residue conservation , which pointed to a similarly conserved re face tunnel . All together , these results suggest that the hydrophobic character of this tunnel is conserved . Several deleterious mutations in human Coq6 , hCoq6 , were recently reported , including R162X , G255R , A353D and W447X , [20] D208H[68] and Y412C . [59] Aligning the hCoq6 sequence with our manually curated alignment , corresponding Coq6p residues were identified and mapped onto the homology model ( S15 Fig ) . Interestingly , the hCoq6 G255R mutation translates to a G248R mutation in Coq6p , and is positioned at the entrance of the putative re face substrate access tunnel 1 identified above , with the arginine sidechain pointing towards the channel lumen ( Fig 9C ) . MD trajectories of the Coq6p-G248R mutant showed a significant reduction of the channel diameter at the choke point ( Fig 9A , red line ) with respect to the wild type ( Fig 9A , blue line ) . The R248 sidechain samples blocking and non-blocking conformations , consistent with the reported reduced activity of the corresponding human mutated enzyme . [20 , 59] We also identified L382 ( Fig 9B ) , a highly conserved residue which faces G248 across the channel , as a key site to introduce another mutation to block this channel . The L382 sidechain is oriented towards the lumen of the access channel and could be mutated to a larger sidechain . Its solvent exposure also suggests that mutation to a polar or charged residue may be possible . We reasoned that a mutation to glutamate , L382E , would be tolerated by the structure while allowing the formation of a salt bridge with the G248R sidechain ( Fig 9E ) . MD calculations of the single Coq6p-L382E mutant ( Fig 9A , green line , and Fig 9D ) showed a significant decrease of the channel diameter when compared to the wild type . However , traversable conformations were occasionally observed during MD runs , allowing us to dock the model substrate . Turning to the Coq6p-G248R-L382E double mutant , MD calculations predicted the formation of a stable salt bridge spanning the channel lumen and causing a complete and persistent blocking of the channel over the 20 ns trajectory ( Fig 9A , purple line ) . Overall , these results show that single mutations partially block the channel , whereas the double mutation causes a complete blockage . To corroborate our in silico predictions , site-directed mutagenesis was performed to test the role of the G248 and L382 residues by studying the single mutants G248R and L382E , as well as the double mutant G248R-L382E . The effect of these different mutations on Coq6p activity was tested in vivo . Contrary to growth on the fermentable glucose medium , growth on the respiratory medium containing lactate-glycerol ( LG ) requires a functional respiratory chain and is thus dependent on ubiquinone . Accordingly , the Δcoq6 strain is unable to grow on LG medium unless it is complemented with a plasmid expressing Coq6p ( Fig 10A ) . The growth provided by Coq6p-L382E was comparable to that of Coq6p , but Coq6p-G248R was slightly compromised ( Fig 10A ) . The Δcoq6 strain expressing Coq6p-G248R-L382E was unable to grow on LG medium showing that the L382E and G248R mutations have an additive effect . Addition of vanillic acid ( VA ) to the growth medium is able to restore Q6 biosynthesis in Coq6-deficient cells by bypassing the deficient C5-hydroxylation reaction , provided that the Q biosynthetic complex is stable which requires a stable Coq6p polypeptide . [17] Non-empty vectors expressing any of the three Coq6p mutant proteins allowed growth on LG medium + VA ( Fig 10A ) , supporting that the Coq6p-G248R-L382E is stable . HPLC-ECD analysis of the Q content correlated with the growth phenotypes and showed that Q was undetectable in cells expressing Coq6p-G248R-L382E ( Fig 10B ) . These cells accumulated 4-HP6 which is characteristic of a C5-hydroxylation defect but forms only when the CoQ synthome is assembled , supporting again that Coq6p-G248R-L382E accumulates in vivo as a stable polypeptide . The Q levels measured in cells expressing Coq6p-G248R and Coq6p-L382E were intermediate ( Fig 10B and 10C ) . Altogether , these results show that the G248R and L382E mutations decrease Coq6p activity to some extent while the combination of both mutations completely inactivates the protein without affecting its stability . These data are consistent with the theoretical prediction of the substrate channel being blocked by the proposed interaction between R248 and E382 residues . Biochemical results provide evidence that Coq6p is a flavoprotein , using FAD as a co-factor . A sequential computational strategy was then adopted comprising the construction of homology models of Coq6p , followed by MD calculations of the FAD-Coq6p complexes . We used these models to get insight into the catalytic site structure and a tunnel system in Coq6p . A specific tunnel was identified and experimentally substantiated through in vivo studies of selected mutants . One important limitation of homology modeling is the availability of templates with sufficiently high sequence identities . While operating in a regime of low sequence identity , the sequence-based searched templates used for Coq6p homology models ( i . e . 4N9X , 4K22 , [24] 2X3N , [27] and 1PBE , [40] ) share the same Rossmann-fold of monooxygenase structures , belonging to the class A of flavoproteins . [69 , 70]Importantly , all three FAD-Coq6p homology models display well-formed α-helix loops as part of the highly conserved GDAxH motif ( highlighted in orange in S4 Fig ) , directly inherited from their respective template in that region . Through MD calculations , FAD in complex with Coq6p recapitulates a number of canonical enzyme-cofactor contacts ( Fig 4 ) . When compared to 2X3N , [27] a FAD-cocrystallized structure , the ribityl chain is H-bonded to the side-chain of D374 , ( D292 in 2X3N ) and the ribose is H-bonded to the side-chain of D61 , ( Q35 in 2X3N ) . Also , the negatively charged pyrophosphate group is electrostatically complemented by the α1 helix dipole of the β/α/β fold at the N-terminus as in 2X3N . Turning to structure-function relationships , studies of pHBH’s catalytic cycle indicate that the substrate enters the active site through a channel , the active site being shielded from solvent . [71] We reasoned that the physical pathway of the aromatic head entry and isoprenyl tail binding is likely to be also a tunnel traversing the Coq6p protein . Calculations on the resting Coq6p models ( prior MD ) show that they share a general multiple channel system , with three distinct tunnels converging to the highly conserved active site ( Fig 6 ) , a buried volume immediately facing FAD . The traversability of Coq6p channel system by the substrate was assessed computationally . Here , we postulate a conformational pre-selection mechanism ( as opposed to induced-fit ) whereby conformations of Coq6p compatible with substrate binding might be accessible through MD calculations on the substrate-free enzyme . [72] The key outcome of these calculations is that re face tunnel 1 may be the Coq6p access channel for its substrate ( Figs 7–9 ) . Interestingly , a recently reported crystal structure of a PHBH-family flavoprotein , namely the 3-Hydroxybenzoate 6-Hydroxylase ( 4BJY ) , [73] features a bound lipid positioned in the re face channel 2 identified in all Coq6p models . The comparison of both enzymes ( S16 Fig ) highlights how the orientation of the helix 12 determines the access of the lipidic chain either through the C-terminus “below” helix 12 ( 4BJY ) or through the channel “above” helix 12 ( Coq6p ) . Turning to Coq6p , the re face tunnel 2 was discarded since no plausible poses for 4HP6 was found in the present work . Still , the aromatic head of 4-HP6 docked in Coq6p ( tunnel 1 ) occupies the same region than the substrate of 4BJY ( as inferred from 4BK1 , a mutant co-crystallized with its substrate and lipid ) , i . e . in a catalytically plausible pose in the vicinity of the FAD isoalloxazine ring , highlighting the similar function of these two proteins belonging to the same family . Importantly , the re face tunnel 1 turned out to be the most conserved one and furthermore it is lined with hydrophobic residues appropriate for contact with the isoprenyl tail of the substrate . This drove us to perform in silico mutagenesis at its bottleneck and propose single and double-mutations to potentially hinder substrate binding . The experimental in vivo findings , i . e . the reduced Q6 biosynthesis of the single and double mutants , provide a posteriori validation of our in silico mutagenesis , and overall of the computational strategy including our conformational pre-selection hypothesis . This work provides the first detailed structural information of an important and highly conserved enzyme of ubiquinone biosynthesis in the absence of crystallographic data . First , it demonstrates that Coq6p is a FAD enzyme and thus belongs to the FAD-dependent mono-oxygenase family . Second , in order to accommodate a bulky hydrophobic substrate , it has evolved a channel and a cavity to specifically direct the substrate towards the catalytic site . Future work should address the catalytic activity of Coq6p since so far we and others failed to find the proper conditions for detecting in vitro enzyme activity . This is not surprising considering the difficulty related to the synthesis of the substrate ( highly hydrophobic ) , to the complexity of the redox system associated to Coq6p activity [16] and the likely requirement for protein partners within CoQ synthome . [12] The availability of a structural model for Coq6p makes it possible to consider further computational approaches regarding its integration in more complex interactions with other proteins and with the membrane . As exemplified here , in silico mutagenesis studies can be coupled to in vivo confirmation to explore functional hypothesis .	Coenzyme Q is an essential redox active lipid present in most living organisms and in tissues of multicellular eukaryotes , which acts as a key component in the mitochondrial respiratory chain . It consists of an aromatic head and a long hydrophobic isoprenyl chain and is synthesized by several enzymes within an obligate protein complex located at the inner mitochondrial membrane . Mutations in several human Q biosynthesis genes are responsible for a range of diseases that may be treated by oral supplementation of Q . Previous studies have established that Coq6 , a predicted flavin-dependent monooxygenase , is responsible for the C5-hydroxylation reaction in the Q biosynthetis pathway , assisted by the ferredoxin Yah1 and ferredoxin reductase Arh1 . We establish here that in Saccharomyces cerevisiae Coq6p uses FAD as a cofactor . We then use a computational approach to construct homology models of Coq6p , followed by MD calculations of the FAD-Coq6p complexes . These models are further used to get insight into the catalytic site structure and a general tunnel system in Coq6p . A specific tunnel is identified and mutations are proposed in silico in order to prohibit substrate access . The results of the in vivo study of the designed mutants substantiate our computational predictions .	[ "Abstract", "Introduction", "Results", "Discussion" ]	[]	2016	Coenzyme Q Biosynthesis: Evidence for a Substrate Access Channel in the FAD-Dependent Monooxygenase Coq6
Previous theoretical studies of animal and human behavioral learning have focused on the dichotomy of the value-based strategy using action value functions to predict rewards and the model-based strategy using internal models to predict environmental states . However , animals and humans often take simple procedural behaviors , such as the “win-stay , lose-switch” strategy without explicit prediction of rewards or states . Here we consider another strategy , the finite state-based strategy , in which a subject selects an action depending on its discrete internal state and updates the state depending on the action chosen and the reward outcome . By analyzing choice behavior of rats in a free-choice task , we found that the finite state-based strategy fitted their behavioral choices more accurately than value-based and model-based strategies did . When fitted models were run autonomously with the same task , only the finite state-based strategy could reproduce the key feature of choice sequences . Analyses of neural activity recorded from the dorsolateral striatum ( DLS ) , the dorsomedial striatum ( DMS ) , and the ventral striatum ( VS ) identified significant fractions of neurons in all three subareas for which activities were correlated with individual states of the finite state-based strategy . The signal of internal states at the time of choice was found in DMS , and for clusters of states was found in VS . In addition , action values and state values of the value-based strategy were encoded in DMS and VS , respectively . These results suggest that both the value-based strategy and the finite state-based strategy are implemented in the striatum . Theoretical studies of decision-making have focused on the dichotomy of whether an environmental model is utilized , i . e . model-free or model-based strategies [1 , 2] . In a typical model-free strategy , called a value-based strategy , the goodness of each action candidate is memorized and learned directly from experienced sequences of state , action , and reward in the form of an action value function [2–5] . The hypothesis that such value-based strategies are implemented in the cortico-basal ganglia circuit[1 , 6] is supported by a growing number of reports of action-value coding neuronal activities in the striatum , the input site of the basal ganglia , in rats [5 , 7 , 8] , monkeys [4 , 9–11] , and humans [12] . By contrast , in a model-based strategy , the goodness of each action candidate is evaluated indirectly using an internal model of environmental state transitions . Recent fMRI studies found BOLD signals correlated with estimated states and state prediction errors in the prefrontal cortex [13–15] . While the value-based and model-based strategies have been helpful in dissecting the process of decision-making , the validity of such concepts and consequent predictions need to be assessed in light of actual animal and human behaviors . For example , animals often utilize a simple “win-stay , lose-switch” ( WSLS ) strategy , in which the same action is repeated if it is rewarded and switched if it is not rewarded [5 , 16] . This strategy does not conform to either the value-based or the model-based strategy . Theoretical studies have shown that optimal behavior under uncertain state observation can be represented as a finite state machine in which an action is selected depending on the agent’s discrete internal state , and the state is updated based on sensory observation and reward feedback [17] . The WSLS strategy is simply realized as a finite state machine with two states . Here we consider the validity of the finite state-based strategy as another class of model-free strategy along with the value-based strategy in modeling animal choice behaviors . We reanalyze a part of the data we published previously [18] , and we show that the finite state strategy fits the choice behavior of rats in a free-choice task more accurately than the value-based strategy and the model-based strategy . We further reanalyze the firing of phasically active neurons ( PANs; putative medial spiny neurons ) recorded from the dorsolateral striatum ( DLS ) , dorsomedial striatum ( DMS ) , and the ventral striatum ( VS ) during the task . We show that the individual states of the finite state strategy are encoded in DMS at the time of choice and that clusters of states are encoded in VS . Furthermore , the action values used in the value-based strategy are also encoded in DMS . These results suggest that both the value-based strategy and the finite state strategy are implemented in the striatum . Next we explore more detailed descriptions of choice behavior using computational models that can predict rat choices based upon past experiences . Along with the Markov models and the value-based strategy tested in our previous study [18] , we tested the model-based strategy and the finite state strategy . While the likelihood of a model fitted to given choice sequences is a useful criterion for comparing models , it is also important to check how the model performs when it runs autonomously . One direct way to check this performance is to compare statistical features of the behavioral sequences produced by the model in a simulation with performance of rats in the actual task ( see Materials and Methods ) . We simulated the Q , FQ , DFQ , and ESE models with constant parameters and the FSA models with 4 , 6 , and 8 states . We excluded the models with variable parameters because the random walk assumption was effective for fitting a model to a given choice sequence , but not for the generation of choice sequence in a free run . We took the number of trials required to reach the block-change criterion ( 80% or more optimal choices in the last 20 trials ) as a measure of the flexibility of adaptation ( Fig 5A–5D ) and the probability that the same action was selected after the rewarded or non-rewarded trial , P ( a ( t+1 ) = a ( t ) \| r ( t ) = 1 ) and P ( a ( t+1 ) = a ( t ) \| r ( t ) = 0 ) , respectively , as a measures of the robustness of the action ( Fig 5E and 5F ) . Statistics were calculated separately for blocks with higher reward probability settings [ ( 90 , 50% ) and ( 50 , 90% ) ] and lower reward probability settings [ ( 50 , 10% ) and ( 10 , 50% ) ] . We tested the hypothesis that data from rats could be generated from each model using the mean of the six statistics ( Fig 5C–5F ) . Only the FSA model with 8 states was not rejected by any statistical test ( the level of the confidence interval for each statistic was set to ( 100–5/6 ) % , so that the chance of at least one false rejection is 5%; Bonferroni Method ) . This result shows that only the FSA model with 8 states sufficiently reproduces the behavior observed in the rats , although it does not exclude the possibility that there are other models better than the FSA model with 8 states . Previous studies have shown that striatal neurons code not only observable behavioral variables , such as action and reward [5 , 7 , 10 , 20–22] , but also hidden variables estimated from behavior using computational models , such as action values [4 , 5 , 7 , 12 , 23] . In our previous study [18] , regression analysis revealed that action values , which were estimated from behavioral data based on the FQ-learning with variable parameters , were coded most strongly in DMS during action execution . In this analysis , we re-analyzed the same neuronal data to examine whether a new class of hidden variables , namely , states and state clusters of the FSA with 8 states , were also coded . However , if we use a regression model that employs only states and clusters as regressors , it would lead to Type I errors ( false positives ) . For instance , the estimate of state 1 is strongly correlated with the left action choice in the same trial , detecting action-coding neurons as state-coding neurons ( Fig 4C ) . To avoid this problem , we first considered a full model including all possible variables ( 30 variables ) that might be coded by striatal neurons ( Poisson regression model , see Materials and Methods ) . Then , we extracted only the important variables to explain the output using lasso regularization [24] ( see Materials and Methods ) . The full model we used was: logμ ( t ) =β0+βbb ( t ) +βaa ( t ) +βrr ( t ) +βa'a ( t−1 ) +βr'r ( t−1 ) +βQLQL ( t ) +βQRQR ( t ) +βQCQc ( t ) +βVV ( t ) +βPLQPL:Q ( t ) +βx1x1 ( t ) +βx2x2 ( t ) +⋯+βx8x8 ( t ) +βx1'x1 ( t+1 ) +βx2'x2 ( t+1 ) +⋯+βx8'x8 ( t+1 ) +βCLCL ( t ) +βCRCR ( t ) +βCWCWSLS ( t ) +βPL:FSACL:FSA ( t ) ( 1 ) where μ ( t ) is the expected number of spikes at trial t in a certain time bin and βi is the regression coefficient for each explanatory variable ( regressor ) . b ( t ) is the monotonically increasing factor , namely , b ( t ) = t , which is inserted to capture the task event-independent monotonic increases or decreases in firing pattern . The remaining regressors are classified into three types: We applied lasso to this full model , which can identify minimally important regressors among many and redundant regressors ( see Materials and Methods ) . When lasso identified certain regressors to explain the activity of a certain neuron , we interpreted this to mean that “the neuron coded the regressors . ” A single striatum neuron tended to code multiple variables in different time bins as shown in Fig 6 . Lasso detected significant populations of neurons that coded observable information ( I ) and estimated information based on the FQ-learning ( II ) , similar to our previous analysis [18] . In addition , this analysis detected neurons that coded states of the FSA model ( III ) . Fig 6A–6C show an example of DMS neurons in which firing rate was significantly correlated with the posterior probability of states of the FSA model . During action selection , firing rate was best explained by the regression model including not only the action , but also x5 ( t ) , in which the FSA model doubts the current belief that left hole is better and wants to choose the right hole ( see Fig 4C ) . Fig 6D–6F show an example of DMS neurons in which firing rate was significantly correlated with the posterior probability of a transited state of the FSA model . The firing rate during the rat’s entry to the left or right hole ( note that the reward or non-reward tone was presented at the onset of the hole poke ) was best explained by the regression model , including not only the action , reward , and x7 ( t ) , but also x7 ( t+1 ) . Here the FSA model believes the right hole is better following an exploratory choice ( see Fig 4C ) . Fig 6G–6I show an example of VS neurons coding the rat’s sub-strategy ( cluster ) . There was a significant , positive correlation between neuronal firing rate during action selection and the posterior probability of the win-stay , lose-switch cluster estimated by the FSA model with 8 states . In our previous study , we detected action-value coding neurons and state-value coding neurons by linear regression analysis , in which action values estimated by the FQ-learning were used as regressors . In this study , we used an augmented regression model ( Poisson regression model ) , including not only variables of the FQ-learning , but also variables of the FSA models . As a result , neurons coding variables of the FQ-learning were still detected ( Fig 7A–7C ) as in our previous analysis [18] , although the performance of the FQ-learning model was worse than that of the FSA model . Significant proportions of neurons in which the firing rates were correlated with action values ( QL or/and QR ) were found in all regions ( Fig 7B ) . Significant proportions of state value- ( Fig 7A ) and chosen value-coding neurons ( Fig 7C ) were found mainly in DMS and VS . A substantial proportion of striatal neurons also coded internal states of the FSA model ( Fig 7D–7F ) . A significant proportion of cluster-coding ( CL , CR , and/or CWSLS ) neurons were found in VS ( Fig 7D ) , which might be similar to the strategy-coding neurons reported in monkey striatum [26] . The proportion of neurons coding x ( t ) in DMS showed a peak during the action execution ( Fig 7E ) . After entry into the left or right hole ( and the reward or no-reward tone was presented ) , populations of x ( t+1 ) in all regions were increased ( Fig 7F ) , consistent with state transition dependence on reward feedback . Some neurons in DMS showed firing correlated with x ( t+1 ) even before presentation of the reward or no-reward tone ( Fig 7F ) , which was possible because the reward was highly predictable ( 90% or 10% ) in one of the actions in each block . Were variables of the FQ-learning and the FSA models separately coded in different neurons ? During action execution ( 500 ms before entry into the L/R hole ) , neurons coding only the variables of the FQ-learning model ( state value , action value , chosen value ) were 6 . 9% ( 14/204 ) in DLS , 8 . 9% ( 10/112 ) in DMS , and 2 . 9% ( 4/138 ) in VS . Neurons coding only FSA-related variables ( sub strategy , x ( t ) , x ( t+1 ) ) were 8 . 8% ( 18/204 ) in DLS , 22 . 3% ( 25/112 ) in DMS , and 14 . 5% ( 20/138 ) in VS . Neurons coding both variables were 2 . 0% ( 4/204 ) in DLS , 7 . 1% ( 8/112 ) in DMS , and 8 . 7% ( 12/138 ) in VS . While VS neurons significantly tended to code variables of both models , in DLS and DMS there were no significant tendencies ( p = 0 . 10 for DLS , p = 0 . 13 for DMS , and p < 0 . 0001 for VS , chi-squared tests ) . Interestingly , not all states were equally coded in the striatum ( Fig 8 ) . During action execution ( Fig 8A ) , only the proportion of state-4- and state-5-coding neurons in DMS and VS ( also state 6 and 8 in DMS ) were statistically significant , and both states preceded an exploratory action in the keep-left and keep-right clusters ( Fig 4C ) . After execution of an action and reward feedback ( Fig 8B ) , representations of most subsequent states appeared in DLS and DMS , while representations of the same state x5 , persisted in VS . Interestingly , states 2 and 7 are major transition targets from states 4 and 5 , and these signals , especially , the signal of state 7 , were prominent in DLS . The finite state-based strategy implemented with N = 8 states showed a significantly higher prediction accuracy ( average likelihood ) for rat choice behaviors than the best reinforcement learning model , the FQ-learning model [5] [18] . Furthermore , we compared statistical features of the time course of learning ( the number of trials to reach 80% optimality ) and the probabilities of repeating the same action after rewarded or non-rewarded outcomes of the rats and the algorithms when faced the same task ( Fig 5 ) . We found that only the FSA model with 8 states could reproduce those features similar to the rats . Therefore the FSA model is the best model to predict rat actions in individual trials and also to reproduce generic features of the time course of learning , although we cannot deny the possibility that there might be an even better model in both respects . The FSA model is conceptually different from the other models . The Q-learning ( FQ-learning ) models and the ESE models are normative models that prescribe behaviors for maximization of rewards , whereas the FSA model is a descriptive model that seeks only to describe the behavior as it appears in the data [27] . The reformulated Baum-Welch algorithm was used not to find the parameters with which the models maximize the reward , but to find the parameters with which the models mimic the choice behavior of rats . The FSA models do not explain why and how the rats learned the procedure ( Fig 4C ) . If an FSA-like algorithm is implemented in the brain , how could the algorithm learn the appropriate choice and transition probabilities to efficiently obtain a reward ? A possible scenario is that rats use the value-based strategy in the beginning of the training . Meanwhile , the finite state strategy monitored behavior to form a procedure that mimicked the value-based strategy without explicit value evaluation . After massive training , the procedure was formed , and the finite state strategy overrode action selection . We speculate that the finite state strategy could be regarded as generalized habit formation . Traditionally , habitual actions are considered automatic responses controlled by simple stimulus-response associations without any associative links to the outcome of those actions [28] . The finite state strategy could be considered as an extended habitual action that depends not only on stimuli , but also internal states . To test this idea , further behavioral experiment will be required . Internal states of the FSA model were represented in the all three subregions of the striatum ( Fig 7E and 7F ) , while it has been reported that habitual actions involve DLS [28–31] . We speculate that retention of internal states required for the FSA model involves the working memory functions of the prefrontal cortex [32] , which can explain the internal state representation in not only DLS , but also DMS and VS , where the prefrontal cortex projects [33] . Analysis of neuronal activities suggests that all striatal areas we recorded , namely , DLS , DMS , and VS , are involved in the finite state strategy . Interestingly , not all states were equally coded in the striatum ( Fig 8A and 8B ) . While codings of x4 ( t ) and x5 ( t ) were found in DMS and VS , coding of x1 ( t ) , x2 ( t ) , x3 ( t ) , and x7 ( t ) was not observed in any areas . Note that x4 ( t ) and x5 ( t ) are the states in which an action is likely to be switched after repeated unrewarded actions at x1 ( t ) , x2 ( t ) , x8 ( t ) or x7 ( t ) . This uneven representation of states suggests that the finite state strategy is implemented in a larger brain circuit that includes the striatum . The requirement of working memory to store the current state suggests the involvement of other brain regions , such as the prefrontal cortex and the hippocampus . Then why are x4 ( t ) and x5 ( t ) are selectively coded in the striatum ? It has been reported that the anterior cingulate cortex ( ACC ) plays an important role in switching behavior evoked by error feedback [34] . The connection from the ACC to the striatum for the execution of switching [35] may be the source of strong coding of x4 ( t ) and x5 ( t ) observed in DMS and VS . Previous studies have reported that action-value signals are represented in the striatum of rodents [5 , 7 , 8] , monkeys [4 , 11 , 23 , 36] and humans [12] , suggesting that the value-based strategy is implemented in the basal ganglia . Consistent with these reports , our previous study [18] reported that state value signals were most strongly represented in VS , and that action value signals were most strongly represented in DMS during action execution . In the present study , we reanalyzed the same dataset as the previous study , with a more complex regression model , including not only action values , but also state values , the chosen value , and variables of the FSA model that best explained animal behaviors . We applied lasso regularization to the augmented regression model , and similar results were reproduced; strong state-value coding in VS ( Fig 7A ) , and a peak of the proportion of action-value coding neurons in DMS during action execution ( Fig 7B ) . In addition , we found that the signal of the chosen value , previously reported in monkeys [23 , 26] and rats [7] , was represented in VS in our dataset ( Fig 7C ) . It has been proposed that DMS is involved in goal-directed actions [28 , 30] based on lesion studies [37 , 38] . Formation of goal-directed action is thought to require an association between actions and outcomes , which is analogous to the action value in reinforcement learning . Accordingly , action-value coding in DMS matches the proposal of goal-directed action in DMS . The action value for the selected action , called the chosen value [7 , 23] , which is necessary for updating action values , was observed in VS . Furthermore , consistent with previous reports in rodents [5] , state-value representation was observed in VS ( Fig 7E ) . These findings suggest that the value-based strategy is implemented in the striatum , although the final action choices are better characterized by the finite state-based strategy . The likelihood of the ESE model for the model-based strategy was much lower than that of the FQ-learning model for the value-based strategy or that of the FSA model for the finite state strategy . Thus , rats may not have estimated the reward setting in our task . In this task , four pairs of reward probabilities were used , but in the previous report in human subjects [13] , only two pairs were used . Therefore , it might be too difficult for rats to estimate one reward setting from four possible pairs . The present results support the notion of a hierarchical structure in the cortico-basal ganglia loops , but suggest specific roles for different loops in implementation of the value-based and finite state-based strategies . Representation of state values and sub-strategies ( clusters ) in VS ( Fig 7A and 7D ) suggests a role for this region in higher-level decisions , namely , selection of sub-strategies depending on the frequency of reward [39 , 40] . Robust coding of action values and states responsible for action switching in DMS ( Fig 7D and 7G ) points to a role for this region in flexible action adaptation . Action coding in DLS was equal to or stronger than that in DMS before movement onset [18] , suggesting a major role for this region in action preparation and initiation . All experimental procedures were performed in accordance with guidelines approved by the Okinawa Institute of Science and Technology Experimental Animal Committee . A part of the dataset used in our previous study [18] was reused in this study . Behavioral and neuronal data were gathered from seven Long-Evans rats . The number of sessions completed by each rat was from 24 to 33 . The average ( + standard deviation ) of the trials per session was 41 . 10 ( + 27 . 58 ) trials . Neurons stably recorded from at least two sessions were 260 in DLS , 178 in DMS , and 179 in VS ( on average , recorded from 2 . 7 sessions ) . From this dataset , phasically active neurons ( PANs; 204 from DLS , 112 from DMS , and 138 from VS ) were extracted based on inter-spike interval statistics . The proportion of inter-spike intervals ( ISIs ) that was > 1 s of total recoding time ( PropISIs>1s ) was calculated for each neuron [41] . Then , neurons for which PropISIs>1s> 0 . 4 were regarded as PANs . Intervals of the six task events ( entry into the center hole , onset of the cue tone , offset of the cue tone , exit from the center hole , entry into the left or right hole , and exit from the left or right hole ) varied by trials . To align event timings for all trials , event-aligned spike histograms ( EASHs ) were proposed by Ito and Doya [18] . First , the average duration for each event interval was calculated . Then , spike timings in a certain event interval for each trial were linearly transformed into corresponding averaged event intervals . Finally , histograms of the number of spikes for each 100 ms time window were calculated ( Fig 6A , 6D and 6G ) . Any decision-making models for a single stimulus ( state ) and binary choice ( action ) can be defined by the conditional probability of a current action given past experiences: PL ( t ) =P ( a ( t ) =L\|e ( 1:t−1 ) ) ( 2 ) where e ( 1:t-1 ) is a simple description of e ( 1 ) , e ( 2 ) , … , e ( t-1 ) . e ( t ) is a set of an action and a reward e ( t ) = {a ( t ) , r ( t ) } , and action a ( t ) and reward r ( t ) can be L or R and 1 or 0 , respectively . Behavioral data are composed of a set of sequences ( sessions ) of actions and rewards . If necessary , we use the index l as the index of sessions , for example a{l} ( t ) . The number of trials for session l is represented by Tl , and the number of sessions is L . To fit parameters to choice data and to evaluate the models , we used the likelihood criterion , which is the probability that the observed data were produced by the model . The likelihood can be normalized , so that it equals 0 . 5 when predictions are made with chance-level accuracy ( PL ( t ) = 0 . 5 for all t ) . The normalized likelihood is defined by Z=[∏l=1L[∏t=1Tlz{l} ( t ) ]]1∑l=1LTl ( 3 ) where z{l} ( t ) is the likelihood for a single trial: z{l} ( t ) ={PL ( t ) ifa{l} ( t ) =L1−PL ( t ) ifa{l} ( t ) =R . ( 4 ) The ( normalized ) likelihood can be regarded as the prediction accuracy , namely , how accurately the model predicts actions using past experiences . Generally , models that have a larger number of free parameters can fit data more accurately and thus show a higher likelihood . However , these models may not be able to fit new data due to over-fitting . For fair comparison of models , choice data were divided into training data ( 101 sessions ) and test data ( 101 sessions ) . Free parameters of a model were determined to maximize the likelihood of training data . Then , the model was evaluated by the likelihood or the normalized likelihood of the test data ( holdout validation ) . Therefore , in this model fitting , each model was fitted to all training set trials from all seven rats with the same free parameters . Fig 2A represents the normalized likelihood for the total of test 101 sessions . For statistical tests of the normalized likelihood between the models ( Fig 2A ) , we compared the normalized likelihood of each session for the same parameters between the models by a paired-sample Wilcoxon test . From the above process , we obtained the likelihood of each trial ( 4 ) in all sessions ( both training and test data ) for each model with the parameters estimated by training data . To compare fitting performance , we averaged the sequences of the likelihoods of the last 20 trials over all blocks with higher or lower reward probabilities ( Fig 2B and 2C ) . To test significant differences between the FSA model and the DFQ model , the Mann-Whitney U test was applied to the likelihoods for every trial . Note that the normalized likelihood depends on the number of trials . If an animal’s choice probability does not change over trials , namely , P ( a ( t ) = L ) = P , and model prediction PL ( t ) is also constant PL , then the expected normalized likelihood for T trials is given by Z^ ( T ) =∑t=0T ( Tt ) Pt ( 1−P ) T−t ( PLt ( 1−PL ) T−t ) 1/T . ( 5 ) This expected normalized likelihood rapidly decreases when the number of trials increases , and when T goes to infinite , it converges to Z^ ( ∞ ) =PLP ( 1−PL ) ( 1−P ) . ( 6 ) For example , let’s assume that a rat’s choice probability is P = 0 . 8 and model A predicts it perfectly by PL = 0 . 8 , the ( normalized ) likelihood is Z^ ( 1 ) =0 . 68 , it’s less than PL , and it decreases to Z^ ( ∞ ) =0 . 61 when T increases . If model B predicts with PL = 0 . 7 , Z^ ( 1 ) =0 . 62 and Z^ ( ∞ ) =0 . 59 , the difference in the normalized likelihood between model A and model B also decreases ( 0 . 07 → 0 . 02 ) when T changes from 1 to infinity . This is the reason why the normalized likelihoods of models shown in Fig 2A ( T = 16856 trials ) are much less than the likelihoods shown in Fig 2B and 2C ( T = 1 trial ) . dth-order Markov models are the simplest non-parametric models . They predict an action at trial t , a ( t ) , from the past d-length sequence of experiences before t , e ( t-d:t-1 ) . The prediction of the dth-order Markov model was given by the following: PL ( t ) =NL ( e ( t−d:t−1 ) ) +1NL ( e ( t−d:t−1 ) ) +NR ( e ( t−d:t−1 ) ) +2 ( 7 ) where Ni ( e ( t − d:t − 1 ) ) is the number of i ( L or R ) chosen after every d-length sequence of the exact same sequence as e ( t-d:t-1 ) in the whole training data [5] . The dth-order Markov model has more than 4d free parameters because there are four types of possible experiences in a single trial ( more precisely , the number of the parameters is 4d+4 ( d−1 ) +⋯+4 . The dth-order Markov model uses the 1st-order Markov model for the prediction of the first trial in a session , and 2nd-order Markov model for the second trial ) . The Markov models are purely descriptive models , but they provide a useful measure to objectively evaluate other models . The DFQ-learning model [5 , 18] , which is an extension of the Q-learning model and which includes the original Q-learning model with certain parameters , is useful to test the Q-learning family . A key component of the DFQ-learning ( and Q-learning ) model is to use action values ( QL and QR ) as predictions of the future cumulative reward that the agent would obtain after selecting left or right , respectively . The model selects an action that has a higher action value with a higher probability: PL ( t ) =11+exp{− ( QL ( t ) −QR ( t ) ) } . ( 8 ) After determining the reward outcome , action values are updated by: Qi ( t ) ={ ( 1−α1 ) Qi ( t−1 ) +α1κ1ifa ( t−1 ) =i , r ( t−1 ) =1 ( 1−α1 ) Qi ( t−1 ) −α1κ2ifa ( t−1 ) =i , r ( t−1 ) =0 ( 1−α2 ) Qi ( t−1 ) ifa ( t−1 ) ≠i , r ( t−1 ) =1 ( 1−α2 ) Qi ( t−1 ) ifa ( t−1 ) ≠i , r ( t−1 ) =0 ( 9 ) where i ∈ {L , R} , α1 is the learning rate for the selected action , α2 is the forgetting rate for the action not chosen , κ1 represents the strength of reinforcement by reward , and κ2 represents the strength of the aversion resulting from the non-reward outcome . This set of equations can be reduced to the standard Q-learning by setting α2 = 0 ( no forgetting for actions not chosen ) and κ2 = 0 ( no aversion from a lack of reward ) . The FQ-model is a version introducing the restriction α1 = α2 . For the Q-learning models , we considered cases of fixed parameters and time-varying parameters . For fixed parameter models , α1 , α2 , κ1 , and κ2 are free parameters . For time-varying parameters , α1 , α2 , κ1 , and κ2 are not free parameters; they are assumed to vary according to the following: αj ( t ) =αj ( t−1 ) +ςjforj∈{1 , 2}κj ( t ) =κj ( t−1 ) +ξjforj∈{1 , 2} ( 10 ) where ζj and ξj are noise terms drawn independently from the Gaussian distribution N ( 0 , σα2 ) and N ( 0 , σκ2 ) , respectively . σα and σκ are free parameters that control the magnitude of the change . The predictive distribution P ( h ( t ) \| e ( 1:t-1 ) ) of parameters h = [QL , QR , α1 , α2 , κ1 , κ2] given past experiences e ( 1:t-1 ) was estimated using the particle filter [4 , 5] . The action probability PL ( t ) was obtained from Eq ( 8 ) with the mean of the predictive distribution of QL ( t ) and QR ( t ) . In this study , 5 , 000 particles were used for the estimation . The ESE model estimates a hidden environmental state , namely , the reward setting from past experience , using the knowledge that reward probabilities should be one of the following: ( 90 , 50% ) , ( 50 , 10% ) , ( 50 , 90% ) and ( 10 , 50% ) ( five trials with zero reward probability inserted in the middle of each session were not considered . ) . The ESE model also assumes that the reward setting is changed with a small probability ε for each trial: P ( s ( t ) \|s ( t−1 ) ) ={1−εifs ( t ) =s ( t−1 ) ε/3ifs ( t ) ≠s ( t−1 ) ( 11 ) where s ( t ) ∈ {1 , 2 , 3 , 4} is the index of reward setting at trial t corresponding to ( 90 , 50% ) , ( 50 , 10% ) , ( 50 , 90% ) and ( 10 , 50% ) , respectively . The prediction of the reward setting at trial t for all s ( t ) is obtained using P ( s ( t ) \|e ( 1:t−1 ) ) =∑s ( t−1 ) =14P ( s ( t ) \|s ( t−1 ) ) P ( s ( t−1 ) \|e ( 1:t−1 ) ) ( 12 ) where P ( s ( t-1 ) \| e ( 1:t-1 ) ) is the prior probability of the reward setting . The prior probability for t = 1 was set to 1/4 for each s . Based on this prediction , action values are given by Qi ( t ) =κ∑s ( t ) =14P ( r ( t ) =1\|s ( t ) , a ( t ) =i ) P ( s ( t ) \|e ( 1:t−1 ) ) ( 13 ) where P ( r ( t ) = 1\| s ( t ) , a ( t ) = i ) is the reward probability for the reward setting s ( t ) and action i . κ is the magnitude of the reward . An actual action , a ( t ) , is selected according to the action probability , which is calculated from Eq ( 8 ) with the action values . After knowing the reward outcome , r ( t ) , the posterior probability of the reward setting for all s ( t ) , was updated using Bayes’ theorem: P ( s ( t ) \|e ( 1:t ) ) ∝P ( a ( t ) , r ( t ) \|s ( t ) , e ( 1:t−1 ) ) P ( s ( t ) \|e ( 1:t−1 ) ) . ( 14 ) The first factor of the right side can be decomposed to P ( a ( t ) , r ( t ) \|s ( t ) , e ( 1:t−1 ) ) =P ( r ( t ) \|a ( t ) , s ( t ) , e ( 1:t−1 ) ) P ( a ( t ) \|s ( t ) , e ( 1:t−1 ) ) ( 15 ) where the first factor on the right side of this equation can be simply written as P ( r ( t ) \| a ( t ) , s ( t ) ) because this factor comes from the reward probability setting of the task and is assumed to be independent of the past experience of rats , e ( 1:t-1 ) . The second factor is the action probability of the agent . Although the agent estimates the current reward setting , s ( t ) , from past experience , e ( 1:t-1 ) , the agent cannot directly observe s ( t ) . In other words , the action probability should be the same for the same past experience , e ( 1:t-1 ) , without being affected by the true hidden state , s ( t ) . Therefore , the second factor can be ignored because it takes the same values for all s ( t ) . Then , Eq ( 14 ) is simplified to P ( s ( t ) \|e ( 1:t ) ) ∝P ( r ( t ) \|s ( t ) , a ( t ) ) P ( s ( t ) \|e ( 1:t−1 ) ) . ( 16 ) Similar to the Q-learning models , we considered the cases of fixed and time-varying parameters . For fixed parameter models , ε and κ are free parameters . For time-varying parameters , ε and κ were assumed to vary by a random walk with the Gaussian distribution N ( 0 , σε2 ) and N ( 0 , σκ2 ) , respectively . σε and σκ are the free parameters that control the magnitude of the change . FSA models are non-parametric models that have internal variables x taking N possible states , x ∈ {1 , 2 , ⋯ , N} . The initial distribution of the state is described by qn=P ( x ( t=1 ) =n ) . ( 17 ) The probability of an action selection depends on the state and is defined by πn ( a ( t ) ) =P ( a ( t ) \|x ( t ) =n ) . ( 18 ) After execution of an action and the subsequent reward outcome , the state is probabilistically moved to another state according to the state transient function: Unm ( a ( t ) , r ( t ) ) =P ( x ( t+1 ) =m\|x ( t ) =n , a ( t ) , r ( t ) ) . ( 19 ) qn , πn ( a ) , Unm ( a , r ) , and N are the free parameters of the FSA models . Considering the probabilistic constraints , ∑n=1Nqn=1 , ∑a={L , R}πn ( a ) =1 , ∑m=1NUnm ( a , r ) =1 , and symmetric constraints ql = qN−l+1 , πl ( a ) =πN−l+1 ( a¯ ) , Ull' ( a , r ) =UN−l+1N−l'+1 ( a¯ , r ) , for l , l’ = 1 , 2 , … , N , where a¯ is the other of actions , and the number of free parameters is ( N/2-1 ) + N/2 + 2N ( N-1 ) = 2N2-N-1 when N is an even number and ( N-1 ) /2 + ( N-1 ) /2 + 2 ( N-1 ) 2 = 2N2 - 3N +1 when N is an odd number . The FSA model can be regarded as an extended version of the hidden Markov model ( HMM ) . However , unlike the HMM , in the FSA model , the state transition probability depends on the action and reward . In the HMM , the Baum-Welch algorithm [19] , a form of the EM algorithm , is used to find the parameters that maximize the likelihood of the given data . We reformulated the Baum-Welch algorithm for the FSA model . 1 . Initialize parameters qn , πn ( a ) , Unm ( a , r ) , so the probabilistic constraints , ∑n=1Nqn=1 , ∑a={L , R}πn ( a ) =1 , ∑m=1NUnm ( a , r ) =1 , and the symmetric constraints , ql = qN−l+1 , πl ( a ) =πN−l+1 ( a¯ ) , Ull' ( a , r ) =UN−l+1N−l'+1 ( a¯ , r ) , are satisfied ( N is a fixed parameter ) . In this study , we set qn = 1/N for all n , πn ( L ) = 0 . 9−0 . 8 ( n−1 ) / ( N−1 ) and πn ( R ) = 1−πn ( L ) for all n , and Unm ( a , r ) =1/N for all n , m , a , and r . 2 . E-step Estimate the posterior probability of the state for all t and l , γn{l} ( t ) = P ( x{l} ( t ) = n\|a{l} ( 1:Tl ) , r{l} ( 1:Tl ) ) , assuming that the data were produced with current parameters . First , estimate αn{l} ( t ) = P ( x{l} ( t ) = n\|a{l} ( 1:t ) , r{l} ( 1:t ) ) , the posterior probability of a state at trial t given the data from 1 to the current trial t . The probability can be obtained iteratively from t = 1 to t by: αn{l} ( t ) =πn ( a{l} ( t ) ) ∑mUmn ( a{l} ( t−1 ) , r{l} ( t−1 ) ) αm{l} ( t−1 ) ∑n'πn' ( a{l} ( t ) ) ∑m'Um'n' ( a{l} ( t−1 ) , r{l} ( t−1 ) ) αm'{l} ( t−1 ) , ( 20 ) where αn{l} ( t = 1 ) = qn . Next , estimate γn{l} ( t ) = P ( x{l} ( t ) = n\|a{l} ( 1:Tl ) , r{l} ( 1:Tl ) ) , the posterior probability of a state , given all data using the already obtained αn{l} ( t ) and an additional variable , χnm{l} ( t ) = P ( x{l} ( t ) = n , x{l} ( t + 1 ) = m \| a{l} ( 1:Tl ) , r{l} ( 1:Tl ) ) . γn{l} ( t ) and χnm{l} ( t ) are obtained in a backward manner from t = Tl ( or Tl-1 for χnm ) to 1 in parallel . For t = Tl , γn{l} ( Tl ) = αn{l} ( Tl ) . Then , χnm{l} ( Tl−1 ) is obtained using χnm{l} ( t ) =γm{l} ( t+1 ) Unm ( a{l} ( t ) , r{l} ( t ) ) αn{l} ( t ) ∑n'Un'm ( a{l} ( t ) , r{l} ( t ) ) αn'{l} ( t ) , ( 21 ) and γn{l} ( Tl−1 ) is obtained using γn{l} ( t ) =∑mχnm{l} ( t ) . ( 22 ) By repeating ( 21 ) and ( 22 ) , γn and χnm are obtained for all trials . 3 . M-step Update model parameters , assuming the state probabilities γn and χnm are true . Given the probabilistic and symmetric constraints , parameters are updated using qn=qN−n+1=∑l=1L[γn{l} ( 1 ) +γN−n+1{l} ( 1 ) ]2L , ( 23 ) πn ( a ) =πN−n−1 ( a¯ ) =∑l=1L∑t=1Tl[γn{l} ( t ) δ ( a{l} ( t ) =a ) +γN−n+1{l} ( t ) δ ( a{l} ( t ) =a¯ ) ]∑l'=1L∑t'=1Tl'[γn{l'} ( t' ) +γN−n+1{l'} ( t' ) ] , ( 24 ) and Unm ( a , r ) =UN−n+1N−m−1 ( a¯ , r ) =∑l=1L∑t=1Tl−1[χnm{l} ( t ) δ ( a{l} ( t ) =a , r{l} ( t ) =r ) +χN−n+1 , N−m+1{l} ( t ) δ ( a{l} ( t ) =a¯ , r{l} ( t ) =r ) ]∑l'=1L∑t'=1Tl−1[γn{l'} ( t' ) δ ( a{l'} ( t' ) =a , r{l'} ( t' ) =r ) +γN−n+1{l'} ( t' ) δ ( a{l'} ( t' ) =a¯ , r{l'} ( t' ) =r ) ] ( 25 ) 4 . Check for convergence of the parameters . If the convergence criterion is not satisfied , return to step 2 . In this study , we stopped the iteration when the maximum change of the parameters was less than 0 . 00001 . To test whether the models can generate behavioral data that have the same statistics as behavioral data , we compared the simulated behavior of the models with the behavioral results . First , as a measure of adaptation speed to the change of reward probabilities , we calculated the mean number of trials in one block for the higher reward probability settings , ( 90 , 50% ) and ( 50 , 90% ) , and for the lower reward probability settings , ( 50 , 10% ) and ( 10 , 50% ) , from all 202 recorded sessions . Because each session consists of four blocks with different reward probability settings , there were 404 higher reward blocks and 404 lower reward blocks in the data . Second , as a measure of the strategy utilized by the rats , the probability that the same action was selected after a rewarded or non-rewarded trial , P ( a ( t+1 ) = a ( t ) \| r ( t ) = 1 ) and P ( a ( t+1 ) = a ( t ) \| r ( t ) = 0 ) , for higher and lower reward probability settings , respectively , was calculated . These four action probabilities were calculated from the last 20 trials in four blocks for each session , and then the mean of these probabilities was calculated from all 202 sessions . Third , we conducted a model simulation for 202 sessions in which the same block sequences as those used in all 202 sessions were applied . Then , the six statistics noted above were calculated from the simulated data: [the number of trials in a block , P ( a ( t+1 ) = a ( t ) \| r ( t ) = 1 ) , P ( a ( t+1 ) = a ( t ) \| r ( t ) = 0 ) ] x [higher , lower reward probability setting] . By repeating this simulation 10 , 000 times , the approximate distribution for each statistic was obtained . Note that the statistics calculated from the rats are random variables . If the hypothesis that the choice behavior of rats was sampled from a certain model is true , then statistics obtained from behavioral measures should fall within the distribution of the statistics ( inside the confidence interval with 1—ε ) calculated by the model . Otherwise , the hypothesis is rejected . We considered six different tests for the same hypothesis , so the chance of at least one false rejection is much higher than ε . Therefore , the confidence interval for each statistic was set to 1—ε/6 , so the chance of at least one false rejection is ε ( Bonferroni Method ) . In this study , ε was set to 0 . 05 . We tested the Q- , FQ- , and DFQ-learning models in addition to the ESE model in which the parameters were fixed . The FSA models with 4 , 6 , and 8 states were also tested . The free parameters that maximize the likelihood of the training data were used for the simulation . The distributions of the statistics in one session ( Fig 5A–5C ) were calculated from 202 behavioral sessions and 10 , 000 x 202 sessions for the models . As a result , the shape of the distribution is smoother for the models than for behavioral data . Linear regression is a popular method to find regressors that can explain the change in neuronal activity , where spikes are assumed to be sampled from a normal distribution . However , in the precise sense , this assumption is not correct because spikes take only non-negative integers . The lower the firing rate of the neuron is , the bigger the gap from the assumption is . Therefore , in the present study , we used Poisson regression assuming that the spikes are sampled from a Poisson distribution ( a distribution of non-negative integer variables ) . In Poisson regression , the expected number of spikes at trial t , μ ( t ) , is predicted by the following exponential function , μ ( t ) =exp ( β0+β1x1 ( t ) +β2x2 ( t ) +⋯+βpxp ( t ) ) ( 28 ) where xi are regressors and βi are regression coefficients . The prediction of the number of the spikes at trial t is represented by a Poisson distribution with the average μ ( t ) , Poi ( y\|μ ( t ) ) =e−μ ( t ) μ ( t ) yy ! . ( 29 ) Optimal regression coefficients are determined so that the objective function , namely , the log likelihood for all trials , l ( β ) =∑t=1Tlog ( Poi ( y ( t ) \|μ ( t ) ) ) ( 30 ) is maximized . For this calculation , a function in MATLAB Statistics and Machine Learning Toolbox “glmfit ( X , y , ‘poisson’ ) ” is available . To select minimum regressors to explain the spikes among many and redundant regressors is , to add a penalty term for large β to the objective function , l ( β ) =∑t=1Tlog ( Poi ( y ( t ) \|μ ( t ) ) ) −λ∑j=1p\|βj\| ( 31 ) where λ is a free parameter called the regularization coefficient , and \|βj\| is the absolute value of βj ( this method is called lasso [24] ) . It has the property that if λ is sufficiently large , some of the coefficients β are driven to zero [42] . In the present study , all regressors were normalized , so that the average was 0 and the variance was 1 . Then , λ was optimized by 5-fold cross-validation for each time bin for each neuron , and regression coefficients were obtained . For these calculations , we used a function in the MATLAB Statistics and Machine Learning Toolbox , “lassoglm ( X , y , ‘poission’ , ‘cv’ , 5 ) ” . The regressors with non-zero coefficients were regarded as candidates for minimum regressors . Then , to calculate a p-value indicating the probability that each candidate could be incorrectly selected , we applied Poisson regression ( MATLAB function , glmfit ) to the regression model including only these candidates as regressors . We selected candidates for minimal regressors that had p-values < 0 . 01 . Proportions of neurons coding variables shown in Figs 7 and 8 are the fraction of neurons for which corresponding variables were regarded as minimal regressors for each time bin . The significance of the proportion ( p < 0 . 05 ) was calculated with a binomial test , assuming that the probability that a regressor could be selected incorrectly was p = 0 . 01 .	The neural mechanism of decision-making , a cognitive process to select one action among multiple possibilities , is a fundamental issue in neuroscience . Previous studies have revealed the roles of the cerebral cortex and the basal ganglia in decision-making , by assuming that subjects take a value-based reinforcement learning strategy , in which the expected reward for each action candidate is updated . However , animals and humans often use simple procedural strategies , such as “win-stay , lose-switch . ” In this study , we consider a finite state-based strategy , in which a subject acts depending on its discrete internal state and updates the state based on reward feedback . We found that the finite state-based strategy could reproduce the choice behavior of rats in a binary choice task with higher accuracy than the value-based strategy . Interestingly , neuronal activity in the striatum , a crucial brain region for reward-based learning , encoded information regarding both strategies . These results suggest that both the value-based strategy and the finite state-based strategy are implemented in the striatum .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[]	2015	Parallel Representation of Value-Based and Finite State-Based Strategies in the Ventral and Dorsal Striatum
Arboviruses cycle through both vertebrates and invertebrates , which requires them to adapt to disparate hosts while maintaining genetic integrity during genome replication . To study the genetic mechanisms and determinants of these processes , we use chikungunya virus ( CHIKV ) , a re-emerging human pathogen transmitted by the Aedes mosquito . We previously isolated a high fidelity ( or antimutator ) polymerase variant , C483Y , which had decreased fitness in both mammalian and mosquito hosts , suggesting this residue may be a key molecular determinant . To further investigate effects of position 483 on RNA-dependent RNA-polymerase ( RdRp ) fidelity , we substituted every amino acid at this position . We isolated novel mutators with decreased replication fidelity and higher mutation frequencies , allowing us to examine the fitness of error-prone arbovirus variants . Although CHIKV mutators displayed no major replication defects in mammalian cell culture , they had reduced specific infectivity and were attenuated in vivo . Unexpectedly , mutator phenotypes were suppressed in mosquito cells and the variants exhibited significant defects in RNA synthesis . Consequently , these replication defects resulted in strong selection for reversion during infection of mosquitoes . Since residue 483 is conserved among alphaviruses , we examined the analogous mutations in Sindbis virus ( SINV ) , which also reduced polymerase fidelity and generated replication defects in mosquito cells . However , replication defects were mosquito cell-specific and were not observed in Drosophila S2 cells , allowing us to evaluate the potential attenuation of mutators in insect models where pressure for reversion was absent . Indeed , the SINV mutator variant was attenuated in fruit flies . These findings confirm that residue 483 is a determinant regulating alphavirus polymerase fidelity and demonstrate proof of principle that arboviruses can be attenuated in mammalian and insect hosts by reducing fidelity . During replication , RNA viruses generate approximately 1 error per 104 nucleotides copied , giving rise to an immense population of genetically distinct but closely related variants [1] , [2] , [3] , [4] . The genetic diversity of these “mutant swarms” is not detected by consensus sequencing , which to-date has been the basis for most studies of viral infection . However , this lack of information on genetic diversity has obscured crucial aspects of virus biology . Although RNA-dependent RNA polymerases ( RdRp ) have a high intrinsic error rate , their mutation rates can be altered to generate both higher and lower fidelity variants ( antimutators and mutators , respectively ) [5] , [6] , [7] , [8] , [9] , [10] , [11] , [12] , [13] , [14] . Thus far , antimutator variants are thought to replicate more slowly , making fewer genomes with greater accuracy; in contrast , mutator variants have been shown to replicate more quickly , synthesizing more viral genomes but introducing many errors during the replication process [8] , [15] , [16] , [17] , [18] . Despite this , overall growth and titers of polymerase fidelity variants are not significantly different when grown in isolation in cell culture; for mutators the negative effects of accumulating deleterious mutations are only noticeable after several rounds of replication [6] , [19] . In recent works , these variants have been useful in exploring how the course of viral infection is affected by either restricted or expanded population diversity [4] , [10] . Current evidence indicates that mutation frequencies of RNA viruses have been optimized over time to be neither too accurate nor too erroneous [4] , [6] , [16] , [20] , [21] , [22] , [23] . It is thought that error-prone replication allows the virus to explore sequence space to gain adaptability and accumulate potentially advantageous mutations . For several RNA viruses , limiting viral population diversity has fitness costs in vivo . Despite similar in vitro growth phenotypes , variants that make fewer errors have reduced titers and exhibit restricted tropism in animal models [19] , . This restriction in tropism may be due to cooperative inter-variant interactions or beneficial minority variants that are missing in a situation with restricted population diversity [19] . It is also proposed that high mutation rates of influenza A may contribute to altered tropism , allowing infection of new hosts [26] . Therefore , it seems that the relatively high error rates of RNA viruses generate a level of diversity that facilitates adaptive fitness advantages . In contrast , there is also an upper threshold to mutation frequencies; if crossed , extreme error rates lead to the accumulation of deleterious mutations and loss of genetic integrity . Evidence for this is demonstrated by treatment of numerous RNA viruses with nucleoside analog mutagens , which increase mutation frequencies and result in extinction by lethal mutagenesis [27] , [28] , [29] , [30] , [31] . Although thus far RdRp mutators have not exhibited growth defects in isolation in vitro , a recent paper showed that HIV mutator and antimutator strains were less fit than wildtype in competition assays [32] . In addition , several studies recently report in vivo attenuation of mutator strains: Coxsackie virus B3 mutator strains present reduced viral titers in key organs and fail to establish persistent infections in mice [6] , and a severe acute respiratory syndrome ( SARS ) coronavirus mutator strain exhibits reduced pathogenesis in several mouse models [7] . Antimutator and mutator variants are valuable tools to study where the threshold of advantageous polymerase error exists for viruses facing different selective pressures . In this respect , arboviruses represent a special evolutionary position due to their need to replicate in disparate hosts , which is accompanied by distinct selective pressures . Arbovirus fitness is not necessarily reduced due to obligate host-cycling ( alternating passages of CHIKV did not limit viral fitness ) , yet it has been shown that evolvability may be reduced due to these evolutionary constraints [33] , [34] , [35] , [36] , [37] , [38] . For alphaviruses , evidence suggests that viral diversity is most restricted in the insect host , due to more stringent population bottlenecks and selective pressures [33] , [34] , [35] , [37] , [39] , [40] . Since minority variants are thought to play important roles in arbovirus pathogenesis , transmission , and emergence [25] , [41] , [42] , [43] , [44] , [45] , the implications of altered polymerase fidelity and mutation rates merit further study . Recently , this question was partially addressed using a chikungunya virus antimutator variant [25] . Chikungunya virus ( CHIKV ) is a re-emerging arbovirus , transmitted by Aedes species mosquitoes . This positive-stranded RNA virus ( family Togaviridae , genus Alphavirus ) has an 11 . 8 kB genome , of which the first 7 . 5 kb encode four nonstructural proteins ( nsP1-4 ) involved in diverse processes including RNA synthesis , immune evasion , and host tropism [46] , [47] , [48] , [49] , [50] , [51] , [52] . In most cases , functions of these proteins are putative in CHIKV and have only been shown in related model viruses , such as Semliki forest virus and Sindbis virus [53] , [54] , [55] , [56] . Nsp4 is the RdRp , responsible for nucleotide incorporation during replication [57] . Previously , we isolated an antimutator strain of CHIKV by passaging virus in ribavirin , an RNA nucleoside analog . Ribavirin causes nucleotide misincorporation by the RdRp , adding selective pressure for an intrinsically more faithful polymerase [14] , [58] , [59] . This antimutator strain harbored a single amino acid change ( 483Y ) in nsp4 . Although 483Y showed no growth defects in vitro , the variant was moderately attenuated in vivo in both mammalian and mosquito hosts [25] . However , no arbovirus mutators have been isolated thus far . To this end , we mutated the conserved cysteine residue at position 483 to obtain several mutators in the arboviruses CHIKV and SINV , confirming this position's importance in determining alphavirus fidelity . We used these novel mutator strains to examine how increased polymerase error affects arbovirus fitness in vitro and in vivo; interestingly , mutator strains presented distinct cell- and host-specific phenotypes . Mammalian cell lines Vero , HeLa , and BHK-21 were maintained in DMEM ( Gibco ) supplemented with 10% newborn calf serum ( NCS , Gibco ) and 1% penicillin-streptomycin ( P/S , Sigma ) , at 37°C with 5% CO2 . Mosquito cell lines C6/36 and U4 . 4 ( Aedes albopictus ) and Aag2 ( Aedes aegypti ) were grown in L-15 media , supplemented with 10% fetal bovine serum ( FBS , Gibco ) , 1% P/S , 1% tryptose phosphate , and 1% non-essential amino acids ( NEAA ) , at 28°C with 5% CO2 . Drosophila melanogaster S2 cells were grown in Schneider's Drosophila media ( Gibco ) , supplemented with 10% FBS , 1% L-Glutamine , and 1% P/S at 25°C . Wildtype CHIKV was generated from the La Reunion strain 06-049 infectious clone , previously described [33] . Nsp4 position 483 mutants were generated by site-directed mutagenesis of the infectious clone using the QuikChange II XL Site-Directed Mutagenesis kit ( Stratagene ) . All newly generated DNA plasmids were Sanger sequenced in full ( GATC Biotech ) to confirm mutagenesis of position 483 and to ensure no second-site mutations were introduced . Select SINV mutants were constructed in the same fashion from the pTR339 wildtype infectious clone [60] . CHIKV and SINV expression plasmids were linearized with NotI or XhoI respectively , purified by phenol-chloroform extraction and ethanol precipitation , and subsequently used for in vitro transcription of viral RNAs using the SP6 mMESSAGE mMACHINE kit ( Ambion ) . RNAs were then purified by phenol:chloroform extraction and ethanol precipitation , quantified , diluted to 1 µg/µl and stored at −80°C . For RNA transfections , BHK-21 cells were trypsinized , washed twice with ice-cold PBS , and resuspended at a concentration of 2×107 cells/ml in ice-cold PBS . Cells ( 0 . 390 ml ) were mixed with 10 µg of in vitro transcribed viral RNA , placed in 2 mm cuvette and electroporated at 1 . 2 kV , 25 µF with infinite Ω in a XCell Gene Pulser ( BioRad ) . Cells were allowed to recover for 10 minutes at room temperature then mixed with 6 ml of pre-warmed media and placed into a T-25 flask . After 48 hours incubation at 37°C , viral titers were determined by standard plaque assay . In brief , 10-fold serial dilutions of each virus in DMEM were incubated on a confluent monolayer of Vero cells for 1 hour at 37°C . Following incubation , cells were overlaid with 0 . 8% agarose dissolved in DMEM and 2% NCS and incubated at 37°C for 72 hours . The cells were then fixed with 4% formalin for 1 hour , the agarose plugs were removed , and plaques were visualized by the addition of crystal violet . Plaque size was quantified by scanning the crystal violet-stained cell monolayer , then quantifying the size of each individual plaque in square millimeters using ImageJ ( http://rsbweb . nih . gov/ij ) . Each virus was then passaged once over a 70–80% confluent monolayer of BHK-21 cells , titered as described above , aliquoted , and stored at −80°C until use . To analyze each virus for reversion at position 483 , viral RNA was extracted for each electroporation and BHK-21 passage using TRIzol reagent ( Invitrogen ) . For CHIKV , this RNA was used to amplify a 3184 bp region corresponding to nucleotides ( 4522–7706 ) , which included position 483 , using the forward primer ( 5′-GATGAGCACATCTCCATAG-3′ ) and the reverse primer ( 5′-GTTTGGGTTGGGATGAACT-3′ ) and the Titan One Tube RT-PCR Kit ( Roche ) . For SINV , a 2225 bp region ( nucleotides 6556–8781 ) was amplified in the same fashion using forward primer ( 5′-ACCAGGCACGAAACACACAGAA-3′ ) and reverse primer ( 5′-ACTGGGCGGAAGTCTGTATGCG -3′ ) . Each PCR product was cleaned using the Nucleospin PCR and Gel Extraction Kit ( Macherey-Nagel ) and Sanger sequenced at position 483/482 to confirm genetic stability . At passage 3 , all viruses used were fully sequenced to ensure no second site mutations . HeLa cells ( 250 , 000 cells/well in 12-well tissue culture plates ) were pre-treated for two hours with either media containing no mutagen , or media containing 200 µM or 400 µM ribavirin ( Sigma ) . Post-treatment , media was removed and the cells were inoculated with virus in DMEM at an MOI 0 . 1 for one hour at 37°C . Following incubation , mutagen-containing media was replaced and cells were incubated for 72 hours at 37°C . Virus was harvested at 72 hours and mean titers were obtained by TCID50 . In brief , a 96-well tissue culture plates was plated for each virus with 1×104 Vero cells/well . Viruses were serially diluted in 8 ten-fold dilutions in DMEM . Each dilution was distributed in a row of the 96-well plate , with each well receiving 100 µl of diluted virus . Viruses and cells were incubated 5–7 days at 37°C with 5% CO2 . Following incubation , cells were fixed with 50 µl of 4% formalin for 30 minutes . All media were removed , and 50 µl of crystal violet was added to each well . Viruses that exhibited significant sensitivity or resistance compared to wildtype at P<0 . 05 or greater at either 200 µM or 400 µM ribavirin were considered potential fidelity variants , and mutation frequencies were estimated ( Table 1 ) . To determine mutation frequencies , all mutants were electroporated in tandem into BHK-21 cells . Supernatants were collected 48 hours later and viral RNA was extracted . For CHIKV , an approximately 800 bp region corresponding to nucleotides 9943–10726 was amplified of the E1 region of the genome using forward primer 5′-TACGAACACGTAACAGTGATCC-3′ and reverse primer 5′-CGCTCTTACCGGGTTTGTTG-3′ . For SINV , the analogous region was amplified using forward primer 5′-TACGAACATGCGACCACTGTTC-3′ and reverse primer 5′-CGCTCGGAGCGGATTTACTG-3′ , and approximately 500 bases of this fragment was included in the analysis . Amplified fragments were purified as described above , and 3 µl of each product was modified by a 3′ A-overhang addition reaction ( 1 µl AmpliTaq Gold 10× buffer , 1 µl 10 mM dATP ) . Modified products were cloned using the TopoTA cloning kit ( Invitrogen ) , and single colonies were picked for sequencing . Mutation frequencies were determined as previously described [61] . Mutation frequencies in mosquito cells were obtained in the same fashion using the samples obtained from C6/36 growth curves ( we determined mutation frequency for a wildtype sample electroporated into C6/36 cells , and there was no difference between samples generated by infection or electroporation; the nonviability of mutators transfected into mosquito cells made it impossible to estimate mutation frequencies in C6/36 by electroporation ) . We sequenced approximately 75 clones per viral population in C6/36 cells . Mutation frequencies from mouse muscle were determined using RNA extracted from homogenized muscle samples from mice that most closely represented the median titer for that variants . For estimating in vivo mutation frequencies , a minimum of 50 clones were sampled per population . To confirm that the presence of RNA or aberrant viral particles in supernatants/homogenates did not affect mutation frequencies , we purified virus on 20% sucrose cushion and re-estimated mutation frequencies; no differences were observed . To estimate the population diversity of variants by deep sequencing , cDNA libraries were prepared by Superscript III from RNA extracted from virus generated in BHK-21 or C6/36 cells , and the viral genome was amplified using a high fidelity polymerase ( Phusion ) to generate 5 overlapping amplicons 2–3 kb in length . PCRs were fragmented ( Fragmentase ) , multiplexed , clustered , sequenced in the same lane with Illumina cBot and GAIIX technology and analyzed with established deep sequencing data analysis tools and in house scripts . Briefly , per-base Phred quality scores were utilized to trim bases with error probabilities higher than 0 . 001 , and sequences with less than 16 bases after trimming were discarded . For this purpose we used the fastq-mcf tool from the ea-utils toolkit at http://code . google . com/p/ea-utils [62] . The alignment step is performed using Burrows Wheeler Aligner [63] and Pileup is performed using SAMtools [64] . Once the pileup is done , an in-house script collects the data per-position and calculates the variance at each nucleotide position by root mean square deviation ( RMSD ) and determines the mean variance and standard error across the whole genome [65] . To estimate population diversity in a phenotypic assay , we performed neutralization assays using viruses which had been passaged 3 times on BHK-21 cells , using the n Neutralizing antibody CHK-102 ( a kind gift from Dr . M . S . Diamond [66] ) . 100 pfu of wildtype and mutator CHIKV strains were incubated for 1 hour at 37°C with serial dilutions of antibody , ranging from 2 µg/ml to . 0001 µg/mL , or left untreated . Virus-antibody complexes were added to pre-seeded confluent monolayers of Vero cells , and allowed to bind at 37°C for 1 hour . Assays were then overlayed with agarose and developed as described above for a plaque assay . Plaques were counted and normalized to the untreated control for each virus . Virus growth was evaluated for WT and all mutant viruses in BHK-21 , C6/36 , U4 . 4 , Aag2 , and S2 cells and titers were determined by TCID50 on Vero cells as described above . Using the 24 hour time point from the C6/36 growth curve , we also performed a cytopathic effect ( CPE ) assay on C6/36 cells on all viruses ( CellTiter 96 AQueuos One Solution Cell Proliferation Assay ( MTS ) kit; Promega ) . We obtained similar titers by standard TCID50 and CPE assay , indicating that viruses amplified on mosquito cells were still equally infectious when titered on Vero cells . For CHIKV , genome copy number was determined by extracting viral RNA from the supernatant at each time point using the TRIzol reagent and performing quantitative RT-PCR ( qRT-PCR ) using the TaqMan RNA-to-Ct kit ( Applied Biosystems ) . Ct values were determined in duplicate based on amplification of nsp4 transcripts using forward ( 5′-TCACTCCCTGCTGGACTTGATAGA-3′ ) and reverse ( 5′-TGACGAACAGAGTTAGGAACATACC-3′ ) primers and probe 5′- [6-FAM] AGGTACGCGCTTCAAGTTCGGCG-3′ as previously published [33] , [67] . To determine genome copy number for SINV , viral RNA was extracted in the same manner and quantitative PCR was performed based on amplification of nsp3 transcripts using forward ( 5′-AAAACGCCTACCATGCAGTG-3′ ) and reverse ( 5′-TTTTCCGGCTGCGTAAATGC-3′ ) primers and the SYBR green PCR master mix ( Applied Biosystems ) . Standard curves were performed in each run using samples of in vitro transcribed CHIKV or SINV RNA . In vitro transcribed RNA was transfected in BHK-21 cells in duplicate , as described above or at 28°C , including RNA from a construct in which the polymerase active site ( GDD ) was replaced with GNN by site-directed mutagenesis to abrogate replication and alongside a mock transfection where no RNA was added . Transfections in C6/36 and U4 . 4 cells were modified by pulsing with 250 V , 50 µF , and 550 Ω . Forty-eight hours post-transfection , supernatant containing progeny virus was collected . Cells were washed twice in PBS and RNA was TRIzol ( Invitrogen ) extracted , quantified and diluted to the same concentration . Samples were prepared in NorthernMax formaldehyde loading dye ( Ambion ) with 1 µl of ethidium bromide , heated to 65°C for 10 minutes , then separated on a 1 . 2% LE agarose ( Lonza ) gel containing 1× morpholinepropanesulfonic acid ( MOPS ) running buffer ( Ambion ) and 6 . 7% formaldehyde . RNA was transferred onto nitrocellulose membrane , cross-linked by ultraviolet irradiation ( UVP ) , and prehybridized at 68°C for 1 hour in ULTRAhyb ultrasensitive hybridization buffer ( Ambion ) . A plasmid used for the expression of CHIKV RNA probes corresponding to the 3′ portion of the E2 glycoprotein was generated by first amplifying the region of the CHIKV genome from 8703 ( 5′-GAAGCGACAGACGGGACG-3′ ) to 9266 ( 5′-GTTACATTTGCCAGCGGAA-3′ ) by PCR and subsequently TOPO-TA cloning the PCR product into the pCRTOPO-II vector . RNA probes complementary to positive strand RNA were labeled with 32P using the MAXIscript SP6 In Vitro Transcription Kit ( Ambion ) , unincorporated nucleotides were removed using illustra MicroSpin S200 HR columns ( GE healthcare ) , and probe was hybridized to the membrane overnight at 68°C . Membranes were washed several times at 68°C with 0 . 1× SSC with 0 . 1% SDS , then imaged using Amersham Hyperfilm MP autoradiography film ( GE Healthcare ) . Quantification was done using ImageJ ( http://rsbweb . nih . gov/ij ) . C57BL/6 mice ( Janvier ) or CD-1 mice ( Charles River ) were housed according to Institut Pasteur guidelines in biosafety level 3 isolators , with the approved experimental protocol #10 . 620 , reviewed by the Institut Pasteur ethics committee under dossier #CETEA 2013-0021 . At 8-days old , litters of C57BL/6 were inoculated with 200 pfu of wildtype or mutant CHIKV viruses subcutaneously ( n = 4/variant ) . Eight-day old CD-1 litters were inoculated with 100 pfu of wildtype or mutant SINV strains in the same fashion , and monitored for symptoms of hind limb paralysis and survival . In addition , seven days post-infection , CHIKV and SINV-infected mice were sacrificed and brains , thigh muscles , livers and blood were harvested and homogenized in 300 µl of PBS at 30 shakes/second for 2 min ( MM300 Retsch ) . RNA was extracted and viral genome copies were determined by qRT-PCR as described . Principal CHIKV vectors Ae . albopictus Providence ( ALPROV , F8 generation ) from La Reunion and Ae . aegypti Paea ( PAE , a lab colony at Institut Pasteur since 1994 ) from Tahiti , in French Polynesia were fed on artificial bloodmeals containing 106 pfu/ml of virus in PBS-washed rabbit blood [68] . CHIKV wildtype and mutators were fed to both Ae . albopictus and Ae . aegypti , and SINV wildtype and mutator 482G were fed to Ae . aegypti . The blood meals were warmed to 37°C and presented to 10 day-old females in membrane feeders , and engorged mosquitoes were incubated for 7 days . Seven days post infection , mosquitoes were dissected to obtain legs and wings , and saliva was obtained by in vitro transmission assay; in brief , mosquitoes were salivated for 30–45 min by placing the proboscis in a pipette tip containing FBS . Following salivation , bodies were frozen . To confirm ingestion , a sample of engorged mosquitoes was immediately homogenized at time 0 . Samples were homogenized as described for mouse tissues , RNA was extracted , and qRT-PCR was performed . A standard curve was generated using serial dilutions of a CHIKV bloodmeal of known titer . Drosophila melanogaster flies ( strain w1118 ) were reared on standard medium at 25°C . Three- to four-day-old female flies were injected with 50 nL of a virus dilution containing 400 pfu in 10 mM Tris-HCl ( pH 7 . 5 ) using a Drummond nanoject injector as previously described [69] . Fly mortality at day 1 was attributed to damage produced by the injection , and these flies were excluded from further analyses . Mortality was monitored daily for 10 days , and every 3–4 days flies were transferred to fresh vials . In all experiments , 30–60 flies per genotype group were injected . Homogenates of individual flies were titrated on by plaque assay on Vero cells , as described above . All experiments were performed in triplicate unless noted otherwise . Statistics , noted where applied , were performed in Microsoft Excel and GraphPad Prism . P-values>0 . 05 were considered non-significant ( ns ) . We previously described a CHIKV antimutator variant that possessed a single amino acid change from a cysteine to a tyrosine at position 483 ( C483Y ) of the RNA-dependent RNA polymerase nsp4 [25] . Since Coxsackie virus B3 mutator strains are situated in a structurally analogous area , we hypothesized that this position plays important roles in modulating intrinsic CHIKV RdRp fidelity [6] . To address this , we substituted each amino acid at position 483 of the CHIKV full-length infectious clone ( Table 1 ) . After three passages in BHK-21 cells , viruses were Sanger sequenced to determine genetic stability . Of the 19 substitutions , 12 were viable and genetically stable ( Table 1 ) . This high number of viable variants indicates that position 483 has structural plasticity and can tolerate a wider range of substitutions than in previous attempts at generating fidelity variants of other RNA viruses [6] , [19] . Interestingly , unstable viruses did not readily revert to wildtype , but mutated to other variants , including the antimutator form of the protein , 483Y ( Table 1 ) . The only strict biochemical requirement we observed was a necessity for uncharged residues , as all variants with charged residues ( 483D , E , H , K , or R ) were unstable or not recoverable . In addition , we observed a general correlation between hydrophobicity of the substituted amino acid and stability or viability of the variant , where hydrophobic amino acids were preferred . Finally , as a first characterization of virus fitness , we measured the mean size of plaques . Variants 483A , G , L , N , Q , T , and W had significantly smaller plaques than wildtype ( Table 1 ) . Because polymerase fidelity variants have altered intrinsic rates of ( in ) correct nucleotide incorporation , they have often been identified by their relative resistance or sensitivity to nucleoside analog RNA mutagens [14] , [19] , [25] . Therefore , we addressed the sensitivity of all 12 genetically stable variants to ribavirin ( Table 1 and Figure 1A ) . Viruses were grown in the presence of either 200 µM or 400 µM ribavirin , or left untreated . We expect antimutator variants ( such as 483Y ) to demonstrate resistance , and mutator variants to demonstrate sensitivity when compared to wildtype . As previously described , the antimutator 483Y demonstrated significantly higher survival than wildtype ( P<0 . 001 , two-way ANOVA ) as did 483M and 483N ( P<0 . 001 for both , two-way ANOVA ) . Additionally , we identified several mutator candidates that were significantly more sensitive to ribavirin ( 483A , G , W , T , Q; P<0 . 05 for all , two-way ANOVA ) . All ribavirin-sensitive variants presented small-plaque phentoypes , as well as variant 483N ( P<0 . 01 for all , Student's t-test ) . Though these variants presented small plaque phenotypes , virus stocks reached wildtype-like titers , with the exception of 483N and 483Q ( Table 1 ) . As observed previously for picornaviruses [5] , [6] , [14] , the ribavirin-resistant and -sensitive phenotypes of these CHIKV variants suggested altered polymerase fidelity . To address this further in a genetic assay , we estimated the mutation frequencies of each variant that demonstrated significantly altered ribavirin sensitivity at either concentration of ribavirin . Viral RNA from the supernatants of BHK-21 cells was extracted , and an approximately 800 nucleotide fragment of the E1 genome was amplified by RT-PCR and TOPO cloned as previously described [61] . We sequenced approximately 150 individual clones per viral population ( corresponding to an average of 122 , 200 nucleotides ) to calculate the mutation frequencies ( Figure 1B and Table 1 ) . Since previous studies with 483Y required >350 clones per population to distinguish more subtle differences in mutation frequencies [25] , we could only statistically confirm the altered fidelities of three mutator strains ( 483A , G and W; P<0 . 05 , P<0 . 001 , P<0 . 01 , respectively , χ2 test ) ( Figure 1B ) . We excluded variants that did not exhibit significant fidelity differences compared to wildtype ( 483M , N , and Q ) . As a complementary approach , we performed deep sequencing on these same virus populations to characterize the relative diversity in these virus populations . In accordance with the mutation frequency data , the mean variance across the whole genome was significantly lower for the antimutator 483Y variant ( P = 0 . 0006 , Mann-Whitney u test ) and significantly higher for the 483A , G and W mutator variants , compared to wildtype virus ( P<0 . 0001 for all , Mann-Whitney u test; Figure 1C ) . Next , we examined growth of these variants in mammalian cells . As seen previously , the antimutator 483Y presented no significant difference in amount of progeny virus ( Figure 2A ) or number of genome copies ( Figure 2B ) . As observed with Coxsackie virus mutators , CHIKV mutator strains ( 483A , G , and W ) generated the same or more genomes than wildtype virus ( Figure 2B ) , but slightly fewer infectious progeny ( Figure 2A ) . Consequently , these mutator variants have a lower specific infectivity than wildtype in mammalian cells ( Figure 1C ) . This is consistent with previously published results showing that mutator variants make more lethally mutagenized RNA [6] , [22] , [28] , [70] . Recently , low fidelity polymerase mutators of Coxsackie virus and exonuclease activity deficient mutators of coronaviruses were shown to be attenuated in mice [6] , [7] . To determine whether this holds true for alphaviruses , we administered a sublethal infection of either wildtype or 483A , G and W viruses to 8-day old C57BL/6 mice . At 7 days of infection , when titers peak and virus is rapidly cleared thereafter , viral loads were determined in different compartments ( muscle , blood , brain , liver ) . Viral loads were significantly lower for all three mutator strains in each tissue ( Figure 3A ) . Since the in vivo mutation frequencies of mutator strains had not been previously reported , we examined the virus populations in the muscle of the wildtype- or the mutator-infected mouse that presented the median viral load . Although we cannot predict whether selection will act differently on these variants in mice to potentially skew the mutation frequencies , they remained elevated to varying degrees for the mutator strains . Interestingly , higher mutation frequencies in vivo correlated with increased attenuation ( Figure 3B ) . Because arboviruses must cycle through both vertebrate and arthropod hosts , and since mutator strains of other RNA viruses were only examined in mammalian systems [6] , [7] , [19] , [24] , we addressed viral replication in three mosquito cell lines: Ae . albopictus C6/36 cells , Ae . aegypti Aag2 cells and Ae . albopictus U4 . 4 cells . The replication profile for the antimutator 483Y was indistinguishable from wildtype in all conditions . On the other hand , the mutator strains 483A , G , and W presented significantly lower infectious progeny in C6/36 ( P<0 . 001 for all , two-way ANOVA; Figure 4A ) , Aag2 ( P<0 . 05 for 483A and G , two-way ANOVA; Figure 4B ) and U4 . 4 ( P<0 . 05 for 483A and W , P<0 . 01 for 483G , two-way ANOVA; Figure 4C ) cells . Unexpectedly , we observed unprecedented reduction in genomic RNA released into the supernatant in all three mosquito cell types ( Figure 4D–F ) . These results are discordant with the existing literature that found mutator polymerases synthesize RNA at faster rates than wildtype [6] , in which case decreases in virus titer resulted directly from the increased mutational burden . Here , the reduced viral titers obtained in mosquito cells seem to result from a host-specific replication defect , rather than the effect of lethal mutation . To further distinguish between these two effects , we examined whether mutation frequencies differed in mosquito versus mammalian cells , comparing wildtype CHIKV to the mutator strains . It is important to note that because mutator strains replicate so poorly in mosquito cells , these strains may present artificially low mutation frequencies . Unfortunately , it is not possible to uncouple replication from mutation frequency in this model . Nevertheless , the mutation frequencies of all viruses , including wildtype ) , were lower in C6/36 cells ( Figure 5 ) than in BHK-21 cells ( Figure 1B ) . Furthermore , the significant differences that existed between mutators and wildtype in mammalian cells were negated in mosquito cells , as evidenced by molecular clone sequencing ( Figure 5A ) and whole-genome deep sequencing ( Figure 5B ) . We thus hypothesized that the negative fitness cost of mutator polymerases in mosquito cells is more closely linked to replication defects . To further confirm this , we generated genetically homogenous in vitro transcribed RNA corresponding to each variant , which do not present the differences in mutation frequencies of virus stocks generated in cell culture . Following transfection of mammalian BHK-21 cells , there were no significant differences in RNA synthesis ( Figure 6A ) or production of infectious virus ( Figure 6C ) ; however , in mosquito C6/36 cells , there was a very marked defect in replication for the mutator variants , compared to wildtype virus or the antimutator 483Y strain ( Figure 6B ) , that correlated with the significant reduction in progeny ( P<0 . 01 for all mutators , one-way ANOVA; Figure 6D ) . Similarly , no detectable infectious progeny was produced following transfection of U4 . 4 cells with the mutator variants ( Figure 6E ) , further confirming the replication defect observed during infection of cells with virus stocks . To exclude the possibility that this replication defect is the result of temperature-sensitivity rather than host-specificity , we performed infections in mammalian BHK-21 cells at 28°C ( mosquito cell temperature ) . We observed no difference in the growth of any variant compared to wildtype ( Figure 6F ) . In addition , we transfected mammalian cells grown at 28°C , and saw no difference in subgenomic RNA synthesis , indicating that the reduced polymerase processivity of mutators in mosquito cells is not due to reduced temperature ( Figure 6B and 6G ) . Finally , we determined whether lower temperature could be responsible for the reduced mutation frequencies we observed in mosquito cells . In mammalian cells at 28°C , mutator 483G makes significantly more mutations than in mosquito cells at 28°C ( P<0 . 05 , χ2 test; Figure 6H ) . In contrast to what we observed in mosquito cells , mutator 483G also made significantly more mutations than WT ( P<0 . 05 , χ2 test; Figure 6H ) . These data indicate that lower temperature is responsible for neither the replication defects nor the reductions in mutation frequencies we observed in mosquito cells . Since host-specific replication defects were observed in mosquito cell culture , we hypothesized that these variants would be even more attenuated in mosquitoes than in mice . We orally infected both Aedes species CHIKV hosts ( Ae . albopictus and Ae . aegypti ) with a blood meal containing either wildtype or the 483A , G and W mutators . Seven days after infection , when CHIKV has reached peak titers , we quantified viral loads in bodies ( infection ) , legs and wings ( dissemination ) and saliva ( transmission ) of individual mosquitoes ( Figure 7 ) . Surprisingly , no significant defect was observed in either Aedes species for any of the variants . To address the possibility that the fitness cost of defective replication , observed in mosquito cell culture , would favor the reversion of these mutant polymerases to wildtype , we deep sequenced virus from the body of an individual mosquito that presented the median titer from each group . Indeed , reversion to wildtype ( or other replication competent variants , such as 483T or 483V ) occurred in 483A ( 81% ) , 482G ( 93% ) and 483W ( 39% ) . Whether position 483 changed to wildtype depended on the genetic distance of the mutated codon from wildtype: for example , W ( TGG ) reverted completely to WT ( TGT ) , while A ( GCT ) reverted predominantly to a combination of V ( 66%; GTT ) and T ( ACT; 13% ) . Interestingly , when we examined higher passages ( passage 3 ) of mutators in C6/36 cell culture , we also observed varying levels of reversion ( ranging from less than 1% to as much as 50% ) , highlighting the strong selective pressure acting against this replication defect . After confirming that polymerase position 483 plays an important role in modulating fidelity in CHIKV , we examined if this residue is a universal fidelity determinant among the alphaviruses . Indeed , this region of the nsp4 gene containing a cysteine is conserved across the alphavirus family ( Figure 8A ) . Thus , we generated the analogous fidelity variants ( 482A , G , W ) in the well-studied , distantly related alphavirus Sindbis virus ( SINV ) . Genetically stable mutants ( 482A and G ) were screened for changes in ribavirin sensitivity . Both showed significantly higher sensitivity than wildtype SINV ( for 482A , at least P<0 . 01 , for 482G , P<0 . 05 , two-way ANOVA; Figure 8B ) . Moreover , the mutation frequencies determined by molecular clone sequencing confirmed the mutator phenotypes suggested by ribavirin screening ( Figure 8C ) : in comparison to wildtype that presented 3 . 3 mutations per 10 , 000 nucleotides , 482A presented 6 . 0 , and 482G presented 6 . 9 ( P<0 . 05 , χ2 test ) . This confirms that this conserved residue is a general fidelity determinant for the alphaviruses . We next addressed whether replication defects also existed for these SINV mutators . In mammalian BHK-21 cells , mutator variants produce near wildtype-like titers of infectious particles ( Figure 8D ) , and the same amounts of extracellular RNA genomes ( Figure 8E ) . Importantly , as was observed for CHIKV strains , the SINV mutators presented more significant drops in virus titers in mosquito C6/36 cells ( P<0 . 001 , two-way ANOVA; Figure 8F ) , that correlated with a significant decrease in extracellular RNA genomes ( P<0 . 001 , two-way ANOVA; Figure 8G ) . Given the similarity of in vitro , host-specific phenotypes of CHIKV and SINV mutators , we hypothesized that SINV mutators would behave as CHIKV mutators in vivo ( exhibiting attenuation in a mouse model and reversion in mosquitoes ) . We inoculated 8-day old mice with wildtype and 482G SINV strains , and observed significantly higher survival in mice infected with the mutator ( 91% compared to 50% for the wildtype , P = 0 . 0474; Figure 9A ) . In addition , only 36% of mice inoculated with 482G exhibited complete hind limb paralysis , compared to 100% of mice infected with wildtype SINV ( P<0 . 0001 , χ2 test; Figure 9A ) . Interestingly , this reduced paralysis correlated with significantly lower titers in the brain at day 7 post-infection ( P<0 . 05 , Student's t-test ) , confirming the attenuation of mutator strains in mammalian models in yet another virus ( Figure 9B ) . We next examined the in vivo phenotype of SINV mutator 482G in Ae . aegypti mosquitoes . As expected , we observed reversion of position 482G to wildtype , and therefore , no differences in titers in the mosquito host ( Figure 9C–E ) . Since SINV has a broader host range than CHIKV , we examined whether the replication defect was mosquito cell-specific , or more general to insects , by infecting Drosophila S2 cells . Interestingly , the mutator strains were replication competent , generating virus titers ( Figure 10A ) and RNA genome copies ( Figure 10B ) at levels comparable to wildtype virus . Finally , we injected Drosophila with SINV wildtype and 482G mutator and followed the kinetics of infection by titering virus in flies for seven days post-infection . In contrast to CHIKV and SINV mutators in mosquitoes , when Drosophila flies were infected with wildtype and mutator strains of SINV , mutator 482G presented significantly lower titers than wildtype on day 3 and 5 ( P<0 . 01 , Student's t-test; Figure 10C ) . Sequencing of virus from 482G-infected flies at day 3 and 5 confirmed that no reversion had occurred . These results indicate that in principle , mutators can be attenuated in insects . Previous work on antimutator CHIKV 483Y suggested this residue could be important for determining intrinsic RdRp fidelity [25] . Although there is no crystal structure available for an alphavirus RdRp , structural models predict that position 483 is located in the same area of the RdRp that generated Coxsackie virus RdRp fidelity variants ( Figure S1 ) [6] . By substituting all other amino acids at this position , more mutator variants were isolated than antimutators , consistent with variants obtained for Coxsackie virus and with variants identified by characterizing the mutation frequencies of previously published reverse transcriptase variants for HIV [6] , [32] . To date , all viable RdRp fidelity variants present error rates that remain within the same order of magnitude as their wildtype counterpart [5] , [6] , [14] , [25] , [71] , [72] . Interestingly , although biochemical assays using purified RdRp of picornaviruses indicate that altering fidelity beyond an order of magnitude is enzymatically possible , these viruses are not viable [73] . Together , these studies suggest that within this viable range , wildtype fidelity sits closer to higher fidelity than lower fidelity . This may be further reflection of how RNA viruses are considered to exist close to a maximum threshold of error [74] . In support of this , in conditions where this reversion does not occur , mutator CHIKV variants present more significant fitness defects in vivo ( Figure 3 ) than antimutator virus [25] . A review of the antimutator and mutator RdRp variant literature in virology reveals the following trends: antimutator strains tend to generate less RNA in vitro , but have higher specific infectivity , and have only been reported to lose fitness in vivo or in competition assays ( Figures 2 and 4 , and [16] , [19] , [25] ) ; while mutators generate more RNA in vitro , but of lower specific infectivity , with more prominent fitness defects in mice [6] . Accordingly , CHIKV mutators showed congruous trends , exhibiting no significant replication defects in BHK-21 cells , but showing marked attenuation in the mouse model . Importantly , we showed that the mutator status of these variants ( higher mutation frequencies ) was maintained in the mouse model at the primary site of CHIKV replication and was likely responsible for the observed attenuation ( Figure 3 ) . However , when we examined CHIKV mutators in the invertebrate host , the previous trends for how mutators behave was reversed . First , the differences in mutation frequencies between wildtype and mutator strains became virtually indistinguishable in mosquito cells ( Figure 5 ) , although it is difficult to draw clear-cut conclusions given the reduced replication rate of mutators . For all viruses , mutation frequency was lower in mosquito cells compared to mammalian cells ( Figure 1 ) . The role that these differences may play in arbovirus evolvability and fitness remain contradictory . Our observations corroborate previous observations in alphaviruses that inter-host cycling slows adaptation [33] , [35] , [75]; while flavivirus studies report that diversity is maintained in the mosquito host [38] , [41] , [76] , [77] , [78] , [79] , [80] . Second , and contrary to expectations , we observed a severe replication defect in three different mosquito cell cultures ( Figure 4 ) , which had never been observed for RdRp fidelity variants in mammalian cell culture . The lower titers of infectious progeny were not the result of accumulation of detrimental mutations as was observed for mutators in mammalian hosts; rather , there was a direct defect in genomic RNA synthesis in mosquito cells ( Figure 4 and 6 ) . Interestingly , similar host-specific replication defects were observed for RdRp mutants of West Nile virus ( although it is unclear if these variants have altered fidelity ) . While differences in host temperature do not seem to be the cause , the cellular host factors implicated or missing in these host cell lines remain to be elucidated . Finally , we could not address whether mutators were attenuated in vivo in mosquitoes; sequencing of virus populations from mosquitoes revealed partial or total reversion of the fidelity-altering residues at position 483 . Although one could expect a variant with severe replication defects to be highly attenuated , it is possible that when coupled to a mutator phenotype , reversion would more quickly and favorably occur when the pressure to increase replication remains , as is the case in mosquitoes that are persistently infected . Whether this defect is general to all mutators in mosquitoes , or whether only amino acids A , G , and W at position 483 bear this curiously coupled mutator/replication effect remains to be seen ( since not all variants at this position were defective , as 483Y has no replication defects in mosquito cells [25] ) . Isolation of additional arbovirus mutators mapping to other residues in the polymerase should resolve this issue . Since the cysteine at position 483 is conserved in the alphavirus genus , we obtained additional arbovirus mutators in Sindbis virus . SINV mutators also showed severe replication defects in mosquito cells , and SINV mutator 482G exhibited the same phenotypes we previously observed in CHIKV mutators in both mice and mosquitoes . However , the wider host range of SINV allowed us to test whether these replication defects occur across all insects or if they were mosquito-specific [81] , [82] . In S2 cells , mutators did not present replication defects , allowing us to test , in principle , whether mutators could be attenuated in an insect model ( Drosophila flies ) . Indeed , in the absence of any in vitro replication defect and resulting pressure to revert , the mutator strain was attenuated in fruit flies . Thus , our results confirm that arbovirus mutators can , in principle , be attenuated in insects . Since the isolation of the first antimutator variant of a RNA virus , the growing body of literature shows that either increasing or decreasing replication fidelity has detrimental effects to virus fitness [6] , [7] , [16] , [19] , [24] , [25] . However , how mutation rates and replication capacity are coupled will require more study , and the degree of attenuation resulting from altering these biochemical properties needs to be more carefully evaluated . A future challenge will be to quantitatively link the measurements of mutation frequencies ( average mutations per nucleotide sequenced ) performed in this work to actual mutation rates ( average mutations per nucleotide site per replication ) [2] , [83] and to in vitro biochemical fidelity ( rates of incorporation of correct and incorrect nucleotides in absence of selection ) [6] , [8] , [73] , [84] , [85] , [86] , [87] . It is possible that the higher mutation frequencies measured for these alphavirus mutator strains are partly skewed by their producing more RNA genomes in shorter replication cycles and thus accumulating mutations more rapidly , rather than incorporating more errors per genome during each replication cycle . Indeed , biochemical studies of single-nucleotide incorporation by other mutator polymerases confirm that mutators are both faster enzymes and have higher frequency of mis-incorporation events per replication . In absence of a biochemical assay for alphaviruses , new technologies using microfluidic single-cell analysis of virus strains during single replication cycles should help correlate mutation frequencies , mutation rates , and enzyme fidelity with more confidence . Recent studies have proposed both antimutator and mutator strains as candidates for rationally designed live attenuated vaccines [6] , [7] , [19] , [25] , [71] . Overall , fidelity variants present attenuated titers in vivo that range from one to several orders of magnitude lower than wildtype virus . Whether this degree of attenuation is sufficient to elicit protective immunity without causing disease will require more careful evaluation in more relevant animal models , as virtually all work has been performed in mice using viruses that are often not natural mouse pathogens . In vitro systems and artificial hosts may alter many of the selective pressures to which a virus would be subjected in a natural host [88] , [89] , [90] , [91] . The present study and other work highlight that intrinsic fidelity and the mutant spectrum are labile and subject to stringent and disparate selective pressures in different hosts [34] , [35] , [75] , [76] , [79] , [82] , [92] , [93] . A more comprehensive understanding of the selective pressures in natural hosts is crucial to predicting how viruses will behave in vivo , and essential to evaluating the feasibility of using fidelity variants as vaccines , whether stand-alone or coupled with other , conventional attenuating mutations . Despite the necessity for further research , from a vaccine development perspective these data support that in principle , mutators can be attenuated in a wider range of hosts and may be viable candidates for live-attenuated vaccines .	Chikungunya ( CHIKV ) is a re-emerging mosquito-borne virus that constitutes a major and growing human health burden . Like all RNA viruses , during viral replication CHIKV copies its genome using a polymerase that makes an average of one mistake per replication cycle . Therefore , a single virus generates millions of viral progeny that carry a multitude of distinct mutations in their genomes . In this study , we isolated CHIKV mutators ( strains that make more errors than the wildtype virus ) , to study how higher mutation rates affect fitness in arthropod-borne viruses ( arboviruses ) . CHIKV mutators have reduced virulence in mice and severe replication defects in Aedes mosquito cells . However , these replication defects result in selective pressure for reversion of mutators to a wildtype polymerase in mosquito hosts . To examine how mutators would behave in an insect model in absence of this genetic instability , we isolated mutators of a related virus , Sindbis virus ( SINV ) . SINV mutators had no replication defect in fruit fly ( Drosophila ) cells , and a SINV mutator strain was stable and attenuated in fruit flies . This work shows proof of principle that arbovirus mutators can exhibit attenuation in both mammalian and insect hosts , and may remain a viable vaccine strategy .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "microbial", "mutation", "virology", "emerging", "viral", "diseases", "viral", "enzymes", "biology", "microbiology", "viral", "replication", "viral", "evolution" ]	2014	Alphavirus Mutator Variants Present Host-Specific Defects and Attenuation in Mammalian and Insect Models
The availability of a robust disease model is essential for the development of countermeasures for Middle East respiratory syndrome coronavirus ( MERS-CoV ) . While a rhesus macaque model of MERS-CoV has been established , the lack of uniform , severe disease in this model complicates the analysis of countermeasure studies . Modeling of the interaction between the MERS-CoV spike glycoprotein and its receptor dipeptidyl peptidase 4 predicted comparable interaction energies in common marmosets and humans . The suitability of the marmoset as a MERS-CoV model was tested by inoculation via combined intratracheal , intranasal , oral and ocular routes . Most of the marmosets developed a progressive severe pneumonia leading to euthanasia of some animals . Extensive lesions were evident in the lungs of all animals necropsied at different time points post inoculation . Some animals were also viremic; high viral loads were detected in the lungs of all infected animals , and total RNAseq demonstrated the induction of immune and inflammatory pathways . This is the first description of a severe , partially lethal , disease model of MERS-CoV , and as such will have a major impact on the ability to assess the efficacy of vaccines and treatment strategies as well as allowing more detailed pathogenesis studies . Since the emergence of MERS-CoV in 2012 , researchers have worked to establish animal disease models to study the pathogenesis of this virus and to develop effective countermeasures . With over 836 cases and at least 290 deaths [1] , there has been a rapid increase in the number of MERS-CoV cases as diagnostics are being more widely applied . While dromedary camels are suspected to be involved in zoonotic transmission of MERS-CoV [2]–[5]; and camel as well as horse DPP4 can efficiently facilitate virus entry [6] , the mechanism ( s ) by which most people acquire MERS-CoV is still unclear . Since targeted attempts to prevent zoonotic transmission are currently not feasible and significant morbidity and mortality still occur in individuals with comorbidities , developing effective prophylactic and therapeutic treatment strategies remains a high priority . Although several treatment regimens have been suggested for use in patients based on in vitro MERS-CoV studies or treatment implemented during the SARS-CoV pandemic , only one such treatment has been tested in vivo against MERS-CoV to date [7] . To enable a better evaluation of MERS-CoV treatment and prevention strategies , an animal model more representative of severe human disease is crucial . Assessment of potential treatments in relevant animal disease models is essential prior to conducting clinical trials . Attempts to generate small animal models of disease have so far not been successful . Mice [8] , [9] , hamsters [10] and ferrets [11] do not support replication of MERS-CoV in what appears to be a receptor dependent manner . To overcome this impediment , it was recently demonstrated that adenovirus vectored transduction of human DPP4 into multiple strains of mice , including various knock-outs , results in productive infection and mild pneumonia [12] . Notwithstanding the aforementioned mouse models , the only naturally permissive MERS-CoV disease model that has been described to date is the rhesus macaque model [13]–[15] . In the rhesus macaque , MERS-CoV causes a transient infection of the lower respiratory tract resulting in mild to moderate clinical disease . While this model is useful , it seems to recapitulate the mild disease observed in some human cases , rather than the more severe or even lethal disease observed in many human cases . Moreover , although this model has been used to assess treatment strategies [7] , [16] , it can be difficult as an evaluative model for therapeutics as clinical signs are mild , the duration of illness is relatively short and virus replication is limited . Dipeptidyl peptidase 4 ( DPP4 , also known as CD26 ) was recently shown to be the cellular receptor for MERS-CoV [17] , and the interaction between the MERS-CoV spike protein and DPP4 was subsequently determined by co-crystallography studies [18] , [19] . Although DPP4 is a relatively conserved protein in general , recent data have shown that receptor specificity is likely a major factor in the species tropism of MERS-CoV [11] , [20] . This suggests that mapping the spike binding region of DPP4 of a species of interest before performing experimental inoculations of animals could provide a more rational approach to identifying MERS-CoV susceptible animal models . Here , we modeled the interaction of the common marmoset DPP4 with the MERS-CoV spike protein and show that no differences exist compared to human DPP4 at the site of interaction . Subsequent inoculation of common marmosets ( Callithrix jacchus ) resulted in severe , even lethal , respiratory disease in inoculated animals , with widespread , coalescing bronchointerstitial pneumonia and high viral loads in the lungs of all animals . Variations in the DPP4 receptor appear to play a major role in the ability of MERS-CoV to infect certain animal species . To predict the ability of MERS-CoV spike protein to bind to marmoset DPP4 , analyses were performed using human DPP4 ( known to bind MERS-CoV spike glycoprotein ) [17] and ferret DPP4 ( unable to bind MERS-CoV spike glycoprotein ) [11] . Comparison of the amino acid alignments of human , ferret and marmoset DPP4 revealed that marmoset DPP4 is 96 . 4% identical to human DPP4 , while ferret DPP4 is 87 . 4% identical to human and 87 . 5% identical to marmoset DPP4 . Recently , the 14 amino acids in human DPP4 that facilitate binding to the receptor-binding domain ( RBD ) of the MERS-CoV spike glycoprotein were identified by co-crystallography studies [18] , [19] . No amino acid differences between human , rhesus macaque and common marmoset were identified within the DPP4 region interacting with the MERS-CoV RBD , whereas between human/rhesus/marmoset and ferret and mouse nine or six amino acid residues were different within this region , respectively ( Fig . 1A ) . The 100% identity of the 14 amino acid residues between human and marmoset DPP4 in the interaction regions indicates that MERS-CoV RBD should bind to marmoset DPP4 . This was further supported by modeling the binding potential between human , ferret and marmoset DPP4 with the MERS-CoV RBD . No significant differences in binding energy were observed between the human ( −981 ) and marmoset ( −978 ) DPP4 – MERS-CoV RBD , whereas the binding energies between ferret DPP4 and the MERS-CoV RBD was significantly higher ( −601 ) . A marmoset DPP4 homology model was built using the human DPP4 structure ( PDB ID: 4KR0 , Chain A ) . This model demonstrated that all amino acid differences between human and marmoset DPP4 were located away from the binding region of DPP4 with the S1 portion of the spike glycoprotein; the nearest residues to this interface that differ are Arg343 , Ile193 and Val279 which are 14A , 10A and 13A away , respectively , from the nearest atom in MERS-S ( Fig . 1B ) . Taken together this suggests that marmoset DPP4 would facilitate binding with the MERS-CoV RBD and that marmosets would be susceptible to MERS-CoV infection . Common marmosets inoculated with MERS-CoV ( hCoV-EMC/2012 [21] ) via the intratracheal , intranasal , oral and ocular routes developed signs of respiratory disease that ranged from moderate to severe ( Table S1 ) . Animals were assigned for scheduled necropsies on 3 ( CM1 , CM2 and CM3 ) and 6 ( CM4 , CM5 and CM6 ) dpi prior to the start of the experiment; three additional animals ( CM7 , CM8 and CM9 ) were monitored for survival ( Fig . 2A ) . Starting on 1 dpi six of nine animals showed increased respiration rates while all animals had increased respiration rates from 2 dpi on . In some of the animals open mouth and/or labored breathing was evident from 3 dpi onwards . All animals showed loss of appetite and decreased levels of activity . Clinical scores were assigned using an established scoring sheet for respiratory disease in common marmosets ( Table S2 ) . Based on this system , peak clinical scores were observed between 4 and 6 dpi with scores returning to baseline by 13 dpi in the two remaining animals ( Fig . 2B ) . Animals exhibited decreased temperatures starting on 3 dpi , returning to a normal range on 9 dpi ( Fig . S1A ) . On 4 dpi , two animals ( CM5 and CM9 ) were euthanized due to the severity of disease as determined by clinical score ( Fig . 2A ) , which included increased respiration rate , open mouth breathing and failure to move following prompting . One of these animals also exhibited the presence of frothy hemorrhagic discharge from its mouth . While most animals showed a decrease in body temperature by 3 dpi; the two animals that were euthanized on 4 dpi were severely hypothermic indicative of the onset of shock at the terminal stages of disease ( Fig . S1A ) . Clinically significant alterations in blood cell counts and chemistry were not noted in any of the animals ( Fig . S1B–J ) in contrast to the rhesus macaque model . As serial blood samples could not be collected on a daily basis on marmosets , due to the small size of the animals , alterations were likely missed as a result of study design . Elevations in liver enzymes ( Fig . S1E–G ) and measures of kidney function ( Fig . S1I , J ) occurred on 3 and 4 dpi , respectively; however , these changes were not outside of the normal range . The animals euthanized on 3 , 4 and 6 dpi all showed hypoproteinemia consistent with high protein pulmonary effusions resulting from alveolar edema ( Fig . S1K , L ) . To monitor for signs of pneumonia animals underwent dorsal-ventral and lateral x-rays during examinations on 0 , 1 , 3 , 6 , 9 , 13 and 20 dpi . All radiographs were normal prior to inoculation on 0 dpi , while on 1 dpi four of nine animals showed mild to marked diffuse interstitial infiltration in the lower lung lobes ( Table S3 ) . On 3 dpi interstitial infiltration of varying severity ( mild to severe ) was noted in all animals bilaterally in the lower lobes . The two animals that were euthanized on 4 dpi showed severe interstitial infiltration with partial to complete congestion of the bronchioles ( Fig . 3 ) . The other animals were not radiographed at this time point as this was not a scheduled examination time point . On 6 dpi the remaining animals had mild to severe interstitial infiltration in the lower lobes . One animal ( CM6 ) had severe infiltration in all lobes with congested bronchioles . By 9 dpi the two remaining animals showed improvement , with lessening infiltration , which was resolved by 13 dpi . Three animals were euthanized on 3 dpi and necropsies were performed . All three animals had relatively comparable gross lesions ( Fig . S2A ) , especially in the lower lobes ( mean affected area 11% of upper lung lobes , 54% of lower lung lobes ) , with multifocal consolidation and dark red discoloration , consistent with interstitial pneumonia . The two animals euthanized due to the severity of disease on 4 dpi showed extensive severe lesions throughout the lungs ( mean affected area 36% of upper lung lobes , 83% of lower lung lobes ) ( Fig . 3 , Fig . S2A ) . In addition , the lungs were firm , failed to collapse and were fluid filled . The lungs from animal CM5 were three times the lung weight to body weight ratio of the lungs from the other sampled animals , which were comparable to CM9 ( Fig . S2B ) . Animals necropsied on 6 dpi revealed lungs comparable to 4 dpi with lesions throughout the lungs ( mean affected area 12% of upper lung lobes , 85% of lower lung lobes ) ( Fig . S2A ) . Lungs from these animals were firm and fluid filled , with fluid leaking from the tissue and had a lung to body weight ratio twice that of 3 dpi animals ( Fig . S2B ) . No other gross pathological changes were observed at necropsy . Lungs from marmosets necropsied at 3 and 4 dpi all showed multifocal to coalescing , moderate to marked acute bronchointerstitial pneumonia ( Fig . 4A , C ) . The pneumonia tended to be centered on small caliber and terminal bronchioles and extended into the adjacent pulmonary parenchyma . Viral antigen was exclusively associated and located throughout regions that contained pathological changes ( Fig . 4B , D , F , H , J ) . The bronchiolar epithelium was frequently eroded , leaving attenuated bronchiolar epithelial cells . Affected bronchioles were filled with small to moderate numbers of macrophages and neutrophils , and occasionally small amounts of fibrin and edema ( Fig . 4E , G ) . The adjacent alveolar interstitium was thickened with congestion , edema and fibrin and moderate numbers of macrophages and neutrophils ( Fig . 4E , Table S4 ) . Alveolar spaces contained moderate to marked numbers of pulmonary macrophages and neutrophils; multifocally there was pulmonary edema , fibrin , and less frequently hemorrhage ( Fig . 4E ) . There were also rare multinucleate syncytia within alveolar spaces ( Fig . 4E ) . At 6 dpi multifocal to coalescing areas of acute pneumonia were still visible; however , there were also extensive areas of type II pneumocyte hyperplasia ( Fig . 4G ) and consolidation of pulmonary fibrin resulting in multifocal hyaline membranes ( Fig . 4I ) . These changes are consistent with a transition from acute to a more chronic reparative stage of pneumonia . Regions that were undergoing tissue remodeling showed evidence of clearing of viral antigen ( Fig . 4 H ) . The distribution of DPP4 in marmoset lungs included type I pneumocytes ( Fig . 5A ) as well as bronchiolar epithelial cells and smooth muscle cells . Consistent with this location of DPP4 , a two-color fluorescent staining for cytokeratin and viral antigen , as well as in situ hybridization to detect viral RNA , identified type I pneumocytes and alveolar macrophages as the primary cell type for MERS- CoV replication ( Fig . 5 B , C ) . One of the surviving animals ( CM7 ) had to be euthanized prior to the scheduled end of the study ( 48 dpi ) and was found to have severe aspiration pneumonia . The lungs from the other surviving animal ( CM8 ) appeared normal at necropsy 55 dpi . Viral antigen was not detected by IHC in either animal indicating that these animals had resolved MERS-CoV infection . All other lesions noted in the remaining tissues , such as interstitial nephritis in the kidney and the presence of giant cells in the adrenal gland , consistent with extramedullary hematopoiesis , were not considered to be clinically significant since they are typical , incidental findings in common marmosets [22] . On 1 , 3 , 6 , 9 , 13 and 20 dpi , nasal and oropharyngeal swabs were obtained from all remaining animals; on 4 dpi before euthanasia , swabs were also obtained from CM5 and CM9 . All but one of the nasal swabs collected on 1 dpi were positive for the presence of viral RNA by qRT-PCR ( Fig . 6A ) . By 3 dpi , viral loads in nasal and oropharyngeal swabs were lower than on 1 dpi . However , swabs collected from CM5 and CM9 at the time of euthanasia at 4 dpi contained the highest viral loads . In the surviving animals , viral loads in the swabs decreased over time and all swabs were negative by 20 dpi ( Fig . 6A ) . A terminal blood sample was collected from animals CM1–CM6 and CM9 . CM5 , euthanized on 4 dpi , and CM4 , euthanized on 6 dpi , were viremic ( Fig . 6B ) , although viral loads in these blood samples were very low and virus isolation attempts on blood samples were not successful ( data not shown ) . Upon necropsy of the animals on 3 , 4 and 6 dpi , tissue samples were collected and subsequently analyzed for the presence of viral RNA by qRT-PCR . High viral loads were detected mainly in respiratory tissues ( Fig . 6C ) . Viral loads in lung tissue samples obtained from the animals on 3 dpi were very high , reaching up to 107 TCID50 eq . /gram . Viral loads in the lungs did not decrease between 3 and 6 dpi . Viral RNA was detected in all tested tissues , albeit not in every animal ( Fig . 6D ) . Viral RNA was not detected in the lungs , or other tissues , from the two surviving animals that were euthanized on 48 and 55 dpi; however , both of these animals seroconverted indicating that they had been exposed to MERS-CoV . Virus isolation attempts were performed on nasal mucosa , trachea , lung and kidney samples . In all animals except CM6 ( 6 dpi ) , virus could be isolated from one or more lung lobes ( Table S5 ) . Virus was isolated from all trachea samples collected on 4 and 6 dpi . Virus was not isolated from any of the kidney samples ( Table 5 ) . To characterize transcriptional changes as a result of MERS-CoV infection , we performed whole transcriptional sequencing on RNA samples extracted from the right lower lobe of marmoset lungs collected at necropsy . Samples were collected from areas of the lung that contained gross lesions . These were compared to the total RNAseq data from the lung of a single uninfected marmoset obtained through the Non-Human Primate Reference Transcriptome Resource ( http://www . nhprtr . org ) [23] . We reasoned that molecules induced by infection that play a role in respiratory pathogenesis would at least double in abundance over the course of infection . Therefore we identified mean differentially expressed ( DE ) transcripts at each time point with at least 2-fold change compared to the uninfected control . Molecular profiles were functionally similar over the course of sampling ( Fig . 7A ) ; however , there were some alterations in functional category enrichment and magnitude of expression relative to the uninfected control . Throughout infection , pathways associated with chemotaxis and cell migration , cell cycle progression , cell proliferation , and fibrogenesis were increasingly activated relative to the uninfected control lung , with higher expression and greater pathway enrichment occurring at 4 and 6 dpi than at 3 dpi . We also observed activation of pathways associated with inflammation , vascularization , endothelial activation , proliferation of smooth muscle cells , and tissue repair , indicating that MERS-CoV infection induces tissue differentiation in marmosets consistent with development of pulmonary fibrosis ( Fig . 7B , Table S6 ) . Despite the observed severe lung pathology , robust antiviral transcriptional responses were induced in the marmoset lungs . Up regulation of innate immune genes such as pattern recognition receptors , interferon-stimulated genes , inflammatory cytokines and signaling molecules , as well as adaptive immune responses such as lymphocyte signaling , proliferation , and differentiation , immunoglobulin production , antigen presentation , and T cell costimulation were observed . We were able to detect many common serum cytokine transcripts using RNAseq , which unfortunately cannot be measured in immunoassays as these are not available for common marmosets . Notably , we detected down regulation of interferon gamma ( IFNγ ) and its receptor ( IFNGR1 ) at multiple time points , and no expression of interferon beta ( IFNβ ) , consistent with reports that MERS-CoV infection attenuates these cytokines [24]–[29] . We did observe up regulation of the type I interferon receptor ( IFNAR1 , IFNAR2 ) and interferon stimulated genes ( ISGs ) induced by type I interferon responses , in addition to the type III interferon receptor ( IFNLR1 ) , ( Table S7 ) . We also observed significant down regulation of IL-8 and IL-18 at all time points , and up regulation of IL-27 for all animals . IL-1β was up regulated in the 4 dpi samples , but down regulated at 3 and 6 dpi . As cytokine transcripts generally have a short half-life , in some cases we were not able to detect DE cytokine transcripts , but were able to detect changes in their receptors or molecules associated with cytokine signaling . We observed up regulation of the receptors for IL-2 ( IL2RB ) , IL-4 ( IL4R ) , IL-6 ( IL-6R ) , IL-17 ( IL17RC , IL17RE ) , IL-22 ( IL22RA1 ) , and IL-27 ( IL27RA ) , and down regulation of receptors for IL-1 ( IL1R1 , IL1RAP ) , IL-12 ( IL12RB2 ) , IL-18 ( IL18R1 ) , and IL-20 ( IL20RA ) ( Table S7 ) . Using the Molecule Activity Prediction ( MAP ) tool in IPA , we were able to generate a network of these molecules showing the predicted activity of cytokines not identified in the dataset at 4 dpi ( Fig . 7C ) . The MAP tool utilizes relationships with molecules neighboring the DE transcripts in this dataset to predict transcriptional behavior . These relationships are based on published work curated in the IPA knowledgebase , and provide insight into the dynamics of all molecules within a given network or pathway generated using the primary transcriptional data from this experiment . While type I interferons , IL-2 , IL-4 , and IL-6 were predicted to be induced , type II interferons and the pro-inflammatory cytokines IL-1 and TNFα were predicted to be at least partially inhibited . While transcriptional induction of proinflammatory cytokines such as IL-1 and TNFα was expected , especially given the extent of inflammation in the lungs , the time points for sampling may have been too late to catch TNFα induction as it is one of the first cytokines induced during inflammation . Its induction is supported by downstream effectors such as NFκB , RelA , and RelB and many accessory proteins that were unregulated at the time points investigated . Moreover , as TNFα signaling tends to be down regulated rapidly , the up regulation of SOCS3 , which is induced by TNFα as part of a negative feedback loop that ultimately results in suppression of TNFα and IL-1 , was observed at all time points . In MERS-CoV-infected common marmosets , clinical disease was more severe than in the rhesus macaque , was of longer duration and resulted in euthanasia of some animals . Viral loads in the lungs were up to 1000 times higher than those in the rhesus macaque lungs ( mean viral load in the lungs 1 . 2×103 in rhesus vs . 1 . 5×106 in marmosets on 3 dpi; 6 . 8×102 in rhesus vs . 2 . 1×106 in marmosets on 6 dpi ) [15] . Two of the six animals that were not euthanized at the scheduled 3 dpi necropsy had to be euthanized due to severity of disease , making this the first lethal MERS-CoV animal model . In both the marmoset and rhesus macaque [15] models , viral replication occurred predominantly in the lower respiratory tract; however , in marmosets MERS-CoV RNA was also detected in the blood . This is suggestive of a more systemic dissemination that was corroborated by the detection of viral RNA in nearly all tested tissues in all infected animals . Taken together , the data from the common marmoset model suggest that this model more closely recapitulates severe , even lethal , human disease caused by MERS-CoV . This differs from the rhesus macaque model for MERS-CoV [13]–[15] , which more closely resembles mild to moderate human disease . While viral RNA was detected in kidney samples from five out of seven marmosets , no histological abnormalities were observed , virus could not be isolated and viral antigen was not detected by immunohistochemistry . Elevations in creatinine and blood urea nitrogen levels were noted on 4 dpi in the two animals that were euthanized , suggesting some degree of kidney involvement during infection; however , the lack of evidence of virus replication in this tissue suggests that this is not a direct viral effect . The acute renal failure in some patients [21] may be a secondary effect of ARDS [30] or other comorbidities and not primarily the result of direct MERS-CoV damage to the kidneys . While evidence of pulmonary fibrosis was not yet observed histologically at the time points investigated , transcriptional evidence of the onset of fibrosis was extensive . Previous studies of severe acute respiratory syndrome coronavirus ( SARS-CoV ) have demonstrated that pulmonary fibrosis was a major mechanism of disease progression , and that lung inflammation caused by infection induced fibrogenic transcriptional programs [31]–[33] . In the marmosets , the role of fibrosis is unknown and needs to be further clarified and supported by data from human cases . We anticipate that studies using larger groups of marmosets , additional control animals , and samples collected at later time points post-infection will confirm the histologic course of fibrosis progression , as well as the transcriptional events underlying these pathogenic processes . Currently , the only small animal MERS-CoV challenge model available , requires transduction of animals with an adenovirus vector expressing human DPP4 [12] . Although this is a very useful model , MERS-CoV infection in this model is highly dependent on the transduction of cells and level of DPP4 expression from the adenovirus vector and thus does not necessarily reflect the natural disease process . Therefore , therapeutics indicated to inhibit MERS-CoV in in vitro studies likely need to be tested in one of the two described nonhuman primate models . As such , marmosets should likely serve as the animal of choice for future therapeutic studies where possible . Not only does the more severe , and potentially lethal disease set a higher bar for protection , it would also allow a greater differentiation to be made between disease in untreated animals versus treated animals , currently a limitation of the rhesus macaque model [7] . The marmoset model also allows the evaluation of intervention strategies at later time points as the disease process in the rhesus model is rapid and quite transient . However , late treatment that targets the virus , as with many countermeasures , is unlikely to be successful once significant lung damage has already occurred as was observed by the lack of success of very late treatment with ribavirin and interferon in human MERS-CoV cases [34] . To enable treatment of patients prior to severe lung injury , future transcriptional studies may yield early indicators of disease progression that can be used as diagnostic or prognostic tests to improve clinical management . The development of the more severe marmoset model will ensure a better pre-clinical analysis of treatments prior to proceeding to clinical trials in humans . As such , this new MERS-CoV disease model is a significant contribution to reducing the impact of MERS-CoV on global public health . All animal experiments were approved by the Institutional Animal Care and Use Committee ( IACUC ) of the Rocky Mountain Laboratories ( RML ) , and performed following the guidelines of the Association for Assessment and Accreditation of Laboratory Animal Care , International ( AAALAC ) by certified staff in an AAALAC-approved facility , following the guidelines and basic principles in the United States Public Health Service Policy on Humane Care and Use of Laboratory Animals and the Guide for the Care and Use of Laboratory Animals . All procedures were carried out under a combination of Ketamine and isoflurane anesthesia by trained personnel under the supervision of veterinary staff and all efforts were made to provide for the welfare and to minimize any suffering in accordance with the “Weatherall report for the use of non-human primates” . Animals were housed in adjoining individual primate cages allowing social interactions , under controlled conditions of humidity , temperature and light ( 12-hour light/12-hour dark cycles ) . Food and water were available ad libitum . Animals were monitored twice daily ( pre- and post-challenge ) and fed a combination of commercial New World monkey chow , rice cereal supplemented with calcium , ZuPreem marmoset , wax worms/larvae and fruit twice daily by trained personnel . Endpoint criteria , as specified by the RML IACUC approved score parameters , were used to determine when animals should be humanely euthanized . Animals were euthanized by exsanguination under deep isoflurane anesthesia . The work with infectious HCoV-EMC/2012 was approved under BSL3 conditions by the Institutional Biosafety Committee ( IBC ) . Sample inactivation was performed according to standard operating procedures approved by the IBC for removal of specimens from high containment . In order to establish a more severe animal model for MERS-CoV , the interaction of the common marmoset DPP4 with the MERS-CoV spike protein was modeled . Subsequently , experimental inoculation of common marmosets was performed to determine whether they would serve as an improved disease model . Nine male common marmosets ( Callithrix jacchus; 2–6 years old ) were randomly assigned a number ( CM1–CM9 ) and subsequently inoculated with MERS-CoV ( strain HCoV-EMC/2012 ) intranasally with 100 µl in each nare , 500 µl orally , 500 µl intratracheally and 50 µl in each eye with DMEM containing 4×106 TCID50/ml ( total dose 5 . 2×106 TCID50 ) . Necropsies of three animals were scheduled on 3 dpi ( CM1–CM3 ) and 6 dpi ( CM4-6 ) . The three remaining animals ( CM7–CM9 ) were not scheduled for euthanasia , but were used to study survival and seroconversion upon inoculation of animals with MERS-CoV ( Fig . 2A ) . The animals were observed twice daily for clinical signs of disease and scored using a clinical scoring system prepared for common marmosets ( Table S2 ) . The in-study euthanasia criteria were established prior to the start of the experiment based on the scoring sheet and euthanasia was indicated at a clinical score of 35 or above ( Table S2 ) . During the course of the study , animals CM5 and CM9 were euthanized on 4 dpi as they reached euthanasia criteria . On 1 , 3 , 6 , 9 , 13 and 20 days post inoculation , clinical exams were performed on anaesthetized animals , x-rays were taken and nasal and oral swabs were collected in 1 ml DMEM with 50 U/ml penicillin and 50 µg/ml streptomycin . Temperature was monitored with IPTT-300 temperature probes ( BMDS ) that were injected interscapularly prior to the start of the experiment . Blood was collected prior to the start of the study and at euthanasia for hematology and blood chemistry analysis . The total white blood cell count , lymphocyte , platelet , reticulocyte , and red blood cell counts , hemoglobin , hematocrit values , mean cell volume , mean corpuscular volume , and mean corpuscular hemoglobin concentrations were determined from EDTA blood with the HemaVet 950FS+ laser-based hematology analyzer ( Drew Scientific ) . Samples of the following tissues were collected: conjunctiva , nasal mucosa , tonsil , mandibular lymph node , salivary gland , trachea , all four lung lobes , mediastinal lymph node , inguinal lymph node , axillary lymph node , mesenteric lymph node heart , liver , spleen , kidney , adrenal gland , pancreas , ileum , colon transversum , urinary bladder , testes , frontal brain , cerebellum and brain stem . Amino acid sequence alignment was generated using the human , ferret and marmoset DPP4s ( accession numbers NP_001926 . 2 , DQ266376 and XM_002749392 respectively ) using CLUSTALW2 [35] . The human DPP4 structure model ( PDB ID: 4KR0 , Chain A ) was used as template to highlight the location of the amino acid differences between the human and marmoset DPP4s . The initial model was built using Nest [36] based on the amino acid alignment and the human DPP4 structure . The resulting structural model was briefly optimized using the TINKER minimization program “minimize . x” with OPLS all-atom force field and L-BFGS quasi-Newton optimization algorithm [37] . For the marmoset and ferret DPP4 , the RBD/DPP4 complex model was generated by merging the RBD domain ( PDB ID: 4KR0 , Chain B ) with the DPP4 model , which was then subjected to the binding energy calculation using an all-atom distance-dependent pairwise statistical potential , DFIRE [38] . The energy difference between the complex and two individual structures - DPP4 and RBD - was taken as the binding energy . HCoV-EMC/2012 [21] was kindly provided by the Department of Viroscience , Erasmus Medical Center , Rotterdam , The Netherlands and propagated once in VeroE6 cells in DMEM ( Sigma ) supplemented with 2% fetal calf serum ( Logan ) , 1 mM L-glutamine ( Lonza ) , 50 U/ml penicillin and 50 µg/ml streptomycin ( Gibco ) ( virus isolation medium ) . VeroE6 and LLC-MK2 cells were maintained in DMEM supplemented with 10% fetal calf serum , 1 mM L-glutamine , 50 U/ml penicillin and 50 µg/ml streptomycin . Histopathology and immunohistochemistry were performed on marmoset tissues . After fixation for 7 days in 10% neutral-buffered formalin and embedding in paraffin , tissue sections were stained with hematoxylin and eosin ( HE ) . To detect HCoV-EMC/2012 antigen , immunohistochemistry was performed using a rabbit polyclonal antiserum against HCoV-EMC/2012 ( 1∶1000 ) as a primary antibody for detection of HCoV-EMC/2012 antigen . Immunohistochemistry for DPP4 ( CD26 ) was performed with a rabbit polyclonal antiserum ( 1∶400 ) ( Abcam ) . To confirm the cell type of infected cells a subset of specimens were stained with an antibody against cytokeratin ( 1∶100 ) ( Dako ) and the rabbit polyclonal against HCoV-EMC/2012 . Fluorescently labeled goat anti-mouse ( AlexaFluor 488 ) goat anti-rabbit ( AlexaFluor 594 ) were used for detection . Tissues from an uninfected control animal was obtained from Primate Biologicals and used to validate all in situ and immunohistochemistry procedures . RNA was extracted from swab and whole blood samples using the QiaAmp Viral RNA kit ( Qiagen ) according to the manufacturer's instructions . Tissues ( 30 mg ) were homogenized in RLT buffer and RNA was extracted using the RNeasy kit ( Qiagen ) according to the manufacturer's instructions . For whole transcriptome sequencing , tissues collected from the right lower lobe of the lung were homogenized in TRIzol reagent and frozen at −80°C . The TRIzol was phase separated and total RNA was further purified from the aqueous phase using the miRNeasy kit ( Qiagen ) according to the manufacturer's instructions . Weighed tissue samples were homogenized in a TissueLyzer II ( Qiagen ) after addition of 1 ml DMEM . Homogenates were centrifuged to pellet cellular debris and 10-fold dilutions of homogenate were made and subsequently inoculated onto VeroE6 and LLC-MK2 cells for virus isolation . After 1 hr , cells were washed once with DMEM and supplemented with virus isolation medium . Cells were scored for cytopathic effect 5 days following infection . For detection of viral RNA , 5 µl RNA was used in a one-step real-time RT-PCR upE assay [39] using the Rotor-Gene probe kit ( Qiagen ) according to instructions of the manufacturer . In each run , standard dilutions of a titered virus stock were run in parallel , to calculate TCID50 equivalents in the samples . Total RNA libraries were constructed using the Illumina TruSeq Stranded Total RNA Preparation Kit ( Illumina ) according to the manufacturer's guide . Input RNA was rRNA reduced using the Ribo-Zero Magnetic Kit ( Human/Mouse/Rat ) , ( Epicentre ) , and rRNA reduction was verified by BioAnalyzer 2100 ( Agilent Technologies ) . Libraries were quality controlled and quantitated using the BioAnalzyer 2100 system and mass measurement using QuBit ( Life Technologies ) . The libraries were clonally amplified on a cluster generation station using Illumina version 2 MiSeq reagents to achieve a target density of approximately 1000 K–1200 K clusters/mm2 on the flow cell . The resulting libraries were then sequenced on an Illumina MiSeq system at single reads of 50 bp , and were processed on the system to generate quality control metrics and FASTQ files with MiSeq Reporter 2 . 3 . 32 . Raw reads were trimmed to 50 base pairs ( bp ) and adapter sequences were removed . We mapped the 50 bp reads to a custom-compiled set of known ribosomal sequences ( human , mouse ) using the short-read aligner software Bowtie v . 1 . 0 . 0 [40] to remove potential rRNA sequences . The remaining unmapped reads corresponding to MERS-CoV genome ( GenBank accession no . JX869059 . 2 ) were also removed using Bowtie . All the remaining reads were mapped to the common marmoset ( Callithrix jacchus ) genome assembly ( Genome assembly: C_jacchus3 . 2 . 1 , GCA_000004665 . 1 ) downloaded from Ensembl ( version 74 ) using STAR v . 2 . 3 . 0 . 1 [41] . Gene level quantification was obtained using HT-seq ( http://www-huber . embl . de/users/anders/HTSeq/doc/overview . html ) with Ensembl annotation ( v74 ) . For the downstream analyses we only retained those transcripts with at least 40 raw read counts in at least one sample . The raw read counts were normalized using edgeR [42] . For the calculation of expression ratios , an offset of 5 was added to the normalized expression levels , i . e . counts per million ( cpm ) values to dump down variations due to low abundances . We identified transcripts with greater than 2-fold change relative to the uninfected control , and used Spotfire DecisionSite ( Tibco ) to perform hierarchical clustering using the unweighted pair group method with arithmetic means ( UPGMA ) method , and generate heatmaps . Functional enrichment analysis was performed using Ingenuity Pathway Analysis ( IPA ) . Raw data were deposited into the Sequence Read Archive ( SRA ) via the Gene Expression Omnibus ( accession number pending ) . GEO Accession #GSE55023 .	The development of vaccines and treatment strategies is aided by robust animal disease models that accurately depict the illness that is observed in humans . Here we describe a new , improved model for MERS-CoV using the common marmoset , whereby the severe , and even lethal , illness that has been observed in many human cases is recapitulated . Prior to the development of this model , the only available animal models for MERS-CoV infection were the rhesus macaque and a mouse model that requires adenovirus-transduced expression of the human version of the protein required for virus entry . The rhesus macaque model more closely mimics the mild to moderate disease observed in some patients—mainly those without significant comorbidities . The increased severity of illness in the common marmoset model is an important advance in the ability to evaluate potential therapeutic agents against MERS-CoV , as discrimination between successfully treated and control animals should be more apparent . In addition , the closer models recapitulate the disease observed in humans , the more likely findings can be eventually translated into use in humans .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "primates", "animal", "models", "animal", "models", "of", "infection", "marmosets", "model", "organisms", "monkeys", "new", "world", "monkeys", "virology", "emerging", "viral", "diseases", "biology", "and", "life", "sciences", "microbiology", "vertebrates", "animals",...	2014	Infection with MERS-CoV Causes Lethal Pneumonia in the Common Marmoset
To navigate their surroundings , cells rely on sensory input that is corrupted by noise . In cells performing chemotaxis , such noise arises from the stochastic binding of signalling molecules at low chemoattractant concentrations . We reveal a fundamental relationship between the speed of chemotactic steering and the strength of directional fluctuations that result from the amplification of noise in a chemical input signal . This relation implies a trade-off between steering that is slow and reliable , and steering that is fast but less reliable . We show that dynamic switching between these two modes of steering can substantially increase the probability to find a target , such as an egg to be found by sperm cells . This decision making confers no advantage in the absence of noise , but is beneficial when chemical signals are detectable , yet characterized by low signal-to-noise ratios . The latter applies at intermediate distances from a target , where signalling molecules are diluted , thus defining a ‘noise zone’ that cells have to cross . Our results explain decision making observed in recent experiments on sea urchin sperm chemotaxis . More generally , our theory demonstrates how decision making enables chemotactic agents to cope with high levels of noise in gradient sensing by dynamically adjusting the persistence length of a biased random walk . Motile cells successfully navigate in external concentration fields of signalling molecules by steering in the direction of local concentration gradients—a process termed chemotaxis . Chemotaxis represents a biological implementation of a gradient-ascent algorithm and is used by bacteria to find food [1] , immune cells to locate infection sites [2] , and sperm cells to follow gradients of chemical cues to find the egg [3 , 4] . The very task of reliably measuring a local concentration gradient with sufficient accuracy is non-trivial at dilute concentrations , since molecular shot noise corrupts concentration measurements [5–8] . To measure a concentration , cells must count individual binding events of signalling molecules , which represents a stochastic Poisson process . Pioneering work on this topic studied the chemotaxis of mobile agents with advanced information processing skills [9] , or even a capacity to compute spatial maps of maximum likelihood of target position [10] . It is an open question how biological cells with limited information processing capability deal with noise during their chemotaxis [11–13] . Here , we present a theory of optimal chemotaxis strategies in the presence of noise , using the framework of Markov decision processes ( MDP ) . First , we compute optimal strategies for the idealized case where cells have perfect knowledge of their distance to the target and their swimming direction . From this , we derive a heuristic for the realistic case where cells only possess a noisy estimate of their positional state . We apply this general approach to chemotaxis along helical swimming paths , which is employed by sperm cells of marine species . Chemotaxis along helical paths represents one of the three fundamental gradient-sensing strategies of biological cells [4] . This strategy , also known as helical klinotaxis , is based on temporal comparison of a concentration signal traced along the swimming path [14–16] . By swimming along a helical path , i . e . circling around a centreline , these cells receive information about the gradient component perpendicular to their direction of net motion . In sperm cells , a chemotactic signalling system processes this information and dynamically adjusts the shape of flagellar bending waves [17] . This feedback loop enables cells to steer in a directed manner , by bending the direction of their helical paths towards the local gradient , see also Fig 1A . Helical swimming represents a stereotypical form of exploratory behaviour , employed by sperm cells and other microswimmers [18] . This strategy is typical for sperm cells from species with external fertilization [16 , 19 , 20] . Helical chemotaxis is qualitatively different from ‘run-and-tumble’ chemotaxis along biased random walks , employed e . g . by swimming bacteria . These bacteria measure only the gradient component parallel to their swimming path [1] , and not the perpendicular component , and thus lack the information required for directed steering responses . The model system of sperm chemotaxis is particularly suited to address optimal navigation in the presence of noise: First , sperm cells have a single objective , to find the egg . In species with external fertilization , evolution presumably optimized the probability to find an egg . Second , recent experiments revealed that sea urchin sperm cells dynamically switch between two different steering modes [16] , thus providing an instance of decision making at the scale of individual cells . To date , the benefit of this decision making is not known . With our theory , we demonstrate a benefit of decision making in sperm chemotaxis , and show that this benefit is directly related to noise in cellular gradient sensing . Our work addresses the intermediate case between the well-understood case of perfect chemotaxis in the absence of noise ( perfectly reliable steering ) , and purely random search strategies that operate in the absence of directed signals ( no steering ) [21–24] . Random search strategies , such as random walks or Lévy walks are relevant far from a target , i . e . outside the spatial range of chemosensation . For such random search problems , ballistic motion maximizes the rate of finding targets if targets are distributed randomly in an unbounded search domain and can be visited only once , whereas Lévy walks maximize this rate when the same target can be revisited [25] . Here , we are interested in navigation in the vicinity of a target , where chemical signals , though noisy , provide information to the navigating cell . We show that even if the signal-to-noise ratio of gradient-sensing is below one , thus impeding reliable chemotactic steering , situation-specific switching between two steering modes can substantially increase the probability to find a target , such as the egg . This applies in particular at intermediate distances from the target , in a ‘noise zone’ , which cells have to cross before they can perform reliable chemotaxis close to the target . Our theory highlights a fundamental relationship between the speed of this chemotactic re-orientation and the strength of directional fluctuations , which result from the amplification of noise in the chemotactic input signal . Recent experiments revealed that during their chemotaxis along helical paths , sea urchin sperm cells switch between two distinct steering modes in a situation-specific manner [16] , see Fig 1 . These two steering modes , termed on- and off-response , are characterized by low and high values of the the rate γ of helix bending in the direction of the local concentration gradient , respectively , see Fig 1C . Cells were observed to employ on-responses when their helix axis pointed in the direction of the concentration gradient , but initiated a transient off-response if their helix axis pointed down the gradient [16] , see Fig 1D . Here , we defined the start of an off-response with ‘high-gain’ steering as the level crossing of γ above its median , and recorded the angle Ψ between the tangent of the helix centreline and the local concentration gradient at the respective times , see S1 Appendix in Supporting Information for details . Note that the relationship between bending rate γ and orientation angle Ψ cannot be explained by the simple geometric relation γ ∼ sin Ψ predicted by a previous theory [15] , see Fig A in S1 Appendix . For simplicity , we will employ an idealized description with two distinct steering modes , each characterized by a constant value of a sensori-motor gain factor introduced below , in contrast to a continuous regulation of this variable . Our theory provides a strong rationale that dynamic switching between steering modes increases the probability to find the egg in the presence of noise . We consider a theoretical description of sperm chemotaxis along helical paths , which describes the feedback loop between swimming , chemotactic signalling , and steering [15 , 26] . We extend this theory by incorporating a situation-specific modulation of the sensori-motor gain factor , which can take two different values in our theory . The sensori-motor gain factor , ρ , controls the strength of chemotactic steering in response to noisy gradient measurements by coupling the output of the chemotactic signalling system to swimming behaviour . Switching between two values of ρ represents a simple implementation of decision making . During chemotactic navigation , a sperm cell measures the concentration of chemoattractant along its swimming path r ( t ) . At low concentrations , the rate b ( t ) at which chemoattractant molecules bind to receptors on the cellular membrane is proportional to the local concentration c ( r ( t ) ) , i . e . b ( t ) = λ c ( r ( t ) ) ( 1 ) with binding constant λ = 7 s−1 pM−1 [27] . The input s ( t ) to the chemotactic signalling system is given by the train of individual binding events with rate b ( t ) ( which represents an inhomogeneous Poisson process with arrival times tj ) s ( t ) = ∑ j δ ( t - t j ) , ⟨ s ( t ) ⟩ = b ( t ) , ( 2 ) see Fig 2A and 2B . We employ a minimal description of chemotactic signalling with a dimensionless output of the signalling system a ( t ) and a dynamic sensitivity p ( t ) [26] , which implements its main characteristics observed in experiments: sensory adaptation and relaxation to a rest state after transient stimulation [28] μ a ˙ = p [ λ c b + s ( t ) ] - a , μ p ˙ = p ( 1 - a ) . ( 3 ) Here , cb sets a threshold of sensory adaption and μ characterizes a time scale of relaxation and adaptation . Dots denote time derivatives . For oscillatory input , s ( t ) = s0 + s1 cos ( Ωt ) , the output a ( t ) oscillates around its steady-state value 1 with amplitude proportional to s1/ ( λcb+ s0 ) . For chemotactic steering , the output of the signalling system , a ( t ) , dynamically regulates the curvature κ ( t ) and torsion τ ( t ) of the helical swimming path κ ( t ) = κ 0 - ρ κ 0 ( a - 1 ) , τ ( t ) = τ 0 + ρ τ 0 ( a - 1 ) . ( 4 ) Curvature and torsion uniquely characterize the time evolution of the swimming path r ( t ) by the Frenet-Serret equations , see S1 Appendix . For constant path curvature and torsion , κ ( t ) = κ0 and τ ( t ) = τ0 , the swimming path would be a perfect helix with radius r 0 = κ 0 / ( κ 0 2 + τ 0 2 ) , pitch 2 π h 0 = 2 πτ 0 / ( κ 0 2 + τ 0 2 ) , and angular helix frequency Ω 0 = v [ κ 0 2 + τ 0 2 ] 1 / 2 , where v denotes a constant swimming speed . In a concentration field , both κ and τ are dynamically regulated in response to the stochastic input signal s ( t ) . The sensori-motor gain factor ρ in Eq 4 sets both the speed of chemotactic steering and of noise amplification , and will be at our focus in the following . The chemotaxis paradigm embodied in Eqs 1–4 is summarised in Fig 2C and 2D: Helical swimming around a centreline R with helix axis perpendicular to a concentration gradient ∇c results in oscillations of the binding rate b ( t ) with the frequency Ω0 of helical swimming . As a consequence , path curvature and torsion oscillate , causing the helix to bend in the direction of the gradient . This decreases the angle Ψ between the helix axis and the gradient direction . Molecular shot noise in concentration measurements adds stochasticity to this directed steering , as discussed next . Eqs 1–4 ( with Eqs . S4-S6 in S1 Appendix ) represent a closed control loop and can be simulated numerically to obtain sperm swimming paths . We use a representative concentration field c ( x ) , established by diffusion from a spherical source representing an egg . Parameters have been chosen to match experiment with swimming speed v = 200 μm/s , helix radius r0 = 7 . 5 μm , helix pitch 2πh0 = 48 . 3 μm , helix period T = 0 . 34s [16] , and egg radius Regg = 100 μm [29] , see S1 Appendix . We use measured values for the chemoattractant content of egg cells and the diffusion coefficient of the chemoattractant [28] . Thus , computed concentrations and corresponding noise levels are representative of physiological conditions in sea urchin . Fig 2E shows swimming paths both in the absence and presence of sensing noise , for a low and a high value of the gain factor ρ in Eq 4 , respectively . For ‘low-gain’ steering , and in the absence of noise , i . e . s ( t ) = b ( t ) , the model sperm cell initially moves closer to the egg ( although it eventually misses the egg ) . If the helix axis is initially perpendicular to the gradient direction , the same occurs for all initial conditions with egg distance R0 = \|R ( t = 0 ) \| in a range T < R 0 < A low , with T ≈ 1 . 0 mm and A low ≈ 3 . 8 mm , see Figure C in S1 Appendix . For initial distances outside this attraction zone , R 0 > A low , swimming paths move away from the egg due to insufficient chemotactic attraction . In a ‘target zone’ defined by R 0 < T , the direction of the concentration gradient changes on short length scales due to the radial symmetry of the concentration field , and helix bending during ‘low-gain’ steering is too slow to follow the gradient . In the presence of noise , swimming paths become stochastic . For ‘low-gain’ steering , with only slight course correction , noise in the input signal hardly affects swimming paths . In contrast , a high gain factor results in fast bending of helical paths , yet it amplifies noise in concentration measurements considerably . This is particularly evident in a ‘noise zone’ spanning intermediate distances R from the egg , where concentration signals are detectable , but the signal-to-noise ratio ( SNR ) of gradient measurements is below one , see Fig 2F . We define the SNR as the ratio between the power of the gradient signal ( here encoded in oscillations of the binding rate b ( t ) with amplitude λ\|∇c\|r0 for swimming perpendicular to the gradient direction ) , and the noise strength of the input signal s ( t ) relative to a single helix period of duration T SNR ( R ) = ( λ \| ∇ c \| r 0 ) 2 / 2 λ c 0 / T , ( 5 ) see S1 Appendix . We introduce the distance N where the SNR equals one . Additionally , we introduce a distance S where only one molecule will be detected per helical turn on average , which marks a spatial limit of chemosensation . These two distances provide a formal definition of the ‘noise zone’ as the range of distances N < R < S bounded by N and S . We computed the probability P ( R0 ) to find the egg for a given initial distance R0 from the egg for a static gain factor ρ in Eq 4 , see Fig 2G ( assuming an isotropic distribution of initial swimming directions ) . In the hypothetical case of noise-free concentration measurements , the success probability is a monotonically increasing function of ρ . For physiological levels of sensing noise , however , we predict an optimal value of the gain factor ρ that maximizes P ( R0 ) , reflecting the competition between responding accurately ( low ρ ) or responding fast ( high ρ ) . The centreline R of helical paths describes a stochastic trajectory with directional persistence . The dynamics of its tangent vector R ˙ / \| R ˙ \| can be formally described as a superposition of ( i ) bending in the direction of the local concentration gradient with bending rate γ , and ( ii ) effective rotational diffusion with rotational diffusion coefficient D , see S1 Appendix . The bending rate γ characterizes a noise-averaged steering response , corresponding to the expectation value b ( t ) of the input signal s ( t ) , whereas the rotational diffusion coefficient D characterizes directional fluctuations of the tangent vector that arise from fluctuations of the input signal around its expectation value . An analytical theory valid in the limit of weak concentration gradients with \|∇cr0\|/c ≪ 1 provides expressions for both γ and D , demonstrating how both quantities scale with the sensori-motor gain factor ρ γ = ρ ε T \| ∇ ⊥ c \|r0 c b + c , ( 6 ) D = ( ρ ε T ) 2 c λ ( c b + c ) 2 . ( 7 ) Here , ε = 2 π κ 0 τ 0 / ( κ 0 2 + τ 0 2 ) is a geometric factor characterizing helical swimming . Eqs 6 and 7 were previously derived for the special case of a linear concentration field [15 , 26] and generalized here to arbitrary concentration fields . Note that the effective rotational diffusion coefficient D depends on the concentration c of signalling molecules . The ratio between the bending rate γ and the effective rotational diffusion coefficient D is directly related to the signal-to-noise ratio SNR defined in Eq 5 γ 2 D = 2 sin 2 Ψ T · SNR . ( 8 ) Eq 8 implies that the speed of steering ( characterized by γ ) and the strength of directional fluctuations due to sensory noise ( characterized by D ) are inseparably coupled . Prompted by recent experiments [16] displayed in Fig 1 , we now address dynamic switching between modes of ‘low-gain’ and ‘high-gain’ steering . We consider sperm navigation as a decision problem , in which a single chemotactic agent , here the sperm cell , can choose between two actions , i . e . ‘low-gain’ or ‘high-gain’ steering , at each state . We ask for a strategy that maximizes the probability to find the egg . We discretize phase space and map the stochastic dynamics of sperm chemotaxis on a finite-state Markov decision process ( MDP ) [30] . A coarse-grained , analytical theory of sperm chemotaxis implies that the distance to the egg R ( t ) = \|R ( t ) \| , and the swimming direction angle Ψ ( t ) , see Fig 2C , are sufficient to describe the dynamics of helical chemotaxis in a radial concentration field due to symmetry [15] . We simulated 104 helical swimming paths , determined R ( t ) and Ψ ( t ) , and then computed transition probabilities in a discretized ( R , Ψ ) -phase space for two values of a constant gain factor , ρ = ρlow and ρ = ρhigh , see Fig 3A and 3B . We obtain respective transition matrices L i j low and L i j high for transitions from one bin labelled i to another bin labelled j , see also Fig D in S1 Appendix . In each case , the transition dynamics is approximately Markovian , see Fig E in S1 Appendix . We additionally introduce an absorbing ‘success state’ if the egg is found , and an absorbing ‘failure state’ if the cell moves beyond a threshold distance , Rth , marking the end of a single search attempt . This Markov chain allows to efficiently determine the probabilities to eventually find the egg . As a control , we compare success probabilities computed using full simulations and predictions from this Markov chain , see Fig F in S1 Appendix . Now , we introduce a MDP , where the model cell can choose in each state between the two actions ‘low-gain’ steering and ‘high-gain’ steering , see Fig 3C . The choice determines the transition probabilities to the next state . We ask for the optimal decision strategy that maximizes the probability to eventually reach the ‘success state’ . An example strategy is sketched in Fig 3D , assigning a choice of steering mode to each state . A fundamental theorem in the theory of MDP states that the optimal strategy can always be chosen to be memoryless , with a hard-wired choice for each state , independent of the history of previous states [31] . We now compute optimal memory-less strategies , and discuss how these depend on the presence of sensing noise . We computed optimal decision strategies for the MDP of sperm navigation , see S1 Appendix for details . We used the open-source probabilistic model checking software PRISM , which offers efficient algorithms to compute optimal strategies even for large MDPs [32] . In Fig 4 , we compare the success probability for the optimal strategy to the success probabilities one would obtain for strategies that choose either always ‘low-gain’ or ‘high-gain’ steering . In the hypothetical case of noise-free concentration measurements , the performance of the optimal strategy is virtually indistinguishable to that of ‘high-gain’ steering , see Fig 4A . In contrast , when accounting for physiological levels of sensing noise , success probabilities for the optimal strategy are substantially higher than success probabilities for ‘low-gain’ and ‘high-gain’ steering , see Fig 4B . This concerns especially intermediate initial distances from the egg , where concentrations are low and sensing noise corrupts concentration measurements . Next , we analysed significance and benefit of optimal decision making at different distances from the egg . We averaged computed strategies for an ensemble of realizations of the MDP , each with transition probabilities obtained by bootstrapping from a large cohort of simulated sperm swimming paths , see Fig 4C and 4D . Greyscale values indicate the frequency that ‘high-gain’ steering is predicted to be optimal for a given state . Thereby , we explicitly harness numerical variations in transition probabilities to extract relevant features of optimal decision strategies . For the case of noise-free concentration measurements , we find two distinct state-space regions , where ‘high-gain’ steering is always favoured . The first region corresponds to the ‘target zone’ , defined as R < T , where the model sperm cell cannot come closer to the egg if it employs ‘low-gain’ steering and initially starts with its helix axis perpendicular to the concentration gradient . A second region is bounded from below by the attraction radius A low ≈ 3 . 8 mm for ‘low-gain’ steering , and by an analogously defined A high for ‘high-gain’ steering with A high ≈ 4 . 8 mm from above . In this region , ‘high-gain’ steering is important to attract cells that would otherwise move away from the egg , see S1 Appendix . In the presence of sensing noise , we consistently find that the optimal strategy chooses ‘low-gain’ steering while moving up-gradient , but chooses ‘high-gain’ steering when accidentally moving down-gradient . This choice is prevalent in the ‘noise zone’ at intermediate distances from the egg , where gradients are detectable , but the SNR ratio is below one . This theoretical prediction matches the observed steering behaviour of sperm cells in recent experiments performed at high chemoattractant concentrations [16] , see Fig 1D . To compare the efficacy of different decision strategies , we introduce the effective chemotactic volume 4 3 π R 3 = ∫ 0 ∞ d R ( 4 π R 2 ) P ( R ) , ( 9 ) which defines an effective chemotactic range R for a given decision strategy . The chemotactic range R sets an effective target size . We find R ≈ 6 . 2 mm for the optimal strategy , while R low ≈ 3 . 8 mm and R high ≈ 4 . 3 mm for a strategy that always chooses either ‘low-gain’ or ‘high-gain’ steering , respectively . Next , we asked at which distances to the target decision making is most important . To quantify respective benefits , we computed chemotactic ranges R ( R c ) as a function of a cut-off distance Rc for hybrid strategies . These hybrid strategies employ the optimal decision strategy only at distances smaller than Rc , but choose always either ‘low-gain’ or ‘high-gain’ steering , respectively , outside this range . In particular , positive values of the derivative ∂ R / ∂ R c reveal at which distances decision making is most beneficial , see Fig 4E and 4F . Distances where this spatial significance measure is positive match exactly those regions where decision strategies are most stable with respect to numerical noise . Thus , two independent spatially-resolved sensitivity measures for optimal strategies give congruent results . While the formalism of MDPs allows us to efficiently compute optimal decision strategies , it is not evident how a biological cell would implement such strategies . In particular , a swimming cell has no direct access to the state variables R and Ψ , but only to the noisy concentration signal s ( t ) . We present a minimal signalling system that implements decision making on the basis of s ( t ) , i . e . on the basis of available information . We introduce a trigger variable q ( t ) that tracks the output of the signalling system a ( t ) with a relaxation time scale η [16] η q ˙ = a - q . ( 10 ) This low-pass filter attenuates fast oscillations of a ( t ) caused by helical swimming in a concentration gradient , yet faithfully retains changes in the baseline of a ( t ) , which occur for either up-gradient or down-gradient swimming ( due to a finite time scale of sensory adaptation [16] ) . In the absence of noise , and for a given concentration field c ( R ) , the signalling variables ( p , q ) are directly related to the state variables ( R , ψ ) as p−1 = λ[cb+ c ( R ) ] and q = 1 + μΩ0h0 pλ\|dc ( R ) /dR\|cosψ ( if we neglect residual oscillations of q ( t ) ) . In the presence of noise , p and q scatter around their expected values , see Fig 5A . Consequently , estimation of state ( R , ψ ) based on ( p , q ) is associated with an error . The accuracy in discriminating between swimming up-gradient ( ψ ≤ π/2 ) and down-gradient ( ψ > π/2 ) decreases as a function of distance R from the egg , see Fig G in S1 Appendix . Estimation of helix orientation angle Ψ can be considered feasible up to a maximal distance R ≈ 3mm , where the accuracy equals 66% ( 100%: perfect discriminability , 50%: complete lack of discriminability ) . We now design a decision rule in terms of p and q ρ ( p , q ) = { ρ low , for q ≥ Θ ( p ) ρ high , for q < Θ ( p ) , ( 11 ) with decision boundary Θ ( p ) yet to be determined . From the optimal decision strategy predicted for the MDP , we compute the relative frequency of ‘high-gain’ steering for each pair of values p and q , using the likelihood of states ( R , Ψ ) for given tuple ( p , q ) , see Fig 5B . We define Θ ( p ) as a piecewise linear fit to the 50%-contour line of this relative frequency , see S1 Appendix . Note that this decision boundary implies ‘low-gain’ steering far from the egg . Helical sperm swimming paths simulated with this decision rule display frequent switching to ‘high-gain’ steering in the noise zone , and only sporadic events of ‘high-gain’ steering in the target zone , see representative example in Fig 5C . Chemotaxis with decision making increases the probability of success for intermediate initial distances to the egg , similar to our analysis of the MDP , see Fig 5D . Note that we use a finite search time of 300s in Fig 5D , which yields lower success probabilities as the corresponding MDP representation , which considers infinite search times . Our simple implementation of decision making is more effective than any constant gain factor , see Fig 5E . While we compute Θ ( p ) for a specific concentration field , the same decision boundary performs superior also in other concentration fields , highlighting a general benefit of decision making , see Fig H in S1 Appendix . We developed a theory of optimal chemotaxis towards a single target in the presence of noise , using sperm chemotaxis along helical paths as application example . We show that a situation-specific switching between two different steering modes—‘low-gain’ and ‘high-gain’ steering—maximizes the probability to find a target , such as an egg , at the centre of a radial concentration field of signalling molecules . The benefit of decision making is causally related to noise in sensory input . If cells could measure concentrations with perfect accuracy , decision making would provide no benefit , compared to exclusive high-gain steering . For physiological noise levels relevant for sperm chemotaxis , ‘low-gain’ steering is chosen in the optimal strategy if the cell is approximately heading in target direction . This minimizes the risk of inadvertently steering in the wrong direction by amplifying noise in the chemotactic input signal . ‘High-gain’ steering is chosen if the net swimming direction is at least perpendicular to the target direction , and the potential benefit of fast steering outweighs the risk of wrong course corrections . The optimal strategy predicted by our theory matches a surprising experimental observation recently made by Jikeli et al . [16] , summarized in Fig 1 . There , it was observed that sea urchin sperm cells switch between ‘low-gain’ and ‘high-gain’ steering depending on their net swimming direction relative to the local concentration gradient . Experiments were performed at high concentrations , corresponding to the target zone in our description . This minimized noise in the experiments . We propose that sperm cells employ decision making not only at the high concentrations tested in experiments , but also at lower , physiological concentrations . Our theory predicts that decision making is most beneficial there . By switching between ‘low-gain’ and ‘high-gain’ steering , sperm cells dynamically adjust the persistence length of the centreline of their helical swimming paths . Thus , in addition to helical chemotaxis with gradual alignment of the centreline with the local gradient direction , these cells simultaneously perform a biased persistent random walk . This random walk can be interpreted as a time-continuous variant of ‘run-and-tumble’ chemotaxis , where the regulation of persistence length is analogous to the regulation of the duration of ‘runs’ . In conclusion , we find that noise in cellular gradient measurements poses a key constraint on chemotactic navigation . We report a fundamental relationship between the speed of chemotactic steering and the strength of directional fluctuations , which are caused by noise in the sensory input . The resultant trade-off between either reliable or fast steering applies to chemotactic motion with directional persistence in general , including chemotaxis by spatial comparison as employed e . g . by eukaryotic cells with crawling motility . We expect that other search problems can described by a Markov decision process in a similar fashion . The heuristic strategy predicted by our theory requires only minimal computational capacities of chemotactic agents and could inspire optimal control designs for artificial microswimmers .	Many cells can navigate upwards a concentration gradient of signalling molecules , a process called chemotaxis . Chemotaxis is used e . g . by sperm cells to find the egg . To measure and compare concentrations , cells count stochastic binding events of signalling molecules that diffuse to cellular receptors . Efficient chemotaxis strategies must be adapted to this molecular shot noise of concentration measurements . We reveal a fundamental relationship between the speed of chemotactic steering and the strength of directional fluctuations that result from the amplification of noise . This implies a trade-off between steering fast and steering reliable . Inspired by recent experiments on chemotaxis of sperm cells of marine invertebrates , we develop a theory that allows to efficiently compute optimal chemotaxis strategies . We show that dynamic switching between either fast or reliable steering can substantially increase the probability for a sperm cell to find the egg . Furthermore , the optimal strategy requires only minimal computational capacities of the chemotactic agent , a key constraint for biological cells . More generally , our work demonstrates a benefit of decision making for chemotaxis in the presence of noise , which could inspire optimal control designs for artificial microswimmers .	[ "Abstract", "Introduction", "Results", "Discussion" ]	[ "cell", "motility", "swimming", "medicine", "and", "health", "sciences", "steering", "decision", "making", "engineering", "and", "technology", "sperm", "chemotaxis", "signal", "processing", "social", "sciences", "neuroscience", "biological", "locomotion", "germ", "cells...	2018	Decision making improves sperm chemotaxis in the presence of noise
Phlebotomus duboscqi is the principle vector of Leishmania major , the causative agent of cutaneous leishmaniasis ( CL ) , in West Africa and is the suspected vector in Mali . Although found throughout the country the seasonality and infection prevalence of P . duboscqi has not been established in Mali . We conducted a three year study in two neighboring villages , Kemena and Sougoula , in Central Mali , an area with a leishmanin skin test positivity of up to 45% . During the first year , we evaluated the overall diversity of sand flies . Of 18 , 595 flies collected , 12 , 952 ( 69% ) belonged to 12 species of Sergentomyia and 5 , 643 ( 31% ) to two species of the genus Phlebotomus , P . duboscqi and P . rodhaini . Of those , P . duboscqi was the most abundant , representing 99% of the collected Phlebotomus species . P . duboscqi was the primary sand fly collected inside dwellings , mostly by resting site collection . The seasonality and infection prevalence of P . duboscqi was monitored over two consecutive years . P . dubsocqi were collected throughout the year . Using a quasi-Poisson model we observed a significant annual ( year 1 to year 2 ) , seasonal ( monthly ) and village effect ( Kemena versus Sougoula ) on the number of collected P . duboscqi . The significant seasonal effect of the quasi-Poisson model reflects two seasonal collection peaks in May-July and October-November . The infection status of pooled P . duboscqi females was determined by PCR . The infection prevalence of pooled females , estimated using the maximum likelihood estimate of prevalence , was 2 . 7% in Kemena and Sougoula . Based on the PCR product size , L . major was identified as the only species found in flies from the two villages . This was confirmed by sequence alignment of a subset of PCR products from infected flies to known Leishmania species , incriminating P . duboscqi as the vector of CL in Mali . In West Africa Phlebotomus duboscqi Neveu-Lemaire is the most important vector of Leishmania major , the causative agent of cutaneous leishmaniasis ( CL ) [1] , [2] . P . duboscqi has been incriminated as the vector of L . major in Senegal [3] and suspected as the vector of CL in Burkina Faso [4] , Niger [5] , [6] , The Gambia [7] , Ghana [8] , Cameroon [9] and Mali [10] , [11] , [12] . The first report of P . duboscqi in Mali was from Hombori in 1906 [13] with additional reports from Timbuctu in 1913 [14] and from Bamako and Nioro in 1943 [10] . Later work by Lariviere [11] and Desjeux [1] found P . duboscqi in all regions of the country . Cutaneous Leishmaniasis is endemic in Mali with cases historically occurring in the districts of Nioro and Segou [11] , [15] . The first published report of CL in Mali concerned two cases identified from Nioro in 1944 [16] . Later studies reported leishmanin skin test positivity rates between 10 and 61% , suggesting that Leishmania is endemic in Mali [15] , [17] , [18] , [19] . Leishmania major was first identified as the causative agent of CL in Mali by isoenzyme analysis of parasites isolated from skin samples taken from a lesion of a tourist visiting Mopti [20] and a local resident living in the same region [21] . Despite the identification of L . major as the causative agent of CL in Mali , and although suspected as the vector , no one has identified the parasite in P . duboscqi . Here , we report on a three year survey to evaluate the diversity of sand flies and the seasonal abundance of P . duboscqi in Kemena and Sougoula , two villages endemic for CL in the District of Baroueli , Region of Segou , in Central Mali . Furthermore , we report for the first time the detection and annual prevalence of L . major parasites in P . duboscqi sand flies collected from the study sites . Sand flies were collected from two neighboring villages , Kemena ( 12°33′ N–6°33′ W ) and Sougoula ( 13°05′ N , –6°53′ W ) , in the Baroueli Health District , Region of Segou , Mali . Both villages have a population size of approximately 1000 inhabitants . Each village is organized into a labyrinth of adjoining compounds within which a single extended family resides in several sleeping , cooking , and storage houses . Houses are constructed of clay bricks plastered with mud and straw , and with thatched or metal roofs . Domestic animals , such as goats , sheep , and chickens are kept within the confines of a family compound while cows are maintained in corrals located around the perimeter of the village . Both villages have a limited infrastructure and lack electricity and running water . The climate consists of three distinct seasons: a dry season from March to June ( temperature range 27–40°C; monthly average rainfall 5 . 2 mm ) , a rainy season from June to September ( temperature range 25–35°C , monthly average rainfall 82 . 42 mm ) , and a third temperate season from October to February ( temperature range 20–35°C; monthly average rainfall 3 . 3 mm ) . Vegetation is sparse and is characterized by the presence of sporadically placed trees such as shea ( Vitellaria paradoxa ) , acacia ( Faidherbia albida ) and neem ( Azadirachta indica ) and small bushes . Most of the land surrounding each village is dedicated for agricultural use . Sand flies were collected using 1 ) dark activated , CDC miniature light traps fitted with double ring fine mesh collection bags ( John W . Hock Company , Gainesville , FL ) , 2 ) sticky traps consisting of single sheets of A4 paper ( 21×29 . 5 cm ) coated on both sides with castor oil and mounted vertically on pegs , onto which randomly impinging sand flies would adhere ( used for the sand fly diversity study only ) , and 3 ) mouth aspirators ( John W . Hock Company , Gainesville , FL ) for collection of resting flies inside of houses used for sleeping . All sand flies were sorted by sex , species and blood meal status , and placed in tubes containing silica gel and cotton until processed . Minimum and maximum temperatures , rainfall and relative humidity for the months of July 2006 to June 2008 were collected from the nearest available weather station in Segou , Mali . Oral informed consent was obtained from head of households for indoor collection of sand flies . Households where consent was given were listed in a written log kept by the entomological team for reference . The head and terminal segments of the abdomen containing the genitalia of each sand fly were carefully removed and placed into 96-well plates containing a solution of lacto-phenol clearing solution ( Bioquip ) . After 24 h incubation at room temperature , the head and terminalia were fixed onto a glass slide , examined using a light microscope and identified using dichotomus keys [22] . From June 2006-July 2008 , the abdomens of blood fed and non-blood fed Phlebotomus females from the same collection location were grouped in pools of no more than 20 individuals and placed in a microfuge tube containing lysis buffer ( 5 . 84 g/L NaCl , 68 . 5 g/L Sucrose , 12 . 10 g/L Tris , pH 9 . 1 , 100 ml EDTA 0 . 5 M solution and 50 ml 10% SDS solution ) . After incubating overnight at 4°C the tissue was macerated using a pestle for 2 min then incubated for 30 min at 65°C . After the addition of 10 µl cold potassium acetate the samples were incubated for 30 min at 4°C and then centrifuged for 10 min at 14 , 000 RPM . The DNA was precipitated using 70% ethanol and resuspended in 100 µl water . The DNA concentration of each extraction was determined using a NanoDrop ( Thermo Scientific Inc . , Wilmington , DE ) . Samples with less than 4 ng/µl of DNA were removed from the sample set . Leishmania DNA was detected by PCR using forward and reverse primers for Leishmania sp . ( Uni21/Lmj4 ) as described in [23] . PCR Primers targeting the sand fly tubulin gene were used as a control for template fidelity ( PpTub-P24F 5′-GCG ATG ACT CCT TCA ACA C and PpTub-P24R 5′-TCA GCC AGC TTG CGA ATA C ) [24] . A representation of PCR products was confirmed by DNA sequencing . Due to difficulties with direct sequencing of the PCR products using the Uni21 and Lmj4 primers , gel-purified PCR products were cloned into the pCR4-TOPO vector using the TOPO TA Cloning Kit for Sequencing ( Invitrogen , Carlsbad CA ) following the manufacturer's instructions . The clones were sequenced directly using the M13 forward and M13 reverse primers . Resulting sequences were analyzed using DNASTAR sequence analysis software ( DNASTAR , Inc . , Madison WI ) . Sequences were compared to published sequences of kDNA from L . major ( Genbank Accession J04654 ) , L . infantum ( AF188701 ) , L . tropica ( Z32841 ) , and L . donavani ( AF167718 ) using BLAST ( http://blast . ncbi . nlm . nih . gov/ ) , aligned to known Leishmania minicircle kinetoplastic DNA using Clustal [25] and edited using BioEdit ( http://www . mbio . ncsu . edu/BioEdit/page2 . html ) . To estimate the prevalence of infection in pooled samples of P . duboscqi females , we used the maximum likelihood estimate ( MLE ) of prevalence accounting for pooling with the confidence interval ( CI ) estimated by exact methods if the number of unique pool sizes was less than or equal to 3 [26] , or otherwise by the skewness-corrected score confidence interval [27]; the estimates and both CIs were calculated using the binGroup R package [28] . To model the sand fly counts or infection rates we used a quasi-Poisson model and tested for significant effects using analysis of deviance and F test [29] . To test for seasonal effects , we tested the overall effect of months after controlling for previous counts and year . In testing for weather effects , we compared models with previous counts , year and months and tested to see if models that additionally added the previous month weather variables ( including 4 weather variables at a time; selecting only one [minimum or maximum] of temperature or wild velocity variables ) significantly improved the fit . For the models of rates , we estimated the number of infected flies of those tested by the MLE of prevalence and used those counts as responses in the quasi-Poisson model with an offset based on the number of flies tested so that the inferences describe effects on the rates [29] . The quasi-Poisson models were performed using R version 2 . 12 [30] . Graphs were made using GraphPad Prism 5 ( Graphpad Software , California , USA ) . From March 2005 to June 2006 , 18 , 595 sand flies were collected in the two villages ( 9 , 887 in Kemena and 8 , 708 in Sougoula ) using all three collection methods . Approximately equal numbers of male and female sand flies were collected ( 9 , 221 M , 9 , 374 F ) . Sixty-nine percent ( n = 12 , 952 ) of sand flies were identified as one of 12 species in the genus Sergentomyia , none of which have been implicated in the transmission of L . major ( Table 1 ) . Of the Sergentomyia , Sergentomyia schwetzi Adler , Theodor and Parrot represented the majority with 47 . 3% of collected specimens , while Sergentomyia antennata Newstead was the second most abundant at 26 . 4% . Ten additional Sergentomyia species were collected: Sergentomyia dubia Parrot , Mornet , and Cadenat ( 12 . 2% ) , Sergentomyia clydei Sinton ( 7 . 9% ) , Sergentomyia africana Newstead ( 3 . 2% ) , Sergentomyia squamipleuris Newstead ( 1 . 7% ) , Sergentomyia affinis vorax Parrot ( 0 . 56% ) , Sergentomyia bedfordi Newstead ( 0 . 49% ) , Sergentomyia fallax Parrot ( 0 . 02% ) , Sergentomyia buxtoni Theodor ( 0 . 13% ) , Sergentomyia darlingi Lewis and Kirk ( 0 . 06% ) , and Sergentomyia christophersi Sinton ( 0 . 01% ) . The remaining 30% of sand flies collected was identified as one of two species of Phlebotomus , the overwhelming majority of which was P . duboscqi ( n = 5 , 643 , 99 . 3% ) . Only 41 Phlebotomus rodhaini Parrot ( 0 . 7% ) were collected ( Table 1 ) . Sticky traps and light traps collected sand flies in about equal numbers ( n = 8 , 290 vs . 8 , 394 ) , yet the majority of Sergentomyia ( n = 7 , 728 , 60% ) were collected using sticky traps whereas only 10% of Phlebotomus ( n = 562 ) were collected using this method . The majority of Phlebotomus ( n = 3 , 380 , 60% ) were collected using light traps . Thirty percent of Phlebotomus ( n = 1 , 701 ) were collected by resting site collection compared to only 1 . 62% ( n = 210 ) of Sergentomyia . Comparing sticky trap and light trap collections from inside and outside houses , the majority of Phlebotomus ( 92% , n = 3 , 641 ) were collected inside dwellings whereas the majority of Sergentomyia ( 71% , n = 9 , 043 ) were collected outside dwellings . From July 2006 to June 2008 , 7 , 950 P . duboscqi ( 3 , 998 female ) were collected . Additionally , a total of 25 P . rodhaini were collected during the two years , 17 of which were collected during one month in Kemena ( October 2006 ) . Comparing the total number of P . duboscqi collected during year one ( July 2006–June 2007 ) and year 2 ( July 2006–June 2008 ) , we found that 1 . 42 times more sand flies were collected in year 2 ( p-value 0 . 0002 , 95% CI: 1 . 19–1 . 69 ) ( Figure 2A ) . We observed a similar effect when comparing the annual collections of female P . duboscqi ( p-value 0 . 0003 ) ( Figure 2B ) . Using the quasi-Poisson model , controlling for year and previous count , we observed a significant seasonal effect reflecting the month to month variation in the total number of sand flies collected ( p-value <0 . 0001 ) . A similar effect was observed when we considered only female sand flies ( P-value <0 . 0001 ) . Monthly collection trends were similar during both collection years . We modeled sand fly counts for each month using January , the lowest seasonal collection month , as a reference . An initial peak with a 3 . 9 fold change from January [FCJan] ( 95% CI: 2 . 6 , 6 . 1 ) was observed in May and July ( 3 . 8 FCJan , 95% CI 2 . 4 , 6 . 0 ) . This was followed by a dip in collections in August ( 2 . 0 FCJan , 95% CI: 1 . 2 , 3 . 3 ) and September ( 1 . 3 FCJan , 95% CI: 0 . 8 , 2 . 2 ) and a second upward trend peaking in November ( 3 . 6 FCJan , 95% CI: 2 . 4 , 5 . 8 ) ( Figure 2A ) . By village , we found that 45% ( n = 3 , 654 ) of all P . duboscqi ( male and female ) were collected in Kemena and 54% ( n = 4 , 276 ) in Sougoula . Using the quasi-Poisson model , controlling for previous count , month , and year we observed a significant difference in total P . duboscqi counts between the two villages ( p-value 0 . 0293 ) with the sand fly counts 1 . 188 times higher , on average , in Sougoula than in Kemena ( 95% CI: 1 . 025 , 1 . 377 ) ( Figure 2A ) . Similar results hold when using only female sand fly counts ( fold-change = 1 . 155 , p = 0 . 1116 , Figure 2B ) . The various weather variables ( relative humidity , rainfall amount , maximum or minimum temperature , and maximum and minimum wind velocity ) were not useful for predicting the observed total or female sand fly collections for either village ( all models had p-value >0 . 53 ) ( Figure 3B ) . The majority of P . duboscqi was collected by resting site collection , particularly during the morning ( 10 . 54 and 14 . 04 female P . duboscqi/person/hour during morning collections in Kemena and Sougoula , respectively , compared to 5 . 12 and 6 . 10 P . duboscqi/person/hour during evening collections ) ( Figure 4 ) . On average , five times more P . duboscqi were collected using light traps placed inside of dwellings than outside in the same compound ( 1 . 74 vs . 0 . 33 and 1 . 78 vs . 0 . 31 P . duboscqi females/trap/night in Kemena and Sougoula , respectively ) ( Figure 4 ) . Virtually no P . duboscqi were collected in the light traps placed outside of the village near natural tree holes ( 0 . 03 and 0 . 12 P . duboscqi females/trap/night in Kemena and Sougoula , respectively ) . A total of 1434 pools ( 3706 total flies; average 2 . 6 flies per pool ) were examined for Leishmaina infection by PCR . Ninety-seven pools were positive for L . major ( Figure 5 ) . Assuming that the sand flies are independently distributed in the pools and the size of the pools is not related to the probability of infection of the pool , we estimate the prevalence of infection to be 2 . 66% , 95% CI: 2 . 20 , 3 . 21 ( Table 2 ) . Infected P . duboscqi were found during each month of the year , although monthly infection estimates varied greatly year to year , being the highest during September 2006 ( 9 . 64%; CI: 4 . 68 , 17 . 34 ) and February 2008 ( 9 . 19%; 95% CI: 5 . 03 , 15 . 27 ) ( Figure 6 ) . After controlling for sand fly count , there was no significant difference in the rates of infection from year 1 to year 2 ( p-value 0 . 2572 ) and neither was there a significant month to month difference ( p = 0 . 2085 ) . The estimated infection prevalence of sand flies was virtually the same for both villages ( 2 . 65% , 95% CI: 1 . 97 , 3 . 51 and 2 . 67 , 95% CI: 2 . 02 , 3 . 44 for Kemena and Sougoula , respectively ) with no significant difference between the two villages ( p-value 0 . 8894 ) ( Table 2 ) . Of the sand flies collected by light traps and resting site collections within compounds , the majority of infected sand flies were collected using light traps versus resting site collection ( 4 . 14% of vs . 0 . 86% ) . The highest estimated prevalence of infected sand flies was collected from compound 5 in Kemena ( 9 . 42%; 95% CI: 5 . 50 , 14 . 92 ) and compound 1 in Sougoula ( 5 . 94%; 95% CI: 2 . 97 , 10 . 57 ) . Comparing the position of the light traps , the estimated infection prevalence of sand flies collected in light traps placed directly outside dwellings was higher than those placed inside dwellings ( 8 . 15% vs . 2 . 07% ) ; no infected sand flies were collected from light traps placed in trees outside of either village ( Table 2 ) . The estimated infection prevalence of flies that were non-blood fed at the time of collection was 4 . 01% versus 1 . 24% for those that were blood fed . Eight representative PCR products from infected wild caught P . duboscqi were sequenced using primers specific to the kinetoplast minicircle DNA of L . major . Sequence analysis confirmed that all the samples were similar to published L . major sequences based on length of the product and primer region identity . Blast analysis indicated a best match to L . major kinetoplast DNA ( Genbank Accession number Z32842 . 1 , E-value 9e-48 ) . Further alignment of the sequences obtained from this study with known Leishmania sequences of other species in Genbank confirmed that the 650 bp fragment size , observed on gel electrophoresis of the PCR products , as characteristic of L . major strains . Killick-Kendrick [31] suggested the following criteria for incrimination of a vector sand fly: proven anthropophilic behavior and isolation and identification from the sand fly of the same species of Leishmania that infects man . Further evidence such as the demonstration that the sand fly feeds on the reservoir host ( if known ) , concordance between the geographic distribution of the suspected sand fly and human disease , proof that the parasite develops in the fly and experimental transmission of the parasite by the bite of the fly can reinforce the incrimination . Based on monthly collections of sand flies over three years in two villages in central Mali , where CL is known to be endemic , we have demonstrated that P . duboscqi is the predominant Phlebotomus species; that it persists throughout the year; that females are primarily collected inside houses in both villages;and that it has an overall infection rate with L . major of 2 . 66% as demonstrated by PCR . This strongly points to P . duboscqi as the primary vector of L . major in Mali . Species of the sub-genus Sergentomyia constitute the majority of sand flies collected in both villages with S . schwetzi being the most abundant . Members of the Sergentomyia genus are known to transmit Sauroleishmania among lizards . Sergentomyia schwetzi is the only Sergentomyia species known to be anthropophilic and was considered a possible vector by Parrot [5] in 1943 . Later , Lawyer [32] concluded that despite the anthropophilic behavior of S . schwetzi , it was not a vector of Leishmania in humans . In this study , only 1 . 6% of all Sergentomyia sand flies were found during resting site collections and the overall majority ( 70% ) was collected outside of dwellings , further supporting the exophilic nature of this sub-genus and the improbability that Sergentomyia sand flies are involved in transmission to humans in our two villages . Two Phlebotomus species were found during the three collection years , P . duboscqi and P . rodhaini , both of which are known vectors of L . major elsewhere in West Africa . While fewer in number than Sergentomyia species , P . duboscqi was predominantly collected by resting site collection from sleeping dwellings and five times more P . duboscqi females were collected in light traps placed inside than outside of dwellings , supporting the anthropophilic nature of this fly . P . duboscqi was collected in similar numbers in both villages year round with two seasonal peaks , May-July and October-November . These results are consistent with Lariviere [11] who reported on the seasonality of 191 P . duboscqi collected in Mali . The collection of P . duboscqi throughout the year is probably the result of having constant monthly temperatures and a relative humidity that does not drop beyond 18% . However , non of the specific weather parameters tested could be significantly correlated with sand fly collections in either village ( Figure 3 ) . Few specimens of P . rodhaini were collected throughout the study period indicating that this species probably does not play a role , or plays a minor role , in the transmission of L . major in Central Mali . To further incriminate P . duboscqi as the vector of Leishmania in our study villages , we tested 3706 specimens ( in 1434 pools ) for the presence of Leishmania DNA by PCR . We found that 97 of the pools tested positive for an overall estimated sand fly infection prevalence of 2 . 66% . None of the infected pools contained P . rodhaini . Since the infection rate in wild-caught sand flies is usually low [31] , PCR was used to permit the efficient screening of a large number of specimens . Having established the infection rate of P . duboscqi in this region , we plan to isolate a viable culture of L . major , necessary to type the strain using traditional methods such as isoenzyme analysis . It is worth noting that in 2006 we established a colony of P . duboscqi collected from our two study villages . Subsequently , females from this colony were used successfully to transmit L . major to an animal model of CL [33] further supporting the status of this species as a competent vector of CL in Central Mali . A recent study by our group [19] found that there is an unexplained discrepancy between the prevalence of leishmanin skin test ( LST ) positivity in our two study villages , Kemena ( 45% LST positive ) and Sougoula ( 20% LST positive ) , despite the fact that the villages are geographically and demographically similar and are only 5 km apart . Furthermore , this discrepancy was consistent over two consecutive annual incidences ( 18% and 17% in Kemena vs . 5 . 7% in both years in Sougoula ) . We hypothesized that the sand fly density and infection prevalence may explain the dissimilar LST results . The two year seasonality study revealed that slightly more female P . duboscqi were collected in Sougoula than in Kemena , yet almost the same percentage of pools were infected in each village ( 2 . 67% vs . 2 . 66% , respectively ) , thus neither abundance nor infection prevalence can explain the disparate LST rates observed in the two villages [19] . Rodent species are well known reservoirs for L . major throughout its distribution range . The contribution of reservoirs to the observed disparity of LST positivity in the two villages remains to be evaluated . In West Africa , including Senegal where P . duboscqi has been incriminated as the vector of L . major , infected Mastomys erythroleucus , Tatera gambiana and Arvicanthis niloticus have been reported [34] , [35] , [36] , [37] . All three species are found in Mali ( T . Schwan , personal communications ) and represent potential reservoirs of L . major in Kemena and Sougoula . Indeed , rodent burrows were observed in many of the houses where light traps were placed . Apart from the potential role of these rodents as reservoirs , their burrows also represent suitable sand fly breeding sites and a source of infected flies . Furthermore , all compounds in our study villages contain goats and chickens living in close proximity to houses used for sleeping which also represent good sand fly breeding sites for uninfected flies . A comprehensive study of the rodent population density and infection prevalence in the two villages is needed to fully understand the infection dynamics in both flies and people . In summary , we have established , for the first time , the diversity of sand flies in two villages endemic for L . major in Central Mali and demonstrated by PCR that P . duboscqi is the primary vector . This work represents the most comprehensive analysis of P . duboscqi , to date , in Mali and further supports the endemic nature of CL in Central Mali . Further investigations of this nature are needed in West Africa .	Female sand flies transmit a parasite called Leishmania that causes a disease called cutaneous leishmaniasis ( CL ) . Several species of sand flies are found in West Africa , but only one species , Phlebotomus duboscqi , has been proven to transmit the parasite . Cutaneous Leishmaniasis has also been reported from Mali , Central West Africa , but the sand fly transmitting the parasite and its annual abundance has not been established , until now . Sand flies were collected during three consecutive years from two neighboring villages in Central Mali , Kemena and Sougoula , where CL is present . P . duboscqi was collected year-round and was the dominant sand fly inside of and surrounding human dwellings . Other sand fly species , known not to be vectors of CL , were primarily found outside the village . Additionally , P . duboscqi females were found infected with L . major , the same Leishmania species identified from human CL cases in Mali . The estimated infection prevalence of P . duboscqi females was 2 . 7% . Interestingly , the sand fly abundance and infection prevalence was similar in the two villages despite a previous report indicating a disparate L . major exposure rate in humans . This study greatly enhances our knowledge of CL transmission in Mali , poorly studied in this country to date .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "medicine", "infectious", "diseases", "ecology", "biology" ]	2011	Seasonality and Prevalence of Leishmania major Infection in Phlebotomus duboscqi Neveu-Lemaire from Two Neighboring Villages in Central Mali
Recent evidence suggests that the metabolism of some organisms , such as Escherichia coli , is remarkably efficient , producing close to the maximum amount of biomass per unit of nutrient consumed . This observation raises the question of what regulatory mechanisms enable such efficiency . Here , we propose that simple product-feedback inhibition by itself is capable of leading to such optimality . We analyze several representative metabolic modules—starting from a linear pathway and advancing to a bidirectional pathway and metabolic cycle , and finally to integration of two different nutrient inputs . In each case , our mathematical analysis shows that product-feedback inhibition is not only homeostatic but also , with appropriate feedback connections , can minimize futile cycling and optimize fluxes . However , the effectiveness of simple product-feedback inhibition comes at the cost of high levels of some metabolite pools , potentially associated with toxicity and osmotic imbalance . These large metabolite pool sizes can be restricted if feedback inhibition is ultrasensitive . Indeed , the multi-layer regulation of metabolism by control of enzyme expression , enzyme covalent modification , and allostery is expected to result in such ultrasensitive feedbacks . To experimentally test whether the qualitative predictions from our analysis of feedback inhibition apply to metabolic modules beyond linear pathways , we examine the case of nitrogen assimilation in E . coli , which involves both nutrient integration and a metabolic cycle . We find that the feedback regulation scheme suggested by our mathematical analysis closely aligns with the actual regulation of the network and is sufficient to explain much of the dynamical behavior of relevant metabolite pool sizes in nutrient-switching experiments . Much is known about the metabolic reactions that lead to the production of biomass and energy in cells . However , understanding the logic of metabolic regulation has been challenging due to the network's scale and complexity . Flux-balance analysis ( FBA ) , a constraint-based computational approach , has been used to show that some microorganisms , including E . coli , maximize their growth rates per molecule of carbon source consumed [1] . FBA uses mass conservation to predict optimal growth rates as well as fluxes [2] . In its simplest form , FBA assumes that cells regulate fluxes to produce biomass at the maximum rate possible given a particular limiting input flux . Recently , FBA has been successfully applied to additional microorganisms [3]–[5] , and to objective functions other then maximizing biomass [6] , e . g . maximization of ATP production [7] or minimization of metabolic adjustment in response to perturbations in metabolic network [8] . Attempts to include regulatory [9] , [10] , thermodynamic [11] , [12] , and environment-specific constraints have resulted in insights into the structure of metabolic networks , e . g . the organization of redundant pathways [13] , [14] . ( For a comprehensive list of FBA achievements see reviews by Kauffman et al , 2003 and Lee et al , 2006 ) . Despite their predictive strength and wide applicability , FBA-based methods are limited; FBA assumes that fluxes are optimal ( thereby assuming perfect regulation ) but does not reveal how these optimal fluxes are achieved . This leaves open the question: how can cells achieve nearly optimal fluxes for efficient growth ? Previously , some complex bio-molecular networks have been successfully analyzed and understood in terms of simple modules [15] , e . g . the eukaryotic cell cycle [16] , [17] . In the same spirit , we address the question of how to achieve optimal growth using several representative modules drawn from real metabolism . In particular we consider four modules , each of which captures an essential feature of the real metabolic network - i ) a linear pathway , ii ) a bidirectional pathway , iii ) a metabolic cycle , and iv ) integration of two different nutrient inputs . Linear pathways , in addition to being common , suggest simple rules for achieving optimal growth . In the second module , representing a bidirectional pathway , metabolites are interconverted , albeit at a cost , with the consequent risk of running a futile cycle ( e . g . , interconversion of fructose-6-phosphate and fructose-1 , 6-bisphosphate ( FBP ) ) . In the third module we analyze a metabolic assimilation cycle . A metabolic cycle can be visualized as a linear pathway where the end product is essential for the first step of the pathway . Two important examples of metabolic cycles are the TCA cycle and the glutamine-glutamate nitrogen-assimilation cycle . Finally , the fourth module addresses the problem of balancing two different inputs , carbon and nitrogen . This module takes into account the ability of microbes to assimilate nitrogen in the form of ammonium via an ATP-independent pathway or a higher affinity ATP-dependent one . When nitrogen is scarce , the ATP-dependent pathway is utilized , whereas when carbon is scarce , it is avoided . For regulation of these modules we invoke only product-feedback inhibition . Since its discovery in the late 1950's , product-feedback inhibition has become recognized as one of the cornerstones of metabolic regulation [18] , [19] . This form of regulation was first hypothesized by Novick and Szilard [20] for the tryptophane biosynthetic pathway from chemostat experiments , and has since been found in almost every biosynthetic pathway [21] . Product-feedback inhibition is a regulatory scheme in which the product of metabolism inhibits its own synthetic pathway . Remarkably , in all four of the modules studied , we find that simple product-feedback inhibition is sufficient to control fluxes so as to enable nearly maximally efficient growth . To test our understanding of the physiological role of product-feedback inhibition , we compared our simple models to actual regulation of the glutamine-glutamate nitrogen assimilation cycle , including its integration with carbon metabolism . We find important similarities between the product-feedback inhibition scheme that we propose based on general principles and the actual regulatory mechanisms present in E . coli . If , as we will argue , simple product-feedback inhibition is enough to achieve nearly optimal growth , why is real metabolic regulation so complex ? Metabolic feedback regulation exists at various levels , such as , control of enzyme mRNA transcription [22] , reversible enzyme phosphorylation [23] , non-competitive allosteric regulation [24] , and competition for enzyme active sites [25] . There are many cases where multiple feedback mechanisms work together , e . g . glutamine synthetase is regulated by a bicyclic cascade of covalent modifications and transcriptionally by the NtrC two-component system [26] . Our mathematical analysis suggests that simple feedback regulation , while adequate for flux control , could lead to large metabolite pools , and that accumulation of these pools may be prevented by multiple regulatory mechanisms working in concert to produce ultrasensitive feedback . Regulation of nitrogen assimilation in E . coli has been studied in great detail , perhaps more carefully than any other metabolic sub-network [25] , [35] , [36] ( see also cites in [25] ) . As nitrogen assimilation involves both a metabolic cycle and nutrient integration , it offers a chance to examine the extent to which actual metabolic networks , beyond the much studied linear or branched biosynthetic pathways , are regulated by feedback inhibition circuits of the sort that we hypothesize above . Our mathematical analysis of metabolic cycle and nutrient integration suggest a simple regulation scheme that allows near optimal steady-state growth . For the nitrogen assimilation GS/GOGAT cycle the analysis suggests feedback inhibition by glutamine and glutamate on GS and GOGAT , respectively . Feedback inhibition of GS by glutamine is well known . It does not involve standard allostery , but instead a bicyclic cascade of covalent modifications [37] . Interestingly , consistent with our suggestion that ultrasensitive feedback might be necessary for adequate control of metabolite pool sizes , it has been proposed that the purpose of this bicyclic cascade is to yield ultra-sensitive feedback [38] . Feedback inhibition of GOGAT by glutamate , in contrast , had not been explicitly considered until recent efforts at data-driven modeling of the network [25] . These efforts revealed that such feedback inhibition is essential to obtain models that match experimental data . Furthermore , examination of older literature reveals biochemical evidence for such feedback inhibition: glutamate and aspartate both inhibit GOGAT activity [39] . The effect of glutamate is an example of standard product inhibition of an enzyme , and was considered initially insignificant due to the high inhibition constant ( i . e . , the feedback is weak ) . However , given the large cellular pool size of glutamate ( mM ) , the high inhibition constant is appropriate ( indeed matching our expectation that large inhibition constant values are required to obtain near-optimal growth , with the associated consequence of large metabolite pool sizes ) . Aspartate is a direct product of glutamate , and provides further feedback essentially as a glutamate surrogate . For the ATP-independent nitrogen flux via GDH the analysis suggests feedback inhibition of GDH by the key nitrogen intermediates , glutamine and glutamate , which is again consistent with biochemical studies of purified GDH enzyme and with the existence of product inhibition of all enzymatic reactions [40] , [41] . A prediction from our analysis is that large changes in metabolite pools will occur upon the onset of nutrient limitation . This also agrees well with experimental observations . For example , consider the dynamics of -ketoglutarate and glutamine , the carbon skeleton and the most nitrogen-rich product of central nitrogen metabolism . -ketoglutarate is part of the TCA cycle , and many TCA cycle metabolites show similar patterns to its temporal response during nitrogen limitation and re-addition [25] . Accordingly , we consider the -ketoglutarate level as an indicator of available carbon ( specifically , carbon in the TCA cycle ) . Glutamine levels have been shown to correlate well with growth rate under nitrogen limitation [36] , and accordingly we consider glutamine levels to indicate available nitrogen . Fig . 3A shows the experimental metabolite pool size dynamics following nitrogen limitation and subsequent upshift for wild-type E . coli , as well as E . coli lacking the covalent modification enzyme responsible for feedback inhibition of glutamine synthetase ( GS ) by glutamine ( glnE ) . The steady-state metabolite pool sizes of the two strains are nearly identical before the nitrogen upshift; however , upon nitrogen upshift , the fold changes in both -ketoglutarate and glutamine are amplified in the feedback-defective strain compared to the WT strain . Moreover , after the nitrogen upshift , large amounts of extracellular amino acids , including glutamine and glutamate , were measured in cultures of the feedback-defective strain consistent with unregulated nitrogen assimilation ( Fig . 3 in Text S1 ) . These observations are consistent with simulations based on our simple feedback model ( Fig . 3B ) . Furthermore , we find metabolite pool dynamics observed under nitrogen-limited growth to also be consistent with our model [42] ( see Text S1 ) . Within our model , the WT strain is described by the module with all three feedbacks present ( Fig . 2D ) , while the feedback-defective strain is described by the same basic module but without the feedback on carbon-dependent nitrogen input flux . As a simulation of the experiment , we started the two modules at steady state in the nitrogen-limited regime , and suddenly increased nitrogen availability by simultaneously increasing the two nitrogen maximum input fluxes and , thereby shifting the modules to the non-nutrient limited ( -limited ) regime . To achieve steady state in our model for the feedback-defectve strain , we assumed a leakage flux for the large nitrogen intermediate pool ( see Text S1 for equations with leakage ) . Some of the system's dynamics , in particular the overshoot of glutamine in the wild-type strain , are not captured by our simple feedback model . Generally , time-delay in the feedback may result in an overshoot in a feedback-inhibited system . This is consistent with the specific implementation of feedback by glutamine on GS: a cascade of covalent modification reactions which occur on the min timescale , with the overshoot occurring in the period where nitrogen assimilation outraces the feedback mechanism . We also compared the growth rate response of the wild-type and feedback-defective strains to relief of nitrogen limitation . Consistent with experimental results , the simulations predicted a bigger increase in the growth rate in the WT strain than in the feedback-defective strain following nitrogen upshift ( Fig . 3C , D ) . In the simulation , the reason for the slower growth in the feedback-defective strain post nitrogen up-shift is excessive drainage of the carbon metabolite pool ( e . g . , -ketoglutarate ) by unregulated nitrogen uptake in the feedback-defective strain . Whether such drainage of a valuable carbon species is the real reason in live cells is not clear , however . An alternative possibility is that the excessive accumulation of glutamine causes osmotic imbalance . This alterative , while not quantitatively included in our model , is nevertheless consistent with the role of feedback inhibition as a homeostatic regulatory mechanism . Understanding metabolism and its regulation have long been central goals of biochemistry . Recently , flux-balance analysis ( FBA ) , a constraint-based computational approach , has been used to predict the optimal metabolic fluxes and growth rates of microorganisms in different environments . In several cases , in particular involving E . coli , the FBA-predicted optima agree remarkably well with experiments [1] , [34] , raising the question “for cells to realize optimal growth how complex must metabolic regulation be ? ” We have addressed this question using a set of representative metabolic modules . We find that , in all the cases studied , simple product-feedback inhibition is enough to achieve nearly optimal growth . Furthermore , the divergence from optimality becomes arbitrarily small as the feedback-inhibition constants are increased . An important trade-off is that larger inhibition constants result in larger pool sizes of non-growth-limiting metabolites , which can be detrimental to growth . However , ultrasensitive feedback mechanisms ( i . e . those with high Hill coefficients ) can substantially restrict these pool sizes; the higher the Hill coefficient of the feedbacks , the smaller the increase in pool size required to achieve the same degree of inhibition . This suggests that the need for ultrasensitive mechanisms to control metabolite pool sizes may account for some of the complexity found in metabolic regulation in real cells at both the transcriptional and post-transcriptional levels . Can we hope to gain insight into real metabolism using the very simple models we studied ? To address this question we examined the nitrogen assimilation network in E . coli , which involves both nutrient integration and a metabolic cycle . First , we found the feedback regulation scheme proposed by our mathematical analysis of representative modules aligns closely with the known regulation of the network . Second , we found reasonable agreement between simulations based on our simple feedback models and actual experimental results , for both wild type and feedback-defective E . coli . Comparing strains with different regulatory schemes allowed us to directly ask the question “is product-feedback inhibition essential for achieving optimal growth ? ” At least in the case of nitrogen up-shift , both our simulations and experimental data argue that it is: the feedback-defective strain grew substantially slower than wild type after the up-shift . One of the central predictions of our feedback framework is that pool sizes will be large for non-growth-limiting metabolites . Since few metabolites are growth-limiting under any nutrient condition , the cells are likely to have large pools of multiple metabolites under a wide range of conditions . Therefore , we need to consider the possible impact of large pool sizes on cell physiology . Can large sizes of metabolite pools be detrimental to the well-being of cells ? In fact , many metabolic intermediates , such as glyoxylate and formaldehyde , are toxic at high concentrations . Even the biosynthetic end-products required for growth ( e . g . amino acids , nucleotides , etc . ) can be detrimental to a cell's growth at high enough concentrations . Metabolites at high concentration can interact nonspecifically with various enzymes and disrupt metabolic reactions [43] . Furthermore , metabolite pools contribute to intracellular osmolarity and consequently to the osmotic pressure inside cells . Dedicated mechanisms to respond to osmotic stress have evolved in microorganisms , reflecting the harmful effects of osmotic imbalance [44]–[46] . For E . coli , the growth rate is maximized in conditions corresponding to external osmotic pressures of around atm [45] , [46] . Furthermore , the turgor pressure has been estimated to be around atm [47] in an AFM study of the magnetotactic Gram negative bacteria Magnetospirillium gryphiswaldense . Consequently , the internal osmotic pressure is thought to be around atm , which corresponds to an effective concentration of mM of solute . Recent measurements have revealed that some metabolite pools can become very large , such as fructose-1 , 6-bisphosphate ( mM ) and glutamate ( mM ) [48] . These large metabolite pools could contribute significantly to the overall internal osmotic pressure of the cells . In general , pools that are large even when growth-limiting will potentially be very large when non-growth-limiting and may cause osmotic imbalance . Such pools in particular may require ultrasensitive feedback mechanisms to restrict their sizes . Experimental manipulation of feedback sensitivities ( e . g . by enzyme mutation , knockout of enzymes involved in covalent modification cascades , etc . ) should help shed light on the role of ultrasensitive feedback mechanisms . Ultrasensitivity is a common feature of feedback inhibition . At the transcriptional level , multiple promoter binding sites along with other cooperative mechanisms like DNA looping yield ultrasensitive responses [49] ( Fig . 4A ) . The response time for transcriptional feedback is limited by protein degradation ( and dilution ) , which in microorganisms is typically of the order of tens of minutes to hours . Metabolite-pool sizes , on the other hand , may change in just few seconds , e . g . the glutamine pool increased by over 10-fold in seconds in the nutrient-switching experiment described above . The fast dynamics of metabolite-pool sizes suggests the need for fast feedback mechanisms . Fast regulation can be realized through various post-translational mechanisms – allosteric regulation of protein aggregates , e . g . ATP molecules bind cooperatively to a homodimer of pantothenate kinase [50] , competition , e . g . Wee1 regulation of Cdk1 [51] , or covalent modifications , e . g . reversible phosphorylation of isocitrate dehydrogenase [52] and the bicyclic cascade of covalent modifications in glutamine regulation [26] ( Fig . 4B ) . Thus , the need for fast ultrasensitive feedback mechanisms may be a key driver of the observed complexity in metabolic regulation . Our study of simple representative metabolic modules is an attempt to identify the design principles underlying the regulatory mechanisms that optimize metabolic function , such as biomass production [53] . In addition to highlighting general lessons in metabolic regulation , our analysis raises new fundamental questions . How many feedbacks are required in a metabolic network , in particular the metabolic network of a real cell ? What principles , in addition to optimal growth and stability , guide the evolutionary selection of feedbacks and feedback mechanisms ? Has the complexity and dynamics of the cellular environment led to additional constraints on feedback strategies ? And finally , given the apparent sufficiency of feedback inhibition , why are other regulatory motifs , such as allosteric enzyme activation , also found in metabolism ? Further experiments in which metabolic feedbacks are eliminated , modified , and/or rewired , in concert with additional theoretical analyses , should facilitate answering these questions . The analyses were carried out using kinetic equations ( Eqs . 3 , 5 , 6 , 7 ) . The equations account for the concentration of each component in the metabolic modules and the steady-state solutions were numerically obtained using MATLAB . For details on the flux-balance analysis ( FBA ) see Text S1 .	Bacteria live in remarkably diverse environments and constantly adapt to changing nutrient conditions . Recent evidence suggests that some bacteria , such as E . coli , are extraordinarily efficient in producing biomass under a variety of different nutrient conditions . This observation raises the question of what physical mechanisms enable such efficiency . Here , we propose that simple product-feedback inhibition by itself is capable of leading to such optimality . Product-feedback inhibition is a metabolic regulatory scheme in which an end product inhibits the first dedicated step of the chain of reactions leading to its own synthesis . Our mathematical analysis of several representative metabolic modules suggests that simple feedback inhibition can indeed allow for optimal and efficient biomass production . However , the effectiveness of simple product-feedback inhibition comes at the cost of high levels of some metabolite pools , potentially associated with toxicity and osmotic imbalance . These large metabolite pools can be restricted if feedback inhibition is ultrasensitive . We find that the feedback regulation scheme suggested by our mathematical analysis closely aligns with the actual regulation of the nitrogen assimilation network in E . coli and is sufficient to explain much of the dynamical behavior of relevant metabolite pool sizes seen in experiments .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "computational", "biology/metabolic", "networks", "microbiology" ]	2010	Achieving Optimal Growth through Product Feedback Inhibition in Metabolism
The pathway for RNA interference is widespread in metazoans and participates in numerous cellular tasks , from gene silencing to chromatin remodeling and protection against retrotransposition . The unicellular eukaryote Trypanosoma cruzi is missing the canonical RNAi pathway and is unable to induce RNAi-related processes . To further understand alternative RNA pathways operating in this organism , we have performed deep sequencing and genome-wide analyses of a size-fractioned cDNA library ( 16–61 nt ) from the epimastigote life stage . Deep sequencing generated 582 , 243 short sequences of which 91% could be aligned with the genome sequence . About 95–98% of the aligned data ( depending on the haplotype ) corresponded to small RNAs derived from tRNAs , rRNAs , snRNAs and snoRNAs . The largest class consisted of tRNA-derived small RNAs which primarily originated from the 3′ end of tRNAs , followed by small RNAs derived from rRNA . The remaining sequences revealed the presence of 92 novel transcribed loci , of which 79 did not show homology to known RNA classes . Trypanosoma cruzi is a protozoan parasite and the causative agent of Chagas' disease , which has substantial health and socioeconomic impact in Latin America [1] . Treatment is currently restricted to a small number of drugs with insufficient efficacy and potentially harmful side effects . The genome of T . cruzi strain CL Brener is complex in terms of sequence repetitiveness and is a hybrid between two diverged haplotypes , named non-Esmeraldo-like and Esmeraldo-like: we refer to them here as non-Esmeraldo and Esmeraldo . Taken together , both haplotypes [2] sum up to approximately 110 Mb distributed over at least 80 chromosomes [3] . Similar to other trypanosomatids , genes are organized into co-directional clusters that undergo polycistronic transcription . Gene rich regions are frequently interrupted by sequence repeats , which comprise at least 50% of the genome [2] . Several gene variants occur in tandemly repeated copies which often collapse in shotgun assemblies [4] . The genome of a different , non-hybrid strain named Sylvio X10 was recently sequenced and partially assembled , showing a core gene content highly similar to CL Brener [5] . The T . cruzi life cycle is complex and consists of several distinct life stages , morphological states and hosts [1] . To achieve successful completion of the life cycle , the parasite must rapidly adapt to different environments by regulating its gene expression [6] . Transcription in T . cruzi often , but not exclusively , starts at strand switch regions [7] , [8] , where long transcripts are produced by RNA polymerase II and matured via trans-splicing and polyadenylation [9] . There is to date no definite model of how and if transcription is regulated , as RNA polymerase II promoters for protein-coding genes have not been identified . Thus , it is thought that gene expression is mainly regulated at the post-transcriptional level [9] . RNA interference ( RNAi ) and related pathways are widespread in animals and other metazoans , participating in a wide range of cellular processes; from chromatin organization to silencing of genes and selfish elements . RNAi relies on small RNA molecules , approximately 20–30 nucleotides in length , to trigger target silencing . In eukaryotes , several different types of small RNAs have been identified . Of these , the best characterized are microRNAs ( miRNAs ) and small interfering RNAs ( siRNAs ) . See Table 1 for a summary of small RNAs discussed in this study . Two proteins are required for small RNA biogenesis and function: Dicer and Argonaute . Among protozoan parasites , the RNAi machinery has either been lost or retained . T . cruzi have lost the canonical RNAi machinery , which has been confirmed both functionally [10] and from the genome sequence [2] , although RNAi is functional in certain other trypanosomatid species [11] , [12] , [13] . In the African trypanosome Trypanosoma brucei , convincing evidence has shown the presence of an active RNAi machinery ( see [14] for references ) and more recently pseudogene-derived small RNAs , which were reported to suppress gene expression [15] . A similar situation has been observed in the Leishmania genus . Leishmania braziliensis possesses a functional RNAi-pathway [16] , whereas other members of this genus do not [12] . Analyses of the T . cruzi genome have revealed lack of both Dicer and Argonaute homologs . However , similar to other trypanosomatids , T . cruzi possess a protein with a solo Piwi domain , but without a PAZ domain . The biological role of this protein is presently unknown , although it has been suggested to represent a member of a novel Argonaute subfamily [17] . To date , little is known about the presence of small non-coding RNAs ( sncRNAs ) in trypanosomatids , which do not depend on RNAi . Recently , one study reported the prediction of sncRNAs in the trypanosomatids [18] , providing evidence of yet uncharacterized sncRNAs in these species . However , comparative genomics suffers from the limitation that it does not facilitate identification of species-specific small RNAs or regulatory elements . Furthermore , another study described small-scale sequencing of small RNAs in T . cruzi , and reported a population of tRNA-derived small RNAs , which was linked to cellular stress [19] . Moreover , studies from another unusual eukaryote , Giardia lamblia , have shown that sncRNA can be highly diverged from metazoan sncRNA [20] , [21] and therefore escape detection using homology searches . Lessons from non-protozoans have taught that novel sncRNAs are often in low abundance and avoid detection using conventional techniques [22] , which do not sequence deep enough to capture the full complexity of the transcriptome . In order to obtain a more complete picture of the short T . cruzi transcriptome , we have performed unbiased deep sequencing and genome wide analyses of the short transcriptome from T . cruzi epimastigotes . The data indicated the existence of an abundance of small RNAs derived from non-coding RNAs and a number of novel expressed loci in the genome . The sequences have been deposited in the DNA Data Bank of Japan under the accession number DRA000396 and are also available for download from http://www . ki . se/chagasepinet/ncrna . html . Epimastigotes from T . cruzi CL Brener were grown exponentially at 28°C in liver infusion tryptose ( LIT ) media [23] supplemented with 10% FBS ( Gibco ) and streptomycin/penicillin ( Gibco ) , pH 7 . 3 . Total RNA was extracted using the TRIzol method ( TRI Reagent , Sigma ) following manufacturers' instructions . The total RNA was converted to cDNA using a standard protocol and size fractioned using a polyacrylamide gel . The sequencing library was generated according to the manufacturers' instructions and sequenced with a 454 instrument ( GS20 FLX ) . The sequence data was stripped of the 3′ ‘CCA’ extension and aligned with the T . cruzi genome assembly [3] using the Burrows-Wheeler Aligner ( BWA ) [24] . BWA was configured to allow up to two mismatches . Repetitive elements were identified using RepeatMasker [25] and RepBase [26] ( r16 . 01 ) . Only repeats longer than 40 bp were considered . Identification of novel expressed genomic loci was performed using clustering of reads which could not otherwise be assigned an identity . Clustering was done on reads satisfying the following criteria: i ) they do not have an annotation ( i . e . tRNA , etc ) ; ii ) have only one valid alignment in the genome ( single mapping ) ; iii ) have an overlap of at least one base with another read . Subsequently , the resulting clusters were filtered using the following criteria: i ) a cluster should contain at least six reads ii ) at least two reads should be unique . The resulting novel non-coding RNAs were manually examined and assigned putative identities using BLAST . A sequence database of trypanosomatid genes was established by extracting sequences containing ‘trypanosoma’ in the header line from the GenBank non-redundant database . Statistical evaluation and charts were performed using the R platform . Homology searches were done using NCBI BLAST . Prediction of tRNA secondary structure was performed using tRNAScan-SE [27] and visualized using VARNA [28] . Prediction of secondary structures of novel small RNAs was done with Vienna RNA [29] . Analysis of putative microRNA targets was performed using TargetScan [30] and GoTermMapper [31] . Scripts were written in Perl and are available on request . Stem-loop real time-PCR experiments and primer design were performed as described in [32] . The quality of RNA samples was assessed on an 1 . 5% TBE-agarose gel . All RNA samples were treated with DNAse I ( Fermentas ) previous to the reverse transcriptase reaction . SnoRNA or 5S rRNA was used as reference RNA for qualitative/quantitative experiments . The following reagents were mixed and subjected to Pulsed RT reaction; 60 ng of T . cruzi RNA ( per RT reaction ) , 0 . 5 mM dNTP , 10X First Strand Buffer , 5 mM MgCl2 , 10 mM DTT , RNAseOUT , 50 units Superscript III RT ( Invitrogen ) and 1 µl of SL-RT specific primer , using the conditions: 30 min at 16°C followed by 60 cycles of 30 s at 30°C , 30 s at 42°C , 1 s at 50°C and a final step of 5 min at 85°C . RNAse H was added and incubated for 20 min at 37°C . The real time PCR reactions were performed in triplicates using 1 µl cDNA , 300 nM forward specific and reverse universal primers and 2X SYBR Green Master Mix ( Roche ) . Cycling conditions: 5 min at 95°C , 40 cycles 95°C-5 s , 60°C-20 s , 72°C-1 s followed by dissociation curve analysis in Strategene Thermal Cycler Mx3000P . Experiments were repeated at least twice per each of two biological samples . Ct values were normalized against 5S rRNA values and the abundance ratio was calculated for each individual Ct value and mean as well as standard deviation were calculated and graphed using SigmaPlot 9 . 0 . The 5S rRNA ( Tc00 . 1047053509455 . 160 ) and C/D snoRNA ( Tc00 . 1047053510739 . 50 ) were used for normalization . Primers used are listed in Table S3 . An epimastigote cDNA library was size fractioned on a polyacrylamide gel and sequenced using 454 sequencing [33] ( Materials and Methods ) . Sequencing resulted in a total of 582 , 243 reads ( 101 , 284 unique ) with a size range between 16 to 61 nucleotides ( Figure 1A ) . The median sequence length of the library was 38 nt . A total of 12 . 2% ( 71 , 309/582 , 243 ) reads occurred as single copy , whereas the remaining reads had a variable copy number between 2 and 41 , 929 . The selected size range should contain only non-coding RNA ( ncRNA ) , as there are no known protein-coding genes in this size range . Further , this size range was selected to avoid spliced leader RNAs . However , degradation products from transcriptional turnover could be present in the sample . Based on two observations we conclude that degradation products were not contaminating the library; i ) degradation fragments should exhibit a random distribution pattern in protein-coding genes , which was not the case , ii ) ribosomal RNA constitute the bulk ( >80% ) of cellular RNA , which was not observed in the sequence data . The sequence data was separately aligned with each of the T . cruzi CL Brener haplotypes; non-Esmeraldo and Esmeraldo ( Figure 1BC , Materials and Methods ) . In addition , reads were aligned with a 38 million base pair collection of unassigned contigs ( Figure 1D ) , which mostly consists of repeats [3] , [34] . This resulted in a total of 90 . 7% aligned reads ( 528 , 228/582 , 243 ) , or expressed in unique reads , 74 . 0% aligned reads ( 75 , 024/101 , 284 ) ( Table 2 ) . Slightly more reads were aligned with non-Esmeraldo , owing to the more complete status of this haplotype assembly compared to Esmeraldo; however the length distribution of the aligned reads were similar ( Figure 1BC ) , indicating both haplotypes might generate similar RNA populations . A total of 9 . 2% ( 53 , 646/582 , 243 ) of the reads could not be aligned with the genome using the default alignment procedure , raising the question if these reads are biologically derived or technical artifacts . The following scenarios are possible; i ) unaligned reads are technical artifacts or enriched with sequencing errors , ii ) unaligned reads represent small RNAs derived from unfinished parts of the genome sequence , iii ) small RNAs have been subjected to chemical modification and RNA editing events . As the T . cruzi CL Brener genome sequence is not complete [2] , [4] it remains possible that at least some small RNAs are derived from unassembled regions . To investigate this , unaligned reads were mapped to the genomic shotgun reads from the genome project , which provided alignment to 0 . 49% ( 2860/53 , 646 ) of the unaligned reads . Examination of a limited number of reads , that failed alignment , found homology to tRNALys . As these reads occurred in a high copy number ( ∼300 ) and mismatches were located in the anticodon loop , this makes it possible that mismatches are not sequencing errors but rather modified nucleosides misinterpreted by the sequencer . In order to differentiate between known and unknown RNA species in the library , we categorized reads into classes using genome annotations . Alignment coordinates were superimposed on genome annotations and each read was categorized into one of the categories in Table 3 if it completely or partially overlapped with the annotation . In cases where a tRNA , snRNA or snoRNA was overlapping a protein-coding gene , the ncRNA gene was preferentially selected . To further improve the classification , reads without annotation were queried against a database of various trypanosomatid sequences ( Materials and Methods ) . For reads with a single alignment location ( single mappers ) , 97 . 4% ( 378 , 446/388 , 551 ) of the reads in non-Esmeraldo and 96 . 7% ( 157 , 280/162 , 622 ) in Esmeraldo were found to correspond to small ncRNAs ( sncRNAs ) derived from tRNA , rRNA , snRNA and snoRNA ( Table 3 ) . tRNA-derived small RNAs ( tsRNA ) was found to be the most abundant type in the library , composing at least 65 . 3% ( 380 , 191/582 , 243 ) of the total sequence data , which we further describe in the next section . This result suggests that the vast majority of small RNA species in T . cruzi epimastigotes are derived from known ncRNA classes . About 2–5% of the aligned sequences could not be classified into known ncRNA classes . It should be noted that this fraction might not represent the entire abundance of novel sncRNA in T . cruzi , as some sncRNA might only be present in a specific life stage or under a certain physiological condition . Furthermore , long novel ncRNAs ( >61 nt ) could exist [18] , as was for example reported in Leishmania infantum [35] . A total of 19 , 893 reads aligned with unassigned contigs , out of which 78% ( 15 , 506/19 , 893 ) represented reads that aligned with rRNA genes ( Table 3 ) . A minor fraction consisted of reads that aligned with tRNA ( 13% ) and snRNA/snoRNA ( 1% ) . This is consistent with the fact that few rRNA genes have been properly assembled [2] , [3] . For both non-Esmeraldo and Esmeraldo , a total of 69 . 1% ( 282 , 036/408 , 008 ) of the small RNAs were assigned to the tRNA category ( considering single mapping reads ) , despite the fact that the library was size selected for sequences shorter than 61 nt and mature tRNAs are between 70–80 nt . A closer inspection revealed the presence of tRNA-derived small RNAs ( tsRNAs ) , a phenomenon reported previously in higher eukaryotes [36] , [37] and lower eukaryotes [38] , [39] , [40] . However , the physiological role , if any , of tsRNA is not well defined ( for review and discussion see [41] , [42] , [43] , [44] ) . T . cruzi tsRNAs were first reported by Garcia-Silva et al . [19] , who found tsRNA to be recruited to cytoplasmic granules and increase under stress conditions . The authors employed a 20–35 nt cDNA library and sequenced 348 clones , and found that 26% of the clones were derived from tRNA and 60% from rRNA . The study also showed a higher representation for 5′end tRNA derived small RNAs , which may be explained by the relatively low number of clones sequenced in this study . In our library , tsRNA had a median length of 38 nt and 88 . 9% ( 250 , 920/282 , 036 ) were derived from the 3′ end of tRNAs ( Figure 1E , Figure 2 , Table 4 ) . Moreover , 75 . 3% ( 189 , 116/250 , 920 ) of the 3′-derived reads contained a ‘CCA’ nucleotide extension; indicating that the majority of 3′ tsRNA are derived from mature tRNA species , as the ‘CCA’ addition is post-transcriptionally added in eukaryotes . However , we cannot rule out that the remaining reads did not lose the ‘CCA’ extension during sample or library preparation . The median length of 38 nt is consistent with the current view of bisectional cleavage of mature tRNA . Despite this , we also identified shorter tsRNA ( <25 nt ) albeit in lower frequency; a total of 1605 tsRNA were 24 nt or less and primarily derived from tRNAGlu , tRNAAsp , tRNATyr , tRNAVal and tRNAArg ( Figure S1 ) . Interestingly , the shorter tsRNA were more often derived from the 5′ arm . Most tRNA isoacceptors were found to be precursors for tsRNA , but with relative different amounts ( Figure S1 ) . The most abundant tsRNA were derived from the 3′ arm of tRNAHis and occurred in 41 , 929 copies and contained the ‘CCA’ extension ( Table 5 , Figure S2 ) . Interestingly , the 3′/5′ ratio of tsRNA was not equal for all tRNA isoacceptors ( Table 4 ) . For example , tRNAGln showed more tsRNA derived from the 5′ arm . A recent study reported the cloning and characterization of tsRNA in the primitive eukaryote Giardia lamblia ( G . lamblia ) [38] , showing that tsRNA are abundantly expressed during the encystation stage and are ∼46 nt long . Consistent with T . cruzi tsRNAs , G . lamblia tsRNAs are derived from most tRNA isoacceptors and predominantly from the 3′ arm . In G . lamblia , tsRNAs from tRNAAsp and tRNAGly were the most frequently cloned , which may indicate species or life stage specific isoacceptor preference . If tsRNA would represent degradation products from tRNA-turnover , it would be expected to find a correlation between the RNA fragment amount and the expression levels of tRNA genes . In the absence of tRNA expression data , we utilized the amino acid usage from the predicted proteome and compared it with the observed tsRNA expression . We found no correlation between the observed tsRNA expression and the amino acid usage ( Pearson's correlation , r = −0 . 05 ) , nor was there a correlation between the genomic copy number of tRNA and tsRNA expression ( Pearson's correlation , r = 0 . 08 ) , suggesting that T . cruzi tsRNAs are not random degradation products from tRNA turnover . As we observed a very high expression of tsRNA from certain tRNA isoacceptors ( e . g . tRNAHis , tRNAArg and tRNAThr ) , but almost no expression from others ( tRNAPhe and tRNAAsn ) , this implies tsRNA are differentially expressed in T . cruzi . Furthermore , we performed experimental validation of a few selected tsRNA ( Figure S3 ) . Consistent with previous reports [36] , [37] , [38] , [39] , we found that the cleavage site was present within the anticodon loop of mature tRNAs ( Figure S2 ) . For shorter tsRNAs , the cleavage site was present in the two other loops , but primarily in the loop of the T-arm . This suggests endonucleolytic cleavage as the responsible mechanism behind tsRNA generation . The precise cleavage supports the idea that tsRNA are generated by a distinct mechanism rather than random degradation . However , as shorter tsRNA were observed , these might require both endonucleolytic cleavage and exonucleolytic trimming in their biogenesis pathway . We observed tsRNA with and without a CCA 3′ terminus , thus the process of tsRNA formation likely targets both pre-tRNAs and mature tRNAs , and therefore takes place either in the cytosol or nucleus , as only mature tRNAs are imported into the mitochondria [45] . An early study by Zwierzynski et al . reported 3′ CCA activity in nuclear extracts [46] , raising questions about the subcellular location of tsRNA biogenesis . The key enzymes involved in tsRNA biogenesis remain to be identified; however , it remains clear that this mechanism is independent of Dicer . It has been hypothesized that tsRNAs inhibit protein synthesis either by depleting the cellular tRNA pool or by a more intrinsic mechanism involving a protein repression complex [43] , albeit there is to date no definite evidence . tsRNAs have been associated with Piwi and Argonaute complexes [47] , [48] , [49] , suggesting that it may guide degradation of target transcripts in RNAi-positive organisms . A recent study reported tsRNAs to guide tRNase Z-mediated cleavage of engineered target sequences and possibly endogenous transcripts [50] , which further supports the idea of these species as functional entities . Small nucleolar RNAs ( snoRNAs ) are present throughout eukaryotes and guide enzymatic modifications of target RNAs in the nucleolus , and can be subdivided into C/D and H/ACA classes based on sequence motifs . Recently , snoRNA-derived small RNAs ( sdRNA ) have been reported in animals [37] , [51] , [52] and in the protozoan G . lamblia [53] and are thought to be generated by a Dicer-dependent mechanism [52] . Metazoan sdRNAs are predominantly ∼17–19 nt and ∼30 nt and generated either from the 5′ ( C/D type snoRNAs ) or 3′ ends ( H/ACA type snoRNAs ) [52] . In both humans and G . lamblia snoRNA-derived small RNAs have been implicated to have miRNA-like functions [53] , [54] . We found that 0 . 26% ( 1413/528 , 228 ) of the total data was represented by snoRNA-derived small RNAs , with a median length of 35 nt , similar for both C/D and H/ACA ( Figure 1G , Figure 2 ) . The observed length of sdRNA is different from metazoan sdRNA and both types were found to have similar number of reads ( n = 770 and n = 643 reads for C/D and H/ACA snoRNA respectively ) . We did not observe the positional bias towards the 3′ end which has been reported for mammalian sdRNA , or a specific alignment pattern suggestive of regulated cleavage . These findings suggested that the observed sdRNAs were generated by a different mechanism compared to those found in metazoans , or less interestingly , represent degradation or break-down products . A total of 0 . 53% ( 2839/528 , 228 ) reads were derived from small nuclear RNAs ( snRNAs ) which were distinct from snoRNAs , with a median length of 40 nt . Interestingly , 82 . 1% ( 2333/2839 ) of the snRNA derived small RNAs were from snRNA U4 and U5 . Two snRNA-derived reads occurred in a high copy number ( ∼100 copies ) . Small RNAs derived from ribosomal RNA have received less attention but are known to exist [37] , [40] and have been reported to increase as a response to oxidative stress . Here , 17 . 2% ( 91 , 206/528 , 228 ) of the aligned sequences represented small RNAs derived from ribosomal RNAs ( rsRNAs ) . rsRNAs could be grouped into three different subpopulations based on their length distribution ( Figure 1F ) ; one population with an average length of 20 nt , a second population with an average length of 33 nt , and a third longer population with an average length of 46 nt . Complete rRNA genes are not present in the current assembly [2] , [3] and it is therefore difficult to conclude if the small RNAs represent degradation products or not . However , the copy number of rRNA-derived small RNAs was highly variable; ranging from 1 ( n = 6337 reads ) to >100 ( n = 117 reads ) , which suggests a mechanism of non-random degradation . A total of 1 . 69% ( 8964/528 , 228 ) of the aligned reads were not derived from known tRNA , rRNA , snRNA , snoRNA or repeats , of which 17 . 4% ( 1565/8964 ) aligned with protein-coding genes and the remaining with intergenic regions ( Figure 1HI ) . In order to find novel ncRNAs , we performed clustering of reads with overlapping alignments ( Materials and Methods ) . These criteria formed 92 loci , consisting of a total 7805 reads ( Table 6 , Table S1 ) , of which 13 loci were identified as known non-coding RNAs using homology searches , which have been missed in the present genome annotation . The remaining 79 loci did not fall into known ncRNA classes and had an average length of 54 nt . None of these had homology to any known RNA class in Rfam or GenBank , albeit seven displayed partial sequence similarity with protein-coding genes and pseudo genes . We performed secondary structure prediction [29] of these unknown RNAs; 26 did not fold at all , 35 folded into non-hairpin structures and 18 folded into hair-pin structures according to predictions . Next we compared the 79 candidates to ncRNAs previously reported from comparative genomics [18] , but failed to find overlap between the two sets of candidates . This result does not exclude the possibility that the previously reported ncRNA are correct , as only 20% ( 15/72 ) was in the size range of our library . Finally we queried our 79 novel ncRNA candidates against other trypanosomatid genomes ( T . brucei , T . vivax , T . congolense , Leishmania spp . ) to test if these sequences are conserved among other trypanosomatids; however , no full length matches were found . These findings suggested that novel RNAs , as identified here , are specific for T . cruzi rather than ubiquitous among trypanosomatids . The remaining 1159 reads did not pass the criteria for clustering and had a median length of 24 nucleotides . These reads were subsequently queried with BLAST against a trypanosomatid sequence database ( Materials and Methods ) ; 335 reads displayed homology to trypanosomatid rRNA genes and 819 with homology to protein-coding genes . For reads with alignment to protein-coding genes we observed no statistical overrepresentation of antisense alignments , and as these did not derive from known ncRNA , the following scenarios are possible; i ) small RNAs with homology to protein-coding genes are spurious transcriptional products , or debris from mRNA turnover , without biological significance , ii ) small RNAs with homology to protein-coding genes are a result of regulated or non-regulated mRNA turnover with biological significance , iii ) small RNAs with homology to protein-coding genes are transcribed from the genome and not derived from mRNA . To address these questions , functional studies will be needed to answer whether these small RNAs are biologically active or debris from the normal cellular turnover . MicroRNA ( miRNA ) is a class of regulatory small RNAs that fine tune gene expression in metazoans . One attractive hypothesis is that intracellular parasites utilize the host microRNA pathway to change the cellular environment for its own needs . Partial evidence exists from Cryptosporidium parvum and Toxoplasma gondii that this may take place [55] , [56] , [57] . None of the small RNAs showed complete or partial homology when compared with human microRNA sequences from [58] . Next , we performed putative target site prediction of the 819 small RNAs . The putative ‘seed region’ ( nt 2–8 ) was extracted from each of the 819 small RNAs and queried using standalone TargetScan against 23-way UTR alignments . A conserved target site was required to be present in the following 7 genomes; Homo sapiens , Mus musculus , Rattus norvegicus , Gallus gallus , Macaca mulatta , Pan troglodytes and Canis lupus familiaris . As a result , a total of 3230 putative target genes were identified . Subsequently , a slimmed gene ontology was used to group the identified genes into a more narrow set of categories . Interestingly , 33% ( 1063/3230 ) of the genes grouped into ‘cellular nitrogen compound metabolic process’ ( GO:0034641 ) , raising the possibility that parasites may modulate the immune response by interfering with the host production of nitric oxide . Furthermore , ‘immune system process’ ( GO:0002376 ) contained 7 . 7% ( 250/3230 ) of the genes . One hypothesis derived from this bioinformatic prediction is that T . cruzi manipulates the host cell environment by secretion of oligonucleotides that mimic human microRNAs . Repeats are an inherent feature of most eukaryotes and have been attributed as an important driving force behind genome evolution [59] . T . cruzi have a significant part of its genome devoted to repeats; inactive and active retrotransposons , microsatellites and large gene families , often arranged in tandem . At least two types of non-Long Terminal Repeat ( LTR ) retrotransposons , designated CZAR and L1 , are potentially active in the T . cruzi genome [60] . The CZAR element consists of two open reading frames and represent a site-specific retrotransposon that inserts into spliced-leader genes [60] . Small RNAs have been implicated in the protection against retrotransposons in both metazoans and protozoa [22] . However , it is presently unknown how RNAi-negative protozoa , such as T . cruzi , protect themselves against the potentially disruptive effects of transposition events . This intriguing question motivated us to look for evidence of small RNAs that target or transcribe from retrotransposons and other repeats . Initially , the T . cruzi genome was searched with RepeatMasker [25] in combination with RepBase [26] to identify all known instances of mobile elements and satellite repeats , which resulted in 13 different types of repetitive elements covering 11–12% of the genome ( Table S2 ) . Twenty base pairs flanking each side of a repeat instance was included . To add more confidence to the analysis , we decided to maximize the number of useable reads by including those that go to multiple locations ( multi mapping reads ) . We used a similar approach to what was described in [61] , where a particular read was allowed mapping to more than one location , but only to one type of element . Reads going to more than one type of element or outside of repeats were removed . This resulted in a total of 0 . 13% ( 782/582 , 243 ) of reads from the library that aligned with various repetitive elements ( Table S2 ) . This suggests that if any of these small RNAs have a role to inhibit or block transposition events , these are present in a very low amount . We found that CZAR contained the highest amount of aligned reads ( n = 446 ) , despite the fact that this element only covered 0 . 21% of the genome . Several instances of the CZAR element were found to have reads in the 5′ or 3′ termini , or in the close vicinity . As reads were mostly found to align sense , these may represent initiation fragments from the transcription of these elements , supporting the idea that at least some CZAR elements are actively transcribed in the genome . The SIRE and TcVIPER have been suggested to represent two classes of dead elements [60] . A low number of reads aligned with TcVIPER ( n = 25 ) and SIRE ( n = 2 ) , possibly suggesting that some transcription of these elements might occur despite their inability to transpose . TcSAT1 is a ∼200 bp satellite repeat and comprises ∼5% of the current draft genome sequence [62] . Conflicting data exist regarding the transcription of TcSAT1 , where Northern blot hybridization experiments indicated no transcription , whereas nuclear-run-on assays and microarrays indicated active transcription ( see [62] for references ) . We found 150 reads aligned with this repeat element , which may represent degradation fragments or small RNAs derived from longer transcripts . Overall , we observed no overrepresentation of antisense reads in any class of repeat elements . However , it is possible that an antisense inhibitory mechanism is present albeit in a very low abundance , which would require deeper sequencing and a more narrow size fraction to be captured . Finally , it is also possible that T . cruzi does not use small RNAs to control transposition . We validated the presence of 12 small RNAs that were found to be abundant in the sequencing data; six tsRNAs ( derived from tRNAAla , tRNATyr , tRNATrp , tRNAGlu , tRNAAsp and tRNAThr ) , four rsRNAs and a repeat-derived small RNA ( Figure S3 ) . Validation was performed by Stem-loop Real Time PCR [32] , which has previously been used to detect microRNAs and is more sensitive than Northern hybridization [63] . Of the 18 selected small RNAs , 12 could be amplified ( Figure S3AB ) . A tsRNA derived from tRNAHis was also detected among our samples ( data not shown ) , however , due to primer dimer formation it could not be properly quantified by real time PCR analysis . To obtain a measure of abundance , the signal intensity from the real time PCR was normalized using full length rRNA and C/D snoRNA ( Materials and Methods ) . A tsRNA derived from tRNAAsp gave the strongest signal of the tested tsRNAs ( Figure S3AC ) . The other five tsRNAs displayed a similar level of expression as the snoRNA-control used in these experiments . Three tsRNAs ( derived from tRNAAla , tRNAGlu , tRNAAsp ) have been detected previously in the T cruzi clone Dm28c by Northern hybridization [19] , thus indicating their presence in independent strains . Of the four tested rsRNAs , rsRNA-2 gave the strongest signal ( Figure S3BC ) . Interestingly , several small RNAs derived from non-coding RNAs can be aligned with protein coding genes in the anti-sense direction . For example , rsRNA-3 and rsRNA-4 can be aligned with two distinct protein-coding genes , along with several other putative small RNAs . A similar situation occurs with MASP genes , where small RNAs derived from the repeat element TREZO [64] can be aligned close to the MASP 3′ UTR , which is the most conserved region among these genes [65] . We validated a small RNA derived from the TREZO element , showing that the abundance is similar to that of the snoRNA control ( Figure S3 ) . TREZO elements cover ∼1–2% of the genome , exhibit site-specificity for insertions and are transcribed [64] , although this is the first report to show they generate such small RNAs . Their putative influence on the MASP family expression needs to be further investigated . In this study , we analyzed the short transcriptome of Trypanosoma cruzi using unbiased deep sequencing and provided a glimpse into the diversity and abundance of small RNAs in this species . Despite the fact that T . cruzi lacks RNA interference , our deep sequencing led to the identification of several new types of small RNAs which have not previously been reported in this important organism . The most common RNA species were small RNAs derived from transfer RNAs , followed by small RNAs derived from ribosomal RNAs . Only 1% of the small RNAs in the library were derived from small nuclear RNAs and small nucleolar RNAs . Our deep sequencing effort confirms that , similarly to other protozoan species and mammalian cell lines , T . cruzi accumulates RNA species from tRNA , rRNA as well as snRNA and snoRNA . A selected set of small RNAs was validated using real time PCR and found to be consistently present in different biological samples , although further experimental work will be needed to provide functional insights into the putative roles of some of these small RNAs . Our sequencing data provide a substantial number of follow-up candidates which might be suitable for detailed experiments . We found no evidence of canonical small non-coding RNAs ( i . e . microRNA and siRNA ) as often found in metazoans; an expected finding , consistent with the absence of the RNA interference machinery and confirms the results from previous studies showing that canonical microRNAs do not exist in Trypanosoma cruzi . About 1 . 69% of the small RNAs in the library were unknown , and we identified 92 novel expressed loci , of which 79 lacked conserved sequence or structural motifs . However , it should be noted that the small RNAs reported in this study may not reflect the complete repertoire , as certain small RNAs may have a life stage specific expression or otherwise only be expressed under a certain physiological condition . Further sequencing efforts will be needed to elucidate the complete set of small RNAs and to completely distinguish biologically non-stable intermediates from stable RNAs . Furthermore , it remains to be elucidated whether small RNAs are generated by a distinct mechanism or produced by RNA decay , although the latter does not exclude the possibility that small RNAs have a functional role . Currently we are undertaking deep sequencing of a smaller size fraction to further understand the composition and complexity of the short transcriptome in this peculiar organism .	Chagas' disease is a major health problem in Latin America and is caused by the protozoan parasite Trypanosoma cruzi . T . cruzi lacks the pathway for RNA interference , which is widespread among eukaryotes , and is therefore unable to induce RNAi-related processes . In many organisms , small RNAs play an important role in regulating gene expression and other cellular processes . In order to understand if other small RNA pathways are operating in this organism , we performed high throughput sequencing and genome-wide analyses of the short transcriptome . We identified an abundance of small RNAs derived from non-coding RNA genes , including transfer RNAs , ribosomal RNAs as well as small nucleolar RNAs and small nuclear RNAs . Certain tRNA types were overrepresented as precursors for small RNAs . Further , we identified 79 novel small non-coding RNAs , not previously reported . We did not identify canonical small RNAs , like microRNAs and small interfering RNAs , and concluded that these do not exist in T . cruzi . This study has provided insights into the short transcriptome of a major human pathogen and provided starting points for further functional investigation of small RNAs and their biological roles .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "and", "Discussion" ]	[ "genetics", "biology", "computational", "biology", "microbiology", "genetics", "and", "genomics" ]	2011	The Short Non-Coding Transcriptome of the Protozoan Parasite Trypanosoma cruzi
The BARD1 protein , which heterodimerizes with BRCA1 , is encoded by a known breast cancer susceptibility gene . While several BARD1 variants have been identified as pathogenic , many more missense variants exist that do not occur frequently enough to assign a clinical risk . In this paper , whole exome sequencing of over 10 , 000 cancer samples from 33 cancer types identified from somatic mutations and loss of heterozygosity in tumors 76 potentially cancer-associated BARD1 missense and truncation variants . These variants were tested in a functional assay for homology-directed repair ( HDR ) , as HDR deficiencies have been shown to correlate with clinical pathogenicity for BRCA1 variants . From these 76 variants , 4 in the ankyrin repeat domain and 5 in the BRCT domain were found to be non-functional in HDR . Two known benign variants were found to be functional in HDR , and three known pathogenic variants were non-functional , supporting the notion that the HDR assay can be used to predict the clinical risk of BARD1 variants . The identification of HDR-deficient variants in the ankyrin repeat domain indicates there are DNA repair functions associated with this domain that have not been closely examined . In order to examine whether BARD1-associated loss of HDR function results in DNA damage sensitivity , cells expressing non-functional BARD1 variants were treated with ionizing radiation or cisplatin . These cells were found to be more sensitive to DNA damage , and variations in the residual HDR function of non-functional variants did not correlate with variations in sensitivity . These findings improve the understanding of BARD1 functional domains in DNA repair and support that this functional assay is useful for predicting the cancer association of BARD1 variants . Variants in BRCA1 and BRCA2 account for a plurality of hereditary breast and ovarian cancer ( HBOC ) cases , and are associated with risks of 50–85% for breast cancer and 15–40% for ovarian cancer [1–4] . BARD1 forms an obligate heterodimer with BRCA1 , which functions as both an E3 ubiquitin ligase [5 , 6] and as a direct mediator of homologous recombination for the recruitment of RAD51 to the sites of DNA double-strand breaks [7–9] . Truncated BARD1 variants have been identified in breast and ovarian cancers [10–12] and germline variants in the BARD1 gene are associated with increased cancer risk [13] . Still , for both BRCA1 and BARD1 , the functional and clinical consequences are often unknown for sequence changes that replace the encoded amino acid residue . Both BRCA1 and BARD1 are tested on clinical gene panels for breast and ovarian cancer susceptibility . Many BRCA1 variants , as well as a few BARD1 variants , have been determined to be clinically pathogenic . However , many more variants , which are generally missense substitutions , do not occur frequently enough in the population to assign a cancer risk and are classified as variants of uncertain significance ( VUS ) . The ClinVar database [14] gathers information on pathogenic and benign variants , but most variants in its database are VUS . A gene panel testing 25 breast cancer-associated genes found 42% of all tests have findings of a VUS in one or more genes , indicating many people have such variants and there is a growing need for their classification [15] . Datasets such as the Cancer Genome Atlas ( TCGA ) gather information on missense variants , but are unable to be used for the accurate prediction of the cancer predisposition of a specific VUS . Assays examining homology-directed repair ( HDR ) function have demonstrated that known pathogenic BRCA1 variants are non-functional in HDR , while benign variants are functional [16–19] . BARD1 consists of an amino-terminal RING domain , three ankyrin repeat domains , and two carboxy-terminal BRCT domains [5 , 20] . Previous work in our lab has examined the HDR function of 29 BARD1 variants , focusing on the RING and BRCT domains [21] . In this study , we identified 76 BARD1 missense and truncation variants that were potentially cancer-associated from a large dataset containing exome-sequencing data on matched germline and tumor samples [19 , 22] , and tested them for HDR function . Several HDR-deficient variants were identified in both the ankyrin repeat and BRCT domains . To examine the effects caused by loss of HDR function , cells expressing HDR-deficient BARD1 variants were treated with DNA damaging cisplatin or ionizing radiation . Cells expressing HDR-deficient variants were more sensitive to DNA damage and formed significantly fewer colonies than cells expressing wild-type BARD1 . Although cells expressing HDR-deficient variants were more sensitive to damage than wild-type cells , quantitative variations in HDR deficiency did not correlate with differences in sensitivity to DNA damage agents . The results of this study reveal functional domains of BARD1 and suggest that the functional analysis of BARD1 HDR activity is predictive of breast and ovarian cancer risk . BARD1 missense variants with potential cancer predisposition were identified in a set of 10 , 389 TCGA cancer samples from 33 cancer types using whole exome sequencing ( Fig 1A ) [19 , 22] . 62 rare germline variants and 14 somatic variants were found with variant calling . The variant allele frequency ( VAF ) and loss of heterozygosity ( LOH ) of germline variants were also examined to identify variants that could be functionally important . Six variants—S339T , T343I , V523A , N450H , G451fs and L239Q—were identified as having significantly higher LOH , indicating they had an increased likelihood of being pathogenic ( Fig 1B ) . At the beginning of this study , variants were selected from a cohort of 4 , 034 samples [19 , 23] that later became part of a larger set of 10 , 389 samples [22 , 24] . Because of changes to selection criteria and data analysis , most , but not all , of the variants analyzed in this study were present in the larger data set , which was used to update the variant calling ( Fig 1C ) . While several variants were not present in our newer data set , they are still likely present in the exome sequencing data . Analyzed variants were present in 24 of the 33 cancer types examined , not just breast or ovarian cancer , as might be predicted for a BRCA1 binding partner ( Fig 1D ) . 76 BARD1 missense variants , a majority of which were located in the ankyrin repeat and BRCT domains or between these domains , were tested for function in the homology-directed repair ( HDR ) assay ( Fig 2A ) . For the HDR assay , a cell line that has two non-functional GFP coding sequences integrated into its DNA is used to examine DNA repair function . One of these GFP-encoding genes contains a recognition site for the rare-cutting restriction endonuclease I-SceI . When the I-SceI expression plasmid is transiently transfected into these cells , a double-strand break is made in one of the GFP sequences . If homology-directed repair uses the second GFP coding sequence as a template to repair across the double-strand break , then the encoded GFP is rendered functional [16 , 25] . We used a HeLa-derived cell clone called HeLa-DR , which has the GFP-encoding recombination substrate integrated at a single site . After transfection of the I-SceI expression plasmid , 10–20% of the cells were converted to GFP-positive [16] . Endogenous BARD1 expression was depleted in HeLa-DR cells by two rounds of transfection of a siRNA that targets the 3’-UTR of the BARD1 mRNA . Simultaneously with the silencing of endogenous BARD1 , BARD1 variants were expressed from transiently transfected plasmids that were resistant to the siRNA . Two days following the first transfection , the siRNA and plasmid were transfected again into the cells along with the plasmid that expresses the I-SceI endonuclease . Three days after the second transfection , the number of GFP-positive cells was determined using flow cytometry ( S1 Table ) . Full HDR activity was observed under conditions of mock depletion of BARD1 by transfection with a control siRNA ( Fig 2A , bar 1 ) and by depletion of BARD1 using the 3’-UTR targeted siRNA with rescue by transfection of a plasmid that expressed wild-type BARD1 ( Fig 2A , bar 3 ) . Cells depleted of BARD1 and transfected with an empty vector had a 25-fold decrease in HDR activity measured as the percentage of GFP-positive cells ( Fig 2A , bar 2 ) . We set the level of GFP expression following a double-strand break to a value of 1 relative to wild-type rescue ( Fig 2A , bar 3 ) to facilitate comparison between experiments . The 76 variants tested were from across the full coding sequence of BARD1 . Variants whose HDR activity was lower than 0 . 6 and whose expression was greater than or equal to endogenous BARD1 were considered to be repair-deficient ( Fig 2A and 2B ) . The eight variants located in the RING domain , as well the 22 in the region between the RING and ankyrin repeat domains , all had HDR activity similar to wild-type . Previous work in our lab [21] examined the HDR activity of 29 BARD1 missense variants , including additional variants in the RING domain . We combined the current HDR results with the previously published results into a single table containing 105 variants ( S1 Fig ) . In this previous work , the variants L44R , C53W , and C71Y in the RING domain were found to be defective in HDR due to defective binding to BRCA1 . Surprisingly , in the current study , four of the 17 variants in the ankyrin domain , which has no known DNA repair function , were found to express full-length BARD1 and be defective in HDR . Variants A460T , L465F , L480S , and P530L had HDR activity lower than 0 . 6 , which was significantly lower than cells expressing endogenous BARD1 . The five variants located between the ankyrin repeat and BRCT domains were proficient in DNA repair with the exception of R565C , whose HDR activity was just below the cutoff of 0 . 6 . Of the 19 missense variants tested in the BRCT domain , which is known to be involved in recruiting and retaining the BRCA1-BARD1 heterodimer to areas of DNA damage , five were found to be defective in HDR [26 , 27] . The variants S660R and G698D had HDR function comparable to cells transfected with empty vector . The variants T598I , P707S , and G753D had activity higher than empty vector but still significantly lower than endogenous BARD1 . A larger fraction of residues conserved across several mammalian species were mutated in repair-deficient variants ( 9/10 ) than in functional ones ( 38/55 ) ( S2 Fig ) . Five truncation variants were also tested , and all were about as equally defective as the empty vector in the HDR assay . Previous work has suggested that filtering using high LOH could be used to identify BRCA1 variants defective in HDR [19] . However , BARD1 variants that were found to have high LOH ( Fig 1B ) were all functional , with the exception of truncation variant G451fs . Testing the HDR function of BRCA1 variants has shown that , with the exception of variants that impact mRNA splicing , known pathogenic variants of BRCA1 are HDR-defective , while known benign variants are not [16 , 17 , 19] . Similarly , the BARD1 variants S241C and E361D , which have been found in patients with breast cancer and are benign according to ClinVar , are functional in HDR ( Fig 2A , blue dots ) . Truncation variants V154fs , S551* , and Q564* , where the asterisk indicates a stop codon , are listed as pathogenic according to ClinVar and were non-functional in the HDR assay ( Fig 2A , red dots ) . Several other variants that were tested have been identified in breast cancer patients and have conflicting interpretations of pathogenicity ( Fig 2A , gray dots ) . We tested whether the level of expression of any BARD1 variants could have affected their HDR activity . Expression of BARD1 variants was examined via immunoblot ( Fig 2B ) . The relative expression of the endogenous BARD1 and siRNA-depleted BARD1 were shown ( Fig 2B lanes 1 , 2 , 14 , 15 , 23 , 24 , 29 , 30 , 33 , 34 , 41 , 42 , 53 , 54 , 65 , 66 , 77 , 78 , 90 , 91 , 100 , 101 ) . Though the expression levels of missense variants differed , they all expressed at higher levels than the endogenous BARD1 . As an example , BARD1 L480S ( lane 63 ) had lower expression than the plasmid encoded wild-type ( lane 55 ) , but both had more intensely labeled bands than the endogenously expressed BARD1 ( lane 53 ) . The variant BARD1 H606D is present in the immunoblots in Fig 2B ( lane 103 ) , but is not listed in Fig 2A because full length protein was not detected , and it was found to contain a nonsense mutation at codon 125 . Similarly , BARD1 L359fs is present in Fig 2B ( lane 50 ) and S3 Fig ( lanes 5 , 14 ) but is not listed in Fig 2A because it was a miss-call during variant selection . Frameshift and nonsense codon variants ( Fig 2B lanes 10 , 60 , 81 , 84 , 112 ) lacked full length BARD1 . Truncation variants G451fs , S551* , Q564* and V767fs expressed truncated BARD1 , while variant protein V154fs was not detected ( S3 Fig ) . We infer that repair defects observed in truncation variants with poor protein expression were due to the absence of protein instead of expressed , non-functional variant protein . We conclude that for the missense variants , a low level of HDR activity was not due to low expression of the BARD1 protein . BARD1 variants A460T , P707S , G753D , and V767fs were selected for further analysis , as they covered a range of HDR activities below 0 . 6 when transiently expressed . The selected variants and wild-type BARD1 were tagged with the His-Biotin-Tobacco Etch Virus ( HBT ) tag [28] and integrated into the FRT site of a HeLa-DR derivative cell line called HeLa-DR-FRT/TR [29] . The advantage of these FRT site-containing cells was that the BARD1 gene was stably expressed from a single site and should have consistent levels of expression . We tested the stably expressed variants in the HDR assay to confirm repair proficiency was the same as the transiently expressed variants . The HDR activities of these variants from Fig 2 are shown in isolation ( Fig 3A ) . The same repair trends were observed in both transiently expressed and stably expressed BARD1 variants ( Fig 3A and 3B ) . BARD1 A460T-integrated cells had the most residual repair activity among these variants , and cells integrated with BARD1 P707S , G753D and V767fs had decreasing levels of repair proficiency respectively . Expression of the integrated BARD1 variants was also greater than or equal to that of endogenous BARD1 ( Fig 3C ) . BRCA1 expression in variant-integrated cell lines was similar in cells expressing endogenous BARD1 and the various defective BARD1 variants ( Fig 3C ) . Thus , none of the changes in HDR observed with these BARD1 variants were attributable to changes in BRCA1 expression . We examined whether the quantitative loss of HDR proficiency correlated with the sensitivity of cells to extrinsic DNA damage . Clonogenic cell sensitivity assays were performed on HeLa-DR-FRT/TR cells expressing integrated BARD1 wild-type and variants , as well as endogenous-only unintegrated cells . Cells were depleted of endogenous BARD1 or BRCA1 and subjected to ionizing radiation ( IR ) ( Fig 4 ) or cisplatin ( Fig 5 ) . Depletion of BARD1 or BRCA1 from the endogenous-only cells ( E/siBARD and E/siBRCA ) examined the effect of a non-rescued HDR defect on sensitivity to IR and provided a baseline for DNA damage sensitivity ( Fig 4A , bottom ) . E/siBARD and E/siBRCA cells formed significantly fewer colonies after IR than control cells ( E/siCON ) . BARD1 variant-integrated cells depleted of endogenous BARD1 ( Variant/siBARD ) all formed significantly fewer colonies following IR than the same cells treated with control siRNA ( Variant/siCON ) ( Fig 4A , top ) . For ease of comparison , we included results from Variant/siBARD , E/siBARD , E/siBRCA and WT/siBARD cells on one graph ( Fig 4B ) . Immunoblots were done to confirm knockdown of endogenous BARD1 ( Fig 4C ) . While expression of each BARD1 variant differed , it was still greater than or equal to that of endogenous BARD1 . E/siBARD , E/siBRCA and Variant/siBARD cells formed significantly fewer colonies than WT/siBARD cells at most irradiation concentrations ( Fig 4B ) . It was expected that increased HDR deficiency would result in increased sensitivity to DNA damage agents . For example , we expected that cells expressing BARD1 V767fs , the most HDR-deficient variant , would form the least number of colonies . Cells expressing BARD1 A460T , the variant with the most residual HDR activity , were expected to have the largest number of colonies compared with the other variants . Interestingly , this was not the trend observed in the results . Instead , HDR-defective BARD1 variants were all equally sensitive to IR , and as sensitive as non-rescued cells depleted of BARD1 or BRCA1 . Similarly , all four BARD1 variants were more sensitive than wild-type to treatment with cisplatin ( Fig 5A ) . Variant/siBARD , E/siBARD and E/siBRCA cells formed significantly fewer colonies than Variant/siCON and E/siCON cells . Variant/siBARD , E/siBARD and E/siBRCA cells also formed significantly fewer colonies than WT/siBARD cells ( Fig 5B ) . As seen with IR , all BARD1 variants and non-rescued cells depleted of BARD1 or BRCA1 were equally sensitive to cisplatin . In addition , BARD1 variant expression remained consistently equal to or greater than that of endogenous BARD1 in cells , indicating that decreased colony formation was not associated with decreased variant expression ( Fig 5C ) . While decreased HDR function resulted in decreased colony formation , quantitative differences in the HDR activity did not correlate to quantifiable changes in sensitivity to cisplatin or IR . In this study , we found: 1 ) from 10 , 389 cancer samples across 33 cancer types , 76 BARD1 missense variants were identified as potentially pathogenic and were selected for functional analysis . 2 ) 16 of the 76 tested variants were defective for HDR , suggesting that these were potentially pathogenic variants . 3 ) Four of the 17 variants tested in the ankyrin repeat domain , for which there was no previously known DNA repair function , were deficient in homologous recombination . 4 ) Five of the 19 variants tested in the BRCT domain , which does have known DNA repair functions , were deficient in homologous recombination . 5 ) Variants that were deficient in HDR rendered the cells sensitive to treatment with DNA-damaging cisplatin or IR . 6 ) Quantitative differences in HDR deficiency among defective variants did not translate to quantitatively different sensitivity to DNA damage . The BRCA1-BARD1 heterodimer is necessary for tumor suppressor function [5 , 30] . Variants that affect binding between BRCA1 and BARD1 have been linked to familial breast cancer or are non-functional in the HDR assay [21 , 31 , 32] . Loss of BARD1 has been linked to increased susceptibility to hereditary breast and ovarian cancer ( HBOC ) and is associated with loss of tumor suppressor activity [4 , 13 , 33–36] . The importance of BARD1 in cancer development indicates how significant it is to determine whether BARD1 VUS are benign or pathogenic . A rise in the quantity of genomic data has led to an increasing number of VUS , uncertain due to their low frequency and conflicting reports of pathogenicity . In this paper , potentially pathogenic BARD1 variants were identified in a dataset of 10 , 389 cancer samples from 33 different cancers [22] . From germline and somatic samples , 76 variants from across the entire BARD1 gene were identified as suggestive for being pathogenic . Analysis from the tumor sequencing data indicated BARD1 S339T , T343I , V523A , N450H , G451fs , and L239Q had significantly increased LOH , suggesting that these variants were more likely to be pathogenic . However , with the exception of the G451fs truncation variant , the other five variants were found to be functional in HDR . Previous work [19] has proposed that HDR-deficient BRCA1 variants could be identified using filtering based on increased LOH . In contrast , we have found that increased LOH does not correlate with HDR deficiency for BARD1 variants . As we are unaware of other studies regarding the relationship between LOH and HDR-deficient BRCA1 and BARD1 variants , these contrary results suggest that increased LOH is not a reliable indicator of non-functional variants . Data suggest that the HDR assay is a more effective method for identifying deficient variants . As the variant expression plasmids used only contain the mRNA coding sequence and we do not have access to patient samples , we cannot accurately examine the mRNA expression levels . However , if the mRNA expression levels of LOH mutants are lower than wild-type BARD1 , this could indicate an HDR deficiency that we do not observe in our protein expression experiments . In addition , the variant A460T , which was non-functional in HDR , was identified as potentially pathogenic during the original analysis of exome sequencing data [19] but was not considered pathogenic after updated analysis . These differences demonstrate the importance of empirical results , such as the HDR assay , for evaluating whether a given variant is predictive of cancer . Conversely , the results from functional assays may provide feedback for the improvement of the bioinformatic interpretation of genomic data . While large scale genomic analyses should be used to identify functionally significant variants , functional assays must still be utilized for more comprehensive characterization of these variants . The 76 BARD1 missense variants were tested in the HDR assay to examine DNA repair function . Variants were considered non-functional if their HDR activity was below 0 . 6 and significantly different from endogenous BARD1 . Previous work in our lab [21] characterized fully non-functional variants as those with HDR activity significantly different from endogenous BARD1 but not empty vector , and intermediate variants as those significantly different from both endogenous BARD1 and empty vector . We have changed our classification standards based on the strength of the reproducibility in this study and for ease of analysis . We now interpret variants as functional or nonfunctional and do not include the intermediate phenotype . This interpretation is supported by the new data that the BARD1 A460T variant would have been ranked as intermediate using the previous interpretation , but assays measuring DNA damage sensitivity using IR and cisplatin indicated that BARD1 A460T was just as sensitive as the V767fs variant , which scored similar to the empty vector in the HDR assay . Based on BRCA1 variant function in HDR , it was anticipated that BARD1 variants with impact on HDR function would map to the RING and BRCT domains . In this study , we tested eight variants in the BARD1 RING domain and none of them were defective for HDR . In a prior study [21] , we had analyzed another nine variants in the BARD1 RING domain; three of these were defective for DNA repair activity . In the current study , we tested 19 variants in the BARD1 BRCT domain , and five of these were defective . BARD1 T598I , S660R , G698D , P707S , and G753D were found to be non-functional in the BRCT domain . The BRCT domain has been shown to interact with the HP1 protein in order to retain both the BRCA1-BARD1 complex and CtIP , which is involved in DNA end resection , at the damage site [26] . The BARD1-HP1 interaction is also necessary for the accumulation of DNA helicase FANCJ at sites of DNA damage [37] . BARD1 L570E/V571E and L570A/V571A variants have been shown to inhibit the interaction between BARD1 and HP1 [26] . The BARD1 BRCT domain is also necessary for binding poly ( ADP-ribose ) ( PAR ) , allowing for rapid recruitment of the BRCA1-BARD1 complex to areas of DNA damage [27] . The variants K619A , C645R and V695L have been shown to disrupt BARD1-PAR interaction [27] . The T598 residue tested in this study ( T598I ) is located on the surface of the protein next to K619 , and could possibly affect PAR binding . However , the relationship between BARD1-PAR binding and HDR is unclear since the BARD1 K619A and V695L variants , as well as several others that disrupt PAR binding , have been shown to be functional in HDR [38 , 39] . BRCA1 also binds to the BARD1 BRCT domain [40] , and previous work in our lab has identified that this binding is affected by the BARD1 G623E variant [21] . Most of the non-functional variants we identified in the BRCT domain , with the exception of T598I , are not located near known binding sites for proteins associated with DNA repair . To our surprise , we found four variants , BARD1 A460T , L465F , L480S , and P530L , were identified as non-functional in the ankyrin repeat domain . Prior to this study , the ankyrin repeat domain had no known reported function in DNA repair . Previous work has shown that a large deletion of the ankyrin repeat domain results in chromosome instability and loss of HDR function [38] . In addition , the BARD1 ankyrin domain has been shown to interact with p53 to mediate apoptosis [41] . The oncoprotein Bcl-3 , which interacts with BARD1 , is also involved in the regulation of NF-κB transcription via ankyrin repeat domain-associated protein interactions [42] . Our results indicate that the ankyrin repeat domain may have functions that are necessary for DNA repair . For both the ankyrin and BRCT domains , the non-functional variants identified may affect BARD1 folding and structure , which could also affect binding to proteins such as BRCA1 or HP1 . For example , the BARD1 P707 and G753 residues are located near one another on the surface of the protein . As coding substitutions at these amino acids result in loss of HDR they may be part of a binding pocket . The identified variants may also indicate binding sites for proteins whose interaction with BARD1 has not yet been discovered . The characterization of non-functional BARD1 variants in areas that are not well-studied helps to further understand the roles of the ankyrin repeat and BRCT domains in homology-directed repair . Many of the BARD1 variants tested have been recorded on ClinVar as having been isolated from patients with breast cancer or hereditary cancer-predisposing syndromes . Variants S241C and E361D , which were functional in HDR , have been identified has likely benign , and truncation variants V154fs , S551* , and Q564* are likely pathogenic . The variants V85L , R194K , I258T , N326S , R565H , and R641Q , which were functional in HDR , have conflicting reports of pathogenicity , with reports indicating they were VUS or likely benign . Since these variants were functional in the HDR assay , we would interpret such variants as likely benign . The non-functional truncating variant V767fs also had conflicting reports of pathogenicity , with reports indicating that it was a VUS or likely pathogenic . The trend observed is this paper is supplemented by the 29 variants that were previously studied [21]—variants V507M and R658C were functional in the HDR assay and are considered benign in ClinVar , and several other functional variants are listed as having conflicting reports of pathogenicity because reports indicate they are VUS or likely benign . Previous work has shown that pathogenic BRCA1 variants are non-functional in the HDR assay , and benign variants are functional [16 , 17 , 19] . Based on the data from ClinVar , this trend appears to be true for BARD1 variants as well , suggesting that the non-functional variants identified in this paper would be likely pathogenic . We also asked whether BARD1 HDR function affected cell sensitivity to DNA damage agents . Cells expressing HDR-deficient variants A460T , P707S , G753D and V767fs , as well as endogenous-only , non-rescued cells depleted of BRCA1 or BARD1 , were more sensitive to treatment with IR or cisplatin than cells expressing wild-type BARD1 . Testing non-rescued cells allowed us to set a standard for the effect of non-functional DNA repair on damaged cells . We had hypothesized that the more HDR-deficient a variant was , the fewer colonies cells expressing that variant would form after damage , indicating a quantitative sensitivity to the DNA damage . The results did not support that expectation . Following treatment with IR or cisplatin , HDR-defective variants were in fact more sensitive to DNA damage , but they were as sensitive to DNA damage as cells depleted of BRCA1 or BARD1 , suggesting that residual repair did not affect sensitivity . The HDR assay indicates which variants are functional and non-functional , but it does not provide information on how this affects cell growth . This paper has shown that loss of homologous recombination results in increased sensitivity to DNA damage , as indicated by decreased colony formation . The HDR results also reveal a correlation between BARD1 variants that cause a loss of DNA repair function with those that are likely cancer predisposing . While we examined in this study a large number of BARD1 variants across the length of the protein , including all three functional domains , many more BARD1 VUS exist . In future work , we hope to mutagenize the BARD1 functional domains on a larger scale , as we have previously done with the BRCA1 N-terminus [29] . Creating a library of all potential BARD1 variants in these functional domains and testing the HDR function of these variants would allow us to identify additional pathogenic variants and regions of interest . The work done in this study helps better understand the role of BARD1 in DNA repair , and how loss of homology-directed repair affects cell growth and sensitivity . Sequencing results from de-identified tumors have already been published , and no additional ethical approval was required . BARD1 missense variants conferring potential cancer predisposition were identified in a cohort of 4 , 034 samples from 12 cancer types [19] that was part of larger set of 10 , 389 TCGA samples from 33 cancer types [22] . Germline single nucleotide variants ( SNVs ) were identified with variant calling on whole exome sequencing data using GATK [43] ( version 3 . 5 , using its haplotype caller in single-sample mode with duplicate and unmapped reads removed and retaining calls with a minimum quality threshold of 10 ) and VarScan [44] ( version 2 . 3 . 8 with default parameters , except where–min-var-freq 0 . 10 , –p value 0 . 10 , –min-coverage 3 , –strand-filter 1 ) operating on a mpileup stream produced by SAMtools ( version 1 . 2 with default parameters , except where -q 1 -Q 13 ) . Germline indels were identified using VarScan and GATK ( same parameters and version as above ) in single-sample mode . Pindel [45] ( version 0 . 2 . 5b8 with default parameters , except where -x 4 , -I , -B 0 , and -M 3 and excluded centromere regions ( genome . ucsc . edu ) ) was also applied for indel prediction . For all analyses , the GRCh37-lite reference was used and an insertion size of 500 was specified whenever this information was not provided in the BAM header . All variants were limited to limited to coding regions of full-length transcripts obtained from Ensembl release 70 plus the additional two base pairs flanking each exon that cover splice donor/acceptor sites . SNVs were based on the union of raw GATK and VarScan calls , while indels were required to be called by at least two out of the three variant callers ( GATK , VarScan , Pindel ) . High-confidence , Pindel-unique calls ( at least 30x coverage and 20% VAF ) were also included . Further , variants were required to have an Allelic Depth ( AD ) ≥ 5 for the alternative allele . Readcount analyses for variants passing these filters were performed in both normal and tumor samples using bam-readcount ( version 0 . 8 . 0 commit 1b9c52c , with parameters -q 10 , -b 15 ) in order to quantify the number of both reference and alternative alleles . Variants were required to have at least 5 counts of the alternative allele and an alternative allele frequency of at least 20% . Of these , rare variants were filtered for , with ≤ 0 . 05% minor allele frequency in 1000 Genomes and ExAC ( release r0 . 3 . 1 ) . Variants passing manual review , with low allele frequencies ( MAF < 0 . 05% ) , and significant LOH were prioritized for characterization . These variants , their cancer type distributions and frequencies are shown in the latest data release of the 10 , 389 samples [22] to the research community on NCI Genome Data Commons ( https://gdc . cancer . gov/about-data/publications/PanCanAtlas-Germline-AWG ) . For transient expression , BARD1 ( NCBI Reference Sequence: NM 000465 . 3 ) wild-type and missense variants were cloned into a pcDNA3 vector backbone containing a rabbit β-globin intron upstream of the BARD1 translation initiation site to drive expression of the 777 amino acid human BARD1 transgene . Variants were cloned using the New England BioLabs Q5 Site-Directed Mutagenesis kit . For stable integrations , BARD1 wild-type and missense variants were cloned into a pcDNA5/FRT/TO vector backbone containing the His-Biotin-Tobacco Etch Virus ( TEV ) ( HBT ) tag [28] . PCR reactions were done using PfuUltra II Fusion HS DNA Polymerase . Vectors and inserts were ligated together using Gibson assembly [46] . Colonies with successful ligations were fully sequenced to confirm expected variants . All variants were verified using Sanger sequencing services provided OSU Comprehensive Cancer Center ( OSUCCC ) Genomics Shared Resource . For examining the HDR function of transiently expressed of BARD1 variants , HeLa-DR-13-9 ( HeLa-DR ) cells were utilized . HeLa-DR cells contain two non-functional GFP coding sequences , one of which is interrupted by an I-SceI restriction endonuclease site . Cells were cultured in DMEM media containing 1% penicillin/streptomycin , 1% GlutaMAX , 10% bovine serum , and 1 . 5 μg/ml puromycin . Cells were seeded in a 24-well plate and transfected with siRNA to the BARD1 3’-UTR ( 5’-AGCUGAAUAUUAUACCAGAdTdT-3’ ) or control siRNA ( 5 pmol ) , and BARD1 wild-type , variant , or pcDNA3 empty vector ( 300 ng ) . All transfections were carried out using Lipofectamine 2000 per the manufacturer’s recommendations . Cells were moved to 6-well plates 24 hours later . 48 hours after the first transfection , cells were transfected with 25 pmol siRNA , 750 ng DNA , and 750 ng of expression plasmid containing the restriction endonuclease I-SceI to induce a double-strand break . If HDR is functional , the break is repaired by gene conversion with the second GFP allele , and cells become GFP-positive [16 , 25] . 72 hours after the second transfection , cells were collected and GFP-positive cells were counted using the FACSCalibur in the OSUCCC Analytical Cytometry Shared Resource . 10 , 000 cells were counted , and the remaining cells were used for immunoblotting . Cells transfected with BARD1 siRNA and BARD1 wild-type plasmid ( wild-type rescue ) , and cells treated with control siRNA and empty vector , served as positive controls . Cells treated with BARD1 siRNA and empty vector were used as a negative control . HDR activity , as defined by the percentage of GFP-positive cells , was normalized to wild-type rescue control and set to 1 . For examining the HDR function of stably integrated BARD1 variants , HeLaDR-FRT/TR cells [29] were used . Cells integrated with pcDNA5-FRT/TO-HBT-tagged BARD1 wild-type or variants ( A460T , P707S , G753D and V767fs ) were seeded in 24-well plates . Cells not integrated with BARD1 variants were used as a negative control . Cells were transfected with siRNA to the BARD1 3’-UTR or control siRNA . All transfections were carried out using Oligofectamine according to the manufacturer’s recommendations . Transfections were carried out on the same time pattern as detailed for transiently expressed BARD1 . For 24-well transfections , 30 pmol siRNA was used , and for 6-well transfections , 50 pmol siRNA and 3 μg of I-SceI expression plasmid were used . Cells were collected and GFP-positive cells counted as detailed for transiently expressed BARD1 . HDR activity , as defined by the percentage of GFP-positive cells , was normalized to cells treated with control siRNA for each individual cell line and set at 1 . Homo sapiens , Felis catus , Canis lupis familiaris , Mus musculus , Monodelphis domestica , and Ovis aries BARD1 protein sequences were aligned using Clustal Omega [47] to examine conserved residues . pcDNA5-FRT/TO-HBT-tagged BARD1 wild-type and variants were co-transfected with plasmid expressing the flippase recombinase in a 1:2 ratio into HeLa-DR-FRT/TR cells to induce integration at the flippase recognition target site [48 , 49] . Transfections were done according to Lipofectamine 2000 manufacturer’s recommendations . 24 hours after transfection , cells were incubated at 30°C for 24 hours and then moved back to 37°C . Integrated cells were selected for with 550 μg/ml Hygromycin-B . For BARD1 variants , replicates were combined , spun down at 1200 rpm for 5 minutes and resuspended in 150 μl of 1X LDS-PAGE dye . Samples were sonicated at 45% for 15 seconds three times . Sample was resolved on 6 or 8% SDS-PAGE gels and transferred to PVDF membrane . Samples were probed with BARD1 ( Bethyl , 1:1000 ) and BRCA1 ( 1:500 ) [50] antibodies . Antibodies to RHA1 ( 1:20000 ) [50] and α-tubulin ( Sigma , 1:20000 ) were used as loading controls . Membranes were incubated with fluorescent ( LI-COR , 1:20000 ) or chemiluminescent ( GE , 1:5000 ) rabbit and mouse secondary antibodies . HeLaDR-FRT/TR cells stably expressing pcDNA5-HBT-BARD1 WT , A460T , P707S , G753D and V767fs , as well as control cells expressing only endogenous BARD1 , were seeded in 24-well dishes and transfected with 30 pmol of siRNA to the BARD1 3’-UTR , BRCA1 3’-UTR or control siRNA . Transfections were done using Oligofectamine as per the manufacturer’s protocol . Cells were transferred to 6-well dishes after 24 hours , and were treated with 50 pmol siRNA 48 hours after the first transfection . 48 hours after the second transfection , 1000 cells of each treatment condition were plated in 10-cm dishes . Remaining cells were saved for immunoblotting to confirm knockdown . After 24 hours , cells were treated with either ionizing radiation ( IR ) or cisplatin . Cells that were treated with IR were subjected to 0 , 1 , 2 , 4 , or 6 Gy using the RS 2000 X-Ray Irradiator . For cisplatin treatment , cells were treated with 0 , 1 , 875 , 3 . 75 , 7 . 5 , and 15 μM of cisplatin for 2 hours , after which cells were washed twice with 1X PBS and fresh media was added . Untreated cells were used as a control . After two weeks of growth at 37°C , cells were fixed with cold methanol and stained with crystal violet . Dishes were coded to blind their treatment and cells were counted using OpenCFU [51] . The log value of the count was used for comparison . All BARD1 variants in the HDR and clonogenic assays were tested in triplicate . For HDR assays using transiently expressed BARD1 variants , HDR activity was normalized to wild-type rescue , which was set to 1 . The Student’s t-test was applied to determine whether BARD1 variant HDR activity significantly differed ( p < 0 . 01 ) from endogenous BARD1 . Variants that were significantly different and below the cutoff of 0 . 6 were considered non-functional . For clonogenic assays , the Student’s t-test was carried out to examine whether BARD1 variant-expressing and endogenous-only cells treated with BARD1 or BRCA1 3’-UTR siRNA formed a significantly different number of colonies than variants treated with control siRNA and cells expressing BARD1 wild-type ( p < 0 . 05 ) .	BARD1 is a breast cancer susceptibility gene encoding a protein that primarily interacts with BRCA1 in DNA repair . Although several BARD1 variants are known to be pathogenic , many more variants do not occur frequently enough to assign a clinical risk . In this paper , we identified 76 potentially cancer-associated BARD1 variants from analysis of over 10 , 000 tissue samples from people with cancer . It has previously been shown that if a BRCA1 variant cannot repair damaged DNA , then it is likely to cause cancer . We tested BARD1 variants for DNA repair function and identified several non-functional variants that were localized in parts of the BARD1 protein not previously associated with DNA repair . Known benign BARD1 variants were found to be functional and known pathogenic variants were non-functional , showing that examining DNA repair function predicted variant pathogenicity . Cells expressing repair-defective BARD1 variants were also more sensitive to DNA damaging agents . These findings help us better understand how BARD1 is involved in DNA repair and show that examining the DNA repair function of BARD1 variants is useful for predicting their cancer risk .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "transfection", "medicine", "and", "health", "sciences", "3'", "utr", "gene", "regulation", "cancer", "risk", "factors", "messenger", "rna", "plasmid", "construction", "oncology", "dna", "damage", "ankyrin", "repeat", "domains", "untranslated", "regions", "dna", "co...	2019	Functional analysis of BARD1 missense variants in homology-directed repair and damage sensitivity
Schistosomiasis is a public health problem in Malawi but estimates of its prevalence vary widely . There is need for updated information on the extent of disease burden , communities at risk and factors associated with infection at the district and sub-district level to facilitate effective prioritization and monitoring while ensuring ownership and sustainability of prevention and control programs at the local level . We conducted a cross-sectional study between May and July 2006 among pupils in Blantyre district from a stratified random sample of 23 primary schools . Information on socio-demographic factors , schistosomiasis symptoms and other risk factors was obtained using questionnaires . Urine samples were examined for Schistosoma hematobium ova using filtration method . Bivariate and multiple logistic regressions with robust estimates were used to assess risk factors for S . hematobium . One thousand one hundred and fifty ( 1 , 150 ) pupils were enrolled with a mean age of 10 . 5 years and 51 . 5% of them were boys . One thousand one hundred and thirty-nine ( 1 , 139 ) pupils submitted urine and S . hematobium ova were detected in 10 . 4% ( 95%CI 5 . 43–15 . 41% ) . Male gender ( OR 1 . 81; 95% CI 1 . 06–3 . 07 ) , child's knowledge of an existing open water source ( includes river , dam , springs , lake , etc . ) in the area ( OR 1 . 90; 95% CI 1 . 14–3 . 46 ) , history of urinary schistosomiasis in the past month ( OR 3 . 65; 95% CI 2 . 22–6 . 00 ) , distance of less than 1 km from school to the nearest open water source ( OR 5 . 39; 95% CI 1 . 67–17 . 42 ) and age 8–10 years ( OR 4 . 55; 95% CI 1 . 53–13 . 50 ) compared to those 14 years or older were associated with infection . Using urine microscopy as a gold standard , the sensitivity and specificity of self-reported hematuria was 68 . 3% and 73 . 6% , respectively . However , the positive predictive value was low at 23 . 9% and was associated with age . The study provides an important update on the status of infection in this part of sub-Saharan Africa and exemplifies the success of deliberate national efforts to advance active participation in schistosomiasis prevention and control activities at the sub-national or sub-district levels . In this population , children who attend schools close to open water sources are at an increased risk of infection and self-reported hematuria may still be useful in older children in this region . Schistosomiasis remains an important public health problem globally with an estimated 200 million cases reported each year [1] . However , 85% of the cases reported annually occur in sub-Saharan Africa and over 150 , 000 deaths are attributable to chronic infection with S . haematobium in this region [2] , [3] . The eggs of S . haematobium provoke granulomatous inflammation , ulceration , and pseudo-polyposis of the vesical and ureteral walls . Hematuria is a very common sign of infection but other signs include dysuria , pollakisuria , and proteinuria . Kidney failure deaths due to urinary tract scarring , deformity of ureters and the bladder caused by S . haematobium infection have become less common due to modern drugs [4] , [5] . Subtle and indirect morbidities such as fatigue , physical or cognitive impairment and effects of co-infections with other infectious diseases like HIV , malaria have received more attention recently [4] . New evidence from a recent review of these studies suggests a causative link between schistosome infection , anti-parasite inflammation , and risk for anaemia , growth stunting and under-nutrition , as well as exacerbation of co-infections and impairment of cognitive development and physiological capacities among infected individuals [6] . The causal relationship between anaemia and schistosomiasis exists even after controlling for other co-infections and dietary factors among pregnant women and children [6]–[9] . The underlying mechanisms proposed range from social determinants to complex immune interactions . In Malawi , schistosomiasis is endemic with S . haematobium being highly prevalent in the southern region while S . mansoni predominates on the central plain and the northern regions [10] . The national schistosomiasis control program estimates that between 40% and 50% of the total Malawian population is infected with schistosomiasis ( National Plan of Action for the Control of Schistosomiasis and Soil transmitted Helminthes five-year plan 2004–2008 ) . Other studies have suggested that these national estimates may have been derived from studies conducted years ago that had some selection bias for high risk schools [11] . A national survey conducted in 2002 among primary school pupils found the prevalence of S . hematobium and S . mansoni infection among school children to be 6 . 9% and 0 . 4% using filtration method of urine and Kato-Katz method for stools respectively [11] . These findings were much lower than expected and it was concluded that schistosomiasis is highly localized in Malawi . This implies that local estimates would be more useful than national estimates in guiding the selection of control strategies to be implemented at district or sub-district level which depends on disease prevalence rate in a community [12] . However , most districts in Malawi do not have local estimates to guide planning and implementation of control interventions at that level . The Ministry of Health through the national schistosomiasis control program is currently undertaking a deliberate effort to have schistosomiasis prevention and control efforts integrated within district plans to encourage ownership and improve sustainability by engaging district teams ( National Plan of Action for the Control of Schistosomiasis and Soil transmitted Helminthes five-year plan 2004–2008 ) . Blantyre district health team initiated and conducted this study aimed at determining the prevalence of schistosomiasis as part of this effort . The findings are expected to be used as baseline for future evaluation and monitoring of control activities . A representative cross-sectional study of urinary schistosomiasis infection among school children in Blantyre district was therefore conducted . To establish the need for intervention in a community , we conducted our baseline survey among children sampled from grade three in elementary schools as it has been shown that this population is intensely affected in endemic communities [13] . The prevalence distribution , factors associated with S . hematobium infection and the reliability of self reported hematuria compared to the “gold standard” parasitological examination among school children are described . This cross-sectional study was carried out in Blantyre district located in southern Malawi . The district has a land area of 2 , 012 km2 and the altitude above sea level ranges from 300 m to 1000 m . According to the national statistical office , it has an estimated population of one million of whom 61% reside in the urban or peri-urban areas . From north-west to south-west boundary of the district runs the Shire River which is the main outlet of Lake Malawi ( Figure 1 ) . The average annual temperature is 27°C whilst the average rainfall is 871 mm . The rainy season is from November to April and the dry season from May to October [14] . All standard three pupils who were attending primary schools in Blantyre district during the time of data collection between May and July 2006 were eligible to participate in our study . Our study sample was selected using a stratified 2-stage probability sampling technique . First , we stratified the district was into 3 ecological risk areas for urinary schistosomiasis which was based on altitude above sea level as follows; high ( <500 m ) , moderate ( 500–1000 m ) and low ( >1000 m ) . Some previous studies have shown that altitude above sea level is known to be associated with schistosomiasis [15] , [16] . We used STATA software v10 . 1 ( StataCorp Ltd , Texas , USA ) to estimate sample size for comparison of proportions between strata . The sample size of 1 , 128 was estimated using a prevalence estimate of 40 . 0% with 80 . 0% power to detect a relative difference of 10 . 0% between strata at a 0 . 05 level of significance . Second we obtained from the district education offices the number of schools and pupils' population estimates in each stratum . Based on the pupils' population distribution within the strata and our prior decision to randomly select 50 standard three pupils from each selected school , we calculated that 23 schools would be required . Studies have shown that Lot-Quality Assurance Scheme approaches provide the ability to identify communities with a high prevalence of schistosomiasis with high levels of sensitivity and specificity , even at very small maximum sample sizes [17] . Finer classification of schools according to categories of prevalence are achieved with moderate sample sizes of ≥15 and have been found to be more accurate with extremely low probabilities of making gross classification errors . Thirdly , sampling was conducted in 2 stages . In the first stage , the primary sampling units were schools that were selected with a probability proportional to number of schools in the strata . A list of schools in each stratum was compiled and using computer generated random numbers schools in each stratum were selected as follows; six schools from the high risk stratum , eight from the moderate risk stratum and nine from the low risk stratum . In the second stage , a random sample of 50 grade three pupils in each selected school was obtained . Children were interviewed by trained Health Surveillance Assistants ( HSAs ) and community health nurses using a questionnaire that was adapted from the 2002 national prevalence survey [11] . To reduce bias and improve the performance of the questionnaires , questions about schistosomiasis were disguised among other health related questions . The questionnaire was pre-tested and modifications were made after discussions with HSAs , teachers and district health office staff . Risk factors included main household water source , child's knowledge of nearby open water sources ( open water source defined as any open water body including lakes , springs , rivers , streams , ponds , swamps and dams ) , frequency of contact with open water sources , dysuria or passing blood in urine within the past month , history of S . hematobium infection and S . hematobium treatment . Other factors included urban or rural location , proximity of school to an open water source and household socio-economic status ( SES ) . We used a method developed by the Centre for Social Research in Malawi ( Kadzandira et al . unpublished report ) and similar methods have been used in other studies to estimate SES [18] . The method combined six variables to formulate the complex indicator of SES which includes housing structures , main occupation of heads of households , and possession of selected assets such as radios , telephones , televisions , etc . For each variable , households were assigned a weight ranging from zero to two . For example , households with a grass-thatched roof were given a weight of zero while households with a plastic paper roof were given a 0 . 3 and households with iron-sheet or tiled roof were given a 1 . 0 ( Table S1 ) . The sum of weights from the six variables determined the SES score of each household . Households were then classified as low SES for those with scores less than 4 . 0 , moderate SES for 4 . 0–6 . 0 and high SES for more than 6 . 0 . There were no significant differences when we analyzed our data using the year 2000 World Bank asset scoring system for wealth ( data not shown ) [19] . In addition to the questionnaire interview , pupils were asked to submit urine in supplied 20 ml screw top plastic containers between 0900 hrs and 1400 hrs . The samples were immediately transported to the University of Malawi College of Medicine microbiology laboratory . Our primary outcome was S . hematobium infection as diagnosed by urinary microscopy examination . Samples were tested for micro-hematuria using urine reagent strips ( Uripath , Plasmatec Laboratory , UK ) and results were scored as negative , + , ++ or +++ as per manufacturer recommendations . 10 mls of urine was filtered using paper filters ( Gelman Sciences , Michigan USA ) and the egg count was recorded per 10 mls of urine . In close collaboration with the district health office , all children who were found to be positive either through microscopic examination or urine reagent strips were referred for treatment . Data were entered into Microsoft Access 2003 . The data were then converted into SAS version 9 . 1 ( SAS Institute , Cary , North Carolina , USA ) which was used for all analyses . Based on the study design of the survey; weighting , stratification , and clustering were taken into account in all statistical analyses using Survey procedures in SAS . The survey analysis procedures use the Taylor series expansion method to estimate sampling errors of estimators based on sample designs [20] . The procedures estimate the variance from the variation among the schools and pool stratum variance estimates to compute the overall variance estimate . A weighting factor was used in the analysis to reflect the likelihood of sampling each student in a stratum . The weight used for estimation is given by the following formula:Where W1 = the inverse of the probability of selecting the school in a stratum and , W2 = the inverse of the probability of selecting a child from the classroom within the school . Bivariate analyses were used to estimate the crude odds ratios and identify variables to be included in the initial multivariate logistic regression model . Chi-square test was used to assess the association between categorical variables and S . hematobium infection in bivariate analysis . Variables that had a p value <0 . 20 were included in the initial multivariate logistic regression model . Using backward elimination method , variables that showed independent association with S . hematobium infection at a significance level of p value <0 . 05 were retained in the model . A significance level of p value <0 . 05 was used in all other analyses . Pearson correlation coefficient was used to test for correlation between continuous variables . The study was conducted using protocols approved by the institutional ethics Review Boards of the University of North Carolina at Chapel Hill , USA and the University of Malawi , College of Medicine . Prior to conducting the study , aims and procedures to be used to collect data were explained to parents or guardians and community leaders including school committees during meetings . Written consent was obtained from the local leaders , children's parents or guardians and assent was subsequently obtained from the children . The mean age of the study population was 10 . 5 years and most of the children ( 86 . 2% ) were aged 8 to 13 years with no significant difference by gender . About 44 . 2% were from poor SES while 21 . 9% of the children were from high socio-economic households . Characteristics of the study sample are shown in Table 1 . S . hematobium eggs were found in 10 . 4% ( 95% confidence interval ( CI ) : 5 . 4–15 . 4% ) of the pupils who submitted urine specimen . Prevalence in schools ranged from 0 . 0% to 46 . 0% and 39 . 1% ( 9/23 ) of the schools had infection prevalence of ten percent or more ( Figure 1 ) . Prevalence varied by altitude above sea level; 17 . 8% of children from <500 m stratum had positive results for S . hematobium ova , 12 . 4% in the 500–1000 m stratum and 3 . 60% in the >1000 m stratum ( p = 0 . 007 ) . Infection was higher in rural areas at 14 . 4% ( 95% CI: 6 . 6–22 . 2% ) compared to the urban areas 3 . 6% ( 95% CI: 1 . 4–5 . 8% ) . 13 . 2% of the boys in the study population were infected compared to 7 . 4% among the girls . Across all age ranges , boys had higher prevalence compared to girls ( Figure 2 ) . The prevalence showed an increasing trend with increasing age and peaked around 11–13 years then started to decline . Boys were more likely to be knowledgeable of existence of open water source in their area compared to girls , 61 . 8% and 55 . 0% respectively ( p = 0 . 03 ) . Among pupils who were knowledgeable of an open water source in their area , children aged 14 years or older were less likely to report playing or swimming in water compared to those less than 14 years old , 79 . 1% and 90 . 8% respectively ( p = 0 . 003 ) . There was no significant difference in the proportions that reported daily/ weekly water contact frequency among boys and girls who reported playing or bathing in open water sources , 85 . 6% and 88 . 1% respectively ( p = 0 . 4 ) . Children who reported previous history of urinary schistosomiasis infection were more likely to report ever being treated for urinary schistosomiasis in the past compared to children who reported no previous history of schistosomiasis ( 27 . 1% and 3 . 7%; p<0 . 001 ) . The overall ova density ( eggs/10 ml of urine ) among the infected in the study population was 10 . 1 ( range 1–80 ) . The prevalence of heavy infection ( ≥50 eggs/10 mls ) among those infected was 2 . 4% . The average ova density was calculated for each school and was compared with school prevalence . There was a strong correlation between prevalence of S . hematobium infection and mean ova density in schools ( Pearson correlation coefficient r = 0 . 65 p = 0 . 005 ) . There was no correlation between age and ova density among those infected . There was no difference in the mean ova density between boys and girls ( mean; 9 . 9 and 10 . 4 eggs/10 mls urine p = 0 . 8 ) . Of the 1 , 150 pupils interviewed , 353 ( 31 . 6% , 95%CI: 23 . 1–40 . 1% ) reported passing blood in urine ( hematuria ) over the past month . Proportion of pupils reporting passing blood in urine ranged from 8% to 70% in schools . Using urine microscopic examination as a gold standard , the sensitivity and specificity of self-reported hematuria was 68 . 3% and 73 . 6% respectively . The negative predictive value ( NPV ) was 95 . 0% while positive predictive ( PPV ) value was very low at 23 . 9% . PPV of self-reported hematuria was higher among boys compared to girls ( 27 . 3% versus 18 . 4% ) and was higher among pupils aged 9 years or more compared to those aged 8 years old or less ( 17 . 0% versus 24 . 8% ) . Table 2 shows the association between risk factors and S . hematobium infection in bivariate analyses . The following factors were significantly associated with infection in bivariate analysis; age 11–13 years , socio-economic status , gender , location , household water source , pupil's knowledge of existence of an open water source , previous history of urinary schistosomiasis and proximity of school to an open water source . Distance from home to an open water source , previous history of schistosomiasis treatment , frequency of water contact and dysuria were not significantly associated with infection . In multivariate analysis , factors that remained significantly associated with S . hematobium infection were male gender ( OR 1 . 81; 95% CI 1 . 06–3 . 07 ) , child's knowledge of an open water source in the area ( OR 1 . 90; 95% CI 1 . 14–3 . 46 ) , previous history of urinary schistosomiasis ( OR 3 . 65; 95% CI 2 . 22–6 . 00 ) distance of less than 1 km from school to nearest open water source ( OR 5 . 32; 95% CI 1 . 66–17 . 07 ) and age 8–13 compared to those age 14 and older ( Table 3 ) . To our knowledge this is the first large-scale survey that has been conducted in Blantyre district to determine prevalence and risk factors associated with S . hematobium infection . The prevalence of S . hematobium infection was 10 . 4% and ranged from 0 . 0% to 46% in schools in our study . Male gender , age , child's knowledge of an open water source in the area , previous history of schistosomiasis and proximity of school to an open water source were independently associated with infection in our study . The last published S . hematobium prevalence estimates for the district were from studies conducted between 1979 and 1981 among pupils from three schools and individuals from two villages where prevalence ranged from 22% to 72% using centrifugation and reagent dipsticks respectively [10] . Our study prevalence estimate is lower than the national estimates of between 40% and 50% currently in use . Our prevalence findings are similar to the national schistosomiasis survey results [11] . The observed differences between the national estimates in current use and the recent observed findings have been attributed to effective schistosomiasis control and availability of anti-schistosomiasis drugs [11] . In addition the differences in size and methodology of studies could partly explain the observed findings . The more recent large-scale surveys are likely to be more accurate compared to previous studies that may have had some selection bias for high risk schools [11] . Also , the wide range of infection prevalence rates among schools in our study illustrates the focal distribution characteristic of schistosomiasis . The association between male gender and S . hematobium found in our study is similar to findings from other studies [21]–[23] . Boys were more likely to be infected and be knowledgeable of the existence of an open water source in their area compared to girls . In other studies , boys had more water-contact compared to girls [23] . This could be partly explained by the fact that boys are usually more adventurous , they are more likely to be knowledgeable of their environments including water bodies and therefore be more likely to play in them compared to girls . The association between gender and S . hematobium infection varies in different communities . Some studies have reported no association between S . hematobium infection and gender [22] , [24] while other studies have reported association with female gender [25] . Our study showed that there was an increasing trend of infection among children from six years to thirteen years with a decline from 14 years . Also , children aged 14 years or more were less likely to report playing or swimming in water in our study . This suggests that children 14 years and older have lower risk of being infected as they are less likely to be engaged in recreational water-contact behaviors compared to younger children . Studies conducted elsewhere have reported similar results , for example , throughout the nine year schistosomiasis study in Kenya , schistosomiasis infection was lower in older children [22] , [26] . Other studies have reported that age-acquired immunity to re-infection contributes to the declining trend in prevalence among children aged 15 years and older [27] . However , we cannot conclude whether age-acquired immunity contributed to our findings since we did not find a negative association between ova load and age . School proximity to an open water source showed a very strong association with infection . Proximity to water sources has been consistently shown to be associated with schistosomiasis infection in other studies [23] , [28] , [29] . In contrast , other studies have found no association or variable influence on schistosomiasis infection and it has been suggested that multiple uses of various water bodies with different transmission levels could explain these findings [30] , [31] . Interestingly , we did not find a significant association between proximity of household to an open water source in our study . This could be partly explained by the following reasons; first , 95 . 0% of pupils who were infected and reported that their homes were far or very far from an open water source attended schools that were less than 1 km from an open water source . This means that while children may not engage in water contact recreational activities close to their homes , they might have been exposed when going or coming from school or other places away from home . Secondly , our questionnaire method of estimating distance from home to nearest water source was more subjective and prone to misclassification . Hence our estimate for the effect of household proximity to water source on infection could have been biased . Communities with high infection rates are usually clustered around contaminated water sources [32] , [33] . This could partly explain why we observed an association between S . hematobium infection and previous history of urinary schistosomiasis since children who reported previous history of schistosomiasis were more likely to report ever being treated for urinary schistosomiasis . We did not find a significant association with reported water contact frequency . Studies have found variable influence of water contact frequency on schistosomiasis infection . This has been attributed to variability in amount of body exposed to water and others have suggested that other factors play major role [22] , [34] . Socioeconomic status of the household was not an independent factor associated with infection in our study population . Conflicting results have been reported in previous studies [35]–[37] . Based on our findings , possibly improving socioeconomic status alone may not significantly reduce the rate of infection among this population . The performance of self reported hematuria in our study is similar findings from other endemic countries with sensitivity and specificity range of 50–100% and 58–96% respectively in moderate and high transmission areas [38]–[40] . Our positive predicted value of self-reported hematuria was low and could be due to several reasons . First , our study was limited since only one urine specimen was collected and we did not encourage the children to conduct exercises prior to urine collection . Studies have shown that repeated examination of urine specimen over consecutive days and exercises prior to urine collection improve egg detection [41] , [42] . It is possible to have an infected child reporting hematuria with no ova detected one urine sample . Also , this could have been exacerbated by the fact that our study population had low infection intensity ( 10 . 1 eggs/10 mls urine ) . Our results probably could have been different if reported hematuria was compared to Circulating Anodic Antigen ( CAA ) from S . haematobium . Secondly , children that either reported passing blood in urine or tested positive for S . hematobium infection were referred for treatment and it might be possible that some children reported hematuria to be referred to health facilities and hence affecting accuracy . This study was initiated and implemented by the district health office . It demonstrates the success of the persistent and deliberate national efforts advocating for active participation at the sub-national or sub-district level in schistosomiasis prevention and control activities in countries where schistosomiasis is endemic . This is crucial for long-term ownership and sustainability of schistosomiasis control efforts . S . hematobium infection was found to be localized in Blantyre district . The district health office in collaboration with the district education offices may consider targeted treatment every two years in schools and surrounding communities with prevalence of between 10% but less than 50% as recommended by WHO in addition to passive treatment in the health facilities and snail control [12] . Even though the predictive value for the self-reported hematuria was low , the sensitivity and specificity was comparable to other previous studies and suggests that the use of questionnaires in this part of Sub-Saharan Africa may still valuable in older children [43] . Further intervention studies to determine the best and cost-effective strategies to provide treatment to children and communities in the affected areas are required . Ecological studies are also needed to identify transmission foci to facilitate implementation of ecologically targeted control measures .	Schistosoma hematobium infection is a parasitic infection endemic in Malawi . Schistosomiasis usually shows a focal distribution of infection and it is important to identify communities at high risk of infection and assess effectiveness of control programs . We conducted a survey in one district in Malawi to determine prevalence and factors associated with S . hematobium infection among primary school pupils . Using a questionnaire , information on history of passing bloody urine and known risk factors associated with infection was collected . Urine samples were collected and examined for S . hematobium eggs . One thousand one hundred and fifty ( 1 , 150 ) pupils were interviewed , and out of 1 , 139 pupils who submitted urine samples , 10 . 4% were infected . Our data showed that male gender , child's knowledge of an existing open water source ( includes river , dam , springs , lake , etc . ) in the area , history of urinary schistosomiasis in the past month , distance of less than 1 km from school to nearest open water source and age 8–10 years compared to those 14 years and older were independently associated with infection . These findings suggest that children attending schools in close proximity to open water sources are at increased risk of infection .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "public", "health", "and", "epidemiology/epidemiology", "public", "health", "and", "epidemiology/infectious", "diseases" ]	2009	Prevalence Distribution and Risk Factors for Schistosoma hematobium Infection among School Children in Blantyre, Malawi
Edwardsiella piscicida is a leading fish pathogen that causes significant economic loses in the aquaculture industry . The pathogen depends on type III and type VI secretion systems ( T3/T6SS ) for growth and virulence in fish and the expression of both systems is controlled by the EsrB transcription activator . Here , we performed a Tn-seq-based screen to uncover factors that govern esrB expression . Unexpectedly , we discovered that RpoS antagonizes esrB expression and thereby inhibits production of E . piscicida’s T3/T6SS . Using in vitro transcription assays , we showed that RpoS can block RpoD-mediated transcription of esrB . ChIP-seq- and RNA-seq-based profiling , as well as mutational and biochemical analyses revealed that RpoS-repressed promoters contain a -6G in their respective discriminator sequences; moreover , this -6G proved critical for RpoS to inhibit esrB expression . Mutation of the RpoS R99 residue , an amino acid that molecular modeling predicts interacts with -6G in the esrB discriminator , abolished RpoS’ capacity for repression . In a turbot model , an rpoS deletion mutant was attenuated early but not late in infection , whereas a mutant expressing RpoSR99A exhibited elevated fitness throughout the infection period . Collectively , these findings deepen our understanding of how RpoS can inhibit gene expression and demonstrate the temporal variation in the requirement for this sigma factor during infection . Edwardsiella piscicida ( formerly included in Edwardsiella tarda ) belongs to the enterobacteriaceae family [1] and is phylogenetically related to Salmonella enterica [2] . Like some species of Salmonella , E . piscicida can also infect a broad range of animal hosts including , fish , amphibians , mammals and humans [3] . The organism is a bane of the aquaculture industry because it infects over 20 species of fish , including important farmed species such as turbot , flounder , eel and catfish , resulting in significant economic losses globally [4–6] . Several E . piscicida virulence determinants , such as adhesins , siderophores , and hemolysin EthA have been uncovered using single mutants ( reviewed in [7] ) and in genome-wide transposon insertion sequencing ( Tn-seq ) -based studies [8] . Like S . enterica , E . piscicida can grow intracellularly [9–10] . The pathogen relies on its type III and type VI secretion systems ( T3/T6SSs ) to translocate a repertoire of ~20 putative and known effectors into host cells to occupy this niche [11–13] . Genome-wide analysis revealed that , among the 33 putative two-component system ( TCS ) encoded in the E . piscicida genome , EsrA-EsrB is indispensable for E . piscicida pathogenicity . This TCS controls the expression of the pathogen’s T3/T6SS machineries and their respective effectors , as well as the expression of an additional ~990 genes , some of which have roles during infection [13–14] . For example , EsrB-activated genes are associated with iron sequestration and uptake ( hemin uptake and siderophore-mediated iron uptake systems ) , while genes for basal metabolism were directly downregulated by EsrB [13] . Although the host signals that activate the EsrA-EsrB TCS are unknown , several regulators , including EsrB , PhoP [14] , PhoR , and Fur [15] , are known to modulate expression of esrB . Mutation of esrAB has been a fruitful strategy for development of live attenuated vaccines against edwardsiellosis in fish [16–17] . Furthermore , in S . enterica , the EsrAB homologs SsrAB play a critical role in regulating virulence , and homologs in Sodalis glossinidius facilitate its endosymbiont lifestyle [18] . Thus , a systematic dissection of the upstream and downstream regulatory networks in which EsrAB is embedded will further our understanding of E . piscicida pathogenicity . This knowledge will also potentially facilitate vaccine development as well expand our knowledge of the evolution of signal transduction networks that govern virulence in diverse Gram-negative pathogens . Here , we used a Tn-seq-based screen to identify upstream regulators of esrB expression . Surprisingly , we found that the alternative sigma factor RpoS ( σS/38 ) inhibits esrB expression and thus exerts negative control over the expression of E . piscicida’s T3/T6SSs . RpoS , like other sigma factors , associates with core RNA polymerase ( RNAP or E ) , enabling RNAP promoter recognition during initiation of transcription [19–21] . In many Gram-negative bacteria , RpoS enables transcription of genes associated with the general stress response and stationary phase metabolism [22–25] and our work suggests that is the case in E . piscicida , as well . While RpoS is also known to negatively regulate gene expression [26–28] , there have been few studies of the mechanisms by which RpoS can inhibit transcription [29] . Using a variety of approaches to investigate how RpoS represses esrB transcription we found that it can antagonize RpoD ( sigma factor 70 ) -mediated transcription of esrB . Notably , the presence of the -6G in the esrB promoter discriminator ( a sequence found between the -10 element and the transcription start site [30] ) was required for RpoS’ repressor function . RpoS R99 , a residue predicted to interact with -6G , was required for this sigma factor to inhibit transcription of esrB , but RpoSR99A still enabled transcription from other promoters . Finally , studies in a turbot infection model indicate that the requirement for RpoS varies during the course of infection . Since EsrB is a key regulator of T3SS and T6SS in E . piscicida , we devised a Tn-seq [8] based screen to identify genes that control its expression . Initially , we created a reporter of the esrB promoter , by fusing the 500 bp segment located upstream of the esrB start codon to a kanamycin ( Kan ) resistance gene ( yielding PesrB-kan ) . This reporter was introduced into a neutral site ( between glms and ETAE_3537 ) on the E . piscicida strain EIB202 chromosome ( Fig 1A ) . Previous studies showed that introduction of DNA into this site does not alter growth [8] . Then , we created a high-density Himar [31] transposon insertion library in this strain ( WT::PesrB-kan ) . The library was cultured in DMEM , a medium that induces the expression of EsrB [13] , in either the absence ( input ) or presence ( output ) of Kan ( Fig 1A ) . High-throughput sequencing was used to identify the sites and enumerate the frequency of insertions in the input and output libraries . Mutants that are present in the input but not the output library should in principle contain insertions in loci critical for EsrB expression; conversely , mutants that are present at greater frequency in the output , represent insertions in genes that ordinarily repress EsrB expression . To estimate an optimal concentration of Kan to use for screening the library , we compared the minimum inhibitory concentrations of WT::PesrB-kan and a derivative with a deletion of esrB ( ΔesrB::PesrB-kan ) . The latter strain was used because EsrB is known to promote esrB expression [14] . The two strains had similar growth in the absence of Kan , but with increasing concentrations of Kan the ΔesrB::PesrB-kan strain exhibited progressive growth defects relative to WT::PesrB-kan ( S1A Fig ) . At a Kan concentration of 600 μg/ml , growth of the WT::PesrB-kan strain was not decreased , whereas growth of the ΔesrB::PesrB-kan strain was markedly inhibited; thus , Kan 600 μg/ml was used for the screen . A plot of the percentage of TA sites disrupted per gene vs the frequency of genes showed that the input library had a high degree of saturation , where the majority of non-essential genes had ~60% of TA sites disrupted ( S1B Fig ) [8] . We compared the transposon distribution profiles in the input and output libraries with the Con-ARTIST pipeline [32] to identify genes that were either under- or over-represented ( \|log2 ( FC ) \| > 1 and P < 0 . 05 ) in the output library; such ‘conditionally depleted’ or ‘conditionally enriched’ genes represent candidate loci that likely promote or inhibit expression of EsrB respectively . There were 23 genes with a greater abundance of insertions in the DMEM+Kan cultures and 16 genes with a diminished abundance of insertions in the Kan-containing medium ( Fig 1B , S1 Table and S2 Table ) . As expected , there were fewer insertions in esrB in the output library then in the output , but the difference did not reach the 2-fold threshold . The screen did not yield other known regulators of esrB expression such as PhoP and Fur , likely because these genes are required for growth in the conditions used for the screen [8] and thus were not included in our analyses . To validate a subset of the screen hits , we picked 7 insertion mutants present in a defined EIB202 transposon library created in our lab that were hits in the screen . qRT-PCR was used to measure esrB transcript levels in the WT and the mutants and in all cases the results were consistent with the findings from the screen ( S1C Fig ) . These observations suggest that many of the 39 genes identified in the screen play a role modulating the expression of the global regulator , esrB . Furthermore , 6 of the hits were in genes encoding hypothetical proteins ( S1 Table and S2 Table ) , suggesting that future studies defining the functions of these proteins will shed light on the pathways controlling esrB . However , it is possible that some of the hits , such as in cpxR [33] , are attributable to the stresses imposed by kanamycin itself . Notably , the abundance of insertions in the gene encoding the RpoS sigma factor ( rpoS ) was greater in the output vs the input library ( FC ratio = 8 . 3 ) , suggesting that RpoS is a repressor of esrB ( Fig 1C ) . The diminished abundance of insertions in the gene encoding the ATP-dependent protease Lon in the output library ( FC = 0 . 16 ) ( Fig 1D ) is consistent with idea that RpoS inhibits esrB expression , since Lon is an established negative regulator of RpoS [34] . To further investigate RpoS control of esrB expression , we constructed an rpoS deletion mutant in the WT::PesrB-kan background ( ΔrpoS::PesrB-kan ) , as well as a strain where this deletion was complemented ( rpoS+::PesrB-kan ) . The ΔrpoS::PesrB-kan mutant exhibited significantly higher ( minimum inhibition concentration [MIC] of 1200 μg/ml ) resistance to Kan than either the WT::PesrB-kan or rpoS+::PesrB-kan strains ( MIC of 600 μg/ml ) , consistent with the idea that RpoS represses the activity of the esrB promoter . Similarly , when the esrB promoter was fused to luxAB ( PesrB-luxAB ) , enabling esrB promoter activity to be measured as fluorescence , there was significantly greater fluorescence detected in the ΔrpoS strain than in WT or rpoS+ complemented strains ( Fig 2A ) , demonstrating that esrB promoter activity is inhibited by RpoS , particularly in cells entering stationary phase at 9 h , when RpoS is highly induced [35–36] . The response regulator EsrB is critical for the expression of T3/T6SS in E . piscicida EIB202 [14] . To begin to assess the consequences of RpoS inhibition of esrB expression on T3/T6SS-related functions , we compared the extracellular protein profiles of several strains including WT , ΔesrB , ΔrpoS and rpoSOE; in the latter strain , rpoS is driven by , the promoter for the 30S ribosomal protein , to enhance expression of this sigma factor [2] . As anticipated , T3/T6SS proteins were over-produced in the ΔrpoS mutant , while there were reduced yields of T3/T6SS proteins in rpoSOE compared to the WT strain ( Fig 2B ) . There was no detectable T3/T6SS secreted products when esrB was deleted from the rpoSOE background , providing additional support for the idea that RpoS repression of genes related to T3/T6SS acts through esrB ( Fig 2B ) . Moreover , RpoS inhibition of T3/T6SS secretion was circumvented in a strain constitutively expressing esrB driven by the Plac promoter ( rpoSOE::Plac-esrB ) , indicating that RpoS repression of virulence factor production is dependent on the esrB promoter region ( Fig 2B ) . We also tested whether the same set of strains used in Fig 2B exhibited auto-aggregation , a phenotype attributable to production of EseB , a T3SS apparatus protein [37] . The pattern of auto-agglutination and production of EseB in these 6 strains ( Fig 2C ) mirrored secretion of T3/T6SS products and is consistent with idea that RpoS exerts negative control over esrB expression . We used RNA-seq to elucidate the RpoS regulon in E . piscicida by comparing the transcriptomes of the WT and ΔrpoS strains . Transcripts of 729 genes were differentially ( \|log2 ( FC ) \| > 1 and P < 0 . 05 ) expressed in the two strains , including 532 genes whose transcripts were apparently up-regulated by RpoS and 197 genes whose transcripts were apparently down-regulated by RpoS ( 2920 genes were not differentially expressed ) ( Fig 2D , S3 Table and S4 Table ) . As expected from the results above , many genes in the T3/T6SS gene clusters had higher transcript levels in the absence of rpoS , consistent with the idea that their expression is down-regulated by RpoS ( Fig 2D , S4 Table ) ; these observations were corroborated with qRT-PCR assays ( Fig 2E ) . As reported in S . enterica [29] , the sdh gene cluster was down-regulated by RpoS . Transcripts of genes related to ferric iron uptake were also less abundant in WT vs ΔrpoS . Since EsrB is known to activate the ferric iron uptake system [13] , this observation is likely also explained by RpoS repression of esrB transcription . Collectively , these protein- and mRNA-based assays are all consistent with the idea that RpoS inhibits the expression of T3/T6SS by repressing the expression of EsrB . RpoS production and activity is directly or indirectly regulated through a variety of mechanisms , including the action of the Lon protease [22–25 , 34 , 38] . Lon over-expression ( lonOE ) led to diminished levels of RpoS ( Fig 3A , top ) and to the concomitant expected elevations in esrB transcripts ( Fig 3A , bottom ) and the amounts of extracellular T3/T6SS proteins and transcripts encoding T3/T6SS proteins EseB and EvpC relative to those detected in the WT ( Fig 3B and 3C ) . Similar amounts of extracellular T3/T6SS proteins were observed in the ΔrpoS lonOE as in the ΔrpoS strain consistent with the idea that lon over-expression modulates T3/T6SS production by depleting RpoS levels . RpoS is a critical alternative sigma factor involved in the response to a variety of stresses including , starvation , low temperature , and reactive oxygen species ( ROS ) [22–24] . When we cultured the WT and ΔrpoS strains harboring a PesrB-luxAB reporter under various stress conditions , we observed that RpoS levels and PesrB activities changed in an inverse fashion ( Fig 3D ) . Taken together , these observations strongly suggest that RpoS mediates a link between environmental conditions and the modulation of expression of esrB and its virulence associated regulon . Next we investigated whether RpoS can interact with the esrB promoter region . Initially , a pull-down assay , where biotin labeled PesrB attached to beads was used as bait , was used to test if RpoS binds this region; beads bound to a biotin labeled portion of the esrB open reading frame ( orf ) were used as a negative control . Lysates from a ΔrpoS strain expressing a functional ( S2A Fig , and S2B Fig ) Flag-tagged RpoS ( ΔrpoS/flag-rpoS ) were incubated with the beads and bound proteins were eluted with NaCl . The Flag-tagged RpoS was eluted from the PesrB bait sequence but not from the esrB orf bait ( Fig 4A ) , showing that RpoS can bind to this promoter . A chromatin immunoprecipitation assay ( ChIP ) was performed to investigate whether RpoS binding to PesrB could be detected in vivo . Protein-cross-linked DNA obtained from ΔrpoS cells expressing flag-rpoS or flag alone was immuno-precipitated using an anti Flag-tag antibody . A PCR assay that amplified PesrB was carried out on the input and precipitated DNA from both strains . The immunoprecipitate from the strain expressing Flag-tagged RpoS and not from the strain expressing the Flag tag contained the PesrB amplification product , whereas the input DNA from both strains contained this product ( Fig 4B ) . In addition , no PCR product was detected in the IP in the absence of the anti-Flag antibody ( Fig 4B ) . Together , these findings strongly suggest that RpoS binds to the esrB promoter region in E . piscicida . We presume that RpoS is binding to the esrB promoter as part of RNAP holoenzyme , since sigma factors are not thought to interact with promoters outside of the context of this macromolecular complex [21] . RpoS is a member of the σ70 family of proteins , and its binding motif is similar to that of RpoD ( σ70 ) [39] . In E . coli and other bacteria , these two sigma factors bind to overlapping sites in the -10 region of the promoter [29 , 40] . The RNA-seq profiles obtained above were used to identify the +1 site of the esrB transcript ( Fig 4C ) . As expected , there was greater abundance of esrB transcripts in the ΔrpoS mutant and the predicted -35 and -10 sequences are similar to known RpoS promoter binding sites [24 , 36 , 40] . In vitro transcription reactions were carried out to begin to dissect the molecular determinants of RpoS repression of esrB transcription . For these assays , we used E . coli core RNAP and E . piscicida RpoD , with or without the addition of E . piscicida RpoS in the reaction mixtures . The esrB promoter ( PesrB ) driving esrB was used as the transcription template . The transcripts generated from the different reaction conditions were assessed using reverse transcription ( RT ) PCR with primers that targeted the esrB ORF region . Addition of E . piscicida RpoD , but not RpoS , to E . coli core RNAP was sufficient to drive transcription from PesrB ( Fig 4D , lanes 4–5 ) . Addition of RpoS to the reaction mixtures abolished RpoD-mediated transcription of esrB ( Fig 4D , lanes 6–7 ) , demonstrating that RpoS is sufficient to repress esrB expression driven from PesrB but not to mediate transcription from this promoter . We also engineered mutant forms of the esrB promoter to test the importance of two promoter elements in enabling RpoS to inhibit esrB transcription . In one mutant , PesrB mut1 , the PesrB -10 box AT-rich region , which is thought to be critical for RpoS binding [24 , 36 , 40] , was replaced with CG nucleotides ( Fig 4E ) . An additional variant of PesrB was constructed where the discriminator sites ( GCC ) found immediately downstream of the -10 box were substituted with TAA nucleotides , yielding PesrB mut2 ( Fig 4E ) . The discriminator region plays a role in proper initiation of transcription and transcription start site selection from σ70 dependent promoters [41–42]; furthermore , in S . enterica serovar Typhimurium the discriminator region of the sdh promoter was required for RpoS repression of sdh expression [29] . Transcription from the PesrB mut1 promoter , containing the mutated -10 box , was not detected ( Fig 4D , lanes 8–10 ) , an expected result given the likely importance of this sequence for either RpoD or RpoS to bind the promoter . Interestingly , when the template DNA containing the mutation of the discriminator sequence ( PesrB mut2 ) was used in the reaction , RpoS no longer repressed transcription; in fact , in this setting , RpoS was sufficient to drive transcription in the absence of RpoD ( Fig 4D , lanes 11–15 ) . Thus , at least in the in vitro context , the sequence of the discriminator region in the esrB promoter is critical for determining whether RpoS functions to inhibit or enable esrB transcription . For in vivo correlations of these in vitro observations , we created luxAB reporter genes driven by PesrB or its variants and introduced them into a neutral chromosomal position in the WT , ΔrpoS and rpoSOE strains . Immunoblots established that RpoS abundance was 2–3 fold higher in the rpoSOE strain than in the WT strain ( S2C Fig ) . As expected , the fluorescence from the reporter driven by PesrB was higher in the absence RpoS and lower when RpoS was overexpressed ( Fig 4F ) . There was little detectable fluorescence in any of the backgrounds from the reporter driven by PesrB mut1 , which is expected since neither RpoD nor RpoS bind to this promoter ( Fig 4F ) [24 , 36 , 40] . Fluorescence from the reporter driven by PesrB mut2 was higher than that driven by PesrB in the WT background , a finding which could be attributed to either absence of RpoS repression and/or to RpoS-mediated activation of transcription from this mutant promoter . The former explanation likely accounts for the elevation in the magnitude of expression from this mutant promoter because its fluorescence was unchanged in the ΔrpoS background . However , there was elevated PesrB mut2 activity observed in the strain overexpressing rpoS , which may be attributable to RpoS contributing to transcriptional activation in this context ( Fig 4F ) . Taken together , these data are consistent with the idea that the discriminator region ( GCC ) in the esrB promoter , which is not essential for RpoS binding , is important for RpoS to interfere with RpoD-mediated transcription of esrB . To deepen our understanding of RpoS inhibition of PesrB expression , we used ChIP-seq to define the RpoS regulon in stationary-phase E . piscicida cells grown in DMEM . This analysis revealed that RpoS bound to 57 loci ( S5 Table ) . Besides esrB , genes enriched by RpoS ChIP included rpoS , sdhC , bglG , and mdtJ ( S3A Fig ) . Using MEME-ChIP [43] , a conserved AT-rich RpoS binding motif and a putative -10 box and -35 box were identified ( S3B Fig ) . Combined with the RNA-seq data , these analyses enabled identification of 21 genes whose expression are likely directly regulated by RpoS; esrB and sdhC were the only candidate targets of direct RpoS repression and 19 candidate RpoS-dependent genes were identified ( Fig 5A ) . We compared the motifs representing the genes activated and repressed by RpoS and found that the -10 box , TAYacT ( -12 to -7 sites ) were similar , whereas the -6 to -4 sites ( relative to -10 box ) were distinct in the motifs derived from the activated and repressed genes ( Fig 5B ) . Examination of the RNA-seq data for genes containing the RpoS binding motif ( S3B Fig ) in their promoter regions revealed an additional 16 candidate genes directly regulated by RpoS ( P < 0 . 001 ) . 4 of these genes are putatively repressed and 12 activated by RpoS ( S4A Fig and S4B Fig ) . The RpoS repressed genes usually contain a GCG discriminator sequence whereas the activated genes harbor a distinct and somewhat more variable discriminator sequence ( often TAA ) ( Fig 5B ) . Notably , all the repressed genes contain a -6G and a -5C ( Fig 5B and S4B Fig ) . Chi-square tests revealed a significant difference ( P < 0 . 001 ) in the occurrence of G and C nucleotides at the -6 and -5 sites of the repressed genes vs the activated genes ( not GC ) ( S4C Fig and . S4D Fig ) , and in the elevated frequency of G vs non -G in the -6 site in the discriminator region of repressed vs activated genes ( P < 0 . 001 ) ( S4E Fig ) . The above analyses suggested that the sequence of the discriminator region of RpoS-regulated promoters , in particular the presence of -6G and -5C could determine if RpoS acts as repressor at the respective promoter . To test this idea , we used site-directed mutagenesis to introduce changes in the -6 to -4 sites as well as in the -10 box in the esrB promoter region fused to a promoterless luxAB reporter . These reporters were introduced into a ΔesrB strain and the resulting bioluminescence was measured ( Fig 5C ) . Consistent with the findings in Fig 4F , modifications in the -10 box ( Mutant 1 ) abolished transcription of PesrB . Substitution of -6G to -6T , A or C alone or together with additional substitutions in -4 and -5 sites all significantly enhanced transcription from PesrB ( Fig 5C ) . However , substitutions in the -5 or -4 sites did not alter transcription of PesrB as long as there was a -6G ( Fig 5C ) . Similarly , substitution of -6G to T in the promoters of two additional RpoS repressed genes , sdhC and 1580 that did not affect the RpoS level in the cells , abolished RpoS repression ( S4F Fig and S4G Fig ) . Taken together , these results demonstrate that the -6G in the discriminator region of RpoS associated promoters is critical for this sigma factor to function as a repressor . We used molecular simulations to model how E . piscicida RpoS interacts with esrB promoter DNA . RpoS was aligned with the Mycobacterium smegmatis RNA polymerase sigma factor σA ( 5VI5 , Chain F ) [44] . The alignment ( which is close , root-mean-square deviation of 2 . 065 Å over 215 Cα atoms ) , places the RpoS DNA binding helices close to the homologous helices in σA . The E . piscicida RpoS residues R99 and L61 are predicted to be in close proximity with -6G , whereas residues D55 , T57 , Q58 and L61 are near -5C ( Fig 6A ) . We focused on R99 and L61 , and constructed strains overexpressing three RpoS substitution variants , rpoSR99A , rpoSL61A , and rpoSL61AR99A in the ΔrpoS background ( Fig 6B , top ) . The PesrB-luxAB reporter was used to monitor the effects of these mutations on PesrB expression . Notably , the R99A substitution abolished RpoS’ capacity to repress PesrB expression , but the L61A substitution did not ( Fig 6B ) ; similar to RpoSR99A , the L61AR99A double substitution did not repress the esrB promoter ( Fig 6B ) . Moreover , in vitro transcription assays with PesrB or its variants ( PesrB mut1 and PesrB mut2 ) as the template in a mixture of the RNAP core enzyme , RpoD and RpoSR99A demonstrated that the R99A substitution mutation in RpoS abolished its capacity to repress esrB transcription; however , unlike RpoS , RpoSR99A could enable transcription from PesrB ( Figs 4D and 6C , lanes 1–4 ) . Thus , since RpoSR99A is capable of supporting transcription , its failure to repress transcription from PesrB is not simply explained by ablation of its capacity to interact with DNA . Moreover , these results strongly suggest that RpoS R99 interaction with the -6G in the esrB promoter discriminator sequence is a critical determinant of whether this sigma factor impedes transcription . Consistent with the finding that rpoSR99A and rpoSL61AR99A did not repress esrB expression , extracellular levels of T3/T6SS secreted proteins were elevated in strains expressing these RpoS variants ( Fig 6D ) . We compared the capacity of WT , ΔrpoS , rpoSOE , and rpoSR99A to resist H2O2 as a way to begin to compare the capacity of RpoSR99A to activate genes in the RpoS regulon . Interestingly , the MIC of the rpoSR99A mutant was higher than WT and identical to that found in rpoSOE , suggesting that RpoSR99A can still function as an activator ( Fig 6E ) . Moreover , qRT-PCR analyses confirmed that overexpression of RpoSR99A led to elevated transcript levels of genes that are ordinarily RpoS-activated as well as those that are ordinarily repressed by RpoS ( Fig 6F ) . These analyses demonstrate that RpoSR99A retains the ability to activate transcription of genes that are usually upregulated by RpoS , but has lost the capacity to function as a repressor , lending additional support for the idea that the interaction of R99 with -6G is a critical requirement for RpoS to function as a repressor . We used in vivo fluorescence imaging to investigate RpoS repression of T3/T6SS expression during E . picicida infection of turbot , a natural host [2] . Luciferase reporters of PeseB- , PevpA- , PrpoS- , and PesrB- expression were introduced into a neutral position on the chromosome of WT , ΔrpoS and rpoSR99A strains , and these strains were inoculated i . p . into turbot fish at the same dose and fluorescence was measured 8 days post infection ( dpi ) . The PesrB-luc fusion did not generate sufficient fluorescence for in vivo monitoring probably because of the low transcript level of esrB ( Figs 2E and 4C ) , but the fusions to the eseB and evpA promoters , whose expression is activated by EsrB , were sufficiently active and serve as indirect measures of esrB expression [13] . As expected , there was little PrpoS-luc activity detected in the ΔrpoS background because of RpoS auto-activation; in contrast , and as observed in vitro ( Fig 7A ) , there was greater PrpoS-luc activity in vivo in the rpoSR99A strain than in the WT strain ( Fig 7A and 7B ) . There was significantly greater fluorescence produced by the PeseB and PevpA fusions in the ΔrpoS and the rpoSR99A strains than in the WT strain ( Fig 7A and 7B ) . These observations mirror the in vitro findings and demonstrate that RpoS represses the EsrB regulon during infection . Furthermore , they show that the RpoS Arg99 residue is required for its repressor activity in vivo during infection . Thus , at least at 8 dpi , RpoS negatively regulates in vivo virulence factor expression . Despite the elevated virulence gene expression in the ΔrpoS mutant , there was ~2x-fewer ΔrpoS CFU recovered from infected fish than the WT and the rpoSR99A strains ( Fig 7B ) , suggesting that RpoS activated genes may also contribute to E . picicida growth at some points during infection . In vivo competition experiments were also carried out to elucidate whether RpoS regulation is required for optimal E . piscicida fitness during infection . The relevant WT comparator strains for these experiments were WT cured of the endogenous R plasmid pEIB202 ( WT ΔP ) , which is known to be proficient at colonization [8] , and the WT with an empty stable pUTat ( WT/pUTat ) [45] . Either of these control strains was inoculated in 1:1 mixtures with different test strains in turbot fish . Each of these strains grew equivalently as assessed in in vitro competition assays in LB ( S5 Fig ) . The ratios of the strain mixtures in livers , the organ with the most robust colonization , were determined 8 dpi . As previously observed [8] , ΔesrB was markedly outcompeted in vivo . The ΔrpoS mutant had a modest ( ~2 . 5 fold ) but significant colonization defect ( Fig 7C ) , suggesting either that over-expression of RpoS repressed genes or absence of expression of RpoS-activated genes is detrimental for optimum growth in vivo . The in vivo growth of rpoSOE , the strain over-expressing RpoS , was more severely attenuated than the ΔrpoS strain ( Fig 7C ) , consistent with the idea that the relief of RpoS repression of the EsrB virulence regulon is critical for the pathogen to grow in vivo . Together , these experiments reveal that RpoS regulation is necessary for E . piscicida optimal growth in vivo . Additional in vivo competition experiments were carried out to more directly assess whether RpoS control of EsrB expression contributes to E . piscicida fitness in vivo . A strain constitutively expressing EsrB in the rpoSOE background ( rpoSOE::Plac-esrB ) competed equally with the WT ( CI~1 , Fig 7C ) , strongly suggesting that the enhanced repression of esrB in rpoSOE accounts for the attenuation of this strain . Conversely , a strain where the native esrB promoter was substituted with PesrB mut1 ( ΔesrB/PesrB mut1-esrB ) exhibited a colonization defect similar to that exhibited by the ΔesrB mutant; this observation is consistent with observations shown above ( Fig 4D ) that this promoter does not support esrB expression . However , the strains containing PesrB mut2 or PesrB mut3 , both of which support esrB expression ( Figs 4D and 5C ) , substituted for the native esrB promoter ( ΔesrB/PesrB mut2-esrB and ΔesrB/PesrB mut3-esrB respectively ) showed no colonization defects ( Fig 7C ) . Notably , the in vivo colonization of the rpoSR99A strain was comparable to that of the WT at 8 dpi . Coupled with the results shown in Fig 7A , these observations provide strong support for the idea that relief of RpoS-mediated repression of esrB expression , and consequent expression of the EsrB regulon ( e . g . T3/T6SS expression ) is critical for E . piscicida growth in vivo . E . pisicicida can cause chronic infections in turbot and during the course of such infections the genetic requirements for fitness are dynamic [8 , 13 , 16] . Prior studies in a zebra fish model revealed that a ΔrpoS mutant did not exhibit significant attenuation 5 dpi [35] . We monitored the fitness of ΔrpoS , ΔesrB , and rpoSR99A mutants relative to WT E . piscicida during a 2-week infection in turbot with time series CI analyses ( Fig 7D ) . Consistent with previous PACE-based analyses of genome-wide fitness profiles during chronic E . piscicida infection of turbot , the ΔesrB strain mutant did not show a defect in growth in vivo until ~5 dpi and after this point its fitness continued to decline ( Fig 7D ) [8] . Remarkably , the ΔrpoS mutant exhibited the inverse pattern; i . e . , it was most attenuated early in infection ( at 1–2 dpi ) , but later , the mutant recovered and by 8–14 dpi it exhibited equal fitness as the WT ( Fig 7D ) . The inverse kinetics of the requirements for rpoS and esrB support a model where rpoS is required early in infection to activate genes required for adaptation to host-derived stresses ( e . g . rpoS , cadA1 , cadB1 , cadB3 , uspB , cspA , cspG , cspH , cspI , speAB , speG , trxC , dps , phoR , csrA ) ; later , presumably at the point when the pathogen begins to occupy the niche where T3/T6SS enable growth , rpoS becomes dispensable , because its repression of esrB inhibits production of these secretion systems . Measurement of RpoS and EseB amounts in liver homogenates from fish infected with WT E . piscicida generally support the idea the requirement for RpoS wanes during the course of infection . The levels of RpoS peaked at ~8 dpi and then declined , whereas EseB levels peaked on ~11 dpi and remained elevated ( S6 Fig ) . The fitness profile of the rpoSR99A strain , which slightly out-competed the WT throughout the 14-days of observation , also supports the idea that repression of esrB must be relieved during the course of infection . As shown above , this variant rpoS is able to promote expression of genes whose transcription require this sigma factor ( Fig 6F ) , but it does not repress esrB . Thus , there may be host signals that lead to inhibition of RpoS expression/activity after E . pisicicida initially establishes itself within the host environment . Taken together , these data suggest that E . pisicicida modulates RpoS’ roles promoting expression of stress adaptation genes and repressing virulence gene expression during the course of chronic infection . Here , we used a genome-wide loss-of-function Tn-seq screen to identify regulators controlling the expression of EsrB , a key activator of E . piscicida virulence . Unexpectedly , we discovered that RpoS inhibits esrB expression , and thus limits production of the pathogen’s T3SS/T6SS . Comparisons of the global transcription profiles of wt and ΔrpoS strains showed that RpoS controls expression , directly or indirectly , of more than 700 genes . Several stress stimuli modulate RpoS abundance and thus likely control esrB expression and E . piscicida virulence . Notably , in vitro transcription of esrB by the RpoD-core RNAP complex ( Eσ70 ) was blocked by RpoS . Furthermore , this inhibitory effect , likely mediated by Eσ38 , was abrogated by mutations in the esrB promoter discriminator or by a single amino acid substitution in RpoS R99 , a residue in the sigma 1 . 2 region ( the first part of RpoS conserved region 2 ) that molecular modeling predicted to be in close proximity to the -6G nucleotide of the esrB promoter discriminator . Collectively , these observations strongly suggest that direct interactions of Eσ38 with the esrB promoter impede transcription of this virulence regulator . In a turbot model , RpoS was required for robust E . piscicida growth during the first few days of infection , whereas EsrB was not; conversely , by 5 dpi RpoS becomes dispensable and EsrB becomes critical . By mediating expression of genes promoting stress responses [35] and inhibiting expression of esrB-controlled virulence genes , RpoS activity allows E . piscicida to co-ordinate expression of diverse cellular pathways ( Fig 8 ) . Thus , our findings suggest that the pathogen interprets variations in host-derived signals during the course of infection to modulate RpoS abundance/activity and thereby fine tunes its physiology for growth in different host environments . RpoS is a key global regulator in many Gram-negative bacteria [20–22 , 24] . In E . piscicida , RpoS was previously shown to be critical for the organism’s adaptation to several stressors , including starvation , high NaCl , H2O2 , as well as serum [35] . Here , using RNA-seq and ChIP-seq , we further refined our knowledge of RpoS control of gene expression in E . piscicida ( S3 Table , S4 Table and S5 Table ) . Expression of more than 500 genes was upregulated by RpoS while ~200 genes were down regulated by this alternative sigma factor . In general , in E . piscicida as in other organisms , RpoS promotes expression of genes activated in stationary phase and facilitates stress responses ( Fig 8 , lower panel; S3 Table ) [24] . For example , CsrA , an important RNA chaperone that functions in stationary-phase processes [46] , was activated by RpoS . Stress response related genes , including uspB and gadBC [47–48] , were also induced by RpoS . The succinate metabolic pathway ( sdhABCD ) , TCA cycle ( citCEFX and acnB ) and hemin iron uptake ( hemNPRS ) in E . piscicida ( Fig 8 , upper panel; S4 Table ) were among the genes most down-regulated by RpoS [22 , 24 , 35] . The succinate pathway is not only an important step in the tricarboxylic acid ( TCA ) cycle , but also serves as an electron donor coupled with the oxidative phosphorylation respiratory chain . The repression of the succinate pathway and TCA cycle by RpoS has also been observed in other bacteria , e . g . Escherichia coli O157:H7 [26] , Legionella pneumophila [49] and S . enterica [29] . Presumably , RpoS represses these metabolic genes via a similar mechanism as its repression of esrB; i . e . , the RpoS R99 residue directly binds to -6G nucleotide in the discriminators of the respective repressed promoters ( Fig 8 , upper panel ) . The mechanism of direct RpoS repression of gene expression uncovered in E . piscicida may be shared among several Gram-negative bacteria since RpoS repressed genes often contain -6G in their respective promoter discriminators , e . g . in S . enterica 4 known RpoS repressed genes contain -6G in their respective promoter discriminators ( S4B Fig ) [29] . RpoS is thought to mediate a trade-off between self-preservation and nutritional competence ( SPANC ) such as in S . enterica [49–50] . Our findings suggest that in E . piscicida , RpoS mediates a different trade-off between stress adaption and virulence; however , it is possible that expression of virulence-associated loci is equivalent to promoting rapid growth ( nutritional competence ) in certain host environments . The sigma factor subunit of RNAP holoenzyme enables this multicomponent enzyme to recognize specific promoters during the initiation of transcription [19–21] . Generally , RpoD ( σ70 ) mediates recognition of promoters carrying out the cell’s housekeeping function [22–25] , while alternative σ factors , like RpoS , mediate transcription of specific subsets of genes in different growth conditions [51–52] . Although sigma factors can enable transcription of repressors or sRNAs that down-regulate expression of target regulons , typically Eσ complexes are not thought to directly block transcription . However , σ factor competition for binding to core RNAP has been thought to explain how one sigma factor can inhibit transcription mediated by another σ factor [28 , 53–56] . An alternative means by which Eσ could impede transcription is by binding to and occluding promoter DNA , preventing initiation of transcription . Our findings are consistent with the latter mechanism: Eσ38 impairs esrB transcription through direct interactions with the esrB promoter , particularly with the -6G in the discriminator . Similar Eσ38-mediated repression at the level of the sdh promoter was also described in S . enterica serovar Typhimurium [29] . Notably , in this enteric pathogen , as in E . piscicida , RpoS interactions with the sdh promoter discriminator proved critical for repression; thus , when Levi-Meyreuis et al . mutated the GCC discriminator in Psdh to TAA , RpoS repression was abolished [29] . Our modeling-based mutagenesis of E . piscicida RpoS extends understanding of the manner in which this alternative sigma factor can block transcription . We show that a particular residue R99 in the sigma 1 . 2 region is essential for repression but not for Eσ38 to initiate transcription . Thus , the manner in which Eσ38 interacts with different discriminator sequences appears to determine the outcome of the interaction ( preventing or initiating transcription ) . Analyses of the data garnered from our RNA-seq and ChIP-seq experiments suggests that there are at least 6 RpoS-regulated promoters at which Eσ38 directly impairs transcription , suggesting that RpoS control of the cell’s transcriptional output is even more varied and subtle than previously thought . Many promoters can be simultaneously recognized by RpoD and RpoS , as these sigma factors share similar recognition motifs in their respective -35 and -10 elements [40] . As expected , the -10 element in PesrB was critical for esrB expression by RpoD or RpoSR99A ( Figs 4D and 6C ) . In vitro transcription studies also revealed that in the presence of RpoD , core RNAP , and the wt esrB promoter sequence , RpoS inhibited transcription in vitro; in the absence of RpoD , Eσ38 did not result in transcription from wt PesrB but it did from PesrB mut2 ( Fig 4D ) and RpoSR99A could drive transcription of PesrB and PesrB mut2 in absence of RpoD as well ( Fig 6C ) . Together , these observations suggest that RpoS interactions with the discriminator modifies the promoter in a manner that renders it resistant to Eσ70 binding/initiation . Additional studies , elucidating precisely how RpoS-promoter interactions prevent transcription are warranted . RpoS has been shown to regulate virulence in several pathogens [57] . In most cases , RpoS is required for virulence . For example , S . enterica serovar Typhimurium rpoS mutants are attenuated , likely because RpoS activates the expression of the plasmid-borne spvR and spvABCD genes , which are important for intracellular growth [58–59] . RpoS modulation of E . piscicida pathogenicity is complex and varies during the course of infection . For the first five days of infection , a ΔrpoS was attenuated , but after that time the mutant was as fit as the WT ( Fig 7D ) . The bases for the reduced fitness of the ΔrpoS mutant requires further definition; however , it is likely that the large set of > 500 genes whose expression is upregulated by RpoS , e . g . stress response genes that promote resistance to host defenses such as H2O2 , facilitate the pathogen’s growth . The early expression of genes ordinarily repressed by RpoS , such as esrB , could in principle also account for the attenuation of the ΔrpoS mutant . However , this does not seem to be the case , since the strain expressing RpoSR99A , which unlike RpoS , does not inhibit esrB expression , was not attenuated early in infection . The inverse kinetics of the requirements for rpoS and esrB during infection ( the esrB mutant became attenuated 5 dpi ) , suggests that the relief of RpoS repression of esrB , and production of T3/T6SS , becomes important only several days after the initiation of infection . Consistent with this idea , we found that ectopic expression of esrB ( from the lac promoter ) could overcome attenuation caused by over-expression of rpoS ( Fig 7C ) . Thus , the level and/or activity of RpoS decreases during the course of infection . Many studies have elucidated the complex cellular factors that govern RpoS levels/activity ( reviewed in [22] ) . Ultimately , environmental conditions , including nutrient availability and stressors , such as hydrogen peroxide , control RpoS activity . Therefore , the relief of RpoS-mediated repression in esrB by 5 dpi ( reflected in the requirement for EsrB at this point ) , strongly suggests that E . piscicida is beginning to occupy a distinct host niche at this point . It will be interesting to further define how the pathogen’s localization , both in terms of host organ and whether it is extra- or intracellular , changes over time . It is tempting to speculate that ~3–5 dpi , when esrB becomes critical for E . piscicida growth , may correspond to the time when the pathogen transitions from growing predominantly extracellularly , in the intestinal lumen and peritoneum , to predominantly intracellularly [7] . It will be fascinating to couple such localization studies with measurements of critical cellular regulators , such as ( p ) ppGpp , known to control RpoS levels/activity [22] , to decipher the molecular factors that govern the activity of this alternative sigma during infection . The bacterial strains used in this study are listed in S6 Table . E . piscicida were statically cultured in Luria-Bertani broth ( LB ) or Dulbecco’s modified essential medium ( DMEM ) at 28°C , while Escherichia coli strains were grown in shaking cultures in LB at 37°C . The E . coli DH5α λpir strain was used to propagate the pir-dependent suicide plasmids and E . coli SM10 λpir strain was used as the conjugation donor to introduce suicide plasmids into E . piscicida . E . coli BL21 ( DE3 ) was used to express recombinant proteins . When required , antibiotics were added at the following concentrations: carbenicillin ( Carb , 100 μg/ml ) , chloramphenicol ( Cm , 25 μg/ml ) , colistin ( Col , 16 . 7 μg/ml ) , kanamycin ( Kan , 100 μg/ml ) , tetracycline ( Tet , 12 . 5 μg/ml ) . The construction of in-frame deletion mutants was accomplished using sacB-based allelic exchange vectors as previously described [14] . Upstream and downstream fragments were amplified by PCR and then the Gibson assembly method [60] was used to ligate these fragments into the suicide vector pDMK which was linearized with XhoI . The vectors were initially propagated in E . coli DH5α λpir and after sequencing and purification , they were introduced into E . coli SM10 λpir and subsequently transferred into EIB202 by conjugation . The single crossover strains were selected on LB agar ( LBA ) medium containing Kan and Col , and the double crossover strains were selected on LBA containing 12% ( w/v ) L-sucrose as previously described [14] . Vectors for complementation , over-expression , and reporters were constructed with the E . piscicida compatible and stable plasmid pUTat as previously described [45 , 60] . The construction of reporter strains with promoterless luciferase ( luc ) , luxAB or Kan resistance gene ( kan ) respectively fused to the promoters of esrB ( PesrB ) , evpA ( PevpA ) , eseB ( PeseB ) , esrB ( PesrB ) , and rpoS ( PrpoS ) and inserted in chromosome were carried out using the same steps employed for generation of in-frame mutants . All the primers used to construct and validate the strains used here are listed in S7 Table . The PesrB-kan fusion was inserted into a neutral site ( between ETAE_3536-ETAE_3537 ) [8] in the WT strain . The transposon insertion sequencing was conducted as previously described [8 , 61–62] . A modified pSC189 [31] , which carried a gene for resistance to tetracycline was used to deliver the Himar transposon; the transposon library was stored in -80°C . Before screening , the library was resuspended in 5 ml of DMEM medium with ( Output ) or without ( Input ) addition of kanamycin , and cultured at 28°C for 24 h without shaking . Then the cultures of each group were plated on LBA medium and cultured at 37°C for 12 h . Finally , all of the colonies in each group were collected from the plates and restored in -80°C . The genomic extraction , library construction and sequencing were conducted as previously described [8] . The library was sequenced on the Hiseq 2500 platform ( Illumina , San Diego , CA ) by GENEWIZ ( Suzhou , China ) . Reads for each output library were normalized based on the input library and the reads per TA site were tallied and assigned to annotated genes or intergenic regions as described [31] . The fold change ( FC ) and Mann-Whitney U test ( MWU ) of each locus are based on comparison of the output and input libraries . The MIC assay was conducted as previously described [61] . ΔesrB::PesrB-kan was constructed as the negative control . A gradient of increasing concentrations of kanamycin were used and all the strains were statically cultured in 96-well plates at 28°C for 24 h . The bacterial growth was also monitored by measuring the optical density at 600 nm ( OD600 ) ( Biotek , Winooski , VT , USA ) . The fluorescence assays were conducted as previously described [60] . The reporter strains were cultured in 50 ml DMEM medium . Every 3 h , 200 μl were removed from each culture and added to 96-wells plates . The cell densities ( OD600 ) were detected with a microplate reader ( Biotek , USA ) and the fluorescence values were detected with a OrionII Microplate Luminometer ( Berthold , Bad Wildbad , Germany ) . Bacterial strains were inoculated into LB medium and subcultured in 50 ml DMEM at 28°C for 24 h . After pelleting the cells at 5 , 000 g for 10 min , protease inhibitors were added to the supernatants which were then filtered through the 0 . 22 μm low-protein-binding Millex filter ( Millipore ) and concentrated to 250 μl using a 10-kDa-cutoff Amicon Ultra-15 centrifugal filter device ( Millipore ) . SDS-PAGE was used to detect the ECPs profiles of EIB202 strains as previously described [13] . For the immunoblot analyses , bacterial cell pellets or concentrated ECP were suspended in PBS to normalize the culture densities based on OD600 measurements . 20 μl of each normalized sample was loaded onto 12% denaturing polyacrylamide gels . The proteins were resolved using electrophoresis and then transferred to PVDF membranes ( Millipore ) . The membranes were blocked in 10% skim milk powder solution , incubated with a 1:2000 dilution of mouse anti RpoS ( Santa cruz ) , Flag-tag ( Beyotime ) , Lon ( Sigma ) , EseB ( GenScript , Nanjing , China ) , RpoB ( OriGene ) or DnaK ( Huabio , Hangzhou , China ) antibodies , and finally incubated with a 1:2000 dilution of anti-mouse peroxidase-conjugated IgG secondary antibodies ( Sigma ) . The ECL reagent ( Thermo Fisher ) was used to visual the blots . Overnight cultures of WT and ΔrpoS were statically subcultured in DMEM at 28°C for 12 h , respectively . RNA samples were extracted with the RNA isolation kit ( Tiangen ) as previously described [13] and incubated with DNase I ( Promega ) for 30 min at 37°C to remove genomic DNA . RNA concentrations were measured with NanoDrop and 1 μg of each sample was used for reverse transcription with PrimeScript II 1st Strand cDNA Synthesis Kit ( TaKaRa ) . The qRT-PCR was conducted on the Applied Biosystems 7500 real-time system ( Applied Biosystems , Foster City , CA ) in triplicate . The comparative CT ( 2-ΔΔCT ) method was used to quantify the relative qualities of each transcript , and the housekeeping gyrB gene was used as an internal control . All the primers used are listed in S7 Table . For mRNA-specific RNA-seq , Ribo-Zero-rRNA ( Epicentre ) was used to remove the rRNA in the RNA samples following the manufacturer’s instructions . The final concentration of RNA samples was determined with the Qubit 2 . 0 Fluorometer ( Thermo Fisher ) . The VAHTS Stranded mRNA-seq Library Prep Kit for Illumina ( VATHS turbo ) was used in the construction of strand-specific RNA-seq libraries , and the sequencing was conducted on the Hiseq 2500 platform to yield 100-base-pair end-reads . Adapter sequences and low-quality bases ( PHRED quality scores ≤5 ) were trimmed by the Trimmomatic package using the default parameters , and truncated reads smaller than 35 bp were discarded . The RNA-seq data processing procedures and statistical analysis were the same as previously described [63] . Overnight cultures of ΔrpoS/flag-rpoS were subcultured in DMEM at 28°C for 24 h . Bacterial pellets were collected and washed using ddH2O , and stored at -80°C . After three cycles of treatment at 80°C for 1 h and ice incubation for 1 h , the pellets were resuspended with 3 ml BS/THES buffer ( THES Buffer: 50 mmol/L Tris HCl ( pH 7 . 5 ) , 10 mmol/L Sucrose ( m/v ) , 140 mmol/L NaCl , 0 . 7% Protease Inhibitor Cocktail ( v/v ) ; 5× BS Buffer: 50 mmol/L HEPES , 25 mmol/L CaCl2 , 250 mmol/L KCl , 60% Glycerol; BS/THES Buffer: 44 . 3% THES Buffer , 20% 5 × BS Buffer , 35 . 7% ddH2O ) . Bacteria were cracked by ultrasonication and supernatants were collected after centrifuge . Biotinylated DNA and NeutrAvidin Agarose Resin beads ( Thermo Fisher ) were mixed and incubated at 25°C for 1 h . Then the probe-labeled beads were washed with TE and BS/THES buffer for two times . The probe-labeled beads and supernatant lysates ( containing 200 ng/μl Poly ( dI:dC ) ) were mixed and incubated with slow shaking at 25°C for 30 min . After 4 washes with BS/THES buffer , proteins were eluted from the beads with a NaCl concentration gradient . Finally , the beads were eluted with ddH2O under 70°C to unbound the Biotin-DNA . The identification of pull-down samples were the same as previously described [64] . Full-length rpoS and rpoD open reading frames were amplified from EIB202 genomic DNA . The PCR products were subcloned into pET28a and transformed into E . coli BL21 ( DE3 ) . The resulting strains were grown with shaking in LB medium at 37°C until OD600 ~ 0 . 6 . Then isopropyl β-D-1-thiogalactopyranoside ( IPTG ) was added to a final concentration of 0 . 5 mM , and the cells were cultured at 16°C for another 16 h . The purification procedure was conducted as previously described [13] with the use of HEPES buffers ( 20 mM HEPES , 250 mM NaCl , x imidazole ) ; the final concentrations of imidazole of binding buffer , washing buffer and elution buffer were 20 mM , 50 mM and 300 mM , respectively . The purified proteins were dialyzed with HG buffer ( 20 mM HEPES , 250 mM NaCl , 5% glycerol ( w/v ) ) for 20 h to remove the imidazole . The purified proteins were stored at -80°C . The pUTat/flag-rpoS and pUTat/flag plasmids encoding RpoS-Flag and the Flag tag alone , respectively , were expressed in the ΔrpoS strain for ChIP assays as previously described [65] . Overnight cultures of each strain were diluted to the same cell density ( OD600 ) and statically subcultured in DMEM medium containing carbenicillin at 28°C for 24 h . Then Rifampicin ( Sigma ) was added at a final concentration of 150 μg/ml , and incubated at 28°C for 30 min . Formaldehyde was used for cross-linking the protein-DNA complexes in vivo and the cross-linking was stopped by addition of glycine solution . The ChIP assay was conducted as previously described [65] . The DNA was purified by phenol/chloroform and precipitated with ethanol . For ChIP validation , DNA fragments were PCR amplified with primer pair esrB-EF/ER ( S7 Table ) . For ChIP-seq , the DNA fragments were used for library construction with the VAHTS Turbo DNA library prep kit ( Vazyme , Nanjing , China ) , and the number of reads per microliter of each library was determined by qRT-PCR . The sequencing procedure was the same as that described for Tn-seq , and the MEME-suite website ( http://meme-suite . org ) was used to identify the RpoS binding motif . Healthy turbot fish ( average weight of ~30 g ) were chosen and acclimatized in the aeration tanks for two weeks with a continuous flow of seawater at 16°C . For in vivo fluorescence detection , the subcultures of reporter strains were diluted to 106 CFU/ml in PBS . The fish were anesthetized with tricaine methanesulfonate ( MS-222 ) ( Sigma-Aldrich ) at a concentration of 80 mg/l . Fish were intraperitoneally ( i . p . ) injected with 100 μl of bacterial suspensions . At 8 days post injection ( dpi ) , the fish were i . p . injected with beetle luciferin substrate ( Promega ) . After 10 min , the fluorescence was detected with a Kodak In-Vivo Multispectral System FX ( Carestream Health ) . Then the livers of the fish were sampled and bacterial colonization was measured by CFU plating . For competition assays , inocula were prepared using fresh cultures of bacteria that were diluted and mixed at 1:1 ratio . The injection dose was 105 CFU/fish . At 8 dpi , the livers from fish in each group ( 5 animals/group ) were removed , homogenized and plated on DHL plates with or without containing chloramphenicol ( Cm ) to enumerate the ratio of the competing strains . The ratios of the bacterial counts were used to determine the CIs . The RNA-seq sequencing data was deposit at SRA ( SRP136988 ) . All animal protocols used in this study were approved by the Animal Care Committee of the East China University of Science and Technology ( 2006272 ) . The Experimental Animal Care and Use Guidelines from Ministry of Science and Technology of China ( MOST-2011-02 ) were strictly followed . All the data and related materials are our original research , and have not been previously published and have not been submitted for publication elsewhere while under consideration .	Edwardsiella piscicida , a major fish pathogen , relies on T3/T6SSs for virulence and the EsrB transcription activator promotes the expression of these secretion systems and many other genes that enable growth in fish . Here , we found that the alternative sigma factor RpoS inhibits expression of esrB thereby diminishing expression of virulence-associated genes . Transcriptome profiling revealed that , as in many other organisms , RpoS enables expression of hundreds of genes , many of which are linked to stress responses , suggesting that RpoS may mediate a trade-off between stress adaptation and virulence . Consistent with this idea , we found that an rpoS mutant was attenuated early , but not late in infection of turbot , whereas an esrB mutant was attenuated late and not early in infection . Molecular analyses demonstrated that RpoS inhibition of esrB expression involves a direct interaction between RpoS and the esrB promoter; in particular , interactions between RpoS residue R99 and the -6G nucleotide in the esrB promoter discriminator appear to be critical for repression of esrB expression . These findings provide new insight into how a sigma factor can impede transcription and demonstrate the temporal dynamics of the requirement for a sigma factor during the course of infection .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "medicine", "and", "health", "sciences", "pathology", "and", "laboratory", "medicine", "gene", "regulation", "pathogens", "microbiology", "dna", "transcription", "bacterial", "diseases", "regulator", "genes", "enterobacteriaceae", "genetic", "elements", "gene", "types", ...	2018	Critical role for a promoter discriminator in RpoS control of virulence in Edwardsiella piscicida
Lowe Syndrome is a developmental disorder characterized by eye , kidney , and neurological pathologies , and is caused by mutations in the phosphatidylinositol-5-phosphatase OCRL . OCRL plays diverse roles in endocytic and endolysosomal trafficking , cytokinesis , and ciliogenesis , but it is unclear which of these cellular functions underlie specific patient symptoms . Here , we show that mutation of Drosophila OCRL causes cell-autonomous activation of hemocytes , which are macrophage-like cells of the innate immune system . Among many cell biological defects that we identified in docrl mutant hemocytes , we pinpointed the cause of innate immune cell activation to reduced Rab11-dependent recycling traffic and concomitantly increased Rab7-dependent late endosome traffic . Loss of docrl amplifies multiple immune-relevant signals , including Toll , Jun kinase , and STAT , and leads to Rab11-sensitive mis-sorting and excessive secretion of the Toll ligand Spåtzle . Thus , docrl regulation of endosomal traffic maintains hemocytes in a poised , but quiescent state , suggesting mechanisms by which endosomal misregulation of signaling may contribute to symptoms of Lowe syndrome . Lowe syndrome is an X-linked disorder caused by mutations in the phosphoinositide-5-phosphatase OCRL ( Oculocerebrorenal Syndrome of Lowe ) . Lowe Syndrome patients display renal proximal tubule dysfunction , glaucoma , cataracts , and neurological phenotypes such as cognitive and behavioral impairments , hypotonia , and epilepsy [1 , 2] . OCRL encodes a 901 amino acid protein with an N-terminal Pleckstrin Homology ( PH ) domain bearing clathrin-binding motifs , a central phosphoinositide-5-phosphatase domain ( with preference for PI ( 4 , 5 ) P2 and PI ( 3 , 4 , 5 ) P3 ) , as well as an ASPM-SPD2-hydin ( ASH ) domain and a catalytically inactive Rho GTPase activating ( RhoGAP ) domain that each mediate interactions with membrane-associated proteins such as Rab GTPases , IPIP27A/B , and APPL [3] . OCRL localizes to multiple membrane compartments and is involved in a range of cell biological processes , including clathrin-mediated endocytosis [4–6] , intracellular trafficking [7–10] , actin cytoskeleton regulation [6 , 11 , 12] , ciliogenesis [13] , and cytokinesis [11 , 14] . However , it remains unclear precisely how these diverse cellular requirements contribute to tissue and organ level pathology in Lowe Syndrome patients . A redundant gene , INPP5B , may partially compensate for loss of OCRL , complicating studies in vertebrate systems [13 , 15 , 16] . By contrast , Drosophila expresses only a single homolog of OCRL , CG3573/dOCRL [14] , and may therefore be a useful model for understanding the functions of OCRL in complex tissues in vivo . dOCRL is required for cytokinesis in cultured S2 cells [14] , but its functions have not yet been examined in vivo . Membrane traffic plays critical roles in regulating signal transduction in many developmental contexts . Signaling cargoes , including both ligands and receptors , are rapidly trafficked through the endocytic system , changing their signaling activities en route [17] . Therefore , mis-regulation of membrane trafficking pathways can lead to drastic alterations in signal output . In Drosophila , the innate immune system is poised to respond quickly and effectively to infection . Mutants in a variety of components of the endosomal trafficking system exhibit various features of immune activation , including increased hemocyte abundance [18–22] , but it has remained unclear which specific immune tissues or pathways are altered or how this leads to hemocyte activation . Here we show that dOCRL controls endosomal traffic in Drosophila larval hemocytes to autonomously restrict immune cell activation . To investigate the role of dOCRL in vivo , we generated null alleles by excision of a P element from the viable , fertile line docrlEY15890 ( S1A Fig ) and isolated two null alleles , docrlΔ3 and docrlΔ4 , which lacked the dOCRL protein product and were larval or pupal lethal when homozygous ( S1B and S1C Fig ) . Lethality was specific to docrl , as it was not complemented by a deficiency removing the docrl locus , and was rescued by a docrl-containing genomic fragment ( S1C Fig ) . Upon dissecting docrl mutant larvae , we observed a striking ( 5–10 fold ) increase in the numbers of circulating hemocytes ( Fig 1A and 1B ) , which are macrophage-like cells that mediate innate immune responses [23] . Notably , docrlΔ3 larvae still accumulated excess hemocytes and died as larvae and pupae when raised under axenic conditions ( S1C Fig ) , suggesting that immune cell activation was not due to over-sensitivity to pathogens . docrl mutants exhibited few actively dividing cells ( marked by phosphorylated histone H3 , S2A Fig ) , suggesting that excess hemocytes do not arise from increased cell division , but instead may reflect reduced hemocyte turnover [24] . Though we detected cytokinetic defects in docrl mutant hemocytes ( S2B and S2C Fig ) , as previously described in cultured cells [11 , 14] , this appeared to be insufficient to counteract the excess hemocyte phenotype . Finally , both a docrl-containing genomic fragment and GAL4-UAS-driven dOCRL-GFP ( expressed with the hemocyte driver He-GAL4 ) restored hemocyte numbers to control levels , indicating that this phenotype is specific to loss of docrl ( Fig 1A and 1B ) . We then examined the localization of hemocytes in control and docrl mutant larvae . Hemocytes are found in circulation as well as along the lateral midline and posterior end of the larva , in sessile or resident pockets that are sites of self-renewal [23 , 25 , 26] . Excess hemocytes in docrl mutants were found broadly in circulation as well as in expanded hematopoietic pockets , suggesting that excess hemocytes do not arise simply from mobilization of the sessile pool ( Fig 1C ) . We also observed the frequent presence of large melanotic masses in the posterior larval body cavity ( Fig 1D–1F ) . Such masses often occur in mutants with excess hemocytes , due to encapsulation of self-tissue in the absence of infection [27] . To test if this was the case in docrl mutants , we visualized genetically marked hemocytes ( He-GAL4 driving UAS-GFP . nls ) directly through the cuticles of live larvae . Melanotic masses were indeed surrounded by GFP-positive blood cells ( Fig 1F ) . However , excess hemocytes were observed in docrl mutant larva with or without melanotic masses , suggesting that the underlying phenotype in docrl mutants is hemocyte over-abundance . We next asked which of several innate immune-implicated tissues require dOCRL to limit hemocyte number: hemocytes themselves , the lymph gland ( the site of hemocyte precursor maturation [28] ) , the fat body ( which mediates the majority of antimicrobial peptide expression [29] ) , nephrocytes ( which mediate clearance of immune-suppressing serpins [30] ) , and muscle ( in which STAT signaling contributes by unknown mechanisms to hemocyte activation [31] ) . First , we tested if tissue-specific RNAi-mediated reduction of dOCRL in these tissues could recapitulate the immune cell phenotype . We were only able to deplete dOCRL by ~40% in hemocytes using available UAS-RNAi lines ( S3A Fig ) , and ubiquitous RNAi using these lines and an actin-GAL4 driver did not recapitulate the lethality of docrl mutants . Further , this level of depletion using drivers specific to hemocytes or other tissues did not cause increased hemocyte activation , suggesting that more complete depletion of docrl is required to cause a phenotype ( S3B Fig ) . We therefore tested if re-expression of dOCRL in specific tissues in the docrl mutant could rescue hemocyte number . Hemocyte number was rescued cell-autonomously by driving dOCRL-GFP ( but not control mCD8-GFP ) in hemocytes with He-Gal4 ( Figs 1B and 2A ) , and in muscle with mef2-GAL4 , indicating that dOCRL in either tissue is sufficient to restrain hemocyte activation . By contrast , expression of dOCRL-GFP in the fat body ( with the strong driver Lsp2-GAL4 ) did not restore hemocyte numbers to wild type levels ( Fig 2A ) . Further , expression of dOCRL-GFP by the driver Dot-GAL4 ( which expresses at high levels in salivary glands , lymph gland , and nephrocytes and only at low levels in hemocytes [32] ) , HH-LT-GAL4 ( which in our hands expresses dOCRL-GFP sparsely in hemocytes ( see below ) [33] ) , Ser-GAL4 ( which expresses in the lymph gland posterior signaling center [34] ) and dome-GAL4 ( which expresses in the lymph gland medullary zone [35] ) , did not significantly rescue hemocyte abundance ( Fig 2A ) . To further test if hemocyte differentiation was affected at the level of progenitors in the lymph gland , we compared lymph glands in control and docrl mutants . Consistent with general immune activation ( but unlike mutants with hyperplasia or hematopoietic defects [34] ) , primary lobes in docrl mutant lymph gland appeared partially disintegrated , while secondary lobes were apparent . This phenotype was rescued by expression of dOCRL-GFP in differentiated hemocytes ( using He-GAL4 ) , suggesting that lymph gland structural defects are secondary to hemocyte-mediated immune activation . Finally , we did not observe noticeable differences in cell proliferation ( phospho-histone H3-positive cells ) or hemocyte differentiation ( P1-positive cells ) in docrl mutant lymph glands ( Fig 2B ) . In summary , our rescue experiments indicate that docrl expression in either hemocytes or muscle is sufficient to rescue hemocyte number to normal levels , while it does not appear to be required in fat body , nephrocytes , or hematopoietic precursors . In light of the profound hemocyte specific rescue and to identify direct cellular functions for dOCRL in hemocytes , we focused our subsequent analyses on hemocytes . Larval hemocyte types include plasmatocytes , which are small macrophage-like cells; crystal cells , which control melanization of foreign bodies; and lamellocytes , which are large , banana-shaped cells involved in encapsulation of foreign bodies . The majority of circulating hemocytes are plasmatocytes , while lamellocytes are rare in unstimulated larvae [36] . However , when we examined hemocyte composition by immunostaining with the antibodies P1 ( which labels plasmatocytes ) and L1 ( which labels lamellocytes ) , we found that 21% of hemocytes in docrl mutant larvae were L1-positive ( Fig 2C and 2D ) . Surprisingly , 62% of these L1-positive cells appeared morphologically similar to plasmatocytes ( Fig 2E ) , suggesting these cells may represent plasmatocytes or pro-hemocytes in the process of differentiating into lamellocytes [37] . Aberrant hemocyte differentiation was fully rescued by a single chromosomal copy of dOCRL , but only partially rescued by expression of dOCRL-GFP with the He-GAL4 driver ( Fig 2D ) , suggesting that lamellocyte differentiation was not completely cell autonomous to He-GAL4-expressing cells ( and may perhaps arise from functions of docrl in muscle [31] and Fig 2A ) , or that it may arise from the 20% of hemocytes that do not express this GAL4 driver [38] . Together , these data indicate that hemocytes in docrl mutants exhibit cell autonomous hyper-activation and partially cell autonomous hyper-differentiation , in addition to greatly increased abundance . To investigate the role of docrl in hemocyte physiology , we examined dOCRL localization in hemocytes by live imaging of endogenously tagged dOCRL ( dOCRL-TagRFPT [39] ) . Flies expressing dOCRL-TagRFPT from the docrl locus , as their only source of dOCRL , are viable and fertile and exhibit no obvious immune phenotype [39] . dOCRL-TagRFPT localized to discrete puncta at the hemocyte plasma membrane and throughout the cytoplasm ( Fig 3A ) . These puncta colocalized most strongly with He-Gal4-driven GFP tagged clathrin light chain ( Clc ) , and moderately colocalized with other compartment markers ( Fig 3B , S4A Fig ) , including endogenously tagged YFP-Rab5 ( early endosomes ) , YFP-Rab7 ( late endosomes ) , YFP-Rab11 ( recycling endosomes ) , and He-GAL4-driven Vps35-GFP ( a component of the endosomal cargo-sorting retromer complex , which has itself previously been implicated in restricting innate immune activation [18 , 22] ) . Interestingly , dOCRL exhibited a qualitatively different pattern of association with different compartments . While dOCRL localized strongly with Clc and diffusely with Rab5 early endosomes and Rab11 recycling endosomes , it exhibited a complementary pattern to Vps35 on endosomes , and accumulated in foci on Rab7 late endosomes ( Fig 3B , S4A Fig ) . To test the functional requirement for this broad distribution of dOCRL , we examined the localization of the primary dOCRL substrate PIP2 in control and mutant hemocytes by live imaging of an mCherry-tagged PH domain of PLCδ , which specifically binds PI ( 4 , 5 ) P2 [40] . In control hemocytes , PH-PLCδ localized at low levels to the plasma membrane and in discrete intracellular puncta ( Fig 3C , S4B Fig ) . By contrast , in docrlΔ3 mutant hemocytes , PH-PLCδ accumulated at higher levels both at the plasma membrane and in intracellular puncta . At the level of the whole cell , we observed a significant increase of mean PH-PLCδ intensity relative to control , perhaps due to stabilization of the reporter in the presence of excess PIP2 ( Fig 3C ) . To better analyze whether aberrant PIP2 associated with a specific endosomal compartment , we compared the accumulation of PH-PLCδ to fluorescently labeled endosomal markers . Levels of PH-PLCδ increased in all compartments examined , similar to the effect seen in whole hemocytes ( Fig 3C and 3D , S4B Fig ) . Together , these data suggest that dOCRL is required to maintain PIP2 homeostasis in diverse endosomal compartments . To test how the dOCRL phosphatase activity contributes to its role in hemocytes , we performed rescue experiments in the docrl mutant using He-Gal4-mediated expression of wild-type dOCRL , a phosphatase-inactive dOCRL ( dOCRLH469R , corresponding to a Lowe Syndrome mutation [41] ) , or the dOCRL phosphatase domain alone [42] . In contrast to full-length dOCRL , which rescued hemocyte abundance ( Figs 1B , 2A and 3E ) , hemocyte-specific expression of either phosphatase-dead dOCRL or the phosphatase domain alone did not suppress excess hemocyte numbers ( Fig 3E ) , though , based on GFP and mCherry fluorescence , they were expressed as well as the wild-type transgene and much more highly than fully rescuing , endogenously tagged dOCRL . These results indicate that phosphoinositide homeostasis is required for innate immune cell quiescence and that the phosphatase activity of dOCRL , together with contributions from other dOCRL domains , is critically involved in this process . PIP2 regulation by dOCRL plays major roles in regulating cellular F-actin assembly [3] . We compared F-actin in control and docrl mutant hemocytes using phalloidin staining , and measured a large increase in F-actin intensity in docrl mutants relative to controls ( Fig 4A and 4B ) . Increased F-actin assembly corresponded to a significantly spikier morphology in docrl mutant hemocytes ( Fig 4C ) . Actin filament polymerization is a major feature of hemocyte activation [18 , 43] , so we considered the possibility that this phenotype could be an indirect result of immune activation in docrl mutants . However , when compared to hemocytes extracted from larvae infected with Micrococcus luteus , docrl mutant hemocytes exhibited much stronger F-actin intensity ( Fig 4D ) , suggesting that this phenotype is only partly due to immune activation and that cell-intrinsic functions of dOCRL in regulating actin polymerization also play a role [14] . These cytoskeletal phenotypes were rescued cell autonomously by expressing dOCRL-GFP with the He-GAL4 or HH-LT hemocyte drivers ( Fig 4B , 4C and 4E ) . Using HH-LT-GAL4 , we noted that only half of circulating hemocytes expressed the rescue construct ( 48% , vs . 80–90% with He-GAL4 ) . We took advantage of this to measure F-actin in rescued compared to non-rescued hemocytes in the same larvae . F-actin levels in rescued hemocytes were indeed lower , despite the persistence of excess hemocytes ( Fig 2A ) , supporting the model that this is a cell-intrinsic phenotype . Interestingly , expression of the phosphatase domain alone partially rescued F-actin levels ( Fig 4F ) , though it was not sufficient to suppress hemocyte number ( Fig 3E ) , and actin still accumulated at plasma membrane ruffles and in intracellular puncta . Thus , we conclude that hemocyte F-actin accumulation is at least partly due directly to loss of docrl in hemocytes , is not merely an indirect result of immune activation , and depends on domains of dOCRL outside its phosphatase domain . To determine the consequences of loss of docrl on endosomal trafficking , we first assayed the abundance and morphology of endosomal compartments . Consistent with the broad increases in PIP2 that we observed in docrl mutant hemocytes , we found defects in compartment abundance , morphology , or behavior throughout the endosomal system ( Fig 5A–5D , S5A Fig ) . In control hemocytes , Clc , Rab5 , and Vps35 compartments formed well-distributed medium-sized puncta , while in docrl mutant cells these compartments fragmented into small punctae at the cell periphery , and also densely accumulated in the perinuclear region ( Fig 5A–5C ) . We also observed loss of larger Vps35-positive tubulo-vesicular structures . Further , Rab5 endosomes were less motile in docrlΔ3 hemocytes than in controls ( S5B Fig ) . We did not detect a marked change in Rab35 compartment structure or distribution in docrl mutants ( S5A Fig ) . The population of Rab7-positive endosomes was strikingly increased in docrl mutant hemocytes compared to controls ( Fig 5C ) , and the level of endogenously tagged Rab7 was significantly elevated ( Fig 5D ) . Interestingly , injury of wild-type larvae with a sterile or M . luteus-coated needle ( both of which activate cellular immune responses [44] ) resulted in enlargement of Rab7-positive endosomes , suggesting that similar to docrl mutants , normal immune responses may elicit changes in endosomal trafficking ( Fig 5E ) . To assess whether these defects in membrane compartment structure correlated with changes in function , we first examined scavenger receptor-mediated endocytosis and phagocytosis , both of which are highly dependent on membrane trafficking and are critical to the functional capacity of hemocytes [45–48] . We tested scavenger receptor-mediated endocytosis by measuring internalized maleylated BSA [49] at confocal slices through the cell interior . At all time points after pulse-chase with mBSA , docrl mutant hemocytes exhibited reduced internalization ( Fig 6A ) . We next tested phagocytosis by measuring the internalization of E . coli . docrl hemocytes readily internalized Alexa-488-labeled E . coli , and the total number of particles/cell was not significantly different than controls ( Fig 6B ) . Though we did detect a significant decrease in the number of E coli particles that were fully internalized in docrl mutants , this small difference is unlikely to account for the overall hemocyte activation phenotype . These results suggest that dOCRL is not critically required for phagocytosis in hemocytes , though they do not exclude a role in kinetics of uptake or phagosome maturation . Next , we tested endolysosomal maturation by labeling control and docrl mutant hemocytes with Lysotracker , which fluoresces in acidified subcellular compartments . docrl mutant hemocytes displayed increased mean Lysotracker intensity and many cells were densely packed with Lysotracker-positive compartments ( Fig 6C ) , compared to control or endogenously tagged dOCRLtRFP-expressing cells , in which Lysotracker typically labeled only a few discrete compartments . The docrl mutant defect was rescued by he-GAL4-driven expression of dOCRL-GFP . Further , partial knockdown of docrl by He-GAL4-driven RNAi caused a significant increase in mean Lysotracker intensity that scaled with driver expression levels ( S6 Fig ) . Together , these results indicate that the endosomal maturation phenotype is cell autonomous to hemocytes . Finally , we tested whether docrl mutant hemocytes exhibited defects in autolysosomal degradation . We employed a dually tagged GFP-mCherry-Atg8a , which allows detection of both nonacidic autophagosomes ( coincident mCherry and GFP ) as well as acidic autolyososomes ( mCherry alone , due to quenching of GFP ) [50] . docrl mutant hemocytes exhibited an increase in mCherry-GFP co-localization , as well as a marked increase in both GFP and mCherry fluorescence . These results are consistent with a failure to fuse autophagosomes with lysosomes , upregulation of autophagy , and/or a failure of lysosomal degradation following lysosome fusion ( Fig 6D ) . Thus , docrl is strongly required for proper regulation of autophagosome-lysosome flux in hemocytes . We next explored which of these dysfunctional endolysosomal compartments could account for docrl-induced hemocyte activation , by individually disrupting them in otherwise wild type animals . We first inhibited the internalization step of endocytosis , by expressing dominant negative versions of clathrin heavy chain and dynamin , as well as a temperature sensitive dynamin mutant ( raised at the restrictive temperature of 29°C , at which uptake of mBSA was blocked ( S7A and S7B Fig ) ) . None of these manipulations caused melanotic masses or excess hemocyte abundance ( Fig 7A ) . OCRL interacts with Rab35 during both early endocytosis and the abscission step of cytokinesis [4 , 11] . However , we did not detect changes in hemocyte number following expression of a dominant negative Rab35 construct ( Fig 7A ) . Finally , we found that hemocyte-specific expression of the dOCRL phosphatase domain , which does not rescue hemocyte number ( Fig 3E ) , robustly rescued mBSA uptake in docrl mutants ( Fig 7B ) . Taken together , these data indicate that defects in the internalization step of endocytosis are unlikely to account for hemocyte activation in docrl mutants . We next tested the role of endosome-lysosome or autophagosome-lysosome fusion in hemocyte activation by depleting the SNARE protein Syntaxin-17 ( Syx17 ) , which blocks autophagosome-lysosome fusion [51] . Syx-17 RNAi did not affect hemocyte number ( Fig 7A ) , suggesting that defects in this pathway are not sufficient to drive immune cell activation in docrl mutants . Similarly , previous work has shown that mutants affecting canonical autophagy do not cause innate immune defects [19] , suggesting that changes in autophagy that we observe in docrl mutants ( Fig 6D ) are unlikely to account for hemocyte phenotypes . We next tested the role of post-endocytic endosomal sorting in hemocyte activation . It has previously been shown that hemocyte-specific depletion of Rab5 and Rab11 increases circulating hemocyte concentration [20] , suggesting again that innate immune cell activation may arise from defects in sorting at these endosomes . To confirm these results and extend the analysis to include additional Rab proteins , we expressed constitutively active or dominant negative Rab5 , 7 , and 11 constructs in hemocytes with he-GAL4 . These mutants lock the GTPase in its GTP ( constitutively active ( CA ) ) or GDP-bound ( dominant negative ( DN ) ) conformation , leading to inappropriate regulation of Rab effectors . Rab5DN ( and to a milder degree Rab11DN ) each led to an increase in circulating hemocyte number ( Fig 7A ) . We further hypothesized that Rab7CA would promote a similar increase in hemocyte number , consistent with the expanded Rab7 compartment in docrl mutants . Indeed , hemocyte specific expression of Rab7CA also led to an increase in hemocyte number ( Fig 7A ) . Thus , manipulating endosomal traffic mediated by Rab5 , Rab7 , and Rab11 in hemocytes recapitulates the docrl mutant phenotype of increased hemocyte number . We next tested additional components of post-endocytic endosomal sorting . Retromer mediates endosomal sorting , and is composed of a cargo-selective trimer of Vps35 , Vps26 , and Vps29 . Vps35 mutants have previously been shown to exhibit immune activation [18 , 22] , and we found that mutations in Vps26 similarly caused melanotic mass formation ( S7C Fig ) . Vps35 and Vps26 associate with endosomes via distinct membrane binding SNX1 containing “SNX-BAR”‘ or SNX3 containing complexes , likely with distinct localization and functions [52] . To clarify the requirement for retromer in hemocyte number , we analyzed snx1 and snx3 mutants . We detected frequent melanotic masses in snx1 , but not snx3 larvae ( S7C Fig ) , suggesting that specific loss of SNX-BAR retromer underlies hemocyte activation in vps35 and vps26 retromer mutants , and supporting the model that endosomal sorting underlies innate immune cell defects . Finally , we asked if hemocyte-specific manipulations of Rab5 , Rab7 , and Rab11 could enhance or rescue the docrl mutant phenotype . Expression of dominant negative Rab5 and Rab11 caused early larval lethality in docrl mutants , but not in controls , suggesting that docrl mutants are sensitized to defects in a Rab5-Rab11 trafficking route . Strikingly , overexpressing a constitutively active transgene ( Rab11CA ) significantly reduced hemocyte number toward normal levels ( Fig 7C ) . However , Rab5CA failed to rescue hemocyte abundance , suggesting that docrl-mediated trafficking defects may occur downstream of Rab5 , but upstream of Rab11 . Finally , Rab7DN moderately suppressed the docrl-dependent increase in hemocytes ( Fig 7C ) . We then examined whether additional docrl mutant phenotypes were suppressed by Rab manipulations . Expression of Rab7DN suppressed both the increase in F-actin levels and the expansion of the Lysotracker-positive compartment , while Rab11CA did not suppress either of these phenotypes ( Fig 7D ) . These results suggest that Rab7DN and Rab11CA may suppress excess hemocyte number in docrl mutants by distinct mechanisms . Neither Rab11CA nor Rab7DN rescue restored docrl mutant adult viability , suggesting that lethality may require more complete rescue , or that it arises from a distinct function of dOCRL . Taken together , these results suggest that the membrane trafficking defect most salient to docrl hemocyte phenotypes is mis-direction of endosomal traffic from a Rab11-dependent route towards a Rab7-dependent route . We explored which innate immune regulatory pathways might be disrupted by endosomal mis-sorting in docrl mutants . Hemocytes mediate immunity in Drosophila via several signaling pathways , including Jun-kinase ( JNK ) signaling [43] , JAK/STAT signaling between hemocytes and muscles [31] , and Toll signaling between fat body and hemocytes [53 , 54] . We examined each of these pathways , and found evidence that each was upregulated . Using the STAT reporter 10XSTAT-GFP , we detected strongly increased GFP signal in muscles in whole larvae ( Fig 8A ) , though not in isolated hemocytes ( Fig 8B ) . In addition , we detected increased activated ( phosphorylated ) JNK in isolated hemocytes ( Fig 8B ) . We did not detect a significant increase in systemic Toll activity ( when compared to the stronger induction by injury ) as measured by qPCR for the antimicrobial peptide drosomycin , which is primarily expressed in fat body [29] ( Fig 8C ) . However , we did observe significant changes in numerous Toll signaling components specifically in hemocytes , including increased accumulation of the Toll transcription factor Dorsal ( S8A Fig ) , increased accumulation of the Tl receptor ( S8B Fig ) , and redistribution of the adaptor MyD88 from the plasma membrane to the cytoplasm ( S8C Fig ) . Further , we detected increased intracellular ( Fig 8D ) and hemolymph ( Fig 8E , S8D Fig ) accumulation of the Toll ligand Spz . Interestingly , hemocyte-specific expression of Rab11DN , Rab7CA , and Syx17RNAi each led to similarly increased intracellular accumulation of Spz ( S8E Fig ) . To test whether manipulation of Tl signaling could rescue the docrl phenotype , we performed genetic epistasis experiments using alleles of Tl , Spz , and the Spz-processing enzyme ( SPE ) . However , none of these manipulations suppressed the docrl-mediated increase in hemocyte abundance ( S8F Fig ) , suggesting either that these alleles do not behave as nulls due to residual or maternal activity , or that multiple other pathways ( including JNK and STAT ) , are sufficient to maintain immune activation in the docrl mutant . We then asked whether Rab-mediated rescue of the docrl phenotype might occur by altered traffic of immune signals . Because Rab11CA autonomously suppressed hemocyte activation ( Fig 7C ) , we reasoned that this rescue should also suppress the relevant docrl-dependent defects in trafficking of signaling molecules . Though changes in Tl signaling are unlikely to fully account for the hemocyte activation observed in docrl mutants ( S8F Fig ) , the correspondence between Rab and dOCRL phenotypes on Spz localization and the availability of suitable reagents prompted us to use Spz as a model cytokine to test the mechanisms of Rab11CA rescue . Indeed , we found that Rab11CA caused Spz-GFP to accumulate in hemocytes , in enlarged intracellular compartments ( Fig 8F and 8G ) . Importantly , this increase in intracellular Spz-GFP corresponded to a concomitant decrease in secreted Spz-GFP , as assayed by Western blot of cell-free hemolymph ( Fig 8H ) . Together , these data suggest the intriguing possibility that aberrant cytokine release is an important determinant of the docrl mutant phenotype , and that a Rab11-dependent sorting route suppresses hemocyte abundance by retaining Spz and/or other relevant cytokines in the cell . OCRL has been implicated in numerous membrane trafficking events , including multiple steps of endocytosis and endolysosomal traffic , autophagy , ciliogenesis , and cytokinesis [3] . It has remained an open question whether individual cell and tissue level pathologies in Lowe Syndrome emerge from defects at specific cellular locations , or from a combination of OCRL-dependent functions . Though we found that docrl mutant hemocytes exhibit many of these cellular defects , our data strongly suggest that endosomal membrane trafficking is the primary role of dOCRL in regulating innate immune cell activation . First , cytokinesis defects are unlikely to account for innate immune cell activation , since relatively few docrl mutant hemocytes fail cytokinesis ( S2B Fig ) , and this would be predicted to reduce rather than increase total hemocyte number . Second , defects in the internalization step of endocytosis are unlikely to cause innate immune cell activation , since independently inhibiting this process via clathrin and dynamin did not increase hemocyte abundance ( Fig 5A ) , and rescuing defective internalization in docrl mutants using the dOCRL phosphatase domain ( Fig 7B ) did not suppress hemocyte abundance ( Fig 3E ) . Third , autophagy defects are unlikely to account for innate immune cell activation , as Syx17 RNAi and canonical atg loss-of-function mutants [19] do not cause an immune cell phenotype . We expanded upon previous findings [20] , and found that recapitulating the trafficking defects of docrl mutants by manipulating retromer or endosomal Rab GTPases was sufficient to produce immune cell dysfunction . Strikingly , the converse manipulation of Rab11 and Rab7 GTPases could rescue the docrl-induced hemocyte phenotype . Expression of Rab7DN suppresses multiple docrl phenotypes , including hemocyte number , expansion of Lysotracker-positive compartments , and increased F-actin , suggesting that it may act by directly suppressing the consequences of loss of docrl . By contrast , Rab11CA suppressed hemocyte number without rescuing Lysotracker or F-actin defects , suggesting that it may bypass the docrl phenotype by a different downstream mechanism . Together , our results suggest that the role of dOCRL in maintaining immune cell quiescence is most critical in post-endocytic endolysosomal traffic , and adds to a growing list of OCRL functions in these compartments [10 , 55 , 56] . Innate immune activation in Drosophila depends on cross talk among multiple tissues , including hemocytes [53] , the fat body [53 , 57] , muscle [31] , and epithelia such as the gut and epidermis [58] . Prior studies have identified hemocyte-autonomous roles for endosomal GTPases in the control of hemocyte number [20] . It remains to be determined whether immune activation in other membrane trafficking mutants , including retromer [18 , 22] and Atg6 [19] , reflect defects in hemocytes or other immune relevant tissues . Our results suggest that excess hemocytes in docrl mutants arise from both hemocyte-autonomous functions of dOCRL , as well as non-hemocyte autonomous functions in muscle ( Fig 2A ) , as re-expression in either tissue is sufficient to rescue hemocyte abundance . In hemocytes , increased F-actin assembly likely reflects both direct effects of dOCRL , as suggested by the sparse rescue with HHLT ( Fig 4E ) , and also secondary effects due to other tissues , the overabundance of hemocytes , or inflammatory signaling pathways such as muscle STAT ( Fig 8A ) . Finally , lamellocyte differentiation may reflect an indirect , non-hemocyte-autonomous role of dOCRL , as hemocyte-specific rescue only partly restores normal lamellocyte differentiation ( Fig 2D ) . This is consistent with the finding that STAT signaling is upregulated in muscle ( Fig 8A ) , and muscle STAT signaling has previously been shown to promote lamellocyte differentiation [31] . Our results suggest that multiple signaling pathways , including hemocyte-specific JNK ( Fig 8B ) and Toll ( S8A Fig ) , and muscle STAT ( Fig 8A ) , likely cooperate to drive excess hemocyte accumulation , differentiation , and activation . It remains poorly understood how membrane traffic regulates these immune signaling pathways , and whether and how membrane traffic might contribute to physiological immune responses . Toll-like receptors are known to be regulated at specific cellular compartments [59] , and work in mammalian cells suggests that endosomal sorting may regulate the localization of cytokine receptors to prevent spurious Toll-like receptor activation [60] . Our studies indicate that investigating endosomal maturation and recycling pathways will provide interesting new insights into innate immune signaling . Though it remains to be unraveled precisely how loss of docrl contributes to activation of each of these signaling pathways , our finding that rescue of hemocyte number by Rab11CA correlates with retention of Spz in the cell raise the possibility that mis-regulation of cytokine accumulation and secretion drive spurious immune cell activation . Previous studies have shown that PIP2-rich membranes are favored for plasma membrane fusion [61] , and the membrane composition of late endosomes determines whether they fuse with the plasma membrane or lysosomes [62 , 63] . Rab11CA may rescue this aberrant release by redirecting Spz from secretion into the hemolymph towards a canonical recycling pathway . In addition to the Tl ligand Spz , the Jak-STAT Unpaired ( Upd ) ligands are promising candidates to contribute to the docrl phenotype . Upd2 and Upd3 are upregulated in hemocytes in response to parasitic wasp infection , and promote STAT activation in muscle , which in turn promotes hemocyte differentiation [31] . Notably , Upd ligands lack signal peptides , and so may be secreted by unconventional trafficking pathways [64] . Additional studies , and new tools to visualize Upd trafficking and secretion , will be required to investigate whether and how dOCRL might regulate Upd secretion . Finally , in the future , it will be interesting to determine whether and how changes in Spz or Upd trafficking participate in normal immune responses . Here , we showed that docrl is required specifically in cells of the innate immune system to suppress spurious activity . One interesting hypothesis is that innate immune cell phenotypes might underlie epilepsy and cystic brain lesions in Lowe Syndrome patients [65] . There is strong evidence for a link between inflammation , innate immunity , and epilepsy [66] . Most strikingly , a recent report implicates that early immune challenge by LPS in mice promotes an astrocytic TLR4-MyD88 signaling pathway that enhances excitatory synaptogenesis and subsequent seizure susceptibility [67] . Several studies also suggest a direct link between OCRL , neuroinflammation , and seizure . Zebrafish ocrl mutants feature cystic lesions in the brain that are enriched in glia , and exhibit seizure susceptibility [15] . Further , pilocarpine seizure induction in mice led to decreased OCRL levels in hippocampal astrocytes [68] . Additionally , a mammalian analogue of Upd , the cytokine IL-6 , plays diverse roles in inflammatory responses [69–71] , is secreted by both astrocytes and microglia in the brain in response to LPS [72] , and contributes to glial scarring following spinal cord injury [73] . Though it remains unclear in many of these cases whether these changes are pathological or compensatory , it will be important to investigate the link between OCRL and seizure in these mammalian systems . Together , these data raise the possibility that loss of OCRL causes seizures in humans due to immune activation in the brain , by mechanisms similar to those we have uncovered in Drosophila . Thus , our finding that dOCRL acts specifically in hemocytes to restrain innate immune cell activation provides a novel line of inquiry into the pathogenesis of the symptoms in Lowe Syndrome patients . All graphs show mean ± SEM . Statistical significance was calculated with Prism software ( GraphPad , La Jolla , CA ) as follows for specific datasets: We used one-way ANOVA followed by pairwise Tukey's test ( Figs 1A , 3E , 4D , 5D , 6C ( intensity ) , 7A and 7C ) , or Kruskal-Wallis test with post-hoc Dunn’s test ( Figs 1B , 1C , 4B , 4C , 4F , 6A , 7D , 8B , 8C , 8G and 8H , S7B , S7C , S8A and S8D Figs ( whole cell & nuclear Dl ) ) . For comparisons between two groups , we utilized Student's t test ( Fig 8D , S6A , S8A ( % nuclear ) and S8C Figs ) or Mann-Whitney U test ( Figs 3D , 4E , 6B and 6C ( PCC ) , S3A , S6A and S8B Figs ) . ANCOVA was used to analyze S6A ( Lysotracker-GFP comparison ) . Chi squared tests were used to calculate differences from expected distributions of cell types ( Figs 2D , 2E , 5E and 6B , S2C Fig ) . In all cases p < 0 . 05 , p < 0 . 01 , **p < 0 . 001 . Drosophila larvae were cultured using standard media at low density at 25°C for all experiments , unless noted otherwise . To generate docrl mutants , P[EPgy2]EY15890 ( 732 bp upstream of the docrl start codon ) was mobilized using a Δ2–3 transposase , in the mus309 mutant background for docrlΔpre , docrlΔ1 , docrlΔ2 , and docrlΔ3 [74] . 600 candidate lines were screened by PCR to identify deletions , which were subsequently sequenced to determine precise molecular coordinates . Line docrlΔpre contains a sequence-verified precise excision of the P-element . dOCRL and VPS35 were cloned using Gateway technology ( Life Technologies , Inc ) into pBI-UASc-gateway-GFP [75] . Constructs were injected into flies ( Genetic Services Inc . Cambridge , MA ) , using Φc31 recombinase at the Attp40 locus [76] . Additional fly strains used have been previously described ( and where noted with BL stock numbers , are available at the Bloomington Drosophila Stock Center ( BL ) or the Vienna Drosophila Stock Center ( v ) ) as follows . Alleles: vps26 ( BL26623 ) , snx1Δ2 and snx3Δ1 [77] , Df ( X ) ED6565 ( BL9299 ) , Dup ( X ) DC402 [78] , Tlr3 ( BL3238 ) , Df ( 3R ) BSC524 ( Tl; BL25052 ) , spz4 ( BL55718 ) , Df ( 3R ) Ex6205 ( spz; BL7684 ) ; spe[6] [79] , Df ( 3R ) BSC491 ( spe; BL24995 ) . Drivers: He-Gal4 ( BL8699 , [38] ) , Lsp2-Gal4 [80] , Dot-Gal4 ( BL6902 , [32] ) , Ser-GAL4 [34] , dome-GAL4 [35] ) , HH-LT [33] , and mef2-Gal4 ( BL2739 ) . Endogenously labeled proteins: dOCRLT-STEP [39] and Rab GTPases [81] . Expression lines: UAS-dOCRL-RNAi ( v110796 ) , Syntaxin17 RNAi ( BL25896 ) , UAS-Spz-GFP [82] , UAS-Toll-Venus [83] , UAS-GFP-mCherry-Atg8a [50] , UAS-Clc-GFP ( BL7109 , Henry Chang ) , UAS-OCRLptase [42] . , UAS-PHPLC-cherry ( BL51658 , [84] ) , UAS-shiTS ( BL44222 ) , UAS-shiK44A ( BL5811 ) , UAS-lacZ ( BL1777 ) , UAS-Rab11QL [85] , rab11N124I [86] . Remaining UAS-Rab constructs including YFP-Rab5QL ( BL9773 ) , Rab7QL ( BL9779 ) , Rab7SN ( BL9778 ) , Rab5SN ( BL42704 ) , Rab35SN ( BL9820 ) were described in [87] . Confocal imaging was conducted on an Andor Revolution spinning disk system consisting of a Nikon Ni-E upright microscope , equipped with 40x ( n . a . 1 . 3 ) , 60X ( n . a . 1 . 4 ) , and 100X ( n . a . 1 . 45 ) oil immersion objectives , a Yokogawa CSU-W1 spinning disk head , and an Andor iXon 897U EMCCD camera . Confocal imaging was used to acquire data related to protein subcellular localization and abundance ( Figs 2 , 3A–3D , 4 and 6 ) . Widefield imaging was conducted on a Ni-E inverted microscope equipped with a Spectra-X LED light engine and a Zyla sCMOS camera , and a 60X ( n . a . 1 . 4 ) or 10x ( n . a . 0 . 3 ) objective . Widefield imaging was used to measure cell relative cell counts and actin accumulation ( Figs 1F , 3E , 3F and 5 ) . Images were collected using Nikon Elements AR software . Fluorescence microscopy image processing and analysis was performed in FIJI ( National Institutes of Health , Bethesda , MD ) . Fluorescence intensity measurements ( area , perimeter , mean and integrated intensity ) were performed on sum intensity projections . Plasma membrane ratio of MyD88 was calculated as follows: Cell profiles of single Z sections taken at the approximate midpoint of the cell body were thresholded in FIJI , and the resulting cell mask was divided into three circumferential regions of approximately equal area . The mean fluorescence intensity in the outer ring ( which captured the plasma membrane signal ) was divided by the mean fluorescence intensity in the adjacent ring ( which captured a representative section of cytoplasm ) . Pearson R was calculated using Coloc2 ( FIJI ) on 3D cell image stacks . Wandering 3rd instar larvae were used for all hemocyte experiments . For injury assays , larvae were pierced with a tungsten needle that was either sterile , or coated with M . luteus ( ATCC 27141 ) . Control ( unpierced ) and injured larvae were then allowed to recover for 4 h on agar plates with yeast paste . Hemolymph was extracted from individual larva or groups of 2–5 larvae . Larvae were collected into PBS , quickly washed with 70% ethanol and then rinsed three times in PBS . Hemolymph was collected by tearing the larval cuticle into PBS with 0 . 01% phenylthiourea . For absolute hemocyte counts ( Fig 1A and 1B ) , hemolymph was loaded onto each side of a disposable hemocytometer ( Incyto C-Chip DHC-N01 ) and allowed to settle for 30 minutes in a moist chamber before quantification of GFP positive cells using a 20x objective on an EVOS FL Cell Imaging System . For relative hemocyte counts ( Figs 1F , 3E , 5A and 5D ) and immunocytochemistry , hemolymph was extracted into PBS from 2 or more pooled larvae , and placed in each well of a multi-chamber microscope slide ( Thermo Scientific Nunc Lab-Tek II 8-chamber slides ) . Each well , containing an independent collection of hemocytes from pooled larvae , represents a single sample . Hemocytes were allowed to settle for 60 minutes in a moist chamber at room temperature and then fixed for 10 minutes in 4% formaldehyde in PBS , washed in PBS , and then permeabilized , stained with primary and secondary antibodies and washed with PBX ( PBS with 0 . 1% Triton-x-100 ) . Slides were mounted with Mowiol with DABCO . Cells were imaged by either confocal or widefield microscopy , and counted from at least 6 fields of view per sample . Lymph glands were dissected in ice-cold PBS and transferred immediately to 1mL of PBS + 1 drop of PBS + 0 . 1% Triton X-100 . All glands dissected within 15 minutes were then fixed for 35 minutes in 4% formaldehyde and processed for immunostaining as described for hemocytes . For live imaging , 1–3 wandering third instar larvae/genotype were bled directly into 50uL M1 medium supplemented with BSA ( 1 . 5mg/ml ) and D-glucose ( 2 mg/ml ) . Hemocytes were allowed to settle 5 minutes , and then coverslips were affixed to a glass slide by double-sided tape ( 3M ) , which simultaneously served as a bridge . A single experiment was considered to be the aggregate of all cells imaged of a single genotype during a single imaging session , and all relevant comparisons were made only between groups imaged on the same day ( over the course of ~3 hours ) imaged with identical settings . To image whole larvae , animals were mounted on double-sided sticky tape and incubated for 20 mins at -20°C before imaging by confocal microscopy . Circulating and sessile hemocyte populations were separated as described [88] and imaged on an EVOS FL Cell Imaging System . A fragment of dOCRL encoding amino acids 1–179 was cloned into pGEX-6P ( GE Healthcare ) . E . coli strain BL21 ( DE3 ) expressing this construct was grown to log phase at 37°C , then induced for 3 h at 37°C with 0 . 4 mM IPTG . Cells were lysed in PBS ( phosphate-buffered saline , pH 7 . 4 ) supplemented with 0 . 5 mM DTT , 0 . 5% Triton-X100 , 0 . 5 mg/ml pepstatin , leupeptin and aprotonin and 1 mM PMSF . Lysates were purified on glutathione agarose ( GE Healthcare ) , washed 4 times with 50 ml of PBS with 0 . 5 mM DTT , and GST was cleaved from dOCRL1-178 at 4° overnight with a ~1:50 molar ratio of Precision Protease ( GE Healthcare ) . Supernatant containing the cleaved protein was further purified by gel filtration on a Superose 12 10/30 column equilibrated in PBS with 0 . 5 mM DTT . Purified protein was flash frozen in liquid nitrogen at a final concentration of 0 . 3 mg/ml , and sent to Cocalico , Inc . for injection into rabbits . Serum from Rabbit #18 was affinity purified against GST-dOCRL1-178 , which was expressed as above , purified on a Profinia system ( Biorad ) with a glutathione affinity column , and eluted with glutathione according to manufacturer’s instructions . Glutathione eluates were gel filtered into PBS on a Sephacryl S-200 16/60 column ( GE Healthcare ) , and conjugated to an Aminolink immobilization column ( Thermo-Pierce ) using the cyanoborohydride method , according to manufacturer’s instructions . 2 mL of α-dOCRL serum from Rabbit #18 were incubated for 2 h on the resin , then washed 5 times with 2 mL PBS , and eluted with 0 . 1 M glycine , pH 2 . 5 . The eluate was then neutralized by adding Tris pH 9 . 0 to a final concentration of 25 mM , and stored at 4°C . Additional antibodies used were α-L1 and P1 ( gift of Istvan Ando ) α-Lamin Dm0 ( DSHB ADL67 . 10 ) , α-Spz ( gift of S . Goto ) , α-MyD88 ( gift of S . Wasserman ) , α-actin ( DSHB JLA20 ) , α-pH3 ( Abcam ab5176 ) , α-pJNK ( Promega V793A ) . To measure phagocytosis , hemocytes were extracted and spread on slides for 30 min as described above . Cells were then washed twice with PBS , and incubated for the indicated times with 100 μl PBS containing 6x106 particles Alexa 488-labeled E . coli ( Life Technologies ) . Cells were then washed quickly 5 times with 500 μl PBS before fixing for 15 minutes in 4% formaldehyde in PBS , washing in PBS , and imaging as above . Receptor-mediated endocytosis was measured essentially as previously described [49 , 89] . Specifically , hemocytes were extracted and spread for 5 mins on slides as described for immunocytochemistry , then pulsed for 45 sec with 5ug/mL Cy5-labeled maleylated BSA ( see below ) in M1 medium ( 150 mM NaCl , 5 mM KCl , 1 mM CaCl2 , 1 mM MgCl2 , 20 mM HEPES , pH 6 . 9; supplemented with BSA ( 1 . 5 mg/ml ) and D-glucose ( 2 mg/ml ) ) , and chased with M1 medium at room temperature . Cells were then fixed and imaged as above . Cy5-labeled maleylated BSA ( Cy5-mBSA ) was prepared as described in [90] and labeled with bis-functional Cy5 according to manufacturer’s instructions ( GE Healthcare ) ) . Images were analyzed by measuring Cy5 intensity at a single central Z-section , in which nearly all signal is internalized mBSA . For Lysotracker uptake assays , hemocytes were extracted and spread for 5–15 mins on slides as described for immunocytochemistry , then incubated for 30 mins with 50 nm Lysotracker Deep Red ( Thermo-Fisher Scientific ) . Cells were fixed and imaged as above . Whole 3rd instar larvae were homogenized in Laemmli sample buffer ( 20 μl/larvae ) and boiled for 1 min . 10 μl ( the equivalent of 0 . 5 larvae ) were loaded for immunoblotting . Larval hemolymph was isolated as follows: A ~2mm slit was made in the bottom of a 500uL Eppendorf tube and the cap was removed . The indicated number of larvae were pierced at their posterior end by a pair of forceps and added to the prepared tube on ice . The tube was placed in a 1 . 7mL Eppendorf tube and spun at 1000xG for 10 sec . 50 μl ice cold PBS with 0 . 01% phenylthiourea was added to hemolymph and centrifuged at 5000 x g for 5 min . 25 μl solution was reserved as cell free hemolymph , and boiled for 1 min in 25 μl Laemmli sample buffer . Boiled samples were centrifuged at 14 , 000 x g for 1 min before loading on SDS-PAGE gels and transferring to nitrocellulose . Blots were labeled with Alexa 680-conjugated secondary antibodies , and detected on a LI-COR Odyssey infrared detection system .	Lowe syndrome is a developmental disorder characterized by severe kidney , eye , and neurological symptoms , and is caused by mutations in the gene OCRL . OCRL has been shown to control many steps of packaging and transport of materials within cells , though it remains unclear which of these disrupted transport steps cause each of the many symptoms in Lowe syndrome patients . We found that in fruit flies , loss of OCRL caused transport defects at specific internal compartments in innate immune cells , resulting in amplification of multiple critical inflammatory signals . Similar inflammatory signals have been implicated in forms of epilepsy , which is a primary symptom in Lowe syndrome patients . Thus , our work uncovers a new function for OCRL in animals , and opens an exciting new avenue of investigation into how loss of OCRL causes the symptoms of Lowe syndrome .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "blood", "cells", "cell", "physiology", "medicine", "and", "health", "sciences", "immune", "cells", "vesicles", "immunology", "developmental", "biology", "cellular", "structures", "and", "organelles", "endosomes", "white", "blood", "cells", "animal", "cells", "life", ...	2017	dOCRL maintains immune cell quiescence by regulating endosomal traffic
Plasmodium relapses are attributed to the activation of dormant liver-stage parasites and are responsible for a significant number of recurring malaria blood-stage infections . While characteristic of human infections caused by P . vivax and P . ovale , their relative contribution to malaria disease burden and transmission remains poorly understood . This is largely because it is difficult to identify ‘bona fide’ relapse infections due to ongoing transmission in most endemic areas . Here , we use the P . cynomolgi–rhesus macaque model of relapsing malaria to demonstrate that clinical immunity can form after a single sporozoite-initiated blood-stage infection and prevent illness during relapses and homologous reinfections . By integrating data from whole blood RNA-sequencing , flow cytometry , P . cynomolgi-specific ELISAs , and opsonic phagocytosis assays , we demonstrate that this immunity is associated with a rapid recall response by memory B cells that expand and produce anti-parasite IgG1 that can mediate parasite clearance of relapsing parasites . The reduction in parasitemia during relapses was mirrored by a reduction in the total number of circulating gametocytes , but importantly , the cumulative proportion of gametocytes increased during relapses . Overall , this study reveals that P . cynomolgi relapse infections can be clinically silent in macaques due to rapid memory B cell responses that help to clear asexual-stage parasites but still carry gametocytes . Due to their ability to establish dormant forms in the liver called hypnozoites , relapsing malaria parasites pose a significant obstacle to malaria elimination [1] . Hypnozoites can become activated , resulting in repeat blood-stage infections , and these relapses account for the majority of P . vivax blood-stage parasitemias [2 , 3] . This raises critical questions about the relative importance of relapses in causing illness , and their importance in transmission . However , there are very few studies that have directly examined relapse biology , especially in the context of pathogenesis , host immunity , and transmission . An improved understanding of relapses and their immunological and epidemiological implications is needed to design effective control and elimination strategies for relapsing malaria parasites . There are no rodent malaria parasites species that form hypnozoites , making these models inadequate for studying relapses . Human studies in endemic areas have limited utility because it is generally difficult to determine whether a blood-stage infection resulted from new , relapsing , or recrudescent infections [4 , 5] . Although approaches such as parasite genotyping , relocation of individuals from P . vivax endemic areas to non-endemic areas , and mass drug administration provide more confidence that a P . vivax infection is due to a relapse , these approaches have caveats [6–10] . For example , new genotypes detected in sequential blood-stage infections could be due to the activation of hypnozoites that did not originally activate and circulate in the blood at the time of an initial sample collection; thus , these parasites could be mistaken as parasites from a new infection even if they originated from hypnozoites in the liver . Additionally , the inability to control for an individual’s infection history complicates the investigation of immune responses during human relapse infections . Nonhuman primate ( NHP ) models lack these barriers and present several advantages for studying relapses . In particular , rhesus macaques infected with Plasmodium cynomolgi , a simian malaria parasite closely related to P . vivax , recapitulate the biological , clinical and pathological features of P . vivax malaria , including hypnozoite formation and relapses [11–15] . Further , P . cynomolgi is now recognized as a zoonotic species in South East Asia , making understanding its biology a priority [16–18] . These species have similar 48-hour intraerythrocytic developmental cycles , antigenic makeup , and infected red blood cell ( iRBC ) modifications that include abundant caveolae vesicle complexes [19–22] . Moreover , genomes and immunological tools are available to support the study of host-parasite interactions and NHP immune responses . Hence , these factors make the rhesus macaque–P . cynomolgi model valuable for the study of relapse biology . We previously showed that P . cynomolgi relapses in rhesus macaques have substantially reduced parasitemias compared to primary infections , and they do not result in anemia or other clinically detectable disease manifestations [12] . The present study explored the host-pathogen interactions that underpin clinically silent P . cynomolgi relapses to garner insights into asymptomatic P . vivax relapses in humans [7 , 23 , 24] . We hypothesized that humoral immunity could explain the lack of clinical disease during relapses since passive transfer of antibodies has been shown to control parasitemia and ameliorate disease during human , NHP , and rodent Plasmodium infections [25–27] . Here , we demonstrate that lack of clinically detectable disease during P . cynomolgi relapses is associated with rapid memory B cell responses and the swift rise of anti-parasite IgG1 antibodies that can mediate clearance . The same humoral immune responses were also associated with protection against subsequent challenge with the same parasite strain about 60 days after radical cure . Interestingly , the immune response reduced the number of sexual stage gametocytes present compared to the primary infection , but the cumulative proportion of gametocytes increased during relapses . This suggests that the immune response generated by the infection primarily targeted the asexual stages . Concordantly , mature gametocyte gene expression was not significantly different between primary infections and relapses . Together , these data show that clinically silent , P . cynomolgi relapses carry gametocytes despite a significant reduction in parasitemia associated with an effective humoral immune response . Overall , this study broadens our understanding of relapsing malaria parasite pathogenesis and infections , with important epidemiological implications relevant to malaria elimination strategies . Six rhesus macaques were inoculated intravenously with 2 , 000 P . cynomolgi M/B strain sporozoites on Day 0 , and the parasitemia and clinical status of each individual monkey was evaluated daily for up to 100 days by light microscopy and complete blood count ( CBC ) analysis , respectively . Blood specimen collection time points are indicated in Fig 1A and 1B . A summary of the clinical and parasitological criteria for each specimen collection is provided in S1 Table . The infections reached patency between days 10–12 ( mean ± SE = 11 . 16 ± 0 . 3 ) after inoculation . Parasitemia peaked between 276 , 981–540 , 156 parasites per microliter of blood ( mean ± SE = 400 , 563 ± 33 , 028 parasites/μl ) between days 17–19 post-inoculation ( Fig 1B ) . After collecting blood samples at the peak , all animals were administered a sub-curative dose of artemether to reduce but not eliminate the blood-stage parasites . This treatment ensured that the animals would remain parasitemic but not develop severe disease . Following the administration of curative blood-stage treatment , relapses were observed at different time intervals for each individual ( Fig 1B ) . As anticipated based on earlier studies , relapse infections had approximately 200-fold lower parasitemia than the primary infections ( p < 0 . 05; Fig 1B and 1C ) . Notably , relapse parasitemias declined below patency 7 to 15 days after their detection in the blood without treatment ( Fig 1B ) . The primary infection induced a myriad of clinical manifestations with elevated temperatures ( mean ± SE = 102 . 3 ± 0 . 44°F ) although not statistically significant and was accompanied by significant decreases in hemoglobin levels and platelet counts compared to when the animals were naïve ( Fig 1D ) . Temperatures remained elevated post-peak for approximately one week after the administration of subcurative artemether treatment ( Fig 1B and 1D ) . The anemia severity was moderate to severe with hemoglobin nadirs ranging from 5 . 7–9 . 0 g/dl ( mean ± SE = 7 . 3 ± 0 . 4 g/dl ) after the peak of parasitemia ( S1A Fig ) . Platelet counts dropped from a mean of 421 , 250 platelets/μl when naïve to an average of 119 , 000 platelets/μl at the peak of the primary infection ( Figs 1D and S1B ) . In stark contrast to the primary infections , anemia , thrombocytopenia , and fever were not observed in the animals during relapses ( Fig 1D , S1A and S1B Fig ) . Clinical severity of malaria has been correlated with inflammatory responses to infecting Plasmodium parasites [28–31] . Congruent with elevated temperatures and clinical presentation , inflammation was highest at the peak of parasitemia when 22 out of 45 cytokines , chemokines , or growth factors tested were significantly increased in the plasma compared to malaria naïve values ( Figs 1E , 1F and S2 ) . Pyrogenic cytokines such as IL-1β , TNFα , IL-6 and IFNγ were significantly increased in plasma at peak parasitemia when the animals were presenting with clinical illness ( Fig 1F ) . Notably , only 6 out of 45 analytes remained significantly elevated after the administration of sub-curative artemether treatment ( S2 Fig ) . In concordance with the clinical presentations , cytokine responses were subdued during relapses compared to the primary infection ( Fig 1E ) . IL-1β , TNFα , IL-6 and IFNγ were not significantly elevated during relapses compared to pre-infection values , and Monokine Induced by Interferon Gamma ( MIG ) was the only cytokine significantly increased during relapses ( Figs 1F and S2 ) . The decrease in inflammation was likely due to the significant reduction in parasitemia that was observed between the initial infections and relapses ( Fig 1B and 1C ) . Overall , these results reaffirmed that P . cynomolgi relapses had substantially reduced parasitemia and did not cause clinical signs of malaria . To identify host responses that may reduce parasitemia and disease during relapses , we performed RNA-Seq on whole blood samples collected at the time points indicated in Fig 1A . The sequencing reads were mapped to concatenated host and parasite genomes , normalized by library size , log2 transformed , and finally , variance due to inter-individual variability was removed via the SNM transformation as previously described [32 , 33] . Unsupervised hierarchical clustering of transcriptional profiles identified three major clusters as indicated by blue , purple , and yellow shading; these clusters captured approximately 30% of the variance associated with changes in gene expression ( Fig 2A ) . Generally , the clinical presentation appears to drive the clustering pattern . The blue cluster consists of samples collected when the animals were not experiencing clinical symptoms , including prior to infection , post peak when the animals were recovering from illness , and during relapses ( Fig 2A ) . In contrast , the yellow cluster consists of samples collected at the peak when clinical signs of malaria were evident , and the purple cluster consists of samples acquired after parasites were detected in the blood but prior to the onset of clinically detectable disease ( Fig 2A ) . These results demonstrated that distinct changes in the host transcriptome occurred during the course of P . cynomolgi blood-stage infections . Samples collected as relapses were resolving ( i . e . relapse resolution ) formed a distinct subcluster compared to the early and peak relapse samples within the blue cluster ( Fig 2A ) . Importantly , this subclustering showed that the transcriptional changes associated with resolving relapse infections were similar across individuals and distinct from the transcriptional changes during primary infections . In agreement with the clustering pattern , differential gene expression analysis identified major transcriptional changes during the primary infection with more muted changes during relapses . Differentially expressed genes ( DEGs ) were determined via ANOVA followed by a t-test post-hoc analysis with Benjamini-Hochberg false discovery rate ( FDR ) correction . Genes with an FDR adjusted p-value of less than 0 . 05 were considered significantly differentially expressed . The DEG analysis was focused on the identification of genes that are differentially expressed between the malaria naïve time point and each subsequent time point ( e . g . malaria naïve vs . peak , malaria naïve vs . relapse resolution , etc . ) . Compared to when the animals were naïve , the largest number of DEGs were identified during the peak of parasitemia , with approximately 3 , 350 and 4 , 713 DEGs upregulated and downregulated , respectively ( Fig 2B ) . The pre-peak and post-peak time points induced comparable changes in host gene expression with over 2 , 000 genes upregulated and over 2 , 000 genes downregulated for each ( Fig 2B ) . In contrast , relapses caused substantially less changes in host-gene expression compared to when the animals were naive ( Fig 2B ) . Only 44 upregulated and 74 downregulated DEGs were identified during the early relapses , and 291 upregulated and 316 downregulated DEGs during the relapse peaks ( Fig 2B ) . In contrast , the relapse resolution infection points had the most DEGs during relapses with 1 , 294 upregulated and 1 , 459 downregulated DEGs ( Fig 2B ) . Next , we focused on genes that were only differentially expressed during relapses . All upregulated and downregulated DEGs during the primary and relapse infections were compared . Six hundred and thirty-five upregulated and 607 downregulated DEGs were determined to be unique to relapses ( Fig 2C and 2D ) . Metacore pathway enrichment analysis based on these upregulated and downregulated DEG sets revealed nine and over 50 significantly enriched pathways , respectively ( S2 and S3 Tables ) . The nine upregulated pathways identified during the relapses are related to B cells , T cells , cell signaling , and antigen presentation ( Fig 2C ) . The pathway with the highest enrichment score was related to B cells , and three out of the nine pathways are related to B cell responses ( e . g . , B cell responses in SLE , B cell signaling in hematological malignancies , and the B cell receptor pathway ) ( Fig 2C ) . The genes that were responsible for the enrichment of these pathways are composed of B cell surface proteins such as the BAFF-R , CD79A , and CD79B in addition to signaling molecules like Btk and VAV-2 ( S2 Table ) . In contrast , the pathways related to T cells , signaling , and antigen presentation are composed of genes such as Akt , MAP kinases , CD86 , T-bet , MHC-II , etc . , which are expressed by a variety of immune cells , including B cells ( S2 Table ) . The top downregulated DEGs during relapses belong to pathways related to inflammation , cell damage responses , and innate immune cells , such as dendritic cells , macrophages and neutrophils ( Fig 2D and S3 Table ) . Together , this analysis showed that relapses induced relatively minor , but unique changes in the host transcriptome compared to the primary infections that were predominantly related to the downregulation of pathways involved in innate immune responses and upregulation of pathways related to B cells . Since the transcriptome data suggested that B cell responses may be involved in ameliorating disease during relapses , we next performed flow cytometry analysis on peripheral blood mononuclear cells ( PBMCs ) isolated during the primary and relapse infections . We utilized a B cell immunophenotyping strategy previously developed for human immunology studies and optimized it for specimens collected from rhesus macaques [34] . With this strategy , B cell subsets in PBMCs are CD19 and CD20 positive and further classified into four subsets based on the surface expression of IgD and CD27 ( Figs 3A and S3 ) . Similar to humans , four B cell subpopulations were identified in PBMCs from rhesus macaques: naïve ( IgD+CD27- ) , unswitched memory ( USM: IgD+CD27+ ) , switched memory ( SM: IgD-CD27+ ) , and double-negative B cells ( DN: IgD-CD27- ) . Surface IgM was present in all subsets , as previously shown for humans , albeit at low frequencies in the SM compartment ( S3B and S3C Fig; [34 , 35] ) . IgG surface staining was evident in the SM and DN populations but not in naïve B cells ( S3B and S3C Fig ) . In summary , this immunophenotyping strategy yielded comparable results for samples acquired from either humans or rhesus macaques . All B cell subsets decreased in the blood at pre-peak ( Fig 3B ) . This decrease was consistent with the pan-lymphopenia observed during the initial infection ( S4 Fig ) . At the peak , naïve and DN B cells stabilized in the periphery whereas SM and USM B cells remained significantly reduced compared to when the animals were naive ( Fig 3B ) . Although USM and SM B cell numbers were reduced , the percentage of Ki67+ USM and SM B cells was significantly increased at the peak , suggesting these cells were activated ( Fig 3C ) . Following subcurative treatment , USM B cells stabilized and SM B cells significantly increased from 263 ± 75 cells/μl ( mean ± SE ) when the animals were uninfected to 761 ± 242/μl ( mean ± SE ) post-peak ( Fig 3B ) . During relapses , USM B cell numbers expanded from 153 ± 187/ μl ( mean ± SE ) when naïve to 686 ± 89/μl ( mean ± SE ) , and SM B cells rose to 1982 ± 590 cells/μl ( mean ± SE ) ( Fig 3B ) . The increase in the absolute numbers was accompanied by an increase in the frequency of Ki67+ USM and SM B cells , consistent with a new expansion of these cells during the relapses ( Fig 3C ) . In contrast , the absolute numbers of naïve and DN B cells were unchanged ( Fig 3B and 3C ) . Notably , there was an increase in frequency of Ki67+ DN B cells during relapses , although the absolute numbers of this population did not increase ( Fig 3B and 3C ) . Parasite-specific IgG and IgM are important for controlling parasitemia . Early during relapses IgG+ SM B cell frequencies were 330 ± 49/μl , which is similar to the naïve numbers ( 225 ± 71/μl ) , but these rapidly increased approximately 8-fold to 1 , 768 ± 567/μl as the relapse infections resolved ( Fig 3D ) . Thus , the expansion in IgG+ SM B cell numbers was induced by the relapse . IgM+ SM B cells also increased in response to the relapses , but the absolute numbers of these cells were smaller in comparison to the IgG+ SM B cells ( Fig 3D ) . Finally , we evaluated if the changes in B cell subsets were correlated with parasitemia during primary and relapse infections . There was not a significant correlation between the number of naïve , DN , USM , and SM B cells and parasitemia during the primary infections ( Figs 3E and 3F and S5 ) . In contrast , USM ( Spearman’s ρ = -0 . 57 , p = 0 . 04 ) and SM ( Spearman’s ρ = -0 . 57 , p = 0 . 03 ) B cells were inversely correlated with parasitemia during relapses ( Fig 3E and 3F ) . Notably , naïve and DN B cell numbers were not inversely correlated with parasitemia during relapses ( S5 Fig ) . Overall , these data illustrate that USM and SM B cells dramatically expand during relapses in response to a new blood-stage parasitemia and may be involved in controlling parasitemia and ameliorating disease . Since B cell responses and antibodies are critical for suppressing parasitemia , we determined whether anti-parasite IgM and IgG were increased during relapses using an ELISA with infected RBC ( iRBC ) and uninfected RBC ( uRBC ) lysates . Total IgM ( tIgM ) increased from 0 . 34 ± 0 . 02 mg/ml to 1 . 93 ± 0 . 18 mg/ml ( mean ± SE ) at the peak of the primary infections and remained elevated post-peak ( S6A Fig ) . P . cynomolgi-specific IgM increased at the peak and remained increased post-peak ( Fig 4A ) . Total and parasite-specific IgG also increased at post-peak ( Figs 4B and S6B ) . The IgG subclass of the iRBC-specific IgG produced during the primary infections was IgG1 ( S6C Fig ) . Neither IgM nor IgG were inversely correlated with parasitemia during the primary infections ( Fig 4C and 4D ) . Notably , we did not observe a difference between the reactivity of IgM that recognized iRBC versus uRBC lysates during the primary infections ( S7A Fig ) . Similar to IgM , uRBC-specific IgG also increased during the primary infections , and a significant difference between IgG recognizing iRBC versus uRBC was also not discernable ( S7B Fig ) . uRBC lysate-specific IgM and IgG were both inversely correlated with hemoglobin levels during the primary infections , suggesting that these antibody responses may be linked with the loss of uRBCs and the development of anemia during P . cynomolgi infections ( S7C and S7D Fig ) [36] . In contrast with the primary infections , relapses did not result in significant changes in tIgM levels ( S6A Fig ) . However , there was an increase in iRBC-specific IgM at the relapse resolutions , but this increase was approximately half of what was detected during the primary infections ( Fig 4A ) . In contrast , iRBC-specific IgG was rapidly produced during relapses and was significantly increased during the peak relapse and the relapse resolution periods ( Fig 4B ) . These values are five-fold higher than those observed during the primary infections ( Fig 4B ) . The increase in IgG occurred alongside the expansion of IgG+ SM B cells that peaked during relapse resolutions , strongly suggesting that this response was important for controlling parasitemia during a relapse ( Figs 3D and 4B ) . As in the primary infections , the iRBC-specific IgG was IgG1 ( S6C Fig ) . Next , we determined if the iRBC specific IgM and IgG were inversely correlated with parasitemia during relapses . Notably , the early relapse time points for monkeys RBg14 and RIb13 were excluded from this analysis because these time points were taken one to two days before a relapse was patent . Including these time points would confound our analysis since the relapse resolution time points were also taken when the parasitemia was below patency . As expected , iRBC-specific IgM and IgG were inversely correlated with parasitemia during relapses ( Fig 4C and 4D ) To determine the functionality of the antibodies generated during the primary and relapse infections with respect to clearance of iRBCs , we performed a phagocytosis assay using a THP-1 monocyte cell line , as previously described [37] . The percentage of THP-1 monocytes that phagocytosed iRBCs after being opsonized with heat-inactivated plasma increased from 5 . 1 ± 0 . 31% from naïve animals to 9 . 3 ± 1 . 1% and 10 . 8 ± 0 . 7% at the peak and post-peak of the initial infections , respectively ( Fig 4E ) . During relapses , the percentage of THP-1 monocytes that phagocytosed iRBCs was even higher ( 16 . 0 ± 1 . 4% ) ( Fig 4E ) . Interestingly , the opsonization of iRBC by the heat-inactivated plasma was positively correlated with the amount of iRBC-specific IgG ( Spearman’s ρ = 0 . 53 , p = 0 . 01 ) and inversely correlated with iRBC-specific IgM ( Spearman’s ρ = -0 . 65 , p = 0 . 0017 ) ( Fig 4F ) . Collectively , these data demonstrate that the anti-parasite antibodies produced during relapses can mediate iRBC clearance , consistent with a key role for IgG+ SM B cells in providing rapid host immunity to suppress parasitemia during relapses . We questioned whether the animals would remain protected months later against a homologous parasite challenge infection and , if so , if the immune responses would be similar to those observed during a relapse . The same cohort of macaques was administered two rounds of radical cure to best ensure elimination of all liver- and blood-stage parasites . Approximately 60 days later they were re-challenged with 2 , 000 P . cynomolgi M/B strain sporozoites . Specimen collections were performed according to the schematic shown in Fig 5A and S4 Table . The reinfections with the homologous strain reached patency between days 9–12 post infection ( mean ± SE = 11 ± 0 . 54 ( Fig 5B ) . Peak parasitemias were substantially reduced compared to the initial infection and ranged from 269–5 , 742 parasites/μl ( Fig 5B and 5C ) . There was no evidence that the decrease in parasitemia was directed against the sporozoite or liver stages since the number of days to patency was similar between the initial infections and homologous reinfections ( S8 Fig ) . Like relapses , homologous reinfections did not cause clinical signs of malaria ( Figs 5D and S9 ) . Changes in cytokine profiles were also minimal with only IL-7 and RANTES differing significantly from pre-homologous values ( Fig 5E ) . The pyrogenic cytokines IL-6 , TNF-α , IL-1β , or IFN-γ did not increase during homologous reinfection and were not significantly different from values obtained during relapses ( Fig 5F ) . Together , this homologous challenge experiment demonstrated that non-sterilizing immunity persisted for at least 60 days after radical cure and that this immunity could control peripheral parasitemia and the clinical manifestations of malaria in the P . cynomolgi model . Since relapses and homologous reinfections had similar clinical presentations , we next employed RNA-Seq analysis on whole blood collected during the homologous challenges to determine if the host responses were similar . To identify DEGs during the homologous reinfections , the primary and post primary time points were compared to the pre-homologous challenge time point by ANOVA followed by a t-test post-hoc analysis with Benjamini-Hochberg false discovery rate ( FDR ) correction . Genes with an FDR adjusted p-value of less than 0 . 05 were considered significantly differentially expressed . As with the relapses , the homologous reinfections induced minimal changes in the host transcriptome ( Fig 6A ) . Forty-five percent ( 18/40 ) of the upregulated DEGs during the homologous reinfections overlapped with the upregulated DEGs during relapses ( Fig 6B ) . Pathway enrichment analysis of all upregulated DEGs during the homologous reinfections again identified pathways related to B cells ( Fig 6C ) . The genes that were enriched in these pathways include those encoding B cell surface proteins such as CD19 and CD20 and B cell signaling molecules such as AKT , BTK , PLC-gamma2 , and VAV-2 ( S5 Table ) . Similar to relapses , these data indicate that the changes in the host responses during homologous reinfections were characterized by pathways involving B cells . The changes in B cell subsets during the homologous challenge experiment were similar to those measured in relapse infections . There was an increase in SM B cells when the homologous infection was resolving at the post-primary time point ( Fig 7A and 7B ) . However , unlike relapses , there was not a significant increase in USM B cells , and DN B cells significantly decreased during the homologous challenges ( Fig 7B ) . The frequency of Ki67+ SM , USM , and DN B cells increased during the homologous reinfections like in the relapses although the USM and DN B cells did not increase in number ( Fig 7C ) . Notably , only IgG+ SM B cells increased during the homologous reinfections whereas both IgG+ and IgM+ SM increased during relapses ( Fig 7D ) . Importantly , USM , SM , and DN B cell numbers were inversely correlated with parasitemia during the homologous infections ( Figs 7E , 7F and S10 ) . Although total IgM was unchanged during the homologous reinfections , IgM recognizing both iRBCs and uRBCs was significantly increased as observed during relapses , albeit at much lower levels than the initial primary infections ( Figs 8A , S11A and S12A ) . Consistent with relapses , total IgG and iRBC-specific IgG were also increased during homologous reinfections ( Figs 8B and S11B ) . Again , the IgG subclass was predominantly IgG1 ( S11C Fig ) . Similar to the relapses , iRBC-specific IgG was inversely correlated with parasitemia during the homologous reinfections , but iRBC-specific IgM was not ( Fig 8C and 8D ) . Notably , the IgG reactivity with iRBC versus uRBC lysates was significantly higher in the homologous reinfections , like the relapses ( S12B Fig ) . As with the relapses , the humoral response during the homologous reinfections was highly effective at opsonizing iRBCs ( Fig 8E ) . The increase in phagocytic activity was again correlated with iRBC-specific IgG ( Spearman’s ρ = 0 . 55 , p = 0 . 0005 ) , but unlike relapses , iRBC-specific IgM was also correlated ( Spearman’s ρ = 0 . 48 p = 0 . 02 ) with opsonic phagocytosis activity during homologous reinfections ( Fig 8F ) . Altogether , these data are consistent with B cell mediated immune responses conferring protection during relapses and homologous reinfections . Next , we questioned how the immunity during a relapse may affect the number and proportion of asexual and sexual parasite stages during relapses and , thus , enumerated the sexual and asexual parasites by microscopy during the primary and relapse infections . During the primary infections , the parasite differentials were performed from patency until the peak of parasitemia . Samples after the peak were excluded from the analysis since these were collected after sub-curative blood-stage drug treatment . The parasites were enumerated during relapses for all days showing patent parasitemia . As expected , the number of gametocytes were significantly reduced during the relapses given the significant reduction in parasitemia during relapses compared to the primary infections ( Fig 9A ) . While the absolute number of gametocytes decreased , the cumulative proportion of circulating iRBCs that developed into gametocytes was significantly increased in the relapses ( Fig 9B ) . In contrast , there was no significant difference between primary and relapse infections in the cumulative proportion of iRBCs containing ring , trophozoite , and schizont stages ( Fig 9B ) . Notably , the percentage of days out of the primary and relapse infections that gametocytes were observed in the blood was also similar ( Fig 9C ) . These data suggested that the immunity during relapses may disproportionately affect asexual stages as opposed to gametocytes . To validate the microscopy results , we examined gametocyte gene expression using parasite transcriptomes obtained from whole blood RNA-Seq data . Samples with less than 100 , 000 parasite reads were removed from the analysis; these included some relapse samples . The post-peak time points from the primary infection were also excluded since these were collected after sub-curative antimalarial treatment . We limited the analysis to P . cynomolgi genes that are homologous to P . vivax gametocyte genes that have been associated with P . vivax transmission in vivo [38–40] . In concordance with the microscopy data , P . cynomolgi homologues of the mature gametocyte genes pvs25 , pvs28 , and pvlap5 had similar gene expression across the primary infections and relapses ( Fig 9D ) . Overall , these results demonstrate that despite the development of effective B cell immunity and reduction of parasitemia during relapses , the relapses maintained detectable levels of gametocytes . In this study , single , sporozoite-initiated infections with P . cynomolgi in a cohort of rhesus monkeys resulted in the establishment of immunity that was capable of suppressing parasitemia during relapses or homologous reinfections initiated 60 days after radical cure . Whole blood RNA-Seq analysis showed that the host response during relapses and homologous reinfections was associated with distinct changes in the host transcriptome related to B cells . This finding was corroborated by flow cytometry and antibody ELISAs demonstrating that class-switched memory B cells rapidly responded during relapses and homologous reinfections along with a concomitant increase in anti-iRBC IgG . Collectively , these data demonstrate that protective , but non-sterilizing , humoral immunity can form after a single P . cynomolgi infection and may be key for preventing disease during relapses . Similarly , studies have found P . vivax-specific B cells can persist for years after an initial infection , and investigations using the P . chabaudi mouse model of malaria have also confirmed that humoral immunity can form and protect against subsequent challenges [41–44] . Therefore , we speculate that additional factors such as the genetic diversity of local P . vivax populations may be responsible for circumventing immunity , leading to symptomatic relapse infections . This seems likely when considering the high genetic diversity of P . vivax relapse infections in endemic areas [45–48] . Additional factors that may contribute to the lack of or subversion of immunity during a relapse include the age of first exposure or the presence of co-infecting pathogens . Antibody responses during Plasmodium infection have been studied for decades , yet the roles of antibody isotypes other than IgG are currently under investigation [49 , 50] . The inverse association of P . cynomolgi-specific IgM antibodies with relapse parasitemia indicate these antibodies may be involved in suppressing parasitemia and preventing the development of disease . USM B cells can differentiate to produce IgM in secondary responses , and these cells expanded during relapses and were inversely correlated with parasitemia during relapses and homologous reinfections [51 , 52] . IgM+ SM B cells also expanded during relapses . Together , our data and recent evidence from rodent malaria models and human samples support a role for IgM+ memory B cells in anti-Plasmodium immunity [53] . Future studies should aim to evaluate neutralizing IgM responses to identify the origin of the B cell subsets that are responsible for their production . Such experiments are needed to delineate if anti-parasite IgM antibodies arise from the memory B cell compartment during recall responses or if they originate from naïve B cells that are stimulated to differentiate and secrete IgM during each blood-stage infection . Identification of the B cell compartment where protective antibodies originate and persist is needed to understand naturally acquired immunity against relapsing malaria parasites and may help to advance the development of a P . vivax vaccine . While IgM may play a role in neutralizing blood-stage parasites during relapses , parasite-specific IgG1 was the predominant isotype produced during both relapses and homologous reinfections . Typically IgG3 rather than IgG1 has been reported as being the dominant subclass in human malaria [54 , 55] . This discrepancy is likely due to differences between the NHP and human immune systems whereby IgG1 mediates the majority of effector functions in NHPs compared to humans where IgG1 and IgG3 contribute similarly [56] . In the data reported here , IgG levels strongly correlated with opsonic phagocytosis activity of P . cynomolgi iRBCs , an important mechanism of peripheral parasite control during blood stage infection . IgG+ SM B cells were the most significantly expanded memory subset during relapses and homologous reinfections and were inversely correlated with parasitemia in both cases , showing the potential importance of the SM B cell compartment during P . cynomolgi relapses and homologous reinfections . This should be taken into consideration with the ‘anti-relapse’ vaccine strategies currently being considered [57] . Despite the beneficial roles of IgG and IgM during P . cynomolgi infection , these antibodies may also contribute to pathogenesis . Removal of uRBCs by the immune system is a substantial contributor to the development of malarial anemia in humans and NHP models [36 , 58–61] . Malarial anemia has been associated with the production of anti-self antibodies that tag uRBCs for elimination in rodent models and in human P . falciparum and P . vivax infections [62–64] . In this study , IgM and IgG antibodies that recognized uRBC lysates were detected during the primary infections , relapses , and homologous reinfections . The peak of the anti-uRBC IgM response occurred when parasitemia plateaued , and these antibodies were inversely correlated with hemoglobin levels when anemia was observed . Together , these data are consistent with a role for anti-uRBC IgM and IgG antibodies in the development of malarial anemia in addition to parasite control . Production of anti-uRBC antibodies could be due to non-specific , polyclonal activation of B cells in response to inflammatory stimuli released by iRBCs . Alternatively , the production of these antibodies may be an adverse , yet necessary , component of the normal immune response against P . cynomolgi that provides benefits like parasite neutralization . Either way future studies should identify the origin and function of ‘anti-self’ antibodies produced during longitudinal Plasmodium infections in NHPs . Human studies with P . vivax have documented that gametocytes are present in symptomatic and asymptomatic relapse infections [7 , 23 , 24] . However , it has remained unclear how gametocytes are affected in the face of ongoing immune responses during a relapse [65] . Our study showed that P . cynomolgi gametocytes are substantially reduced during asymptomatic P . cynomolgi relapses , but the cumulative proportion of gametocytes increases relative to asexual stages of the parasite . The asexual stages predominated in circulation , and their relative proportions remained similar between the primary and relapse infections . These results argue that the reduction in the number of gametocytes is likely due to removal of asexual parasites , thereby , preventing their development into gametocytes , rather than anti-gametocyte immunity during relapses . In essence , our data are consistent with the host developing immunity to reduce parasitemia and prevent disease while the parasite manages to produce gametocytes that remain in circulation for ingestion by mosquitoes . This situation is advantageous for relapsing malaria parasites because the establishment of non-sterilizing immunity minimizes the chances of the host succumbing to infection while allowing for continued opportunities for transmission . Importantly , this scenario is fitting with the biology of P . vivax gametocytes since these are detectable soon after patency and , thus , could be transmitted before relapse parasitemia is substantially reduced [66] . Future studies should address whether the potent humoral immune response during P . cynomolgi relapses in rhesus macaques may also possess transmission-enhancing properties as shown previously in Toque monkeys ( Macaca sinica ) infected with P . cynomolgi [67] . While our study provides the most comprehensive analysis of P . cynomolgi relapses to date , it is not without limitation . Although our data strongly support the premise that humoral immunity is important in suppressing parasitemia during relapses to ameliorate disease , other cell types are likely involved . For example , future analysis of monocytes and T cells would be useful , particularly since the decreased inflammation observed during relapses and homologous reinfections could result in improved T cell help for B cells , thereby , increasing the effectiveness of the humoral immune response in subsequent exposures . Nonetheless , our study demonstrates the importance of humoral immunity because the increase in IgG during relapses and homologous reinfections occurs nearly twice as fast as in primary infections . Also , the lower increase in the IgM response during relapses and reinfections compared to the primary infections is consistent with a strong secondary response . Second , this study was not designed to test the infectiousness of gametocytes to mosquitoes . However , this would be an important addition to P . cynomolgi relapse investigations based on our results . Lastly , although P . cynomolgi infections of rhesus macaques are a valuable experimental surrogate for human P . vivax infections , there are differences that may influence the development of immunity . For example , P . cynomolgi parasitemias in rhesus are typically higher than P . vivax parasitemias in humans , which could lead to establishment of durable immunity faster in the rhesus macaque—P . cynomolgi infection model . On the other hand , experimental infections with P . vivax in neurosyphalitic patients have demonstrated appreciable homologous immunity months to years after one infection [68–70] . The results from those studies argue that the data presented here have a high degree of relevance to P . vivax infections in humans . In conclusion , our studies with P . cynomolgi in rhesus macaques and studies on human P . vivax infections collectively provide strong evidence that relapses and homologous reinfections do not necessarily result in clinically detectable disease [41 , 68–70] . Instead , it is becoming clear that relapses and potential reinfections with the same parasite variant can be clinically silent , and we have shown that this is , at least in part , due to potent humoral immunity that forms after an initial infection . This is highly significant considering that we have shown that clinically silent P . cynomolgi relapses continue to harbor gametocytes . If individuals in endemic communities have clinically silent relapses , they will not seek treatment . Meanwhile , they may serve as a source of gametocytes that may remain infectious to mosquitoes . The number of clinically silent relapse infections and their infectiousness to mosquitoes remains largely unknown and should be evaluated carefully in the future . As a next step on the path to eliminating P . vivax and other relapsing malaria parasites , empirical studies that identify the factors that influence relapse pathogenesis , immunity , and infectiousness to mosquitoes are needed , and the P . cynomolgi-macaque models can be used for investigations in each of these areas . Nonhuman primate cohort infections were performed at the Yerkes National Primate Research Center ( YNPRC ) at Emory University , an Association for Assessment and Accreditation of Laboratory Animal Care ( AAALAC ) international-certified institution . Freshly isolated , salivary gland sporozoites were generated for each infection using additional rhesus monkeys at the Centers for Disease Control and Prevention ( CDC ) . Rhesus monkeys utilized for experiments were of Indian origin , male , 7–13 kg , and 5–6 years of age . All male animals were used for experiments to eliminate the female menstrual cycle as a contributor to the development of anemia . All procedures including blood collections , infections with malaria parasites , clinical interventions , etc . were reviewed and approved by Emory University’s and/or CDC’s Institutional Animal Care and Use Committees . During the experimental procedures at YNPRC , the animals were housed socially in pairs in compliance with Animal Welfare Act regulations as well as the Guide for the Care and Use of Laboratory Animals . Plasmodium cynomolgi M/B strain parasites were used for all experiments as previously described [12] . Two thousand freshly isolated , salivary gland sporozoites were generated , isolated , and administered intravenously as previously described to initiate the initial infections with relapses and homologous reinfections [12] . Subcurative antimalarial treatments consisted of a single dose of artemether at 1 mg/kg administered intramuscularly ( IM ) . Curative blood-stage treatments consisted of a 7 day regimen of artemether administered IM with the first dose at 4 mg/kg and subsequent doses at 2 mg/kg . Radical cure consisted of a combination treatment with artemether and primaquine . Artemether was first administered IM for 8 days using an initial dose of 4 mg/kg followed by subsequent doses at 2 mg/kg . At the conclusion of the artemether treatment , primaquine was administered orally in peanut butter at 2 mg/kg for 7 days . Blood was collected in EDTA at pre-defined time points as indicated in the experimental schematics in Figs 1A and 5A . Parasitemia and hematological parameters were evaluated daily by light microscopy and complete blood counts ( CBC ) analysis , respectively . For the daily parasitemia and CBC assays , blood was collected into a pediatric capillary tube using a standardized ear-prick procedure as previously described [12] . Blood specimens utilized for transcriptomic analysis were bled directly into Tempus tubes according to the manufacturer’s suggested protocol . Plasma was collected from each time point prior to isolation of PBMCs for flow cytometry analysis as described below . Bone marrow samples were not utilized for the experiments presented here . Daily parasitemia was determined as reported in Joyner et al . 2016 [12] . Briefly , thick and thin blood film preparations from capillary or venous blood were prepared and allowed to dry . Thin films were fixed with 100% methanol , and thick films left unfixed . The thick and thin films were then stained using a Wright’s-Gurr stain . For thick film preparations , parasites were enumerated by determining the number of parasites that were observed within 500–2000 white blood cells ( WBCs ) depending on the parasite density . The number of parasites per the number of WBCs was then calculated and multiplied with the leukocyte count as determined by the CBC to yield parasites per microliter . For days where parasitemia was too high to enumerate using thick blood films , the thin blood film was used; this was typically when parasitemia was greater than 1% . The number of parasites out of 1000–2000 RBCs were determined and the percent parasitemia calculated by dividing the number of parasites counted by the number of total RBCs counted and multiplying by 100 . The percentage of infected RBCs was then multiplied against the RBC concentration as determined by the CBC analysis to determine parasites/μl . Parasitemia was determined by two expert microscopists independently through the course of each infection . If discrepancies were observed between the two readers , a third , independent microscopist counted the slides . The two most similar values were then averaged to determine the parasitemia at any point during infection . The parasite stages present at the selected times during the infections were determined by thick or thin film microscopy by counting 10 to 100 parasites and noting their stage . Each slide was examined for at least 15 minutes before stopping . The proportion of rings , trophozoites , schizonts , and gametocytes , were then calculated and used to determine the frequency of each parasite-stage per microliter of blood using the number of parasites per microliter as determined above or as proportions ( % ) for area under the curve ( AUC ) analysis as described below . Area under the curve ( AUC ) for parasitemia ( parasites/μl ) and proportions ( % ) of parasite stages during the primary , relapse , and/or homologous reinfections were calculated using Riemann sums by determining the trapezoidal area between each data point during the indicated time periods . The formula used to calculate AUC was as follows: ( DataPoint1+DataPoint22 ) ×Δtime . Complete blood counts ( CBCs ) were performed prior to infections and daily after inoculation using capillary and/or venous blood . If values from the CBC were considered abnormal ( e . g . low platelet counts or observation of nucleated RBCs ) , the values obtained by the hematology analyzer were either confirmed or adjusted based on a manual differential or manual platelet count . For manual differentials , the phenotype ( e . g . monocyte , neutrophil , nucleated RBC , etc . ) was determined , and the percentage of each subset calculated . If there was a discrepancy with the CBC based on the differential , the percentage of monocytes , lymphocytes , and granulocytes was adjusted to ensure accuracy . If nucleated RBCs were present , the number of nucleated RBCs was determined and subtracted from the leukocyte count and added to the RBC count . Rectal temperatures were also obtained when animals were sedated for sample collections . Notably , two pre-infection values were collected prior to the initial infections to ensure accurate naive measurements were obtained since abnormal values may occur before an NHP becomes used to daily interaction . These values were averaged to obtain the malaria naïve values used for the analysis of the clinical data . A custom , nonhuman primate multiplex cytokine assay was designed and purchased from eBioscience/Affymetrix , which is now a part of Thermofisher . These kits were performed according to the manufacturer’s suggested protocol except for one modification . Instead of diluting plasma 1:1 with sample dilution buffer , the samples were not diluted prior to running the assay . This was altered after initial experiments demonstrated that many analytes were not within the dynamic range of the standard curves if additional dilutions were performed . Samples were fully randomized prior to performing the multiplex kit to minimize plate- and well-specific effects . All multiplex data was analyzed using the ProcartaPlex Analyst software available through Thermofisher . Concentrations of cytokines in the plasma were determined and used for downstream analyses . Total RNA was extracted using the Tempus RNA isolation kit ( Fisher Scientific; Cat#:4380204 ) . Globin transcripts were depleted using GLOBINclear Human Kit ( Fisher Scientific; Cat#:AM1980 ) according to the manufacturer's instructions . Libraries were prepared using the Illumina TruSeq mRNA stranded kit ( Illumina Inc . ; Cat#:20020595 ) as per manufacturer’s instructions . 1 ug of Globin depleted RNA was used for library preparation . ERCC ( Invitrogen; Cat#:4456740 ) synthetic spike-in 1 or 2 was added to each Globin depleted RNA sample . The TruSeq method ( high-throughput protocol ) employs two rounds of poly-A based mRNA enrichment using oligo-dT magnetic beads followed by mRNA fragmentation ( 120–200 bp ) using cations at high temperature . First and second strand cDNA synthesis was performed followed by end repair of the blunt cDNA ends . One single “A” base was added at the 3’ end of the cDNA followed by ligation of barcoded adapter unique to each sample . The adapter-ligated libraries were then enriched using PCR amplification . The amplified library was validated using a DNA tape on the Agilent 4200 TapeStation and quantified using fluorescence based method . The libraries were normalized and pooled and clustered on the HiSeq3000/4000 Paired-end ( PE ) flowcell on the Illumina cBot . The clustered PE flowcell was then sequenced on the Illumina HiSeq3000 system in a PE 101 cycle format . Each sample was sequenced to a target depth of 100 million pairs ( 50 million unique fragments ) with exception of Time point 2 samples that were sequenced to 200 million pairs ( 100 million unique fragments ) . Raw FASTQ files from the RNA-Seq experiments of all animals at all time points were aligned to the P . cynomolgi [21] and M . mulatta ( version 7 . 8 ) [71] reference genomes using the Spliced Transcripts Alignment to a Reference tool ( STAR , version 2 . 4 . 1c ) . The aligned features were further quantified and mapped using the High-Throughput Sequencing tool version 0 . 6 . 1p1[72] using only the P . cynomolgi reference to select parasite-specific transcripts . All sequencing and transcript mapping results were deposited to the NCBI GEO and SRA databases under the accessions GSE104223 ( E23 ) and GSE104101 ( E24 ) . Raw count data from MaHPIC Experiments 23 , 24 , and 25 were all library size normalized using the ‘DESeq2’ package for R [73] . Prior to normalization for library size , genes of extremely low read count ( <10 reads across all samples ) were filtered . RNA-Sequencing data taken during initial infections , homologous reinfections , and heterologous reinfections ( not presented here ) were normalized together . Data structure was then examined with principal component analysis . Individual animal effects were removed using Supervised Normalization of Microarrays with the ‘SNM’ package for R [32] . LIMMA was then used to assess gene expression changes during each infection and between infection stages [74] . Fraction of gene expression variance explained by unsupervised Ward’s hierarchical cluster analysis was determined by finding the ratio of between-cluster variance ( B ) to total variance ( T ) , which is in turn the sum of between-cluster and within-cluster variance ( W ) . Where j = 1 , 2 , … are the different clusters , i = 1 , 2 , … are the different genes measured in each sample , k = 1 , 2 , … are the different samples , nj is the number of samples in cluster j , gj¯ ( i ) is the average value of gene i in cluster j , g¯ ( i ) is the average value of gene i across all samples , and gk ( i ) is the value of gene i in sample k . For the within-cluster variance equation , the cluster j that is used for each step of the summation is the cluster to which sample k belongs . Only samples in which parasites were detected by microscopy and at least 100 , 000 total reads ( corresponding to at least 90 parasite/ μL ) were analyzed . For the comparison of initial infections versus relapses , the pre-peak and peak infection stages were used for comparison with the relapse time points . The early relapse or peak relapse points were used to represent relapses . For animals in which both early and peak relapse samples had sufficient parasitemia and parasite reads the earlier sample was selected to represent relapse . Notably , if there was not an early relapse infection stage , the peak relapse point was used for analysis . ROh14 did not have a relapse and thus was not analyzed . All samples were library size normalized together and log2-transformed using DESeq2 . Changes in gene expression were then assessed by using a linear mixed effect model with a Tukey-Kramer HSD post hoc analysis . P-values < 0 . 05 were considered statistically significant . Plasma was isolated prior to performing the PBMC isolation by centrifuging the blood samples at 400 × g followed by pipetting off the plasma . After removing the plasma , the blood pellet was resuspended in two times the original volume of blood that was received . This modification of the procedure did not appear to alter the viability or yield of PBMCs . After this step , the manufacturer’s suggested protocol was followed . After each isolation , each monkey’s PBMCs were washed two times in sterile PBS followed by enumeration on a Countess II fluorescent cell counter . The viability of the PBMCs was simultaneously assessed by Trypan Blue exclusion assay . PBMC viability was always ≥ 90% . 5×105–2×106 PBMCs were aliquoted into flow cytometry tubes for staining with fluorescently conjugated antibodies . The variation in number of PBMCs used for each staining procedure was due to leukopenia that developed during the acute , symptomatic infections . After aliquoting into individual FACS tubes , cells were washed once more in PBS prior to re-suspending in antibody cocktails comprised of the antibodies indicated in S6 Table for the initial infections with relapses and S7 Table for the homologous reinfections . Notably , some markers listed are not presented in the manuscript , but are provided to convey the complete panel configurations used in each experiment . All staining procedures were two-step . For surface IgG staining , the IgG was prepared in a separate cocktail and added first , followed by a 30-minute incubation , washing PBS by centrifugation at 400 × g , and then resuspending in a cocktail that contained the other antibodies in the panels . For intracellular staining , cells were initially surface-stained with the cocktail followed by incubation in eBioscience FoxP3 fix perm buffer ( Thermofisher ) overnight at 4°C for intracellular markers . After fixing overnight , the cells were washed according to the manufacturer’s procedure and then incubated for 45 minutes at 4°C with antibodies against intracellular markers . The cells were then washed twice in the fix/perm buffer provided by the kit and resuspended in 100–200 μl of PBS depending upon cell yield . All samples were acquired on an LSR-II flow cytometer using standardized acquisition templates and rainbow calibration particles for voltage control . Compensation controls were run at each acquisition . Data were initially compensated in FlowJo version 10 . 1 followed by exporting to Cytobank for gating . Cell population level statistics were then exported from Cytobank for further analysis . Absolute numbers for each B cell subset was determined by calculating the percentage of each subset out of the mononuclear cells in the sample and multiplying the percentage with the mononuclear cells/μl value obtained from the CBC at each time point . The mononuclear cells/μl was obtained by adding the lymphocyte/μl value to the monocyte/μl value . If two values for absolute numbers were obtained due to a population being present in both panels ( e . g . switched memory ) , the values were treated as technical duplicates and averaged to obtain the final value used for each analysis . Corning high-binding microtiter plates were coated with Anti-Monkey IgG+IgA+IgM ( Rockland Immunochemicals ) or Anti-Monkey IgM ( Life Diagnostics ) diluted in ELISA coating buffer ( Abcam ) to 0 . 6 ug/ml and 5 ug/ml , respectively . The plate was incubated overnight at 4°C followed by washing four times with PBS containing 0 . 05% Tween-20 ( PBS-T ) . After the final wash , the plate was blotted dry and blocked using serum-free Sea Block ( Abcam ) for two hours at RT followed by four washes in PBS-T . Plasma samples from the different infection points were diluted 1:100 , 000 for total IgG or 1:10 , 000 for total IgM in 10–33% serum-free Sea Block and then added to each well . The plate was then incubated at RT for 2 h followed by washing four times with PBS-T . After blotting dry , HRP-conjugated anti-IgG ( Jackson Immunoresearch ) or HRP-conjugated anti-IgM ( Jackson Immunoresearch ) diluted 1:30 , 000 or 1:20 , 000 in 10–33% Sea Block in PBS , respectively , were added to each well and incubated for 1 h at RT in the dark . After incubating , the plate was washed four times with PBS-T and 100 μl of High Sensitivity TMB Substrate ( Abcam ) was added to each well and allowed to develop for 3–5 minutes . One hundred microliters of Stop Solution ( Abcam ) was added to stop the reaction . The absorbance at 450 nm was then measured , and total IgM or total IgG antibody concentrations were calculated based on a 4-PL standard curve using purified IgM calibrators from Abcam’s Monkey Total IgM ELISA kit and using purified IgG Monkey Calibrators from Rockland Immunochemicals . Concentrations of total IgG and IgM were used for downstream analyses . Rhesus macaques were inoculated with cryopreserved , blood-stage P . cynomolgi B/M strain parasites to generate schizonts for lysate preps described below . Briefly , a vial of cryopreserved , blood-stage parasites were removed from the liquid nitrogen and quickly thawed in a 37°C water bath . After thawing , saline solutions of different concentrations were added drop-wise to slowly change the osmotic pressure while preventing RBC lysis . The number of ring-stage parasites were then enumerated using light microscopy as described above and inoculated intravenously into a rhesus macaque . The infections were followed daily for each monkey until parasitemia reached 3–10% ring-stage parasites . At this time , blood containing predominantly rings was collected in sodium heparin , washed , and depleted of leukocytes and platelets by passing over a glass bead column and through a Plasmodipur filter . The parasites were then matured ex vivo to 3–8 nucleated schizonts under blood-gas conditions ( 5%:5%:90%;O2:CO2:N2 ) in RPMI supplemented with L-glutamine , supplemented with 0 . 25% sodium bicarbonate , 50 μg/ml hypoxanthine , 7 . 2 mg/ml HEPES , 2 mg/ml glucose , and 10–20% Human AB+ serum . When mature , the schizonts were isolated by a 1 . 093 g/ml Percoll density gradient . The parasite layer was then isolated and washed 4 times in sterile RPMI , aliquoted , and stored in vapor phase liquid nitrogen until needed . Aliquots of parasite or uninfected RBC pellets were removed from liquid nitrogen storage , thawed quickly in a 37°C water bath and placed back into the liquid nitrogen tank for ten minutes . This procedure was repeated three more times . After the final thaw , 1 volume of PBS was added followed by vigorous vortexing for 1–2 minutes . The aliquot was then centrifuged at 3 , 000 × g for 10 minutes at 4°C . The supernatant was then removed and placed into another sterile tube and the pellet discarded . This process was repeated three more times . After the final centrifugation , the protein concentration was determined using a Pierce BCA assay according to the manufacturer’s protocols . The lysates were then diluted to optimal concentrations for ELISAs in PBS , aliquoted , and stored at -80°C until needed . Corning high-binding microtiter plates were coated with schizont lysate or uninfected RBC lysate diluted in ELISA coating buffer ( Abcam ) to 5 ug/ml . The plate was incubated overnight at 4°C followed by washing four times with PBS containing 0 . 05% Tween-20 ( PBS-T ) . After the final wash , the plate was blotted dry and blocked using serum-free Sea Block ( Abcam ) for two hours at RT followed by four washes in PBS-T . Plasma samples from the different infection points were diluted 1:100 in 10–33% serum-free Sea Block and then added to each well . The plate was then incubated at RT for 2 h followed by washing four times with PBS-T . After blotting dry , horseradish-peroxidase ( HRP ) conjugated anti-IgG ( Jackson Immunoresearch ) or HRP-conjugated anti-IgM ( Jackson Immunoresearch ) diluted 1:30 , 000 or 1:20 , 000 in 10–33% Sea Block in PBS , respectively , were added to each well and incubated for 1 h at RT in the dark . After incubating , the plate was washed four times with PBS-T and 100 μl of High Sensitivity TMB Substrate ( Abcam ) was added to each well and allowed to develop for 3–5 minutes . One hundred microliters of Stop Solution ( Abcam ) was then added to stop the reaction . The absorbance at 450 nm was then measured , and the OD450 of iRBC and uRBC-specific IgG and IgM were used for downstream analyses . Corning high-binding microtiter plates were coated with schizont lysate diluted in ELISA coating buffer ( Abcam ) to 5 ug/ml . As a positive control , duplicate wells were coated with recombinant expressed rhesus IgG1 , IgG2 , or IgG3 ( NHP Reagent Resource ) diluted to 1 ug/ml in ELISA coating buffer . The plate was incubated overnight at 4°C followed by washing four times with PBS containing 0 . 05% Tween-20 ( PBS-T ) . After the final wash , the plate was blotted dry and blocked using serum-free Sea Block ( Abcam ) for two hours at RT followed by four washes in PBS-T . Plasma samples from the different infection points were diluted 1:100 in 10% serum-free Sea Block and then added to each well . The plate was then incubated at RT for 2 h followed by washing four times with PBS-T . After blotting dry , mouse anti-rhesus IgG1 , IgG2 , or IgG3 ( NHP Reagent Resource ) diluted 1:10 , 000 , 1:1 , 000 , and 1:10 , 000 in 10% Sea Block in PBS , respectively , was added to each well and incubated at RT for 1 h . Following four washes in PBS-T and blotting dry , HRP-conjugated anti-mouse IgG ( Jackson Immunoresearch ) diluted 1:10 , 000 in 10% Sea Block in PBS was added to each well and incubated for 1 h at RT in the dark . After incubating , the plate was washed four times with PBS-T and 100 μl of High Sensitivity TMB Substrate ( Abcam ) was added to each well and allowed to develop for 30 minutes . One hundred microliters of Stop Solution ( Abcam ) was then added to stop the reaction . The absorbance at 450 nm was then measured , and optical densities ( ODs ) were used for downstream analyses . We adapted a previously established phagocytosis assay for P . cynomolgi [75] . Briefly , the THP-1 monocytic cell line was obtained from ATCC and maintained in vented 75cm2 culture flasks at 10% CO2 in RPMI-1640 supplemented with 10% fetal bovine serum , 2mM L-glutamine , 10mM HEPES , 1mM sodium pyruvate , 4500 mg/L glucose , and 1500 mg/L sodium bicarbonate . The cells were maintained at a density of 1 × 105 cells/ml of culture and were not allowed to exceed 1 × 106 cells/ml . Plasmodium cynomolgi strain M/B were thawed , matured in vitro to schizonts , and isolated as described above . Purified schizonts were incubated with 5 ug/ml dihydroethidium ( DHE ) for 20 min at 37°C , followed by 3 washes in THP-1 media before use in the assay . After labeling with DHE , the schizonts were opsonized in heat-inactivated plasma from different specimen collections for 45 minutes at RT in the dark . While the parasites were opsonizing , THP-1 cells were harvested and an aliquot of THP-1 cells was incubated with 5 uM Cytochalasin D for 1 h at 37°C to serve as a negative control . THP-1 cells were then added to each well to a final Effector Target ratio of 1:20 and incubated at 37°C for 3 h . The cells were then transferred to FACS tubes , washed twice with THP-1 media , and then lysed with ACK lysing solution for 10 min at RT in the dark . Cells were then re-suspended in PBS and acquired immediately on a BD LSR-II using a standardized acquisition template . All statistical analyses were performed using a linear mixed-effect model with Tukey-Kramer HSD post-hoc analysis . For the statistical model , each animal was treated as a random effect with time points as fixed effects . All data were transformed as necessary to ensure the best model fit , and the best model fits were typically obtained with a log10 , log2 , or arcsin transformation . All data went through rigorous validation protocols and are publicly deposited in public repositories . All clinical data associated with the experiment have been publicly released on PlasmoDB http://plasmodb . org/plasmo/mahpic . jsp ( see http://plasmodb . org/common/downloads/MaHPIC/Experiment_23/ and http://plasmodb . org/common/downloads/MaHPIC/Experiment_24/ for the datasets in this manuscript ) . All RNASeq results have been publicly released as described in the Materials and Methods . Flow cytometry , multiplex cytokine assays , and ELISA are publicly available at ImmPort as part of study SDY1409 .	Plasmodium vivax contributes significantly to global malaria morbidity and remains a major obstacle for malaria elimination due to its ability to form dormant stages in the liver . These forms can become activated to cause relapsing blood-stage infections . Relapses remain poorly understood because it is difficult to verify whether P . vivax blood-stage infections in patients are due to new infections or relapses in most cases . Here , we use a nonhuman primate model of Plasmodium vivax malaria in concert with state-of-the-art immunological and molecular techniques to assess pathogenesis , host responses , and circulating gametocyte levels during relapses . We found that relapses were clinically silent compared to initial infections , and they were associated with a robust memory B cell response . This response resulted in the production of antibodies that were able to mediate clearance of asexual parasites . Despite this rapid immune protection , the sexual-stage gametocytes continued to circulate . Our study provides mechanistic insights into the host-parasite interface during Plasmodium relapse infections and demonstrates that clinically silent relapses can harbor gametocytes that may be infectious to mosquitoes .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "blood", "cells", "medicine", "and", "health", "sciences", "parasite", "groups", "immune", "cells", "plasmodium", "gametocytes", "immunology", "tropical", "diseases", "vertebrates", "parasitic", "diseases", "parasitic", "protozoans", "animals", "parasitology", "mammals", ...	2019	Humoral immunity prevents clinical malaria during Plasmodium relapses without eliminating gametocytes
Cells and tissues are exposed to stress from numerous sources . Senescence is a protective mechanism that prevents malignant tissue changes and constitutes a fundamental mechanism of aging . It can be accompanied by a senescence associated secretory phenotype ( SASP ) that causes chronic inflammation . We present a Boolean network model-based gene regulatory network of the SASP , incorporating published gene interaction data . The simulation results describe current biological knowledge . The model predicts different in-silico knockouts that prevent key SASP-mediators , IL-6 and IL-8 , from getting activated upon DNA damage . The NF-κB Essential Modulator ( NEMO ) was the most promising in-silico knockout candidate and we were able to show its importance in the inhibition of IL-6 and IL-8 following DNA-damage in murine dermal fibroblasts in-vitro . We strengthen the speculated regulator function of the NF-κB signaling pathway in the onset and maintenance of the SASP using in-silico and in-vitro approaches . We were able to mechanistically show , that DNA damage mediated SASP triggering of IL-6 and IL-8 is mainly relayed through NF-κB , giving access to possible therapy targets for SASP-accompanied diseases . Age-related diseases can be held accountable for the major part of morbidity and mortality in an ageing population . Additionally they cause a large proportion of yearly health costs [1] . Cellular senescence is one of the most prominent events that is likely to contribute to ageing . It refers to the irreversible cell cycle arrest that is essential when cells encounter detrimental changes . Once in permanent arrest , these cells are normally cleared by the immune system before they are able to do any harm to the organism [2] . However , some of these cells persist and develop a secretory phenotype releasing a variety of factors among which pro-inflammatory cytokines , chemokines and extracellular matrix degrading proteases are included . Together these shape the senescent-associated secretory phenotype or SASP [3–5] . While the SASP can cause chronic inflammation in tissue , it can also reinforce senescence in autocrine and paracrine manner [6 , 7] . This feature of the SASP not only keeps senescent cells in their growth arrested states but it promotes senescence spreading to healthy bystander cells . Therefore , the SASP contributes to the accumulation of senescent cells during ageing , but also supports the emergence of age-related chronic diseases and tissue dysfunctions by elevating inflammatory processes [6 , 8] . Major soluble factors that facilitate this bystander-infection of healthy cells are IL-6 and IL-8 . Both have been shown to be important in the maintenance and spreading of oncogene- and DNA-damage-induced senescence [3] . Also , both have been shown to be highly overexpressed by senescent cells and are known to locally and systemically play important roles in the regulations of a variety of processes in the aging body [3 , 4 , 9] . IL-6 , in fact , most likely contributes to organ dysfunction during aging thus promoting frailty [8] . To allow for a deeper understanding of the SASP and the dynamics of its complex interactions a computational model of the Regulatory Network ( RN ) [10] and subsequent simulations can be insightful . RNs can be described by different mathematical models such as differential equations , Bayesian networks , and Boolean networks among others [11] . The Boolean network model [12 , 13] , as opposed to other model approaches , can be based on qualitative knowledge only . In gene-gene interaction , for example , the expression of a gene is regulated by transcription factors binding to its regulatory regions . The activation of a gene follows a switch-like behavior depending on the concentration of its transcription factors . This behavior allows common approximation of the possible states of a gene to be active or inactive [14 , 15] . Ultimately , this can be encoded as Boolean logical values: true ( “1” ) or false ( “0” ) . The interactions between genes , e . g . whether a factor acts as an activator , repressor or both can be described by functions . These Boolean functions are the basis to simulate dynamic behavior , i . e . changes over time . As every regulatory factor has two possible states ( active or inactive ) in a Boolean network model , 2x possible state combinations ( i . e . gene activation patterns ) exist for x genes . For any activation pattern , iterative updates of genes in the network through consecutive application of the Boolean rules eventually lead to sequences of gene activation patterns that are time-invariant , called attractors . These attractors can correspond to observed expression profiles of biological phenotypes or can be used to create hypotheses to further evaluate in wet-lab experiments [16 , 17] . Different update strategies for the Boolean functions exist . Using a synchronous update strategy means applying all Boolean functions simultaneously , also assuming that regulatory factors interact independently of one another and that their interaction has a similar time scale resolution . Relaxing these assumptions leads to the concept of asynchronous updates where each Boolean function of is updated separately one at a time in any order . This allows a more direct modelling of different time scales . The asynchronous update strategy also usually generates trajectories that are different from those of synchronous Boolean networks . The state transition graph of an asynchronous Boolean network becomes a Markov chain which requires the additional definition of transition probabilities in each node of the state graph . Interestingly , point attractors ( those with one state ) in asynchronous Boolean networks are the same as those in synchronous Boolean networks . However , these networks can also show loose/complex attractors [18] which are part of active research [19 , 20] . Another extension of Boolean networks are probabilistic Boolean networks , which may define more than one Boolean function for regulatory factors where each function has a specific probability to be chosen for update . Although this concept may closer represent a biological system , it again requires parameter estimation for the probabilities . However , estimation of the probabilities naturally demands large amounts of interaction specific data which is , for larger networks , neither economically , nor experimentally viable . In our case , we decided to focus on synchronous Boolean networks , partly due to their proven usability , and their ability to reveal key dynamical patterns of the modelled system . However , to strengthen our models’ hypothesis , we additionally performed in-silico experiments with an asynchronous update scheme ( S1 Text ) . Synchronous Boolean networks have been used to model the oncogenic pathways in neuroblastoma [21] , the hrp regulon of Pseudomonas syringae [22] , the blood development from mesoderm to blood [23] , the determination of the first or second heart field identity [24] as well as for the modeling of the Wnt pathway [25] . The qualitative knowledge base that is necessary to reconstruct [26] a Boolean network model consists mostly of reports on specific interactions that describe local regulation of genes or proteins . Boolean network models utilize this knowledge about local regulations to reconstruct a first global mechanistic model of SASP . In summary , such a model allows to generate hypotheses about regulatory influences on different local interactions . These interactions , in turn , can be tested in wet-lab in order to validate the generated hypothesis and assess the accuracy of the proposed model . Here , we present a regulatory Boolean network of the development and maintenance of senescence and the SASP incorporating published gene interaction data of SASP-associated signaling pathways like IL-1 , IL-6 , p53 and NF-κB . We simulated the model and retrieved steady states of pathway interactions between p53/p16INK4A steered senescence , IL-1/IL-6 driven inflammatory activity and the emergence and retention of the SASP through NF-κB and its targets . This Boolean network enables the highlighting of key players in these processes . Simulations of knock-out experiments within this model go in line with previously published data . The subsequent validation of generated in-silico results in-vitro was done in murine dermal fibroblasts ( MDF ) isolated from a murine NF-κB Essential Modulator ( NEMO ) -knockout system in which DNA damage was introduced . The NEMO knockout inhibits IL-6 and IL-8 homologue mRNA expression and protein secretion in MDFs after DNA damage in-vitro , possibly enabling at least a lowering of the contagiousness for neighboring cells and the pro-tumorigenic potential of the SASP . The model presented in this article allows a mechanistic view on interaction between the proinflammatory and DNA-damage signaling pathways and thereby helps to gain insights into the dynamics of the SASP . Furthermore , it enables to generate extensive hypotheses about possible knockout targets that can be experimentally tested and verified in-vitro . To the best of our knowledge , this report is the first one that combined in-silico simulation of the SASP with its laboratory based experimental validation . The reconstruction of a Boolean network model for SASP requires screening for many candidate interactions in published literature and data . Although the model , after reconstruction , may be reduced in the number of components [20 , 27 , 28] , it would potentially hide some of the interaction targets and regulatory factors with regard to the signaling cascade . The regulatory factors defined in this model are beneficial if one wants to extend the model and include additional related signaling pathways . The subsequent model must accurately correspond to the current understanding of the process at hand , i . e . , able to predict well-known phenotypes of SASP . Biological phenotypes represent a long-term behavior of a biological system based on interaction of regulatory factors . In the same sense , attractors are the long-term behavior of a Boolean network model based on the Boolean rules of modelled regulatory factors . Hence , there is a natural correspondence between biological phenotypes and attractors in the Boolean network . In the following , we use figures that depict the signaling cascade towards an attractor as well as the attractor itself . The interpretation of these attractors in the context of SASP further allows generation of hypotheses that can be tested in a biological system . The information for the reconstruction of these networks was collected from published data . An overview of the genes incorporated in this model and their interaction can be found in Fig 1 . The corresponding Boolean rules are listed in Table 1 . The network depicts processes following a cell cycle arrest inducing action , such as DNA damage and other cellular stresses . Here , we analyze SASP under strong DNA damage and do not distinguish between different levels of DNA damage . We first analyzed if our model can render steady states for cell cycle progression when there is no stress signal input . Our data show a normal cell cycle progression with active CDK2 and CDK4 , as well as phosphorylated Rb and hence an active E2F . No other signaling pathways that are implemented in this model were activated which can be seen as normal cell cycle progression ( Fig 2 ) . Upon the outside signal DNA damage , we observe first the activation of the DNA damage response with a subsequent activation of p53 and p16INK4A signaling , leading to a stop in cell cycle progression and at a later time point to permanent cell cycle arrest . Simultaneously NF-κB signaling gets activated by the DNA damage response through NEMO , giving rise to beneficial but also detrimental effects of NF-κB like the senescence associated secretory phenotype ( Fig 3 ) . After entering p53/p21 and p16INK4A mediated permanent cell cycle arrest upon DNA damage , the activation of NF-κB leads to an increase of IL-1 , IL-6 as well as IL-8 expression among others [29–33] . Our model shows the direct activation of these cytokines and chemokines by NF-κB after its activation through the DNA damage response and NEMO ( Fig 3 ) . The NF-κB pathway has been studied extensively and there are knockout mice available for all proteins of the pathway , however some of them are embryonically lethal due to the importance of NF-κB signaling in regulating development and apoptosis . We therefore focused on published in-vitro knockout and overexpression phenotypes . IL-6 and IL-8 are extremely important in maintaining and spreading the SASP in an autocrine as well as paracrine fashion . Hence , we followed the question what knockouts and/or overexpressions the Boolean network model suggests to inhibit the expression of IL-6 and IL-8 under the assumption of existing DNA damage . These simulations are included in S1 Text . RelA binds with p50 to form a transcriptionally active heterodimer ( called NFkB in this model ) . In its inactive state , it is bound with the inhibitor of kappa B ( IκB ) and resides in the cytoplasm . Upon NF-κB activation , the inhibitor is phosphorylated by the inhibitor of kappa B kinases ( IKK ) and degraded which releases the RelA/p50 heterodimer to translocate to the nucleus and regulate the transcription of target genes . To investigate the role of RelA on the expression of IL-8 , we set NFkB = 0 , simulating the ablation of the transcriptionally active heterodimer ( Fig 4 ) . The predictions of the model simulations are consistent with knock-out experiments where the absence of RelA caused a significant reduction in IL-8 production in human fibroblast ( IMR-90 ) [7] . We also simulated the overexpression of IκB by constantly activating IκB ( IkB = 1 ) and could show an effect comparable to the knock-out of RelA ( Fig 5 ) . In our model the overexpression of IκB leads to the inhibition of IL-8 and IL-6 expression which is in line with a previously published report , where the overexpression of a non-degradable IκBα completely abolishes IL-8 production , among other soluble factors , in human epithelial and cancer cell lines [34] . Another promising knockout described by our network is inhibitor of nuclear factor kappa-B kinase subunit gamma also known as NEMO , which is able to prevent IL-6 and IL-8 expression after DNA damage activated the DNA damage repair apparatus and cell cycle progression has been stopped in-silico ( Fig 6 ) . In studies with murine NEMO knockout models it has already been shown that murine embryonic fibroblasts ( MEFs ) isolated from these mice show reduced NF-κB activity and IL-6 secretion upon stimulation with typical NF-κB activators like IL-1 and TNF [35] . Apart from being important for the assembly of the IKK-complex , NEMO also acts as a shuttle relaying the ATM-mediated DNA damage apparatus to cellular response mechanisms . Upon DNA damage ATM can bind NEMO and trigger its translocation from the nucleus to the cytoplasm where it activates NF-κB signaling [36] . This in turn will help cells avoid clearance through apoptosis , increasing the number of long-term senescent cells in tissues and organs of the organism and might also increase and sustain the inflammatory potential of the SASP . In order to evaluate proposed knockouts NEMO was depleted from murine dermal fibroblasts ( MDFs ) using a NEMO-floxed mouse line . These MDFs were isolated from murine skin and subsequently transfected with a Cre-recombinase coding plasmid including a fluorescence reporter construct ( Fig 7 ) . To purify NEMO knockout MDFs , these cells were FACS sorted two days post-transfection ( S1A Fig ) . Successful NEMO knockout was assessed by PCR ( S1B Fig ) and western blot ( S1C Fig ) . To study the effect of DNA damage , overnight-starved MDFs were treated with 25 μM etoposide , an established DNA damage and senescence inducer , for 3 h followed by a 24 h incubation period [37] . Afterwards cell media supernatant was taken and total RNA was isolated . We first measured p21 mRNA expression as an indicator for DNA damage and cell cycle arrest . Without a significant reduction of cell viability ( Fig 8A ) , p21 mRNA expression was upregulated more than twofold in etoposide treated compared to untreated MDFs ( Fig 8B ) . NEMO is of high importance for DNA damage mediated nuclear translocation of the NF-κB signaling molecule p65 . As shown by immunofluorescence staining of untreated NEMO wildtype MDFs compared to etoposide treated wildtype and knockout MDFs , the translocation of p65 into the nucleus upon DNA damage is significantly increased in wildtype whereas it is brought down to the level of untreated wildtype MDFs when NEMO is knocked out ( Fig 8C ) . As we have observed the effect of a NEMO knockout on the nuclear translocation of p65 and thereby activation of NF-κB , we further explored the possible suppressive effect on IL-6 and IL-8 activation . To achieve this we isolated total RNA and analyzed the mRNA expression of IL-6 and the murine homologues of IL-8 CXCL1 ( KC ) , CXCL2 ( MIP-2 ) and CXCL5 ( LIX ) . Upon DNA damage , we observed a significant reduction in IL-6 mRNA expression with a strong downregulation in untreated knockout compared to untreated wildtype . An even stronger downregulation in etoposide treated NEMO knockout compared to wildtype MDFs was detected . Taken together a NEMO knockout could reduce DNA-damage mediated IL-6 mRNA expression by almost tenfold ( Fig 9A ) . Next , we measured the secretion of IL-6 . While there is nearly no secretion of IL-6 in untreated wildtype as well as knockout MDFs , a strong increase in IL-6 secretion occurred in etoposide treated wildtype MDFs , whereas the NEMO knockout MDFs only shows a small increase in secretion with a more than hundredfold reduction when compared with etoposide treated wildtype cells ( Fig 9B ) . We additionally analyzed the mRNA expression of three murine IL-8 homologues to assess the impact of a NEMO knockout on DNA damage mediated IL-8 expression . We found that all three chosen homologues were significantly downregulated in NEMO knockout MDFs compared to wildtype MDFs after DNA damage . The total expression of IL-8 homologues mRNA in NEMO knockout MDFs was reduced by at least fivefold when compared to treated wildtype MDFs ( Fig 9C ) . There is detectable secretion of IL-8 homologues in untreated wildtype and NEMO knockout MDFs , however the secretion strongly rose upon etoposide treated in wildtype cells whereas there is no detectable increase in the NEMO knockout MDFs . This effect was similarly found for the studied IL-8 homologues KC and MIP-2 ( Fig 9D ) . However , we did not find any significant alteration in the expression of two housekeeping genes , such as beta-actin and 18s rRNA in the NEMO knockout MDFs , compared with NEMO wildtype ( S2A Fig ) . In addition , we also did not observe any significant alteration in the expression of a wide array of genes that were predicted by Boolean network not to be changed after NEMO knockout ( S2B Fig ) . These data show the importance of NEMO and NF-κB signaling for the activation of IL-6 and IL-8 in the case of DNA damage . In early stages DNA damaged and cell cycle arrested MDFs most likely activate secretory SASP signaling through NF-κB rather than other stress pathways . In the model of DNA damage and proinflammatory signaling presented here we collected and combined previously published knowledge on major regulators of the SASP . Using this model , we identified attractors fitting cell cycle progression and cell cycle arrest as they physiologically occur . This suggests reliability of this model in terms of reproducibility of current biological knowledge . The network model allows us to time- and cost-effectively generate hypotheses and predict gene knockouts that may influence the outcome of the SASP in-vitro . In the process of modeling , we first created individual models of DNA damage and proinflammatory signaling . In a next step , we fused these two sub-networks to the model presented here . In S1 Text , we analyzed the impact of integrating both pathways in one Boolean network model . Our results indicate that there is not only an effect of DNA damage in the proinflammatory signaling but also vice versa . On one hand , we deduce a stabilization of the DNA damage response network as the integration of both sub-networks leads to a reduction of possible attractors ( 87 to 19 ) . On the other hand , the inner dynamics of each sub-network stay intact , showing biologically reproducible signaling cascades ( e . g . Fig 4 ) . In the simulation without DNA damage , only activation of cell cycle regulation genes that facilitate cell cycle progression were observed [38] . In contrast , when we entered DNA damage into the network , we detected early activation of the DNA damage response ( DDR ) followed by a p53/p21 mediated cell cycle arrest and at a later time point the activation of proinflammatory signaling through NF-κB [39 , 40] . We utilized the Boolean network to simulate knockout and overexpression states that have the power to inhibit both IL-6 and IL-8 activation , such as knockouts of ATM and RelA or the overexpression of IκBα , that have previously been published to decrease IL-8 or IL-6 expression and secretion in-vitro [7 , 9 , 34] . One of the most prominent knockout suggestions obtained was that of NEMO , which acts as an essential modulator of NF-κB signaling and is a major link between DDR and NF-κB signaling [41] . Therefore , it is a suitable target to prevent NF-κB activation , while maintaining the repair potential of the DDR . Taken together these in-silico data suggest NF-κB to be one of the major SASP activators in response to DNA damage activating all three mediators of proinflammatory signaling depicted in this network . For the sake of manageability , the model presented here was limited to a core set of pathways involved in senescence and the SASP . Of course , the value of the results could still be enriched by adding even more components and additional pathways , such as a more detail view on CEBP-signaling , growth factor signaling and the expansion of cell cycle related signaling . This would enable to simulate an even deeper level of signaling involved in the SASP . Another factor that was not viewed in this work is the influence of the intensity levels and timing of expression and stimuli on the outcome of the SASP . Physiologically occurring DNA damage , for example , is not an all or nothing event but rather comes in different levels and lengths of damage that can trigger a multitude of different reactions in the cell . In future works , it would be interesting to add these into the model . Such extension would allow simulations of the exact amount and timing of damage needed to trigger full-blown SASP rather than senescence . Furthermore , it would possibly reveal at which point the cell decides that it is beneficial to trigger SASP signaling in order to warn the system of the damage and initiate clearance as opposed to trying to repair itself . IL-6 and IL-8 reinforce senescence in an autocrine and paracrine way , concomitantly preventing senescent cells from exiting cell cycle arrest and forcing neighboring cells into senescence themselves [3 , 42] . Persistent DDR activity , that is also known to induce IL-6 and IL-8 secretion [9] , could be shown in various premalignant and malignant lesions in-vivo , and is hypothesized to be one the main causes of aging [9 , 43 , 44] . Due to this ability to promote invasiveness of cancer cells and the spreading of senescence to neighboring cells IL-6 and IL-8 are of special interest [3 , 45] . While it is probably not detrimental to transiently activate the respective signaling pathways , the long-term persistence of unrepairable DNA damage leads to a lasting activation of NF-κB through the DDR mechanisms and thereby to a prolonged stimulation of IL-6 and IL-8 . Ultimately , this initiates and perpetuates a vicious cycle from which cells cannot escape and causes the development of the SASP . To explore and validate previously generated in-silico results in-vitro , we isolated murine dermal fibroblasts from NEMO-floxed mice and transfected these with a Cre-recombinase plasmid to deplete NEMO . Contrary to NEMO knockout MDFs we observed RelA enrichment in the nucleus in DNA damaged wildtype cells . This suggests that mainly NEMO is responsible for the forwarding of DNA damage signals from the DDR to NF-κB signaling . We were particularly interested in achieving inhibition of IL-6 and IL-8 expression and secretion in-silico and in-vitro . As we could show in our in-vitro results , DNA damaged NEMO knockout cells did not reveal any induction of IL-6 or IL-8 homologue mRNA expression , suggesting that DNA damage-triggered IL-6 and IL-8 expression is mainly conferred by NF-κB signaling . This was confirmed on protein level , showing a strong decrease in secretion of both IL-6 and IL-8 homologues in NEMO knockout MDFs . In conclusion , abolishing NEMO is sufficient to not only block the signaling from DDR to NF-κB but also to decrease expression and secretion of two of the most prominent and established SASP mediators IL-6 and IL-8 . The question arises why damaged senescent cells have to start expressing and secreting factors that are detrimental to themselves , surrounding cells and tissues . The secretion of many SASP factors can be explained firstly by the attempt to clear senescent cells from tissue by cells of the innate immune system and secondly as a warning to the microenvironment that there is a danger in the near vicinity . Senescent cells secrete different factors that attract phagocytic immune cells and induce proteolytic enzymes to facilitate their migration through the extracellular matrix [46] . As long as damaged cells can be cleared in early phases the SASP is probably beneficial for the organism , however once the immune system cannot keep up with the emergence of damaged cells , detrimental effects accumulate and tissue takes damage [2 , 47] . In this phase , it would be beneficial to have the possibility to counteract the SASP and give the immune system time to catch up . In summary , we could illustrate that in-silico identification of genes with mechanistic contribution in the regulation of the SASP , confirmed under experimental conditions in-vitro , is a highly suitable approach and holds substantial promise to identifying therapeutic targets to delay or even prevent the detrimental SASP effects on tissue homeostasis and overall ageing . Using our Boolean model , we were able to reproduce published data in-silico and generate various knockout proposals to shut down two of the most detrimental effectors of the SASP . This is of major clinical relevance in terms of tissue aging . In fact , SASP factors like IL-6 and IL-8 have been correlated with inflammaging not only driving the aging process itself , but also promoting aging associated morbidity , frailty and mortality [48] . We additionally were able to validate and prove one of the most prominent knockout suggestions in-vitro , keeping in mind that there might always be detrimental off-target effects when altering a major signaling pathway like NF-κB . However , targeting NEMO and its interaction partners , as already shown in studies of inflammatory arthritis and diffuse large B-cell lymphoma , may hold promise for the development of new therapies for age-related pathologies in which senescence and the SASP play a role [49 , 50] . Murine dermal fibroblasts from an inducible connective tissue-specific NEMO-deficient mouse model were used for in-vitro experiments . This mouse line ( Col ( I ) α2-CreERT+;NEMOf/f ) was generated by crossing Col ( I ) α2-CreERT transgenic mice [51] with NEMO floxed mice [35] . These mice were backcrossed to C57BL/6J for at least 6 generations . They were maintained in the Animal Facility of the University of Ulm with 12 h light–dark cycle and SPF conditions . The breeding of the mice and all experiments were approved by the animal ethical committee ( approval number , Tierversuch-Nr . 1102 , Regierungspräsidium Tübingen , Germany ) . For mice genotyping standard PCR techniques were used . The sequences of the primers used in this manuscript are summarized in S1 Table . Briefly , DNA was isolated from the tail tip of an individual mouse using a commercial kit ( Easy DNA kit , Invitrogen ) . Purified DNA was later dissolved in TE and used for PCR amplification . The PCR products were run in QIAxcel Advance system ( Qiagen ) using the program AM320 and then documented digitally . Murine dermal fibroblasts ( MDFs ) were isolated from ear skin of young mice and cultured as previously described [52] . DNA damage was induced by adding etoposide to cell culture media at a concentration of 25 μM for 3 hours after overnight serum-starvation . Supernatants subsequently removed and cells were rinsed with PBS before adding fresh culture media . Cells and/or media were used 24 h later for further analysis . Recombineering technology [53] was used to constract plasmids containing CDS of both Cre recombinase and fluorescence reporter , mRuby2 or only mRuby2 . pCAG-Cre vector ( a gift from Connie Cepko , Addgene plasmid # 13775 ) was used for the recombineering . In the first construct , the aim was to insert the T2A-mRuby2 sequence before the stop codon of Cre recombinase and in the second construct , the aim was to replace the Cre ORF with mRuby2 ORF . In brief , synthetic DNA fragments were synthesized either as gBlock ( IDT ) or as GeneArt string ( Thermo Scientific ) . Four DNA fragments were synthesized , the first one contained 5’ 50 bp homology regions to the vector ( targeting 50 nucleotide upstream of Cre ORF stop codon ) , chloramphenicol and ccdB cassettes and 3’ terminal 50 bp homology regions to the vector ( targeting 50 nucleotide downstream of last amino acid coding codon of Cre ORF , i . e . , condon preceding the Cre ORF stop codon ) . The second synthetic fragment contained 5’ 50 bp homology regions to the vector ( targeting 50 nucleotide upstream of Cre ORF start codon ) , chloramphenicol and ccdB cassettes and 3’ terminal 50 bp homology regions to the vector ( targeting 50 nucleotide downsteram of Cre ORF stop codon ) . The third synthetic fragment contained 5’ 50 bp homology regions to the vector ( same as fragment 1 ) , T2A sequence-mRuby2 ORF and 3’ terminal 50 bp homology regions to the vector ( same as fragment 1 ) . The fourth synthetic fragment contained 5’ 50 bp homology regions to the vector ( same as fragment 2 ) , mRuby2 ORF and 3’ terminal 50 bp homology regions to the vector ( same as fragment 2 ) . E . coli containing pCAG-Cre was processed for electrocompetent using standard methods and these electrocompetent E coli , containing pCAG-Cre were electroporated with a dual inducible expression plasmid pSC101-ccdA-gbaA ( a gift from Prof . A . Francis Stewart ) and selected for ampicillin 100μg/ml and tetracycline 3 . 5μg/ml at 30°C . Next day , 4–5 colonies were expanded and the expression of recombineering proteins , λphage redα , redβ and redγ and recA ( redgbaA ) was induced by L-rhamnose ( 1 . 4mg/ml ) . After 1 h of L-rhamnose treatment , the induced E . coli were processed for electrocompetent and then electroprorated either with synthetic DNA fragments 1 or 2 . After 1 h of recovery in SOC medium , the electroporated E coli , were plated in LB-agar containing ampicillin 100μg/ml , tetracycline 3 . 5μg/ml , chloramphenicol 25μg/ml and 1 . 4mg/ml L-arabinose . L-arabinose addition induced the expression of ccdA , the antidote of ccdB in that only recombined plasmid containing E . coli can survive . Thereafter colonies from fragment 1 and fragment 2 electroporated E . coli plates were picked and expanded for the verification of first recombinant product using restriction digestion analyses . The corresponding colony was expanded and redgbaA expression was induced by L-rhamnose for 1 h . The induced E . coli containing either recombined DNA fragment 1 or fragment 2 were made electrocompetent for the second round of recombineering . The E . coli , containing recombined DNA fragment 1 then electroporated with synthetic DNA fragment 3 . The E . coli , containing recombined DNA fragment 2 were electroporated with synthetic DNA fragment 4 . The recovered electroporated E . coli were plated in LB-agar containing ampicillin 100μg/ml and incubated at 37°C overnight . Colonies from both plates were picked , expanded and verified for the second recombinant products . The correct plasmids were sequenced and verified through commercial services ( Sequiserve , Germany ) . Plasmid preparation was performed using a commercially available kit ( Qiagen plasmid plus kit , Qiagen ) . This plasmid ( pCAG-Cre-T2A-mRuby2 ) can be obtained from the authors on request and was deposited in the Addgene repository ( Accession ID 102989 ) . Early passage MDFs with a floxed NEMO allele were transfected with a Cre expressing vector using an electroporation-based transfection method ( Amaxa , Lonza Group ) . Transfer of the plasmid was performed using a commercial kit with the AMAXA program N24 ( Nucleofector Kits for Mouse or Rat Hepatocytes , Lonza ) . Successful NEMO knockout was assessed by PCR as explained before . Two days after transfection cell populations were purified using the mRuby2-based reporter system included in the previously described Cre-expressing vectors . Gating was set for living cells and singlets , sorting was based on mRuby2 expression in the PE-channel . FACS-sorting was performed with a FACSAria III system ( BD Biosciences ) and analysis was done on FACSDiva and FlowJo ( Tree Star ) software . Cells were fixed in 4% PFA in PBS for 15 min and thereafter treated with 0 . 1% Triton X-100 for 10 min at room temperature . Blocking was performed in 5% BSA for 1 h at room temperature . Anti-p65 ( #8242 , 1:200 , Cell Signaling ) and anti-γH2A . x ( ab22551 , 1:200 , Abcam ) were used as primary antibodies overnight at 4°C . Incubation with the secondary antibody Alexa 488 goat anti-mouse ( for γH2A . x , 1:500 ) and Alexa 555 goat anti-rabbit ( for p65 , 1:500 ) was performed at room temperature for 1 h . Western blot analyses were performed as described earlier [54] . In brief , murine dermal fibroblasts were lysed in RIPA lysis buffer ( 25mM Tris-HCl pH 7 . 6 , 150mM NaCl , 1% NP-40 , 1% sodium deoxycholate , 0 . 1% SDS ) supplemented with protease and phosphatase inhibitors ( Thermo Scientific ) . Cells in RIPA were sonicated using sonopuls HD 2070 and MS72 microtips ( Bandelin ) . The sonicator setting was 50% power 3 cycles and 10 sec for three times . Following sonication , the lysate was centrifuged for 15 min at 14000 rpm and 4°C . The supernatant was collected and protein concentration was measured by Bradford Assay ( Biorad ) . 50μg of protein from each lysate was resolved in 4–20% SDS-PAGE , followed by transfer to nitrocellulose membrane and probing the membrane with anti-NEMO antibody ( 1:1000 , Abcam ) . The membrane was incubated with goat anti-rabbit IgG coupled with HRP for 1 hr ( Jackson ImmunoResearch ) . Thereafter the membrane was developed by LumiGLO chemiluminescence reagent ( Cell Signaling Technologies ) using Fusion FX7 Geldoc system ( Vilber Lourmat ) , followed by stripping with Restore Plus Western blot Stripping Buffer ( Thermo Scientific ) and re-probed with anti-β-actin antibody coupled with HRP ( 1:12000 , Santa Cruz ) , finally developed the membrane using LumiGLO . Twenty-four hours after treatment , total RNA was isolated from cultured murine dermal fibroblasts using a commercial kit ( RNeasy Mini Kit , Qiagen ) as described by the manufacturer . Two μg of RNA per sample were reverse transcribed using illustra Ready-To-Go RT-PCR Beads ( GE Healthcare ) . Quantity and quality of total RNA and cDNA was assessed using Nanodrop 1000 ( Thermo Scientific ) and QIAxcel Advance system ( Qiagen ) . The 7300 real time PCR system ( Applied Biosystem , Life Technologies ) was used to amplify cDNA using Power SYBR green mastermix ( Applied Biosystems , Life Technologies ) . Sequences for primers used in all experiments and genotyping are provided in S1 Table . After etoposide treatment cells were supplied with fresh culture media . Culture media was taken for analysis of secreted IL-6 and murine IL-8 homologues ( KC and MIP-2 ) 24 h after treatment . Media was stored at -80°C until analysis . Concentrations of secreted IL-6 and murine IL-8 homologues after DNA damage were determined using commercial kits ( Mouse IL-6/KC/MIP-2 Quantikine ELISA Kit , R&D ) as described by the manufacturer . The influence of a NEMO knockout was compared to wildtype controls based on IL-6 , IL-8 homologue and p21 mRNA expression as well as IL-6 and IL-8 homologue protein secretion . The sample size for all experiments was 5 per group . The expression and secretion of the two groups was tested using unpaired two-tailed t-test . Furthermore , the influence of the NEMO knockout compared to wildtype controls on the nuclear translocation of p65 was measured by the percentage of fluorescence intensity in the cell nucleus as well as cytoplasm ( sample size = 10 ) . The fluorescence intensity was tested using unpaired two-tailed t-test . The exact p-values are depicted in the respective figures . The figures show mean values . Error bars correspond to the standard error of the mean . In a first step , IL- and DNA-damage pathways included in the Boolean model of SASP were reconstructed individually . To generate the independent gene regulatory networks of inflammatory and DDR signaling , we collected peer-reviewed literature that is considered relevant in the context of SASP ( see Table 1 ) . This literature reports data about the local interaction of key genes regulating each pathway . The information was collected in murine and human experimental in-vivo and in-vitro studies . In order to control the complexity of model we restricted the set of regulatory factors in the model to the most relevant for SASP and to those being important components of each pathway . The modeled pathways were chosen based on the requirement in the onset and maintenance of the SASP shown in studies related to senescence and the SASP . In total 80 publications were used to determine the relationships between the individual components of the model ( Table 1 ) . After the reconstruction of Boolean network models of inflammation and DNA damage response , both were combined into a larger network . The impact of combining the two network models instead of simulating them independently is shown by additional analysis in S1 Text . Simulations based on specific environmental ( input ) conditions were performed to find the corresponding attractors . Furthermore , to identify possible interaction targets , i . e . , to generate testable hypotheses about interventions , we fixed corresponding regulatory factors to either 0 or 1 ( modelling of knockout or overexpression , similar to [55] ) and reran the simulations ( S1 Text ) . Given an interaction target , we looked for the attractors that positively influence the DNA damage response phenotype . Network figures were drawn with Biotapestry ( www . biotapestry . org ) . Simulations of the Boolean network were performed with the package BoolNet [12 , 56] in R ( www . r-project . org ) . This model contains two external signals ( DNA damage and Activated Oncogenes ) . These signals do not coincide with genes within the network , but represent different stimuli from external or internal sources that are known to activate the DNA damage response and/or cell cycle arrest signaling through either p16INK4 or p53/p21 .	The senescence associated secretory phenotype is developed by cells undergoing permanent cell cycle arrest . This phenotype is characterized by the secretion of a variety of factors that facilitate tissue breakdown and inflammation and is therefore theorized to , in part , be causal for aging and age-related diseases . In recent years the SASP has been implicated in a variety of chronic inflammatory diseases . Due to these advances , it is imperative to better understand the dynamics of this cellular phenotype and to find ways to disrupt it . We have developed a Boolean network incorporating the major signaling pathways of the SASP that allows us to specifically investigate interactions of the pathways and genes involved . We validated our model by reliably reproducing published data on the SASP . We utilized our model to uncover components that directly control the detrimental effects of the senescence associated secretory phenotype that are largely caused by IL-6 and IL-8 , two major factors of the SASP in establishing and spreading senescence as well as causing local inflammation . In subsequent in-vitro experiments , we were able to verify our computational results and could suggest NEMO as one potential target for therapy of SASP-related diseases .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "medicine", "and", "health", "sciences", "senescence", "gene", "regulation", "cell", "cycle", "and", "cell", "division", "cell", "processes", "fibroblasts", "dna", "damage", "physiological", "processes", "developmental", "biology", "organism", "development", "network", ...	2017	A model of the onset of the senescence associated secretory phenotype after DNA damage induced senescence
Leprosy is endemic in large part of Brazil with 28 , 761 new patients in 2015 , the second largest number worldwide and reaches 9/10 . 000 in highly endemic regions and 2 . 7/10 . 000 in the city of Fortaleza , Ceará , Northeast Brazil . For better understanding of risk factors for leprosy transmission , we conducted an epidemiologic study supplemented by 17 locus VNTR and SNP 1–4 typing of Mycobacterium leprae in skin biopsy samples from new multibacillary ( MB ) patients diagnosed at a reference center in 2009 and 2010 . Among the 1 , 519 new patients detected during the study period , 998 ( 65 . 7% ) were MB and we performed DNA extraction and genotyping on 160 skin biopsy samples , resulting in 159 ( 16% ) good multilocus VNTR types . Thirty-eight of these patients also provided VNTR types from M . leprae in nasal swabs . The SNP-Type was obtained for 157 patients and 87% were of type 4 . Upon consideration all VNTR markers , 156 different genotypes and three pairs with identical genotypes were observed; no epidemiologic relation could be observed between individuals in these pairs . Considerable variability in differentiating index ( DI ) was observed between the different markers and the four with highest DI [ ( AT ) 15 , ( TA ) 18 , ( AT ) 17 and ( GAA ) 21] frequently demonstrated differences in copy number when comparing genotypes from both type of samples . Excluding these markers from analysis resulted in 83 genotypes , 20 of which included 96 of the patients ( 60 . 3% ) . These clusters were composed of two ( n = 8 ) , three ( n = 6 ) , four ( n = 1 ) , five ( n = 2 ) , six ( n = 1 ) , 19 ( n = 1 ) and 23 ( n = 23 ) individuals and suggests that recent transmission is contributing to the maintenance of leprosy in Fortaleza . When comparing epidemiological and clinical variables among patients within clustered or with unique M . leprae genotypes , a positive bacterial index in skin biopsies and knowledge of working with someone with the disease were significantly associated with clustering . A tendency to belong to a cluster was observed with later notification of disease ( mean value of 3 . 4 months ) and having disability grade 2 . A tendency for lack of clustering was observed for patients who reported to have lived with another leprosy case but this might be due to lack of inclusion of household contacts in the study . Although clusters were spread over the city , kernel analysis revealed that some of the patients belonging to the two major clusters were spatially related to some neighborhoods that report poverty and high disease incidence in children . Finally , inclusion of genotypes from nasal swabs might be warranted . A major limitation of the study is that sample size of 160 patients from a two year period represents only 15% of the new patients and this could have weakened statistical outcomes . This is the first molecular epidemiology study of leprosy in Brazil and although the high clustering level suggests that recent transmission is the major cause of disease in Fortaleza; the existence of two large clusters needs further investigation . Leprosy , caused by infection with Mycobacterium leprae remains a significant public health problem in many developing countries . The disease presents a wide spectrum of clinicopathologic forms that ranges from tuberculoid leprosy ( TT ) to borderline forms and lepromatous leprosy ( LL ) and lesions involve skin and peripheral nerves . Disease can be paucibacillary ( PB ) or multibacillary ( MB ) with the most severe LL form involving organs such as liver , spleen and bone marrow and the bacterial burden in such patients is massive and causes severe deformities when not treated . Multi-drug therapy using dapsone , rifampicin and clofazimine was implemented in the 1980s and has considerably reduced disease prevalence , but that is not the case with incidence , implying that leprosy is still being transmitted to a considerable extent [1] As M . leprae cannot be cultured on artificial media , molecular techniques have been used for better characterization of the organism [2 , 3 , 4 , 5] , including the deciphering of the genome sequence [6] . Single nucleotide polymorphisms ( SNPs ) analysis allowed studies on phylogeography of leprosy , evolving models for the global spread of M . leprae [7 , 8] . Besides SNPs , Variable Number Tandem Repeats ( VNTRs ) are used for genotyping and a certain relation between the number of certain VNTR alleles and SNP-Type has been observed [9] , showing that VNTR typing adds to our knowledge on spread of leprosy . Multiple-locus variable number tandem repeats analysis ( MLVA ) of a set of micro- and mini-satellites of M . leprae is a fingerprinting procedure for differentiation at the strain level [10 , 11 , 12 , 13] and useful during transmission studies , to distinguish reactivation from re-infection [14] and to study bacterial population structure on different levels and countries , as described for Brazil [15 , 16] , China [9 , 17 , 18 , 19] , India [20 , 21 , 22 , 23 , 24] , Philippines [25 , 26] , Thailand [27 , 28] , Mexico [29] , Colombia [30 , 31] and the United States [32] . Fortaleza is the capital of Ceará , a state located in northeastern Brazil . In 2015 , 80 . 5% of the 184 municipalities in the state diagnosed new patients of leprosy and 10% were classified as being hyperendemic , defined by having an incidence of higher that 4/10 . 000 . Ceará is one of the poorest regions of the country , reporting 1 . 743 new leprosy patients in the same year , including 528 in the capital , representing an incidence rate of 2 . 7/10 , 000 inhabitants . MB is detected in two thirds of these patients and 5 . 9% of the total patients reported in the state are younger than 15 years of age , both of which are indicators of ongoing and recent transmission seems of the disease [33] . Previous genotyping of M . leprae strains in Brazil , from a set of unrelated patients from the Southeast region of the country demonstrated a high VNTR based genetic variability in predomintly SNP-Type 3 background [15] . Later , it was observed that SNP-Type 4 is much more frequent in the North-northeast part of the country [6] . Although preliminary data on use of genotyping to add to transmission studies have been presented in Ceará [34] , Mato Grosso [35] and Pará [36] , no full reports exist on molecular epidemiology studies of leprosy and the rearch for risk factors for recent transmission in Brazil; therefore our study addresses this gap . Fortaleza is the capital and also largest city of the state of Ceará , and the fifth largest city ( 314 , 930 km2 ) in Brazil with 2 , 627 , 482 inhabitants in 2017 . It has 120 neighborhoods and the highest population density among the country's capitals . Although Fortaleza has the tenth highest GDP in the country and the highest in the Northeast region , it has the typical uneven distribution of wealth observed in most of Brazil’s major cities . Besides being an important industrial and commercial center , it is the second most desired tourist destination in Brazil and fourth in number of visitors [37 , 38] . This study was designed to better understand the clinical and epidemiological characteristics of leprosy in the city of Fortaleza . A cross-sectional study was conducted from November 2008 to December 2010 and during this period , all new leprosy patients diagnosed by trained dermatologists of the National Reference Center of Dermatology Dona Libânia ( CDERM ) were invited to participate in the study . This tertiary reference center serves about 80% of the almost 800 new leprosy patients diagnosed annually in Fortaleza and is the most important reference center for skin disease , including leprosy , in that city [39] . Patients were diagnosed by clinical evaluation; microscopic evaluation of bacillary index of acid fast bacteria in slit skin smears ( SSS ) analysis and histopathological evaluation of biopsy specimens . Patients were classified according to Ridley-Jopling criteria based on histological study and bacterial indices ( BI ) [40] . All new patients responded to a detailed questionnaire that included demographic , epidemiologic , socioeconomic , environmental and behavioral components . In addition to the questionnaire , data for the patients were introduced and maintained by registered health workers in the SINAN database ( http://portalsinan . saude . gov . br ) . A second skin biopsy and nasal swab was collected for genotyping of M . leprae in a subset of all diagnosed patients . The skin biopsy samples were collected using a 5 mm punch . Tissue for histopathology was treated with formol and embedded in paraffin while the tissue for genotyping was placed in a sterile 1 . 5 mL tube and stored at -20°C . The DNA was extracted by using the DNeasy Blood & Tissue kit ( Qiagen Biotecnologia do Brasil Ltda , SP , Brazil ) following the manufacturer's guidelines . Nasal swabs were collected from patients who also provided a second skin biopsy for genotyping , by gently rubbing a swab previously wetted with Tris-EDTA buffer ( pH 8 . 0 ) , in one side of each nostril over the lateral conchae . After collection , each swab was immersed in a sterile and labeled tube and stored at -20°C until processing as described by Lima et al . [41] Genotyping by MLVA of 17 VNTRs was performed as described by Kimura et al . [13] and based on four multiplex PCRs that generated 17 amplicons . The allele for each VNTR locus is the copy number of the repeats which was determined by denaturation of amplicons and capillary gel electrophoresis on the sequencer ABI 3130 Genetic Analyzer , using the internal molecular weight sizing standards ( LIZ 500 ) . The copy number of each locus was calculated based on the size of the PCR amplicon using the Peak Scanner software ( Applied Biosystems do Brasil ) and comparing to previously calibrated M . leprae strain NHDP63 . To study reproducibility of the assay , DNA from five M . leprae samples from Brazil was sent to CSU for comparative analysis of the alleles . For differentiation of four genotypes of M . leprae based on three SNPs , we used a procedure that combined PCR-restriction enzyme analysis ( REA ) and direct sequencing as described by Sakamuri et al . [26] . Differentiation of genotypes 1/2 from 3/4 was obtained by submitting to BstUI mediated PCR-RFLP analysis of the locus at nucleotide position 2 , 935 , 685; digestion occurs in case of genotype 3/4 and lack of digestion for genotype 1/2 . Differentiation of genotypes 3 and 4 is obtained by SmlI mediated PCR-RFLP at nucleotide position 14 , 676; digestion indicates SNP-Type 4 and lack of SNP-Type 3 . Differentiation of SNP-Type 1 or 2 was performed by direct PCR sequencing as described by Monot et al . [7] . The copy number of all alleles were introduced into Microsoft Excel files and imported into Bionumerics software ( version 7 . 6; Applied Maths; Sint Martens Latem , Belgium ) . Definition of clustering was based on comparison of the copy number of the VNTRs using two different stringencies: either considering those that presented identical copy number for all 17 alleles , or considering those that had identical copy number in 13 alleles , excluding the four most variable loci . A similarity matrix was constructed using the categorical similarity coefficient and the unweighted pair group method with arithmetic mean ( UPGMA ) . This was the basis for a complete linkage tree , a circular top score UPGMA tree and a range of minimum spanning trees ( MST ) . The cartographic bases and the population used were obtained from the Brazilian Institute of Geography and Statistics ( http://www . ibge . gov . br/ ) . The coordinates were obtained using a global positioning system ( GPS ) and stored in a geographic database ( BDGeo ) . Data were used to generate graphics , satellite imagery processing , to establish topological relations between the graphic elements and their attributes , spatial analysis and visualization through thematic maps . We evaluated the spatial analysis Kernel density estimation ( KDE ) using a fixed radius of 2 km . Analyses were performed in ArcGis ( http://www . esri . com/ ) and TerraView ( http://www . dpi . inpe . br/menu/Projetos/terraview . php ) . In TerraView it was possible to build a dual Kernel or Kernel ratio , based on the number of patients and the population [42] . We used the interpolator points by Inverse Distance Weighting ( IDW ) to estimate the cell values using a weighted linear combination of a set of sampling points . The satellite image in Fig 1 was generated using the sensor Sentinel 2 of the European Space Agency ( ESA ) ( https://sentinel . esa . int/web/sentinel/user-guides/sentinel-2-msi ) with Open Access CC-BY License ( http://open . esa . int/ ) . For evaluation of the association of the demographic , clinical and environmental/behavior variables and having a clustered or a unique M . leprae genotype , chi squared and Fisher exact tests were used . Mann-WhitneyU test was used for evaluation of differences between a single characteristic in individuals with clustered genotypes or unique patterns . An informed consent form was signed by the participants of the study , authorizing the collection of clinical samples . The present study was approved by the Ethics Committee of CDERM and the national ethical committee . At CDERM , 830 ( 284 PB and 546 MB ) and 689 ( 237 PB and 452 MB ) new leprosy patients were diagnosed respectively in 2009 in 2010 , totaling 1519 in the study period and among these , 998 were MB patients ( 65 . 7% ) . Recruitment was conducted only on two days per week , which was further reduced in December , January and July and on holidays . This resulted in the collection of a second biopsy specimen for genotyping from 301 MB patients only of whom we received 160 ( only 92 from 2009 and 68 from 2010 ) . This resulted in M . leprae genotyping of 16 . 8% and 15% of the newly diagnosed MB patients respectively in 2009 and 2010 . From these 160 patients , 101 also had nasal swab collected . Because the questionnaire was developed for a larger case-control study evaluating risk factors for leprosy and the patients within the study presented here only partially overlapped with the larger study , we accessed data from the SINAN database ( http://portalsinan . saude . gov . br ) for 61 of the 159 patients ( 31% ) . Of the 160 patients biopsies submitted to M . leprae genotyping , 159 yielded high quality MLVA-based and 157 SNP-based genotypes and are presented in S1 Table and S1 Fig . Initially , 134 M . leprae were defined as SNP-Type 4 ( 85% ) , 15 as SNP-Type 3 , three as SNP-Type 1 and six samples could be characterized only to the SNP-Type 1 or 2 level because of insufficient material for sequencing . Four isolates with SNP-Type 3 were grouped within the MLVA-based clusters of isolates with SNP type 4 so we suspected wrong classification due to partial digestion during PCR-RFLP . Two samples had sufficient material left to repeat and both were indeed confirmed as being SNP-Type 4 , resulting in 136 SNP-Type 4 ( 86% ) and 13 ( 8 . 2% ) SNP-Type 3 . The MLVA-based typing results are presented in S1 Table and all but eight strains yielded the complete 17 locus-based genotypes ( 95% ) , five isolates failed in the amplification of one locus while another two lacked alleles for five and six alleles respectively . The latter strain was clustered with another but not included in the analysis of clusters at high cluster stringency . Two isolates presented a double peak , one at ( TA ) 18 and another at ( AT ) 15 and these alleles were not considered for analysis . The differentiating power and allele distribution of the satellites is presented in Table 1 , and varied between 0 and 0 . 93 , and 1 and 22 , respectively . Three markers–[ ( GGT ) 5 , 6–3 and 21–3] were invariable , four [ ( AT ) 15 , ( TA ) 18 , ( AT ) 17 and ( GAA ) 21] were highly discriminatory with a Hunter-Gaston discriminatory index ( HGDI ) above 0 . 8 , and the rest had a HGDI of less than 0 . 8 . Interestingly , the copy number distribution pattern of 18–8 is different from that of the other markers and presented a bimodal pattern ( Table 1 ) . Regarding cluster analysis , when using the highest stringency including 17 markers , we observed 157 different genotypes formed by 154 singletons and three clusters of two patients each , resulting in an overall cluster level of 3 . 8% ( 6/157 ) . Upon analysis of the data of the patients within each of the three clusters and or of those with unique genotypes , no particular risk factor for belonging to one of or any cluster was identified . However , when excluding the four markers with HGDI > 0 . 8 , 83 different genotypes were detected , 63 unique ones and another 20 found in 96 patients , resulting in a cluster rate of 60 . 4% ( 96/159 ) . The two largest clusters were composed of 23 and 19 patients and the remaining clusters were composed of one of six , two of five , one of four , six of three and eight of two patients . Again , no clear patient characteristic was detected that could explain the formation of an individual cluster but when analyzing the data of those belonging to a genotype cluster or not , some significant associations and tendencies for clustering were observed . We observed a significant association of clustering being BI positive ( p = 0 . 037 ) or having worked ( p = 0 . 049 ) with someone with leprosy ( note: working together has p = 0 . 25 ) . Surprisingly , the variable ‘having lived with someone who had leprosy’ demonstrated an inverse relation with clustering ( p = 0 . 065 ) . Although not significant , an association was observed between clustering and disability ( p = 0 . 445 ) , mainly because of a tendency to have more grade 2 disability among clustered patients ( 13 . 1% against 5 . 1% ) and longer time between observing first lesion and diagnosis/disease notification ( p = 0 . 14 ) . Another unusual finding was that alcohol consumption was significantly associated with non-clustering , i . e . , of having unique genotypes ( p = 0 . 047 ) ( Table 2 ) . Note however that some of these associations occurred with the number of patients for some categories being < 5 and a detailed relation between clustering and variables is presented in S2 Table . We also performed chi-square analysis of patients and other characteristics of the 18 patients from cluster 12 and 23 patients from cluster 14 ( totalling 41 among 159 = 25 . 8% in cluster ) and observed no significant association of the clustered cases when compared to the rest of any of the variables . Additionally , we plotted the date of diagnosis of the patients , clustered cases and those belonging to the two major clusters ( 12 and 14 ) on a monthly based time scale of the study period and although some higher frequency of diagnosis was observed between March and June of 2009 , no particular independent increase in clustering was observed during the study period ( S2 Fig ) . Among 38 patients , MLVA patterns were also available from nasal swabs , with the exclusion of alleles 6–3 and 18–8 , not performed in this sample type and as described recently , difference in copy number of the alleles with highest DI was observed in the M . leprae genotypes when comparing both samples in a considerable number of patients [43] . Upon inclusion of the genotypes of M . leprae present in nasal swabs in the analysis , we observed that eight genotypes from nasal swab were part of some cluster , increasing the number of clustered patients by 10 . One cluster with a genotype shared by M . leprae in skin biopsy of five patients increased to eight when considering genotypes in nasal swab while three new clusters were observed , one composed of the genotype observed in the nasal swab of two patients , and two others composed of a genotype that was observed in the nasal swab and skin biopsy of two patients each ( S3 Table ) . Data to perform spatial analysis was available for 156 patients and demonstrated clearly a higher density in the western part of Fortaleza ( S3A Fig ) . The Figure also presents the number of patients per neighborhood ( S3B Fig ) , the population estimated by the 2010 census by neighborhood ( S3C Fig ) , KDE using georeferenced homes of patients ( S3D Fig ) , KDE using neighborhood ( S3E Fig ) and dual Kernel using neighborhood ( S3F Fig ) in Fortaleza . The neighborhood with the highest number of patients is Granja Lisboa . The result obtained by using KDE in the neighborhoods showed , as expected , the same clusters , both centered in the neighborhoods Bom Jardim , Bonsucesso , Granja Lisboa and Granja Portugal ( S3D and S3E Fig ) . However , when applying dual Kernel analysis , two clusters are observed , one including Granja Lisboa and Siqueira ( southwest region ) , the same as observed using Kernel ( neighborhood ) , and the second centered in the neighborhood Jacareacanga in northern Fortaleza . The distribution of the patients among the non-clustered patients and for each cluster across the neighborhoods of the city is presented in S4 Table and demonstrates that nine groups with clustered genotypes had at least two patients in the same neighborhood ( groups 2 , 7 , 8 , 9 , 12 , 14 , 16 , 17 and 19 ) . Overall , patients with the same M . leprae genotype are spread across the city , except for the biggest cluster 14 showing two pairs of two very nearby patients ( Fig 1 ) . When performing KDE analysis using a distance of 2 km concentrating on the distribution of the patients from the two largest clusters , association was observed with some neighborhoods . The cluster formed by 19 patients was associated with Jacareacanga , Canindezinho , Conjunto Esperança and Manoel Sátiro , while those of the cluster with 23 isolates with Bonsucesso and Vila Pery ( Fig 2 ) . However , when performing the same type of analysis with the 62 patients with unique genotypes , we observed association with Granja Lisboa , Granja Portugal and Bom Jardim . Note however that the number of patients with unique patterns is about three times higher than those in each of the two biggest clusters . Finally , we plotted distribution of patients and performed spatial analysis according the number of lesions and number of bacilli observed by bacilloscopy ( S4 Fig ) . The number of lesions varied between 0 and 88 ( total of 2120 , medium value 13 . 6 and standard deviation of 13 . 2 ) and bacterial indices were between 0 and 6+ ( total 421 , median value 2 . 7 and SD 1 . 99 ) . Although we observed that patients with high number of lesions or high bacillary load were spread over the city , two neighborhoods that were associated with cluster 12 ( Canindezinho and Conjunto Esperança ) and 14 ( Bonsucesso ) presented patients with high BAAR . When UPGMA based dendrograms including all 17 satellites and with or without including SNP-Type were constructed , most isolates belonged to two major groups . The isolates that were not of the SNP-Type 4 were observed at the outer limits of the tree ( S1 Fig ) . For evaluation of the bacteriological population structure and the influence of inclusion of loci on cluster formation and tree topography , we constructed a MST including either all 17 satellites or gradually removing the most variable ones . As observed in organizing the allele number in an Excel file , the same three clusters of genotype pairs were observed in the MST when including all 17 loci , and the 20 clusters when omitting the four most variable markers ( Fig 3 ) . Depending on the number of VNTRs included for MST construction , we observed either two or three major groups and gradually excluding VNTRs with the highest variability , we observed that AC9 or/and AC8b are the main drivers for maintaining separate groups; omitting these markers resulted in a population with a large central cluster of 88 isolates with 11 branches . Most of the isolates have indeed a 6- or 7-copy number of these alleles and leaving out these markers coincides with the observation of clusters formed by different SNP-Types ( S5 Fig ) . In 1991 , the WHO adopted a resolution for elimination of leprosy by the year 2000 and implementation of MDT resulted in a significant reduction of prevalence . Between 2002 and 2012 , a 65% reduction in the prevalence ( from 4 . 33 to 1 . 51 patients/10 , 000 ) was achieved in Brazil but leprosy is unevenly distributed within the country with pockets of incidence levels of more than 10/10 , 000 [44] . The Northeast region is the poorest of the country reporting a third of the newly diagnosed patients and a detection rate that is twice that of the average in the country and the State of Ceará is one of the poorest states in the region . Over 10% of its municipalities classified as hyperendemic and the capital , Fortaleza , considered a priority for leprosy control , having the highest demographic density in the country and one of the municipalities in the state with the highest detection rates [45] . In addition , 5 . 9% of the new patients are less than 15 years of age and only half of the contacts are being investigated for disease [33] . Transmission of leprosy is assumed to be from person to person through the respiratory system or damaged skin , with risk for developing disease being higher if a family member had disease and even more when these presenting the LL form [46 , 47] . However , new patients often mention lack of contact with other leprosy patients , suggestive of unrecognized transmission routes [48] , including exposure to an environmental source such as water , soil , plants and animals [49] but no study unequivocally demonstrated the mechanism of leprosy transmission [50] . Since the report on the existence of genetic variability [4 , 5] and of the genome sequence of M . leprae [6] , analysis of SNP-Types and micro- and mini-satellites added to our knowledge about genetic variability of M . leprae and its biology , such as existence of geographic or family associated genotypes [18 , 19 , 23] , genetic divergence between bacilli inhabiting different tissue [20] and differentiation between relapse and re-infection [14] . Although studies on genetic variability of M . leprae have been conducted in several regions endemic for leprosy , mostly detailed epidemiologic information is missing except for a study in Qiubei , China , demonstrating intra-familial strain types [19] and regional differences in clustering [18] . No prospective molecular epidemiology study with detailed epidemiologic and clinical data have been reported except for a study reporting transmission of dapsone resistant M . leprae in Cebu , the Phillipines [51] . We hereby confirm the high prevalence of SNP-Type 4 in the northeast of Brazil as reported previously [16] and probably due to introduction of leprosy by slave traffic from West Africa . Isolates with SNP-Type 3 are partly 3I , as defined by the gyrA97 SNP ( SNP7614 ) [52]; and our earlier observation during studies on drug resistance [14 , 53] . We also observed a surprising strong correlation between SNP- based and VNTR based genotypes suggesting that in certain populations , microsatellites are also deeply rooted into the bacterial population structure . Only by omitting GTA9 and AC8a from the analysis , the relation between VNTR and SNP-Type was disrupted . Association between certain VNTRs and SNP-Type has been demonstrated before [15 , 31] but might be more pronounced here due to the very high level of SNP-Type 4 in our study population . Because of the high level of SNP-Type 4 in the studied population , it would have been interesting to characterize the M . leprae isolates to the sub-SNP-Type level but no DNA was left to perform this . The influence of stringency of definition of genotype clustering for interpretation of transmission and phylogeny has been clearly demonstrated for tuberculosis [54] but not extensively for leprosy [26] . The difference in clustering level using two stringencies in the present study is remarkable ( 3 . 8% vs . 60 . 4% ) and we believe that the high clustering level represents recent transmission and therefore being the major drive for developing leprosy in Fortaleza . Although clusters are generally small , we also observed two larger ones and clusters of considerable size have also been described in China [18] , in the Philippines [26 , 27] and among those shared between humans and armadillos in the US [32] . The choice of stringency for definition of clustering in the present study is partly based on the fact that the four markers with Simpson Index >0 . 85 also were mostly presenting allele differences in the genotypes of M . leprae present in the nose and in the skin . Those were also among those disfavoring MLVA analysis for M . leprae genotyping as described by Monot et al . [8] . One weakness of our study is that we have no epidemiologic links that proovef that the 13 VNTR-based clustering is indicative for intense leprosy transmission in the present setting but this is probably due to lack of healthy household contacts ( HHC ) in the present sampling and low representativity of sampling . Extensive MIRU-VNTR genotyping data from M . tuberculosis show that the most variable MIRUs can be omitted without much loss of transmission links [54] . The number of M . leprae bacilli in the human body can reach 1012 so differences in copy number due to higher number of replication cycles during development of leprosy are imaginable . Finally , the presently used VNTR-based stringency is still higher than that those used by Sharma et al . [32] and Lavania et al . [24] . Sharma et al . related the SNP-VNTR type 3I-2-v1 genotype among 80 . 3% of the armadillo samples from the South of the US and 22/52 human patients were infected with M . leprae presenting one of two major genotypes . Interestingly , Lavania et al . [24] , using an identical typing approach observed 66 different patterns among 70 leprosy patients . Although sample representativity and other variables might strongly influence clustering levels , the difference between both studies is striking and might also be due to differences in transmission dynamics . Some markers that were included [ ( TA ) 10 and 18–8] were not used for genotype definition in the before mentioned studies but in the present study had a HGDI of 0 . 38 and 0 . 22 , respectively . This again suggests the need for regional evaluation of VNTRs for local M . leprae genotyping for developing "lower cost" genotyping in the mostly poorer endemic regions . However , having in mind the huge amount of information obtained from the standardized 24-MIRU-VNTR procedures for phylogenetic studies of M . tuberculosis , we here suggest the use of 17 STRs or even more for better understanding of transmission and phylogeny of M . leprae on a larger scale . The comparison of M . leprae genotypes present in skin biopsy and nasal secretion is described and discussed in detail elsewhere [43] . While all isolates presently presented four copies of ( GGT ) 5 , one nasal swab sample presented six copies of this allele [43] and although other alleles than that of four copies are described with very low frequency in Brazil [16] , they are more frequent in countries like Thailand [27] and the Philippines [25] . Contrary to the single allele with two copies of 23–3 described by Lima et al . [43] , in 8% of our patients , a single copy of this marker was observed . A further finding by Lima et al . was the observation that some individuals presented differences in copy number in five to seven loci , including less variable ones , being highly suggestive for multiple infection or more extensive intra-patient strain evolution . In addition and more importantly for transmission studies , our data show that inclusion of the genotypes from nasal swabs may have consequences for clustering outcome . Because the hypothesis is that the nose is a port of entry and exit of M . leprae , the genotype in nasal swabs could contribute to the transmission links suggested by genotyping M . leprae in skin biopsies . We therefore suggest that more studies including both samples are needed to understand transmission dynamics . However , as stated elsewhere , there is no guarantee that M . leprae in the nasal swab is representative for disease but very recently , molecular evidence for an important role of the nose in leprosy transmission was presented by Araujo et al . [55] . High levels of recent transmission in Fortaleza is also evidenced by the observation of two large clusters of about 20 patients and may indicate the existence of two main lineages of M . leprae strains differing in four alleles ( AC8b , GTA9 , AC9 and AC8a ) in Fortaleza . This might be related to some undetected factor causing more transmission of these strains but unfortunately , our study did not allow their definition and might depend on a social network approach as demonstrated in molecular epidemiology studies of tuberculosis [56] . Alternatively , these strains might have higher transmissibility , undescribed so far in leprosy but proven for some lineages of M . tuberculosis . Our earlier observation that reinfection or strain selection of M . leprae isolates of SNP-Type 4 was very frequent in relapse patients in Rio de Janeiro , a region predominant for SNP-Type 3 could be an example of that [14] . Identifying behavioral and environmental risk factors for developing leprosy is a difficult task because of the long incubation time of the disease ( 2–5 years for tuberculoid leprosy and 8–12 years for lepromatous leprosy ) . It is not easy to determine time and duration of exposure and onset of infection and risk factors for disease might change over time . Among 165 municipalities in the state of Ceará , a 300-fold difference in disease incidence was observed and associated with poverty , inequality , uncontrolled urbanization , population growth and low level of education [57] . The same group [44] also looked for socioeconomic , environmental and behavioral factors associated with leprosy in a case control study in four municipalities including that of Fortaleza; low education level , experience of food shortage at any time in life , frequent contact with natural bodies of water and infrequent changing of bed linen were associated with leprosy . Another study in this city concentrated on infection with M . leprae in the absence of clinical disease and demonstrated that higher levels of anti PGL-1 in patients without known contact with leprosy patients are much higher than reported elsewhere in the literature [58] . More recently , nasal carriage of M . leprae by PCR was observed in 67% of HHC but interestingly , 28% of persons living in richer part of the city were also positive . This is probably due to complex interaction between the populations at high and low risk for infection by leprosy . Domestic service and daily migration of the poor in houses of the upper class and richer parts of the city is still common [41] . An earlier spatial analysis in Ceará showed the highest density of disease is among the most urbanized and economically highest developed [59] . Our spatial analysis on genotype distribution did not demonstrate a distribution of clustering that was different from disease distribution in Fortaleza in general , showing that with the present data , there do not seem to be clear hot spots of ( recent ) transmission in the city . However , some neighborhoods were associated with the two biggest clusters , being group 12 ( Jacareacanga , Canindezinho , Conjunto Esperança and Manoel Sátiro ) and group 14 ( Bonsucesso and Vila Pery ) . We also observed that three of these neighborhoods ( Bonsucesso , Canindezinho and Conjunto Esperança ) presented patients with high BI ( note that only MB cases were submitted to genotyping ) and in a recent study on the social , educational and economic development of neighborhoods in Fortaleza , both were indicated as being among the poorest in the city ( www . ipece . ce . gov . br/publicacoes/Perfil%20Socioeconomico%20Fortaleza%20final-email . pdf ) . In addition , very recent data also demonstrate that both neighborhoods are hyperendemic ( > 4/10 . 000 ) for leprosy with high incidence in children less than 15 years of age ( 0 . 5-1/10 . 000 ) [60] . Some limitations of our study is that our sampling occurred during a relatively short period of time , that genotyping was performed only on 15% of the new MB patients and that PB patients were omitted from analysis . This might mask transmission links due to factors other than contact with MB patients and explain why a considerable proportion of the new patients were not aware of earlier contact with patients . Nonetheless , the most significant association with clustering was having positive bacilloscopy , which is in agreement with the long standing idea that transmission of leprosy is caused by close contact with MB patients . However , significance of this finding is weakened because the mean BI between groups with clustered and unique genotypes is almost the same , but again , only MB patients were submitted to genotyping . Definition of being MB or PB in the present study is based on Ridley-Jopling method and our results are in favor for maintaining this technique as part of the diagnostic procedure , contrary to the current recommendation of WHO to define PB and MB patients only on basis of number of lesions and nerve involvement . The significant association of clustering with patients having had contact with another case at work but not at time of diagnosis present could be due to the long incubation time for developing leprosy; however , a low number of patients reported contact at work . Although we could not establish a relation of cluster with the nature of the work or localization of the workplace , this needs further investigation because some working places harbor a large number of persons including undetected leprosy cases during long periods and could be hot spots of transmission . Some examples are metallurgic and car assembly factories , areas of civil construction , handicraft fairs and offices . Social interactions and the physical , residential and occupational environments have been suggested to be more conducive to transmission of a community in Qiubei , China [18] . This finding is not in line with our observation that having lived with a leprosy patient is associated with belonging to a non-cluster and to explain this , further research , eventually using whole genome sequencing is warrented . HHC have been described to be at higher risk for developing leprosy in several conventional epidemiologic studies but also in studies that performed M . leprae genotyping , including China [17] , Thailand [27] , Colombia [31] and India [24] . Although investigation of HHC is part of the leprosy program in Brazil , this is not always being performed and in Fortaleza in particular , this seems to be the case in about 50% of the patients [43] . The lack of association between clustering and house hold in the present study is probably due to the inclusion of new patients only and without contact investigation and inclusion of patients from the same house hold . Nonetheless , our observation of inversed association of sharing home with a leprosy case and cluster is surprising and needs to be better investigated . Another puzzling finding was the significant association between alcohol use and having M . leprae with a unique genotype . Several studies associated alcohol ( ab ) use as a risk factor for leprosy , including a case control study in Mato Grosso state [61] , Maranhao state [62] and with treatment abandonment in Tocantins [63] . This finding needs further investigation but again , the low number of patients in some analytical cells due to the paucity of biopsied patients and lack of specific questionnaire data could be partly responsible . Another issue are the different protocols used for collecting information about alcohol ( ab ) use . We also observed that some characteristics that are usually associated with higher risk for leprosy also had a tendency to be more pronounced in clustered patients . This was the case of clustering among males and later diagnosis at a later stage due to more reluctance to seek care among men as widely in Brazil . We also observed a tendency to have a higher disability grade in clustered patients . Higher disability grade reflects longer incubation time , bacillary load and time before diagnosis , therefore being able to infect more individuals . This is in concordance with the longer time delay between first observation of lesions and disease diagnosis reported in clustered patients . We conclude by referring to a very recent study that evaluated temporal trends in leprosy in Fortaleza for the period 2001 to 2012 [59] . Although there was a steady decrease in the number of new patients , from hyperendemic ( ≥4/10 , 000 ) in 2001 to highly endemic ( 2<4/10 , 000 ) in 2012 , the number of new patients in children less than 15 years old was steady and there was also noted a steady increase in the number of MB and of lepromatous patients since 2005 . Such data indicate both ongoing recent transmission including to children and late diagnosis in adults , reflected also by the rise in grade 2 disability ( from 6% to 9% in new patients ) . Given the chronic nature and natural history of the disease it is unlikely that there will be an improvement of these trends in the near future . Low levels of education , unfavorable socioeconomic conditions , and delayed presentation to the health system are factors that are generally associated with late diagnosis . This is in agreement with our data of high clustering levels and , demonstrating that recent transmission of leprosy is a serious problem in Fortaleza . The realization of a prospective molecular epidemiologic study in a complex setting like Fortaleza is difficult but we hope that a new study of longer duration , with higher intake of patients , collecting both skin biopsy and nasal swabs or biopsy , inclusion of HHC , a more detailed questionnaire including social network studies that might allow definition of risk factors for belonging to the same cluster , and finally investment in DNA extraction and more sensitive genotyping that allows inclusion of PB patients . As a final comment , we believe that , although whole genome sequencing of M . leprae genomes is still challenging because of the need of bacterial DNA enrichment , the technical expertise needed and the considerable cost , inclusion in future studies might be beneficial for better understanding of leprosy transmission .	Leprosy is a transmissible disease that is still endemic in several countries including in Brazil , a country with highly variable region associated incidence of disease . Fortaleza is a city in Northeast Brazil with high incidence and conventional epidemiology studies are suggestive for high levels of recent transmission . Genotyping of M . leprae allows the recognition of individuals that have been infected with the same strain ( called a cluster ) and therefore being suggestive for belonging to the same transmission network . In the present work , by analyzing genotypes of M . leprae in the skin lesion of multibacillary patients , we made observations that improve our knowledge on interpretation of genotypes and clusters and confirm the high levels of recent transmission in the city of Fortaleza . This is one of the few studies that used molecular epidemiology to look for risk factors for recent transmission of leprosy and to our knowledge , the first in Brazil . Our data support further investigation of the workplace as a source of infection , preferentially by a study designed on a larger number of patients and including analysis of M . leprae present in the nose .	[ "Abstract", "Introduction", "Materials", "and", "methods", "Results", "Discussion" ]	[ "biogeography", "mycobacterium", "leprae", "medicine", "and", "health", "sciences", "ecology", "and", "environmental", "sciences", "biopsy", "population", "genetics", "tropical", "diseases", "geographical", "locations", "surgical", "and", "invasive", "medical", "procedure...	2017	Genotyping of Mycobacterium leprae for better understanding of leprosy transmission in Fortaleza, Northeastern Brazil
Iron is essential for life . Accessing iron from the environment can be a limiting factor that determines success in a given environmental niche . For bacteria , access of chelated iron from the environment is often mediated by TonB-dependent transporters ( TBDTs ) , which are β-barrel proteins that form sophisticated channels in the outer membrane . Reports of iron-bearing proteins being used as a source of iron indicate specific protein import reactions across the bacterial outer membrane . The molecular mechanism by which a folded protein can be imported in this way had remained mysterious , as did the evolutionary process that could lead to such a protein import pathway . How does the bacterium evolve the specificity factors that would be required to select and import a protein encoded on another organism’s genome ? We describe here a model whereby the plant iron–bearing protein ferredoxin can be imported across the outer membrane of the plant pathogen Pectobacterium by means of a Brownian ratchet mechanism , thereby liberating iron into the bacterium to enable its growth in plant tissues . This import pathway is facilitated by FusC , a member of the same protein family as the mitochondrial processing peptidase ( MPP ) . The Brownian ratchet depends on binding sites discovered in crystal structures of FusC that engage a linear segment of the plant protein ferredoxin . Sequence relationships suggest that the bacterial gene encoding FusC has previously unappreciated homologues in plants and that the protein import mechanism employed by the bacterium is an evolutionary echo of the protein import pathway in plant mitochondria and plastids . TonB-dependent transporters ( TBDTs ) are 22-stranded β-barrel proteins integrated in the bacterial outer membrane . They contain a plug domain inserted inside the β-barrel that can be actively removed by the inner membrane protein TonB to provide access for the import of vitamins , chelated metals , and other cofactors essential to bacterial viability [1] . These various cofactors need to be small enough to pass through the internal channel of the TBDT , a channel whose internal diameter is approximately 20 Å [1–3] . TBDTs also serve as an outer membrane channel to initiate the import of bacteriocins , a class of protein antibiotics that are larger than the channel diameter but that have evolved to unfold in response to engagement with their specific target TBDT [4 , 5] . The genes for bacteriocins have coevolved with the genes for the TBDTs , producing an exquisitely specific and deadly system of protein transport to control bacterial population growth [6] . In the case of bacterial pathogens , iron is acquired from host tissues , in a process referred to as iron piracy [7] . This process too is mediated by TBDTs . In some cases , the iron is captured by small siderophores , the genes for which coevolved with those of the TBDTs such that the size of the siderophore conveniently fits within the constraints imposed by the channel formed by the TBDT [1–3] . However , there are also several well-documented examples in which iron is acquired by pathogens through direct extraction from host proteins [7 , 8] . Proteins such as ferredoxin , hemoglobin , and others are bigger than the TBDT channel . Nonetheless , under iron limitation , pathogens like Pectobacterium can remain viable and virulent using proteins such as ferredoxin , pirated from their plant host to supply iron [9] . How this important protein transport pathway is mediated by the bacterium had remained unknown and difficult to interpret from evolutionary considerations . The gram-negative bacterium Pectobacterium is a pathogen causing soft-rot disease in a range of plant types [10–12] . Strains of Pectobacterium across the globe are distinguished geographically , leading to suggestions that they have evolved to specifically rot-endemic plants [13] . Plants have complex , species-specific mechanisms for withholding iron from pathogens such as Pectobacterium [14–16] , and successful pathogens therefore need to develop species-specific countermeasures [7 , 17] . Defying these and other evolutionary relationships , either through agricultural development or inadvertently , plant species can cross continental barriers . By way of example , Australia was isolated from other land masses for approximately 100 million years and is , therefore , highly vulnerable to noxious weeds , given endemic plants have not adapted for competition , and endemic animals might not graze on introduced weed species . Thus , noxious weeds like Angled onion ( Allium triquetrum ) that have been inadvertently introduced into Australia gain a selective advantage in some environments [18] . However , a strain of Pectobacterium was recently isolated on these introduced weeds as a potential biological control agent , as it appears inactive against endemic plants yet highly active against Angled onion [19] . This Australian isolate of Pectobacterium carotovorum subsp . cartorvorum Waldee was shown to produce severe soft-rot symptoms by invading both cortical and vascular tissue of the weed [19] . To characterize this important isolate of Pectobacterium , we used whole genome sequencing . Inspired by previous studies on iron acquisition systems required for soft-rot by the phytopathogens , the genome-encoded complement of TBDTs was analyzed , and a fusA-fusC operon was identified . The fusC gene was found to be closely related to an orphan protein subfamily of plant proteins annotated as M16 proteases , which , in plants , are involved in organelle biology , including subreactions of the protein import pathway . The crystal structure of FusC from Pectobacterium demonstrated the structural characteristics of an M16 protease and further revealed a plausible import mechanism for ferredoxin , an iron-bearing protein from plants . We suggest a Brownian ratchet mechanism by which the phytopathogen binds and imports the plant protein in order to scavenge iron and thereby potentiate rot of the plant host . Species of Pectobacterium across the globe are distinguished geographically because they have evolved to specifically rot-endemic plants [13] . Phytopathogens can travel with introduced plant species , and whole genome sequencing of the recently identified Australian biocontrol isolate showed genome characteristics of a large group of isolates from China . This is in contradistinction to other Pectobacterium isolates from the Southern Hemisphere ( “New Zealand , ” “South Africa , ” “South America” ) , which form a wholly distinct clade ( Fig 1A ) . Recently , a ferredoxin receptor , FusA , was identified in the Scottish P . atrosepticum isolate SCRI1043 , encoded in an operon that functions in iron acquisition [9] . To determine whether FusA and other TBDTs for iron acquisition are conserved globally in Pectobacterium spp . , hidden Markov models ( HMMs ) were constructed . Genome sequence analysis using the HMM built to detect TBDTs revealed that the Australian isolate RMIT1 encodes 23 TBDTs , many of them typical of iron acquisition systems ( Table 1 ) . The most distantly related of these TBDTs shares 82% sequence identity to the ferredoxin receptor FusA from P . atrosepticum SCRI1043 [9] . Inspection of sequenced Pectobacterium genomes revealed that in each case downstream of the fusA gene is a gene called fusC , previously noted by Walker and colleagues as encoding a protein belonging to the M16 family of proteases [9] . Characteristic features of the M16 metalloprotease family of proteins include being a Zn2+-dependent yet ATP-independent protease , having a conserved architecture of two homologous 50-kDa domains that encompass the catalytic chamber , and using electrostatic-mediated interactions to capture substrate proteins [20–22] . The architecture predicted in the conceptual translation of the FusC protein of Pectobacterium suggests it would be part of the M16 protease family . In addition to being present in all complete genome sequences from isolates of Pectobacterium , the fusA-fusC genes are also conserved in two other species of plant pathogen—namely , Dickeya and Klebsiella . However , assessment of the wealth of complete genome sequence data on isolates of Klebsiella spp . , which can be commensals and pathogens of plants and can also cause disease in humans [23] , showed only a sporadic presence of fusC ( S1 Fig ) . The M16 family of proteases is found broadly across the domains Bacteria and Eukarya . To understand the sequence relationships within this important family of proteases , cluster analysis of sequences ( CLANS ) was used to classify family members in plants and bacteria . The CLANS analysis highlighted the known mitochondrial and plastid M16 protease subfamilies of mitochondrial processing peptidase ( MPP ) , stromal processing peptidase ( SPP ) , and presequence protease ( PreP ) ( Fig 1B ) , which function in protein import into these organelles [24] . Interestingly , the analysis revealed that plant genomes also encode a previously undescribed M16 protease subfamily that we refer to as plant FusC ( Fig 1B ) . The nomenclature is appropriate given that the FusC from Pectobacterium and Klebsiella are most closely related to the novel plant protein ( Fig 1B ) . A number of M16 proteases previously characterized from bacteria , such as PqqL and YhjJ isoforms , are similar to the plant FusC , while the PqqF- and PtrA-type bacterial M16 proteases are more distinct , according to the CLANS data ( Fig 1B ) . M16 proteases are typically either a dimer of approximately 50-kDa two-domain subunits [25] or an approximately 100-kDa monomer composed of four domains [26] . SignalP [27] predicts that FusC from Pectobacterium has a 25-residue N-terminal signal sequence that would direct it into the periplasm and , once processed , would produce a mature protein of approximately 100 kDa . Recombinant FusC was engineered for expression in the periplasm of Escherichia coli , and the purified protein was shown to be a monomer in solution ( S2 Fig ) . FusC was crystallized and the structure determined . The structure of the FusC monomer demonstrates conserved features seen in the MPP dimer ( Fig 2A ) . The crystal structure demonstrated that FusC has the four characteristic domains of an M16 protease , creating the entire “clamshell” structure from a single polypeptide ( Fig 2A ) . These four domains of an M16 protease fit together through interactions between the β-sheet and α-helix protruding from the opposing domain , creating half of the clamshell ( Fig 2B ) . The first of these domains ( Fig 2C , S3 Fig ) would correspond to the catalytically active β-subunit of MPP [28] , housing the active site residues with a metal ion coordinated by two histidine residues ( H80 and H84 ) and a glutamic acid ( E165 ) . The second domain has an analogous pocket but does not have residues that could coordinate a metal ion for catalysis ( S3 Fig ) . In FusC , domain 3-domain 4 ( corresponding to the α-subunit of MPP , which is catalytically inactive [28] ) are likewise unable to form an active site ( S3 Fig ) . In protein import pathways , the concept of a Brownian ratchet is invoked to explain a means to energize protein movement across a membrane , particularly when some degree of protein unfolding is involved [31] . Such a mechanism requires that a factor on the trans side of the membrane can clamp onto a short segment of the substrate protein to transform the nondirectional Brownian motions of the receptor-bound substrate into a vectorial motion across the membrane [32] . Being in the periplasm , FusC would be located on the trans side of the bacterial outer membrane and so was tested for the characteristics required of a factor mediating a Brownian ratchet . Small-angle X-ray scattering ( SAXS ) analysis of FusC demonstrated that the protein has interdomain flexibility in solution ( Fig 4A ) . To explore the possibility of FusC providing a Brownian ratchet for ferredoxin import , it was subjected to size-exclusion chromatography coupled with small-angle X-ray scattering ( SEC-SAXS ) . In the absence of ferredoxin , the scattering curves observed for both FusC and FusC ( E83A ) are identical ( Fig 4A and 4B ) . The maximum dimension ( Dmax ) of Apo-FusC in solution is 129 Å ( Fig 4C–4F ) . This corresponds with the dimensions of two M16 protease clamshells stacked end on end , suggesting FusC can adopt a fully open conformation . This observation contrasts with the partially closed clamshell structure observed in the FusC:ferredoxin crystal structure , which has a Dmax of 94 Å ( Fig 4G ) . The Dmax of FusC and its flexibility in solution suggest that its clamshell domains can adopt a range of conformations from fully open to at least partially closed . When FusC was mixed with a 3:1 molar excess of ferredoxin prior to SEC-SAXS , no changes were observed in its scattering profile . When the equivalent experiment was run using the FusC ( E83A ) mutant in the presence of ferredoxin , the scattering of FusC ( E83A ) altered , with a decrease in the biomodal nature of the population of FusC ( E83A ) in solution ( Fig 4A ) . The ferredoxin-induced perturbance of the FusC ( E83A ) spectra was small , suggesting that the FusC:ferredoxin complex disassociates over the time course of the size-exclusion chromatography ( SEC ) experiment . To test this hypothesis , the binding kinetics of FusC ( E83A ) with ferredoxin were determined using isothermal titration calorimetry ( ITC ) ( Table 2 ) . The Kd of the FusC ( E83A ) :ferredoxin complex was found to be 1 . 9 μM ( Fig 4E ) . Given this affinity and the prospect of dissociation in this range , static-SAXS was utilized to further interrogate FusC:ferredoxin complex formation in solution . Ferredoxin was titrated into FusC ( E83A ) at a range of concentrations from 0:1 to 10:1 molar ratio of ferredoxin to FusC ( E83A ) , and a change in scattering of FusC was observed . With increasing ferredoxin concentration , the scattering of FusC ( E83A ) shifted from a biomodal to monomodal distribution that saturated at a 3:1 molar ratio of ferredoxin to FusC ( Fig 5A–5F ) . While the mass of FusC increased with ferredoxin binding , the Dmax of the particle remained constant , demonstrating that FusC ( E83A ) remains flexible even when bound to ferredoxin . Considering both the observed crystal contacts and the various SAXS analyses , we conclude that FusC can clamp onto ferredoxin and that it does so via a C-terminal segment of its ferredoxin substrate . Phytopathogens can have myriad strategies to dominate their necrotic environments . Soft rot of host tissues depends on specific protein-secretion systems to deliver proteins with hydrolytic activity into the tissues , and growth and replication of bacteria in this rot are dependent on access to iron from the rotting tissue [33 , 34] . Under conditions of iron limitation , growth of a model species of Pectobacterium was shown to be supported by plant ferredoxin as an iron source [35] . This has led to the understanding that a pathway for iron acquisition from host proteins is a survival strategy for Pectobacterium , and the outer membrane receptor responsible for ferredoxin import into P . atrosepticum isolate SCRI1043 is the TBDT FusA [9] . How ferredoxin , a tightly folded protein stabilized by its iron cofactor , could be imported across the bacterial outer membrane into the bacterial periplasm had remained a mystery of cell biology . Given the size of ferredoxin ( minimal diameter 35 Å ) and the unchangeable internal diameter of the channel in the β-barrel protein FusA ( maximal diameter of 34 Å after removal of the plug domain ) , at least some degree of unfolding of ferredoxin would be required for its import across the Pectobacterium outer membrane . How does a bacterium , even a specialist phytopathogen , evolve a means to transport a plant iron–containing protein across its outer membrane ? Clues come from the “greasy slide” mechanism by which oligosaccharides can be imported through the maltoporin LamB [36] , but , in every case known , translocation of a protein across a membrane requires complex , energy-requiring processes , catalyzed by factors that can recognize the protein substrate of interest . Our analysis suggests that Pectobacterium spp . have acquired , either by horizontal gene transfer or by convergent evolution , the FusC subclass of M16 proteases related to the plant FusC proteins found in their hosts . In other Brownian ratchet models , protein translocation across a membrane is achieved through Brownian motion “breathing” of a terminal segment of the substrate protein [31 , 32] . Such a mechanism could be invoked for the C-terminal segment of ferredoxin , which was observed fixed to the noncatalytic F1 site on FusC . This type of Brownian motion in the C-terminal segment may be enhanced once ferredoxin contacts the “glove” of FusA [9] . By analogy to the process of protein import into mitochondria [31 , 32] , we suggest a model whereby a noncatalytic ferredoxin-binding site in FusC would assist driving ferredoxin import as a Brownian ratchet ( Fig 6 ) . We are unaware that any of the other clades of bacterial M16-peptidases have such a function . The concept of a Brownian ratchet provides for a net displacement of a segment of polypeptide through a translocation channel by capturing the polypeptide on the trans side of the membrane [37] . This capture requires some form of clamp to lock on to the substrate polypeptide , which would otherwise diffuse freely backward and forward through the channel [31 , 32 , 37] . Even for very large polypeptides , theoretical considerations show that the driving force provided by a Brownian ratchet is more than enough to ensure vectorial translocation across the membrane [37] . In the specific case of a small protein like ferredoxin , with a diameter only slightly greater that the translocation pore itself , unfolding of the C-terminal segment would both decrease the substrate diameter and provide a segment of polypeptide to be captured on the trans ( periplasmic ) side of the channel ( Fig 6 ) . Like the M16 proteases found in bacteria , plants and other eukaryotes also have M16 proteases . Two well-characterized examples , MPP and SPP , work as processing peptidases in protein import pathways . The MPP ( formerly MPP/PEP ) binds substrate proteins in transit through the translocon of the inner mitochondrial membrane [28 , 38 , 39] . In plastids , the SPP binds the precursor form of ferredoxin and clips the stromal targeting sequence to generate the mature form of ferredoxin , which can then be folded into the active iron-sulfur protein to participate in photosynthesis [40] . While the function of PreP in degrading peptides , damaged , and nonnative proteins in organelles is well characterized [24] , there is as yet no information on the function of the plant FusC that was discovered in our study . It is tempting to speculate that plant FusC might be involved in the turnover of ferredoxin , given the extensive quality control that is required to limit levels of other oxidatively damaged proteins in the process of photosynthesis [41] and the equivalent role of the malarial parasite M16 protease falcilysin that catalyzes turnover of the iron protein hemoglobin [42] . The components that mediate protein import into eukaryotic organelles , such as mitochondria and plastids , evolved from proteins present in their bacterial ancestors [43–45] . We suggest that the import of ferredoxin into Pectobacterium makes use of a mechanism analogous to that used in eukaryotic organelles , even using structurally related elements and a concept borrowed from the more complex protein import systems of plastids and mitochondria . Enterobacteriaceae M16 protein sequences ( Dataset S1 ) were obtained from InterPro IPR011765 “Peptidase M16 , N-terminal , ” and redundancy was minimized using CD-HIT with a 0 . 98 cutoff [46] . FusC-related plant sequences were initially identified through a BLAST search using Pectobacterium FusC as a query sequence . In order to cover a broad phylogenetic distance , plant M16 protein sequences were manually chosen such as to include species of green algae , moss , and monocotyledon and dicotyledon plants ( S3 Table ) . A uniform , representative set of plant species ( i . e . , the same genome projects ) was used for the MPP , SPP , plant FusC , and PreP sequences . The plant sequences were combined with the enterobacterial M16 sequences , and the sequences were classified by an all-against-all BLAST , clustered based on pairwise similarities , and visualized with CLANS with a E-value cut off of 1 × 10−15 [47] . While it has not been studied in detail , sequence characteristics of the plant FusC protein from Arabidopsis thaliana can be found at the genome project site ( https://www . arabidopsis . org/servlets/TairObject ? type=locus&name=AT5G56730 ) . The genome of P . carotovorum subsp . cartorvorum Waldee RMIT1 was sequenced by the Wellcome Trust Sanger Institute genome sequencing facility . Illumina sequencing libraries ( Illumina , San Diego , CA , United States of America ) were prepared with a 450-bp insert size and were sequenced ( Illumina HiSeq2000 ) . The long paired-end reads ( 100 bp ) were assembled [48] ( https://github . com/sanger-pathogens/ ) and annotated using prokka [49] . Whole genome sequence data were annotated with prokka ( http://www . vicbioinformatics . com/software . prokka . shtml ) and yielded the following summary statistics: Contigs: 99—Bases: 4995997—tmRNA: 1—rRNA: 4—CDS: 4461—tRNA: 55—repeat_region: 3 . To analyze the phylogenetic relationship of P . carotovorum subsp . cartorvorum Waldee RMIT1 compared with other related strains , a tree was rendered using iTOL ( http://itol . embl . de ) . Whole genome sequence data for the other Pectobacterium spp . were accessed ( https://www . ncbi . nlm . nih . gov/genome/genomes/1799 ) via NCBI . Ultimately , a total of 37 sequenced genomes were used for constructing the phylogenetic trees . To map plant FusC sequences in respect of phylogenetic relationships of globally sampled Klebsiella genomes , the genome sequence data and associated metadata—including species , sample source , location , and date of isolation—were obtained from the NCBI Pathogen Detection project database via micro-react . org ( https://microreact . org/project/ncbi-klebsiella ) and visualized using Figtree v1 . 4 . 3 ( http://tree . bio . ed . ac . uk/software/figtree/ ) . To identify Klebsiella strains carrying peptidase homologues , we used NCBI tblastn with an E-value cutoff of 1e−05 to query a representative M16 peptidase sequence ( genbank ID: WP049102367 . 1 ) against the NCBI nucleotide collection databases , and the location of the positive BLAST query results were mapped onto the Klebsiella phylogeny . TBDT sequences were collected from UniProt and literature [50 , 51] , duplicates were removed , and the remaining 87 samples were aligned with T-Coffee [52] to generate the formatted alignment results as input file for HMM training . Then , we used HMMER [53] to train the HMM and employed this model to search against the Pectobacterium genome dataset . Finally , 23 putative sequences with high confidence were retrieved ( Table 1 ) . After comparing them with the original dataset , we found no overlap . We added these putative sequences into the original dataset and trained a new HMM to iteratively search against the Pectobacterium genome dataset . No further putative sequences were retrieved in this subsequent search . A DNA fragment encoding the FusC open reading frame lacking the signal peptide was amplified to contain NcoI and XhoI restriction sites and cloned into a modified pET20b vector with sequence that would add an N-terminal polyhistidine tag followed by a TEV cleavage site to the encoded recombinant protein . The resulting vector was designated pETFusC . To express a FusC ( E83A ) mutant , whole plasmid mutagenesis was utilized [54] . To amplify pETFusC incorporating a single base pair change from “A” to “C” ( at the codon indicated in bold ) , PCR was used with the overlapping oligonucleotide primers: Forward—GTAGCGCACATGGTCGCACACATGGTTTTTCGTG; Reverse—CACGAAAAACCATGTGTGCGACCATGTGCGCTAC . The product from this reaction was digested with DpnI for 1 h at 37 °C and then transformed into E . coli Top10 cells . Plasmid DNA was prepared from these colonies and sequenced to confirm the A to C mutation: the resulting plasmid was designated pETFusCE83A . Both vectors were transformed into E . coli BL21 ( DE3 ) C41 for protein expression . Protein expression was performed in terrific broth ( 12 g tryptone , 24 g yeast extract , 61 . 3 g K2HPO4 , 11 . 55 g KH2PO4 , 10 g glycerol ) containing 100 mg . ml−1 ampicillin for plasmid selection . Bacterial cultures were incubated at 37 °C until OD600 of 1 . 0 , then protein expression was induced with 0 . 3 mM IPTG , and incubation continued for 14 h at 25 °C . Cells were harvested by centrifugation and lysed using a cell disruptor ( Emulseflex ) in Ni-binding buffer ( 50 mM Tris , 500 mM NaCl , 20 mM imidazole [pH 7 . 9] ) containing 0 . 1 mg . ml−1 lysozyme , 0 . 05 mg . ml−1 DNase1 , and cOmplete Protease Inhibitor Cocktail ( Roche ) . The resulting lysate was clarified by centrifugation and applied to Ni-NTA agarose , followed by washing with 10× column volumes of Ni-binding buffer and elution of protein with a stepwise gradient of Ni-gradient buffer ( 50 mM Tris , 500 mM NaCl , using steps of 25 , 50 , 125 , and 250 mM Imidazole [pH7 . 9] ) . Fractions containing eluted recombinant protein were pooled and applied to a 26/600 Superdex 200 size-exclusion column equilibrated in SEC buffer ( 50 mM Tris , 200 mM NaCl [pH 7 . 9] ) . Fractions from SEC were pooled and incubated with 0 . 5 mg/ml−1 TEV protease and 1 mM DTT at 20 °C for 4 h to cleave the N-terminal 10×His tag . This solution was then passed through a 1-ml Ni-NTA agarose column , with the flow-through containing purified , TEV-cleaved FusC collected . Purified FusC was concentrated to 10 mg/ml and then snap frozen for storage at −80 °C . As a model ferredoxin , the protein from the plant A . thaliana was produced using vector previously described [9] , with the protein purified using the same protocol as for FusC . Purified FusC alone or in combination with ferredoxin ( 3:1 molar ratio ferredoxin to FusC ) was screened for crystallization conditions using commercially available screens ( approximately 800 conditions ) . No crystals were obtained for FusC alone; however , crystals grew in a number of conditions with FusC in combination with ferredoxin . Crystals of FusC only grew in both the presence of both ferredoxin and 3−5 mM EDTA . The presence of FusC and ferredoxin in the crystals was confirmed by collecting crystals with well solution and melting them in H2O for analysis by SDS-PAGE . A 10-kDa protein corresponding in size to intact ferredoxin was confirmed to be ferredoxin by mass spectrometry ( Fig 4A ) . An initial crystallization condition of 0 . 1 M Bis-Tris propane , 0 . 2 M NaK phosphate , 20% PEG 3 , 350 pH 6 . 5 was chosen for optimization . Crystals were cryoprotected by soaking for 10 min in crystallization solution plus 20% glycerol with or without a small quantity of powdered ethylmercury phosphate and flash cooled in liquid N2 . Native data were collected at 100 °K at the Australian synchrotron and processed in the space group P22121 to 2 . 7 Å . Single-wavelength anomalous dispersion ( SAD ) data from ethylmercury phosphate–soaked crystals were collected using the same parameters as native data , yielding the same space group with data processed to 2 . 3 Å . The catalytically inactive FusC ( E83A ) mutant was crystallized in the presence of ferredoxin as with wild-type FusC , however , without the 3−5 mM EDTA required for crystallization of wild-type FusC . Crystals grew in 0 . 2 M NaK phosphate , 20% PEG3350 . Crystals were flash cooled and data collected as per wild-type FusC with diffraction data processed to 1 . 9 Å with the same space group and unit cell parameters as wild-type protein . Phenix autosol was used to locate heavy atom sites in HgSAD data to perform phasing and density modification [55] . The best phasing solution consisted of 13 Hg sites , most of which had very low occupancy ( <0 . 2 ) , with 4 higher occupancy sites between 0 . 35 and 0 . 45 . This partial derivatization provided poor initial phases; however , density modification yielded interpretable maps . The FusC:ferredoxin complex structure was built using programs from the Phenix package in addition to manual building using Coot [55 , 56] . This FusC model was then used for molecular replacement , using Phaser , of both FusC and FusC ( E83A ) datasets [57] . These models were then built and refined [58] . Purified concentrated FusC ( E83A ) and ferredoxin were dialyzed extensively against the same reservoir of ITC buffer ( 20 mM HEPES , 50 mM NaCl [pH 7 . 5] ) . The concentration of proteins post dialysis was determined using absorbance at 280 nM and BCA assay . Proteins were then diluted in ITC buffer to reaction concentrations of 20 μM for FusC ( E83A ) and 200 μM ferredoxin . Protein concentrations were estimated again , following dilution . All ITC experiments were performed using a MicroCal VP-ITC calorimeter . Ten μl ferredoxin was sequentially titrated into the ITC reaction chamber containing FusC ( E83A ) . Negative heats of binding were observed , and 27 titrations were performed per reaction , allowing the ferredoxin concentration in the chamber to reach saturation . Control reactions were performed , with titration of either buffer into FusC ( E83A ) or ferredoxin into buffer . These reactions yielded no significant heats of dilution . All reactions were repeated 3 times; binding parameters were calculated from the average of the two best reactions . Representative heats of binding from a single reaction are presented . SEC-SAXS was performed using Coflow apparatus at the Australian Synchrotron [59 , 60] . Purified FusC or FusC ( E83A ) +/− ferredoxin were analyzed at a preinjection concentration of 100 μM for FusC and 300 μM ferredoxin . Chromatography for SEC-SAXS was performed at 22 °C with a 10/30 Superdex S200 column , at a flow rate of 0 . 4 ml/min in 50 mM Tris , 100 mM NaCl , 5% glycerol , and 0 . 2% sodium azide [pH7 . 9] . The inclusion of glycerol and azide were essential to prevent capillary fouling due to photo-oxidation of buffer components . Scattering data were collected for 1-s exposures over a q range of 0 . 0 to 0 . 3 Å−1 . A buffer blank for each SEC-SAXS run was prepared by averaging 10–20 frames pre- or postprotein elution . Scattering curves from peaks corresponding to FusC or FusC ( E83A ) were then buffer subtracted , scaled across the elution peak , and compared for interparticle effects . Identical curves ( 5–10 ) from each elution were then averaged for analysis . Data were analyzed using the ATSAS package , Scatter , and SOMO solution modeler [61] . For static-SAXS—titrations of ferredoxin into FusC ( E83A ) —the purified proteins were buffer matched in 50 mM Tris , 100 mM NaCl , 5% glycerol [pH 7 . 9] , and a series of samples were prepared with FusC at a constant concentration of 30 μM , and ferredoxin at a range of concentrations 0–300 μM . A series of blanks was also prepared with ferredoxin in buffer at a range of concentrations 0–300 μM . Scattering for samples was collected at 22 °C over a q range of 0 . 0 to 0 . 4 Å−1 , with exposures of 1 s , and scattering from the blank of the corresponding ferredoxin concentration was subtracted . Data were analyzed as for SEC-SAXS experiments . Sedimentation velocity ( SV ) was carried out in a Beckman Coulter Optima analytical ultracentrifuge using an An-50 Ti 8-hole rotor . FusC ( 370 μl ) at concentrations ranging from 0 . 25 to 2 mg . ml−1 was loaded into a 12-mm path-length centerpiece and centrifuged at 40 , 000 rpm for approximately 6 h at 20 °C . Scans were collected every 20 s using absorbance optics ( at 230 , 240 , and 280 nm; a radial range of 5 . 8–7 . 2 cm; and radial step size of 0 . 005 cm ) , using a buffer containing 50 mM Tris , 200 mM NaCl , pH 7 . 9 . Data were analyzed with SEDFIT , using the continuous c ( s ) distribution model [62] . SEDNTERP was used to calculate the partial specific volume , the buffer density , and viscosity at 15 °C and 20 °C .	Earth’s carbon cycle depends on saprophytic microbes to rot old or diseased plant matter and recycle carbon from that biomass . Some bacteria ( phytopathogens ) have evolved to cause disease and rot in even healthy plants and may have utility as biological control agents against noxious weeds . To understand the mechanisms driving each of these scenarios has significance in environmental engineering and agriculture . Access to iron is a limiting factor for bacteria-mediated plant rot . Here , we show how a plant-pathogenic bacteria has reevolved a mechanism , analogous to the protein import pathways that evolved in plant plastids and mitochondria , to import the plant iron–bearing protein ferredoxin from plant tissue . The study is based on structural and biophysical characterization of a key M16 family protease , FusC , resident inside the bacterial outer membrane .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "crystal", "structure", "markov", "models", "enzymes", "condensed", "matter", "physics", "enzymology", "membrane", "proteins", "crystals", "plant", "science", "mathematics", "outer", "membrane", "proteins", "materials", "science", "plant", "pathology", "crystallography", ...	2018	FusC, a member of the M16 protease family acquired by bacteria for iron piracy against plants
The vertebrate cranium is a prime example of the high evolvability of complex traits . While evidence of genes and developmental pathways underlying craniofacial shape determination is accumulating , we are still far from understanding how such variation at the genetic level is translated into craniofacial shape variation . Here we used 3D geometric morphometrics to map genes involved in shape determination in a population of outbred mice ( Carworth Farms White , or CFW ) . We defined shape traits via principal component analysis of 3D skull and mandible measurements . We mapped genetic loci associated with shape traits at ~80 , 000 candidate single nucleotide polymorphisms in ~700 male mice . We found that craniofacial shape and size are highly heritable , polygenic traits . Despite the polygenic nature of the traits , we identified 17 loci that explain variation in skull shape , and 8 loci associated with variation in mandible shape . Together , the associated variants account for 11 . 4% of skull and 4 . 4% of mandible shape variation , however , the total additive genetic variance associated with phenotypic variation was estimated in ~45% . Candidate genes within the associated loci have known roles in craniofacial development; this includes 6 transcription factors and several regulators of bone developmental pathways . One gene , Mn1 , has an unusually large effect on shape variation in our study . A knockout of this gene was previously shown to affect negatively the development of membranous bones of the cranial skeleton , and evolutionary analysis shows that the gene has arisen at the base of the bony vertebrates ( Eutelostomi ) , where the ossified head first appeared . Therefore , Mn1 emerges as a key gene for both skull formation and within-population shape variation . Our study shows that it is possible to identify important developmental genes through genome-wide mapping of high-dimensional shape features in an outbred population . Understanding the evolutionary processes that have generated and maintained morphological diversity in nature is a long-standing goal in biology . The cranium and mandible of vertebrates is a good example of such diversity . The fact that the cranial and mandible bones have to be integrated with the brain and sensory systems , as well as with the respiratory and digestive systems , makes this structure a prime example of both high integration and high evolvability . Although information about genes and developmental pathways involved in shape determination keeps accumulating , we are far away from understanding the genotype-phenotype map translating genetic variation into craniofacial shape variation [1] . To approach this question , here we aim to identify the genetic factors underlying such morphological differences . Previous experimental work has explored the genetic basis of craniofacial variation in a range of species , including Darwin’s finches [2–4] , cichlids [5 , 6] , dogs [7–9] , and mice [10–16] . There has also been some recent work on natural facial variation in humans [17–20] . Much of this work has been made possible by developments in geometric morphometrics , which provide the techniques for quantifying subtle shape variation [21] . Combined with increasing availability of genomics resources for mice , this has made possible genome-wide studies of natural shape variation in mice [12] . Early work in mice has focused mainly on the mandible . This is a well-established model for the study of complex traits because mandible shape can be approximated in 2 dimensions [22 , 23] . Several quantitative trait loci ( QTL ) studies have investigated mandible shape variation , and have identified several genomic regions underlying 2D variation in this trait , mostly in crosses of inbred laboratory strains [13 , 24 , 25] . Although the skull has received less attention due to its higher complexity and the difficulty of defining appropriate phenotypes ( 2D vs 3D ) , recently Burgio et al . [15] , Pallares et al . [12] , and Maga et al . [10] have successfully identified genomic regions underlying 3D skull variation in mice . In our previous study [12] , we used genome-wide association ( GWAS ) based on natural recombinants from a hybrid zone between two subspecies of the house mouse . This enabled us to identify candidate skull and mandible shape loci with much higher resolution than conventional QTL studies in mice ( e . g . [10] ) . Here , we approach the question from a micro-evolutionary perspective by analyzing within-population shape variation . The utility of studying phenotypic variation at the within-population level is well acknowledged , as it permits one to focus on within-species genetic contributions to phenotypic variation [26] . We use a population of “Carworth Farms White” ( CFW ) outbred mice , whose suitability for genome-wide mapping was previously described [27–29] . Recently developed genomic resources for this population allow for QTL mapping on autosomal chromosomes ( see Methods ) . The CFW mice were originally derived from a small number of Swiss mice , and have been maintained for dozens of generations as an outbred colony with a large breeding population that avoids crosses between closely related individuals [27 , 30 , 31] . Importantly , the mice used in this study show little evidence for population stratification or cryptic relatedness , which simplifies the analysis and interpretation of genetic variation contributing to quantitative traits . The high number of recombination events in the history of this population has resulted in small linkage blocks , which , together with the above mentioned features , result in high mapping resolution [27] . Heritability of each skull and mandible PC was calculated using the standard additive polygenic model ( S2 and S3 Tables and Fig 1 ) . The heritability values we report here are “SNP heritability” [32]; that is , the estimate of the proportion of phenotypic variance explained by all available SNPs . All skull and mandible PCs exhibit substantial contributions from the additive genetic variance component; only two PCs have heritability values lower than 20% . Mandible size has SNP heritability of 36 . 4% ( 95% confidence interval 16 . 4–56 . 4 ) , and skull size of 35 . 4% ( 15–55 . 8 ) . To summarize the heritability for mandible and skull shape with a single statistic , we calculated a weighted average of the SNP heritability of individual PCs ( see Methods and Fig 1C and 1D ) . These “total heritability” values are 43 . 6% for mandible shape and 42 . 4% for skull shape . This statistic is equal to the proportion of the bar chart that is shaded dark gray in Fig 1C and 1D . We also checked whether the proportion of total phenotypic variation explained by each PC is correlated with our SNP heritability estimates . This correlation is not strong , but significant; mandible , r2 = 0 . 14 , p-value = 0 . 034; skull , r2 = 0 . 16 , p-value = 0 . 034 ( Fig 1A and 1B ) . Partitioning the variance by chromosome shows almost all chromosomes contribute to shape variation ( Fig 2 ) . We find also a correlation with chromosome size , as expected , but this is only statistically significant for mandible shape ( Fig 2B ) . In our previous study we found a highly significant correlation for both , mandible and skull shape [12] . The weaker correlation in the present study is likely due to lower and somewhat uneven marker coverage , which is in itself not strongly correlated with chromosome length ( S2 Fig ) . In particular , chromosome 16 is underrepresented with respect to marker coverage; the fact that this chromosome contributes very little to the phenotypic variance ( Fig 2 ) could be due to technical limitations regarding SNP identification , or to this chromosome indeed harboring little variation . Out of the 22 PCs used to map skull shape , 12 PCs had at least one significant QTL; and out of the 20 PCs used in the mapping of mandible shape , 7 PCs had a significant QTL ( see Fig 3 and Tables 1 and 2 ) . 17 QTLs were identified for skull shape variation ( Table 1 ) , and eight QTLs for mandible shape variation ( Table 2 ) . One QTL was associated with mandible centroid size , and no QTLs were identified for skull centroid size . The shape traits associated with the “peak SNPs” ( SNPs with lowest p-value ) are depicted in Figs 4 and S4–S8 . In some cases multiple QTLs were found in the same chromosome ( chr2 , 5 , 9 , 11 , 13 ) but associated with different PCs ( Fig 3 ) . Four QTLs were associated with more than one PC; interestingly , two of them were associated with PCs from skull and mandible ( chr5 and chr9 ) . Together the 17 QTLs identified for skull shape explain 11 . 4% of skull variation . The 8 QTLs for mandible shape together explain 4 . 4% of mandible variation . The effect size of individual SNPs ranges from 0 . 02 to 1 . 13% of the total phenotypic variation ( Fig 5 and Tables 1 and 2 ) . The single QTL found for mandible size explains 4 . 1% of size variation . We compiled a list of 115 protein-coding genes within the craniofacial shape QTLs ( S4 and S5 Tables ) . For most of the regions compelling candidate genes could be identified based on previously reported craniofacial phenotypes or previous evidence for a role in bone morphogenesis . Table 3 lists the functional information available for these genes . Most candidate genes listed in this table are transcription factors or known regulators of developmental signaling cascades . The total heritability estimated in this study , ~43% for craniofacial shape and ~36% for craniofacial size , corresponds to SNP heritability estimates . In humans , SNP heritability is considered an underestimate of the narrow sense heritability because it does not take into consideration rare alleles [90] . However , in the CFW population used here , rare alleles are expected to be uncommon due to the bottleneck when the population was started , the limited number of generations since the bottleneck , and the modest effective population size . Thus , SNP heritability estimates in this population may be closer to narrow sense heritability . Using a population of wild derived mice and a 3D approach , craniofacial SNP heritability was estimated as 65% for shape and 72% for size [12] . Using a pedigree of wild caught mice and a 2D approach , the heritability of mandible shape and size was found to be 0 . 61 and 0 . 49 , respectively ( Siahasarvie and Claude , personal communication ) . Regardless of the method or the experimental design , the heritability estimates for mandible size and shape in mice are high . It remains to be seen if the same pattern is true for the skull; pedigree-derived data need to be collected . In a recent study of wild soay sheep , the SNP heritability of mandible length was estimated to be 53% [91] . Human studies estimate a narrow sense heritability of ~0 . 8 for facial morphology [18] . Although more data are needed , a pattern emerges from these studies: the form ( i . e . shape and size ) of craniofacial structures is a highly heritable trait . The resolution achieved here was much higher than the resolution from traditional F2 crosses . Although the resolution was still not high enough to conclusively pinpoint individual genes , it was nonetheless possible to explore all genes within the QTL regions and often we identified a single candidate gene for which previous relevant phenotypic information existed . Moreover , 77% of the regions overlap with previous studies ( see Tables 1 and 2 ) ; 7 of them with QTL regions derived from a backcross [10] , 19 of them overlap with some of the ~ 4 , 000 enhancers active during craniofacial development in the mouse [11] , and 1 region overlaps with a GWAS using a wild-derived population of mice [12] . Such overlaps cannot be explained by chance only ( S3 Fig ) . Enhancers are DNA sequences that positively regulate the expression of nearby genes; therefore the high overlap with the enhancer dataset could indicate that some of the SNPs identified in this study tag a causal variant located in the regulatory region of the candidate genes . The possibility of representing visually the shape traits associated with each SNP ( see Figs 4 and S4–S8 ) allows the identification of specific craniofacial regions affected by the candidate genes . This information will become very valuable in future studies exploring the developmental role of such genes in craniofacial shape determination . Many of the candidate genes are genes with reported craniofacial phenotypes . However , most of them were previously not quantitatively assessed and therefore knowledge of their specific effects on craniofacial shape variation requires a geometric morphometrics analysis of mutant mice . Such an analysis can be done in heterozygous knockout mice for the gene of interest . In this way the genetic alterations and their phenotypic effects are less drastic compared to the knockout of both alleles , and therefore closer to natural variability within populations [92] . Several other genes are new candidates for craniofacial shape determination; they are involved in diverse processes of bone formation but have not been directly implicated in craniofacial development . We found two pairs of genes involved in the same signaling pathway; Sh3pxd2b and Gab1 are part of the epidermal growth factor signaling pathway–EGF; Sh3pxd2b regulates EGF-mediated cell migration [78] , and Gab1 is involved in EGF-mediated cell growth [52] . Mn1 and Cldn18 are involved in the RANK-RANKL-OPG signaling pathway; Mn1 regulates RANKL expression by stimulating RANKL’s promoter [66] , and Cldn18 regulates RANKL-induced differentiation of osteoclasts [35] . Among the candidate genes , Mn1 is a particularly interesting one . It was originally discovered for being involved in a myeloid leukemia phenotype and it was therefore named meningioma 1 [93] . This gene has the largest effect size in our screen ( Tables 1 and 2 ) and is associated with many PCs in the skull and in the mandible ( regions 4 , 8 , 12 in Table 1 , regions 1 and 2 in Table 2 , and S11 Fig ) , thus being also the most pleiotropic gene in our study . Knockout studies of Mn1 revealed that the leukemia phenotype of the gene is only a by-product of a particular fusion with another gene , while the core function of Mn1 lies in regulating the development of membranous bones of the cranial skeleton [65] . Intriguingly , Mn1 is an orphan gene specific to bony vertebrates ( Euteleostomi ) ( S12 Fig ) , a taxon characterized by the formation of bones and a suture-structured head skeleton . The origin of such orphan genes is connected to the emergence of evolutionary novelties [94] , and the Mn1 knockout phenotype in mouse suggests that it plays a crucial function in the emergence of a vertebrate novelty–the bony head . Hence , Mn1 has the hallmarks of a key gene in the genetic architecture of craniofacial development and shape determination . The fact that it also emerges out of our genome-wide analysis lends credence to the notion that the approach is suitable to detect relevant genes even for highly polygenic phenotypes . There are long standing discussions about how to deal experimentally with polygenic traits and their implications for understanding the evolution of such traits [95 , 96] . Genome-wide association studies have certainly moved us forward in this respect . Even relatively simple quantitative phenotypes like human height have a highly polygenic nature [97 , 98] . Still , when a sufficiently powerful experimental design is used , key regulatory pathways influencing this phenotype can be identified [97 , 98] . The natural variants of these pathways have individually small effects , but knockouts of these genes can have large effects . Here we have shown that we have a similar scenario regarding craniofacial shape , which is a complex phenotype with a highly polygenic architecture . It is encouraging to see that even under such seemingly adverse genetic conditions , we can still identify credible candidate genes previously studied in loss of function experiments . This implies that genes occupying central positions in developmental pathways may also be the ones that carry enough natural variation to allow mapping through GWAS . At the same time we identified regions without any previous information related to craniofacial development ( skull regions 9 and 14; mandible regions 3 , 5 , 8 and 9 ) ; such discoveries contribute new information to the genetics underlying skull and mandible shape determination , and dedicated efforts should be made to understand the phenotypic effect of the genes and regulatory elements falling in such regions . Human studies required ~25 , 000 individuals to explain 3–5% of height variation with genome-wide-significant SNPs [99] , and ~250 , 000 to explain 16% [97] . Using ~5 , 400 individuals , only 5 loci were significantly associated with facial morphology in humans [18] . We have explained 4–11% of craniofacial variation using only ~700 outbred mice . Given the development of semi-automatic tools to speed up the phenotyping of shape traits ( e . g . [100] ) , it seems feasible to increase both the number of animals involved , as well as to apply these tools to different mapping contexts . Hence , we are becoming more confident that an understanding of the biology behind craniofacial development will become possible . All procedures were approved by the University of Chicago Institutional Animal Care and Use Committee ( IACUC ) in accordance with National Institute of Health guidelines for the care and use of laboratory animals . Mapping population Male mice from the CFW mouse colony , maintained by Charles River Laboratories , were used for genome-wide association mapping . Upon their arrival at the University of Chicago , the mice were subjected to behavioral and physiological tests over the course of 2011 and 2012 ( additional phenotype data from these tests are included in a separate manuscript that is being prepared for publication ) . At the end of these experiments , the mice were sacrificed and their heads were stored in ethanol . The average age at the time of sacrifice was 13 weeks ( ranging from 12 to 14 weeks ) . Skulls and mandibles were measured in a subset of 720 mice between 2013 and 2014 at the Max Planck Institute for Evolutionary Biology in Plön , Germany . Mouse heads were scanned using a computer tomograph ( micro-CT—vivaCT 40; Scanco , Bruettisellen , Switzerland ) at a resolution of 48 cross-sections per millimeter . Using the TINA landmarking tool [101] , 44 three-dimensional landmarks were positioned in the skull , and 13 in each hemimandible ( S1 Fig and S1 Table ) . The semi-automatic landmark annotation extension implemented in the TINA landmarking tool was used to reduce digitation error and accelerate the phenotyping process [100] . The raw 3D landmark coordinates obtained in TINA tool were exported to MorphoJ [102] for further morphometric analyses . The symmetric component of the mandible and skull were obtained following [103] . In short , for mandible a full generalized Procrustes analysis ( GPA ) was performed with the landmark configurations of the right and left hemimandibles . The GPA eliminates the variation due to size , location , and orientation of the specimens , and generates a new dataset that only contains shape variation . For each individual , we recorded an average of the right and left resulting configurations , which represents the symmetric component of shape variation . For skull , a mirror image of the landmark configuration of each individual was generated , and a full GPA was performed with the original and mirror configurations . Again , the resulting configurations were averaged to obtain the symmetric component of shape variation . The new landmark coordinates generated by the GPA are called “Procrustes coordinates” . To define shape features , we computed in MorphoJ principal components ( PCs ) from the n x 3k covariance matrix of Procrustes coordinates , where n is the number of samples and k is the number of landmarks; 3k represents the number of Procrustes coordinates ( n = 590 , k = 13 for mandible , and n = 710 , k = 46 for skull ) . PC loadings computed in this analysis define the phenotypes used in the QTL mapping . Differences in age , spanning 2 weeks , did not correlate significantly with shape variation , so we did not use age as a covariate in subsequent analyses . In a separate project that will be presented in more detail in a later publication , areal BMD ( aBMD ) of the isolated femur was examined . Unexpectedly , we found that CFW mice appear to be predisposed toward abnormally high aBMD . This is a characteristic of the CFW mice that does not appear to be shared with commonly used inbred lab strains . A qualitative analysis of mice with high BMD showed substantial differences in mandible , and modest differences in the skull compared to mice with normal BMD . We therefore assessed covariation of BMD with shape measurements , separately for the skull and the mandible . For the skull , we found a small correlation between BMD and shape ( r2 = 1 . 4% , p ( 10 , 000 permutations ) < 0 . 001 ) . However , no individual PC corresponds to these shape differences due to BMD . Therefore , BMD was not used as covariate for skull trait mapping since it would have little to no effect on our ability to map QTLs for skull shape . For the mandible , there was a stronger correlation between shape and BMD ( r2 = 6% , p ( 10 , 000 permutations ) <0 . 001 ) . BMD accounts for 29% of the variation in the first PC , 8% of the variation in the third PC , but little to no variation in the remaining PCs ( maximum r2 is 1 . 4% for PC6 ) . Therefore , we computed mandible shape residuals by removing the linear effect of BMD; we used these residuals as input to the PCA , then PC loadings from this PCA analysis were used as phenotypes in the QTL mapping for mandible shape . The standard measure of size in geometric morphometrics is the centroid size ( CS ) . This is the measure we used for mapping . Centroid size is defined as the square root of the sum of the squared distances of a set of landmarks from the center of gravity or centroid [104] . The CS for mandible was defined as the average of the CS of right and left hemimandible . The skull CS was calculated using all landmarks from right and left sides [103] . All these calculations were done in MorphoJ . The mice were genotyped using a genotyping-by-sequencing ( GBS ) approach [105] . In separate work , we have shown that GBS protocols can be used in combination with existing mouse genomics resources and software toolkits to obtain high-quality genotype data at a large number of genetic markers . In short , GBS libraries were prepared by digesting genomic DNA with the restriction enzyme PstI and annealing barcoded oligonucleotide adapters to the resulting overhangs . Samples were multiplexed 12 per lane , and sequenced on an Illumina HiSeq 2500 using single-end 100-bp reads . By focusing the sequencing effort on the Pstl restriction sites , we obtained high coverage at a subset of genomic loci . The 100-bp single-end reads were aligned to the Mouse Reference Assembly 38 from the NCBI database ( mm10 ) using bwa [106] . We used a GBS-adapted version of the “best practices” pipeline of GATK [107–109] to discover variants and call genotypes . For the Variant Quality Score Recalibration ( VQSR ) step , we calibrated variant discovery against ( 1 ) whole-genome sequencing ( WGS ) data ascertained from a small set of CFW mice , ( 2 ) SNPs and indels from the Wellcome Trust Sanger Mouse Genome project [110] , and SNPs available in dbSNP release 137 . GBS yields highly variable coverage across samples at the same cut site , hence variants with highly variable genotyping call rates . Therefore , to augment the set of SNPs with available genotypes , we used IMPUTE2 [111] to estimate missing genotypes and improve low confidence genotype calls . In total , we identified 92 , 374 autosomal SNPs . - For QTL mapping , we took an additional step to filter out SNPs with low "imputation quality" assessed by inspecting the IMPUTE2 genotype probabilities ( more precisely , any SNP in which less than 95% of the samples have a maximum probability genotype greater than 0 . 5 ) , and SNPs with minor allele frequencies less than 2% . After completing this filtering step , we ended up with a final panel of 80 , 027 SNPs used to map QTLs on autosomal chromosomes . 720 mice were used for mapping loci associated with skull traits ( shape and size ) . Due to the correlation between BMD and mandible shape , to assess support for mandible QTLs we used only the 592 mice for which BMD measurements were available . We mapped QTLs for all PCs explaining at least 1% of total phenotypic variation in the sample; this includes 22 PCs capturing 84% of skull shape variation , and 21 PCs capturing 94% of mandible shape variation . Each PC was analyzed separately . To map size variation , the centroid size of mandible and skull was used . Note that the use of PCs restricts the findings to SNPs associated with the shape directions represented by such PCs; therefore genetic variants not aligned with the PC directions will not be detected with this approach . We used the linear mixed model ( LMM ) implemented in GEMMA [112] to map the phenotypes , and at the same time to correct for the residual population structure that might still be present in the mapping population . The support for association with a given SNP is based on the p-value calculated from the likelihood-ratio test in GEMMA . “Proximal contamination” refers to the loss in power to detect a QTL when the causal marker is included in the calculations of the kinship matrix [113 , 114] . In human genome-wide association studies with smaller sample sizes , this loss in power is expected to be minor [115] . However , in this study we expect that proximal contamination will have a larger impact on the genome-wide association analysis , particularly for genetic variants with larger effects , due to extended patterns of linkage disequilibrium in the CFW mouse population . To address this reduction of power due to proximal contamination , we took a ‘leave one chromosome out’ approach in which each chromosome is analyzed using a kinship matrix defined using all SNPs except SNPs on the chromosome being scanned [114 , 116] . A genome-wide significance threshold was calculated separately for each of the phenotypes used in the mapping ( 43 PCs and centroid size ) . A commonly used approach for assessing significance is to estimate the null distribution of p-values by randomly permuting the phenotype observations while keeping the genotypes the same . Such a procedure is technically not appropriate here because it fails to account for the lack of exchangeability among the samples , sometimes resulting in inflation of false positives [117 , 118] . However , since cryptic relatedness appears to have a small impact on association tests , a naive permutation test that assumes independence of the samples should provide an acceptable means to estimate the rate of false positive associations . This approach is supported by previous experiments we have performed in advanced intercross lines showing that improperly accounting for hidden relatedness in the permutations still produces a reasonable estimate for the significance threshold , despite the fact that advanced intercross lines have complex patterns of familial relationships [116] . Therefore , individual phenotypes were permuted 1 , 000 times , the distribution of minimum p-values was calculated , and the significance threshold was defined as 95% of this distribution . The average 95th percentile for all phenotypes was 8 . 9 x 10−7 ( -log ( p ) of 6 . 04 , ranging from 5 . 97 to 6 . 16 ) . This average threshold is depicted in Fig 1 , but the exact threshold calculated separately for each phenotype was used to determine significance of the associations . The LD pattern around the significant SNPs was used to define the QTL regions . A correlation value r2 ≥ 0 . 8 between the “peak” SNP ( SNP with the smallest p-value ) and the neighboring SNPs was used to select SNPs belonging to the QTL region . Genes falling within the QTL region were investigated using the MGI database [119] and literature search to suggest interesting gene candidates . Note that the choice of QTL region is inherently arbitrary , and it is possible that causal gene variant ( s ) underlying the QTL are not found within the QTL region as it is defined here . Numerous SNPs were statistically associated with various PCs; however , the effect of the SNP might go beyond the specific PC and affect other aspects of shape . Once we have identified genetic associations with individual PCs , we estimate the effect of the SNP on total shape . This is accomplished by fitting a standard multivariate regression model to the shape vectors ( 3k Procrustes coordinates ) and SNP genotypes . This multivariate regression was implemented in MorphoJ . We report effect size as the proportion of shape variance explained by the SNP . Overlap with genetic loci reported in previous studies was assessed by defining 500-Kb and 1-Mb windows around the “peak” SNP—that is , the SNP with the lowest p-value—in each of the 26 QTLs identified in this study . This window size was chosen to correspond to the mean size of the QTL regions , which is 0 . 89Mb ( see above for the way QTL regions were defined ) . Once the “true” overlap was determined , 26 genomic regions of 500 Kb and 1 Mb were randomly chosen from the genome and the overlap with previous studies was re-calculated . This was repeated 1 , 000 times to exclude the possibility that the global pattern of overlap was due to chance ( S3 Fig ) . SNP heritability of skull and mandible shape and size were estimated under the null linear model in GEMMA . We showed previously that for these traits , the null model and the Bayesian model implemented in GEMMA yield similar estimates [12] . These heritability estimates are defined as the proportion of phenotypic variation that can be explained by the SNPs used in the mapping; this estimate is often called “SNP heritability” [32] . We used a weighted sum over all PCs to summarize the “total heritability” of craniofacial shape . Each of the weights in this average is given by the proportion of total variation in the original phenotype explained by the PC ( S2 and S3 Tables ) . By averaging over the individual heritability estimates across selected PCs , this yields a scalar value representing SNP heritability of skull and mandible shape . Shape is inherently a multivariate trait , and different shape directions might have different heritabilities [120 , 121] . Here we are not interested in which directions are more heritable than others; our goal is to capture how additive genetic variance contributes to overall phenotypic variation . From this perspective , the “total heritability” value not only informs about the role of genetics in trait variation , but also allows for comparison with other studies provided that the shape data are projected onto the same PCs [122 , 123] . The proportion of phenotypic variation explained by each chromosome was calculated using the restricted maximum-likelihood analysis implemented in GCTA ( Yang et al . 2011 ) . The first 10 principal components of the kinship matrix were included as covariates . An individual REML analysis was done for each chromosome ( option–reml–grm–qcovar ) . Due to the small sample size of this study ( ~700 mice ) it is not possible to fit all the chromosomes at the same time , which results in an inflation of the individual chromosomal estimates . We therefore used the relative ( dividing by the variation explain by all chromosomes together ) and not the absolute contribution of each chromosome to the total phenotypic variation . Because PCs were used as phenotypes , additional calculations were needed to estimate the chromosomal contribution to the global phenotypes–skull and mandible shape . The additive variance per chromosome per PC was multiplied by the proportion of phenotypic variation represented by that PC . Finally , the values for each chromosome were summed across all PCs . The full code and data reproducing the steps of our analyses are available for download at http://dx . doi . org/10 . 5061/dryad . k543p .	Formation of the face , mandible , and skull is determined in part by genetic factors , but the relationship between genetic variation and craniofacial development is not well understood . We demonstrate how recent advances in mouse genomics and statistical methods can be used to identify genes involved in craniofacial development . We use outbred mice together with a dense panel of genetic markers to identify genetic loci affecting craniofacial shape . Some of the loci we identify are also known from past studies to contribute to craniofacial development and bone formation . For example , the top candidate gene identified in this study , Mn1 , is a gene that appeared at a time when animals started to form bony skulls , suggesting that it may be a key gene in this evolutionary innovation . This further suggests that Mn1 and other genes involved in head formation are also responsible for more fine-grained regulation of its shape . Our results confirm that the outbred mouse population used in this study is suitable to identify single genetic factors even under conditions where many genes cooperate to generate a complex phenotype .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[]	2015	Mapping of Craniofacial Traits in Outbred Mice Identifies Major Developmental Genes Involved in Shape Determination
Accurate quantitative assessment of infection with soil transmitted helminths and protozoa is key to the interpretation of epidemiologic studies of these parasites , as well as for monitoring large scale treatment efficacy and effectiveness studies . As morbidity and transmission of helminth infections are directly related to both the prevalence and intensity of infection , there is particular need for improved techniques for assessment of infection intensity for both purposes . The current study aimed to evaluate two multiplex PCR assays to determine prevalence and intensity of intestinal parasite infections , and compare them to standard microscopy . Faecal samples were collected from a total of 680 people , originating from rural communities in Timor-Leste ( 467 samples ) and Cambodia ( 213 samples ) . DNA was extracted from stool samples and subject to two multiplex real-time PCR reactions the first targeting: Necator americanus , Ancylostoma spp . , Ascaris spp . , and Trichuris trichiura; and the second Entamoeba histolytica , Cryptosporidium spp . , Giardia . duodenalis , and Strongyloides stercoralis . Samples were also subject to sodium nitrate flotation for identification and quantification of STH eggs , and zinc sulphate centrifugal flotation for detection of protozoan parasites . Higher parasite prevalence was detected by multiplex PCR ( hookworms 2 . 9 times higher , Ascaris 1 . 2 , Giardia 1 . 6 , along with superior polyparasitism detection with this effect magnified as the number of parasites present increased ( one: 40 . 2% vs . 38 . 1% , two: 30 . 9% vs . 12 . 9% , three: 7 . 6% vs . 0 . 4% , four: 0 . 4% vs . 0% ) . Although , all STH positive samples were low intensity infections by microscopy as defined by WHO guidelines the DNA-load detected by multiplex PCR suggested higher intensity infections . Multiplex PCR , in addition to superior sensitivity , enabled more accurate determination of infection intensity for Ascaris , hookworms and Giardia compared to microscopy , especially in samples exhibiting polyparasitism . The superior performance of multiplex PCR to detect polyparasitism and more accurately determine infection intensity suggests that it is a more appropriate technique for use in epidemiologic studies and for monitoring large-scale intervention trials . Gastrointestinal parasites including soil-transmitted helminths ( STH ) cause considerable morbidity worldwide , especially in resource-poor communities . The chronic effects on health are predominately attributed to the burden of disease rather than mortality [1] . The majority of the global burden is considered due to the five main STH–Ascaris lumbricoides , hookworms ( Necator americanus and Ancylostoma spp . ) , Trichuris trichiura and Strongyloides stercoralis [2–4]; with a significant burden also due to protozoan infections . Polyparasitism is especially widespread , and the impact of this is likely to be more severe than single parasite infections [5–7] . Reliable diagnostic techniques suitable for accurate and sensitive identification of parasites in terms of infection intensity are essential in order to determine effectiveness of disease control programs . This is because morbidity and transmission pressure of helminth infections are directly related to both the prevalence and intensity of infection [8 , 9] . Several microscopy-based techniques are available and widely used for the identification and quantification of STH eggs . The Kato-Katz ( KK ) thick smear technique , originally developed for diagnosis of schistosomiasis [10] , is currently the most widely used microscopic technique , and is considered the gold standard by the World Health Organization ( WHO ) for assessing both prevalence and intensity of infection in helminth control programmes [11] . A major drawback of the KK is that multiple samples with multiple slides per sample are required to be examined over several days to reach high levels of sensitivity and quantitative accuracy , especially in light infections [12] . Moreover , immediate and skilled processing is required to reduce chance of false-negative results , particularly for hookworms , due to fast clearance on slides [7 , 13] . Alternative methods for microscopic diagnosis of STH include using concentration steps , such as formalin-ether sedimentation and flotation techniques such as McMaster [14] , simple sodium nitrate methods [15] , FLOTAC [16] or mini-FLOTAC [17] . These have some advantages including increased sensitivity over KK . For example , in a recent study sodium nitrate methods proved superior for detecting low egg burdens in samples compared to quadruple KK smears and resulted in higher EPG [15] . Comparisons of KK to FLOTAC also showed equivalent [18 , 19] or superior [20 , 21] sensitivity of FLOTAC techniques . These flotation based techniques have drawbacks in terms of being labour intensive , having poor reproducibility owing to operator error , and for most tests requiring centrifugation steps [16 , 19] . In addition , microscopic-based techniques lack the ability to assign species-level identification of helminth eggs ( e . g . those of hookworms , and Ascaris ) . Diagnosis of S . stercoralis is particularly challenging , as only a small number of larvae are released in stool regardless of infection intensity , with Baermann sedimentation or agar plate culture methods thought to provide greatest specificity and sensitivity [22 , 23] . Serological diagnostic methods are available for STH but their use is limited due to poor specificity in endemic areas [24] . Stained faecal smears and faecal concentration methods allow for diagnosis of protozoa . However , these too have their limitations with regards to poor sensitivity and the inability to differentiate protozoan parasite stages to a species level . For example , diagnosis of E . histolytica infections by microscopy misses 40% of infections [25]; and it is not possible to visually differentiate pathogenic E . histolytica from non-pathogenic Entamoeba dispar . Thus , only E . histolytica specific stool antigen detection tests are approved for diagnostic use by the WHO [26] , although shown to have poor sensitivity compared to PCR-based methods [27 , 28] . More efficient coproantigen capture enzyme linked immunosorbent assay ( ELISA ) based assays can be used for diagnosis of Cryptosporidium and Giardia , however there have been reports of false-positive and false-negative results [29] . Polymerase chain reaction ( PCR ) -based techniques are assuming a dominant place in modern diagnostic microbiology . For STH diagnostics they have been shown to be more sensitive than microscopy , particularly at low infection intensities [11 , 30] . For detection of protozoal infections , PCR-based techniques have shown significantly higher sensitivity compared to microscopy and/or immunodiagnostic techniques [31–37] . Adapting PCR assays to multiplex real-time platforms enables simultaneous detection of multiple parasites , thereby minimizing reagent costs and processing time [12 , 30 , 38] . In addition , such assays can be adapted to be quantitative , which is a significant advantage in STH infections where parasite burden rather than presence or absence of infection is a key determinant of morbidity . Multiplex real-time PCR methods for the diagnosis of intestinal parasites [39] have been applied in epidemiological surveys in parasite endemic areas of Ghana , Togo [13] , Bangladesh , and recently in the Philippines [40] as well as in hospital samples from symptomatic patients in the Netherlands [38] , and Malaysia [41 , 42] . Despite numerous studies incorporating multiplex PCR methods , very limited data has been published to date with thorough quantitative comparisons between this method and microscopy-based techniques . The aim of the current study was to evaluate the use of two multiplex PCRs for the detection and quantification of intestinal protozoa and helminths as tools to support large scale epidemiologic and treatment efficacy studies . This method was compared with results obtained with sodium nitrate flotation and zinc sulphate centrifugation for determining prevalence and intensity of intestinal parasite infections in villages in Timor-Leste and Cambodia . A specific aim was to evaluate qPCR as a method for determining infection intensity , an important parameter in both epidemiologic and anthelmintic efficacy studies . Samples were sourced from two separate parasitic surveys . Single faecal samples from 467 individuals enrolled in the WASH for WORMS interventional trial in Timor-Leste ( Registered with the Australian New Zealand Clinical Trials Registry; Trial registration: ACTRN12614000680662 ) [43] and from a separate study of 213 individuals enrolled in a cross sectional study of intestinal parasites in northern Cambodia [44] . A single faecal sample per individual was collected within a maximum of 12 hours of defecation and divided into two aliquots and stored separately . One sample in 5% w/v potassium dichromate solution for PCR analysis , and the other sample in 10% formalin for microscopy . Samples from Timor were transported at room temperature to Queensland Berghofer Institute of Medical Research for extraction ( QIMRBerghofer ) and PCR and Timor-Leste National Lab for microscopy . Cambodian samples were transported at room temperature to the University of Queensland ( UQ ) Gatton campus for microscopy and DNA extraction , and DNA transported on ice to QIMR Berghofer for PCR . All faecal samples were examined microscopically and enumerated for Ascaris , hookworms , and Trichuris eggs using a simple sodium nitrate flotation as previously described [45]; and for the presence of protozoa cysts and oocysts using zinc sulphate centrifugal flotation [35] . Full methods available in S1 Methods . Specific parasitologic diagnosis of S . stercoralis infection was not undertaken due to resource limitations that precluded agar plate culture . Samples stored in 5% potassium dichromate were first subject to centrifugation at 2 , 000 g for 3 min , followed by removal of the preservative supernatant . Samples were re-suspended to 50 ml with phosphate buffered saline ( 1 X PBS ) and the centrifugation repeated . Supernatant was again decanted off , and the sample pellet was subsequently stored at 4°C for up to a month . DNA extraction was performed using the Powersoil DNA Isolation Kit ( Mo Bio , Carlsbad , CA USA ) . Minor modifications were made to the manufacturer’s protocol following optimization with Trichuris vulpis eggs . Prior to extraction , samples were spiked with a known quantity of a positive control PCR target , namely a plasmid containing equine herpes virus ( EHV ) insert from the glycoprotein B gene [46] . Full DNA extraction protocol is available in S1 Methods . Extracted DNA was run in two pentaplex real-time PCR reactions . The first was a quantitative assay for N . americanus , Ancylostoma spp . ( A . duodenale , A . ceylanicum ) , Ascaris spp . , T . Trichiura and EHV; the second was a semi-quantitative assay for E . histolytica , Cryptosporidium spp . , G . duodenalis , S . stercoralis and EHV [12 , 13 , 24 , 34 , 42 , 47] . Details of the primers and probes are listed in Tables 1 and 2 . However an alternate S . stercoralis forward primer was used after optimization of this assay with S . stercoralis positive DNA samples from the Northern Territory ( Australia ) ( Deborah Holt—Personal communication ) . The Rotor-Gene 6000 ( Qiagen , Melbourne , VIC AUS ) was used for all PCR assays , with reactions set up for both PCR reactions as previously described [30] with minor modifications . Assay protocol , optimisation and preparation of PCR controls is available in S1 Methods . A sample processing summary is displayed in Fig 1 . A Ct cut-off of 31 for Ascaris was established based on the limit of detection on a previously published conventional PCR , to ensure reproducibility of results [48] . The limit of detection of all other assays in terms of the maximum Ct-value considered to be positive was set at 35 . All PCR assays were validated at independent laboratories ( Queensland Medical Laboratory , Australia; The Task Force for Global Health , USA; St . Vincent’s Hospital , Sydney ) , and additional positive E . histolytica ( n = 2 ) and E . dispar ( n = 1 ) control samples tested to ensure PCR specificity for the pathogenic species , E . histolytica . Further confirmation of Entamoeba spp . presence in all controls was provided by a genus specific conventional PCR [49] . To produce a calibration curve to interpolate EPG values from PCR Ct-values a series of seeding experiments were conducted for Ascaris spp . and N . americanus infections . Ascaris suum eggs were purchased from Excelsior Sentinel Inc . ( Ithaca , NY ) , supplied in unembryonated form and stored in 5% potassium dichromate at room temperature for shipping . Hookworm eggs were freshly isolated from three N . americanus infected stool samples , kindly provided by Alex Loukas ( James Cook University ) . Ascaris and hookworm eggs were purified separately [50] , and multiple 10 μL aliquots counted under a microscope following staining with Lugol’s iodine solution to determine the concentration of eggs in each sample . Purified eggs were pooled and suspended in a total of 5 ml PBS and diluted to produce a range of concentrations of eggs . Ascaris eggs were prepared in triplicate in 2 ml screw top tubes in concentrations ranging from 200 , 000 EPG to 5 EPG , with an additional 200 mg negative control faecal sample added to each sample . Hookworm eggs were similarly prepared but concentrations ranged from 6 , 000 EPG to 250 EPG , and were only performed in duplicate due to a smaller number of available egg numbers . Both Ascaris and hookworm eggs at each of the concentrations were subject to DNA extraction and multiplex PCR as stated previously . PCR Ct-values were converted to intensities based on assumed 100% reaction run efficiency , provided by the Rotorgene Q software ( Multiplex PCR Intensity = 10−0 . 298*Ct +9 . 81 ) . This allowed the interpolation of the relation between the log transformed EPG and log transformed PCR intensity to determine a relationship between Ct-values and EPG . Kappa statistics were used for comparison of multiplex PCR and microscopy determined prevalence . Analysis was performed using SPSS ( IBM Corp . ) , Excel 2008 ( Microsoft ) , and GraphPad Prism version 6 . 0 ( GraphPad Software Inc ) . The WASH for WORMS interventional trial in Timor-Leste complies with the provisions contained in the National Statement on Ethical Conduct in Human Research and was approved by the University of Queensland Medical Research Ethics Committee ( #2011000734 ) , the ANU Human Research Ethics Committee ( protocol: 2014/311 ) , and the Timor-Leste Ethics Committee of the Ministry of Health ( reference 2011/51 ) . The Cambodia sample collection study protocol [44] was approved by the National Ethics Committee for Health Research , Ministry of Health , Cambodia ( NECHR , #192 , ) Ethics Committee of the Cantons of Basel-Stadt and Baselland ( EKBB , #18/12 ) . Written informed consent was obtained for all participants . Optimized oligonucleotide concentrations used throughout testing are listed in Table 1 . Experimental comparisons of plasmid control dilutions series’ were undertaken between singleplex and multiplex PCR reactions ( Fig 2 ) . No or minimal effect on sensitivity and efficiency of one PCR on another was found in each multiplex PCR . Example data of standard EHV Ct-values results is shown in S1 Fig . Field samples excluded due to inhibition of the EHV control were subject to repeat analysis . Comparison of the prevalence of parasite infection between multiplex PCR and microscopy was made individually for each data set ( Fig 3; Table 2; S1 Dataset ) . The observed multiplex PCR prevalence was consistently higher across nearly all target organisms in both regions studied . Multiplex PCR detected almost three times ( 2 . 9 , 435/151 ) the number of hookworm infections , 1 . 2 ( 260/219 ) times more Ascaris infections and 1 . 6 ( 115/70 ) times more Giardia infections than microscopy at both study sites . S . stercoralis was however detected in four of the Cambodia microscopy samples , which tested negative in multiplex PCR . The number of samples that tested negative in microscopy was twofold ( 2 . 34 ) higher than the number testing negative in multiplex PCR , indicating that a large proportion of infections were missed by microscopy . Direct comparisons between diagnostic techniques on individual samples from all regions were undertaken using Kappa agreement statistics ( Table 3 ) . Results show good , moderate and fair agreement for Ascaris , Giardia , and hookworms , respectively . Only target organisms with over 20% prevalence were analysed . For all parasites analysed , multiplex PCR identified a large number of positive samples not detected by microscopy ( Ascaris 68 , Hookworm 299 , Giardia 69 ) , whilst a small number of microscopy positive samples were not identified as positive by multiplex PCR ( Ascaris 27 , Hookworm 15 , Giardia 24 ) . Along with a higher detection rate of all individual target organisms , the multiplex PCR approach was also superior in terms of sensitivity in detecting samples with multiple infections ( Fig 4 ) . Increased polyparasitism was detected in multiplex PCR in comparison to microscopy , as similar levels of single parasite infections were detected but more than double ( 2 . 4 times ) the number of dual parasite infections . This trend is further compounded with three parasites , with PCR detecting over 17-fold ( 17 . 3 ) the number of infections than microscopy . An additional three samples were found to harbour four parasites ( Ascaris , N . americanus , Ancylostoma spp . , and Giardia ) , only detected using multiplex PCR . Coinfection with the two genera of hookworms , N . americanus and Ancylostoma spp . that was detected in the multiplex PCR resulted as well in an increased prevalence of polyparasitism . Infection intensities as determined by microscopy data for nematodes with greatest prevalence is presented in Fig 5 . According to WHO guidelines all infections from both study regions for both hookworm and Ascaris would be classified as low intensity infections ( Ascaris < 5 , 000 EPG; hookworm < 2 , 000 EPG ) . The infection intensity frequency distribution of parasites with at least 20% prevalence for Timor-Leste and Cambodia is shown in Fig 6 . This Multiplex PCR Ct frequency distribution graph shows the majority of infected individuals harbouring high relative infection burdens ( low Ct-values ) . The infection intensity of G . duodenale was similar in both Timor-Leste ( Average Ct 24 . 5; range 16 . 2–34 . 3 ) , and Cambodia ( Average Ct 23 . 2; range 16 . 0–32 . 0 ) . The few positive Cryptosporidium sp . samples in Timor-Leste were all of similar infection intensities ( Average Ct 30 . 9; range 29 . 0–33 . 0 ) , whilst the four positive T . trichiura Cambodian samples all were of low infection intensities ( Average Ct 32 . 7; range 31 . 5–33 . 7 ) . A comparison of quantitative results from both microscopy and multiplex PCR was attempted . However , data are not presented as no statistically relevant relationship was found for either Ascaris or hookworms from both study regions . Similarly , no statistical correlation was found between PCR-determined positive samples with low infectivity levels and negative microscopy results , or vice versa . Standard curves obtained from well-defined controls presenting relationship between PCR determined Ct-values ( converted to infection intensity ) as a measure of EPG derived from standard microscopy practices are shown in Fig 7 . The interpolation of Ct-value to EPG derived from this experimental data is also presented within Fig 7 , along with the Ct-value range in which this can be reliably used to estimate EPG from multiplex PCR data . A preliminary example of the use of the resulting calibration curve is shown using Timor-Leste field data in S2 Fig . The present study further demonstrates and validates the suitability of multiplex PCR for detection and quantification of Ascaris , hookworms and Giardia in stools obtained in large scale surveys . The two multiplex PCRs in Ascaris , hookworm and Giardia endemic regions of Timor-Leste and Cambodia resulted in significantly higher levels of prevalence compared to microscopy alone [30] [13 , 36] [51] . Multiplex PCR also showed a greater ability for detection of co-infections , and provided more accurate and reliable infection intensity data; both of key importance to the assessment of disease burden due to the elevated risk of morbidity [52] , A large disparity in the prevalence rates between techniques was noted for the detection of hookworm , Ascaris and Giardia when compared on a per sample basis . Multiplex PCR in particular detected a large number of positive samples not detected in microscopy; a possible result of the failure of microscopy to detect polyparasitism , present in nearly half ( 49 . 1% ) of all positives samples by multiplex PCR . A small number of samples were also deemed positive by microscopy but negative in multiplex PCR . This may be due to variation in the dispersion of larvae , eggs and oocytes within the subsamples taken , due to the nonhomogeneous nature of the stool [53] , as well as the non-uniform nature of their excretion in stool [30 , 54] , causing variation in both techniques . Differences between techniques may also be due to errors leading to false positive results . Such errors are less likely in multiplex PCR due to rigorous controls , whilst limited controls can be implemented with microscopy , which relies heavily on the technical expertise of the user . This is potentially the case of the four Strongyloides positive samples detected only by microscopy , as larvae resemble hatched hookworm larvae . Alternatively further PCR optimization may be required as there have been previous reports of PCR sensitivity issues for detection of Strongyloides , as only low levels of larvae are present even in heavy infections [55] . Microscopy sensitivity issues of parasite detection , particularly in samples exhibiting polyparasitism suggest further compounding issues in accurately determining infection intensity levels . Data comparing PCR Ct-values to microscopy determined EPG values have been previously been reported , showing a broad range of Ct-values for each microscopy EPG value; especially for microscopy negative samples [13] . This inability to link the quantitative field data using different techniques despite statistically sound prevalence agreements suggests that multiplex PCR is superior for use in diagnostic testing of STH for survey work where accurate intensity data is required . This focus on infection intensity is a major strength of the present study when considering disease burden , as STH infection prevalence alone does not provide a measure of potential morbidity , which is related directly to infection intensity [9] . Despite this , the majority of studies have focused on direct prevalence comparisons between techniques , with limited reports on the quantitative abilities of the techniques [30] , although some correlation data has been provided [13] . The WHO currently considers prevalence the main measure in STH control programmes , with suggestion of including intensity data only if available , with treatment aims to target medium to heavy STH infections [56] . Providing this more accurate infection intensity data using the multiplex PCR technique may produce data required to establish more effective STH control programmes . The WHO infection intensity estimates may also require re-evaluation with improved intensity data to be useful in PCR-parasite intensity surveys . These estimates were established based using the KK technique , and suggest only low intensity infections when applied to the current study microscopy data ( 53 ) , whilst multiplex PCR data indicated high intensity infections ( low Ct-values ) in the majority of individuals . In the present study , the full relationship between Ct-values and EPG was assessed using standard curves obtained from well-defined controls , with carefully prepared measurements . The feasibility of such an approach has been shown and further indicates that PCR quantification is likely to be more accurate and that additional detailed studies should be undertaken in the future . At this stage insufficient experimental data are available to be able to produce accurate predictions across the whole range of possible Ct-values as the current interpolation is limited to within the microscopy determined EPG range in which it was tested . Additional validation of the Ascaris EPG interpolation is required and at higher infection intensities before it can be reliably used with field samples and significant further testing of hookworm samples is essential to produce accurate predictions of EPG . Determining this mathematical relationship between Ct-values , EPG , and by corollary adult worm burden , and assuring its accuracy represents a goal in future research , vital to progress PCR methodologies as the gold standard of intestinal nematode detection . The potential shown here to convert this PCR data to the more widely understood “EPG” format may allow this technique to provide more widely accepted and reportable STH surveillance . However , attempting to enumerate egg counts by PCR without a truly quantitative gold standard for comparison does however , represent a limitation of this work . Despite proving successful in parasite diagnosis and providing valuable information on infection intensity , there are limitations of multiplex PCR as a diagnostic tool . The use of PCR preservatives such as potassium dichromate and ethanol temporarily arrests further egg development , and are thought to allow accurate quantitative data after many months of storage . However extensive testing into the effect of such preservatives on the genome copy numbers for the target nematodes has not been performed but is crucial for accurate interpolation to an egg count . Further compounding the issue is the ability to relate nematode egg counts to DNA intensity , when copies of DNA increase once eggs have embryonated . Literature does indicate that once embryonated the ITS1 gene target of the PCR assays remains at a constant level [57] , suggesting that accurate quantification is possible . Further testing on preservation methods is essential in terms of the effect on the quantitative accuracy over time , both those for microscopy and multiplex PCR techniques . The maximum storage time in potassium dichromate is currently undetermined , however reports have suggested a one month storage period in potassium dichromate for Giardia for optimal detection [33] . Recent reports on formalin preserved samples suggest a 15 day storage window in which microscopy can be completed to gain accurate quantitative hookworm data , with additional decline in the ability to detect parasites after one month [52] . Microscopy on samples in this study was not reliably performed within this optimum time period; potentially resulting in the lower prevalence and intensity data for hookworms . The main limitation of multiplex PCR in national control programmes is the capacity to implement it in parasite endemic low-resource settings , with the current need to send samples to well equipped labs for analysis . Microscopy based techniques can however , be undertaken with less resources and more affordable equipment . The material cost of processing samples and running both multiplex PCRs was estimated as AU$12 . 37 per sample ( AU$6 . 05 per extraction; AU$3 . 16 per multiplex PCR ) . This also represents a limitation when compared to flotation based microscopy , costing just AU$1 per sample ( labour not included ) . The trade off of the additional costs and the inability for onsite analysis compared to the higher sensitivity and ability to detect multiple infections as well as the more accurate intensity data is an issue for consideration in design of epidemiologic studies and for clinical trials of effectiveness of interventions . The use of multiplex real-time PCR for intestinal parasite diagnosis has proved to be more sensitive , and is more likely to detect mixed parasite infections than standard microscopy techniques . The real benefit of multiplex PCR is in its ability to more accurately determine infection intensity and the potential to report results in more understandable ‘EPG’ terms , which will prove to be inherently more useful in determining the success of de-worming and intervention trials .	Gastrointestinal parasites including soil-transmitted helminths cause considerable morbidity worldwide , especially in resource-poor communities . Large-scale epidemiologic and treatment efficacy studies are regularly undertaken to determine the optimum ways to reduce or eliminate parasites from endemic communities , thereby reducing the burden of disease . Accurate and sensitive tests for detection of soil transmitted helminths and protozoa are of great importance to the success of such trials . Increasingly recognised is the importance of accurately determine the infection intensity , as morbidity and transmission pressure of helminth infections are directly related this and not just to prevalence . A vast majority of studies use standard microscopy methods which , although well accepted , may not be as accurate as more recently developed molecular techniques such as multiplex PCR . Therefore , there is need for further evaluation of multiplex PCR techniques and their ability to detect infections and provide infection intensity data . In the current study real-time PCR showed a higher sensitivity for the detection of intestinal helminths and protozoa especially in cases of mixed infections as well as more accurate determination of infection intensity compared to microscopy .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "invertebrates", "medicine", "and", "health", "sciences", "giardia", "parasitic", "diseases", "parasitic", "protozoans", "animals", "necator", "americanus", "health", "care", "ascaris", "protozoans", "molecular", "biology", "techniques", "morbidity", "research", "and", ...	2016	Application of a Multiplex Quantitative PCR to Assess Prevalence and Intensity Of Intestinal Parasite Infections in a Controlled Clinical Trial
In Latin America , there are 13 geographically isolated endemic foci distributed among Mexico , Guatemala , Colombia , Venezuela , Brazil and Ecuador . The communities of the three endemic foci found within Mexico have been receiving ivermectin treatment since 1989 . In this study , we predicted the trend of occurrence of cases in Mexico by applying time series analysis to monthly onchocerciasis data reported by the Mexican Secretariat of Health between 1988 and 2011 using the software R . A total of 15 , 584 cases were reported in Mexico from 1988 to 2011 . The data of onchocerciasis cases are mainly from the main endemic foci of Chiapas and Oaxaca . The last case in Oaxaca was reported in 1998 , but new cases were reported in the Chiapas foci up to 2011 . Time series analysis performed for the foci in Mexico showed a decreasing trend of the disease over time . The best-fitted models with the smallest Akaike Information Criterion ( AIC ) were Auto-Regressive Integrated Moving Average ( ARIMA ) models , which were used to predict the tendency of onchocerciasis cases for two years ahead . According to the ARIMA models predictions , the cases in very low number ( below 1 ) are expected for the disease between 2012 and 2013 in Chiapas , the last endemic region in Mexico . The endemic regions of Mexico evolved from high onchocerciasis-endemic states to the interruption of transmission due to the strategies followed by the MSH , based on treatment with ivermectin . The extremely low level of expected cases as predicted by ARIMA models for the next two years suggest that the onchocerciasis is being eliminated in Mexico . To our knowledge , it is the first study utilizing time series for predicting case dynamics of onchocerciasis , which could be used as a benchmark during monitoring and post-treatment surveillance . Human onchocerciasis is caused by the filarial worm Onchocerca volvulus , which is transmitted by the bites of blackflies of Simulium species [1] . The symptomatology of onchocerciasis disease is characterized by clinical manifestations such as onchocerca skin diseases , onchocercomata , lymphadenopathy , and ocular lesions , including the irremediable terminal effect of blindness [2] . Onchocerciasis is the major cause of blindness and dermatitis in endemic areas , and it remains as an important public health problem in Africa . In Latin America , there are six countries ( Brazil , Colombia , Ecuador , Venezuela , Guatemala , and Mexico ) with scattered and small endemic onchocerciasis foci , where a population of 470 , 222 individuals is currently estimated to be at risk [3] . The discovery of onchocerciasis in America was in 1915 by Rodolfo Robles in Guatemala; hence it was first named as Robles's disease . In Mexico , the first cases of onchocerciasis were documented in 1923 in Chiapas , originated as a consequence of active seasonal migration of coffee workers from the endemic areas between Guatemala and Mexico . The regions in Chiapas and Oaxaca of Mexico are associated with the presence of abundant vector populations [4] . The focus of Oaxaca and Northern Chiapas represented the expansion of onchocerciasis from Southern Mexico or Guatemala [5] . The onchocerciasis control program in Mexico was first established in 1930 and has worked continuously up to date . During 1930–1946 , a sporadic larval vector control campaign using Creolin was carried out to eliminate vector populations from breeding sites together with nodulectomy ( removal of nodules ) campaigns [6] . The administration of diethylcarbamazine ( DEC ) began in 1947 , followed by a sporadic application of DDT to eliminate the vector populations in 1952 . In 1990 , DEC was supplanted by ivermectin ( Mectizan; Merck & Co . , Inc . , Whitehouse Station , NJ ) [4] . In 1992 , the Onchocerciasis Elimination Program for the Americas ( OEPA ) was launched [7] , and has successfully coordinated the efforts of the affected countries in Latin America . In 1989 , the onchocerciasis program in Mexico started the treatment with ivermectin only for symptomatic individuals [4] . Later in 1997 , ivermectin distribution was implemented twice a year for most eligible residents from all at-risk communities , followed by the distribution four times a year in the Chiapas foci as from 2003 upward , which was a successful strategy to accelerate the interruption of the parasite transmission [8] . As OEPA is preparing to wind up activities as from 2012 , it is important to predict the possibility of future occurrence of new cases of the disease because there still exists the possibility of recrudesce due to the existence of potential infected vector population or multiple vectors [9] . The time series analysis has been applied in the field of epidemiological research on infectious diseases for the prediction of epidemiological spread tendency , which provided valuable information for making decisions in the control of such diseases [10]–[13] . For instance , the ARIMA models [14] as well as Seasonal Auto-Regressive Integrated Moving Average ( SARIMA ) [15] models were used to analyze time series data containing ordinary or seasonal trends of dengue cases to develop a forecasting model in endemic areas of Rio de Janeiro , Brazil and French West Indies , Guadeloupe respectively . The univariate Auto-Regressive Integrated Moving Average ( ARIMA ) models are a kind of time series analysis for forecasting a time series data [16] . ARIMA models need a stationary series , that is , the mean and variances of the series are independent of time . Stationarity can be accomplished by data transformations and differencing . Once the series is stationary , it is fine-tuned through adding auto-regressive ( AR ) orders ( lags of the differenced series ) and/or moving average ( MA ) orders ( lags of the forecast errors ) as needed to remove any last hints of autocorrelation from the forecast errors . The usefulness of ARIMA models resides mostly in providing an estimate of the variability to be expected among future observations and depends on past values and random errors [10] . Herein , the cases of onchocerciasis reported by the Mexican Secretariat of Health ( MSH ) during the past two decades were analyzed using time series analysis . We adopted the ARIMA approach for describing the case dynamics of onchocerciasis in the endemic foci and predicted the tendency of occurrence of onchocerciasis cases in the immediate future . The official norm NOM-032-SSA2-2002 of MSH has defined that a case of onchocerciasis should comply with at least one of with the following requirements: demonstration of microfilariae through microscopic examination of superficial skin snips , identification of adult worms by removing nodules , observation of microfilariae in the cornea and anterior chamber of the eye , positive PCR and hybridization from skin snips or nodules . The individual should also present typical clinical manifestations of the disease , and inhabit or have resided in areas of active transmission . Monthly data of onchocerciasis cases between 1988 and 2010 were obtained from the MSH web site ( http://www . dgepi . salud . gob . mx/anuario ) . Preliminary information on cases in 2011 was obtained from the weekly bulletin web site of MSH ( http://www . dgepi . salud . gob . mx/boletin/ ) . Time series analysis for identifying significant predictors as well as for forecasting monthly onchocerciasis cases were carried out using the statistical analysis ARIMA model . The data in 1990 for Oaxaca and in 2001 for Chiapas was not available . The cases of onchocerciasis in Chiapas and Oaxaca from 1988 to 1993 were recorded every two months , which could result in data bias ( one month with 0 cases after one month with data ) . Considering that the month without data does not indicate no case occurrence but the cases not reported , and then the cases of that month accumulated in the data of next month , we thus decided to adjust the data by assigning the half part of cases of a month to the previous zero-case month . Because of disease control activities ( ivermectin distribution ) , cases of infection have been greatly reduced , giving rise to an abundance of zeros in the monthly case data . It needs to stabilize the variance of the series before seeking the best model that fits each dataset . The square root ( sqrt ) transformation was applied to stationarize our datasets . After stabilizing the variance , the descriptive method procedure was performed for plotting the onchocerciasis data through the autocorrelation function ( ACF ) and partial autocorrelation function ( PACF ) to identify the order of differentiation as well seasonal and non-seasonal effects . The residuals of the models fitted were inspected with the ACF and PACF plots and further verified with the Ljung-Box test . The best ARIMA model was selected for analysis according to the lowest Akaike Information Criterion ( AIC ) . The ARIMA models were represented by the form as ( p , d , q ) ( P , D , Q ) S , where p is the order of auto-regression , d is the order of differencing ( or integration ) , and q is the order of moving-average for non-seasonal series . P , D , Q are their seasonal counterparts , and S is the seasonal period . If the parameters p and q or P and Q are together present in the non-seasonal or seasonal series , the model was termed as mixed ARIMA model . We estimated the parameters of ARIMA models with the “arima” function implemented in the software R [17] that compute the exact likelihood via a state-space representation of the ARIMA process by using the Kalman filter [18] , [19] , “skipping” the missing observations in the computations , obtaining the maximum likelihood estimators of the model parameters . The model's fitted values were also graphically compared with the observed data . The fitted model was adopted to out-of-sample predict onchocerciasis cases for the next two years in the foci using the one-step ahead approach , that is , a forecast generated for the next observation only . For example , as the observed value for January 1998 was obtained in Oaxaca region , the data were updated to January 1998 , re-estimated the parameters of the ARIMA model , and computed the next 1-step ahead predicted value , February 1998 . This process was continued until the end of the year 1999 . The software R ( version 2 . 11 . 1 ) was used for all statistical analyses and graphic displays [17] . The automatic algorithms implemented in software R were also used to aid in the selection of the ARIMA models [20] . The cases of onchocerciasis in Mexico from 1988 to 2011 were summarized in Table S1 . The total number onchocerciasis cases in Chiapas , Oaxaca and other regions of Mexico during 1988 to 2011 were 15 , 584 cases . The highest number of cases was in 1988 with 3 , 197 cases and afterwards the number decreased gradually , with the lowest number in 2010 with just 15 cases . The recorded cases were predominantly from two regions , Oaxaca and Chiapas ( Figure 1 ) , and some sporadic cases from other regions . In the Oaxaca focus , the total reported cases were 1 , 628; the number of cases was highest in 1991 and later decreased marginally . The last case in Oaxaca was recorded in 1998 . Therefore , this disease had been successfully eliminated from the Oaxaca region . The Chiapas foci had a total of 13 , 849 cases reported . The case number remained high before 1990 and maintained a little lower level from 1991 to 1997 , with the second peak in 1994 . Then the recorded cases stably reduced , until 12 cases in 2011 . There were 107 cases reported in other states ( mainly Northern Mexico ) during 1988–1993 , 2005–2007 , and 2009–2011 , which were imported cases of onchocerciasis according to case definition of MSH . In Oaxaca , there were no reported cases since 1999 . This observation gave a good example for us to test if the Time Series Analysis describes well the dynamics of infection cases and predicted the approximate time of disease elimination in Oaxaca . The plot of sqrt-transformed onchocerciasis cases for Oaxaca showed a decreased trend since 1990 ( Figure 2 A ) . The plot of ACF has positive autocorrelations out to a high number of lags , suggesting a nonstationary time series and a need for differencing ( Figure 2 B ) . After stabilizing the series with the first order difference , we inspected the ACF and PACF plots , which suggests that non-seasonal and seasonal parameters are needed in the model ( Figure S1 A , B ) . The negative ACF cutoff at lag 2 , associated with the slow decay of PACF at lags 2–3 suggests that MA orders are needed in the model but the positive correlation at lag 12 associated with the PACF cutoff at lag 12 suggests that non-seasonal AR orders could be also added . In addition , the negative ACF cutoff at lag 32 associated with the slow decay of PACF at lags 28–31 suggested seasonal MA orders . Consequently , we fitted several ARIMA models with different Auto-Regressive orders , AR ( p , P ) and Moving Average orders , MA ( q , Q ) , and excluded any models in which the residuals were not significant and had high AIC values . Thus , the best-fit model for Oaxaca was a mixed ARIMA ( 0 , 1 , 2 ) × ( 0 , 0 , 1 ) 12 ( AIC = 388 . 28 ) seasonal non-stationary model . All coefficients of ARIMA models for Oaxaca were significant ( Table 1 ) . The plots ACF and PACF of the residuals were almost located within the confidence limits ( Figure 3 A , B ) . The Ljung–Box statistic test did not reject the null hypothesis of independence in the residuals time series ( P value = 0 . 93 ) . The model plot that fitted actual plot of dynamics of case data was shown in Figure 4 A . This model was then adopted for two-years-ahead prediction using the 1-step ahead approach . The forecast values for Oaxaca , showed a markedly decreasing trend and zero cases would occur from January 1998 to December 1999 ( Figure 4 A ) , corresponding to the fact that the last case was reported from Oaxaca in 1998 . The predicted result that matches the observations in Oaxaca focus allows us to apply the same methodology to the Chiapas foci . Figure 2C showed the time series profile from 1988–2011 of onchocerciasis cases with the sqrt-transformation . As in Oaxaca , the plot of ACF showed a need for differencing because of a slow decay of positive autocorrelations out to a high number of lags ( Figure 2 D ) . The ACF and PACF plots produced with the first order difference also suggest that non-seasonal and seasonal parameters are needed in the model ( Figure S1 C , D ) . The negative ACF cutoff at lag 1 , associated with the slow decay of PACF at lags 1–5 , suggests non-seasonal MA orders . The positive ACF at lags 12 and 15 associated with the PACF sharp cutoffs at lags 12 and 15 also suggests that non-seasonal and seasonal AR orders could be added . Then , several ARIMA models with different AR and MA orders were fitted , excluding any models in which the residuals exhibited higher autocorrelation , non-significant coefficients and high AIC values . The best-fitted model obtained for Chiapas was a mixed ARIMA ( 1 , 1 , 1 ) × ( 1 , 0 , 1 ) 12 ( AIC = 960 . 77 ) . All the coefficients of the ARIMA model for Chiapas were significant ( Table 1 ) . The plots ACF and PACF of the residuals show no remaining temporal correlation ( Figure 3 C , D ) . The Ljung–Box statistic test did not reject the null hypothesis of independence in the residuals time series ( P value = 0 . 34 ) . Graphically the best-fitted model followed closely the decreasing trend of the observed series in Chiapas ( Figure 4 B ) . The model was then used for two-years-ahead prediction using the 1-step ahead approach . It showed that the cases would continuously and markedly decrease in the recent years and the annual zero case could occur at the period from January of 2012 to December of 2013 in the Chiapas foci ( Figure 4 B ) . The key aspect in the control of onchocerciasis disease in Latin America is the treatment with the drug ivermectin available to all the people at risk [4] , [8] , [21]–[24] . The onchocerciasis program in Mexico began treatment with ivermectin in 1989 , initially treating only symptomatic individuals in hyperendemic communities [4] . In 1994 , annual mass ivermectin treatment to eligible residents ( i . e . , those who were 5 years older and who had resided in the endemic community ) in the at-risk communities was initiated . From 1997 , the strategy was modified to provide mass treatments twice a year to every eligible resident in the at-risk communities [8] . In Oaxaca , the new cases had been controlled effectively and the last case occurred in 1998 . The situation in Chiapas is more complicated . It needs to note that the infection cases in Chiapas maintained a high platform before 1997 and there was a marked reduction from 953 cases in 1996 to 573 cases in 1997 and remained on a decreasing trend up to date . In 2003 , the biannual treatment strategy was modified in the majority of the formerly hyperendemic communities of Southern Chiapas focus by increasing treatment frequency to four times a year in order to accelerate the interruption of parasite transmission [25] . After this modification , the incidence was maintained at a low level ( <100 cases ) and reduced stably until the 12 cases in 2011 . Thus , this observation indicates treatment of endemic communities with regimental distribution of Mectizan was a key to the elimination of onchocerciasis in Mexico , which is in agreement with a recent study [6] . At present , the transmission of onchocerciasis in Oaxaca and the Northern Chiapas has been eliminated , but the transmission was only interrupted in the Southern Chiapas focus as recently declared by OEPA [26] . Ivermectin treatment has been halted in the Southern Chiapas focus in 2012 . Thus , all foci in Mexico are under epidemiological surveillance post-treatment which is within the fourth phase for certification of the elimination of onchocerciasis . The above description shows that the elimination of onchocerciasis in humans is an arduous task . The evaluation of the current state and the prediction of future situations are germane for evaluating epidemiological patterns . Several mathematical models have been developed to simulate the onchocerciasis future in specific endemic zones [27]–[29] . These methods take into account mainly: the life cycle of the parasite , the type of treatment , the phenotypic characteristics of the vector , the microfilarial load in the skin , and the biting rate of the vectors . Two onchocerciasis transmission models were predominant in use that incorporated some of the above-mentioned variables [28] , [30]–[32] . These mathematical models were applied to explore the epidemiological consequences and the effects of control interventions on the parasite population dynamics . One of the statistical analyses useful to make predictions is the time series analysis that has been used to study vector-borne diseases [15] , [33] , [34] . In the present study , the ARIMA models based on statistical concepts were fitted to onchocerciasis data collected from the endemic regions to predict the cases for the coming years . The Oaxaca focus has recorded the systematic data and there is no case since 1999 , which provided a good example for practicing the method on describing the variation of cases in the focus and predicting the annual absence of infection case . A mixed ARIMA model was fitted to imitate the case variation data and predicted the absence of cases between 1998–1999 , which coincides with the observed data , suggesting that the method is reliable . We then employed the similar ARIMA method for data from Chiapas , in which the mixed ARIMA model fitted well the observed data . Thus , this methodology could be considered to apply in other regions as the surveillance system for onchocerciasis . On the other hand , one goal of the onchocerciasis program in Mexico was to interrupt transmission of the parasite by the year 2012 [4] . Our model predicted that the values less than 1 case annually were located in the years 2012–2013 . According to the World Health Organization ( WHO ) [35] , [36] and OEPA [37] criteria to declare a place free of onchocerciasis , a reduction of new infections to an incidence rate of less than one new case per 1 , 000 individuals ( <0 . 1% ) and an absence , or near absence , of infective-stage larvae of O . volvulus in the vector population ( i . e . , a rate of less than one infective fly per 1 , 000 parous flies ) must be documented in such area . If the current trend of onchocerciasis cases is sustained , the declaration of onchocerciasis elimination in Mexico would be in the not far future . In the present study , sqrt transformation was chosen for stationarizing the series . Actually , Anscombe transform [38] ( Figure S2 ) and the natural log ( ln ) ( Figure S3 ) have been also tried . As approaching to elimination , cases of infection have been greatly reduced , giving rise to an abundance of zeros in the monthly case data . With the application of ln-transformation ( the most common use to stabilize the variance ) , we cannot use the number “1” to replace the zero in the dataset , as done in the previous report [38] , because the model will consider it as one real infection case and the cases could never be less than 1 in the prediction . Thus , log-transformations of the data were performed by adding a constant ( <1 ) to the entire dataset . The results , based on the datasets transformed respectively by each of the three methods , were evaluated by the root-mean-square of the errors ( RMSE ) . The RMSE values with the application of sqrt or Anscombe transform were similar and much smaller than with ln- transformation ( Table S2 ) , indicating that both sqrt and Anscombe transform are better than ln- transformation , which corresponds to the comment that logarithmic transformation is generally not recommended to apply to count dataset [39] , particularly to small count dataset . Although there is almost no difference between sqrt and Anscombe transform in present study , sqrt could be more practical way to stationarize such datasets because sqrt is much easier to operate . One phenomenon needs to be mentioned . The predicted plot in both Oaxaca and Chiapas showed that all the data in the coming years are less than 1 case but with 95% prediction interval beyond 1 case . A prediction interval is always wider than a confidence interval because it is not only related to the value of the population mean , but also the data scatter . When it approaches the elimination of disease , the number of cases show as 0 , 1 , or >1 , not as continuous data less than 1 but approaching to 0 . It means that as much closer to elimination , the distribution data become much more discretized . In this situation , it could be difficult to expect the prediction of annual cases less than 1 with the 95% prediction intervals within one case , even though the zero cases were treated as 0 . 1 . On the other hand , our test for the focus in Oaxaca demonstrated our prediction is acceptable despite of the 95% prediction intervals beyond 1 case . This phenomenon mentioned here could be common for times series analysis as approaching to elimination of diseases . Our analysis thus provided a good reference for such prediction of similar diseases . We realize that the present model is adopted for predicting the cases under the same situation in the recent future . Since 2012 , the mass treatments with ivermectin have been halted . Does it mean our prediction does not work ? If ivermectin is the principal reason leading to reduction of cases and the transmission has been really interrupted , it is very possible that the tendency of case occurrence could keep reduction until to zero case . In other words , the mass treatments with ivermectin still keep the influence ( the consequence of the treatment ) and our prediction model has the condition to work . Otherwise , the mass treatments with ivermectin could need to be re-continued . To this sense , the application of this prediction model could be used as a benchmark during monitoring and surveillance after mass treatment has been withdrawn . We are aware of the possible limitations of the present study . The data used in the current study rely on total clinical cases of onchocerciasis reported by the surveillance system of MSH , which may underestimate the true number of cases as earlier posited by various researchers [38] , [40] , [41] . Another limitation is the heterogeneity of the data used that could affect the time series analysis . However , the application of this method in Oaxaca focus indicated that our data analysis was adequate . Therefore , the time series analysis applied herein is acceptable . In conclusion , onchocerciasis in Mexico was a serious public health problem in the past . ARIMA models predicted an extremely low ( zero ) expected cases of onchocerciasis for the next two years , implying that onchocerciasis is being eliminated . These results showed that time series analysis could be a practical method for predicting onchocerciasis case tendencies and could be used as a benchmark for monitoring and surveillance on the post ivermectin-mass-treatment duration . To our knowledge , it is the first study utilizing time series analysis for predicting the case dynamics of onchocerciasis .	Mexico is one of the countries where human onchocerciasis ( river blindness ) can be found in Latin America . In 1989 , the onchocerciasis program in Mexico started the treatment with ivermectin only for symptomatic individuals and then mass distribution of ivermectin was initiated for all eligible residents from 1994 , either annually , twice or four times a year in endemic foci , coordinated by Mexican Secretariat of Health ( MSH ) . In our study , we used a statistical method to analyse the cases of the disease reported by MSH from 1988 to 2011 . The analysis showed that the cases of the disease have marginally decreased since 1999 . The results also predicted an extremely low number ( absence ) of cases between 2012 and 2013 in the Chiapas region , the last endemic area , suggesting that disease is on a trend towards elimination in Mexico . Meanwhile , it could provide a benchmark for surveillance after mass treatment has been halted in 2012 .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "medicine", "public", "health", "and", "epidemiology", "epidemiology", "infectious", "disease", "epidemiology", "epidemiological", "methods", "public", "health" ]	2013	Time Series Analysis of Onchocerciasis Data from Mexico: A Trend towards Elimination
Toll-like receptors ( TLR ) and cytokines play a central role in the pathogen clearance as well as in pathological processes . Recently , we reported that TLR2 , TLR4 and TLR9 are differentially modulated in injured livers from BALB/c and C57BL/6 ( B6 ) mice during Trypanosoma cruzi infection . However , the molecular and cellular mechanisms involved in local immune response remain unclear . In this study , we demonstrate that hepatic leukocytes from infected B6 mice produced higher amounts of pro-inflammatory cytokines than BALB/c mice , whereas IL10 and TGFβ were only released by hepatic leukocytes from BALB/c . Strikingly , a higher expression of TLR2 and TLR4 was observed in hepatocytes of infected BALB/c mice . However , in infected B6 mice , the strong pro-inflammatory response was associated with a high and sustained expression of TLR9 and iNOS in leukocytes and hepatic tissue respectively . Additionally , co-expression of gp91- and p47-phox NADPH oxidase subunits were detected in liver tissue of infected B6 mice . Notably , the pre-treatment previous to infection with Pam3CSK4 , TLR2-agonist , induced a significant reduction of transaminase activity levels and inflammatory foci number in livers of infected B6 mice . Moreover , lower pro-inflammatory cytokines and increased TGFβ levels were detected in purified hepatic leukocytes from TLR2-agonist pre-treated B6 mice . Our results describe some of the main injurious signals involved in liver immune response during the T . cruzi acute infection . Additionally we show that the administration of Pam3CSk4 , previous to infection , can attenuate the exacerbated inflammatory response of livers in B6 mice . These results could be useful to understand and design novel immune strategies in controlling liver pathologies . Accumulative evidences demonstrated that the liver has specific immunological properties and contains a large number of resident and non-resident cells that participate in the regulation of inflammatory and immune responses [1] , [2] . Kupffer cells are among the first cells that orchestrate the inflammatory response under many pathological conditions and they produce pro-inflammatory cytokines and several chemokines after pathogen stimulation . Interestingly , while TNFα and IL6 released by Kupffer cells are involved in hepatic inflammation and liver cell death , paradoxically they also mediate regeneration of the liver after injury [1] , [3] . Importantly , hepatic infiltration of neutrophils participates in early response to cellular stress , and their activation is critical for host defence but can also cause additional tissue damage . Thus , proteases and reactive oxygen species ( ROS ) released by neutrophils can result in mitochondrial dysfunction and eventually in necrotic cell death [4] . Hepatocytes are the most abundant cells in the liver , and it has been shown that they have an important role not only in detoxification but also in controlling systemic innate immunity via production of secreted PRRs and complement components , while also acting as antigen presenting cells [2] , [3] . Primary culture of hepatocytes express mRNA for all TLRs and respond to TLR2 and TLR4 ligands [1] . Recently , it has also been reported that hepatocytes are desensitized by LPS in a TLR4 signalling-dependent manner [5] . However , LPS response is mediated by several hepatic cell populations , which are part of a cellular network involved in the hepatic wound healing and regenerative response [1] , [6] , [7] . Although , the liver is the target of a wide range of microbes including Listeria , Salmonella and Plasmodium species , there are few data related to the implication of Trypanosoma cruzi experimental infection in liver and the relevance of the innate immune response in this organ [8] , [9] . The T . cruzi parasite , an obligate intracellular protozoan , is the causative agent of Chagas disease and represents an important public health burden in Latin America . Nowadays , this infection affects at 9 million people , and more than 30 , 000 new cases occur every year ( WHO; 2007 . http://www . who . int/mediacentre/news/notes/2007/np16/en/index . html ) . Chagas disease is characterized by two distinct phases , the acute phase which lasts 2–4 months , involves a number of parasites detected in the blood stream as well as in host tissues followed by a life-long chronic phase . Parasite persistence eventually leads to severe complications in cardiac and gastrointestinal tissue . However , T cruzi also infects the reticuloendothelial system including the liver , spleen and bone marrow . Hepatic affection during acute infection by T . cruzi has been demonstrated both in humans and experimental animals [10]–[12] . Despite extensive experimental investigations , the proper mechanism developed by the host to limit acute T . cruzi infection has not been fully elucidated . We have recently reported severe hepatic injury in B6 mice infected with T . cruzi , Tulahuén strain . The high mortality in B6 mice was associated with an unbalanced pro-inflammatory cytokine profile , a decreased TLR2 and TLR4 and an increased TLR9 expression in the liver compared to infected BALB/c mice [13] . In this study , we hypothesized that the injurious and non-controlled local inflammatory response in the liver depends on the adequate activation of immuno-modulatories signalling , such as TLR2 and TGFβ cytokine , on both non-immune and immune liver cells during T . cruzi infection . With the purpose to gain more knowledge about the mechanisms involved in hepatic injury during T . cruzi infection , we analyzed the ability of hepatic leukocytes to produce cytokines and inflammatory mediators , such as NO and ROS , which could directly and/or indirectly trigger tissue injury . In parallel , we also investigated liver iNOS and nicotinamide adenine dinucleotide phosphate ( NAPDH ) oxidase gp91 and p47 phox expression , the main inducers of NO and ROS , respectively . In addition , we evaluated the induction of TLR2 , TLR4 and TLR9 in leukocytes infiltrating the liver and hepatocytes during T . cruzi acute infection . Taking into account that the use of TLR-agonists is a potential therapeutic approach and a promissing strategy to treat infectious diseases [14]–[16] , we also evaluated if Pam3CSK4 ( TLR2-TLR1 agonist ) pre-treatment before challenge with T . cruzi is able to modulate the strong liver inflammatory response elicited in B6 mice . Our results show for the first time that , high and sustained levels of NO , ROS and pro-inflammatory cytokines would be dangerous signals to the liver during T , cruzi acute infection . The administration of Pam3CSk4 , previous to infection , clearly attenuated the exacerbated inflammation in liver of B6 mice during acute T . cruzi infection , Tulahuén strain . BALB/c and B6 mice were purchased from the Faculty of Veterinary Sciences ( National University of La Plata , Bs As , Argentina ) and maintained according to the National Research Council's guide for the care and use of laboratory animals . Protocols were approved by the Animals Experimentation Ethics Committee , Faculty of Chemical Science , National University of Cordoba . Six to eight week-old females of both mouse strains were infected intraperitoneally with 5×103 T . cruzi blood-derived trypomastigotes , Tulahuén strain . Non-infected mice for each strain were used as control . Parasites were maintained by serial passages from mouse to mouse . Six to eight week-old B6 mice were pre-treated 24hs before to infection with 10 ug of Pam3CSK4 . At the next day the mice were intraperitoneally infected with 5000 parasites . In addition , non-infected mice were challenged with the same doses of TLR2-ligand and controlled along the acute phase , 24 days post infection ( dpi ) . The different groups consist of infected mice , pre-treated/infected mice , pre-treated/non-infected mice and non-infected mice . Plasma glutamic-oxaloacetic transaminase ( GOT ) and glutamic-pyruvic transaminase ( GPT ) activities were measured as tissue damage index , employing a commercial kit ( Wiener Lab ) and performed according to appended protocol . Liver specimens were fixed in 4% paraformaldehyde and embedded in paraffin . Sections deparaffinized and rehydrated were permeabilized , blocked using Fc block , and incubated with primary antibody , rabbit polyclonal anti-TLR2 at 1/50 , rabbit polyclonal anti-p47-phox at 1/200 , goat anti-mouse TLR4 at 1/100 , goat polyclonal anti-gp91at 1/400 , or monoclonal anti-mouse TLR9 1/100 dilutions ( Santa Cruz Biotechnology ) in PBS containing 2% ( w/v ) BSA . After washing , the sections were incubated with FITC-conjugated anti-rabbit ( 1/200 ) , TRITC-conjugated anti-goat ( 1/200 ) , and FITC-conjugated anti-mouse ( 1/100 ) . Nuclei were examined using DNA-binding fluorochome Hoechst 33258 ( 2 ug/ml ) . Slides were observed with a NIKON ECLIPSE TE-2000 U Microscope equipped with UV and fluorescence broadband . Hepatic leukocytes were isolated from both infected and non-infected mouse strains as previously described [13] . Briefly , liver in RMPI medium containing 100 U/ml heparin plus 2% of FBS was passed through 100-µm nylon meshes; red blood cells were removed using lysis buffer ( SIGMA ) and finally hepatocytes were separated from leukocytes using isotonic 37% Percoll gradient ( SIGMA ) . For flow cytometry , 1×106 cells were stained using a standard protocol and the following Abs were used: FITC-labeled rat anti-mouse Ly-6G-Gr1 and F4/80 ( eBioscience ) , FITC-labeled hamster anti-mouse CD3e mAb ( BD Pharmingen ) , PE-labeled rat anti-mouse TLR2 and TLR4 and biotin anti-mouse TLR9 mAb ( eBioscience ) . Stained samples were acquired using the Ortho Diagnostics System ( Johnson and Johnson Company ) flow cytometer , and data were recorded and analyzed using FlowJo software 5 . 7 . 2 ( Tree Star , Inc . ) . RNA was isolated from 50 mg of hepatic tissue using TRIZOL reagent ( Invitrogen ) as previously described [13] . The cDNA obtained was amplified by PCR using specific primers for gp91 and p47-phox . As an endogenous control GAPDH primer was used . Each primer pair was tested to reach the appropriate condition of amplification . The primer sequence ( sense and antisense ) from 5′ to 3′: gp91 ( GAAGACTCTGTATGGACGGC and GCCTGTGTCATTGTGATTTCCT ) ; p47-phox ( GACCTGTCGGAGAAGGTGGT and CTTCACGGGCAGTCCCATGA ) GAPDH ( ACCACCATGGAGAAGGCCGG and CTCAGTGTAGCCCAAGATGC ) . Relative expression levels of the transcripts were estimated by standardization with internal control of GAPDH gene , and evaluated by densitometry using the GELPRO 3 . 1 analysis program . Hepatic leukocytes ( 2×106/ml ) from infected ( 14 and 21dpi ) and non-infected mice ( purified as described above ) were cultured with RPMI medium ( Sigma ) containing 10% FBS and 40 ug/ml gentamycin for 72 h . ELISA assay was performed as previously described [13] for IL1β , IL4 , IL6 , IL10 , IL12 , TGFβ , TNFα and IFNγ cytokines . Standard curves were generated using recombinant cytokine for each of them ( eBioscience ) . Supernatants ( 100 µl ) were mixed with Griess reagent ( Sigma , USA ) at a ratio of 1∶1 and incubated at room temperature for 15 min . The absorbance was measured at 540 nm using a microplate reader and nitrite concentration was calculated in µM using a standard curve of sodium nitrite . Livers were washed and lysed for 30 min on ice in lysis buffer ( 1% Triton X-100 , 0 . 5% sodium deoxicholate , 9% SDS , 5% dithiothreitol ( DTT ) , 1 mM sodium ortovanadate , 10 ug PMSF , 30 ug aprotinin in PBS ) . Aliquots of tissue lysates , diluted in SDS sample buffer , were separated on a 10% SDS-PAGE and western blot assays were performed as previously described [13] . The following antibodies were used: rabbit polyclonal anti-p47-phox at 1/200 dilution or goat polyclonal anti-gp91 ( Santa Cruz Biotechnology ) at 1/800 in PBS containing 1% ( w/v ) BSA . The assays were revealed using the ECL chemiluminescent system ( Amersham Pharmacia Biotech ) according to the manufacturer's instructions . Statistical significance among groups was assessed by ANOVA . To compare different experimental conditions an analysis of variance ( Two-way or one-way ANOVA ) with Bonferroni's post-hoc test was performed . A p-value <0 . 05 was considered significant . During the acute phase of the inflammatory response , a variety of pro-inflammatory cytokines are released which can induce activation , danger or regenerative signalling . The cytokine production was evaluated in supernatant of purified hepatic leukocytes by ELISA . A strong inflammatory profile , characterized by high levels of IL6 , TNFα , IL12 , and IL1β was found in B6 mice at 14 and 21 dpi compared to control mice ( Figure 1A–D ) . BALB/c mice also showed an increase in IL6 and IL12 levels , but only at 14dpi ( Figure 1A and C ) . In addition , TNFα and IL1β secretion by hepatic leukocytes from infected BALB/c did not produce any changes compared to non-infected mice ( Figure 1B and D ) . The inflammatory response in liver was similar to our previously reported results using splenic leukocytes [13] . However and as opposed to spleen cells , purified liver leukocytes from B6 did not produce IL10 and only induced a slight increase in TGFβ ( Figure 1E and F ) . Interestingly , hepatic leukocytes from infected BALB/c produced high levels of IL10 and TGFβ , two important immunoregulatory cytokines at 14 and 21dpi ( Figure 1E and F ) . On analyzing the polarizing cytokines , we observed that T . cruzi infection induced a mixed Th1/Th2 profile with high levels of IFNγ and moderated production of IL4 in BALB/c , whereas high levels of IFNγ but not IL4 was detected in infected B6 mice ( Figure 1G and H ) . These results indicate that T . cruzi acute infection induces a polarized Th1 profile in liver of B6 mice . It is widely known that NO , which is produced by the enzyme iNOS , is a key mediator of a variety of biological functions , such as microbicidal activity and immunosuppression . However , NO is also associated with the most important immunopathologies , including septic shock [17] , [18] . Taking into account that NO metabolite and Th1-biased response are involved in T . cruzi induced myocarditis [19] we quantified the iNOS induction and NO production as a possible mechanism involved in liver injury during this infection . As shown in Figure 2A , an increased iNOS expression was observed in hepatic tissue from B6 mice at 14dpi by Western blot . However , tissue from infected BALB/c did not reveal any changes in iNOS induction when compared to non-infected mice ( Figure 2A ) . Nitrite production was increased in cultured hepatic leukocyte supernatants from both mouse strains at 14 dpi , but this was significantly lower in BALB/c than in B6 . Additionally , B6 liver leukocytes showed a sustained nitrite production up to 21dpi ( Figure 2B ) . NADPH oxidase is a highly regulated membrane-bound enzyme complex found in a variety of phagocytic and non-phagocytic cells . The core enzyme consists of five subunits: p40-/p47-/p67-phox ( cytosolic ) and p22-/gp91-phox in plasma membrane . Upon stimulation , the cytosolic subunit p47phox is phosphorylated , and the entire cytosolic complex migrates to the membrane , where it associates with p-22-/gp91 to form active NADPH oxidase . Activation of NADPH oxidase leads to release of ROS to the extracellular environment , or at intracellular membranes . ROS production is important for killing pathogens , but an over production may be also detrimental to tissue homeostasis . In this work , we comparatively analyzed the induction of gp91 and p47-phox , two major components of the NADPH oxidase . The hepatic tissue from both mouse strains showed a significant increase in gp91 and p47-phox mRNA and proteins at 14 , 21 and 24dpi compared to non-infected control ( Figure 2C ) . Interestingly , p47-phox mRNA and protein expression was higher in B6 than in BALB/c at 24dpi ( Figure 2C ) . Supporting this result , the activation of NADPH oxidase complex that involve co-localization in the plasma membrane of g91 and p47-phox subunits was observed only in infected B6 at 24dpi ( Figure 2D ) . Normal and injured livers are enriched in innate immune cells , which have a significant impact on hepatic health or disease . Recently , we demonstrated that during T . cruzi acute infection the predominant leukocytes in liver were Gr1 , F4/80 and CD3 positive cells [13] . These three populations accounted for more than 80% of the infiltrated liver cells from both infected BALB/c and B6 mouse strains . In the present work , we analyzed the TLR2 , TLR4 and TLR9 kinetic expression in CD3+ ( Figure 3A ) , Gr1+ ( Figure 3B ) and F4/80+ ( Figure 3C ) hepatic leukocytes through double marking and employing flow cytometry . We observed that the absolute number for each double positive cell population expressing TLR2 and TLR4 was increased in both mouse strains at 14 and 21dpi compared to non-infected control mice . However , this increase did not show significant differences between both infected mouse strains ( Figure 3A–C ) . Interestingly , double positive cells for TLR9 in CD3 , Gr1 and F4/80 leukocytes were increased up to 21dpi only in infected B6 mice ( Arrowhead in Figure 3A–C ) . Previously , we reported a decrease of TLR2 and TLR4 and an up-regulation of TLR9 in liver tissue of infected B6 mice , whereas BALB/c showed higher expression of TLR2 and TLR4 and a lower expression of TLR9 at the end of the acute phase [13] . Therefore , we think that other liver cells would be involved in the TLR2 and TLR4 differential expression observed between these mouse strains . Interestingly , the immunofluorescent analysis of TLR2 , TLR4 and TLR9 expression showed higher TLR2 and TLR4 expression in hepatocytes from infected BALB/c compared to infected B6 mice ( Figure 4A–B ) . These results clearly show a marked difference in the expression of TLR2 and TLR4 in hepatocytes from BALB/c and B6 during T . cruzi acute infection . Taking into account that the TLR2 upregulation in hepatic tissue of infected BALB/c mice could be contributing to an anti-inflammatory effect and that TLR2 has been proposed as an immunomodulator receptor during T . cruzi infection [20] , we evaluated whether pre-treatment with Pam3CSK4 , a TLR2-TLR1 agonist , 24hs before infection with T . cruzi is able to improve the immune response balance and the outcome of the acute infection in B6 mice . Interestingly , the treatment of B6 mice with TLR2 agonist significantly reduced the GOT/GPT ratio compared to infected mice alone ( Figure 5A ) . This result was very exciting because it indicates a marked diminution of liver damage during acute infection . Furthermore , the number of inflammatory foci ( Figure 5B ) and the mortality ( Table 1 ) in mice pre-treated with TLR2 agonist were significantly lower than infected B6 mice and not treated mice . This result was correlated with lower transaminase activities ( Pearson coefficient , r = 0 . 9 ) and consequently with a decreased tissue injury . Testing the hypothesis that pre-treatment with Pam3CKS4 before T . cruzi infection would modulate the exacerbated inflammatory response in B6 mice , we demonstrate a reduction of IL6 , IL12 , TNFα and IL1β levels in pre-treated infected B6 mice compared to controls of infected mice alone ( Figure 6A–D ) . In parallel , an increase of TGFβ concentration was detected in mice treated with TLR2 agonist ( Figure 6E ) . Notably , NO secretion was also reduced in mice pre-treated compared to infected mice alone at 14dpi ( Figure 6G ) . Our results suggest that the pre- treatment with Pam3CSK4 one day before infection reduces the inflammatory environment and the hepatic damage of B6 mice during T . cruzi infection . In the present work , we focus on the study of local immune response in hepatic tissue during T . cruzi infection . Our results clearly demonstrate that liver leukocytes from infected B6 mice produce increased and persistent levels of TNFα , IL6 , IL12 and IL1β pro-inflammatory cytokines , associated with undetectable or low IL10 and TGFβ production . Previously , we observed a enhanced apoptosis and necrosis in liver of this mouse strain [13] . Recently , it was reported that anti-TNFα treatment decreases hepatic apoptosis of C57BL/6 mice infected with trypomastigotes of Tulahuén strain , providing evidence about the role of this cytokine in the induction of liver injury [12] . By contrast , in livers from BALB/c mice , the pro-inflammatory profile was effectively countered by IL10 and TGFβ regulatory cytokines . In this mouse strain , infection induced a mixed Th1/Th2 phenotype while a dominant Th1 profile was observed in livers of infected B6 mice . The Th1 response contributes to an effective parasite clearance although an excessive inflammation may be deleterious for the host . We postulated that a mixed cytokine prototype in BALB/c might be insufficient for controlling parasite but might favour tissue repair . Similarly , a mixed Th1/Th2 and a dominant Th1 phenotype was observed in heart of infected BALB/c and B6 mice respectively [21] . Nitric oxide and ROS are two key inflammatory mediators involved not only in pathogen clearance but also in tissue injury . Nitric oxide is produced by different isoforms of NO synthase , among them the inducible isoform ( iNOS ) that is activated by IFNγ and TNFα [22] , two cytokines that were prominent in our model . Hemodynamic instabilities and tissue damage caused by NO may be related to different NO synthase isoforms [23] as well as to differences in cellular concentration and/or temporal expression [17] , [24] . In our work , we demonstrated a stronger expression of hepatic iNOS and a higher NO production by liver leukocytes of infected B6 compared to BALB/c mice . Interestingly , this process was concomitant with high levels of pro-inflammatory cytokines detected in B6 mice . On the other hand , results presented here demonstrate that p47 and gp91-phox components of NADPH oxidase were increased in livers of both infected mouse strains compared to non-infected . However , an enhanced and sustained p47-phox expression was observed in livers from B6 mice . Strikingly , the co-expression of gp91 and p47-phox on the plasma membrane was found only in liver from infected B6 at 24dpi . Several authors have described that ROS can induce cell death by either apoptosis or necrosis in liver pathologies [25] , [26] . Thus , the activation of NADPH oxidase enzymatic complex would be a key player in the liver damage , probably as an instrument contributing to liver apoptosis and necrosis during infection in B6 mice . TLRs have been shown to be critical for the protection against infections . Recently , Padilla et al . 2009 , demonstrated that insufficient TLR activation contributes to the slow development of CD8+ T cell responses in T . cruzi infection [16] . However , the breakdown of homeostasis by infection can deregulate certain pathways , such as TLR-cytokines , and plays a key role in the pathogenesis of the liver disease . Our results show a similar TLR2 and TLR4 expression on hepatic macrophages , granulocytes and T cells of infected BALB/c and B6 mice . However , the TLR9 expression showed a clear difference in hepatic leukocytes from infected B6 and BALB/c mice . We found that only leukocytes from infected B6 mice maintained high expression of TLR9 during the acute phase . These results support the hypothesis that sustained TLR9 signalling might contribute to excessive and harmful inflammatory response in infected B6 mice . In accordance with our results , a crucial role of TLR9 in IL12 and IFNγ production during T . cruzi infection was recently shown [15] , [27] . Furthermore , it is noteworthy that uninfected BALB/c and B6 mice did not show differences in their TLR9 expression , yet the persistent levels of this receptor in infected B6 could involve some lost factors , that are present in BALB/c , which should modulate the excessive inflammatory response in liver . Interestingly , we found that TLR2 and TLR4 are differentially modulated in hepatocytes of infected BALB/c and B6 mice , suggesting that these innate immune receptors would play a role not only in immune cells but also in liver parenchyma cells . In this sense , it has been postulated that TLR signalling in parenchyma cells would be a key mechanism to prevent death caused by excessive cytokine release [28]–[30] . There are increased evidences demonstrating the potential role of TLR-ligands treatment as therapeutic approach and they have shown to be highly effective in the protection against protozoan , among them T . cruzi [15] , [31] , [32] . In our study we observed that pre-treatment with Pam3CSK4 , a TLR2-TLR1 agonist , before infection induced a marked reduction of pro-inflammatory cytokines , nitrite and transaminasa levels and a decrease in the number of hepatic inflammatory foci and consequently in the mortality of infected mice . Together these results suggest that TLRs and probably others PRR should contribute to the induction of anti-parasite immune response , but an altered TLR regulation could lead to tissue damage . Interestingly , TLRs sense the endogenous molecules released by dying cells and lead to fast IL1β release through inflammasome activation [33] . Notably , in our model the IL-1β cytokine was up regulated in infected B6 mice and it would be a very interesting idea to explore activation of inflammasome platform during acute phase of T . cruzi infection . In our study we postulated that the inadequate integration of signals involving molecular ( TLRs , cytokines , NO and ROS ) and cellular ( immune and parenchyma cells ) components influences the outcome of local immune response during this parasite infection . Moreover , the differential TLR and cytokine modulation in the liver , induced by T . cruzi infection , emphasize the importance of local innate immune response in hosts with different genetic background and could contribute to the understanding and design of novel immune strategies in controlling liver pathologies .	Trypanosoma cruzi , an obligate intracellular protozoan , is the etiological agent of Chagas Disease that represents an important public health burden in Latin America . The infection with this parasite can lead to severe complications in cardiac , liver and gastrointestinal tissue depending on the strain of parasite and host genetics . Recently , we reported a fatal liver injury in T . cruzi infected B6 mice . However , the local immune response against this parasite is poorly understood . This work highlights some of the molecular and cellular mechanisms involved in liver pathology during the acute phase of infection . Using two mouse strains with different genetic backgrounds and responses to infection , B6 and BALB/c , we found that infected B6 mice develop a strong pro-inflammatory environment associated with high TLR9 expression . Conversely , infected BALB/c mice showed a more balanced inflammatory response in liver . Moreover , higher TLR2 and TLR4 expression were found only in hepatocytes from BALB/c . These data emphasize the importance of an adequate integration of signalling between immune and non-immune cells to define the outcome of infection . In addition , the pre-treatment with TLR2-agonist reverts the strong pro-inflammatory environment in T . cruzi infected B6 mice . These results could be useful in the understanding and design of novel immune strategies in controlling liver pathologies .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "infectious", "diseases/protozoal", "infections", "infectious", "diseases/neglected", "tropical", "diseases", "immunology/immunity", "to", "infections", "immunology/innate", "immunity" ]	2010	Importance of TLR2 on Hepatic Immune and Non-Immune Cells to Attenuate the Strong Inflammatory Liver Response During Trypanosoma cruzi Acute Infection
The interplay of microbiota and the human host is physiologically crucial in health and diseases . The beneficial effects of lactic acid bacteria ( LAB ) , permanently colonizing the human intestine or transiently obtained from food , have been extensively reported . However , the molecular understanding of how LAB modulate human physiology is still limited . G protein-coupled receptors for hydroxycarboxylic acids ( HCAR ) are regulators of immune functions and energy homeostasis under changing metabolic and dietary conditions . Most mammals have two HCAR ( HCA1 , HCA2 ) but humans and other hominids contain a third member ( HCA3 ) in their genomes . A plausible hypothesis why HCA3 function was advantageous in hominid evolution was lacking . Here , we used a combination of evolutionary , analytical and functional methods to unravel the role of HCA3 in vitro and in vivo . The functional studies included different pharmacological assays , analyses of human monocytes and pharmacokinetic measurements in human . We report the discovery of the interaction of D-phenyllactic acid ( D-PLA ) and the human host through highly potent activation of HCA3 . D-PLA is an anti-bacterial metabolite found in high concentrations in LAB-fermented food such as Sauerkraut . We demonstrate that D-PLA from such alimentary sources is well absorbed from the human gut leading to high plasma and urine levels and triggers pertussis toxin-sensitive migration of primary human monocytes in an HCA3-dependent manner . We provide evolutionary , analytical and functional evidence supporting the hypothesis that HCA3 was consolidated in hominids as a new signaling system for LAB-derived metabolites . The interplay of microbiota and the human host is physiologically crucial in health and diseases . Lactic acid bacteria ( LAB ) are microorganisms present in many foods and the intestine of most mammals . There are extensive reports about the beneficial role of LAB on the immune system [1] . Short-chain fatty acids ( SCFAs ) and lactate are known metabolites of LAB that have been shown to play an important role in the maintenance of the gut barrier function [2] . SCFAs can induce effects in the host through activation of specific G protein-coupled receptors ( GPCRs ) expressed in intestinal epithelial cells and immune cells that are located in the intestinal mucosa [3] . The present study focuses on GPCRs belonging to the family of hydroxycarboxylic acid receptors ( HCAR ) which are regulators of immune functions and energy homeostasis under changing metabolic and dietary conditions . At least two HCAR subtypes are present in mammalian genomes . HCA1 ( formerly GPR81 ) is activated by lactate and HCA2 ( formerly GPR109a , HM74A , PUMA-G ) is activated by the ketone body D-3-hydroxybutyrate ( D-3HB ) but also by the SCFA butyrate . Both receptors mediate anti-lipolytic effects in adipocytes through Gi-protein coupling [4] . Further , HCA2 is known to be expressed in enterocytes , colonocytes and several types of immune cells including neutrophils and macrophages mediating anti-inflammatory effects [3] . A third HCAR subtype , HCA3 ( formerly GPR109b , HM74 ) , was recently identified in the human genome but is absent in the mouse genome [5] . The amino acid sequence of the human HCA3 differs from HCA2 in 16 positions and an extended C terminus ( Fig 1A ) . These differences are sufficient to change agonist specificity of HCA3 towards being activated by the fatty acid β-oxidation intermediate 3-hydroxyoctanoate ( 3HO ) , but not by D-3HB , likely also mediating anti-lipolytic effects under fasting conditions [6] . Further , aromatic D-amino acids were found to activate HCA3 and elicit chemotactic responses in human neutrophils [7] . However , a plausible hypothesis why HCA3 function is of advantage in humans is currently missing . Here , we reconstructed the evolutionary history of HCARs and experimentally show that HCA3 is functionally present in humans and all other great apes . After gene duplication and distinct structural changes HCA3 gained the ability to recognize 3HO as an agonist but lost D-3HB specificity . Moreover , we identified D-phenyllactic acid ( D-PLA ) , a metabolite produced by LAB as the so far most potent naturally occurring agonist acting at HCA3 . High levels of D-PLA can be found in LAB-fermented products , such as Sauerkraut ( S1 Table ) . LAB fermentation is an ancient process that happened even before humans took advantage of it . It is known that global transitions affected the last common ancestor of early hominoids causing changes in its diet , rendering ingested fruits and leaves more likely to be fermented before ingestion [8] . We provide functional and phylogenetic evidence supporting the hypothesis that increased intake of LAB-fermented food likely posed a positive selective pressure maintaining HCA3 function in hominids . We further hypothesize that HCA3 presence was advantageous in the interplay between ingested and gastrointestinal microbiome and the hominid host by taking over functions in the immune system . Feedback regulation of the energy metabolism is vital for organisms exposed to variable dietary supply . Receptors for intermediates of the energy metabolism , such as hydroxycarboxylic acids , already appeared in early vertebrate evolution [9] . Mining of public sequence databases revealed the presence of at least one HCA1 ortholog in the genome of cartilaginous , lobe- and ray-finned fishes , amphibians , and mammals but not in any sauropsidian species ( birds , reptiles ) ( Fig 1B , S1 Fig , S2 Table ) . HCA2 is present in all mammals and arose from an HCA1 gene duplication in early mammalian evolution ( Fig 1B , S2 Fig , S2 Table ) . We found that HCA3 , the evolutionarily youngest HCAR , is present in the genomes of all great apes and siamang , but absent in all other gibbon genomes investigated so far [10] . Because automated genome assembly can cause problems in assigning highly homologous sequences , we manually reanalyzed the genomic sequence traces for the presence of HCA2 and HCA3 and verified the findings by amplifying , cloning and sequencing HCAR from great apes , siamang , and white-cheeked gibbon . Our analysis showed that HCA3 arose from a duplication of HCA2 before the split of great apes and gibbons ( Fig 1B ) but underwent pseudogenization in most gibbon species . Subtypes resulting from gene duplication can have several fates , which include but are not limited to pseudogenization or gain of new function . In the latter case one copy may accumulate mutations and acquire unique functionality without risking the fitness of the organism , which is ensured by the remaining homolog . To test whether the persistence of HCA3 in great apes and some gibbons caused changes in evolutionary constraints of HCAR subtypes we performed Phylogenetic Analysis by Maximum Likelihood ( PAML ) [11] . Ape HCA1 evolved with an evolutionary rate ( ω = 0 . 189 ) significantly ( p = 0 . 0258 ) different compared to all other mammalian HCA1 ( ω0 = 0 . 105 ) whereas no significant difference in evolutionary constraint of ape HCA2 compared to other mammalian HCA2 orthologs was detected ( S3 Table ) . We further tested whether HCA3 evolved with a different evolutionary rate than HCA1 and HCA2 in species carrying all three HCAR subtypes . We found that HCA3 evolved with a significantly ( p = 0 . 0064 ) higher evolutionary rate ( ω = 0 . 257 ) than HCA1 and HCA2 ( ω = 0 . 103 ) , but significantly different from 1 ( S3 Table ) . Our data indicates that HCA3 is not drifting into neutrality ( pseudogenization ) in great apes but rather gained new functionality , a hypothesis we further addressed using functional analyses . To study whether structural differences contribute to the distinct functional properties of the three HCAR we heterologously expressed the human HCA1 , HCA2 and HCA3 in CHO-K1 cells and performed functional assays using the dynamic mass redistribution technology ( DMR; Corning Epic System ) . Compared to classical second messenger assays , DMR assays provide the advantage that cellular responses are recorded time-resolved and independently of the activated signaling cascades . Receptor activation is monitored kinetically , thus revealing potential differences in activation kinetics mediated by different agonists . As reported , lactate only activates HCA1 [12 , 13] , D-3HB activates HCA2 and with a delayed signal onset also HCA3 [14] and 3HO activates HCA3 but not HCA2 ( Fig 1D ) [6] . All HCAR are Gi protein-coupled receptors , i . e . activation of the receptor by an agonist leads to inhibition of adenylyl cyclases , thus resulting in a decrease in intracellular cAMP levels . Concentration-response curves of different agonists can be determined using cAMP inhibition assays . This extends the functional characterization since it reveals information about constitutive HCAR activity ( basal cAMP ) and allows the quantification of the agonistic activity at the respective HCAR by determination of EC50 and Emax values . EC50 values reflect the potency of an agonist , i . e . the concentration of an agonist required to produce 50% of its maximal effect; the lower the EC50 , the higher the potency of an agonist . Emax ( efficacy ) is the maximum effect induced by the agonist , i . e . when Emax is reached increasing the agonist concentration will not produce a greater magnitude of the effect . We performed cAMP inhibition assays on all hominoid HCARs as well as mouse HCA1 and HCA2 and analyzed them with the already established endogenous agonists lactate , D-3HB and 3HO ( Table 1 , Fig 2 ) . Considering the observed differences in expression levels ( Table 2 ) , we detected no significant differences in Emax and EC50 values upon stimulation with the respective endogenous agonists when comparing within HCA1 , HCA2 , and HCA3 orthologs ( Table 1 ) . However , the fact that D-3HB stimulates HCA3 orthologs ( Figs 1D , 2A and 2C ) to some extent supports its evolutionary origination from HCA2 . In contrast , 3HO had no activity on HCA2 orthologs , indicating a gain of functionality in great ape HCA3 after gene duplication ( Figs 1D , 2D and 2E ) . D-3HB activates siamang HCA2 with about 3-5-fold lower potency when compared to the human ortholog ( Fig 2B ) and siamang HCA2 is activated by 3HO at high concentrations ( Fig 2E ) . These findings and the fact that HCA3 is only present in siamang but not in any other gibbon species initiated further PAML analyses . We found that HCA2 and HCA3 of siamang evolve with an ω that is not significantly different from 1 ( S3 Table ) . This indicates a loss of constraint on siamang HCA2 and HCA3 . Further , when compared within all other available gibbon HCA1 and HCA2 orthologs , we found that siamang HCA1 evolves under purifying selection , but indeed siamang HCA2 exhibits an ω ( ω = 0 . 643 ) , not significantly different from 1 . Thus , we found that , in contrast to all other HCAR orthologs , the agonist profiles of siamang HCA2 and HCA3 are less distinguishable and therefore less conserved . In sum , our combined evolutionary and functional analyses support a loss of evolutionary constraint on the siamang HCA2 and HCA3 orthologs ( Fig 2 , S3 Table ) . Sequence alignment of HCA3 orthologs revealed that from all amino acid positions which differ from HCA2 seven amino acid positions are conserved in all HCA3 orthologs ( Fig 1B , 1C and S3 Fig ) . Further , we found that Ser91 and Thr173 ( referring to positions in human HCA3 NP_006009 . 2 ) are great ape-specific HCA3 positions , while Tyr86 and Trp142 are only conserved in human , chimpanzee , bonobo , and gorilla HCA3 ( Fig 1B , 1C and S3 Fig ) . Since 3HO activated all hominoid HCA3 orthologs with comparable potency , we next analyzed whether potencies of the HCA3 agonists D-phenylalanine ( D-Phe ) and D-tryptophan ( D-Trp ) [7] , vary between orthologs . We confirmed HCA3-specificity of both agonists ( S4 Fig ) and found that both , D-Phe and D-Trp , activate human HCA3 with highest potency when compared to all other ape orthologs ( Table 3 , S4 Fig ) . Although Irukayama-Tomobe et . al showed both aromatic D-amino acids to be HCA3-specific agonists , they did not identify relevant sources of D-Phe and D-Trp to put their findings in a physiological context [7] . By comprehensive research of literature , we here draw the link to fermented foods and beverages which are a likely source for D-Phe and D-Trp in concentrations sufficiently high to activate HCA3 . Several classes of bacteria produce and secrete aromatic D-amino acids ( e . g . Acetobacter , Bifidobacterium , Brevibacterium , Lactobacillus , Micrococcus , Propionibacterium , Streptococcus ) [15–17] and for all of them residence in the human gastrointestinal tract has been demonstrated [18] . Moreover , D-amino acids were found to be present in body fluids and certain tissues in the μM range [19–22] . Connecting this previously established knowledge led us to further investigate whether other microbial metabolites activate HCA3 . LAB are known to produce metabolites structurally related to 3HO and D-Phe , such as 3-hydroxydecanoate ( 3HDec ) and D-PLA , respectively [23 , 24] . PLA has been detected in μM concentrations in fermented food such as Sauerkraut for which remarkably stable microbial associations over time and region have been shown ( Literature summarized in S1 Table ) [25] . Moreover , LAB represent 0 . 01–1 . 8% of the total bacterial community found in the human intestine where some of them are colonizers and others are passengers [26] . Additional support for the link between LAB and D-PLA is provided by analyses of patients with short bowel syndrome ( SBS ) with their microbiota known to be imbalanced to Lactobacillus ( L . mucosae , L . acidophilus , L . fermentum ) [27 , 28] as the major resident bacteria ( increased from ≤ 1% up to 60% of the fecal flora ) . These patients exhibit highly increased urinary levels of D-PLA [29–33] . To our knowledge , the present study is the first in which D-PLA , L-phenyllactic acid ( L-PLA ) and indole 3-lactic acid ( ILA ) , a metabolite derived from Trp metabolism , were functionally tested for their agonistic activity at the human HCA3 . Using the DMR technology we found that all three metabolites activate the human HCA3 , but not HCA1 and HCA2 ( Fig 3A ) . D-PLA and ILA ( each 1 μM ) induced a response comparable to that of 100 μM 3HO , the known endogenous agonist indicating an about 100-fold higher potency of the LAB-derived compounds ( Figs 1D and 3A ) . Next , we functionally analyzed all HCA3 orthologs with the newly identified agonists using cAMP inhibition assays to determine their potencies ( Table 3 ) . Unfortunately , the individual potencies of D- and L-ILA could not be determined since only the DL isomeric mixture was commercially available . L-PLA activates HCA3 orthologs in the μM range . However , D-PLA activates HCA3 with an EC50 value of 150 nM , thus being an about 35-fold more potent agonist than the L-enantiomer and about 10-fold and 240-fold more potent than the known agonists 3HO and D-Phe , respectively ( Tables 1 and 3 ) . Further , both D-PLA and D-Phe exhibit higher potencies at the HCA3 of human , chimpanzee and gorilla when compared to the orangutan ortholog ( Fig 3B , 3C and S4 Fig ) . In contrast , the more hydrophobic agonists 3HDec , D-Trp and ILA are less potent at the chimpanzee HCA3 ortholog when compared to the other great ape HCA3 orthologs ( Fig 3B , 3C and S4 Fig ) . The Trp metabolite ILA acted at human HCA3 with a potency comparable to that of D-PLA . Thus , D-PLA is the most potent agonist at human HCA3 for which sufficient presence in LAB-fermented food has been previously described ( S1 Table ) . Accumulating evidence suggests that only a small number of LAB species are true inhabitants of the human intestinal tract . The majority of LAB is derived from fermented food , the oral cavity or more proximal parts of the gastrointestinal tract , like the esophagus [34] . However , we asked whether the gene for D-lactate dehydrogenase ( EC 1 . 1 . 1 . 28 ) , the enzyme that has been shown to produce D-PLA [35–37] , is present in the human microbiome . We found that genes encoding D-lactate dehydrogenase are very common among bacteria and are also found in some of the very common and abundant gut bacteria ( S5 Fig , S4 Table ) . However , the question remains , if these D-lactate dehydrogenases are capable of converting the aromatic amino acids to the respective lactic acid in the human intestine . Even if this ability is restricted to LAB possessing this gene , they are still detectable to some degree in the human intestinal microbiome [38] . To further substantiate D-PLA resorption from gut and its urinary elimination , a pharmacokinetic study was performed determining PLA levels using Liquid Chromatography Mass Spectrometry ( LC-MS ) in human plasma and urine samples . As shown in Fig 4A , PLA plasma levels are increased already 30 minutes after oral application of 100 mg D-PLA and subsequently decline rapidly due to renal excretion as determined in concomitantly collected urine samples . PLA plasma levels reached concentrations ( >20 μM ) capable to maximally activate the human HCA3 ( Fig 3B ) . To verify this , we tested whether PLA-containing plasma and urine samples elicit an HCA3-specific , pertussis toxin ( PTX ) sensitive decline in intracellular cAMP levels of transiently transfected CHO-K1 cells ( Fig 4B , S5 Table ) . Thus , plasma and urine samples were diluted and assayed by addition to transfected CHO-K1 cells . Indeed , we found that PLA containing urine and to a lesser extend plasma activated specifically and concentration-dependent HCA3 transfected cells ( Fig 4B , S5 Table ) . Interestingly , we also found basal plasma levels of PLA of about 0 . 4 μM before oral D-PLA administration ( Fig 4A , S6 Fig ) . Our established LC-MS measurement method does not discriminate between L-PLA and D-PLA . Blood levels of L-PLA in the μM range are only found in patients with phenylketonuria [39] . L-PLA is 35-fold less potent than D-PLA at HCA3 ( Table 3 ) and plasma samples from basal conditions ( 0 . 4 μM ) were sufficient to inhibit forskolin-induced cAMP formation via HCA3 ( S5 Table ) , potentially due to presence of D-PLA derived from alimentary microbiota sources . Sauerkraut is known to contain high levels of LAB as well as D-PLA . A recent study showed that Sauerkraut exhibits remarkably stable microbial associations over time and regions [40] . Therefore , we asked whether ingestion of Sauerkraut ( 5–6 g per kg body weight ) can cause an increase in plasma and urinary D-PLA levels . Using LC-MS we found that levels of PLA in plasma ( before = 0 . 2 μM; after = 0 . 9 μM ) and urine ( before = 0 . 7 μM; after = 9 . 1 μM ) are increased 2 h postprandial in four individuals ( Fig 4C , S6A Fig ) . Further , we found that plasma and urine containing increased concentrations of PLA after Sauerkraut ingestion induced an HCA3-specific , PTX-sensitive inhibition of intracellular cAMP levels ( Fig 4D , S6B Fig , S5 Table ) . This clearly indicates that fermented food , such as Sauerkraut , can be a source of D-PLA leading to functionally relevant concentrations of this bacterial metabolite in humans . Until recently , mRNA expression datasets based on microarrays constituted the main source for expression data . However , due to their high nucleotide sequence similarity ( 97% ) these datasets did not enable the distinction between human HCA2 and HCA3 . High quality , paired-end , RNA-Sequencing datasets are in principle suitable to distinguish between those highly similar sequences . Indeed , public resources provide quantitative expression data ( FPKM: fragments per kilobase million , TPM: transcripts per kilobase million ) for both HCA2 and HCA3 . Quantification of transcripts in RNA-Sequencing projects is based on counting small RNA fragments mapped to a gene ( read lengths 70–100 bp ) . Because long nucleotide sequence parts of HCA2 and HCA3 transcripts are identical it was not clear whether the used bioinformatic tools are precise enough to properly distinguish and quantify human HCA2 and HCA3 transcripts . Therefore , we manually reanalyzed a publicly available dataset ( Bioproject: PRJNA326727 ) ( S6 Table , S7 Fig ) . We mapped all relevant reads to HCA2 or HCA3 and excluded those matching to both receptors . We found no significant differences between the provided FPKM values and our manually curated read counts ( S7 Fig ) . Mining of publicly available RNA-Sequencing data reanalyzed and provided as TPM values revealed highest expression of HCA3 in immune cells such as neutrophils and monocytes and a pattern distinct from that of HCA2 ( S7 Fig , S7 Table ) . Expression of HCA3 mRNA , especially in adipose tissue , skin and lung , is considerably lower compared to HCA2 ( S7 Fig ) . However , both receptors are up-regulated in ulcerative colitis and inflammatory bowel disease patients compared to controls ( S6 Table , S7 Fig ) . As described above , HCA3 is expressed in a variety of human immune cells including macrophages , neutrophils , and monocytes ( S7 Table ) . First , we used freshly isolated human peripheral blood mononuclear cells ( PBMCs ) , which include lymphocytes ( T cells , B cells , and NK cells ) , monocytes , and dendritic cells . We observed a significant reduction in cAMP levels for all HCA3 agonists except of 10 μM L-PLA in PBMCs ( S6C Fig ) . However , PTX sensitivity of this reduction could only be observed for 3HO and D-PLA , but not the other HCA3 agonists ( S6C Fig ) . We assume that this is likely due to inter-individual variation in cell counts of the different cell types in PBMCs . Thus , we isolated human monocytes and performed cAMP inhibition assays . We found that 10 μM 3HO as well as 1 μM D-PLA and ILA induced a decrease of cAMP in human monocytes that can be blocked by PTX ( Fig 4E ) . Using a transwell migration assay we showed that D-PLA can trigger chemotactic responses with high potency in isolated monocytes ( Fig 4F ) . Both , a PTX-sensitive reduction in cAMP levels and migratory responses were also observed for D-Phe and D-Trp with freshly isolated human monocytes , however , 1000-fold higher concentrations were required ( S6D Fig ) . Finally , we performed siRNA-mediated knock-down of HCA3 in freshly isolated human monocytes ( S6E Fig ) and determined cAMP levels upon stimulation with 3HO and D-PLA in comparison to scrambled negative control siRNA ( siNC ) -transfected cells ( Fig 4G ) . In contrast to control ( siNC ) -transfected monocytes the reduction in intracellular cAMP levels was abolished in siHCAR3-transfected monocytes demonstrating HCA3 specificity of the D-PLA induced effect ( Fig 4G ) . A recent paleogenetic study provides a comprehensive summary of existing evidence for large-scale ecological transitions occurring during hominid evolution that caused changes of the habitat accompanied by changes in diet and the microbial environment [8] . The exposure to an altered microbial environment most likely posed a selective pressure and required host adaptation but may also have provided the advantage to access new ecological niches . In the present study , we provide evidence for such a unique example , where genetic events potentially improved the availability of a new food repertoire under changed ecological conditions that possibly triggered the fixation of a duplicated gene with new functions in hominids . Through analyses of the evolutionary history of the HCA receptor family , we now show that HCA3 , being absent in non-hominoid primates and all other vertebrates , resulted from a gene duplication that occurred before the split of gibbons from great apes ( Fig 1B ) . In apes , we found gradual fixation of amino acid positions in HCA3 ( Fig 1C ) . We discovered new LAB-derived agonists acting at all ape HCA3 orthologs . Both , D-PLA and ILA , activate human HCA3 with an about 10-fold higher potency when compared to the endogenous agonist 3HO ( Tables 1 and 3 ) . Further , our functional analyses of great ape HCA3 orthologs revealed , that potencies of D-PLA are significantly higher in human , chimpanzee and gorilla compared to orangutan ( Fig 4C ) . This led us to do an extensive literature search to identify potential global environmental changes that conceivably coincided with dietary changes and are archaeologically accepted to have occurred when the last common ancestor of human , chimpanzee and gorilla lived on earth . As mentioned at the beginning of the discussion , a recent study showed that about 10 million years , ago a mutation in the class IV alcohol dehydrogenase appeared in the common ancestor that humans share with gorillas and chimpanzees , which allowed an increased tolerance of alcohol [8] . In this study , a scenario was projected in which the large-scale ecological transitions caused the common ancestor to move out of the trees and lead a more terrestrial life [8] . Carrigan et al . conclude that this new life style rendered fruit picking directly from the trees to be less likely but increased the probability that food that had started to ferment was collected from the ground [8] . This is suggested to have posed the selective pressure that increased the tolerance to dietary alcohol before human-directed fermentation [8] . In this context , not only the increased activity of ethanol-metabolizing enzymes ( e . g . ADH4 ) might have provided a selective advantage but also an immune system that can sense high levels of D-PLA ingested with lacto-fermented food ( including plants , fish , meats , milk ) [35] . Our functional data combined with the described archaeologically documented evolutionary context gives rise to the hypothesis that in the last common ancestor of human , chimpanzee and gorilla , increased ingestion of fermented food was likely associated with increased ingestion of LAB accompanied by increased levels of D-PLA . D-PLA is a metabolite , produced in high concentrations by LAB in the process of lactic fermentation and has a broad antimicrobial activity against bacteria and fungi [35] . We demonstrate , using LC-MS that D-PLA is quickly absorbed from the gastrointestinal tract upon ingestion of LAB-fermented Sauerkraut , can reach physiologically relevant plasma concentrations and is renally eliminated in humans ( Fig 4A and 4C ) . These diet-induced PLA plasma concentrations are sufficient to activate HCA3 in human monocytes and potentially modulate human immune functions ( Fig 4E–4G ) . The beneficial effects of LAB-fermented food on human health are numerous , ranging from protection of the intestinal barrier , limitation of inflammatory cytokine production to improved fasting insulin levels and glucose turnover rates among many others [41] . HCA3 is expressed in many different immune cells but also adipose tissue ( S7 Table ) . We hypothesize that the LAB-mediated , HCA3-dependent physiological impact likely extends to influences on the human host energy storage . This would also be plausible in an evolutionary context , since it is known that the nutritional and functional properties of food are enhanced by LAB-fermentation due to the transformation of substrates and the formation of bioactive or bioavailable end-products [25] . In summary , our work provides multifaceted evidence supporting the hypothesis that HCA3 evolved as a GPCR activated by LAB metabolites . Our data suggests that this might have improved tolerance to the increased ingestion of LAB in the last common ancestor of human , chimpanzee and gorilla . As LAB-derived D-PLA is specifically recognized by HCA3 expressed in monocytes ( Fig 4H ) and numerous studies describe increased hypo-responsiveness of the immune system through anti-inflammatory processes for D-PLA and LAB ( S1 Table ) , we speculate that HCA3 plays a role in mediating at least some of those effects ( Fig 4H ) . In the present study , we identified a new set of microbial-derived metabolites acting as classical signaling molecules in the human host through recognition by a specific receptor . The question remains how D-PLA affects monocyte functions . Future studies shall address whether HCA3 activation by D-PLA impacts phagocytic capacity of monocytes or influences differentiation of monocytes to macrophages . Does D-PLA , through activation of HCA3 , prime monocytes to either increase of pro-inflammatory host response to concomitantly ingested pathogenic bacteria or reduction of pro-inflammatory response to LAB ? At last , our study opens-up the interesting question of how D-PLA ingested with LAB-fermented food can influence energy storage in HCA3-expressing adipocytes . The studies on humans and with human materials were conducted in accordance with the Declaration of Helsinki and with the recommendations of “Ethik-Kommission an der Medizinischen Fakultät der Universität Leipzig” with written informed consent from all blood donors . The protocol was approved by the aforementioned committee ( 313/14-ek ) . HCA1 , HCA2 and HCA3 of 54 selected vertebrate species were aligned to infer the phylogenetic model of HCAR evolution as shown in Fig 1 ( S1 Fig ) . Mammalian HCA1 and HCA2 were aligned to infer a phylogenetic tree of 88 mammalian species . All nucleotide alignments were generated with the ClustalW algorithm ( Bioedit Sequence Alignment Editor 7 . 0 . 9 [43] ) followed by manual trimming where gaps were deleted . Phylogenetic evolutionary history was inferred by the Maximum Likelihood method based on the General Time Reversible model using MEGA6 [44] . The resulting tree ( S2 Fig ) with a 75% cut-off value was used for PAML analyses . Tests of selection ( ω = dN/dS ) were accomplished by maximum likelihood using a codon-based substitution model implemented in PAML version 4 . 2 [11] . Branch models [45] that allow ω to vary among branches in the phylogeny were applied to determine ω ratios along particular lineages . Likelihood ratio tests ( LRT ) were performed to test nested competing hypotheses ( S3 Table ) . CHO-K1 cells were grown in Dulbecco’s Modified Eagle Medium: Nutrient Mixture F-12 ( DMEM/F12 ) supplemented with 10% fetal bovine serum ( FBS ) , 100 U/ml penicillin and 100 μg/ml streptomycin . Cells were maintained at 37 °C in a humidified 5% CO2 incubator . For transient transfection Lipofectamine 2000 ( Life Technologies , Darmstadt , Germany ) was used . Cells were split into 25 cm2-cell culture flasks ( 0 . 9 x 106 cells/flask ) and transfected with a total amount of 3 μg of plasmid the following day . All 134 RNA-sequencing samples belonging to Bioproject: PRJNA326727 , ( GEO Accession: GSE83687 ) [47] were manually reanalyzed and compared to available FPKM values as stated in S6 Table and S7 Fig . Two large collections of metagenomic data from the human gut microbiome were searched for D-lactate dehydrogenase ( EC 1 . 1 . 1 . 28 ) , the enzyme that catalyzes the conversion of D-lactate to pyruvate , but has also been shown for some LAB to convert phenylpyruvate to D-PLA . Metagenomic data from 1 , 885 human stool samples from the curated Metagenomic Dataset ( https://github . com/waldronlab/ curatedMetagenomicData ) [48] were searched for 1 , 443 EC 1 . 1 . 1 . 28-related UniRef100 entries . In addition , genes with the K03778 annotation were searched in the Integrated Gene Catalogue ( IGC ) [49] reference data set for the human gut microbiome , which comprises data from just over 1 , 000 individuals . Detailed information , a representative tree of the detected genes and a list of all entries are summarized in the Supplementary Information ( S5 Fig , S4 Table ) . 100 mg of D-PLA were dissolved in 30 ml H2O and ingested by a healthy male subject ( 78 kg , 1 . 80 m ) . 2 . 5 ml of EDTA blood samples were taken right before and 30 min , 60 min , and then every hour up to 7 h after ingestion . In a subsequent , chronically separate experiment the same healthy male subject adapted to a lacto-vegetarian diet , including uptake of 500 g per day of fresh ( non-pasteurized ) Sauerkraut ( Spreewaldhof , Germany ) for three days . 2 . 5 mL of EDTA blood samples were taken before diet , 2 h postprandial over three days and 5 days after having stopped the above described diet . Further , urine samples were taken over the time course . Similarly , three healthy female subjects ingested 5–6 g of fresh ( non-pasteurized ) Sauerkraut per kg body weight for one day . 2 . 5 mL of EDTA blood samples and urine samples were taken before ingestion and 2 h postprandial . Fresh whole blood samples were centrifuged for 10 min at 4 °C and 800xg to isolate plasma , which was stored in 500 μl aliquots at -80 °C until further analyses . Urine was collected and the total volume was determined . 10 μl of human plasma and urine were treated with 90 μl precipitating agent ( acetonitrile including the internal standard phenyl-d5-lactate at 10 ng/ml ) , thoroughly mixed and centrifuged for 5 min at 13 , 000xg . Supernatants were transferred to autosampler vials and 10 μl were injected into the Liquid Chromatography Mass Spectrometry ( LC-MS ) system . It consisted of a Prominence UFLC system from Shimadzu ( Duisburg , Germany ) and a QTRAP 6500 from SCIEX ( Framingham , MA , USA ) . Chromatographic separation took place on a Luna HILIC column ( 3 μm , 50 x 2 mm ) from Phenomenex ( Darmstadt , Germany ) via gradient elution at a flow rate of 0 . 4 ml/min . The mobile phase was 15 mmol/l ammonium acetate buffer ( pH 6 ) in acetonitrile and water . Electrospray ionization was applied in negative mode . Mass transitions were m/z 165>103 and m/z 170>108 for phenyl-lactate and phenyl-d5-lactate , respectively . Quantitation was carried out using a semi-quantitative approach following Eq 1 with c and A as the concentration and the peak area of the analyte ( An ) and the internal standard ( IS ) . The response factor ( RF ) was determined from signal responses in standard solutions and equaled 1 .	Although it has been known for 15 years that HCA3 is present in humans and other hominids but absent in all other mammals , no study so far aimed to understand why HCA3 was functionally preserved during evolution . Here , we take advantage of evolutionary analyses which we combine with functional assays of hominid HCA3 orthologs . In search for a reasonable scenario explaining the accumulated amino acid changes in HCA3 of hominids we discovered D-phenyllactic acid ( D-PLA ) , a metabolite produced by lactic acid bacteria ( LAB ) , as the so far most potent agonist specifically activating HCA3 . Further , oral ingestion of Sauerkraut , known to contain high levels of D-PLA , caused subsequent plasma concentrations sufficient to activate HCA3 . Our data interpreted in an evolutionary context suggests that the availability of a new food repertoire under changed ecological conditions triggered the fixation of HCA3 which took over new functions in hominids . These findings are particularly important because they unveiled HCA3 , which is not only expressed in various immune cells but also adipocytes , lung and skin , as a player that transfers signals of LAB-derived metabolites into a physiological response in humans . This opens up new directions towards the understanding of the versatile beneficial effects of LAB and their metabolites for humans .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "blood", "cells", "medicine", "and", "health", "sciences", "immune", "cells", "body", "fluids", "immunology", "vertebrates", "animals", "mammals", "urine", "primates", "metabolites", "gibbons", "digestive", "system", "white", "blood", "cells", "animal", "cells", "ev...	2019	Metabolites of lactic acid bacteria present in fermented foods are highly potent agonists of human hydroxycarboxylic acid receptor 3
Efficient HTLV-1 viral transmission occurs through cell-to-cell contacts . The Tax viral transcriptional activator protein facilitates this process . Using a comparative transcriptomic analysis , we recently identified a series of genes up-regulated in HTLV-1 Tax expressing T-lymphocytes . We focused our attention towards genes that are important for cytoskeleton dynamic and thus may possibly modulate cell-to-cell contacts . We first demonstrate that Gem , a member of the small GTP-binding proteins within the Ras superfamily , is expressed both at the RNA and protein levels in Tax-expressing cells and in HTLV-1-infected cell lines . Using a series of ChIP assays , we show that Tax recruits CREB and CREB Binding Protein ( CBP ) onto a c-AMP Responsive Element ( CRE ) present in the gem promoter . This CRE sequence is required to drive Tax-activated gem transcription . Since Gem is involved in cytoskeleton remodeling , we investigated its role in infected cells motility . We show that Gem co-localizes with F-actin and is involved both in T-cell spontaneous cell migration as well as chemotaxis in the presence of SDF-1/CXCL12 . Importantly , gem knock-down in HTLV-1-infected cells decreases cell migration and conjugate formation . Finally , we demonstrate that Gem plays an important role in cell-to-cell viral transmission . Five to 10 million people are infected with the HTLV-1 retrovirus ( Human T-cell Leukemia Virus Type 1 ) worldwide [1]; and 1–6% of infected people will develop either Adult T-cell Leukemia ( ATL ) [2] , a malignant lymphoproliferation of mature activated T-cells , or inflammatory disorders , such as Tropical Spastic Paraparesis/HTLV-1 Associated Myelopathy ( TSP/HAM ) [3] , [4] . The long period of clinical latency between primo-infection and ATL outcome suggests that the pathogenesis is a complex multistep process [5] . The exact mechanism by which infected individuals develop ATL is still debated , although the Tax viral oncoprotein has clearly been associated with cell transformation . Indeed , Tax not only activates viral expression but it also triggers a wide range of cell-signaling pathways , reprograms cell cycle , interferes with control checkpoints and inhibits DNA repair [6] , [7] . Tax immortalizes/transforms T-lymphocytes in vivo and in vitro [8] and promotes tumors in transgenic mice [9] . In contrast to HIV-1 , HTLV-1-infected cells produce few free viral particles [10] . Therefore , efficient transmission of HTLV-1 from infected to uninfected T-cells relies on cell-to-cell contacts , both in vitro and likely in vivo [7] , [11] . Transmission of HTLV-1 particles to an uninfected T-lymphocyte occurs through two mutually non-exclusive models: ( i ) formation of a virological synapse between an infected lymphocyte and an un-infected lymphocyte ( involving cytoskeleton reorganization and reorientation of the microtubule-organizing center ( MTOC ) in the infected T-lymphocyte ) [12] or ( ii ) formation and transfer of a viral biofilm-like structure [13] . In the first situation , Tax promotes MTOC polarization and is found at the cell-to-cell junction and around the MTOC , together with the viral gag p19 protein [14] , [15] , while viral particles might be transferred within the synaptic cleft [16] . In the biofilm model , HTLV-1 particles are stored and attached together on the outer cell surface in a virus-induced extracellular matrix composed of collagen , agrin and cellular linker proteins ( tetherin and galactin3 ) . During cell-to-cell contacts , these embedded extracellular virions spread and infect target cells . Tax plays also an important role by increasing the production of collagen , a component of this matrix [13] . Whether other cellular genes are targeted by Tax and enhance cell-to-cell transfer is currently a matter of investigation . Using GeneChip Affymetrix technology , we recently identified a series of genes up-regulated in T-lymphocytes transduced by Tax-expressing lentivirus [17] . Among these genes , we focused our attention towards those that are important for cytoskeleton dynamic and thus possibly modulate cell-to-cell contacts . Gem , a member of the small GTP-binding proteins within the Ras superfamily [18] , was found to be up-regulated following Tax expression [17] . Gem , together with Rad , Rem , and Rem2 , form a subfamily of atypical proteins of the Ras family , often termed RGK ( for Rad and Kir/Gem ) . Indeed , these proteins present significant variations , compared with other Ras family proteins in the regions involved in binding phosphate moieties of guanine nucleotides . The conserved threonine residue ( T35 ) of the switch 1 region ( often referred to as the effector region ) that coordinates magnesium ion complexing the beta and gamma phosphates of GTP ( Mg-GDP/GTP binding ) is absent , and the DXXG region of switch 2 which contributes to nucleotide binding and GTPase catalysis is profoundly modified to DXWE . As a consequence , Gem markedly prefers GDP over GTP , and more importantly , Gem has an undetectable intrinsic GTPase activity [18]–[20] . Activation or deregulation of Ras signaling pathways causes cell growth , differentiation and survival and can ultimately lead to oncogenesis and cancer [21] . Gem expression is induced in human peripheral blood T-cells by mitogens such as PHA or PMA [22] and promotes cell shape remodeling by restructuring actin cytoskeleton and microtubule networks . In epithelial cells , Gem induces cell elongation , with disappearance of actin stress fibers , loss of focal adhesion and enhancement of the actin cortical network [23] . Using a library of randomized hybrid ribozymes , Gem was identified as an actor in cell invasiveness , a process that is essential for tumor metastasis [24] . In neuroblastoma cells , Gem antagonizes Rho kinase-induced neurite retraction and morphological differentiation [18] . Therefore , we hypothesize that Gem could play an essential role in HTLV-1 transmission . We demonstrated here that Gem is highly expressed at the RNA and protein levels in Tax-expressing cells and HTLV-1-infected cell lines . We further delineated the mechanism of Tax-induced gem transactivation and show that Tax recruits CREB and the CBP co-activator onto a CRE sequence present on the gem promoter . Gem protein co-localizes with F-actin and is involved in spontaneous cell migration and chemotaxis . Importantly , we observed that gem knock-down decreases the rate of HTLV-1-infected cell migration and conjugate formation . Most importantly , we demonstrated that Gem expression in HTLV-1-infected cells is involved in viral transmission , from infected cells to target cells through enhanced cell-to-cell contacts . Using transcriptional profiles analyses , we recently identified genes that are up-regulated in T-lymphocytes transduced by Tax lentivirus [17] . In order to identify genes that could play a role in cell transmission , we focused our retrospective analysis on genes that are involved in cytoskeleton remodeling . Interestingly , gem mRNA was strongly up-regulated ( 28 fold ) in Tax expressing vs . control cells . We also performed a database search on HTLV gene expression profiles and uncovered that gem up-regulation was also listed in previous studies that used high-throughput approaches on Tax expressing lymphocytes , HTLV-1-infected cells lines or HTLV-1 immortalized cells [25]–[27] ( Table S1 ) . These observations led us to investigate in more details the mechanisms and biological consequences of Tax-induced gem transactivation . We first confirmed the microarray analyses by performing RT-PCR experiments and demonstrated that gem expression is strongly up-regulated in 293T cells transduced with Tax lentiviral particles but not in cells transduced with a control lentivirus ( Figure 1A lane 2 vs . 3 ) . Gem protein was also strongly overexpressed in 293T and MOLT4 cells transduced with Tax-expressing lentivirus , while it was undetectable in the absence of Tax expression ( Figure 1B and 1C ) . As controls , Tax and GAPDH protein levels were evaluated by western blot . We also used the Tax inducible JPX-9 T-cell line to confirm the effect of Tax on Gem expression ( Figure 1D ) . Tax expression was induced by addition of ZnCl2 in the culture media for 24 h or 48 h . Interestingly , in that lymphocytic cell line , Gem expression also follows Tax induction . We then assessed Gem expression in a series of HTLV-1-infected cell lines ( C8166 , C91/PL , Hut102 and MT2 ) . All cell lines tested showed a high level of Gem protein expression , while in non-infected T-lymphocytes ( CEM , Jurkat and MOLT4 ) Gem expression was not detectable ( Figure 1E ) . As controls , Tax and GAPDH protein levels were evaluated by western blot . Altogether , these results demonstrate that Tax expression is sufficient to induce Gem expression both at RNA and protein levels in different cell types . The sequence of gem promoter contains a cAMP Responsive Element ( CRE ) [28] . This sequence is very similar to the Tax Responsive Elements ( TRE ) present in the U3 region of the 5′ viral Long Terminal Repeat ( Figure 2A ) . To determine whether the CRE sequence in the gem promoter is required for Tax-mediated transactivation , we compared transactivation levels of reporter plasmids encompassing either the full-length gem promoter ( Gem-luc ) or the CRE-deleted promoter ( ΔCRE-Gem-luc ) ( Figure 2A ) . Reporter plasmids were transfected in 293T cells previously transduced with lentiviral particles allowing Tax expression . In contrast to the control vector , Tax significantly activated transcription from the wild-type gem promoter ( p = 4 . 10−4 , ANOVA and Tukey post-hoc test ) , but not from the ΔCRE promoter , thus indicating that the CRE sequence is indeed necessary for Tax-mediated transactivation ( Figure 2B ) . It is well established that Tax activates the viral promoter ( 5′LTR ) through its ability to bind CREB and its co-activators CREB Binding Protein ( CBP ) and p300 [29] . To determine whether Tax activates transcription from the gem promoter through a similar mechanism , we performed Chromatin immunoprecipitation assays ( ChIP ) using extracts from C8166 cells ( Tax-expressing cells ) . This allowed us to demonstrate that Tax , CREB and CBP were present on the gem promoter ( Figure 2C lanes 3 , 4 and 5 ) . The HTLV-1 LTR ( lower panel ) was used as control for complex binding . These results indicate that gem transcription results from a direct interaction between Tax , CREB and CBP on the CRE sequence present in the gem promoter . Gem was previously reported to be involved in cell cytoskeleton remodeling [23] . To determine whether Gem colocalizes with F-actin , a Gem-expressing vector was transfected and immunofluorescence experiments were performed in different cell types ( Figures 3 , Figure 4 and Figure 5 ) . Gem promoted cell elongation , with a disappearance of actin stress fibers and a loss of focal adhesion points ( Figure 3A ) , which were still present in cells transfected with backbone vector ( Figure 3B , see actin stress fiber ( white arrows ) and focal adhesion points ( yellow arrows ) ) . Importantly , Gem colocalized with cortical actin filaments in non-T and in T cells ( Figure 3A , Figure 4 and Figure 5 ) . Gem-expressing cells presented various shapes and lengths with very long extensions as well as membrane protrusions ( Figure 4 ) . As previously described for N1E-115 mouse neuroblastoma cells and NIH3T3 cells [30] , Gem expression induced unusual dendritic morphology with ( Figure 4 , right panel “4” ) or without ( left panel “1” ) a large “synaptic bouton” morphology . Flattened bipolar ( “2” ) or multipolar ( “3” ) cells , notably described as an intermediate phenotype in N1E-115 cells expressing a mutated MLC ( 18D , 19D ) were also observed . Due to its role in cytoskeleton reorganization and reorientation of the MTOC during the formation of the virological synapse [12] , [14] , [15] , we hypothesized that Tax could also colocalize with Gem in the actin-rich membrane protrusions . A Flag-Tax plasmid was transfected together with a HA-Gem plasmid and immunofluorescence experiments were performed . Tax and Gem colocalization could not be observed ( Figure 5B ) . Even at high magnification ( Figure 5B ) , Tax cytoplasmic speckles ( visible in green ) did not merge with Gem signal ( visible in red ) . This result was also confirmed by co-immunoprecipitation experiments , which did not demonstrate any interaction between Tax and Gem ( data not shown ) . Cell mobility follows a perpetual reorganization of the actin cytoskeleton , where monomers of actin polymerize to form fibers [31] . A previous report demonstrated that HTLV-1-infected T-lymphocytes have increased mobility [32] , which could be beneficial for cell-to-cell viral transmission by increasing the probability of encountering uninfected target cells . Moreover , HTLV-1-infected T-lymphocytes induce blood brain barrier ( BBB ) disruption . This allows infected cells to enter into the central nervous system ( CNS ) [33] . Because Gem promotes reorganization of the cytoskeleton , we hypothesized that it could play a role in the observed enhanced cell motility of HTLV-1-infected cells . We therefore performed a series of Transwell migration assays . The ability of cells to transfer from the upper chamber to the lower chamber of a Transwell permeable filter was measured and the percentage of migrating cells was then quantified by flow cytometry . We first observed that C91/PL and Hut102 ( HTLV-1-infected cell lines ) had a higher mobility than non-infected MOLT4 T-cells ( Figure 6A , p<0 . 001 , ANOVA and Dunnet post-hoc test ) . Interestingly , the ability of HTLV-1 cells to migrate seemed to correlate with Gem expression levels ( Figure 6B ) , suggesting that HTLV-1-induced migration increase was associated with Tax-induced Gem overexpression . To test this hypothesis , Gem siRNA or control siRNA were transfected into C91/PL cells . As expected , Gem expression dramatically decreased in Gem siRNA transfected cells but not in cells transfected with irrelevant siRNA ( Figure 7A , lanes 6–8 vs . 2–4 ) . As controls , Tax and actin western blots were performed and did not show any significant change ( Figure 7A lower panels ) . Using Transwell migration assays , we then measured the spontaneous migration of C91/PL after transfection of Gem siRNA or control siRNA ( Figure 7B ) . A 33% decrease in migration was observed in C91/PL transfected with Gem siRNA ( Figure 7C , p = 0 . 018 , Student's t-test ) . As a control , Gem western blot was performed using cell extracts from cells transfected with control or Gem siRNA ( Figure 7D ) . Altogether , these results demonstrate that Gem plays an essential role in the enhanced cellular migration of HTLV-1-infected cells . To test whether Gem is sufficient to increase cell motility , we then performed a wound-healing assay ( Figure 8 ) . Seventy-two hours post-transduction with control or Gem lentiviral particles , Hela cell monolayers were wounded and a first image was taken ( time = 0 h ) . Of note , the presence of an IRES-GFP in the lentiviral vectors allows the monitoring of transduced ( ie . GFP-positive cells ) cells . Pictures were then taken after 3 h , 6 h , 9 h and 12 h , and distance between the 2 fronts of migration was measured ( Figure 8A–B ) . Interestingly , after 12 h , Gem-expressing cells migrated 33% faster than cells transduced with the control lentiviral particles ( Figure 8B ) . Of note , 100% healing would mean that the wound is totally closed . The difference between control and Gem expressing cells was visible at each time point ( Figure 8B ) . As a control , Gem expression was determined by western blot ( Figure 8C ) . These results demonstrate that Gem expression is sufficient to increase Hela cells spontaneous migration . Since T-cells are the primary targets of HTLV infection in vivo , we then assessed the role of Gem in chemokinesis ( spontaneous cellular migration ) and in chemotaxis ( in presence of chemoattractant ) after MOLT4 transduction with Gem or control lentiviral particles . The ability of cells to transfer from the upper chamber into the lower chamber of a Transwell permeable filter was measured in absence or presence of a chemoattactant ( SDF1/CXCL12 ) . Percentage of migrating cells was then quantified by flow cytometry ( Figure 9A ) . We first assessed the localization of Gem in MOLT4 transduced cells and observed colocalization of Gem with cortical actin ( Figure 9B ) . This result is consistent with results observed in Figures 3A , 4 and 5 . An 8-fold increase in migration was observed in T-cells transduced with Gem lentiviral particles compared to controls ( Figure 10A , p<0 . 0001 , Student's t-test ) . In the presence of SDF1/CXCL12 ( Figure 10C ) , 40% of Gem transduced MOLT4 cells migrated into the lower chamber , while only 15% of the MOLT4 cells transduced with the control lentiviral particles migrated ( p<0 . 0001 , Student's t-test ) . As a control , Gem expression was monitored by western blot ( Figure 10B and 10D ) . These experiments demonstrate that Gem expression is sufficient to increase T-cell chemokinesis and chemotaxis . Efficient transmission of HTLV-1 from infected to uninfected T-cells relies on cell-to-cell contacts . Tax and p8 ( cleavage product of p12 ) were previously shown to be involved in formation of cell-to-cell conjugates [14] , [34] . In order to determine whether Gem plays a role in this process , Hut102 ( HTLV-1-infected ) were transfected with Gem siRNA or control siRNA before culturing them with Jurkat target cells that have been labeled with dye ( Figure 11A ) . A 40% reduction ( p = 0 . 05 , Mann-Whitney test ) in the number of T-cell conjugates ( Figure 11A , white arrows ) was observed when Jurkat cells were incubated with Hut102 cells transfected with Gem siRNA ( Figure 11B ) . Similar results were obtained with MT2 cells ( data not shown ) . These results indicate that Gem is directly involved in the formation of T-cell conjugates and is , therefore , likely to play a role in viral transmission from infected cells to target cells . Finally , to assess whether Gem could be involved in viral transmission , we performed a cell-to-cell viral transfer experiment . Hut102 cells were transfected with Gem siRNA or control siRNA . Forty-eight hours post transfection they were co-cultured for a short period of time ( 20 minutes ) with Jurkat target cells that have been previously labeled with dye . We then monitored by flow cytometry the level of intracellular gag p19 in the Jurkat cells . p19 signal monitors viral transfer from infected cells toward target cells ( Figure 11C ) . These results demonstrate that a reduction in Gem expression directly impacts viral transmission , decreasing the number of cells infected by 20% ( p = 0 , 02 , Student's t-test ) . This result is consistent with the 40% reduction in conjugate formation observed in figure 11B . As control , Gem and beta-tubulin levels were monitored by western blot analysis . Of note , a 22% decrease in Gem level expression was observed in cell transfected with Gem siRNA ( Figure 11C lower panel ) . To determine if the decrease in viral transfer impacted productive infection , we transfected C91/PL cells with control or Gem siRNA and a plasmid expressing GFP . Twenty-four hours after transfection , GFP positive cells were harvested by cell sorting of the transfected cultures . These cells were then co-cultured for one hour with a reporter cell line BHK1E6 , containing the lacZ gene driven by the HTLV-1-LTR promoter . Cells transfected with control siRNA were capable of transmitting virus to target cells resulting in expression of β-galactosidase . In contrast , no β-galactosidase positive cells were measured when Gem protein expression was knocked down ( Figure 11D , graph ) . As a control Gem expression was monitored by western-blot ( Figure 11D , lower panel ) . To be sure that the decrease in viral transmission was not due to a decrease in viral production and release following siRNA transfection , Gag p19 was monitored by ELISA ( Zeptometrix Corporation ) from 24 h to 72 h in the cell culture supernatant of HTLV-1 cells treated with Gem siRNA or control siRNA ( Figure S1 ) . We did not observe any significant difference in viral production in cells treated with Gem siRNA . This confirms that the decrease in viral transmission seen above is not due to a decrease in viral production but to the direct role of Gem in viral transfer . Altogether , these results indicate that Gem is not only involved in the formation of T-cell conjugates but also in viral transmission from infected cells to target cells . In contrast to HIV-1 , HTLV-1-infected cells produce few free viral particles [10] . Indeed , efficient transmission of the virus from an infected cell to an uninfected T-lymphocyte relies mostly on cell-to-cell contacts via the formation of a virological synapse and/or of a viral biofilm-like structure [12]–[15] . In both models , Tax plays an important role . Thus , one can assume that blocking the formation of cell conjugates would have a major impact on cell transmission and proviral load evolution . We previously identified a set of genes that are up-regulated in T-cells transduced by HTLV-1 Tax encoding lentivirus [17] . Here , we performed a retrospective analysis , focusing our attention on genes that could potentially be involved in HTLV viral transmission , and identified gem . Interestingly , data mining allowed us to retrieve previous reports which also listed gem mRNA among the genes that are up-regulated gene in Tax-transformed lymphocytes ( Tesi ) , Tax-inducible JPX-9 cells , HTLV-1-transformed ( C8166 and Hut102 ) or -immortalized cells ( Bes , ACH . WT and Champ ) [17] , [25]–[27] . Our results also demonstrate that Gem is strongly expressed at the protein level following Tax expression in T-lymphocytes and in all HTLV-1-infected cell lines tested . The fact that Gem expression is directly linked to Tax expression was confirmed by the reporter gene and Chip assays , demonstrating that together with CREB and CBP , Tax activates transcription from the gem promoter through a CRE sequence . Small GTPases of the Rho family are pivotal regulators of several aspects of cell behavior , such as cell motility , cell proliferation and apoptosis . They play a central role in many motile responses that involve the actin cytoskeleton and/or microtubule network , from neurite extension to phagocytosis and cancer-cell invasion [35] . Our results show that Gem strongly colocalizes with F-actin and are consistent with its direct involvement in actin cytoskeleton dynamic . Polymerization of actin soluble units in filaments is a major mechanism for the cell movement [31] . Interestingly , Kress et al . , showed that fascin , a protein that stabilizes filamentous actin and concentrates in cellular protrusions , such as filopodia , during cell migration , is overexpressed following Tax expression . This protein is overexpressed in HTLV-1-infected cell lines and involved in HTLV-1 cellular invasion [25] . Moreover , Varrin-doyer et al . , also demonstrated that the collapsing response mediator protein 2 ( CRMP2 ) was overexpressed in HTLV-1-infected T-lymphocytes in vitro , showed that CRMP2 was involved in migration of these cell clones and demonstrated that Tax was partly involved both in modulation of CRMP2 level and lymphocyte migratory rate . In addition , CRMP2 remodels T-lymphocyte microtubule cytoskeleton and partially co-localized with Tax at the cell-to-cell contact point in HTLV-1-infected cells [32] . We did not observe such co-localization between Gem , Tax and actin within the cells . However , expression of Gem either directly or through Tax induction was sufficient to increase lymphocyte migratory rates , and Gem suppression in HTLV-1-infected cells severely reduced cellular migration . These results demonstrate a direct role of Gem in HTLV-1-infected lymphocytes migration and it would now be interesting to define whether their shape was also modified . Interestingly , it has been demonstrated that HIV-1 is also able to modify actin dynamics in order to promote HIV infection . In fact , Yoder et al . , demonstrated recently that during HIV infection the actin cytoskeleton in resting T-cells is a post-entry barrier for HIV-1 [36] , [37] . However , the virus has developed a strategy to overcome this restriction . Binding of the gp120 envelope glycoprotein to the CXCR4 chemokine co-receptor induces activation of a cellular protein named cofilin . Cofilin is a critical factor for depolymerizing the cortical actin filaments ( F-actin ) , therefore allowing migration of the viral material into the cell . Gem is a negative regulator of the ROCK-I ( ROKβ ) Rho kinase . When overexpressed , Gem inhibits ROCK I-induced neurite retraction and ROCK-mediated phosphorylation of myosin light chain ( MLC ) and myosin light chain phosphatase ( MLCP ) [30] . Interestingly , CRMP2 is phosphorylated by another Rho kinase: ROCK-II ( ROKα ) . CRMP2 normally promotes axon outgrowth , possibly through its ability to promote microtubule assembly . It has been shown that CRMP-2 phosphorylation is involved in the regulation of the neurons growth cone morphology [38] . Fascin is also a target of Rho-kinases . Rho activity modulates the interaction of Fascin with the p-Lin-11/Isl-1/Mec-3 kinase 1 and 2 ( LIMK1/2 ) , which allow the recruitment of this complex to actin filaments . This interaction modulates actin filopodia dynamics and promotes its stabilization [39] . Thus , CRMP2 and Fascin have been proposed to function down-stream of Rho kinases [35] , [39] , whereas Gem is an upstream negative regulator of ROCK-I Rho kinase [30] . Altogether , these results suggest that Fascin , CRMP-2 and Gem are different actors that play a role in the regulation of HTLV-1-infected cell motility and are therefore involved in viral transmission . However , our results show that Gem could have additional functions during HTLV-1 infection . First , it is involved in the formation of conjugates between infected and uninfected T-lymphocytes . Of note , Tax and p8 were also previously shown to be involved in the formation of cell-to-cell conjugates that involve cytoskeleton reorganization [14] , [31] . In the HIV-1 situation , the virus is also transmitted through the production of free viral particles by the infected cells [40] . In contrast , HTLV-1 transmission occurs mainly through cell-to-cell contacts [11]–[13] , [16] . Thus , we hypothesize that the enhanced cell migration upon HTLV-1 infection increases the probability of encountering and infecting new target cells through the creation of cell-to-cell contacts . Thus our results indicate that HTLV-1-infected cells , which express Gem , are likely to migrate faster and to form more conjugates with target T-cells and thus to transmit more efficiently the virus . It would now be of interest to assess simultaneously the relative contribution of Fascin , CRMP-2 and Gem in HTLV-1-infected cells motility and viral transmission . In conclusion , our results demonstrate that HTLV-1 Tax protein promotes Gem expression in vitro and in vivo . As a consequence , Gem modifies cell morphology and increases migration of HTLV-1-infected cells as well as formation of cell-to-cell conjugates . Finally , we demonstrate that Gem plays an important role in cell-to-cell viral transmission . 293T , HeLa , Jurkat and MOLT4 cells were obtained from ATCC . HTLV-1-infected cell lines and JPX-9 cell line were obtained from Dr Antoine Gessain ( Pasteur Institute , Paris ) . 293T and HeLa cells were cultured in DMEM-GLUTAMAX-I ( Gibco , Invitrogen ) complemented with 10% fetal bovine serum ( FBS ) ( Gibco , Invitrogen ) and antibiotics ( penicillin-streptomycin , PAA ) . MOLT4 , Jurkat , JPX-9 and HTLV-1-infected cell lines ( C8166 , C91/PL , Hut102 and MT2 ) were cultured in RPMI-GLUTAMAX-I ( Gibco , Invitrogen ) , complemented with 10% fetal bovine serum ( FBS ) ( Gibco , Invitrogen ) , and antibiotics ( PAA ) . Cells were maintained at 37°C in 5% CO2 . Tax expression was induced by adding 120 µM of ZnCl2 in the culture media of JPX-9 cells [41] . Production of Lenti-IRES-GFP ( named “Lenti-control” thereafter ) , Lenti-Tax-IRES-GFP ( named “Lenti-Tax” thereafter ) and Lenti-HA-Gem-IRES-GFP ( named “Lenti-Gem” thereafter ) lentiviral particles as well as cell transduction procedures were performed as previously described [17] . Stocks of lentiviral particles were stored at −80°C . Gem cDNA was amplified from pMT2T-Gem vector [42] and cloned either in frame with a HA tag into the pSG5M vector [43] or into the pSDM101 vector [17] . Tax-1 cDNA was amplified from pSG5M-Tax-1 vector [43] and cloned in frame with a Flag tag into the pSG5M vector . Seventy-two hours post-transduction , total RNAs were extracted from 293T cells using the RNeasy mini kit ( Qiagen ) . To avoid DNA carryover , RNA samples were treated with DNase I RNase-free DNAs set ( Qiagen ) . Five hundred nanograms of total RNA were used as a matrix for RT-PCR using the one step RT-PCR kit ( Qiagen ) . PCRs were performed with an annealing temperature of 52°C for Gem primers ( Fwd: [5′-GAACTAGGCTCATCAGAATCGTGAC-3′]; Rev: [5′-GAGCTGTGACATACAAGGGTCAACC-3′] ) or of 59°C for GAPDH primers ( [17] ) . ChIP assay was carried out using 10 µg of antibody to Tax ( Tab 172 , NIH ) , CREB or CBP ( Santa Cruz biotechnologies ) antibodies as described previously [44] , [45] . C8166 cells ( 5×107 ) were cross-linked and sheared by sonication to get 200- to 800-bp DNA fragments . Chromatin extracts were pre-cleared with salmon sperm DNA and magnetic protein G beads ( Sigma ) . Supernatants were diluted 10-fold with ChIP dilution buffer , and appropriate antibodies were added and incubated overnight at 4°C . Immune complexes were then collected by addition of magnetic protein-G beads and washed stepwise . Cross-linking was reversed and DNA was purified by proteinase K ( Sigma ) treatment , phenol extraction , and ethanol precipitation . PCR was performed using primers specific for the HTLV-1 LTR ( nucleotides −160 to −139 [5′-CCACAGGCGGGAGGCGGCAGAA-3′] and nucleotides −102 to −79 [5′-TCATAAGCTCAGACCTCCGGGAAG-3′] ) or Gem promoter ( nucleotides −759 to −738 [5′-GAGATGCTGCTGATTGGATGC-3′] and nucleotides −136 to −115 [5′-CCTCTGCAGCAAACTCGGAGT-3′] ) . Lentivirus-transduced cells , HTLV-1-infected or siRNA-transfected cells were collected and washed with PBS . Proteins were extracted using WCE buffer ( 50 mM Tris-HCl pH 8 , 120 mM NaCl , 5 mM EDTA , 0 . 5% NP40 , 1 mM PMSF , 1 mM DTT , 50 mM NaF , 0 . 2 mM Na3VO4 ) in the presence of protease inhibitors ( Complete-EDTA-free , Roche ) . Protein amounts were quantified using the Bradford reagent assay ( Biorad ) and 40–70 µg were loaded into 4–12% NU-PAGE gels ( Invitrogen ) . Following electrophoresis and protein transfer onto PVDF membranes , membranes were blocked for 1 h in a 5% milk/PBS-Tween 0 , 05% solution and incubated overnight with primary antibody ( 1∶2000 anti-Gem ( [42] ) , 1∶400 anti-p24 ( TP-7 clone , Zeptometrix ) , 1∶4 , 000 anti-Flag M2 ( Sigma ) , 1∶40 , 000 anti-β-actin ( AC74 , Sigma ) , anti-GAPDH ( Novus , NB300-322 ) , 1∶1000 ) and anti-β-tubulin ( Santa Cruz , H-235 ) . The next day , membranes were washed and incubated either with anti-rabbit or with anti-mouse horseradish peroxidase-conjugated secondary antibodies ( GE Healthcare ) . Membranes were then developed using the SuperSignal West Pico ( Pierce/Thermoscientific ) or ECL plus kit ( GE Healthcare ) . HeLa cells were transfected with 200 ng of pSG5M-HA-Gem or pSG5M backbone plasmid using Effectene reagent ( Qiagen ) following manufacturer's instructions . Peripheral Blood Lymphocytes ( PBLs ) were electroporated with 5 µg of pSG5M-HA-Gem . Gem transduced MOLT4 and Gem electroporated PBLs were seeded on 0 . 01% poly-L-lysine ( Sigma ) coated coverslip before fixation in 4% PFA solution ( Sigma ) . Twenty-four ( Hela ) or 48 hours later ( PBLs ) , cells were fixed in 4% PFA solution ( Sigma ) for 20 min , neutralized with NH4Cl for 10 min and permeabilized with 0 . 5% Triton X-100 ( Sigma ) for 5 min . Following PBS washes , cells were incubated with a 5% PBS/milk solution then with anti-HA ( Sigma , H6908 , 1∶150 dilution in 5% PBS/milk ) or anti-Flag M2 ( Sigma ) antibody ( 1∶1000 dilution in 5% PBS/milk ) for 1 h at room temperature . F-actin was stained with Rhodamine Phalloidin toxin ( R415 300U , Life Technologies ) . Samples were then incubated with FITC-conjugated goat anti-IgG rabbit ( Vector Laboratories ) ( 1∶100 ) or with CY3-conjugated goat anti-IgG mouse ( Amersham Biosciences ) at a 1/1000 dilution in 5% PBS/milk for 1 h at room temperature . For the immunofluorescence of Gem performed in transduced Molt4 , samples were incubated with DyLight 649-conjugated sheep anti-IgG rabbit ( STAR36D649 AbD serotec ) . Nucleic acids were stained with DAPI-containing mounting medium ( DAPI Fluormount G , Southern biotech ) and cells were visualized and signals acquired with a spectral Leica SP5 confocal microscope . Images analyses were performed with the Fiji software [46] . We used Plot Profile tool in Image-J software to create a plot of intensity values across features in our images . Forty-eight hours post-transduction with control or Tax lentiviral particles ( MOI = 1 ) , 293T cells were transiently transfected , either with a Gem-luc or a Delta-CRE-Gem-luc ( 500 ng ) plasmid [28] using Polyfect reagent ( Qiagen ) . Transfections were carried out in the presence of a phRG-TK vector ( 10 ng ) in order to normalize the results for transfection efficiency . Reporter activities were assayed 24 h post-transfection using the dual-luciferase reporter assay system ( Promega ) as previously described [47] . Luciferase assays were performed with the GLOMAX microplate luminometer ( Promega ) . One hundred twenty thousand Hela cells were seeded in 12-well plates . Twenty-four hours later , cells were transduced with either control ( Lenti-IRES-GFP ) or HA-Gem lentiviral particles ( MOI = 1 ) . Seventy hours post-transduction , a wound was performed in the cell monolayer when cells reached confluence . Pictures were then taken every 3 hours up to 12 h post-wounding with a Carl Zeiss Axiovert 135 . Distances between the 2 fronts of migration were measured using Image-J software ( NIH , USA ) . MOLT4 or Jurkat cell lines were transduced with control ( Lenti-IRES-GFP ) or HA-Gem lentiviral particles ( MOI = 5 ) . Forty-eight hours later , 5×105 transduced cells were loaded in the upper chamber of 5 µm polycarbonate transwell filters ( Corning ) . Complete RPMI medium containing or not SDF1/CXCL12 chemo-attractant ( 150 ng/ml ) was added in the lower chamber . Twenty-four hours later , quantification of the number of migrating cells was performed using TruCount flow tubes ( BD Biosciences ) and flow cytometer Facscalibur4c+HTS ( BD biosciences ) . HTLV-1-infected cells ( C91/PL , Hut102 ) or non-infected control cells ( Jurkat , MOLT4 ) were transfected with 75 nM of Gem siRNA ( ON-TARGETplus smart Pool Gem siRNA L-008717-00-0020 , Thermo Scientific ) or siRNA control ( ON-TARGETplus Non-targeting Pool , Thermo Scientific ) using Hiperfect reagent ( Qiagen ) . 5×105 transfected cells were loaded in the upper chamber of 5 µm polycarbonate transwell filters ( Corning ) whereas complete RPMI medium was added in the lower chamber . Quantification was performed as described above . Jurkat T-cells were stained ( CellTracker Red CMPTX 577 nm , 2 µM , Invitrogen ) in RPMI-GLUTAMAX-I ( Gibco , Invitrogen ) in the absence of FBS for 30 min at 37°C . Cells were then washed and incubated in RPMI complete media for 45 min . HTLV-1-infected cell lines ( C91/PL , Hut102 and MT-2 ) were transfected with 75 nM of Gem siRNA or control siRNA ( Thermo Scientific ) using Hiperfect reagent ( Qiagen ) . After twenty-four hours , these cells were mixed with pre-stained Jurkat-T cells ( ratio 1∶1 ) on 0 . 01% poly-L-lysine ( Sigma ) treated Lab-Tek ( Nunc ) and incubated for 1 h at 37°C . Cells were then fixed ( 4% Formalin , Sigma ) for 20 min at room temperature and mounted in DAPI Fluormount G ( Southern biotech ) . Signals were acquired with an Axioimager Z1 microscope ( Zeiss ) . At least 15 pictures were taken for each condition and more than 1000 HTLV-1-infected cells were counted per condition . Image analyses were performed with the Image-J software . Data are presented as the mean from 3 independent experiments . Hut-102 cells were transfected either with Gem siRNA ( ON-TARGETplus smart Pool Gem siRNA L-008717-00-0020 , Thermo Scientific ) or with control siRNA ( ON-TARGETplus Nontargeting Pool , Thermo Scientific ) . Forty-eight hours post-transfection , these cells were co-cultured with CellTracker Red ( 2 µM , Invitrogen ) CMPTX-labelled Jurkat target cells for 20 min . Donor and target cells were mixed at a concentration of 105 cells/mL each . Cells were then fixed with 4% paraformaldehyde , permeabilized with triton 0 , 05% , and stained for intracellular p19 expression ( Zeptometrix Corporation ) . Analysis was performed with a FACSCalibur flow cytometer ( BD biosciences ) . Transfer of infectious virus was measured using BHK1E6 cells that contain a lacZ reporter gene driven by the HTLV-1 LTR promoter [48] . C91/PL cells were transfected with 75 nM control or Gem siRNA and 0 . 05 µg of pMax-GFP plasmid using the Human T cell Nucleofector Kit , program O-017 ( Lonza , Basel , Switzerland ) as described by the manufacturer . Twenty-four hours after transfection , GFP positive cells were harvested by cell sorting on a FACSAria III cell sorter ( BD Biosciences , San Jose , CA ) . GFP+ cells were rested for one hour in complete media and then 1×105 cells were co-cultured with a monolayer of 1×105 BHK1E6 cells , in a 6-well poly-lysine coated plate . After one hour of co-culture , GFP+ cells were removed , BHK1E6 cells washed three times with phosphate buffered saline ( PBS ) and cultured in fresh media for 24 hours . The next day , monolayers were washed twice with PBS and assayed using a β-galactosidase Staining Kit according to manufacturer's directions ( Active Motif , Carlsbad , CA ) . After staining , β-galactosidase expressing cells were counted by brightfield microscopy . Analyses were performed in GraphPad Prism software Version 5 . 0b . When considering spontaneous migration of lymphocytes , chemotaxis to SDF-1 or cell-to-cell transfer , 1-tailed Student's t-tests and one-way ANOVA ( with Dunnet post-hoc test ) were used to compare the mean between 2 and 3 groups respectively . For quantification of wound healing assay or conjugate formation , Mann-Whitney test were performed . For quantification of promoter activation , i . e . luciferase assays , repeated one-way ANOVA ( with Tukey post-hoc test ) were used . P values less than . 05 were considered significant . One million HTLV-1-infected cells ( C91/PL ) were transfected with 75 nM of Gem siRNA ( ON-TARGETplus smart Pool Gem siRNA L-008717-00-0020 , Thermo Scientific ) or control siRNA ( ON-TARGETplus Nontargeting Pool , Thermo Scientific ) using Hiperfect reagent ( Qiagen ) . Twenty-four , forty-eight or seventy-two hours post-transfection , supernatants were collected and centrifuged at 2 000 rpm for 3 min . Supernatants were diluted in RPMI complete media ( 1∶100 ) and p19 levels were assessed using the RETROtek HTLV p19 Antigen ELISA kit ( Zeptometrix Corporation ) following manufacturer instructions .	HTLV-1 was the first human oncoretrovirus to be discovered . Five to ten million people are infected , and 1–6% will develop either Adult T-cell Leukemia , or Tropical Spastic Paraparesis/HTLV-1 Associated Myelopathy ( TSP/HAM ) . HTLV-1 infects primarily T-cells , but dendritic cells were also found to carry proviruses . Contrary to HIV-1 , cell-free HTLV-1 viral particles are poorly infectious . Thus , efficient viral transmission relies on formation of virological synapses or formation and transfer of viral biofilm-like structures . The Tax viral transactivator plays a key role in both modes of transmission . Using transcriptomic analyses , we recently identified cellular genes that are deregulated following Tax expression in T-cells . We focused our attention on genes that are important for cell architecture and are thus likely to modulate cell-to-cell contacts and motility . We found that Gem was highly upregulated both at the RNA and protein levels in Tax-expressing cells and HTLV-1-infected cell lines . We further show that Tax binds cellular co-activators and transcription factor and activates transcription from the gem promoter . We demonstrated that Gem is involved in cellular migration of HTLV-1-infected cells . Importantly , gem knockdown decreases the rate of HTLV-1-infected cell migration and cell-to-cell conjugate formation . We also show that Gem plays an important role in HTLV-1 transmission through cell-to-cell contacts , the most efficient mode of viral infection .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "viral", "transmission", "and", "infection", "viruses", "and", "cancer", "virology", "gene", "expression", "biology", "microbiology", "molecular", "cell", "biology" ]	2014	Gem-Induced Cytoskeleton Remodeling Increases Cellular Migration of HTLV-1-Infected Cells, Formation of Infected-to-Target T-Cell Conjugates and Viral Transmission
Replicating recombinant vector vaccines consist of a fully competent viral vector backbone engineered to express an antigen from a foreign transgene . From the perspective of viral replication , the transgene is not only dispensable but may even be detrimental . Thus vaccine revertants that delete or inactivate the transgene may evolve to dominate the vaccine virus population both during the process of manufacture of the vaccine as well as during the course of host infection . A particular concern is that this vaccine evolution could reduce its antigenicity—the immunity elicited to the transgene . We use mathematical and computational models to study vaccine evolution and immunity . These models include evolution arising during the process of manufacture , the dynamics of vaccine and revertant growth , plus innate and adaptive immunity elicited during the course of infection . Although the selective basis of vaccine evolution is easy to comprehend , the immunological consequences are not . One complication is that the opportunity for vaccine evolution is limited by the short period of within-host growth before the viral population is cleared . Even less obvious , revertant growth may only weakly interfere with vaccine growth in the host and thus have a limited effect on immunity to vaccine . Overall , we find that within-host vaccine evolution can sometimes compromise vaccine immunity , but only when the extent of evolution during vaccine manufacture is severe , and this evolution can be easily avoided or mitigated . Live vaccines replicate within the host . As true of any reproducing population , these within-host vaccine populations may evolve . For live vaccines that do not transmit , any within-host evolution is a dead end and might thus seem to be irrelevant to vaccine function . But if the process is fast enough , or the vaccine population replicates long enough , the vaccine population may evolve to a state where it is ineffective or virulent—either change would be bad . The two main types of live viral vaccines are attenuated and recombinant-vectored . Most live virus vaccines in use today are attenuated , their reduced virulence typically achieved by adapting the wild-type virus to a new environment ( e . g . replication in a novel cell line or low temperature ) , with a consequent reduced replication rate in humans . The use of attenuated vaccines is too risky for pathogens such as HIV , and a safer alternative is to develop a live , recombinant vector vaccine where one or a few pathogen genes with immunogenic activity ( proteins that elicit protective immunity ) are expressed from a benign virus vector . The expected consequences of within-host evolution differ between these two types of vaccines ( Table 1 ) . Evolution of an attenuated vaccine is likely to be a reversion toward the wild-type state , the rate of this process depending heavily on vaccine design and the duration of vaccine virus replication in the host ( reviewed in [1] ) . To a first approximation , reversion toward the wild-type state should lead to the vaccination more closely resembling natural infection [2] , such as higher virus densities , side-effects and disease , and possibly an increased immune response . Within-host evolution of an attenuated vaccine might also predispose the virus to better transmission—also reflecting the wild-type state—but this outcome is not assured: viral adaptation to different tissues within the host may hamper growth in and dissemination from tissues important in transmission ( e . g . , [3] ) . The expected consequences for evolution of a recombinant-vectored vaccine are fundamentally different [4] . In most cases , the antigen against which immunity is sought comes from a foreign transgene inserted into a competent viral vector without replacing any vector genes . Vectors in development include adenovirus , VSV ( vesicular stomatitis virus ) and CMV ( cytomegalovirus ) . The vector genome carries out all viral amplification and transmission functions , and the transgene does not contribute to any process benefiting vector reproduction . From an evolutionary perspective , the transgene is both dispensable and potentially costly: selection may favor loss of the transgene and thus loss of vaccine’s ability to elicit immunity against the antigen encoded by the transgene . This evolution therefore generates something akin to infection by the wild-type vector . As vectors are typically chosen to be avirulent for immune competent hosts , vaccine evolution will result in no more than a harmless infection that does not generate immunity to the antigen encoded by the transgene . Considerable attention has recently been given to the evolution of attenuated vaccines and designs that retard their evolution . Evolutionary stability of attenuated vaccines seems attainable by engineering designs , including the introduction of hundreds of silent codon changes , genome rearrangements , and some types of deletions ( comparisons and reviews are provided by [1 , 5 , 6] ) . Far less thought has gone into the consequences of evolution for recombinant vector vaccines or of strategies to minimize this evolution . Although recombinant vector vaccines are not yet in widespread use , many are under development [7 , 8] , and their success may rest on understanding within-host evolution . Here we explore how the combination of evolution during the process of vaccine manufacture and during its within-host dynamics following vaccination could affect the immune responses elicited by a recombinant vector vaccine and reduce its efficacy—the specific interaction between evolution and immunity . We consider viral vaccines and focus on vaccines that cause short-duration ( acute ) infections . The ideas we discuss also apply to live vaccines of bacteria and other pathogens . Our overall message is that while vaccine evolution may occur , it is either unlikely to be a problem ( i . e . , compromise the generation of immunity ) , or it is easily mitigated . When vaccine evolution does limit the adaptive immune response , we identify ways of escaping such outcomes . Our analysis rests on mathematical models , but most results can be explained intuitively ( perhaps only in hindsight ) , with the main results illustrated graphically; many analyses are relegated to Supporting Information . Our analysis assumes that vaccines replicate within the host untill cleared by host immunity; we exclude vaccines that reproduce for just a single infection cycle ( e . g . , Modified Vaccinia Virus Ankara ) , as they have no significant opportunity for evolution . Our models are numerical analyses of ordinary differential equations . The equations are given in supporting information ( S1 Appendix ) . The were numerically evaluated and graphed in R ( S1 File , a Markdown file ) , sometimes also evaluated in Mathematica ( S2 File ) . The key question is whether evolution of the vaccine virus ( henceforth just ‘vaccine’ ) meaningfully affects immunity to the antigen encoded by the foreign transgene ( henceforth just ‘antigen’ ) . The potential for vaccine evolution is easy to understand . Through mutation , any large vaccine population will contain mutants that inactivate or delete the foreign transgene , and those revertants will then grow amidst the vaccine . Vaccine inferiority may accrue in two different ways: the transgenic insert and its expression may intrinsically impair vaccine growth , and adaptive immunity to the foreign antigen may impair the vaccine’s growth but not the revertant’s during an infection . It is easy to appreciate how and why the vaccine may be inferior to the revertant , and this can result in an increase in frequency of the revertant . However , the relationship between this evolution and the extent of immunity to the vaccine antigen is more complex . We thus explain some of the factors that affect how this evolution translates into a reduction in immunity to the antigen , and why in some circumstances , substantial evolution can result in little change in immunity to the antigen , while in different situations it can result in a substantial reduction . We now employ quantitative models to evaluate the intuitive ideas presented above . Given the high dimensionality of the problem , we are especially interested in how well intuition works and whether generalities are observed across large regions of parameter space . A flow diagram of the elements and interactions within the host reveals the complexity of the model ( Fig 3 ) and facilitates understanding the dynamical equations . V and W are the respective vaccine and revertant densities , with intrinsic growth and death rates governed by four parameters ( not illustrated ) . The model also includes variables for resources ( R ) , innate immunity ( Z ) , adaptive immunity to vector ( Y ) , and adaptive immunity to antigen ( X ) that are both influenced by and influence V and W . In the following sections , we explore the dynamics of these interactions with simulations and present results graphically ( the results presented do not allow resource limitation to influence dynamics; trials where resource limitation matters were conducted but are not shown ) . Equations and parameter values are provided in S1 Appendix . Resource limitation and innate immunity yield qualitatively similar results , so trials with resource limitation are not illustrated in the main text . The equations apply only to within-host processes; any pre-host evolution is subsumed into inoculum composition . The models assist us by forcing us to specify assumptions for how the viruses and immunity interact , and by allowing us to rigorously explore outcomes in different scenarios . However , there is uncertainty in the model structure , many parameter values are unknown , and different viruses will behave somewhat differently . Consequently , we focus on broad generalities that arise from many simulations and illustrate these for a few specific cases , reserving Supporting Information files for further details . The presentation below briefly discusses the dynamics of individual trials for illustration but then moves to contour plots that reveal differences in outcomes as the key parameters are changed . The model used here incorporates the structure of earlier models that described immune responses [36–38]; parameter values used here were chosen as described in some of these earlier studies . The results above suggest that vaccine evolution is only likely to compromise immunity if there is substantial pre-host or within-host evolution and if this evolution depresses vaccine virus in the host . As the short duration of infection limits within-host evolution , one means of achieving vaccine efficacy is to control the inoculum . Two ways of controlling the inoculum are to control its composition and to control its size . Pre-host evolution can be reversed by purifying the inoculum after the fact or by taking care to start with a pure isolate and limiting growth ( e . g . , Fig 1 ) . The benefit of suppressing revertant frequency in the inoculum is evident in Fig 7: the magnitude of immunity to the vaccine increases by orders of magnitude as the initial frequency of the revertant is decreased . The effect is strongest at low inoculum levels , pointing to the other solution—increase inoculum size . Intuition also suggests that the deleterious effects of evolution can be reduced by increasing the inoculum size , provided the composition does not change: to achieve a threshold antigen level , a large inoculum requires less growth than a small one . Less growth reduces the potential for evolution—in the extreme , a large enough inoculum requires no vaccine growth , as with killed vaccines . These conjectures are supported by Fig 7: when the revertant frequency in the inoculum is high , increasing the inoculum size appreciably increases the magnitude of immunity; a much reduced benefit is seen when revertant frequency is low , likely because there is less evolutionary interference from the revertant . These results suggest parallel benefits from reducing the frequency of the revertant in the inoculum and increasing the dose . Consideration of the gains from each could help choose an economically feasible strategy , since both purifying the inoculum and increasing its dose are likely to incur financial costs . Whether and how well controlling the inoculum will work in practice will depend on details . Solutions may be quantitative rather than absolute . Intuition is useful for guidance but needs to be confirmed by formal analyses , guided by data from the specific implementation . Any live viral vaccine may evolve within the host . The potential for attenuated viruses to revert to wild-type virulence is well appreciated [1 , 2] , even if it presents a problem for relatively few vaccines ( e . g . , attenuated polio , [41] ) . There is also a potential for live , recombinant vector vaccines to evolve—our focus in this paper—with the main concern being loss or reduced expression of the transgenic insert [4 , 42] . If such a vaccine were to evolve fast enough or long enough that it lost the insert , vaccine efficacy might well suffer . We find that evolution during manufacture ( pre-host evolution ) can play a more important role than within-host evolution in reducing vaccine efficacy , and furthermore that it may be the more easily mitigated . We developed and analyzed models to explore ways in which vaccine evolution could lead to a reduction in vaccine efficacy . An intrinsic fitness advantage of the revertant virus , expected because transgene expression is likely to have metabolic and other costs , will lead to vaccine being gradually overgrown by revertant . This is only likely to cause a reduction in the immunity to the vaccine antigen if it leads to a reduction in the absolute amount ( as opposed to merely a reduction in relative frequency ) of the vaccine virus . There are in fact several mechanisms by which an ascending revertant population may suppress vaccine: revertant can reduce the amount of the vaccine virus in the host if the revertant uses resources required for virus replication or if the vaccine virus is cleared by the innate or adaptive responses elicited by the revertant . The clear and positive message from our study is that vaccine evolution , if it proves to be a problem for immunization , should be easily mitigated by manipulating the vaccine inoculum . Critical to understanding and addressing this problem is recognizing that the vaccine may evolve both within the host and also during manufacture , whereby the inoculum already carries modest to high levels of revertant . The composition of the inoculum can have a large effect on within-host evolution and immunity . By limiting the amount of revertant in the inoculum , and also by boosting the inoculum level , it should usually be possible to limit the amount of within-host vaccine evolution and ensure that immunization is effective . We emphasize , however , that this solution will typically not work for transmissible vaccines and vaccines that establish long term infections within the host . Furthermore , using a large inoculum may seem to defeat the purpose of using a live vaccine . There may be cases in which vaccine evolution is so rapid that controlling the inoculum is not sufficient . The solution in this case is to develop or engineer the vaccine with less of a disadvantage . The timing and tissues of antigen expression , location of the transgene in the vector genome , and the size of the transgene may all influence intrinsic fitness effects [10 , 11 , 19 , 43 , 44] . Directed evolution approaches might also improve vaccine efficacy: one simple approach in reducing an intrinsic cost might be to adapt the vector in vitro to host cells expressing the antigen in trans , allowing compensatory mutations to evolve in response to the antigen before the transgene is cloned into the genome . This adapted vector would then be used as the vaccine backbone . Another simple approach would be to compete several different vaccine designs in vitro and pick the design with highest retention of the transgene . Any approach using in vitro adaptation needs to avoid adapting the vector to the extent that it compromises ability to grow in vivo . Most of these possibilities are ways to reduce pre-host evolution and reduce revertant concentration in the inoculum . One may hope that vaccine designs which reduce pre-host evolution also reduce within-host evolution . Measuring the intrinsic fitness effect of the transgene is likely to be an important step in vaccine design . For assessing vaccine evolution , the relevant biological realms are within the host and in vitro . In vitro growth environments are the more easily studied and may reveal much about a vaccine’s intrinsic propensity to evolve loss of antigen expression . There are various ways intrinsic fitness effects and their evolutionary consequences might be studied . Vaccine growth in tissue culture may reveal some aspects of intrinsic fitness effects and should be relatively easy to study . Deletion of the transgene per se would be detectable by PCR , and the fitness advantage of revertant over vaccine could be measured from changes in revertant frequency . The quantitative relevance of an in vitro estimate to in vivo growth would be unknown , but the measure should allow qualitatively comparing engineering designs that improve intrinsic vaccine fitness . If vaccine reversion were due to down regulation of the transgene instead of deletion , fitness estimation would require knowing the mutations responsible and monitoring their frequencies . Use of culture-wide antigen levels to measure fitness might provide a sense of whether vaccine evolution would lead to reduced antigen levels in vivo , but it would be less sensitive in measuring evolution than is measuring mutation frequencies . Evolution is not the only consideration in designing a recombinant vector vaccine , and the model helps us identify vaccine properties that promote efficacy . First the vaccine should elicit an immune response that rapidly clears the pathogen ( i . e . the rate constant for clearance of the pathogen , call it kP , is high ) . Second , the vaccine should elicit a large response to this antigen . This requires that the antigen rapidly elicits immunity ( i . e . has low ϕX , and in terms of immunology it should be an immunogenic antigen ) , and also requires a high vaccine viral load to generate a large response . Engineering this requires tackling a trade-off between avoiding vaccine clearance ( i . e . having a low kX ) but allowing for rapid clearance of the pathogen ( having a high kP ) . Vaccines designed to express the antigen in a form that is different from that in the pathogen might help solve this problem . Thus , to elicit immunity to influenza , one might design secreted forms of the hemagglutinin or neuraminidase proteins . A recombinant hemagglutinin protein that is secreted rather than on the virion surface would prevent the antibody response to this protein from clearing the recombinant vector vaccine ( have low kX ) without compromising the clearance of the influenza virus pathogen which has hemagglutinin on its surface ( i . e . has high kP ) . In this manner our model allows the identification and tuning of parameters that affect vaccine efficacy , and a comprehensive search of parameter space would identify ideal combinations of vaccine properties . In vitro assays may be useful in measuring intrinsic fitness effects , but in vivo—in the patient—is the ultimate environment for studying within-host evolution and its effects . Not only are the dynamics of viral spread different between in vitro and in vivo environments , but most immune components will be in play only in vivo . Furthermore , those components may vary across tissues within the host . Sampling across this heterogeneity in vivo will be challenging but may be necessary to know whether , when , and where vaccine evolution is a problem . If revertant remains a minority of the population , we expect that vaccine evolution can be ignored . Perhaps in vitro studies of vaccine evolution will provide most of the information relevant to in vivo evolution , but it is too early to know . We have focused on recombinant vector vaccines that cause acute infections . Necessarily , our recommendations are based on simple models that are caricatures of the complex within-host dynamics of acute infections . Simple models are appropriate at this stage because of uncertainties at many biological levels , and under these circumstances simple models frequently generate more robust results than do complex models [45 , 46] . The generation of innate and adaptive responses can be modeled with different assumptions than used here , and those alternative processes may affect the conclusions . For example , time-lags in the activation of cells may dominate the time for the generation of an innate immune response , with virus density having a consequently smaller role than assumed here ( as can be seen in [47] and modeled in [30] ) . We have modeled that responses to different antigens are generated independently of each other and do not compete . We have done so because vaccines are likely to cause relatively mild infections during which the densities of pathogen and immune cells do not reach sufficiently high levels required for competitive interactions to be important . The adaptive immune response may be more influenced by recruitment which is followed by a period of proliferation even in the absence of antigen [48–50] . Both these scenarios would minimize the impact of evolutionary changes in the vaccine on the amount of immunity generated to the transgene . Finally , it is easily appreciated that there are realms we do not consider , such as within-host spatial structure [51] and recombinant vector vaccines based on viruses such as cytomegalovirus that cause persistent infections [52] or that are transmissible . Spatial structure may limit the impact of vaccine evolution on immunity ( e . g . , prevent mutants from taking over the entire population ) . In contrast , vaccines that cause persistent infections or are transmissible are likely to be more severely affected by evolution than are vaccines causing acute infections , as there is a longer timeframe for evolution to operate . With so little experience from recombinant vector vaccines , we can merely guess how commonly the neglect of within-host evolution will compromise vaccine efficacy . Given that simple steps can be taken to reduce vaccine evolution , vaccine development programs should at least entertain the possibility that evolution can underlie failure . Avoiding vaccine evolution may be easier than developing an entirely new vaccine .	Recombinant vector vaccines are live replicating viruses that are engineered to carry extra genes derived from a pathogen—and these extra genes produce proteins against which we want to generate immunity . These vaccine genomes may evolve to lose the extra genes during the process of manufacture of the vaccine or during replication within an individual , and there is a concern that this evolution might severely limit the vaccine’s efficacy . The dynamics of this process are studied here with mathematical models . The potential for vaccine evolution within the host is somewhat limited by the short-term growth of the vaccine population before it is suppressed by the immune response . We find that evolution is a problem only when the process of manufacture results in the majority of the vaccine virus being revertant . We show that increasing the vaccine inoculum size or reducing the level of revertant in the vaccine inoculum can largely avoid the loss of immunity arising from evolution .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "organismal", "evolution", "viral", "vaccines", "medicine", "and", "health", "sciences", "immunology", "microbiology", "vaccines", "microbial", "evolution", "infectious", "disease", "control", "infectious", "diseases", "evolutionary", "immunology", "viral", "evolution", "...	2019	Recombinant vector vaccine evolution
Plant protein kinases form redundant signaling pathways to perceive microbial pathogens and activate immunity . Bacterial pathogens repress cellular immune responses by secreting effectors , some of which bind and inhibit multiple host kinases . To understand how broadly bacterial effectors may bind protein kinases and the function of these kinase interactors , we first tested kinase–effector ( K-E ) interactions using the Pseudomonas syringae pv . tomato–tomato pathosystem . We tested interactions between five individual effectors ( HopAI1 , AvrPto , HopA1 , HopM1 , and HopAF1 ) and 279 tomato kinases in tomato cells . Over half of the tested kinases interacted with at least one effector , and 48% of these kinases interacted with more than three effectors , suggesting a role in the defense . Next , we characterized the role of select multi-effector–interacting kinases and revealed their roles in basal resistance , effector-triggered immunity ( ETI ) , or programmed cell death ( PCD ) . The immune function of several of these kinases was only detectable in the presence of effectors , suggesting that these kinases are critical when particular cell functions are perturbed or that their role is typically masked . To visualize the kinase networks underlying the cellular responses , we derived signal-specific networks . A comparison of the networks revealed a limited overlap between ETI and basal immunity networks . In addition , the basal immune network complexity increased when exposed to some of the effectors . The networks were used to successfully predict the role of a new set of kinases in basal immunity . Our work indicates the complexity of the larger kinase-based defense network and demonstrates how virulence- and avirulence-associated bacterial effectors alter sectors of the defense network . Plant immunity is generated by the activation and coordination of several protein kinase-based signal transduction pathways into cellular defense responses [1 , 2] . Kinases modify the activity status of other proteins through specific biochemical modifications ( substrate phosphorylation ) or by recruiting proteins in signaling complexes . Signaling pathways transmit pathogen signals from the cell periphery to intracellular compartments and trigger changes in gene expression , hormone-based signaling , and defense compound production [3] . To survive in plant tissues and ensure spread to other plants , pathogens must overcome plant defenses and redirect their energetic and nutrient resources . The constant tug-of-war between plants and pathogens has generated a complex immune system in plants , and equally multifaceted assault and endurance mechanisms in pathogens . Plant pathogens such as the gram-negative flagellated bacterium Pseudomonas syringae can colonize a broad range of plants , an ability at least partly determined by an extensive and versatile effector repertoire [4 , 5] . P . syringae subverts the basal immunity in part by attacking components of signaling pathways activated by pathogen-associated molecular patterns ( PAMPs ) or secreted effectors . PAMP-triggered immunity ( PTI ) is induced by PAMP perception by pattern recognition receptors ( PRRs ) , some of which are receptor-like kinases ( RLKs ) . Upon PAMP recognition , PRRs activate membrane-associated receptor-like cytosolic kinases ( RLCKs ) , cytosolic mitogen-activated protein ( MAP ) kinase ( MAPK ) cascades , and other cytosolic kinases , including Ca2+-dependent kinases [6] . Effector-triggered immunity ( ETI ) , the second layer of immunity , is activated by direct or indirect recognition of effectors , followed by activation of signaling pathways and induction of defense responses and programmed cell death ( PCD ) . However , most intracellular effectors are not recognized by the plant and instead are thought to impair the plant’s ability to sustain an efficient immune response , a condition described as effector-triggered susceptibility ( ETS ) [7] . Work with Arabidopsis , tomato ( Solanum lycopersicum ) , and Nicotiana benthamiana has identified specific P . syringae effectors that inactivate plant kinases [8 , 9] . For example , the AvrPto effector binds membrane-associated kinases , including the PRRs FALGELLIN-SENSING2 ( FLS2 ) , EF-TU RECEPTOR ( EFR ) , and the BRI1-ASSOCIATED RECEPTOR KINASE1 ( BAK1 ) co-receptor to disrupt PTI and promote bacterial virulence [10] . The AvrPto effector also induces host resistance in some tomato genotypes by interacting with the Pto kinase , which activates Pseudomonas resistance and fenthion sensitivity ( Prf ) resistance protein , resulting in ETI [11] . Another effector called HopAI1 represses PTI through its interactions with the cytosolic MAPKs , MPK3 and MPK6 [12] . Recent work suggests that pathogen effectors may interact with not only a few targets but with multiple host targets , indicating that the breadth of effector–plant interactions are only beginning to be understood . Proteome-scale interactomics [13 , 14] revealed an impressive number of putative effector-interacting proteins alongside fundamental properties of plant–pathogen interaction networks , such as effector convergence on network hubs . Furthermore , using a different methodology ( i . e . , global transcriptional profiling of Arabidopsis defense-related mutants coupled with modeling ) [15] identified regulatory relationships between immune-related subnetworks and highly interconnected network components . However , in these studies , the physical layout of the underlying plant cellular networks targeted by pathogens remained out of the reach of the analytic and experimental methodologies utilized . To better understand how protein kinases contribute to basal immunity , we first sought kinase targets that interact with multiple effectors , a characteristic of defense-associated host proteins [13 , 14] . Five P . syringae effectors ( AvrPto , HopA1 , HopAI1 , HopAF1 , and HopM1 ) were selected based on two main criteria: high prevalence among the P . syringae isolates [5] and a known ability or potential to suppress defense responses [10 , 16–19] . HopA1 disrupts the formation of a protein complex involved in activating basal immunity and ETI [20] . HopAF1 interacts with the methylthioadenosine nucleosidase proteins MTN1 and MTN2 to disrupt ethylene ( ET ) production [18] . HopM1 binds HopM1 interactor 7 ( MIN7 ) , disrupting vesicle trafficking and reducing callose deposition [19] . These five effectors suppress different parts of the cellular immune response in plants , suggesting that they may interact with distinct host proteins . Here , we identify targets of bacterial effectors in plant cells , perform an in-depth functional analysis of a set of multi-effector–interacting kinases to superimpose effector-specific pathways over the plant–effector interaction space , and characterize the properties of the plant defense network . We developed a multipronged approach consisting of identification of in vivo pairwise interactions between 279 tomato kinases and five effectors from the model tomato pathogen , Pseudomonas syringae pv . tomato ( Pst ) ( HopA1 , HopAI1 , HopAF1 , AvrPto , and HopM1 ) . Next , we characterized the role of 35 multi-effector–interacting kinases in PTI , ETS , ETI , and PCD . We created new methodologies for data integration and generated signaling networks to facilitate visualization of the protein kinase networks involved in defense ( Fig 1 ) . This network-centric approach allowed us to compare signaling networks associated with different levels of plant immunity and led to identification of novel defense-associated kinases . The approach and the results obtained are described in the sections that follow and in the Supporting information ( S1 Materials and Methods ) . To better understand how diverse effectors may be targeting plant protein kinases , we tested interactions between 279 tomato kinases [21] and five effectors . A total of 1 , 170 pairwise kinase–effector ( K-E ) interactions were tested in tomato protoplasts using the split luciferase complementation assay ( SLCA ) , where luciferase activity indicates reconstitution of the N-terminal and C-terminal domains of the enzyme fused to interacting bait or prey proteins . The split luciferase complementation ( SLC ) data analysis is described in the Supporting information ( S1 Materials and Methods ) . Interactomics primary data are available at https://figshare . com/s/35c4aab65174c67a496e ) ; the MATLAB code for the SLC data analysis is provided in S3 Data . Several controls were included in each SLC experiment to ensure reproducibility of the method across replicates , including a positive control ( protoplasts with full-length luciferase ) , a negative control ( untransformed protoplasts ) , and a reference interaction set between the AvrPto effector and the Pto kinase , which have been shown to interact in planta [22] . In addition , an AvrPtoI96A ( AvrPto with an Ile to Ala mutation ) and Pto kinase pair were included as a control for interaction strength because the I96A mutation inhibits effector function and interaction with the Pto kinase [10] . The SLCA screen was reproducible with low variability of luminescence signals across technical replicates ( S1A Fig ) . High correlation was observed for the signals for full-length luciferase and controls ( reference set of positive and negative interactions ) among plates ( S1B Fig ) ; the K-E signals and the control sets did not show correlation , indicating a lack of a measuring bias in the protocol ( S1C Fig ) . The signals from the AvrPto–Pto and AvrPtoI96A–Pto interactions were highly correlated with an average 3-fold reduction in signal for the AvrPtoI96A–Pto interaction . A multiple regression model of Pto–AvrPtoI96A versus Pto–AvrPto and Luciferase signals has R2 = 0 . 887 ( adjusted R2 = 0 . 886 ) and regression coefficients of 7 . 17 × 10−3 ( Luciferase ) and 3 . 06 × 10−1 ( Pto–AvrPto ) ( S1D Fig ) . Moreover , the luminescence signal of the K-E pairs showed a wide dynamic range , indicating that there are no physical limitations in measuring the luminescence produced in the SLCA ( S1E Fig ) . Out of the 279 kinases tested for interactions with AvrPto , HopA1 , HopAI1 , or HopAF1 , 133 ( 48% ) interacted with at least one effector and were named Kinase Effector Interactors ( KEIs ) . No significant interactions were identified for HopM1 out of the 30 tested kinases , suggesting that this endomembrane-specific effector [19] may not associate with kinases . The K-E interaction network contains 321 significant interactions of 133 kinases with four effectors and includes previously confirmed K-E interactions ( Fig 2A; S1 Table ) . Among the 133 kinases , approximately 70% are multi-effector interactors , out of which 38 interact with all four effectors and 24 are shared by HopA1 , HopAI1 , and HopAF1 ( Fig 2B ) . To estimate the relative affinity of K-E interactions , a metric called the “interaction strength coefficient” ( normalized signal fold change ) was used to quantify the difference in reconstituted luciferase activity between each tested pair and the reference interactions . On average , the interactions of KEIs with HopA1 or HopAI1 were twice as strong when compared with AvrPto , possibly due to the better reconstitution of the luciferase , higher affinity , or low dissociation of K-E complexes ( Fig 2C ) . Notably , two known interaction pairs ( HopAI1–MPK6 and HopAI1–MPK4 ) were the strongest among all control interactions tested for these MAPKs ( inset of Fig 2A ) . The distribution of the fold change interaction values off all K-E interactions tested is shown in S2A Fig . An analysis of the candidate KEIs along the spectrum of kinase classes [23] revealed that the effectors interacted mostly with leucine-rich repeats ( LRR ) -type RLKs and RLCKs from Class 1 ( 42% ) , kinases from Class 2/Raf-like ( 59% ) , and Class 4/MAPKs and calcium-responsive kinases ( 47% ) ( Fig 2D; S2B Fig ) . A group of KEIs was selected for functional characterization based on their ability to putatively interact with multiple effectors . The 35 focus KEIs ( S2 Table ) were silenced in N . benthamiana , a relative to tomato and host for Pseudomonas syringae DC3000 strains lacking the HopQ1-1 avirulence gene , due to its amenability for transformation and high efficiency of gene silencing [24 , 25] . Virus-induced gene silencing ( VIGS ) constructs containing a fragment of an Escherichia coli gene ( EC1 ) served as a negative control . After confirmation that KEI expression was silenced , the plants were inoculated with an effectorless Pst strain ( D29E ) [26] and four single-effector strains expressing AvrPto , HopA1 , HopAF1 , or HopAI1 in the D29E background ( S3 Table; S3A Fig; S1 Data ) . In the EC1 control , the presence of some effectors ( AvrPto , HopAI1 , or HopAF1 , but not HopA1 ) in D29E led to a moderate but significant increase in Pst growth compared to D29E ( Fig 3A ) , indicating that these effectors can contribute to Pst virulence in isolation from the broader repertoire . Among the 35 KEIs , seven KEIs influenced D29E growth compared with the EC1 control , indicating a role in basal immunity ( Fig 3B ) . The majority of these KEIs—including RLKs ( KEI188/LYK4 , KEI72/SOBIR1 , KEI156 , and KEI161/RKL1 ) , RLCKs ( KEI149/PTI1-like ) , and the Ca2+-regulated KEI255/CIPK25—promoted bacterial growth when silenced , while silencing of one kinase ( KEI339 ) inhibited D29E growth . In comparison , silencing of 17 KEIs caused a significant change in the growth of single-effector strains compared with the EC1 control ( Fig 3C–3F; S3B and S3C Fig ) . Silencing of SOBIR1 , a key component of PTI [27 , 28] , affected the growth of D29E and the HopA1- and AvrPto-carrying strains . Moreover , silencing of KEI342/SlBAK1 ( one of the tomato BAK1 homologs that may facilitate basal immunity [29] ) or of KEI327/SlMPK1 ( a kinase with high similarity to AtMPK6 and a possible role in PTI [30] ) interfered with the growth of single-effector strains exclusively . Interestingly , the majority of KEIs required for D29E response were RLKs . On the other hand , cytosolic kinases were preponderant in plant response to D29E + HopAI1 , + HopAF1 or + AvrPto , showing a 4- , 2 . 8- , and 2 . 3-fold increase , respectively , relative to the RLKs ( Fig 3G ) . Among the KEIs with significant contributions to bacterial growth , 52% participated in plant response to D29E + HopA1 and + HopAF1 and 36% to D29E + AvrPto and HopAI1 , while only 12% were necessary for defense against D29E ( S3D Fig ) . KEIs had significant positive or negative effects on the growth of Pst strains , indicating that KEIs promote either immunity or ETS , but not both ( S3E Fig ) . Overall , when mapping the sign of variation ( Fig 3H ) , most KEIs classified as RLK/RLCKs promoted basal immunity , while KEIs promoting ETS mainly included kinases from the other cytosolic and MAPK-like ( 60% and 30% , respectively ) . Some protein kinases have been shown mediate cellular response to multiple types of stresses [31] . To determine if KEIs are similarly involved in multiple response pathways , we tested the focus KEIs in ETI and MAPK-mediated PCD responses . The HopQ1-1 effector is an avirulence factor in N . benthamiana , in which it is recognized by an unknown R protein [32] . To test the ETI in KEI-silenced plants , we quantified the size of the necrotic lesion [32] triggered by the inoculation with a D29E + HopQ1-1 strain as a proxy for quantification of PCD ( S4A Fig; S1 Data ) . Eleven out of the 35 tomato KEIs tested were required for PCD , including RLKs ( KEI37/LYC10 , KEI161/RKL1 , and AtFLS2 ) , RLCKs ( KEI7/PBL8 ) , MAPKK kinases MAP3Ks ( KEI20/SlCTR1 ) , SnRKs ( KEI250/CIPK6 ) , GSK3/Shaggy-like ( KEI272/SK13 ) , and KEI339 ( Fig 4A; S4B Fig ) . Silencing of KEI72/SOBIR1 , a known positive cell death modulator [33] , KEI7/PBL8 , and KEI20/SlCTR1 impaired PCD the most . KEI160/NtIRK , previously associated with antiviral defense and regulation of the R-gene–mediated PCD in N . benthamiana [34] , facilitated PCD . The results indicate that many KEIs mediate ETI-associated PCD by exerting exclusively positive regulatory roles . To test the role of KEIs in the development of MAPK-dependent PCD , KEI-silenced plants were infiltrated with constitutively active MKK7 or MKK9 . Both MAPK kinases ( MAP2Ks ) are known to participate in multiple immune-related processes [35 , 36] , and their prolonged expression induces activation of MPK3 and MPK6 and PCD [1] . Lesion size was significantly altered in 23 of the tested KEI-silenced lines following MAP2K expression . MKK7-triggered , MKK9-triggered PCD was modified in several lines , seven of which were required for both MKK7- and MKK9-mediated pathways ( Fig 4B and 4C ) . The RLKs and RLCKs functioned as both positive and negative PCD regulators , compared with other classes ( Fig 4D ) . Notably , silencing of the PCD negative regulator KEI342/BAK1 [37] inhibited both MKK7- and MKK9-PCD . To determine how these phenotypes may be related , we performed correlation analyses between bacterial growth and lesion size measurements across the KEI lines ( Fig 4E ) , as described in S1 Materials and Methods . A significant correlation ( R > 0 . 6 ) was observed across bacterial growth assays , but no correlation was found between bacterial growth and PCD treatments , suggesting these responses utilize distinct signaling pathways . To obtain a global view of the link between the KEIs’ structural class and their contribution to immune phenotypes , we plotted a phylogenetic tree and visualized significant contributions to immunity based on our assays ( Fig 4F and S2 Data ) . Interestingly , the phylogenetic tree highlighted the clear difference in structural class between the PTI- versus ETS-promoting kinases ( RLKs/RLCKs versus MAPKs , CIPK/SnRKs , ribosomal protein S6 kinase ( S6K ) , AGC kinases , and glycogen synthase kinase3 [GSK3]-like , respectively ) . The involvement of KEIs in multiple stress responses prompted the development of functional signaling networks to understand how defense networks are modified during different immune responses . To construct the networks , we calculated the co-occurrence frequency of the focus KEIs in various functional assays to evaluate the degree of KEIs phenotype overlap , indicative of functional association among KEIs . Using a set of logical rules and prior information ( Fig 5A ) , a KEI signaling network was generated with the nodes ( KEIs ) ordered hierarchically within the canonical structure of a signaling pathway: RLK → RLCK → RAFs/MAPKs → cytosolic kinases ( Materials and methods; S1 Materials and Methods ) . Indirect evidence positioned RAFs upstream of MAPK cascades and at a similar hierarchical level with MAP2Ks [2 , 31 , 38 , 39] . Directed edges weighted by co-occurrence values link the nodes . To generate the signaling network , KEIs with similar phenotypes were grouped in modules , in which the position of nodes within the same hierarchical level or kinase structural class was based on their regulatory strength rank and sign of regulation , while the redundant edges between successive network levels were removed . Using these criteria , we collapsed the weighted composite graphs into a minimal network providing an overview of the KEI pathways critical to immune-related plant phenotypes ( Fig 5B ) . Next , we generated networks representing the response of the minimal network under our eight experimental conditions , called stimuli-specific networks ( SSNs ) , to reveal how the signaling network responded ( Fig 5C ) . A comparison of the infection-response networks demonstrated that most of the D29E network is maintained across single-effector networks , with the exception of avirulence-inducing HopQ1-1 . Addition of these single effectors affected the network topology , such that a larger number of RLK and RLCKs played a role . The networks doubled in diameter ( the average length of shortest paths between all pairs of nodes ) and had longer distance ( shortest path index ) between any two nodes , relative to the D29E network , indicative of activation of a more diverse and complex defense network ( Fig 6A , S5A Fig ) . For example , KEI104-BSK7 promoted immunity against strains containing HopA1 , HopAF1 , and AvrPto , but not the D29E strain . Interestingly , addition of virulence-promoting effectors ( HopAI1 , HopAF1 , and AvrPto ) activated ETS pathways , including cytosolic kinases KEI339 , KEI323/S6K2 , and KEI318/SRK2C . In the PCD networks , ( HopQ1-1 , MKK7 , and MKK9 ) , the signaling pathways were markedly distinct . While the MKK7 network is primarily comprised of KEIs that repress cell death , the MKK9 and HopQ1-1 networks were comprised of KEIs that promote cell death , densely populated with several RLK modules feeding into many cytosolic KEIs , and with most of the components functioning as negative regulators of the PCD; few nodes were shared . In the MKK7 and MKK9 networks , the diameter and path length indices were similar to HopQ1-1 ( Fig 6A; S5A Fig ) . Signaling flow through SSNs converged onto a set of KEIs associated with the global control of transcription and translation , ion and nutrient homeostasis , and extracellular acidification . Examples include KEI250/CIPK6 [40–42] , KEI33/CIPK11 [43] , KEI255/CIPK25 [44] , KEI323/S6K2 [45] , and KEI311/KIN10 [46] . To measure relative importance of the nodes in the SSNs we used maximum clique size algorithm ( MCC ) [47] , which finds clusters of the largest size in a given network; sub-graphs of essential nodes were derived based on their MCC rank for the PTI , ETS ( Fig 6B ) and ETI , PCD ( Fig 6C ) . Cytosolic kinases from the RLCK and MAPK-like families were preponderant essential nodes in both MCC-ranked graphs . Another parameter measuring centrality in networks is the betweenness centrality ( BC ) index , also regarded as a measure of the control potential of a node within a network [48] . Among all SSNs , the average BC indices were higher for MKKs and HopQ1-1 networks , indicating the importance of individual nodes on signaling outcome ( S5B Fig ) . In contrast , the signaling networks associated with PTI and ETS ( D29E , HopA1 , HopAI1 , HopAF1 , and AvrPto ) were smaller and had fewer high-control nodes ( low-centrality nodes ) , implying decreased efficiency in signal transmission . To determine if the SSN networks could be used to predict the performance of genes in the defense response , we tested 18 KEIs for their role in basal immunity in N . benthamiana . In this assay , immunity is first induced by inoculation with a non-pathogen ( P . fluorescens ) , followed by inoculation with the ETI-inducing P . syringae [49] . In the region where the inoculation areas overlap , little visible cell death develops , likely because of induced defense responses , limiting bacterial proliferation and secretion of the HopQ1-1 avirulence protein [50] . Most KEI-silenced lines had significantly increased cell death in the area infiltrated with both strains , indicating an impaired immune response as compared with the EC1 control and known immunity-promoting kinases ( BAK1 and FLS2 ) ( Fig 6D; S1 Data ) . Six of the eight highly MCC-ranked KEIs , including KEI149/PTI1-like , KEI104/BSK7 , KEI86/PBL5 , and KEI7/PBL8 , had statistically significant phenotypes; others , including KEI156 , 151/BIR2 , 160/IRK , 323/S6K2 , 304/LeMKK3 , and the PTI-promoting 72/SOBIR1 , also exhibited significant differences in the cell death intensity compared with controls ( Fig 6E ) . KEI91 LRR RLK , which had no significant phenotypes in the ETS , ETI , or PCD phenotyping , showed control-level cell death . The known and curated protein–protein interactions ( PPIs ) for the Arabidopsis homologs of these KEIs were extracted from public databases and used to generate a network ( Fig 6F ) , as described in the Supporting information ( S1 Materials and Methods—KEI signaling network analysis ) . The network has a PPI enrichment p-value of 3 . 1 × 10−8 , indicating that most of these kinases are biologically connected among themselves . To further test our SSNs predictions , we overlapped the information from our eight orthogonal phenotyping assays over the PPI network . The nodes connected by 70% of the edges ( 17 out of 25 ) co-occurred in various SSNs , indicating they may also be functional partners . Notably , 65% of edges connecting the interacting and functionally related KEIs co-occurred in more than two SSNs . For example , the interacting pairs KEI327/MPK6 and KEI323/S6K2 were part of the HopA1 , HopAI1 , and AvrPto SSNs . Overall , these results indicate the predictive potential of the SSNs for mapping plant defense networks and their response to perturbations . Plant immunity is generated as a result of numerous coordinated cellular processes . The study of inducible plant immunity requires approaches that reveal the organization and dynamics of the overall system and generate predictions on how molecular-level interventions can modify plant phenotypes . Building and characterizing biological networks , as a system-level approach to study plants , is starting to prove its effectiveness in predicting the function of cellular components and identifying biochemical and functional relationships among them [15 , 51–54] . Here , we describe a network-driven integrative analysis of the plant immune system , which includes in vivo plant–pathogen interactomics and a comprehensive study of kinase targets and identification of signal-specific networks . Some of the findings revealed by our approach included ( 1 ) that some effectors may bind several tomato kinases and that a proportion of kinases can interact with multiple effectors , ( 2 ) defense-associated kinase networks contain both shared and specific nodes involved in basal immunity , ETS , ETI , and PCD , ( 3 ) effector-triggered kinase networks are larger and more complex compared with a basal-defense network; however , they have fewer nodes with high centrality than unperturbed networks , and ( 4 ) previously uncharacterized kinases are essential for promoting bacterial resistance in N . benthamiana . A comprehensive characterization of the kinases identified in this study can provide insights into the underlying molecular mechanisms of defense and on the sensitivity and response to perturbations of plant defense networks , and will help identify targets for genome editing in crops . Our K-E screen predicts that interactions between plant proteins and pathogen effectors occur with a relatively low specificity when compared , for example , with receptor–ligand interactions . These observations confirm previous assumptions regarding effector promiscuity in target selection [13 , 55 , 56] and are supported by work demonstrating the functional interchangeability of P . syringae effectors [4] . By associating with multiple elements of a pathway , an effector may increase its chances to interfere successfully with the plant immune response . Furthermore , it may be evolutionarily beneficial for effectors to maintain the ability to interact with diverse partners to ensure functionality in new plant hosts with divergent immune signaling pathways [57–61] . On the other hand , HopM1 did not interact with any of the tested tomato kinases , suggesting a degree of target selectivity for some effectors . Indeed , target selectivity is further indicated by the fact that not all members of a kinase family interacted with the same effector . While this may be due to our experimental system , because effectors are rarely present individually and high expression of both kinase and effectors likely increased the chances for false positives , the effectors interacted with a mostly shared set of defense-associated kinases , suggesting the functional relevance of these interactions . Thus , while these interactions will have to be confirmed by additional methods , our findings indicate the effectiveness of using effector interactions as a starting point for genetic characterization . Interestingly , several effectors interacted with both positive and negative modulators of immunity , demonstrating that interaction alone is not sufficient to predict the role of a host target in defense ( HopA1 and PK3 or BAK1 ) . During pathogenesis , effectors have additive or synergistic effects on promoting virulence in plants , and the impact of individual effectors on immunity is typically minor or nonsignificant [62] . The overall contribution of individual effectors is likely dependent on both the relative importance of individual host targets within the defense network and on the status of the network itself , as different sectors are inactivated by other effectors . In addition , essential kinases such as BAK1 often play dual roles , depending on the status of other regulatory kinases in the cell . For example , AtBAK1 is essential for activation of PTI , but overaccumulation of AtBAK1 or loss of its negative regulator AtBIR1 can also activate immunity [63] . In biological networks , elimination of highly connected nodes ( hubs ) increases the diameter of the network [64] and has a deleterious effect on the characteristic path length and network integrity [65] compared with the removal of low-connectivity nodes . In this study , effectors appeared to neutralize hubs and nodes with high control potential in the network , thus having a detrimental effect on the structural integrity of the plant immune network . In addition , networks expanded in the presence of effectors , which may indicate plant deployment of new signaling sectors during purturbation . Interestingly , several susceptibility-linked KEIs were identified across defense networks . These KEIs may act as negative immune regulators or could be recruited to subvert plant pathways for the benefit of the pathogen [66 , 67] . Our results postulate that the composition and topology of plant signaling networks are determined by the plant’s ability to identify damage from effectors and activate compensatory pathways . Conversely , effector strategies to increase pathogen virulence consist in blocking/inactivating the sensor layers ( RLK/RLCK modules ) and recruiting kinases in the lower layers of the network for increasing pathogen fitness . Comparison of the MKK7 and MKK9 networks suggests an antagonistic relationship between the pathways activated by these MAP2Ks , whereby activation of one may cause inhibition of the other . MKK7 is a positive regulatory component of the immune response and systemic acquired resistance , operating via salicylic acid ( SA ) synthesis [68] , while MKK9 positively regulates ET signaling through increasing ETHYLENE-INSENSITIVE3 ( EIN3 ) receptor stability [69] . The complex functional relationship between SA and ET , comprising both synergistic [70 , 71] and antagonistic [72] interactions , provides additional strength to this model . Together , our network-centered approach has revealed the effect of individual effectors on signaling network topology and has facilitated the identification of novel immune kinases . However , several questions remain , including how effectors work together to modify the host immune network and if this information can be used to accurately predict the outcome of plant–pathogen interactions . A combination of systems biology approaches and genome editing has the potential to help address these questions and further the development of resistant plants for agricultural production . The coding region of the effector genes without the stop codon was cloned into the pENTR/SD/D-TOPO . The sequence for HopAI1 was amplified from P . syringae pv . tomato T1 using primers 5′-caccatgctcagtttaaagctgaacacccag and 5′-gcgagtccagggcggtggcatcag . All other effectors were obtained from P . syringae pv . tomato DC3000 . Hrp promoter-driven effectors fused at the C terminus with the HA tag were generated in the destination vector pCPP5372 [73] using Gateway cloning . pCPP5372 carrying different effectors was mobilized into DC3000D29E , a derivative of DC3000D28E lacking HopAD1 , by triparental mating using the helper plasmid pRK2013; Trans-conjugants were selected on KB medium with appropriate antibiotics . DC3000D28E::ShcM HopM1 has been described previously [74] . Bacteria were maintained on King’s B medium at 37 °C . Cloning of the tomato KEIs and the SLC method were described previously [21] . To create clones for VIGS of orthologous kinases in N . benthamiana , tomato gene sequences were analyzed using bioinformatics tools available at solgenomics . net; the VIGS tool and the optimal gene fragment with the fewest off-targets were used to design primers . Gene fragments were amplified from N . benthamiana cDNA , cloned into the TOPO pER8 Donor vector using the manufacturer’s protocol , subcloned into the TRV2 expression vector , and transformed into Agrobacterium GV2260 for expression in planta [75] . Each interaction was tested in 4 to 16 independent assays , and the reconstituted luminescence was recorded at six time points . The decision to test over the minimum of four times was taken for the pairs showing significant levels of interaction when compared with the reference sets , while up to 16 assays ( four biological replicates ) were used for K-E pairs showing variability or low interaction levels . The interactions were corrected for multiple testing with a false discovery rate ( FDR ) of of 0 . 05 . The analysis of SLCAs is described in S1 Materials and Methods . The KEI-silenced lines were produced by syringe-infiltrating leaves of 2-week-old N . benthamiana plants with the TRV2-KEI Agrobacterium clones along with TRV1-containing Agrobacterium at a 1:1 ratio as described [75] . The EC1 and FLS2 constructs [49] served as controls and were included in each round of KEI line testing . KEI lines were grown ( 16 light , >50% humidity ) in 6-inch-diameter pots for 3 weeks before testing . All functional assays were done using the third and fourth fully expanded leaves . Bacterial growth was tested in infiltrated leaves at 6 dpi . Each plant was tested once with each strain , and three plants were tested per round of KEI-silenced line production . Each KEI-silenced line was tested over a minimum of 3 and maximum of 18 trials alongside the controls EC1- and FLS2-silenced lines , resulting in between 9 and 56 biological replicates per KEI–P . syringae strain combination . The analysis of bacterial growth assays is described in S1 Materials and Methods . Two to three leaves of KEI-silenced lines were syringe infiltrated with Agrobacterium carrying the MKK7DD or MKK9DD as described previously [1] . The D29E + HopQ1-1 strain was applied by syringe inoculation at a level of 3 × 108 CFU/mL . For both MAP2Ks and HopQ1-1 induction of PCD , the area of infiltration was marked , and the intensity of PCD was quantified over the 3 days after infiltration ( dpi ) , as in [1] . The PCD was scored as 1 = 0%–25% , 2 = 26%–50% , 3 = 51%–75% , and 4 = 76%–100% of the infiltration area demonstrating necrosis . The data presented are using values at 1 dpi for HopQ1-1 and 2 dpi for MAP2K treatments . All three treatments were applied to the same leaf , and three plants were infiltrated per each round of VIGS . A minimum of three biological replicates were performed; 9 to 34 plants were tested per treatment . The analysis of PCD assays is described in S1 Materials and Methods . We used Clustal Omega tool [76] to align the sequences and iTOL [77] to build the phylogenetic tree of the 35 analyzed KEIs ( Fig 4F and S2 Data ) . A sharable link is provided at https://itol . embl . de/tree/13018201144195851480695765# . We constructed a table of KEIs’ role in immune response using as rows the KEIs ( clustered along the gene family structure using phylogenetic analysis ) and as columns the classes of immune responses: PTI , ETS , ETI , MKK7ox , and MKK9ox . Each kinase was represented by an importance score , defined as the cumulative phenotype strength in each analyzed process: IS ( Ki , P ) =∑s∈P-log10 ( pVal ( Ki , s ) ) ;P , immunereponseclass;s , stress;Ki , kinase . The heat map displays the importance scores of kinases versus immune processes ( red: positive effect on immune response; blue: negative effect on immune response ) . The map shows a pattern of distinct RLKs having positive effect in PTI , ETS , and ETI ( presumably by triggering immune responses ) , while cytosolic kinases have a negative effect on all immune response processes ( possibly by contributing to pathogen growth and spread ) . The network inference model uses the following matrices: ( 1 ) networks effect matrix: a matrix containing the decisions of phenotype testing on stress assays for each KEI; ( 2 ) phenotype effect matrix: a matrix containing the phenotype effect ( positive or negative ) of the stress assays for each KEI; and ( 3 ) co-occurrence matrix: a matrix containing the number of co-occurrences of pairs of kinases in treatments . In addition , it contains a set of structural constraints and rules for network structure inference . The objective of our pathway inference method is to minimize the maximum co-occurrence pattern divergence for nodes included in the same pathway . We use a co-occurrence pattern similarity measure defined as S ( A , B ) = ( \|A⋂B\|\|A\| ) x ( \|A⋂B\|\|B\| ) and pattern overlap measure OA ( A , B ) = ( \|A⋂B\|\|A\| ) . We developed a method to infer the signaling graph ( the corresponding adjacency matrix ) from the network effect matrix , co-occurrence matrix , and phenotype effect matrix subject to the structural constraints rules . The method consists of the following steps: T1 , T2 , and T3 are control parameters with geometric rate convergence to equilibrium values .	Some bacterial pathogens secrete virulence factors called effectors , which influence host tissues during infection . The impact of such bacterial effectors on the transmission of immune signals in plants remains poorly understood . In this study , we developed an integrative network approach to discover interactions between bacterial effectors and a class of host signal-mediating enzymes called protein kinases . We also characterized the functions of the targets of these kinases in order to understand how bacterial effectors might disrupt the flow of information in signaling pathways within plant cells . We show that plants activate larger signaling networks when inoculated with pathogens that produce effectors . We also find that plant signaling networks are specific to individual effectors and that the networks include kinases with both positive and negative effects on plant resistance to pathogens . We propose that the topology of immune signaling networks is determined by the plant’s ability to activate compensatory pathways in response to the effectors’ network-disruptive actions . Conversely , pathogens may increase their virulence both by disrupting host signaling at the membrane-located end of the signaling network and by recruiting cytosolic kinases . This work provides a framework for the study of plant–pathogen communication and could be used to prioritize targets for improving resistance in crops .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "cell", "death", "medicine", "and", "health", "sciences", "pathology", "and", "laboratory", "medicine", "pathogens", "cell", "processes", "immunology", "signaling", "networks", "microbiology", "cell-mediated", "immunity", "network", "analysis", "plants", "flowering", "p...	2018	Integrative network-centric approach reveals signaling pathways associated with plant resistance and susceptibility to Pseudomonas syringae
Human IgG1 antibody responses are associated with protection against Schistosoma haematobium infection and are now a target for schistosome vaccine development . This study aimed to investigate the relationship between total IgG and the IgG subclasses and the monocyte IgG receptor , known as FcγRIIIa or CD16 , in schistosome exposed people . Systemic levels of schistosome-specific anti-adult worm total IgG and IgG subclass titres were measured by ELISA in 100 individuals from an S . haematobium endemic area in Zimbabwe and , using parametric statistical methods and regression analysis , related to the levels of CD16 expression on individuals' circulating monocytes , determined via flow cytometry . Monocyte CD16 expression rose with parasite-specific total IgG and IgG1 in healthy participants , but not in schistosome infected patients . Similar to parasite-specific IgG and IgG1 , CD16 expression in healthy individuals is associated with protection against schistosome infection . This relationship indicates a mechanistic link between the innate and adaptive immune responses to helminth infection in protection against infection . Further understanding the elements of a protective immune response in schistosomiasis may aid in efforts to develop a protective vaccine against this disease . An estimated 200 million people worldwide are infected with helminths of the genus Schistosoma , with the heaviest burden of disease occurring in sub-Saharan Africa , where both Schistosoma haematobium and Schistosoma mansoni are endemic , causing significant morbidity amongst affected communities [1] . Infection and disease are controlled by treatment with the drug praziquantel ( PZQ ) , and the World Health Organization ( WHO ) recommends protective chemotherapy via mass drug administration ( MDA ) with PZQ in endemic areas [2] . There is mounting pressure to develop a vaccine against schistosomiasis , which would provide long term protection to the 650 million people at risk of exposure [3] , and pre-empt the development of drug resistance . Current vaccine development research focuses on determining which naturally developed immune responses are associated with protective immunity that develops in the context of endemic exposure to infection , and investigate ways of inducing those responses artificially whilst avoiding a pathological response [4] , [5] . While significant progress has been made in characterising humoral and cellular responses in experimental models , relatively less work has been conducted relating the innate and adaptive arms of the immune system in schistosome infected versus uninfected humans . In particular , there is a paucity of studies simultaneously determining cellular and related humoral responses associated with natural protection against schistosome infection . Experimental studies have shown links between innate cells from the myeloid lineage and resistance to helminth infection . For example , murine macrophages and are involved with tissue repair and fibrosis [6] , [7] , as well as in limiting pathology by regulating Type 2 cytokine production [8] , [9] and inhibiting T cell proliferation [10] . This current study focused on circulating monocytes , myeloid cells related developmentally to macrophages , which are present in the blood vessels and are thus easily accessible for investigation in humans . Studies from several decades ago showed a direct role ex vivo for human PBMC-derived monocytes in the killing of schistosomula [11]–[13] . Similar to macrophages , monocytes display phagocytic capabilities and express varying levels of the FcγRIIIa ( also known as the CD16 receptor ) [14] , which is related to distinctions in their phenotype and function in a range of pro-inflammatory conditions [15] , [16] . The Fcγ receptors have a critical role in immune regulation , acting as a link between the humoral and innate cellular arms of the immune response [17] . In humans , the CD16 receptor exhibits high affinity binding to the Fc portion of IgG antibodies , with high affinity binding demonstrated to IgG1 and IgG3 , which leads to phagocytosis , release of inflammatory mediators and clearance of immune complexes [14] . The importance of the interaction between IgG and Fcγ receptors has been demonstrated in experimental models , whereby there is a diminished macrophage effector function induced after IgG1-mediated phagocytosis in Fcγ chain knock-out mice [18] . Furthermore , S . mansoni infection exacerbated granuloma formation and fibrosis in both Fcγ receptor and in B cell deficient mice [19] , highlighting the importance of antibody signalling via the Fcγ receptor in protection against pathology associated with schistosomiasis infection . However , there are few studies relating IgG subclasses to the Fcγ receptors in human schistosomiasis . To address this knowledge gap , the present study focuses on the relationship between CD16 and the IgG subclasses . Our previous studies , and those of others , have shown that , in humans , schistosome-specific IgG1 and IgG3 antibodies are associated with natural resistance to infection [20]–[22] . Induction of helminth-specific IgG1 and IgG3 through vaccination is now preferred over IgE to avoid generating IgE-mediated pathological responses to vaccination [4] , [23] and , in particular , this study focuses on these protective subclasses . This study , therefore , investigated the relationship between levels of the IgG receptor , FcγRIIIa ( CD16 ) and schistosome specific IgG subclasses in uninfected , healthy individuals versus schistosome infected patients . The healthy individuals comprised of young people , who had yet to acquire schistosome infection , and older people who were putatively resistant to infection , as they were infection free despite being lifelong residents of the schistosome endemic areas and experiencing regular exposure to infective water . The study focused on adult schistosome-specific IgG responses since adult worms reside in the circulating blood , and thus are in direct contact with monocytes in this compartment . In addition , our studies , as well as those of others , have highlighted the importance of the adult worm stage in stimulating protective immune responses [24]–[27] . Ethical and institutional approval was granted by the Medical Research Council of Zimbabwe and the University of Zimbabwe's Institutional Review Board . Local permission for the study was granted by the Provincial Medical Director . The study design , aims and procedures were explained in the local language , Shona , prior to enrolment . Participants were free to drop out of the study at any time and informed written consent/assent was obtained from all participants and/or their guardians prior to taking part in the study and to receiving antihelminthic treatment . The study presented here was part of a larger on-going immuno-epidemiological study based in Mashonaland East , Zimbabwe where S . haematobium is endemic [28] . The area has a low prevalence of soil transmitted helminths ( STH ) and Schistosoma mansoni [29] , and the residents are subsistence farmers with frequent contact with infected water for purposes of bathing , washing and collecting water . Recruitment into the study was school based and the wider community was also invited to participate . Residential history , antihelminthic treatment history and water contact habits of the participants were captured through questionnaire . Following sample collection , participants were offered treatment with the antihelminthic drug praziquantel at the recommended dose of 40 mg/kg of body weight [3] . In order to be included in this study participants had to meet the following criteria: 1 ) be lifelong residents of the study area to allow age to be used as a proxy for history of exposure to schistosome infection , 2 ) have provided a minimum of two urine and two stool samples on consecutive days for parasite detection , 3 ) not have previously received antihelminthic treatment , 4 ) be negative for co-infection with malaria , STH , S . mansoni and HIV and 5 ) have provided a blood sample for serological and cellular assays . Further to this , participant's PBMC sample must have yielded at least 106 cells to allow enough cells for all experimental conditions . From an initial cohort of 633 recruited individuals , 68 were excluded for not meeting criteria 1–4 above and a further 184 did not provide sufficient blood sample for both serological assays and cell phenotyping . From the remaining 381 individuals , a cohort of 100 individuals was further selected to allow for , as far as possible , equal numbers of females to males and an even distribution of ages and infection prevalence . Individuals with one or more S . haematobium eggs found in their urine samples were classified as infected . The final study group was divided into three age groups and is described in Table 1 . From each participant a stool and urine sample was collected on three consecutive days and examined microscopically for the presence of S . haematobium eggs in urine , and S . mansoni and STH eggs in stool using standard techniques [30] , [31] . A random sample of 100 stool samples was also processed via the formol ether concentration technique , and these confirmed Kato Katz diagnosis [32] . Up to 20 millilitres of venous blood was collected from each participant in heparinised tubes or silicone –coated tubes ( both from BD Biosciences , San Jose , CA ) , for purposes of processing for PBMC purification ( heparin tubes ) , or serum ( silicone tubes ) using routine methods . An additional drop of blood was collected from each participant for microscopic detection of malaria parasites and for HIV detection using DoubleCheckGold HIV 1&2 Whole Blood Test ( Orgenics Ltd . , Yavne , Israel ) . Peripheral blood mononuclear cells ( PBMC ) were isolated from the remaining tubes via density gradient centrifugation using Lymphoprep ( Axis-Shield , Cambridgeshire , UK ) . Isolated PBMCs were cryopreserved and stored in liquid nitrogen in Zimbabwe prior to freighting to Edinburgh in dry shippers where the overall viability of isolated PBMCs was estimated in a sample of 84 individuals' PBMCs using propidium iodide ( PI ) ( Sigma-Aldrich , Dorset , UK ) exclusion . Mean PI uptake in this sample was 15 . 3% ( +/−1 . 06% ) , which is within the range considered viable . Schistosome soluble worm antigen preparation ( SWAP ) -specific antibody serum levels for total IgG , IgG1 , IgG2 , IgG3 and IgG4 were quantified using antibody ELISA . Lyophilized SWAP ( Theodor Bilharz Institute , Giza , Egypt ) was reconstituted as recommended by the manufacturer and as described by Mutapi et al . [25] . ELISAs were conducted as previously reported [22] , using 5 µg/ml of SWAP antigen in carbonate bicarbonate buffer to coat all ELISA plates , and adding sample at a 1∶100 dilution in 5% skimmed milk . Secondary IgG HRP-conjugated antibody was added at a 1∶1000 dilution for total IgG and IgG1 , and at a 1∶500 dilution for IgG2 , IgG3 and IgG4 . The colorimetric reaction was left for 10 minutes for total IgG , and 15 minutes for the IgG subclasses , and quantified with an ELISA reader at 405 nm . Each antibody ELISA was performed in duplicate on the same day for all samples with positive and negative controls on each plate . Cryopreserved PBMCs were thawed as previously described [33] , and resuspended at 5×106 cells/ml in PBS . Cells were incubated with 10% FCS at 4°C for 10 minutes prior to staining for 30 minutes with Alexa488 conjugated anti-CD14 ( clone M5E2 ) , PE-Cy7 conjugated HLA-DR ( clone L243 ) ( both from BD Biosciences , San Jose , CA ) , and Pacific Blue conjugated anti-CD16 ( clone CB16 ) ( eBiosciences , San Diego , CA ) . Unbound antibodies were washed off and cells were resuspended in PBS prior to acquisition of at least 50 , 000 live events on a BD FACS LSR II ( BD Biosciences , San Jose , CA ) . Compensation was performed prior to acquisition of each experiment using BD FacsComp beads ( BD Biosciences , San Jose , CA ) . Analysis was performed using FlowJo software ( TreeStar , USA ) and mean fluorescence intensity ( MFI ) was calculated for CD16 . To ensure that only CD14 positive cells representing monocytes were analysed , only cells expressing both HLA-DR and CD14 were selected for analysis , in a gating strategy previously described [28] , [34] . Briefly , a live gate to include all leukocytes was drawn based on forward scatter ( FSC ) and side scatter ( SSC ) , HLA-DR positive cells were gated to exclude any CD16+ expressing natural killer ( NK ) cells as well as other non-MHC expressing cells . Monocytes were defined as CD14 or CD16 expressing cells . The expression level of CD14 and CD16 was reported as MFI . All statistical analyses were conducted using the statistical package SPSS version 19 ( IBM Corp , NY , USA ) . Due to the possibility of gender and age dependent exposure patterns in this population [35] , [36] , appropriate statistical techniques were necessary to adjust for this variation prior to investigating the relationship of interest [37] . Parametric statistical modelling in the form of analysis of variance ( ANOVA ) and linear regression was therefore used . Data were transformed in order to meet assumptions of parametric tests . Surface marker expression ( measured as MFI ) was log transformed ( log10 ( x+1 ) ) . Antibody level ( after subtraction of the blank control ) was square root transformed . Categorical variables were sex ( male/female ) , infection status ( uninfected/infected ) and age group ( 5–10 years [age group where infection is rising] , 11–15 years [age group where infection is peaking] or >16 years [age group where infection is declining] ) . To determine the extent of changes in the proportion of monocytes relative to the rest of the PBMCs , ANOVA with sequential sums of squares ( SS ) was used with the total number of live monocytes as the dependent variable , and the independent variables were sex ( male or female ) , age group ( 5–10 years , 11–15 years or >16 years ) and infection status ( uninfected or infected ) . Monocytes from different individuals show varying levels of CD14 and CD16 expression intensity dependent on various factors including age [38] , [39] and presence of inflammation [40] . Therefore , to test the hypothesis that CD14 and CD16 receptor expression levels changed with infection and age group , a multivariate analysis of variance ( MANOVA ) with sequential SS was used with receptor expression ( CD14 and CD16 ) as the dependent variable , and the independent variables were sex , age group and infection status . The model was extended to include an interaction term between age and infection to assess patterns of surface receptor expression co-dependent on age and infection status . The relationship between CD16 expression and age group dependent on infection status was investigated further using a partial correlation , controlling for variation due to sex , and followed with Fisher's z transformation , which tests for a significance in the difference between two correlations . The relationship between each of the anti-SWAP antibody titres ( total IgG , IgG1 , IgG2 , IgG3 , IgG4 ) with infection and age group was tested using a univariate ANOVA with sequential SS , entering sex and age group before infection in the analysis . In addition , the interaction between infection and age was tested and all appropriate post-hoc tests were conducted . Finally to determine the relationship between CD16 expression on monocytes and antibody levels according to infection status , a linear regression analysis was used . The relationship between antibody levels and CD16 expression according to infection status was investigated after accounting for sex and age differences , and significant interactions were followed up with a correlation analysis of the relationship between the two variables in uninfected and infected individuals separately . For all statistical tests significance of p≤0 . 05 was considered significant . The proportion of monocytes in PBMCs did not vary significantly with host age or infection as shown in Figure 1A . Thus , any changes observed in subsequent analyses will be a result of changes in monocyte phenotype between individuals , rather than as a result of changes in monocyte proportions , for example due to migration of monocytes into or out of the vasculature . The expression levels of CD14 and CD16 on monocytes were investigated with relation to infection status . CD16 expression levels were shown to vary dependent on age and infection status ( Table 2 and Figure 1B ) . Thus , while expression of CD16 on monocytes was similar between infected ( schistosome patients ) and uninfected ( healthy ) individuals in the younger age groups , with increasing age there was an increasing intensity of expression in the healthy individuals compared to a decreasing intensity of expression with age in the infected individuals as shown in Figure 1B . Consistent with the heterogeneous relationship between age and infection status , the correlation coefficient was significantly different ( z = 2 . 97 , p = 0 . 003 ) with schistosome patients having a positive relationship ( r = 0 . 350 , p = 0 . 006 ) and healthy individuals having a negative relationship ( r = −0 . 287 , p = 0 . 085 ) . The oldest age group demonstrated significant differences in monocyte CD16 expression dependent on infection status ( Figure 1B ) . In contrast , levels of the monocyte marker CD14 did not vary with any of the investigated variables ( Table 2 and Figure 1C ) . The relationship between SWAP-specific total IgG , as well as the IgG subclasses and infection status dependent on age was investigated . Total IgG , IgG1 , IgG2 and IgG4 , but not IgG3 , significantly varied with host age , with total schistosome-specific IgG levels lowest in the age of peak schistosome infection , but the adult worm-specific subclasses IgG1 , IgG2 and IgG4 highest in the oldest age group ( Table 3 ) . Overall , levels of parasite-specific IgG and IgG1 varied between healthy participants compared to schistosome infected people , with infected individuals having greater antibody levels compared to uninfected individuals ( Table 3 ) . However , only for total IgG , the effects of infection status varied with host age as indicated by the significant age group-infection status interaction term ( Table 3 ) . Figure 2 demonstrates the gradual increase in levels of schistosome-specific total IgG in healthy participants , while in schistosome infected patients , total IgG levels did not change significantly with age . For total IgG , IgG1 , IgG3 and IgG4 , but not IgG2 , the youngest age group showed significant differences in antibody levels dependent on infection , with the infected individuals showing greater antibody levels compared to the uninfected individuals ( Figure 2 ) . The relationship between intensity of CD16 expression on all monocytes and IgG levels against SWAP was investigated to test the hypothesis that innate cell ( monocyte ) phenotype is related to schistosome-specific acquired immune markers ( IgG ) . As both CD16 expression and IgG antibody titres showed significant relationships with infection status , the population was partitioned by infection status prior to analysing the relationship between IgG and CD16 by regression analysis . This analysis showed that in healthy , uninfected individuals , expression levels of CD16 rose significantly with levels of total IgG and IgG1 after allowing for variation due to age and sex ( Table 4 ) . Expression levels of CD16 in infected individuals did not show a relationship with parasite-specific total IgG or IgG1 . In addition , levels of IgG2 , IgG3 and IgG4 did not show a relationship with CD16 expression levels in either infected or uninfected participants ( Table 4 ) . Human schistosome-specific IgE and IgG1 responses have been shown to be associated with resistance to infection [41]–[43] . In order to understand naturally developed protective immune responses that can be targets for artificial induction through vaccination , previous studies in schistosomiasis have focussed on describing the interaction between IgE and cellular mediators of protective effector responses , such as eosinophils and macrophages . However , following Phase 1b vaccination trials on the human hookworm vaccine candidate , it was found that inducing an IgE response in naturally exposed people caused a pathological immune response , compromising the vaccine's safety [23] . Subsequently , there has been a shift from developing IgE-mediated helminth vaccines towards vaccines that induce the IgG1 and IgG3 subclasses . This study focused on determining the relationship between schistosome-specific IgG subclasses and IgG FcγRIIIa ( CD16 ) on human monocytes , a cell type related developmentally to the macrophage and which has been shown to have a significant role in immune responses to experimental schistosome infection . Monocytes have previously been demonstrated to be involved in immune response to schistosome larvae [11] , [12] , [44] , and as circulating cells interacting with adult schistosome stages in the venules , monocytes are a highly relevant , but under studied innate immune cell type in human schistosomiasis . The relationship of increasing IgG subtypes with age , as well as the higher expression of IgG against adult worm antigen in infection , has previously been reported [20] , [45] . Anti-SWAP IgG1 and IgG4 are both dominant antibody subclasses in human schistosomiasis , while IgG2 and IgG3 are detected at lower levels [46] . In particular , IgG1 is associated with protection or developing immunity [43] and IgG4 is associated with infection [47] . These schistosome specific responses observed here in both the older uninfected and infected groups confirm that both age groups have been exposed to schistosome infection . Although the age at which this exposure occurred in the uninfected group is not clear , the increasing levels of all the antibody subclasses with age represent the cumulative exposure to adult schistosome antigens throughout the population's lifetime . In human immuno-epidemiological studies , the uninfected , older individuals are classified as putatively resistant to infection via an immune-mediated mechanism [43] , [48] , and the gradual increase in parasite-specific total IgG and IgG1 in this group is consistent with schistosome-specific immune responses associated with protection [20] , [22] . The presence of infection in the older individuals who are lifelong residents of this schistosome endemic area suggests that they are carrying chronic infections , a fact that has been corroborated by other studies conducted in the same area and including some of the same participants [49] . In particular , chitinase 3-like 1 protein , a marker of inflammation that has been linked to schistosome-related hepatic fibrosis [50] , was found to be highest in the older people harbouring infection [49] . Indeed , with increasing age , and thus duration of exposure to schistosomes , differences in the immune system become more apparent , and studies investigating myeloid derived dendritic cells from members of the same residential area have , similar to results presented here , shown age related changes with infection [33] . In this study , members of the younger patient cohort will have had a shorter infection history compared to the older age groups , and therefore have experienced much less schistosomiasis associated pathology [5] , [42] , [51] . The differences in CD16 expression dependent on age and infection may therefore be indicating an altered immune activation status in relation to schistosome infection , and may indicate a potential link between CD16 expression and pathology . Indeed , the pattern of increasing CD16 expression with age , observed here in the healthy individuals , has previously been noted in other populations [38] , [39] , and CD16 expression has previously been reported to be upregulated with monocyte maturation and activation [15] , [52] . The absence of a relationship between monocyte CD14 expression levels and schistosome infection is likely related to CD14 being a lipopolysaccharide ( LPS ) receptor , which is involved in immunity against bacterial challenge [53] , [54] , and therefore playing a less significant role in schistosome-specific responses induced by adult worm antigens . In contrast to the significant relationship between CD16 and infection status shown here , we found no relationship between infection intensity and either CD14 or CD16 expression ( data not shown ) , indicating that it was the presence of infection in the host that was important in this relationship rather than the burden of infection , a pattern that has been demonstrated in other immunological and pathological features of human schistosomiasis [49] , [55] . The positive correlation observed between the protective IgG antibodies ( total IgG and IgG1 ) and monocyte CD16 in uninfected individuals , indicates that the CD16 expression level on monocytes may be associated with protection against infection , in association with an activated monocyte phenotype . Observations from research into monocyte involvement in human HIV infection report on CD16 - IgG mediated ADCC activity [56] , [57] , and there may be a similar mechanism mediating protection in schistosomiasis , in particular involving CD16-IgG1 interactions . However , the precise role of any ADCC mechanism warrants further investigation . Importantly , consistent with results from mechanistic mouse experimental studies of schistosomiasis [19] , this relationship suggests a link between the innate and adaptive arms of the immune system in the response to schistosomiasis , which may be important for furthering vaccine research efforts . The lack of association between the other IgG subclasses , IgG2 , IgG3 , and IgG4 may be due to antibody properties such as length of memory response [58] , [59] and lack of affinity between the Fcγ receptor and the antibody , as is the case with IgG4; or the relatively low levels of expression associated with schistosome infection , as is the case with IgG3; or a combination of these factors , as may be the case for IgG2 . Although the study cannot determine causality from the observations made here , transfection studies of a macrophage cell line demonstrated that chronic inflammation inhibits FcyRIIIa ( CD16 ) glycosylation , in turn reducing the ability for CD16 receptor activation following IgG binding [60] . Thus , chronically infected individuals in this study may have deficiencies in their CD16 receptor contributing to continued infection . In addition to the Fcγ receptor , CD16 , monocytes also express CD32 ( also known as FcγRIIb ) an inhibitory receptor [61] , and CD64 ( also known as FcγR1a ) , a high affinity receptor for IgG [62] . Differential expression of these receptors may further indicate the capability of monocytes to activate in response to infection , and may enlighten on the role of other IgG subtypes in schistosome infection . Altogether , our study demonstrated that monocyte CD16 expression is associated with protection against schistosome infection . The level of CD16 in healthy individuals is positively associated with levels of total IgG and IgG1 , antibodies which have previously been associated with resistance to infection . Conversely , in schistosome infected patients who are lifelong residents of a schistosome endemic area , CD16 expression is significantly reduced . This decrease in expression of a monocyte activation marker , combined with the lack of association with protective IgG , may be a result of an altered immune activation state in chronic schistosomiasis infection .	Schistosomiasis is a parasitic disease caused by the parasite Schistosoma spp . Over 240 million people are infected worldwide , mainly in Sub-Saharan Africa , but an efficacious , protective vaccine has yet to be found . Protection against schistosome infection in individuals living in endemic areas is mediated by antibodies . In particular , IgG1 antibody has been shown to be protective against infection in individuals living in endemic areas , and eliciting IgG1 production has become a cornerstone of vaccine development efforts . However , little is known about the mechanisms by which IgG1 induces protection . The cell surface molecule CD16 is an IgG antibody receptor expressed on monocytes and binds preferentially to IgG antibody subclasses . The work presented here thus investigates the relationship between IgG levels and the monocyte CD16 receptor in a population endemically exposed to infection with schistosomes . We present results linking CD16 expression with IgG1 levels , whereby uninfected individuals have a positive relationship between IgG1 and CD16 expression levels , while schistosome infected individuals did not show any statistically significant relationship between the two . Thus we provide evidence to suggest a mechanistic link between the innate and adaptive immune response in parasitic infection , associating monocyte CD16 expression with a protective immune response .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "blood", "cells", "humoral", "immunity", "innate", "immune", "system", "immune", "cells", "antigen-presenting", "cells", "immunology", "microbiology", "parasitology", "antibodies", "immune", "system", "proteins", "white", "blood", "cells", "animal", "cells", "proteins",...	2014	CD16 Expression on Monocytes in Healthy Individuals but Not Schistosome-Infected Patients Is Positively Associated with Levels of Parasite-Specific IgG and IgG1
Kaposi sarcoma-associated herpesvirus ( KSHV ) causes several tumors and hyperproliferative disorders . Hypoxia and hypoxia-inducible factors ( HIFs ) activate latent and lytic KSHV genes , and several KSHV proteins increase the cellular levels of HIF . Here , we used RNA sequencing , qRT-PCR , Taqman assays , and pathway analysis to explore the miRNA and mRNA response of uninfected and KSHV-infected cells to hypoxia , to compare this with the genetic changes seen in chronic latent KSHV infection , and to explore the degree to which hypoxia and KSHV infection interact in modulating mRNA and miRNA expression . We found that the gene expression signatures for KSHV infection and hypoxia have a 34% overlap . Moreover , there were considerable similarities between the genes up-regulated by hypoxia in uninfected ( SLK ) and in KSHV-infected ( SLKK ) cells . hsa-miR-210 , a HIF-target known to have pro-angiogenic and anti-apoptotic properties , was significantly up-regulated by both KSHV infection and hypoxia using Taqman assays . Interestingly , expression of KSHV-encoded miRNAs was not affected by hypoxia . These results demonstrate that KSHV harnesses a part of the hypoxic cellular response and that a substantial portion of hypoxia-induced changes in cellular gene expression are induced by KSHV infection . Therefore , targeting hypoxic pathways may be a useful way to develop therapeutic strategies for KSHV-related diseases . Kaposi sarcoma-associated herpesvirus ( KSHV ) is the etiologic agent for several hyperproliferative disorders and tumors , including Kaposi’s sarcoma ( KS ) , primary effusion lymphoma ( PEL ) and a form of multicentric Castleman disease ( MCD ) [1–4] . Like other herpesviruses , KSHV has two patterns of gene expression: latent , in which only a small subset of genes are expressed; and lytic , in which the full repertoire of genes are expressed and viral progeny are produced [5] . A number of recent studies have shown that hypoxia and hypoxia-inducible factors ( HIFs ) are important in the KSHV life cycle and the pathogenesis of KSHV-induced diseases [6–8] . Two of the tumors caused by KSHV , KS and PEL , preferentially arise in relatively hypoxic environments: the extremities and pleural effusions , respectively [9 , 10] . Cells respond to hypoxic environments by a rapid up-regulation in their levels of two main HIFs , HIF-1 and HIF-2 , which in turn enter the nucleus and activate HIF-responsive genes by binding to hypoxia response elements ( HRE ) in their promoter regions [11 , 12] . Hypoxia and HIFs can also up-regulate levels of the cellular microRNA ( miRNA ) , miR-210 , which in turn affects a number of target genes to promote adaptation to hypoxia [13 , 14] . Interestingly , exposure of KSHV-infected PEL cells to hypoxia or to HIFs has been shown to induce lytic KSHV replication [7] and also to directly up-regulate certain KSHV genes , including the lytic switch gene replication transcription activator ( RTA ) [6] , the open reading frame ( ORF ) 34 to 37 cluster of lytic genes [15] , and latency-associated nuclear antigen ( LANA ) [16] . KSHV infection can induce a number of changes in the gene expression pattern of target cells that facilitate viral infection , persistence , avoidance of host antiviral strategies , and when appropriate , viral replication . Some of these changes are induced by KSHV proteins , which can modulate a number of intracellular signaling pathways and the production of growth factors and cytokines [17 , 18] . Other changes are caused by KSHV-encoded miRNAs , which are generated from a dozen precursor miRNAs [19–21] . It has been shown that among the changes mediated by KSHV are the induction of a hypoxic phenotype and the increase in the levels of and activation of HIFs under certain conditions [16 , 22 , 23] . These changes are mediated by several KSHV-encoded proteins , including latency-associated nuclear antigen ( LANA ) [16] , viral interferon regulatory factor 3 ( v-IRF3 ) [24] , and viral G protein-coupled receptor ( v-GPCR ) [22] . However , whether or not these three KSHV proteins are necessary to activate HIF is unknown . Our group recently used next-generation sequencing ( NGS ) analysis to explore the changes in cellular mRNA and miRNA expression profiles induced by latent KSHV infection using the SLK tumor line [25] . Under normoxic conditions , KSHV infection modulated many mRNAs including a number that had been previously reported to be affected by hypoxia . Also , expression of a number of human miRNAs was different in KSHV-infected SLKK cells as compared to uninfected SLK cells , including a small but statistically significant up-regulation of miR-210 [25] . In the current study , we extended these findings by using NGS to assess the changes in mRNA and miRNA expression profiles induced in this same SLK line by exposure to hypoxia , and we compared the findings to those induced by chronic KSHV infection . In addition , we studied the changes in mRNA and miRNA expression profiles after exposure to hypoxia in SLKK cells to those in KSHV-uninfected SLK cells in order to understand how hypoxia and KSHV may interact to modulate cellular gene expression . To assess the changes in mRNA gene expression induced by hypoxia in KSHV-uninfected SLK cells , next-generation sequencing was performed on mRNA libraries from hypoxic and normoxic uninfected SLK samples in triplicate . Out of a total of 30 , 808 mRNAs analyzed , 519 annotated genes were significantly differentially expressed ( P ≤0 . 05 , fold change ≤-2 or ≥2 ) in hypoxic vs . normoxic SLK cells ( Fig 1A ) . The most abundant genes ( average read count > 100 ) that were differentially expressed are depicted in Fig 1B . Importantly , among the 210 up-regulated genes were a number that have been described previously to be increased in hypoxia , such as BNIP3 , BNIP3L , DDIT4 , LDHA , SLC2A1 ( GLUT1 ) and STC2 ( Fig 1A and 1B ) [26–31] . As a confirmation that expected cellular mRNAs were responding to hypoxia , we assessed the expression of vascular endothelial growth factor ( VEGF ) , a well-described hypoxia-responsive gene [32] , in normoxic and hypoxic SLK cells . By RNA-seq , VEGF expression was up-regulated 2 . 1 fold but that change was not considered significant as P = 0 . 06 , which is above our threshold of P≤0 . 05 . However , we separately confirmed by RT-qPCR that VEGF was significantly up-regulated ~2 . 5 fold by hypoxia ( P<0 . 005 , Fig 1C ) . Taken together , these results provide confidence that changes seen in the SLK cells under hypoxic treatment were in fact due to hypoxia . The same SLK hypoxic samples that were analyzed for mRNA-Seq were also sequenced for small RNA-Seq to assess differential miRNA expression . A total of 112 out of 1 , 378 miRNAs were found significantly deregulated in hypoxia compared to normoxia ( P ≤0 . 05 , FC ≤-2 or ≥2; Fig 1D , top left and right sectors ) . Perhaps the most striking observation was that of the 112 miRNAs deregulated in response to hypoxia , a large majority ( 99 miRNAs , or 88% ) were down-regulated . Of the 112 hypoxia-regulated miRNAs , also known as hypoxamiRs [33] , 47 were expressed with an average miR read count greater than 1 , illustrated by our heat map analysis; this showed a clear distinction between the triplicate hypoxic and normoxic samples ( S1A Fig ) . Ten ( 21% ) of the 47 hypoxamiRs with a read count greater than 1 were up-regulated , while 37 ( 79% ) were down-regulated by hypoxia . After a ranking based on read abundance , we determined the top 5 most abundant up- and down-regulated hypoxamiR in SLK cells ( Fig 1E ) . As previously described for other cell types [34–36] , of the most abundantly expressed miRNAs , miR-210 was the most significantly up-regulated by hypoxia in SLK cells with a nearly 13-fold increase ( Fig 1E ) . Interestingly , we also observed an 8-fold increase of miR-210 host gene ( MIR210HG ) , the precursor to miR-210 , by RNA-Seq . This is consistent with previous findings showing that at least one functional HIF-1 binding site ( hypoxia-response element , HRE ) is located in the promoter region of miR-210 host gene [37] . A previously described hypoxamiR , miR-146a-5p , was also induced by hypoxia in SLK cells ( 2-fold ) , and has been shown to be a NF-κB-dependent gene which leads to the down-regulation of the inflammatory response [38] . MiRNAs miR-1185-5p , miR-668 , and miR-5001-3p were all up-regulated ( 2 to 4 fold ) . On the other hand , miR-6723-5p , miR-222-5p , miR-4484 , miR-2682-3p , and miR-1257 were the most down-regulated miRNAs ( Fig 1E ) . Our small RNA-Seq study of uninfected SLK cells exposed to hypoxia confirmed previously published findings [37 , 38] and allowed for a global view of hypoxia-regulated miRNAs that will be compared to the effects of KSHV infection . Ingenuity Pathway Analysis ( IPA ) was used to analyze the large number of deregulated genes in hypoxic SLK cells , in order to identify the principal pathways altered by hypoxia ( S2A Fig ) . Among the top 15 pathways most significantly modulated by hypoxia were the Wnt/β-catenin signaling , the neuregulin signaling , prostanoid biosynthesis , the epithelial-mesenchymal transition pathway , and a number of cancer signaling pathways . Also , an integrated approach was used to correlate miRNA and mRNA differential profiles of the RNA-Seq data ( S3A Fig ) . In hypoxic SLK cells , 426 differentially expressed mRNAs were predicted targets of the 52 modulated miRNAs . The IPA microRNA Filter analysis further revealed that 65% ( 278 out of 426 mRNAs ) of these predicted targets inversely correlated with the changes in miRNA expression ( up-regulated miRNA and down-regulated target mRNA , or down-regulated miRNA and up-regulated target mRNA ) . Thus , substantially more than half of the changes occurring in the mRNA targets under hypoxic stress were consistent with predicted changes in the levels of various miRNAs , suggesting that at least some of these mRNA changes were a result of the changes in miRNA expression . This percentage is similar to that ( 73% ) which was obtained when miRNAs and mRNA targets were similarly analyzed in KSHV-infected SLKK cells as compared with uninfected SLK cells [25] . When we restricted the IPA analysis to targets with a high confidence prediction or experimentally observed responses , 35 miRNAs were identified that were paired to 108 mRNA targets ( S3A Fig ) . Specifically , eight miRNAs that were up-regulated by hypoxia were predicted to target 43 mRNAs that were down-regulated , and 27 miRNAs that were down-regulated by hypoxia were predicted to target 65 mRNAs that were up-regulated . These miRNA-mRNA pairs are listed in S3B Fig . Notably , changes in miR-210 were correlated with changes in homeobox A1 ( HOXA1 ) , a well-described target of miR-210 [37] , and CLUH , a regulator of mitochondrial biogenesis [39] , among others . KSHV infection and KSHV-encoded proteins have previously been shown to increase HIF transcription and its responsiveness to hypoxia mimics [16 , 23 , 40 , 41] . With this background , we wanted to explore the relationship between the genes affected by hypoxia and those affected by KSHV infection in SLK cells . To this end , we compared the cellular changes induced by hypoxia in SLK cells to our previous analysis of the differences between KSHV-infected SLKK cells and uninfected SLK cells [25] . SLKK cells are chronically infected by KSHV , and the virus is in a tightly latent state [42] . In Fig 2 , we compare the effects of hypoxia in SLK cells with those of chronic KSHV infection ( SLKK vs . SLK ) at both miRNA and mRNA levels . Of the 210 genes that were significantly up-regulated by hypoxia ( P≤0 . 05 , FC ≤-2 and ≥2 ) , 49 ( 23% ) were also significantly up-regulated by KSHV infection ( Fig 2A ) . Also , of the 309 genes that were down-regulated by hypoxia , 128 ( 41% ) were also down-regulated by KSHV infection . Overall , of the 519 hypoxia-regulated genes in SLK cells , 177 ( 34% ) were similarly regulated by KSHV . The genes up-regulated by both hypoxia and KSHV infection are listed in S1 Table and include genes that are known to play roles in hypoxia and viral pathogenesis; i . e . nuclear enriched abundant transcript 1 ( NEAT1 ) [43 , 44] , integrin alpha 5 ( ITGAV ) [45 , 46] and baculoviral IAP repeat-containing 3 ( BIRC3 ) [47 , 48] . While there was considerable overlap , of the 210 genes that were up-regulated by hypoxia ( P ≤0 . 05 and FC≥2 ) , a substantial number ( 155 ) remained unchanged with KSHV infection and 6 were down-regulated by KSHV ( P ≤0 . 05 and FC ≤-2 ) ( Fig 2A ) . To get a better sense of how these hypoxia-responsive genes were affected by KSHV infection , we plotted the KSHV-induced changes in the genes that met our cut-off criteria for up-regulation by hypoxia ( P ≤0 . 05 and FC ≥2 ) ( Fig 2B ) . As can be seen , while a number of genes were significantly up-regulated under both conditions ( 49 red dots in Fig 2B ) , a substantial number of genes up-regulated by hypoxia were also up-regulated with KSHV infection but just not to the level sufficient to meet the cut-off criteria ( P>0 . 05 ) ( grey dots to the right of the y-axis in Fig 2B ) . Only 6 genes were up-regulated by hypoxia but significantly down-regulated by KSHV infection ( black dots in Fig 2B ) ; interestingly , these included the high-affinity glucose transporter GLUT1 , a HIF-1α-target gene also known as SLC2A1 that plays a role in glycolysis and the Warburg effect [49] . Consistent with this finding , it has recently been reported that GLUT1 is down-regulated by KSHV in rat mesenchymal cells and that this down-regulation promotes cell survival and oncogenic transformation [50] . Looking at the genes affected by KSHV , a relatively smaller proportion were similarly affected by hypoxia; 177 out of 1 , 559 ( 11% ) of the genes deregulated by KSHV were similarly changed by hypoxia ( S4 Fig ) . Overall , these results suggest that KSHV similarly affects many of the genes affected by hypoxia , although the extent of the change with KSHV infection is often less . In addition , the results demonstrate that while the hypoxic response comprises a substantial portion of the changes in cellular gene expression induced by KSHV , KSHV induces a number of other changes in gene expression unrelated to hypoxia . A similar comparison was carried out on the data obtained from the small RNA-Seq differential analysis ( and Taqman analysis in regard to miR-210 ) . There were only ten deregulated miRNAs that overlapped between hypoxia and KSHV infection ( Fig 2C ) . Of these ten miRNAs , two ( miR-4671-3p and miR-210 ) were up-regulated , while the other eight were down-regulated; the most abundant down-regulated miRNAs were miR-548b-3p and miR-1270 . In order to further explore the relationship between the cellular genes affected by KSHV infection and those affected by hypoxia , we analyzed the reported results of two studies of human umbilical vein endothelial cells ( HUVECs ) [51 , 52] . KS spindle cells are thought to be derived from endothelial cells , and HUVECs can thus potentially provide more insight into the pathogenesis of KS than SLK cells . One study used microarray to analyze the genes whose expression was modulated by de novo KSHV infection of HUVECs ( 48 hrs post-infection ) [51] , while the other used RNA sequencing to analyze HUVECs exposed to hypoxia ( 1% O2 ) for 48 hrs [52] . Overall , there were 8 , 863 genes that were analyzed in both studies , of which 350 were dysregulated by de novo KSHV infection and 1 , 137 were dysregulated by hypoxia ( using a 1 . 5 fold cut-off for both parameters ) . Of the 350 genes that were dysregulated by de novo KSHV infection in HUVECs , 49 ( 14% , including 40 up-regulated and 9 down-regulated ) were similarly dysregulated by hypoxia ( S2 Table ) . It is worth pointing out that unlike the SLKK cells , in which KSHV infection was almost completely latent , KSHV de novo infection of HUVECs involves expression of both latent and certain lytic KSHV genes and in addition involves a cellular response to acute viral infection . Even so , the genes affected by the response to hypoxia comprised a considerable proportion of the genes modulated by de novo KSHV infection . Furthermore , we compared the reported expression profile of KS lesions with known hypoxic gene signatures . Cornelissen et al . [53] identified 76 key host genes that are dysregulated in AIDS-KS lesions as compared to normal tissue . Comparing these results to studies of hypoxia , 22 ( 29% ) of these genes are either known HIF-1 targets , have been reported to be similarly dysregulated by hypoxia , or were similarly dysregulated in the SLK/SLKK cell model ( S3 Table ) . Of note , one of the genes up-regulated in hypoxia and AIDS-KS , PKM2 , has been separately shown to regulate the KS angiogenic phenotype by acting as a coactivator of HIF-1 and increasing the levels of HIF-1 angiogenic factors , including VEGF [54] . Overall , there was a consistent pattern of substantial overlap between KSHV infection and hypoxia . We were further interested in studying how hypoxia affected mRNA and miRNA expression in KSHV-infected cells . To this end , KSHV-infected SLKK cells were exposed to 1% O2 for 24hrs , and deep sequencing of mRNA libraries of these hypoxic and control normoxic SLKK cells was performed ( each n = 3 ) . Differential analysis using Cufflinks showed that 268 annotated genes were differentially expressed; 215 genes were up-regulated and 53 were down-regulated ( P≤0 . 05 , FC ≤-2 and ≥2; in blue in Fig 3A ) . In order to confirm that hypoxic induction occurred in the SLKK cells , quantitative real time PCR was again undertaken to determine the expression of VEGF in hypoxia . In the SLKK RNA sequencing , VEGF was up-regulated 3 fold but did not make our significance threshold of P≤0 . 05 ( P = 0 . 07 ) . However , as seen in the quantitative real time PCR data in Fig 3B , there was a ~5 . 5 fold induction of VEGF upon exposure to hypoxia ( P<0 . 001 ) , confirming that the cells were in fact under hypoxic stress . Of interest , N-Myc downstream regulated 1 ( NDRG1 ) was the most significantly up-regulated gene in hypoxic SLKK cells ( Fig 3A ) , with a 4 . 7 log2 fold change increase . These results differed from those found in both normoxic and hypoxic SLK cells , in which NDRG1 was poorly expressed and was thus not significantly changed . NDRG1 expression has been shown to be increased by a variety of environmental stresses , including hypoxia , in either normal or tumor cells , and is involved in caspase activation and apoptosis [55–57] . Also , this is one of the genes that was up-regulated by both hypoxia and de novo KSHV infection in HUVECs ( S2 Table ) . Another gene of interest in SLK and SLKK cells was stanniocalcin-2 ( STC2 ) . This was one of the mRNA most significantly up-regulated by hypoxia in both SLKK cells and SLK cells , with a 3 . 8 and 3 . 9 log2 fold change increase , respectively ( Figs 1A and 3A ) . STC2 promotes cell proliferation , epithelial-mesenchymal transition ( EMT ) and invasiveness in hypoxia; traits that likely promote viral persistence and malignant progression [26 , 58] . After selection based on the highest read count abundance with P≤0 . 01 , we determined the 5 most abundant up- and down-regulated mRNAs in hypoxic SLKK cells and then ordered them based on fold change ( Fig 3C ) . Of these mRNAs , DDIT4 , IGFBP3 , BNIP3 , PGK1 and LGALS1 were up-regulated , while VCAM1 , CCND1 , PLAU , CXCL1 and GDF15 were down-regulated . Some of these genes , BNIP3 , DDIT4 and IGFBP3 for example , were similarly changing in SLK cells under hypoxic stress . It was previously shown that KSHV infection leads to enhanced transcription of HIF and increases the induction of HIF-1 and HIF-2 by a hypoxic mimic [23] . Therefore , we wanted to assess the levels of HIFs across all four conditions ( i . e . normoxic-uninfected , normoxic-infected , hypoxic-uninfected , and hypoxia-infected samples ) in order to determine the degree to which HIF levels might play a role in the observed changes . We were not able to detect HIF-1α in normoxic SLK or SLKK cells , and we could not reliably detect HIF-2α under any conditions . However , consistent with the results of Carroll et al . in endothelial cells infected with KSHV de novo , [23] , HIF-1α was substantially more up-regulated by hypoxia in SLKK cells than SLK cells ( S5 Fig ) , consistent with a relative increase in HIF activity in SLKK cells . Looking at hypoxia-induced changes in miRNA expression profiles in SLKK cells using the same cutoffs ( P≤0 . 05 , FC ≤-2 and ≥2 ) , 79 miRNAs were significantly deregulated by hypoxia ( Fig 3D ) . Of these , miR-210 was again the most significantly up-regulated miRNA ( P = 2 . 0x10-4 and log2 fold change = 3 . 7; Fig 3D and 3E ) , similar to uninfected SLK cells . Of these 79 hypoxamiRs ( miRNAs regulated by hypoxia ) , 72 made the abundance cut-off ( average miR count ≥1 ) , which allowed for an unbiased , unsupervised clustering of hypoxic vs . normoxic SLKK samples ( S1B Fig ) . Of these 72 miRNAs , 51 were down-regulated , while 21 were up-regulated , including miR-210 . We selected the 10 most abundant hypoxamiRs ( Fig 3E ) . Of these , miR-210 , miR-4320 , miR-193a-5p , miR-139-5p and miR-3126-5p were up-regulated while miR-222-5p was the most down-regulated miRNA . We also looked at the effects of hypoxia on KSHV-encoded miRNAs in SLKK cells . Hypoxia was previously shown to affect the expression of LANA mRNA [59] , and since at least some KSHV miRNAs share the promoter region with LANA [60] , it was possible that their expression might also be affected . However , as seen in Fig 3F and S4 and S5 Tables , hypoxia did not change the expression of any KSHV miRNAs as assessed by deep sequencing . Also , comparing hypoxic and normoxic SLKK cells , hypoxia did not affect the overall fraction of viral miRNA ( ~9% ) as compared to the total miRNA read count ( S4 Table ) . In order to validate these findings , Taqman assays were undertaken to assess the expression of the four most abundant miRNAs in SLKK cells ( Fig 3G ) . These included miR-K12-3-5p , miR-K12-4-3p , and miR-K12-8-3p , that are expressed during latency , as well as miR-K12-10a-3p , a KSHV miRNA under the Kaposin promoter that is expressed both during the latent and lytic phases [61] . In addition , we utilized Taqman assays to quantify the expression of four additional KSHV miRNAs that are less abundant in SLKK cells ( miR-K12-1-5p , miR-K12-2-5p , miR-K12-6-3p , and miR-K12-11-3p ) . None of these eight KSHV miRNAs showed a significant change in expression between normoxia and hypoxia in SLKK cells ( Fig 3G ) . These results suggest that while certain cellular miRNAs such as miR-210 are dramatically affected by hypoxia and many are down-regulated by hypoxia , none of the KSHV viral miRNAs are affected . Because miR-210 was consistently increased in response to hypoxia and also increased in response to KSHV infection [25] , a more detailed investigation of its regulation was undertaken . The relative expression of miR-210 was evaluated by Taqman assay under four different conditions: normoxic SLK cells , normoxic SLKK cells , hypoxic SLK cells and hypoxic SLKK cells ( each n = 3 ) . As previously shown by Taqman assay [25] , KSHV infection alone increases miR-210 levels ( ~2 . 5 fold increase ) ( Fig 4A ) . In addition , there was a strong induction of miR-210 in hypoxic SLK cells ( 15–20 fold increase ) , and levels were highest of all in hypoxic SLKK cells . Based on this observation and miR-210’s role in other cancers , we hypothesize that an increase in miR-210 in KSHV-infected cells , either in normoxia or in hypoxia , may help promote cellular changes that favor infected cells and the development of KSHV-induced cancers . ISCU , a mitochondrial iron sulfur scaffold protein , is a miR-210 target that is important in cellular metabolism . In MCF7 cells , it has been shown that miR-210-mediated down-regulation of ISCU contributes to a shift to glycolysis , enhanced cell survival , and an increase in iron uptake required for cell growth [62] . To investigate whether ISCU is in fact down-regulated by miR-210 in our cell model , transfection of miR-210 miRNA mimics and miR-210 miRNA inhibitors was performed in SLKK cells , either in normoxia or in hypoxia . These miRNA inhibitors are small single-stranded RNA molecules that bind to and inhibit specific endogenous miRNA molecules in order to down-regulate miRNA activity . As seen in Fig 4B , hypoxia significantly increased miR-210 expression by ~6–8 fold in SLKK cells . As expected , the transfection of miR-210 mimics or inhibitors substantially increased or decreased miR-210 relative expression , respectively ( Fig 4B ) . Note that after transient transfection with miRNA mimics or antisense inhibitors , the majority of transfected RNA is vesicular and is therefore not accessible for loading into Argonaute [63] . However , the levels of RISC-associated miRNA mimics are close to those of endogenous miRNAs [63] . This data allowed for a further analysis of the effects of miR-210 on expression of its putative target , ISCU . To this end , we evaluated the levels of ISCU by qRT-PCR in SLKK cells under these same conditions ( Fig 4C ) . ISCU was found significantly down-regulated after transfection of miR-210 mimics and after exposure to hypoxia ( NT control , mimic-ctrl , and anti-ctrl ) . Finally , ISCU was significantly up-regulated as compared to hypoxic anti-ctrl when miR-210 inhibitors were transfected , as had been described in other cell systems [62] . ISCU protein expression was similar but not completely identical to that of ISCU mRNA ( Fig 4D ) . In particular , ISCU protein was significantly reduced when miR-210 mimics were transfected , and significantly increased when miR-210 inhibitors where transfected , demonstrating that it is modulated by miR-210 in SLKK cells . However , ISCU protein expression in hypoxia did not decrease as much as its mRNA counterpart . These results provide evidence that ISCU is a target of miR-210 in KSHV-positive SLKK cells and that miR-210-mediated repression of ISCU could function to mediate cell survival . We next compared the cellular response to hypoxia in KSHV-infected versus KSHV-uninfected cells , to explore the effects of latent KSHV infection on the response to hypoxia . When looking at mRNA expression changes due to hypoxia in SLKK and SLK cells , we found that a total of 65 mRNAs , including key hypoxic genes such as BNIP3 , LOX , VCAM-1 , IGFBP3 , PLOD2 , and miR-210 host gene ( MIR210HG ) , responded in a similar manner; either induced ( 51 genes ) or repressed ( 14 genes ) by hypoxia in both cell lines ( Fig 5A and 5B ) . However , a majority of genes whose expression was significantly modified by hypoxia in SLK cells did not meet the cut-off for change in hypoxia-exposed SLKK cells ( Fig 5A ) . To explore this further , we assessed the effects of hypoxia in SLKK cells of genes that were significantly up-regulated by hypoxia in SLK cells ( Fig 5B ) . As seen in Fig 5B ( red dots ) , a significant portion ( 24% ) of the genes that were significantly up-regulated in SLK cells were also up-regulated in SLKK cells . Most of the other genes ( grey dots in Fig 5B ) that were up-regulated by hypoxia in SLK cells trended towards up-regulation in SLKK cells , but just did not make our cut-off values ( P≤0 . 05 , FC ≤-2 or ≥2 ) . Interestingly , there were no genes induced by hypoxia in SLK cells but significantly repressed in SLKK cells . This was in contrast to the previous comparison between the effects of hypoxia and KSHV infection in SLK cells , in which some genes up-regulated by hypoxia were significantly down-regulated by KSHV infection ( Fig 2B ) . Also , there was a strong positive correlation between the hypoxia-induced mRNA changes in SLK and SLKK cells ( Pearson correlation coefficient of r = 0 . 72 ) . Among the 51 mRNAs up-regulated in both SLK and SLKK cells , BNIP3 , a pro-apoptotic gene known for being one of the most hypoxia-responsive genes [64 , 65] , had its mRNA levels increased 4 fold due to hypoxia in both infected and uninfected cell lines . MIR210HG , the host gene for miR-210 , was also one of the responders in common between the two cell lines . There were 14 genes that met our cut-off values for down-regulation in both SLK and SLKK cells ( Fig 5A ) . VCAM-1 was one of the significantly down-regulated mRNAs in both cell lines , with -1 . 4 and -2 . 8 log2 FC in SLK and SLKK cells respectively . This is consistent with the fact that hypoxia has been reported to reduce VCAM-1 expression [66 , 67] . Looking at miRNA expression profiles , only 12 miRNAs were similarly regulated by hypoxia in SLK and SLKK cells ( Fig 5C and 5D ) : 2 up-regulated and 10 down-regulated . As expected based on the RNA-Seq studies of SLK and SLKK individually , fewer miRNAs were up-regulated by hypoxia in both SLK and SLKK cells than were down-regulated . The up-regulated miRNAs comprised only miR-210 , and miR-3074-3p ( Fig 5D ) . Of the 10 down-regulated miRNAs , miR-222-5p had been previously reported down-regulated by hypoxia in MCF7 cells [68] . However , a few down-regulated miRNAs were seen in both SLK and SLKK cells that have not been previously reported down-regulated by hypoxia , including miR-4484 and miR-663b ( Fig 5D ) . Ingenuity Pathway Analysis ( IPA ) was used to identify the key molecular networks that are altered by hypoxia in SLKK cells . The top 15 most significantly altered pathways are illustrated in S2B Fig . They include a number of pathways known to be altered by hypoxia such as a number of cancer signaling pathways , glycolysis , Wnt/β-catenin signaling , IL-8 signaling , and retinoic acid receptor ( RAR ) activation [69–71] . Merging the small RNA-Seq with mRNA-Seq differential analyses , IPA paired 53 miRNAs with 176 mRNA targets , and a remarkable 82% ( 145 mRNAs ) of these targets were concordant under hypoxia ( either up-regulated miRNA and down-regulated target , or down-regulated miRNA and up-regulated target; Fig 6A ) , again suggesting that the miRNA changes may be in many cases driving the changes in the target genes . Looking only at the miRNAs and targets that were either paired with a high level of confidence or experimentally observed in the literature ( confidence filter ) , 32 hypoxamiRs remained and were paired with 67 targets ( Fig 6A ) . Fig 6B lists the seven up-regulated miRNAs and their 13 down-regulated targets , as well as the 25 down-regulated miRNAs and their 54 targets . Of note , miR-224-5p is up-regulated in hypoxia and is predicted to down-regulate pentraxin 3 ( PTX3 ) , also known as TNF-inducible gene 14 ( log2FC = -1 . 7 ) . PTX3 is known to play a dual role in both protecting cells against pathogens and controlling autoimmunity . Interestingly , three of the down-regulated miRNAs ( miR-3064-5p , miR-3175 , and miR-3201 ) and some of their respective up-regulated targets in SLKK cells exposed to hypoxia were similarly deregulated in hypoxic SLK cells ( S3B Fig ) . Viruses , and especially those that cause chronic infections , are attuned to and respond to changes in their target cells . At the same time , viral infection leads to a number of changes in these target cells , some mediated by the virus and others as part of the host response to infection . Previous studies have shown that hypoxia and HIFs can affect KSHV biology and KSHV-induced tumor formation [6 , 7 , 59] . Moreover , KSHV infection and LANA , a KSHV-encoded latent gene , have been shown to induce the up-regulation of HIF [16] and at least one HIF-responsive gene ( VEGFR1 ) in newly infected endothelial cells [23] . Also , several HIF-responsive genes related to angiogenesis or metabolism have been shown to be elevated in immortalized HUVECs chronically infected with KSHV [70] . However , these data do not provide a complete picture of the cellular response to KSHV infection and the role that hypoxia plays . To further investigate these relationships , we used RNA-Seq analysis to assess the changes induced by hypoxia in SLK cells and compare them with the changes we had previously found in chronically KSHV-infected SLK cells ( the SLKK line ) [25]; this paper represents the first global survey of these effects . We found that there was a substantial overlap in the alteration in gene expression induced by KSHV infection and hypoxia . In particular , more than a third ( 34% ) of the genes seen differentially expressed under hypoxia were similarly up- or down-regulated by KSHV latent infection . Also , a majority of the 155 genes that made our cut-off for up-regulation by hypoxia but not for KSHV ( Fig 2A ) demonstrated a trend towards up-regulation in KSHV infection ( Fig 2B ) . By contrast , only 5 genes up-regulated by hypoxia showed a down-regulation in KSHV-infected cells . Looking at this from the perspective of KSHV-modulated genes , 11% of these genes are genes modulated in the same way by hypoxia . Overall , these results provide evidence that KSHV commandeers the cell response to hypoxia and that hypoxia-related changes in gene expression comprise a substantial portion of the response to KSHV infection . While levels of HIF-1 were greater in hypoxia-exposed SLKK cells than hypoxia-exposed SLK cells , it was noteworthy that a hypoxic signature was observed in normoxic SLK cells in spite of the fact that we could not detect elevated levels of HIF-1 protein . Further studies would be required to determine to which degree the overlap between the hypoxic response and the KSHV infection is due to KSHV-induced up-regulation of HIF , especially under normoxic conditions . Cells respond to hypoxia through a number of changes , primarily mediated by HIFs . The systems and pathways that are induced by hypoxia include glycolysis and glucose uptake , growth factor signaling , immortalization , resistance to apoptosis , angiogenesis , and genetic instability [72 , 73] . It is notable that KSHV has such an effect on hypoxia-related genes , and raises the question as to what evolutionary advantage it provides the virus . Looking at specific genes that are up-regulated by both hypoxia and KSHV infection in our study , a number have activities that can facilitate chronic viral infection or thwart innate anti-viral mechanisms . For instance , baculoviral IAP repeat containing 3 ( BIRC3 ) , also known as cellular inhibitor of apoptosis 2 ( cIAP-2 ) , is a hypoxia-responsive gene that can confer resistance to apoptosis by interfering with caspases [48] . Here , BIRC3 was induced in SLK cells by both hypoxia ( log2 fold change of 1 . 7 or 3 . 2 fold ) and by KSHV infection ( 14 . 9 fold change ) ( S1 Table ) . Interestingly , KSHV protein K15 also up-regulates BIRC3 [47] , which suggest that the up-regulation in KSHV infection can occur through both HIF-dependent and independent mechanisms . By interfering with apoptosis , BIRC3 can inhibit the ability of KSHV-infected cells to destroy themselves ( and the infecting virus ) [74–76] . Another gene that is up-regulated by both hypoxia and KSHV infection is NEAT1 , a host long non-coding RNA involved in nuclear paraspeckle formation . There is evidence that NEAT1 plays an important role in regulating gene expression [77 , 78] . NEAT1 has also been found to increase the survival of cancer cells [43] , and may thus benefit KSHV infection by preventing the death of KSHV-infected cells . Up-regulation of genes such as BIRC3 or NEAT1 would presumably be beneficial for most viruses , and in fact several other chronic viruses , including hepatitis B virus , human papillomavirus , Epstein Barr virus , and human T cell lymphotropic virus ( HTLV-1 ) have also been reported to activate HIF and certain HIF-responsive genes . At the same time , the modulation of HIF-responsive genes by KSHV is particularly robust . Endothelial cells are important target cells for KSHV infection , and it is possible that the virus evolved to up-regulate hypoxia-responsive genes in part because several of these genes , including STC2 and IGFBP3 ( see S1 Table ) , promote angiogenesis and the growth of endothelial cells [26 , 79 , 80] . An unintended consequence of the activation of these and other HIF-responsive genes may be tumorigenesis leading to the development of Kaposi sarcoma and other KSHV-induced tumors . As described in our previous paper [25] , we used SLK and SLKK cells in this study because of the extremely tight control of KSHV latency in these cells and because they provide the ability to compare the effects of hypoxia and KSHV infection on KSHV infected and uninfected cells . SLK cells were originally thought to be derived from a KS biopsy and have a number of endothelial-like features [81]; however , this line was subsequently found to be indistinguishable from Caki-1 , a clear-cell renal carcinoma line [81] . Nonetheless , this line remains a valuable tool to study the effects of latent KSHV infection . To further explore whether there was substantial overlap in the cellular response to KSHV infection and to hypoxia in systems that were biologically more similar to KS , we used published datasets to analyze gene changes in HUVECs and in KS lesions , and found evidence in these systems as well . In the de novo infected HUVECs system , it is noteworthy that a strong hypoxic signature was seen in the cellular response in spite of the fact that lytic KSHV genes and an acute cellular response to viral infection are present . However , there may be limitations to that system in that HUVECs were infected de novo for 48hrs , which corresponds to an early lytic activation . In that phase , KSHV is establishing initial infection and cells are reacting to this infection [82] . Therefore , it is not surprising that cellular genes expressed during this short-lived and transient phase might be different from those that will be differentially expressed during latency and that hypoxia-related genes comprise a smaller percentage of the genes that are differentially expressed . Additionally , compiled results from a number of studies show that a substantial proportion of genes whose expression is modulated in KS lesions are genes that are responsive to hypoxia . Taken together , the prominence of a hypoxic gene activation in KSHV infection and the potential of HIFs to activate KSHV genes may provide a rationale for targeting hypoxic pathways as a potential therapeutic strategy for KSHV-related diseases . Indeed , HIF-1 dysregulation is known to fuel both angiogenesis and tumor metabolism in KS [54] . Furthermore , HIF-1 suppression has been found to lower viral interleukin-6 levels , lytic replication of KSHV , and proliferation of PELs [83] , supporting that HIF-1 is indeed an important factor in the maintenance of KSHV infection and the survival of PELs . A somewhat different picture emerged when we examined the miRNA changes induced by KSHV infection and hypoxia . The majority of the deregulated miRNAs in either hypoxia or in KSHV infection were repressed ( 88% by hypoxia and 73% by KSHV infection ) [25] . However , there was relatively little overlap in the miRNA expression profiles affected by hypoxia and those affected by KSHV infection . A notable exception to this was miR-210 , which was significantly up-regulated by both hypoxia and KSHV latent infection; this was confirmed using independent Taqman validation assays . miR-210 is directly up-regulated by the binding of HIF-1α to an HRE in its promoter region and has previously been shown to be the main miRNA induced by hypoxia [33 , 35] [84] . While miR-210’s effects on KSHV were not investigated here , miR-210 has been extensively studied and is known to be an important factor in regulating the immune response , glycolysis , and tumorigenesis , all areas which may impact the biology of the virus [85 , 86] . Additionally , miR-210 has been shown to be up-regulated by EBV infection [87] and HIV infection [88] , and miR-210 helps maintain virion production in HBV infection [89] . An important target of miR-210 is ISCU , a mitochondrial iron sulfur scaffold protein involved in cellular metabolism [62] , and as seen here , changes in ISCU are consistent with its regulation by miR-210 in KSHV-infected cells ( Fig 4C ) . Of note , inhibition of ISCU by miR-210 in MCF7 cells has previously been shown to lead to a shift to increased glycolysis , an enhanced cell survival rate , and an increased iron uptake required for cell growth [62] . Such changes could help promote the Warburg effect that has been described in KSHV-infected cells and favors their cell growth and survival . The exact role of miR-210 in KSHV infection and disease pathogenesis remain to be established in future studies . By contrast to miR-210 , the majority of miRNAs affected by hypoxia or by latent KSHV infection were down-regulated . The mechanism leading to these changes is not clear , but could be related to changes in post-translational miRNA biogenesis . In this regard , it has been reported that the key proteins of the miRNA biogenesis , Dicer and Drosha , can be down-regulated in hypoxia , hence reducing miRNA processing [90] . However in the present experiments , neither Dicer nor Drosha mRNAs were found significantly down-regulated by hypoxia in the RNA-Seq data; nevertheless , the levels of these miRNA processing proteins might still be affected by post-transcriptional alterations . Additional studies will be needed to understand this general miRNA down-regulation by hypoxia . Certain cells newly infected by KSHV , such as endothelial cells , have been reported to increase glycolysis , also known as the Warburg effect , and a number of genes involved in glycolysis are up-regulated by HIF-1 ( but not HIF-2 ) [32 , 91] . While the increase in miR-210 levels could contribute to the Warburg effect , it was noteworthy that a number of genes involved in glycolysis were up-regulated by hypoxia ( including GLUT1 , hexokinase 2 , phosphoglycerate kinase , pyruvate dehydrogenase kinase-1 , and lactate dehydrogenase-5 ) , but were not by KSHV infection . In fact , GLUT1 , which transports glucose into the cell , was one of the 5 down-regulated genes in SLKK cells ( Fig 2B ) . Interestingly , while GLUT1 is up-regulated in endothelial cells acutely infected by KSHV [54] , it has recently been reported to be down-regulated in acutely infected rat mesenchymal cells [50] . It should be noted that SLKK cells are chronically infected and that KSHV is tightly latent in these cells; additional studies will be needed to clarify the effects of KSHV infection on genes involved in glycolysis . KS preferentially arises in the feet , which have relatively low oxygen tension , and PEL usually arises in pleural spaces , which have no blood vessels and are also hypoxic [9 , 10] . Previous studies by our group and others have shown that hypoxia and HIFs can activate a number of KSHV genes , including RTA , LANA , and ORFS 34–37 . Thus , the combination of KSHV infection and hypoxia appears to be important in the pathogenesis of KSHV-associated tumors . To explore these relationships further , we analyzed the effects of hypoxia on KSHV-infected cells . There was substantial overlap in the affected mRNAs in SLK and SLKK cells exposed to hypoxia , especially considering the genes that were up-regulated . However , in many cases , genes that were moderately up-regulated by hypoxia in SLK cells fell below our significance threshold for up-regulation in hypoxia-exposed SLKK cells ( Fig 5B ) , and none of the genes that were up-regulated by hypoxia in SLK cells were down-regulated by hypoxia in SLKK cells ( Fig 5A ) . In our RNA-Seq analysis . we did not directly compare the overall changes in mRNA and miRNA in uninfected SLK cells in normoxia to SLKK cells in hypoxia . Such a comparison was made for miR-210 ( Fig 4A ) and revealed that the effects were cumulative . Also , this analysis showed that the induction of miR-210 by hypoxia in SLKK cells was less than that in SLK cells , largely because of the baseline elevation in the SLKK cells . This suggests that the reason that less genes overall were deregulated by hypoxia in SLKK cells than in SLK cells ( 519 vs . 268 ) may be because of a similar shift in baseline gene levels by KSHV . KSHV encodes its own miRNAs , which are formed from 12 precursor miRNAs . These precursor miRNAs are on mRNA transcripts downstream of LANA , and share promoter regions with Kaposin . LANA , Kaposin , and most of the KSHV miRNAs are produced during latency , and some of the miRNAs ( e . g . miR-K12-10 and -12 ) are up-regulated during chemical induction of KSHV lytic infection [92] . Since LANA and the lytic switch gene RTA can be up-regulated by hypoxia and HIF [59] , we thought it plausible that expression of at least some of the KSHV-encoded miRNAs might be similarly induced by exposure to a low-oxygen environment . However , the levels of these mature miRNAs were remarkably unchanged by hypoxia ( Fig 3F and 3G ) . Some of the KSHV miRNAs are extremely abundant and the combination of a longer half-life of miRNAs compared to the mRNAs [93] might explain this observation . Another factor might be the length of hypoxia , which might be much longer in the context of KSHV infection in humans , compared to the experimental hypoxia treatment in our assays ( 24hrs ) . We did not specifically analyze the levels of the miRNA precursors , and it is possible that an increase in these precursors was counterbalanced by a suppression of miRNA maturation , which also contributed to the overall decrease in cellular miRNAs ( Fig 3D ) . Additional studies will be needed to clarify this; however , the results indicate that a modulation of KSHV-encoded miRNAs does not have a substantial effect on the response of KSHV-infected cells to hypoxia . In summary , this global gene survey demonstrates that there is a substantial overlap between the genes affected by hypoxia and by KSHV infection , and shows that induction of a hypoxic gene response is a substantial component of the modulation of cellular genes by KSHV infection . However , by contrast to mRNA expression , there was relatively little overlap in the modulation of miRNAs by KSHV infection and hypoxia , a notable exception being an up-regulation of HIF target miR-210 , which has been previously shown to be important in the angiogenic response and in tumor formation . This study provides a more thorough understanding of the role of hypoxia and the interplay between KSHV infection and HIF/hypoxia , and lays the groundwork for further analyses of these interactions . Human KS-derived SLK and SLKK cells ( also known as SLK+rKSHV . 219 ) [51 , 81 , 94 , 95] were a gift from Dr . Don Ganem ( UCSF , CA ) . They were expanded on receipt , frozen in liquid nitrogen , and stored in a cryogenic tank until used in the experiments described hereafter . Cells were thawed and maintained in Dulbecco’s Modified Eagle medium supplemented with 10% v/v fetal bovine serum ( Sigma-Aldrich , St Louis , MO ) and 1% Penicillin/streptomycin/glutamine solution ( Gibco , Carlsbad , CA ) . Additionally , KSHV-positive SLKK cells were periodically grown under selection with 10 μg/mL puromycin to maintain the viral episome . Not counting the time during which cells were frozen , SLK cells had been kept in culture for less than 3 months before being used for RNA isolation ( RNA-Seq ) , transfection assays , or de novo KSHV infections . SLK cells infected with recombinant rKSHV . 219 ( SLKK cells ) were originally selected over time using puromycin in order to reach a high latent viral expression . Therefore , again not counting the time that they were frozen , the SLKK cells were kept in culture for a longer period of time ( 6 months to a year ) before being used for RNA isolation ( RNA-Seq ) and transfection assays . All experiments were done in triplicate independent cell cultures maintained at 37°C in humidified 5% CO2 . Exposure of cell cultures to 1% oxygen was undertaken in an InVivo2 Hypoxia Work Station ( Ruskinn Technology , UK ) . This was undertaken in parallel with cells maintained in normoxic conditions ( 21% O2 ) . All cells were harvested for protein and RNA 24hrs post-treatment . All experiments were done at least in triplicate from independent cell cultures . SLK and SLKK cells were seeded at a concentration of 6 to 9x105 cells in a10cm dish in DMEM medium with 10% FBS containing no antibiotics; this allowed cells to achieve 50% confluence the following day . The culture medium was then removed and the cells were transfected with the miRNA mimics ( i . e . miR-210 ) , miRNA inhibitors ( i . e . anti-210 ) or negative controls ( Dharmacon , Lafayette , CO , USA ) at 10nM final concentration , using Dharmafect 1 ( Dharmacon ) . Dharmafect 1 is a potent transfection reagent that allows the negatively charged membrane to interact with the liposome/nucleic acid complex . The miRNA transfection experiments were performed in a 2 . 4mL Opti-MEM/Dharmafect mixture and 9 . 6mL DMEM medium ( Invitrogen ) . The cells were incubated in normoxia or hypoxia ( 1% O2 ) for 48hrs at 37°C/5% CO2 before performing RNA isolation and nuclear/cytoplasmic extraction . Total RNA was extracted from cells using miRVana miRNA isolation kit according to manufacturer’s instructions ( Ambion , Life Technologies , Carlsbad , CA ) . RNA abundance and integrity were determined after isolation using a Nanodrop-ND-1000 spectrophotometer ( Thermo Fisher Scientific , Waltham , Massachusetts , USA ) and an Agilent 2100 Bioanalyzer ( Agilent Technologies , Santa Clara , CA ) , respectively . Only samples of total RNA with an RNA integrity number ( RIN ) >9 were further used for RNA-sequencing and small RNA-sequencing . All samples were stored at -80°C . For miRNA expression , stem-loop qPCR was performed using TaqMan Universal master mix ( Applied Biosystems ) and the following microRNA assays: 000512 for miR-210-3p , 197204 for miR-K12-1-5p , 197192 for miR-K12-2-5p , 008316 for miR-K12-3-5p , 197240 for miR-K12-4-3p , 008459 for miR-K12-6-3p , 241994 for miR-K12-8-3p , 008504 for miR-K12-10a-3p , and 008562 for miR-K12-11-3p . Reference RNA RNU43 ( assay 001095 ) was used as endogenous control to normalize expression . Thermal cycling conditions included an enzyme activation step ( 95°C for 10 min ) and 40 cycles of amplification at 95°C for 15s followed by 60°C for 1min . For gene expression , cDNA synthesis was performed on total RNA using the Superscript II reverse transcriptase kit , 0 . 5mM dNTP set and 50 ng/μL random hexamer ( Invitrogen ) . qPCR was performed using the FastStart Universal SYBR Green/ROX master mix ( Roche Applied Science , Mannheim , Germany ) . 18S transcript was used as endogenous control to normalize expression . Thermal conditions included an enzyme activation step ( 95°C for 10min ) , 40 cycles of amplification at 95°C for 15 sec and 60°C for 1 min , and melting curve analysis following instrument standard instructions . The following primer sets were used: 5’-CCTTGCTGCTCTACCTCCAC -3’ and 5’- AGCTGCGCTGATAGACATCC-3’ for VEGF , and 5’-GCCCGAAGCGTTTACTTTGA-3’ and 5’-TCCATTATTCCTAGCTGCGGTATC-3’for 18S . All qPCR reactions were performed on an ABI 7300 Real Time PCR instrument ( Applied Biosystems ) . Each experiment was performed in triplicate . Change in miRNA or mRNA expression was determined based on the delta Ct method [96] . Error bars in relative expression plots represent the standard error of the mean and P-values were calculated using two-tailed Student t-test unless otherwise stated . * , , and * indicate P ≤0 . 05 , P ≤0 . 01 , and P ≤0 . 001 , respectively . Cells were lysed and cytoplasmic and nuclear extracts were prepared using the NE-PER extraction kit with 1X halt-protease inhibitors cocktail ( both Pierce , Waltham , MA ) and EDTA . Protein extracts were stored at -80°C . 20μg of protein were mixed with warm 2X lithium dodecyl sulfate ( LDS ) such that the LDS to sample ratio was 1:4 ( v/v ) . Sample were denatured for 5 minutes at 95°C and subjected to SDS-PAGE ( 4–12% NuPAGE Bis-Tris ) with 1X MOPS running buffer diluted with dH2O ( all Invitrogen ) . The proteins were transferred onto a nitrocellulose membrane by an iBlot apparatus for 8 minutes ( all Invitrogen ) . The membrane was blocked with 5% w/v nonfat dry milk in 1X TBST ( 10nM Tris-HCl , pH 8 . 0 , 150nM NaCl , and 0 . 05% Tween 20 ) . The blot was incubated with mouse antibodies to HIF-1α ( 1:500 , BD Biosciences , San Jose , CA ) or ISCU ( 1:800 , ab180532 , Abcam ) overnight and goat anti-mouse IRDye 800CW antibody ( Licor Biosciences , Lincoln , NE ) at a 1:10 , 000 dilution for 30 minutes; being washed twice with washing buffer in between incubations . Scanning was performed using the Li-Cor Odyssey CLx imaging system coupled with the Image Studio software . Images were later adjusted for contrast and intensity using PowerPoint 2011 version 14 . 6 . 3 for Macintosh ( Microsoft Co , Redmond , WA ) . Small RNA libraries were constructed as previously described [68] using the TruSeq RNA Sample Prep kit ( Illumina inc . , San Diego , CA ) . Briefly , after running 2 μg of total RNA in a 15% urea-TBE gel ( Invitrogen , Life technologies , Carlsbad , CA ) for 1 hr at 200V , the 20 to 30 nucleotide RNA fraction was excised and eluted in 0 . 3 M NaCl . Following separation of the elute from the gel debris using a Spin-X-column ( Thermo Fisher Scientific ) , the small RNA samples were precipitated in 100% ethanol and 1mg/mL glycogen , incubated at -80°C for 30 min , centrifuged at 14 , 000rpm for 25 min , washed with 75% ethanol , air dried and resuspended in RNAse-free water . Illumina TruSeq libraries were then prepared according to the manufacturer’s protocol and the final RNA library concentration was measured by a Qubit 2 . 0 fluorometer using Qubit dcDNA HS Assay kit ( Life Technologies ) . We verified the size of the products contained in the libraries using a high sensitivity DNA chip and an Agilent 2100 Bioanalyzer ( Agilent Technologies ) . Finally , a total of six miRNA libraries ( three SLK and three SLKK samples ) were sequenced using the Illumina HiSeq platform . Total RNA of samples used for small RNA-Sequencing was treated with Turbo DNA-free DNase I and Dynabeads ( both from Ambion , Life Technology ) to deplete samples from the residual DNA and to isolate the polyadenylated mRNA transcriptome , respectively . PolyA+ RNA libraries were then prepared with the ScriptSeq v2 RNA-Seq kit ( Epicentre , Madison , WI ) . The final concentration and size distribution of the RNA libraries were measured by using a Nanodrop-ND-1000 spectrophotometer ( Thermo Fisher Scientific ) , and by running a DNA 100 chip on an Agilent 2100 Bioanalyzer ( Agilent Technologies ) . Finally , a total of six polyadenylated mRNA libraries ( three SLK and three SLKK samples ) were sequenced using the Illumina HiSeq platform . For small RNA sequencing , reads were aligned against known miRNAs from miRbase ( version 19 . 0 ) using the software package SHRIMP [97] ( http://compbio . cs . toronto . edu/shrimp/README ) with miRNA specific settings ( -o 1 -n 2 -r 30% -h 50%—local -Q—qv-offset 32—sam ) . To process paired-end sequencing , reads were aligned separately covering the mature miRNA both on the forward and reverse read and the obtained number of matches were averaged . The match counts were normalized and tested for differential expression using the edgeR package [98] with default settings . For mRNA sequencing , adapter sequences were trimmed using the Fastx toolkit . Reads were aligned against two target genomes , human ( hg19 ) and KSHV ( Genbank accession number NC_009333 . 1 ) using TopHat [99] to generate spliced alignments . Transcripts were assembled using Cufflinks and Cuffdiff [100] in order to reveal differentially expressed genes . To be more conservative and eliminate infinite fold differences in cases where there was a nil denominator , fold changes were calculated adding 0 . 01 read to both the numerator and denominator . Significant mRNA fold change was determined by an adjusted P-value lower than 0 . 05 based on the Benjamini and Hochberg multiple testing correction . In order to visualize miRNA-Seq profiles , sequencing reads were converted to bedGraph format using BEDtools [101] . The output files were then uploaded and displayed using the UCSC Genome Browser ( http://www . genome . ucsc . edu ) . R ( http://www . R-project . org ) suite software was used for statistical analyses , heat maps and scatter plots . Pathway analysis and inverse correlations between expression levels of differentially expressed miRNAs and their respective target mRNAs were analyzed using Ingenuity Pathway Analysis ( IPA , Ingenuity Systems , Redwood City , CA; www . ingenuity . com ) . This in silico analysis software reveals enrichment for molecular networks and signaling pathways . Moreover , we generated medium to high confidence miRNA target predictions and as well as experimentally observed miRNA-mRNA interactions using the IPA tool called “MicroRNA Target Filter” that integrates multiple target prediction algorithms such as TargetScan , TarBase , miRecords and the Ingenuity Knowledge Base . This allowed for the integration of miRNAs with mRNA targets , predicted and validated , that were both differentially expressed in SLKK vs . SLK cells . Opposite expression pairing between miRNA and mRNA levels was implemented to further refine the analysis . Further filtering options were applied such as the confidence parameter . We compared two publicly available datasets from published HUVECs studies [51 , 52] . One study used microarray to analyze the genes modulated by de novo KSHV infection ( 48 hrs post-infection ) [51] , while the other used RNA sequencing to analyze HUVECs exposed to hypoxia ( 1% O2 ) for 48 hrs [52] . We parsed the data based on expression fold change ( FC>1 . 5 ) , and an additional P-value cut-off of 5x10-5 was used on the RNA-seq study [52] Overall , there were 8 , 863 genes that were detected in both studies , of which 350 and 1 , 137 were dysregulated by de novo KSHV infection and hypoxia , respectively . Additionally , we compared 76 genes known to be dysregulated in AIDS-KS lesions [53] with our results in the SLK/SLKK model and other published hypoxia studies [32 , 102–117] . There were 22 out 76 genes similarly regulated in AIDS-KS and hypoxia . Raw miRNA and mRNA data are available on the NCBI Gene Expression Omnibus ( GEO ) database under the series accession identifier GSE79032 .	Kaposi sarcoma-associated herpesvirus ( KSHV ) is an oncogenic herpesvirus known to cause several tumors and hyperproliferative disorders . While there has been reports of KSHV activating and increasing hypoxia-inducible factors ( HIFs ) , this is the first report investigating and establishing the extent to which KSHV has evolved to reproduce the effects of hypoxia . We demonstrate that the cellular changes in gene expression induced by KSHV infection include many of the changes induced by hypoxia . This has substantial implications for the biology of KSHV and the pathogenesis of KSHV-associated cancers . To achieve this , we used mRNA-sequencing and small RNA-sequencing in combination with bioinformatics analysis , and orthogonal assays such as qRT-PCR and Taqman assays to determine the effects of hypoxia on miRNA and mRNA expression . We showed that not only was there a 34% overlap between the hypoxic response and KSHV infection , but also that miRNA miR-210 , a HIF-target known to have anti-apoptotic , angiogenic , and oncogenic properties , was independently and additively increased by KSHV infection and hypoxia . Furthermore , we explored the effects of hypoxia on KSHV miRNAs and consistently observed that none of the KSHV miRNAs are affected by oxygen deprivation . These studies suggest that KSHV harnesses a part of the hypoxic cellular response and that a substantial portion of hypoxia-induced changes in cellular gene expression are induced by KSHV infection . Previous studies have shown that hypoxia and HIFs activate KSHV-encoded genes , including several involved in tumor formation . These findings suggest that targeting HIFs or hypoxia pathways could block this positive feedback loop between KSHV and hypoxia and thus might be a useful strategy to treat KSHV-related tumors or other diseases . We believe these are important findings with broad implication for the understanding of the biology of KSHV and other oncoviruses and the pathogenesis of KSHV-induced tumors . As such , it should be of interest to the broad community of investigators and clinicians interested in the biology of oncoviruses , virus-induced cellular changes , and the pathogenesis and therapy of virus-induced tumors .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "sequencing", "techniques", "medicine", "and", "health", "sciences", "pathology", "and", "laboratory", "medicine", "gene", "regulation", "pathogens", "rna", "extraction", "microbiology", "pulmonology", "viruses", "micrornas", "dna", "viruses", "hypoxia", "molecular", "b...	2017	RNA Sequencing Reveals that Kaposi Sarcoma-Associated Herpesvirus Infection Mimics Hypoxia Gene Expression Signature
Coincidence detector neurons transmit timing information by responding preferentially to concurrent synaptic inputs . Principal cells of the medial superior olive ( MSO ) in the mammalian auditory brainstem are superb coincidence detectors . They encode sound source location with high temporal precision , distinguishing submillisecond timing differences among inputs . We investigate computationally how dynamic coupling between the input region ( soma and dendrite ) and the spike-generating output region ( axon and axon initial segment ) can enhance coincidence detection in MSO neurons . To do this , we formulate a two-compartment neuron model and characterize extensively coincidence detection sensitivity throughout a parameter space of coupling configurations . We focus on the interaction between coupling configuration and two currents that provide dynamic , voltage-gated , negative feedback in subthreshold voltage range: sodium current with rapid inactivation and low-threshold potassium current , IKLT . These currents reduce synaptic summation and can prevent spike generation unless inputs arrive with near simultaneity . We show that strong soma-to-axon coupling promotes the negative feedback effects of sodium inactivation and is , therefore , advantageous for coincidence detection . Furthermore , the feedforward combination of strong soma-to-axon coupling and weak axon-to-soma coupling enables spikes to be generated efficiently ( few sodium channels needed ) and with rapid recovery that enhances high-frequency coincidence detection . These observations detail the functional benefit of the strongly feedforward configuration that has been observed in physiological studies of MSO neurons . We find that IKLT further enhances coincidence detection sensitivity , but with effects that depend on coupling configuration . For instance , in models with weak soma-to-axon and weak axon-to-soma coupling , IKLT in the axon enhances coincidence detection more effectively than IKLT in the soma . By using a minimal model of soma-to-axon coupling , we connect structure , dynamics , and computation . Although we consider the particular case of MSO coincidence detectors , our method for creating and exploring a parameter space of two-compartment models can be applied to other neurons . Neurons that spike selectively to multiple subthreshold inputs that arrive within brief time windows are coincidence detectors . Coincidence detection is a fundamental neural computation that allows the brain to extract information from the temporal patterns of synaptic inputs . In the cortex , neurons have biophysical specializations compatible with coincidence detection [1–7] , but some have questioned whether temporally-precise computations are possible in cortex due to highly variable neural activity therein [8 , 9] . In the early auditory pathway , the existence of coincidence detector neurons and their functional importance are widely valued [10–12] . Principal cells of the medial superior olive ( MSO ) in the mammalian auditory brainstem are a canonical example: they receive inputs originating from both ears [13 , 14] and are sensitive to microsecond-scale differences in the timing of arriving inputs [15–18] . These coincidence detector neurons are critical for sound-source localization [19] and likely play important roles in other aspects of binaural ( two-eared ) hearing such as sensitivity to interaural correlation [20 , 21] Temporally-precise neural coincidence detection requires specialized neural dynamics and circuitry . Coincidence detector neurons should have fast membrane dynamics with time-scales of integration shorter than the intervals between volleys of synaptic inputs [2 , 22] . Inputs to coincidence detectors should also be brief and well-timed to precisely convey timing information . The requirements for effective coincidence detection in the auditory system are exceptionally stringent because auditory neurons must process inputs with temporal information at kilohertz-scale and higher [23–25] . Auditory brainstem circuitry is equipped with a suite of specializations to promote coincidence detection [26] . Afferent inputs to MSO cells are reliable and temporally-precise [27 , 28] , dendritic processing in MSO further enhances coincidence detection [25 , 29–31] , and voltage-gated currents that are partially active near resting voltage make MSO cells extremely fast and precise processors [25 , 32] . Voltage-gated currents are also sources of dynamic negative feedback that contribute to the remarkable coincidence detection capabilities of these neurons . In MSO neurons , activation of low-threshold potassium ( KLT ) current and inactivation of sodium current are two identified sources of dynamic negative feedback [33–36] . In response to , say , a pair of brief excitatory inputs , these feedback mechanisms will transiently raise the spiking threshold after the first input , and thereby reduce the chance that the neuron will spike in response to the second input unless the inputs arrive nearly synchronously ( within a coincidence detection time window ) . We investigate the extent to which a structural specialization—namely , the coupling between the input region ( soma and dendrite ) and the output region ( axon and axon initial segment ) —can further optimize coincidence detection sensitivity . To do this , we develop a two-compartment neuron model as a minimal description of input-output coupling and systematically explore the effects of coupling configuration on coincidence detection sensitivity . The two-compartment formulation is motivated by observations that spike generation in MSO likely occurs in the axon or axon initial segment [37 , 38] . Furthermore , sodium channels in the soma are inactivated near resting potentials [39] and spikes are small and graded in the soma [37] , suggesting the soma does not participate in spike generation . Indeed , an absence ( or small amount ) of sodium current in the soma appears as a general design principle for temporally-precise auditory neurons [40] . Studies of coincidence detector cells in the avian auditory brainstem have shown that a passive soma can enhance coincidence detection [41] and that the distribution of voltage-gated channels in the axon initial segment undergoes activity-dependent modulation [42 , 43] to improve coincidence detection , perhaps in an optimal manner [44] . A two-compartment formulation neglects the helpful contributions of dendritic processing to coincidence detection , but the role of dendrites has been considered in detail in previous studies [25 , 29–31] . In this study , we systematically relate forward coupling ( soma-to-axon ) and backward coupling ( axon-to-soma ) strengths to model parameters . We explore this parameter space and find the coupling configurations that enhance coincidence detection sensitivity . Specifically , we identify strong soma-to-axon coupling as a natural configuration for neural coincidence detection because it engages sodium inactivation as a mechanism that transiently increases spike threshold on the time-scale of synaptic inputs and prevents firing to inputs that do not arrive concurrently . Moreover , the combination of strong soma-to-axon and weak axon-to-soma coupling generate spikes more efficiently ( requires fewer sodium channels ) and with shorter refractory periods than other models . This feedforward configuration enhances high-frequency coincidence detection and represents distinct advantages over one-compartment point neuron models that cannot exhibit this asymmetric coupling configuration . We observe that KLT current provides additional benefits for coincidence detection sensitivity , but these benefits depend on coupling configuration and where KLT current is located in the two-compartment structure . For instance , coincidence detection sensitivity in neurons with weak soma-to-axon coupling can be substantially improved if KLT current is co-localized with spike-generating currents . We select passive properties to match known physiological characteristics of MSO neurons , so our observations apply directly to those canonical coincidence detectors . Nonetheless , our method for systematically exploring the parameter space of coupling configurations can be applied to study the relationships between structure , dynamics , and computation in other neurons that are well-described by a two-compartment idealization [24 , 45–50] . In particular , a useful aspect of our work is that we show how to explicitly construct two-compartment models that satisfy the constraint of having nearly identical passive dynamics in the input compartment . After a description of our model and method , we present our results as follows . First , we test coincidence detection sensitivity of the two-compartment model using simulated synaptic inputs . We then use simpler inputs ( direct current ) and phase plane analysis to explain the effects of coupling configuration on coincidence detection sensitivity . These main results concern interactions between cell structure and spiking dynamics . It is known , though , that KLT current is prominent in MSO neurons at subthreshold voltage levels and enhances their coincidence detection sensitivity . To complete our study , therefore , we repeat our study of coincidence detection sensitivity with simulated synaptic inputs , but this time with dynamic ( voltage-gated ) KLT current included in the model . We construct and analyze a minimal description of a neuron that separates the input region ( soma and dendrites ) from the spike generating region ( axon and axon initial segment ) of a cell . This two-compartment model [45] has the form: A 1 C m V 1 ′ = - A 1 G 1 ( V 1 - E l k ) - g c ( V 1 - V 2 ) - I K L T , 1 - I i n A 2 C m V 2 ′ = - A 2 G 2 ( V 2 - E l k ) - g c ( V 2 - V 1 ) - I K L T , 2 - I N a . ( 1 ) The dynamical variables Vi ( i = 1 , 2 ) describe the membrane potential in each compartment . Passive parameters in the model are membrane capacitance per area ( Cm ) , axial conductance ( gc ) , reversal potential of leak current ( Elk ) , compartment surface area ( Ai ) , and membrane leak conductance density ( Gi ) . Parameters subscripted with i can take different values in the two compartments ( i = 1 , 2 ) . To simplify notation , we will often omit the explicit reference to membrane area and instead use the notation ci = Ai Cm and gi = Ai Gi for i = 1 , 2 . The first compartment ( abbreviation: Cpt1 ) receives input current ( Iin ) and the second compartment ( abbreviation: Cpt2 ) is the site of spike-generating sodium current ( INa ) . In some simulations we also include dynamic ( voltage-gated ) low-threshold potassium current ( IKLT ) . These currents are described in more detail below . We use standard neurophysiological measures of passive activity in the soma to determine some parameters , and vary other parameters to create a family of two-compartment models distributed in a two-dimensional parameter space . We select model parameters so that , regardless of coupling configuration , the passive dynamics in Cpt1 are nearly identical regardless of the strength of coupling between compartments . This novel formulation allows us to meaningfully and systematically probe the dynamics of the model . The two parameters that define coupling configuration are introduced below . They describe strength of forward coupling ( Cpt1 to Cpt2 ) and backward coupling ( Cpt2 to Cpt1 ) . The properties we match to experimental measurements include resting potential in the soma ( Vrest ) , input resistance for input to the soma ( Rin ) , and exponential time constant ( τexp ) with which soma voltage returns to rest following a brief perturbation . We use the following values based on in vitro measurements of gerbil MSO neurons [32]: Vrest = −58 mV , Rin = 8 . 5 MΩ , and τexp = 340 μs . We first match these properties using a model with passive dynamics ( by setting INa and IKLT to zero ) . After identifying parameter relations that satisfy these constraints , we discuss how to introduce sodium and KLT currents . In the passive model , the resting potential is identical to the reversal potential and we have Vrest = Elk . We now determine the remaining parameters based on the values of Rin and τexp . It is convenient to rescale the voltage equations by gi + gc and to introduce terms that represent the deviation of voltage from rest: Ui = Vi − Elk ( for i = 1 , 2 ) . This yields the following equations for passive and subthreshold dynamics ( INa and IKLT removed , for now ) : τ 1 U 1 ′ = - U 1 + κ 2 → 1 U 2 - J i n τ 2 U 2 ′ = - U 2 + κ 1 → 2 U 1 ( 2 ) The rescaled input current is denoted Jin = Iin/ ( g1 + gc ) . The time constants τi = ci/ ( gi + gc ) describe the passive dynamics of the ith compartment ( i = 1 , 2 ) when the other compartment is held at its resting voltage . We have also introduced in Eq 2 the two parameters that describe coupling strength . The forward coupling parameter is κ1→2 . Formally , it is the ratio of voltages U2/U1 at steady state in response to a constant current applied to Cpt1 . Similarly , the backward coupling parameter is κ2→1 . It is the ratio U1/U2 at steady state in response to constant current applied to Cpt2 . These quantities are attenuation factors that take values between zero ( complete attenuation ) and one ( no attenuation ) . We find it more intuitive to refer to these constants as measures of coupling strength—values near zero represent weak coupling , values near one indicate strong coupling . We will refer throughout to κ1→2 as the forward coupling parameter ( soma-to-axon coupling ) and κ2→1 as the backward coupling parameter ( axon-to-soma coupling ) . The relationship between coupling parameters and conductance parameters in Eq 1 are: κ 1 → 2 = g c g 2 + g c κ 2 → 1 = g c g 1 + g c . ( 3 ) Interestingly , the effect of one compartment on the other depends only on axial conductance and conductance in the target compartment . That is , the forward coupling parameter κ1→2 depends on gc and g2 but not on g1 . Similarly , the backward coupling parameter κ2→1 does not depend on g2 . Moreover , axial conductance is the same for current flowing in either direction , so differences in forward and backward coupling are solely due to differences between leak conductance in the two compartments . Next , we will show how to invert these equations to simply and uniquely define all passive model parameters for any combination of κ1→2 and κ2→1 . We only require prior knowledge of Rin and τexp ( experimentally measurable parameters ) and the assumption that the area of Cpt1 is much larger than the area of Cpt2 . We denote the ratio of compartment areas as α = A2/A1 . We will always use α = 0 . 01 in this study . This assumption is plausible for cells with input regions that are much larger than spike-generating regions , and is consistent with previous models of auditory coincidence detector neurons [24 , 38] . We find the axial conductance ( gc ) by expressing it in terms of input resistance and the coupling coefficients . By setting U 1 ′ = U 2 ′ = 0 in Eq 2 , we find the steady state relations U 1 s s = κ 2 → 1 U 2 s s - J i n and U 2 s s = κ 1 → 2 U 1 s s . From these relations , and the steady-state input resistance ( applying Ohm’s Law ) , it follows that R i n = - U 1 s s / I i n = 1 / [ ( 1 - κ 1 → 2 κ 2 → 1 ) ( g 1 + g c ) ] . Solving for gc and after some substitutions we find g c = κ 2 → 1 R i n ( 1 - κ 1 → 2 κ 2 → 1 ) . ( 4 ) The remaining parameters determine the passive dynamics of V1 , so they depend on τexp . In a one-compartment passive model , τexp is identical to the membrane time constant and its value is c1/g1 . In a two-compartment model , coupling between the compartments introduces a second time-scale that can influence the rate at which V1 returns to rest after a brief perturbation ( Rall’s equalization time constant [51] , and see also [52] ) . Some additional analysis is required , therefore , to relate τexp to model parameters . We invoke the assumption that the input region is much larger than the output region ( A1 ≫ A2 , or , equivalently , take 0 < α ≪ 1 ) and observe that this can create a separation of time-scales in the passive dynamics of U1 and U2 . The ratio of time constants in the two compartments is τ2/τ1 = ( c2/c1 ) [ ( g1 + gc ) / ( g2 + gc ) ] . After some substitutions , and using the assumption that Cm is identical in both compartments , we find that τ 2 τ 1 = α κ 1 → 2 κ 2 → 1 ( 5 ) We restrict ourselves to coupling configurations for which κ1→2/κ2→1 does not exceed ten , so that τ1 is an order of magnitude larger than τ2 ( recall we use α = 0 . 01 ) . In this scenario , we can segregate the passive dynamics into a slow variable ( U1 ) and a fast variable ( U2 ) . The ratio of time-constants is a small parameter which we denote by ϵ = α κ 1 → 2 κ 2 → 1 . For ϵ close to zero , we can make the approximation that U2 evolves instantaneously ( on the fast time-scale ) to its U1-dependent steady-state value of U 2 s s = κ 1 → 2 U 1 . On the slow-time scale , U2 takes this instantaneous value and the dynamics of U1 are ( to leading order in the small parameter ϵ ) : τ 1 U 1 ′ = - ( 1 - κ 2 → 1 κ 1 → 2 ) U 1 + J i n + O ( ϵ ) ( 6 ) In other words , in cases when τ1 ≫ τ2 , the passive dynamics in Cpt1 are approximately linear and one can match the time-scale of U1 to the experimentally-observed membrane decay time by setting τ1 = τexp ( 1 − κ1→2 κ2→1 ) . This slaving of U2 to U1 is valid for describing passive subthreshold dynamics . If sodium current is included , then the dynamics are non-linear and spike-generation is possible . On the slow time-scale , spike-generation is represented by a discontinuous jump to a fixed point at higher values of U1 and U2 . To summarize our method: we use values for three standard neurophysiological measures of ( passive ) soma dynamics ( Rin , τexp and Elk ) , we choose α ( the ratio of surface areas A2/A1 ) to be small ( α = 0 . 01 in all simulations ) , and we let the two coupling constants define a two-dimensional parameter space of soma-axon coupling . For any coupling configuration , we can then uniquely determine the passive parameters in Eq 1 . The parameter relationships , described above , are: g c = κ 2 → 1 R i n ( 1 - κ 1 → 2 κ 2 → 1 ) g 1 = g c ( 1 κ 2 → 1 - 1 ) g 2 = g c ( 1 κ 1 → 2 - 1 ) c 1 = τ e x p ( 1 - κ 1 → 2 κ 2 → 1 ) ( g 1 + g c ) c 2 = α c 1 . ( 7 ) Using these parameter relations guarantees that passive dynamics in Cpt1 remain nearly identical as we explore neural dynamics and coincidence detection sensitivity in the two-dimensional parameter space of coupling strengths . We only consider coupling configurations in which forward coupling is stronger than backward coupling ( κ1→2 ≥ κ2→1 ) . This corresponds to an assumption that voltage signals propagate forward from the soma to the axon with less attenuation than signals that backpropagate from the axon to the soma . This condition is appropriate for MSO neurons , since in vitro recordings show weak backpropagation of action potentials to the soma and dendrites [31 , 37] . Notice , from Eq 3 , that the condition κ1→2 ≥ κ2→1 is met if g1 ≥ g2 . This is reasonable in our model because we assume A1 ≫ A2 and do not expect that the density of leak conductance in each compartment ( G1 and G2 ) would differ by orders of magnitude . A voltage-gated low-threshold potassium current ( IKLT ) is prominent in MSO neurons and thought to improve coincidence detection sensitivity [53] . We model IKLT in the ith compartment with the equation I K L T , i ( V i , w i ) = g K L T w i 4 z ∞ ( V i ) ( V i - E K ) - g K L T w ∞ 4 ( V r e s t ) z ∞ ( V r e s t ) ( V r e s t - E K ) , ( 8 ) where the reversal potential is EK = −106 mV . We include the second term so that the addition of KLT current does not alter the resting potential ( IKLT , i = 0 when Vi = Vrest , for i = 1 or 2 . ) . Equivalently , one could adjust Elk to counterbalance the amount of KLT current active at rest , and omit this correction term . The dynamics of the activation variable wi are as in [32] at a temperature of 35°C: w i ′ = w ∞ ( V i ) - w i τ w ( V i ) w ∞ ( V i ) = 1 1 + e ( V i + 65 ) / 6 τ w ( V i ) = 0 . 46 ( 100 6 e ( V i + 75 ) / 12 . 15 + 24 e - ( V i + 75 ) / 25 + 0 . 55 ) ( 9 ) The inactivation variable is slow ( time-scale of several hundred milliseconds ) , so we make the simplification that its value is fixed at the steady state z∞ ( Vrest ) where the steady state function is [32] z ∞ ( V ) = 0 . 78 1 + e ( V + 57 ) / 5 . 44 + 0 . 22 . ( 10 ) Next we discuss how we include dynamic IKLT in the two-compartment model . We omit subscripts , for ease of presentation , but the same method applies for dynamic KLT current in either compartment . We first find the passive leak conductance in the relevant compartment using Eq 7 . Call this glk . We then reduce this conductance by some amount , typically 10% . In other words , we set the leak conductance in the relevant compartment ( gi ) to 0 . 9glk . Lastly , we set gKLT to the value that preserves the total conductance in the compartment at the resting potential . In some simulations , we leave the KLT conductance fixed at its resting value . We refer to this case as frozen KLT—the KLT current acts as a leak current and the subthreshold dynamics are the same as the original passive model . In other simulations , we allow KLT conductance to depend on voltage . We refer to this case as dynamic KLT . To include dynamic KLT in Cpt1 , for example , we would choose gKLT so that it satisfies the equation g 1 + g K L T w ∞ 4 ( V r e s t ) z ∞ ( V r e s t ) = g l k and allow the KLT activation variable w to evolve according to Eq 9 . The second compartment represents regions of the cell in which spikes are generated , presumably the axon initial segment or other excitable regions in the axon [38] . We use a reduced model of sodium current , adapted from earlier models of auditory brainstem neurons [32 , 54] , to produce spikes: I N a ( V 2 , h ) = g N a m ∞ 3 ( V 2 ) h ( V 2 - E N a ) - g N a m ∞ 3 ( V r e s t ) h ∞ ( V r e s t ) ( V r e s t - E N a ) , ( 11 ) where the sodium reversal potential is ENa = 55 mV . The second term is included so that INa = 0 when V2 is at its resting value . Equivalently , one could adjust Elk to counterbalance the amount of sodium current active at rest , and omit this correction term . Setting INa = 0 at rest simplifies analysis of the model and is appropriate for MSO neurons . Sodium current in MSO neurons is small at rest , with most sodium channels inactivated at the resting membrane potential [39] . We assume that activation of sodium is sufficiently fast to justify the approximation that the gating variable m instantaneously reaches its voltage-dependent equilibrium value m∞ ( V2 ) [34] . The gating variable h governs inactivation of the sodium current and has dynamics as in [32]: h ′ = h ∞ ( V 2 ) - h τ h ( V 2 ) h ∞ ( V 2 ) = 1 1 + e ( V 2 + 65 ) / 6 τ h ( V 2 ) = 0 . 24 ( 100 7 e ( V 2 + 60 ) / 11 + 10 e - ( V 2 + 60 ) / 25 + 0 . 6 ) ( 12 ) These are the same as in [54] , but with temperature adjusted to 35°C , and also note the resting membrane potential in our model is -58 mV as opposed to -65 mV in [54] Our primary objective is to determine the effects of coupling configuration on coincidence detection . Maximal conductance gNa determines excitability and spike threshold in the model neuron and thus also influences coincidence detection sensitivity [22] . Rather than setting gNa to an arbitrarily chosen value , we explore a range of gNa to determine the best possible coincidence detection sensitivity over this range of gNa , for each coupling configuration . We explain our method for choosing gNa values in more detail below . We generate synaptic inputs to the two-compartment model using a model of the auditory periphery [55] . This model includes the effects of cochlear filtering and nonlinearities , inner hair cell activity , and synaptic transmission , and generates auditory nerve spike trains . As inputs to this model , we use sine waves that represent pure tone sounds . We perform simulations with frequencies ranging from 200 Hz to 700 Hz at a level of 70 dB . The neuron model receives two streams of auditory nerve inputs representing ( conceptually ) inputs from the two ears , see the schematic in Fig 1 . The sine waves to the two ears are presented either with identical timing to generate coincident inputs , or with a time delay to simulate non-coincident inputs . MSO coincidence detectors receive a small number of synaptic inputs [56] , so we use the auditory nerve model to simulate five independent input sequences of spike times per ear . Each auditory nerve spike time creates an excitatory post-synaptic conductance ( EPSG ) described by a double exponential function [18]: g s y n ( t ) = 125 . 25 ( e - t / 0 . 18 - e - t / 0 . 1 ) . ( 13 ) Conductance gsyn is in units of nanosiemens and t is milliseconds . These EPSGs are transformed into synaptic current ( EPSCs ) according to the equation I s y n ( t ) = g s y n ( t ) ( V 1 - E s y n ) , ( 14 ) where the reversal potential for the excitatory current is Esyn = 0 mV . We set the constant scaling factor in the definition of gsyn ( t ) ( Eq 13 ) so that a single excitatory input depolarizes V1 by roughly 6 mV , a value consistent with measurements of MSO neurons’ responses to synaptic excitation in vitro [57] . Our goal is to identify and understand essential aspects of how structural and dynamical features of neurons affect coincidence detection . Although we do not incorporate a complete description of neural processing in the MSO-pathway , we view this setup to be adequate for probing coincidence detection in an MSO-like two-compartment model using quasi-realistic stimuli . Notably , we do not include spherical bushy cells in the cochlear nucleus that may enhance temporal precision of afferent inputs to MSO neurons [27] , nor do we include inhibitory inputs that appear to modify time-difference tuning of MSO coincidence detectors [57–63] , but see also [18] . In some simulations , as we will make clear in the context of the Results , we use more simplistic inputs such as steps or ramps of current injected directly into the input compartment to study response characteristics of the model . We measure firing rate ( spikes per second ) generated by the two-compartment model in response to coincident inputs ( identical sine wave stimuli to the two ears of the auditory nerve model ) and non-coincident inputs . In some simulations we generate non-coincident inputs by using sine wave stimuli that were anti-phase to the two ears of the auditory nerve model . For example , for a 500 Hz stimuli , the two sine waves would have a 1 ms time difference . In this construction of non-coincident inputs , the time difference between the sine waves shortens with increasing frequency . To confirm that this dependence of time difference on frequency does not bias our results , we also perform simulations in which we generate non-coincident inputs by using sine wave stimuli to the auditory nerve model with a fixed time difference of 500 μs . To measure coincidence detection sensitivity , we compute the difference in firing rates for responses to coincident and non-coincident inputs . We compute firing rates ( spikes per second ) by counting the number of spikes generated in trials of length 250 ms . We then calculate mean and standard error of firing rates from 100 repeated trials . Large-scale calculations to sweep over parameter space were performed using Matlab simulation code executed on computers managed by the Ohio Supercomputer Center . The ordinary differential equations defining the two-compartment model were solved numerically using the Matlab command ode15s ( a variable-step , variable-method solver useful for stiff systems ) . Simulation code is available at https://github . com/jhgoldwyn/TwoCompartmentModel . A good coincidence detector neuron would be one with a large difference in firing rates for these two conditions . Firing rate difference measures of coincidence detection sensitivity have been used in related studies [64 , 65] . Other measures have been considered , including Fisher information [38] , width of time-difference tuning curves [62] , and quality factor ( similar to d-prime ) [22] . The right measurement of coincidence detection sensitivity remains , as these alternatives reveal , an open question ( and one wrapped up in ongoing debates regarding the nature of the neural code for sound source location [66] ) . One justification for comparing in-phase to out-of-phase firing rates is that it is relevant to a system that uses a two channel representation of auditory space in which sound location is represented by the difference in firing rates between two populations of cells tuned to distinct time-differences [67] . We suggest an additional perspective based on an analogy to signal classification theory and the receiver operating characteristic ( ROC ) [68] . Consider a coincidence detector neuron responding to a periodic ( sine wave ) stimulus . Each cycle of the stimulus evokes a volley of synaptic inputs that may or may not be temporally aligned with one another . The task of the coincidence detector neuron is to respond ( generate a spike ) if the synaptic inputs arrive within a brief time window and to not respond ( not spike ) if the synaptic inputs are dispersed in time . From this perspective , a coincidence detector neuron is an observer of its own synaptic inputs and it signals the presence of coincident inputs by generating a spike . Chance [69] has articulated a similar approach for measuring synaptic efficacy . Extending the analogy , for each two-compartment model ( parameterized by coupling configuration ) , we construct ROC curves by plotting hit rate ( firing rate to coincident inputs ) against false alarm rate ( firing rate to non-coincident inputs ) for varying values of the sodium conductance gNa . Sodium conductance controls the overall excitability of the model and operates as the threshold parameter in ROC analysis . To compare coincidence detection sensitivity across coupling configurations , we simulate the model for a range of gNa values and define coincidence detection sensitivity to be the maximum firing rate difference . In this way , we identify the gNa level for which the neuron , acting as an observer of its inputs , is the best possible coincidence detector . In other words , we identify the gNa value that maximizes hits ( spikes generated in response to coincident inputs ) while minimizing false alarms ( spikes generated in response non-coincident inputs ) , for a given coupling configuration and stimulus . We illustration this calculation in Fig 2A ( cartoon only , not simulation data ) . We provide S1 Fig . to illustrate this calculation with simulation data . The gNa values that produced similar spiking activity for different coupling configurations could vary by orders of magnitude , so we required a reasonable way to determine the appropriate range of gNa values to use . For each coupling configuration , we set a reference value for gNa , denoted g N a r e f , by finding the smallest gNa value for which a pair of coincident EPSGs could evoke a spike . The rationale for this definition of g N a r e f is that , for any gNa value larger than g N a r e f , the model neuron could possibly spike in response to two non-coincident EPSGs . For gNa significantly larger than g N a r e f , the model neuron could even spike in response to a single EPSG . By restricting gNa to values near g N a r e f , we ensured that the model neurons were operating as coincidence detectors and not responding to multiple inputs that lacked temporal alignment . Values of g N a r e f in the parameter space of forward and backward coupling strengths are shown in Fig 2B . Values of g N a r e f increase if dynamic KLT current is present in the model . Replacing 10% of glk with KLT conductance increased reference gNa values by 5% to 30% higher depending on coupling configuration , see S2 Fig . We then measure coincidence detection sensitivity using gNa values that ranged from 0 . 2 to 2 . 2 times g N a r e f , in increments of 0 . 05 times g N a r e f . From the firing rate differences measured across this range of gNa values , we identify the maximum firing rate difference and use this as our measure of coincidence detection sensitivity . As a result of this process , we obtain different best gNa values depending on input frequency , coupling configuration , and KLT currents . An implication of this approach is that neurons should modulate sodium conductance based on stimulus parameters and physiological conditions . We do not pursue this idea here , but see [42 , 44] for relevant studies and S3 Fig . for supporting results showing changes of best gNa with stimulus frequency . For further analysis of the dynamics of the two-compartment model , we performed bifurcation analysis using the Auto feature in XPPAUT . XPPAUT is a software package for analyzing and simulating dynamical systems , and uses continuation methods to perform numerical bifurcation analysis [70] . As discussed in Materials and methods , the forward coupling parameter ( soma-to-axon , κ1→2 ) and backward coupling parameter ( axon-to-soma , κ2→1 ) characterize the passive dynamics of a two-compartment model ( see Eq 3 ) . By defining model parameters according to Eq 7 , we construct a family of models with nearly identical passive dynamics in Cpt1 . Recall we select model parameters based on reported properties of MSO neurons: steady state input resistance is 8 . 5 MΩ , the decay time constant is 340 μs , and the resting potential is −58 mV [32] . We also assume the capacitance per unit area is the same in both compartments and that the surface area of Cpt1 was 100 times larger than the surface area of Cpt2 ( α = 0 . 01 ) . Using a typical value of membrane capacitance Cm = 0 . 9 pF/cm2 [71] , we find that capacitance in Cpt1 is C1 = 40 pF . The area of Cpt1 is A1 = 4444 μm2 , which is a plausible value for the surface area of the soma and dendrite regions of an MSO neuron [31 , 72] . In order to maintain identical passive dynamics in Cpt1 across coupling configurations , the leak conductance ( g1 , g2 ) and axial conductance ( gc ) vary with the values of the coupling parameters as shown in Fig 3 . Leak conductance in Cpt1 increases as forward coupling strength increases but decreases as backward coupling strength increases ( Fig 3A ) . Leak conductance in Cpt2 exhibits the opposite dependence on coupling strength: g2 decreases as forward coupling strength increases and increases as backward coupling strength increases ( Fig 3B ) . The axial conductance connecting the two compartments depends primarily on the strength of backward coupling—the contour lines in Fig 3C are nearly horizontal except in cases of strong forward and backward coupling ( upper right corner ) . The upper left in each panel of Fig 3 is empty because we only consider coupling configurations for which forward coupling is not weaker than backward coupling ( κ1→2 ≥ κ2→1 ) . To explore how coupling configuration modifies neural dynamics , we will often compare three models near the edges of the coupling parameter space . These are a weakly-coupled model ( κ1→2 = 0 . 3 , κ2→1 = 0 . 2 ) , a strongly-coupled model ( κ1→2 = 0 . 8 , κ2→1 = 0 . 7 ) , and a forward-coupled model ( κ1→2 = 0 . 8 , κ2→1 = 0 . 2 ) . The locations of these three models in the coupling parameter space are shown as colored stars in Fig 3 . We remark that complete coupling ( κ1→2 = κ2→1 = 1 ) is equivalent to a one-compartment point neuron model because voltages in the two compartments are the same in this case . Our parameterization method is designed to maintain the same voltage response in Cpt1 ( V1 ) regardless of the coupling configuration . In fact , due to the strong separation of time scales between the two compartments ( recall Eq 5 ) , the voltage in Cpt1 is governed approximately by linear dynamics with time constant τexp ( see Eq 6 ) and the voltage in Cpt2 is V2 ≈ Elk + κ1→2 ( V1 − Elk ) . These approximations are valid to leading order in the small parameter ϵ = α κ 1 → 2 κ 2 → 1 . We remind the reader that these calculations are performed in the case of passive dynamics—i . e . for a model without spike-generating sodium current ( gNa = 0 ) and with frozen low-threshold potassium current ( IKLT acts as a leak current , and in fact is equivalent to gKLT = 0 , see Materials and methods ) . Simulations of passive two-compartment models illustrate how the parameterization method results in models with nearly identical V1 dynamics ( Fig 4B ) . We use darker colors to show time-courses of voltage in Cpt1 ( V1 ) and lighter colors to show time-courses of voltage in Cpt2 ( V2 ) . V1 responses are nearly identical regardless of coupling configuration and attenuation of V2 responses depends on κ1→2 . Time-courses of V1 and V2 shown here are responses to 500 Hz coincident inputs . The three coupling configurations in this figure are indicated in the schematic in the top row ( from left to right in Fig 4A: weakly-coupled , forward-coupled , and strongly-coupled , as defined previously ) . We include spiking in the model by adding sodium current to Cpt2 ( see Eq 11 ) . As described in Materials and methods , we define g N a r e f for each coupling configuration as the minimum level of gNa at which two coincident inputs evoke a spike . We find it helpful to normalize gNa to these reference values when comparing across coupling configurations . In Fig 4C , we show responses to 500 Hz coincident inputs with gNa set to g N a r e f . This results in gNa = 6291 nS for the weakly-coupled model , gNa = 398 nS for the forward-coupled model , and gNa = 2003 nS for the strongly-coupled model . We do not include dynamic KLT current in these simulations . These responses to identical inputs ( same trains of synaptic inputs for each model , regardless of coupling configuration ) reveal that spiking dynamics differ depending on coupling configuration . Coupling configuration dramatically changes the amplitude and shapes of spikes—the peaks of V2 in the weakly-coupled model are near 40 mV whereas the peaks of V2 in the forward-coupled and strongly-coupled models are near 0 mV ( note the different axes ) . Spikes in the weakly-coupled model tend to be all-or-none , but the forward-coupled and strongly-coupled models can have more graded spike amplitudes . Not surprisingly , amplitudes of backpropagated spikes in Cpt1 depend on κ2→1 . For models with weak backward coupling ( the weakly-coupled and forward-coupled models ) , V2 spikes cause small V1 deflections that can appear similar to subthreshold V1 activity . For the strongly-coupled model , in contrast , V1 tracks V2 more closely and shows larger amplitude backpropagated spikes in Cpt1 . Importantly , the number and timing of spikes also changes with coupling configuration . In these traces , the forward-coupled model has two more spikes than the weakly-coupled and strongly-coupled models ( see the extra spikes at 10 ms and 20 ms for the forward-coupled model ) . This anticipates our main result: coupling configuration affects spike generation and can alter the sensitivity of neurons to coincident inputs . If we view this example simulation using the analogy to signal detection theory and ROC analysis , described previously , we can say that the forward-coupled model correctly identifies two more coincident events ( has two more hits ) than the weakly-coupled and strongly-coupled models . In a first set of simulations , we study the two-compartment model with passive subthreshold dynamics . The low-threshold potassium ( KLT ) current is frozen and included as part of the leak current ( see Materials and methods ) . The only voltage-gated current in this set of simulations is the spike-generating sodium current in Cpt2 . We quantify coincidence detection sensitivity by finding the maximum firing rate difference between coincident and non-coincident inputs for each coupling configuration ( as described in Materials and methods ) . In Fig 5 , we report results for three stimulus frequencies ( from left to right: 300 Hz , 500 Hz , and 700 Hz ) . We construct non-coincident inputs in two ways: in Fig 5A we use out-of-phase sine wave inputs to the auditory nerve model . Time delays for out-of-phase inputs vary with frequency , so we also test coincidence detection sensitivity using sine wave inputs that are misaligned in time by a fixed 500 μs time difference in Fig 5B . Larger firing rate differences reflect better coincidence detection . In our analogy to a signal classification task , we say a large firing rate difference indicates a detector with a high hit rate in response to coincident inputs and a low false alarm rate in response to non-coincident events . We make two observations that we will explore in greater detail below . First , coincidence detection sensitivity improves with increases in forward coupling strength . That is to say , as one moves from left-to-right within each panel of Fig 5 , the firing rate difference increases . Second , the combination of strong forward coupling ( large κ1→2 ) and weak backward coupling ( small κ2→1 ) enhances coincidence detection for high-frequency stimuli ( notice the large firing rate differences in the lower right corner of panels for 500 Hz and 700 Hz stimuli ) . In Fig 6 , we further detail how coincidence detection sensitivity changes with stimulus frequency . Strong forward coupling enhances coincidence detection for all frequencies above 200 Hz ( the weakly-coupled model has the smallest firing rate difference across these frequencies ) . An advantage for the forward coupling configuration ( combination of strong forward coupling and weak backward coupling ) emerges for stimulus frequencies above about 400 Hz . These observations hold ( qualitatively ) regardless of whether we use out-of-phase stimuli for non-coincident inputs ( Fig 6A ) or time-delayed stimuli ( Fig 6B ) . We provide tuning curves in S4 Fig . for additional views of how coincidence detections sensitivity depends on stimulus frequency , gNa , and coupling configuration . We emphasize that the comparisons we are making are across coupling configurations . We are not suggesting , in particular , that the non-monotonic trend in Fig 6A indicates an optimal frequency for coincidence detection . This non-monotonic trend is due in part to a ceiling effect at low frequencies ( in-phase firing rates do not exceed 200 Hz for a 200 Hz stimulus ) , and also masks the fact that the proportion of coincident inputs that trigger a spike ( the hit rate , in the signal detection analogy ) decreases with increasing stimulus frequency . Why does strong forward coupling improve coincidence detection ? And why does the specific combination of strong forward and weak backward coupling ( the forward-coupled model ) enhance high-frequency coincidence detection ? We provide explanations below . First , we will demonstrate that strong forward coupling endows the two-compartment model with two properties that are advantageous for neural coincidence detection: phasic firing and sensitivity to input slope . Second , we will show that the specific combination of strong forward coupling and weak backward coupling shortens the refractory period of the two-compartment model . Neurons with short refractory periods can faithfully and rapidly respond to high-frequency sequences of coincident inputs . Neurons with longer refractory periods are disadvantaged when performing high-frequency coincidence detection . They may miss opportunities to generate spikes in response to coincident inputs , and thus their hit rate ( when thinking of these neurons as signal detectors ) may be depressed . In the preceding sections , we have detailed the advantages of strong forward coupling generally , and weak backward coupling for high-frequency stimuli , for coincidence detection sensitivity in a two-compartment neuron model . With the exception of the spike-generating sodium current , the two-compartment model we have considered to this point has been passive . We questioned how our findings would change if additional voltage-gated currents were included . Of particular interest in the context of neural coincidence detection in the MSO is the low threshold potassium ( KLT ) current . This current is prominent in MSO neurons and enhances their coincidence detection sensitivity [33] . We therefore repeated our test of coincidence detection sensitivity with dynamic KLT conductance ( see Materials and methods ) . In Fig 12 we show coincidence detection sensitivity measured from responses to three input frequencies ( from left to right: 300 Hz , 500 Hz , and 700 Hz ) . In the top row , 10% of the total conductance in Cpt1 at rest is dynamic KLT conductance . In the bottom row , 10% of the total conductance in Cpt2 at rest is dynamic KLT conductance . The format of each panel is similar to Fig 5 with the color scale in each panel representing the maximal firing rate difference between in-phase and out-of-phase inputs . Upon comparing these results to our previous results using a passive model ( frozen KLT , Fig 5 ) , we observe some differences . Dynamic KLT in the input region ( Cpt1 , top row ) improves coincidence detection sensitivity for all model configurations . While models with strong forward coupling and weak backward coupling ( lower right corner of each panel ) remain as effective coincidence detectors , the optimal configuration shifts to models with stronger backward coupling . For the 300 Hz stimulus , for instance , the largest firing rate differences are achieved for models with strong forward and strong backward coupling ( upper right corner ) . Models with strong backward coupling can more effectively make use of the KLT current because V2 spikes propagate back into Cpt1 to activate the KLT current . Dynamic KLT in the output region ( Cpt2 , bottom row ) also improves coincidence detection sensitivity for all model configurations , but the greatest increases are in models with weak coupling . As a result , coincidence detection sensitivity is nearly uniform across all model configurations , especially in responses to lower and higher frequency inputs . Dynamic KLT in Cpt2 tends to provide the most benefit for models with weak coupling because it provides a secondary source of voltage-gated , dynamic , negative feedback in these models for which sodium inactivation does not suffice to establish dynamics conducive to coincidence detection , including phasic responses to steady inputs and slope-sensitivity ( recall Figs 7 and 8 ) . We compare coincidence detection sensitivity across a range of stimulus frequencies for models with the weak , forward , and strong coupling configurations and that include dynamic KLT , see Fig 13 . As above , we test models with dynamic KLT conductance in Cpt1 ( Fig 13A1 and 13B1 ) , and models with dynamic KLT conductance in Cpt2 ( Fig 13A2 and 13B2 ) . Results for models with dynamic KLT are shown in thick lines . For reference , we also include our earlier results using the frozen KLT model ( thin lines , same as results shown in Fig 6 ) . The results are consistent with our observations from Fig 12 . In particular , we find that dynamic KLT conductance improves coincidence detection sensitivity ( relative to the passive model ) for all coupling configurations and nearly all stimulus frequencies . Moreover , the benefit of KLT depends on coupling configuration . Dynamic KLT added to Cpt1 ( soma ) improves coincidence detection sensitivity the most for the strongly-coupled model . Dynamic KLT added to Cpt2 ( axon ) improves coincidence detection sensitivity the most for the weakly-coupled model . Specializations that support temporally-precise coincidence detection in MSO neurons include voltage-gated currents active at membrane potentials near resting values [53] , fast and well-timed excitatory synapses [27] , and dendritic structure ( bipolar dendrites , that segregate inputs from opposite ears onto opposite dendrites ) [29 , 31] , see also [25] for review . In this work , we showed that soma-axon coupling is an additional structural specialization that can enhance neural coincidence detection . By performing a thorough search through the space of coupling configuration , we found that strong forward ( soma-to-axon ) coupling improved coincidence detection sensitivity . And , moreover , the asymmetric forward-coupled configuration of strong forward coupling and weak backward coupling was the optimal configuration for coincidence detection in response to higher frequency inputs ( 500 Hz to 700 Hz ) . We identified advantages of strong forward coupling for neural coincidence detection including phasic and slope-sensitive spiking dynamics , and ( for the forward-coupled model ) short refractory periods . These advantages depended on the action of the sodium inactivation gating variable , which is the sole source of voltage-gated , dynamic , negative feedback in the version of the two-compartment model with frozen KLT . KLT current is an additional source of negative feedback and one known to be prominent in MSO neurons in soma and dendrites regions [31] , as well as axon regions [38] . We found that dynamic KLT current improved coincidence detection for nearly all coupling configuration and stimulus frequencies . There were notable interactions between coupling configuration and KLT current . Dynamic KLT current improved coincidence detection in neurons with strong forward and backward coupling so that this strongly-coupled configuration ( the configuration most similar to a one-compartment model ) became optimal for coincidence detection in response to intermediate-frequency stimuli . In addition , dynamic KLT current localized to Cpt2 ( the spike-generator region of the neuron model ) could rescue coincidence detection sensitivity in neurons with weak soma-axon coupling so that this weakly-coupled configuration could exhibit coincidence detection sensitivity on par with other models . There was some loss of efficiency when dynamic KLT current was included ( g N a r e f values increased by 5% to 30% ) , but these differences are modest compared to the order of magnitude differences in g N a r e f across coupling configurations . Our work adds to the characterization of MSO neurons as coincidence detector specialists equipped with an array of features—structural and dynamical—that enhance their temporal precision . Strong forward coupling appears to be a natural configuration for coincidence detection . Moreover , the specific combination of strong forward coupling and weak backward coupling is advantageous for high frequency coincidence detection . The benefits of these coupling configurations can be supplemented with appropriately-targeted KLT current . The need for multiple , complementary mechanisms that enhance coincidence detection MSO neurons has been explored previously , but usually for one-compartment models ( Na inactivation and KLT activation as two sources of negative feedback [36]; KLT current and hyperpolarization-activated cationic current as currents that regulate input resistance [32] ) . The effect of KLT is considerable and well-studied , so here we emphasized the role of structure ( soma-to-axon coupling ) . The intrinsic advantage of the forward-coupling configuration may help maintain coincidence detection sensitivity in scenarios in which KLT is less effective ( early in development [73] , for instance , or when KLT is inactivated by long time-scale channel gating dynamics that are not in our model [32] ) . Other features of MSO neurons such as inhibitory synaptic inputs [57 , 59–63] and dendritic structure [29–31] specialize these cells for coincidence detection . The variety of physiological tools MSO neurons use to perform coincidence detection emphasizes the exceptional nature of the temporally-precise computations these neurons perform . Inhibitory inputs in MSO primarily target cell bodies [74] so careful study of the inhibition , excitation , and cell-structure would be an interesting avenue of future research . Previous studies using point-neuron models and dynamic clamp electrophysiology demonstrate that the interaction of inhibition and KLT current can shift the location of the peak of time difference tuning curves [57 , 61 , 63] . We would expect similar effects in a two-compartment model , as long as the soma-targeting inhibition and KLT current are not electrically isolated . These effects may require , therefore , a prominent KLT current in the soma ( co-localized with inhibition ) or a strong soma-to-axon coupling configuration to enable soma-targeting inhibition to interact with axonal KLT current ( if it is present ) . We tested coincidence detection sensitivity with out-of-phase inputs and inputs with a fixed 500 μs delay . We observed qualitatively similar results for both types of inputs . The latter stimulus may be more relevant in studying neural coincidence in the context of sound localization . Time differences in this context would be created by differences in travel times of sounds arriving at the two ears ( interaural time differences ) and are limited by animal head size . In humans , for instance , maximal interaural time differences created by head size are approximately 700 μs . Our model is phenomenological—the two-compartments are lumped representations of input and output regions . We do not , therefore , resolve structural details of dendrites or spike initiation zones ( see [38] for an example of the latter ) . Nonetheless , we can make qualitative observations that relate our findings to MSO physiology . Action potentials in the MSO are likely generated in a spike initiation zone near the soma , and back-propagated action potentials in the soma are small and graded [37] . This indicates a strict electrical segregation of the soma and dendrites from the axonal initiation zone , ( in the words of [37] ) . In the context of our model , this corresponds to weak backward coupling ( small value of κ2→1 ) . Backpropagation of signals into the dendrites is further attenuated due to the low input resistance of these neurons and the strong effects of voltage-gated potassium current [31] . Additionally , current injection into the soma reliably evokes action potentials that propagate into the axon [37] . This suggests a configuration in which the soma has a strong effect on the spike-generator ( minimal attenuation , large value of κ1→2 in our model ) . Taken together then , it appears that MSO neurons may be structured in a forward-coupled manner , consistent with our observations that this configuration confers advantages for coincidence detection by engaging sodium inactivation as dynamic negative-feedback mechanism , by promoting rapid resetting of the spike generator ( shortening the refractory period , which enables high-frequency spiking ) , and by enabling efficient spike generation ( smaller sodium conductance required ) . The complete picture of MSO excitability and axonal structure is , of course , more complicated than our two-compartment description . Recent computational simulations provide evidence that spike generation may occur throughout MSO axons ( initial segment and multiple nodes of Ranvier ) [38] . Spike generation at more distal locations on the axon can preserve excitability in response to high-frequency stimuli by preventing inactivation of sodium channels [38 , 44] . Studies of coincidence detector neurons in related structures in the avian auditory show that excitability of these neurons can be adjusted via modulation of ion channel density in spike generator regions [42–44] . This raises intriguing questions about plasticity in the spike initiation zone , and dynamic regulation of the soma-to-axon connection . We have formulated a family of two-compartment models to investigate neural coincidence detection in MSO neurons . We showed that parameters in this two-compartment framework can be chosen in a principled manner to explore the range of coupling configurations , while maintaining similar passive dynamics in the input region . With this approach , we identified how structure ( the nature of soma-axon coupling ) affected dynamics in the spike-generator region , and , in turn , how these differences in dynamics affect the sensitivity of coincidence detector neurons to synaptic inputs . Our approach provides a unifying view of structure and function in neurons performing an identified computation . It is one that should find applications in studies of other neurons . Coincidence detector neurons in the auditory brainstem of owls , for instance , have been modeled as two-compartment structures [41] . The two-compartment idealization has also been useful for investigating dynamics of bursting [45 , 49 , 75] , bistability [46] , oscillations [47] , and resonance [50] in neurons , and could also describe signaling between a ( large ) dendrite region and a ( small ) dendritic spine . Our framework for creating , and systematically exploring , a parameter space of soma-axon coupling configurations can be used to shed further light on the relationship between structure , dynamics , and function in these and other neural systems .	Brain cells ( neurons ) are spatially extended structures . The locations at which neurons receive inputs and generate outputs are often distinct . We formulate and study a minimal mathematical model that describes the dynamical coupling between the input and output regions of a neuron . We construct our model to reflect known properties of neurons in the auditory brainstem that play an important role in our ability to locate sound sources . These neurons are known as coincidence detectors because they are most likely to respond when they receive simultaneous inputs . We use simulations to explore coincidence detection sensitivity throughout the parameter space of input-output coupling and to identify the coupling configurations that are best for neural coincidence detection . We find that strong forward coupling ( from input region to output region ) , enhances coincidence detection sensitivity in our model and that low-threshold potassium current further improves coincidence detection . Our study is significant in that we detail how cell structure affects neuronal dynamics and , consequently , the ability of neurons to perform as temporally-precise coincidence detectors .	[ "Abstract", "Introduction", "Materials", "and", "methods", "Results", "Discussion" ]	[ "pharmacologic", "analysis", "action", "potentials", "medicine", "and", "health", "sciences", "dendritic", "structure", "auditory", "nerves", "nervous", "system", "membrane", "potential", "electrophysiology", "neuroscience", "simulation", "and", "modeling", "compartment", ...	2019	Soma-axon coupling configurations that enhance neuronal coincidence detection
Bacterial growth depends crucially on metabolic fluxes , which are limited by the cell’s capacity to maintain metabolic enzymes . The necessary enzyme amount per unit flux is a major determinant of metabolic strategies both in evolution and bioengineering . It depends on enzyme parameters ( such as kcat and KM constants ) , but also on metabolite concentrations . Moreover , similar amounts of different enzymes might incur different costs for the cell , depending on enzyme-specific properties such as protein size and half-life . Here , we developed enzyme cost minimization ( ECM ) , a scalable method for computing enzyme amounts that support a given metabolic flux at a minimal protein cost . The complex interplay of enzyme and metabolite concentrations , e . g . through thermodynamic driving forces and enzyme saturation , would make it hard to solve this optimization problem directly . By treating enzyme cost as a function of metabolite levels , we formulated ECM as a numerically tractable , convex optimization problem . Its tiered approach allows for building models at different levels of detail , depending on the amount of available data . Validating our method with measured metabolite and protein levels in E . coli central metabolism , we found typical prediction fold errors of 4 . 1 and 2 . 6 , respectively , for the two kinds of data . This result from the cost-optimized metabolic state is significantly better than randomly sampled metabolite profiles , supporting the hypothesis that enzyme cost is important for the fitness of E . coli . ECM can be used to predict enzyme levels and protein cost in natural and engineered pathways , and could be a valuable computational tool to assist metabolic engineering projects . Furthermore , it establishes a direct connection between protein cost and thermodynamics , and provides a physically plausible and computationally tractable way to include enzyme kinetics into constraint-based metabolic models , where kinetics have usually been ignored or oversimplified . The biochemical world is remarkably diverse , and new pathways and chemicals are still discovered routinely . Even for extensively studied model organisms like E . coli , efforts to exhaustively map metabolic networks are only nearing completion on the stoichiometric level . Our understanding of metabolic fluxes , their dynamic regulation and their connection to cell fitness is far from perfect [1] . Furthermore , the rational design of novel and efficient metabolic pathways remains a substantial challenge and metabolic engineering projects require considerable efforts even for relatively simple metabolic tasks . Among the different possible criteria [2] , one key to understanding the choices of metabolic routes , both in naturally evolved and engineered organisms , may be enzyme cost . Quite often , cells use metabolic pathways in ways that seem irrational , as in the case of aerobic fermentation ( known as the Crabtree effect in yeast or the Warburg effect in cancer cells [3] ) . However , apparently yield-inefficient fluxes can sometimes be explained by an economic use of enzyme resources [4 , 5] . It is posited that pathway structures that require too much enzyme per unit flux will be out-competed during evolution and will not be efficient for biotechnological applications . Thus , a quantitative analysis of resource investment in enzyme production , predicting the amount of enzyme needed to support a given flux , would be valuable in aiding the rational design of metabolic pathways . To understand why specific enzymes or pathways occupy larger or smaller areas of the proteome [6] , we could proceed in two steps , determining first the metabolic fluxes and then enzyme levels needed to realize these fluxes . Metabolic fluxes can be measured through isotope-labeled tracer experiments in combination with computational modeling . Methods for flux prediction ab initio rely on mechanistic aspects ( chemical mass balances and kinetics ) and economic aspects ( cost and benefit of pathway fluxes ) and combine them in different ways . Constraint-based methods like Flux Balance Analysis ( FBA ) determine fluxes by requiring steady states—i . e . , fluxes must be such that internal metabolite levels remain constant in time—and assuming that natural selection maximizes some benefit function ( e . g . , maximal yield of biomass ) . Several optimality criteria for fluxes can be combined by multi-objective optimization [1 , 7] . In some cases , the second law of thermodynamics is used to put further constraints on fluxes or metabolite levels [8–11] . Some extensions of FBA [12–14] use metabolite log-concentrations as extra variables and constrain fluxes to flow only in the direction of thermodynamic driving forces , i . e . , towards lower chemical potentials . Thermodynamics links between flux directions and reactant concentrations , and thus physiological bounds on metabolite levels are translated into restrictions on flux directions . These links between fluxes and metabolite concentrations hold independently of specific reaction kinetics . The relationship between fluxes and metabolite concentrations can be used also in the opposite direction—i . e . given all flux directions , certain metabolite profiles can be excluded [14] . The set of feasible metabolite profiles can be depicted as a polytope in the space of metabolites’ log-concentrations . To further narrow down the metabolite concentration profiles , the Max-min Driving Force ( MDF ) method [15] chooses profiles that ensure sufficient driving forces , thus keeping reactions distant from chemical equilibrium . Typically , constraint-based models bypass the non-linearity of enzyme kinetics by focusing on the feasible flux space and assess the relative benefits of different flux distributions . Thus , such models do well in simulating binary perturbations such as reaction knockouts or nutrient deprivation . On the other hand , they were not designed to predict the necessary enzyme levels and the cost of making and maintaining the enzymes , and therefore perform poorly at these tasks . Here we ask: how can we estimate the amount of protein required to sustain a given flux through a reaction or pathway ? It is often assumed that the flux through a reaction is proportional to the enzyme level . FBA methods use this assumption to translate enzyme expression , as a proxy for protein burden , into flux bounds or linear flux cost functions [16] . For practical reasons ( computational tractability and lack of detailed knowledge ) , flux costs are often represented by the sum of absolute fluxes [17 , 18] . To obtain better proxies of protein demand and related cellular burdens , fluxes have been weighted by “flux burdens” that account for different catalytic constants kcat [2 , 19] , protein size and lifetime [20] , or equilibrium constants [17] . In reality , however , enzyme demand does not only depend on fluxes , but also on metabolite levels , which in turn are determined by the non-linear kinetics of all active enzymes and transporters . Therefore , it is not only the choice of numerical cost weights , but the very relation between enzyme amounts and fluxes that needs to be clarified . For a simple estimate , we can assume that each enzyme molecule works at its maximal rate , the catalytic constant kcat . In this case , enzyme demand is given by the flux divided by the catalytic constant [2 , 19] . To translate enzyme demand into cost , the different sizes or effective lifetimes of enzymes can be considered [20] . The notion of Pathway Specific Activity [2] applies this principle to the efficiency of entire pathways ( assuming that enzyme levels are optimally distributed ) , and provides a direct way to compare between alternative pathways . However , by assuming that enzymes operate at their maximal capacity , we underestimate the true enzyme demand ( see Fig 1 ) . Enzymes typically do not operate at full capacity . This is due to backward fluxes , incomplete substrate saturation , allosteric regulation , and regulatory post-translational modifications . Below , we will refer to allosteric regulation only , but other types of post-translational regulation , e . g . , by phosphorylation , could be treated similarly . The relative backward fluxes depend on the ratio between product and substrate concentrations , called the mass-action ratio . Whenever the mass-action ratio deviates from its equilibrium value , the equilibrium constant , this deviation can be conceptualized as a thermodynamic driving force . The driving force determines the relative backward flux and thus affects reaction kinetics and enzymatic efficiency [21 , 22] . With smaller forces , the relative backward flux increases , enzyme usage becomes less efficient , and enzyme demand increases [4 , 23]—a situation that , in models , can be avoided by applying the MDF method . In fact , a cost increase due to backward fluxes can be included in the principle of minimal fluxes in FBA [17] . However , metabolites do not only affect thermodynamic forces , as acknowledged in thermodynamic FBA , but also affect kinetics as reactants and allosteric effectors . While the relative backward fluxes depend on thermodynamic forces , the forward flux depends on the availability of substrate molecules . At sub-saturating substrate levels , enzyme molecules spend some time waiting for substrate molecules , thus reducing their average catalyzed flux . Likewise , the presence of reaction product can reduce the fraction of enzyme molecules available for catalysis in the direction of pathway flux . Thus , converting metabolic fluxes into enzyme demand can be difficult because enzymes may not realize their maximal capacity . Since reduced enzyme efficiency is mostly due to metabolite concentrations , enzyme and metabolite profiles must be considered together . This quickly becomes a cyclic inference problem because steady-state metabolite levels depend again on enzyme profiles . Since many metabolites ( e . g . , co-factors like ATP ) participate in multiple pathways , enzyme demands may be coupled across the entire metabolic network . Moreover , there may be many possible enzyme and metabolite profiles that realize the same flux distribution . To determine a single solution , one can make the assumption that the most reasonable enzyme profile for realizing a given flux is the one with the minimum associated cost . This assumption may be justified if we focus on biological systems shaped by evolution , or on engineered pathways that should be efficient . A direct optimization of enzyme levels can be difficult , but there is a tractable approach in which metabolite levels are treated as free variables , which determine the enzyme levels , and therefore enzyme cost . This approach , together with a minimization of metabolite concentrations [24] , has been previously applied to predict enzyme and metabolite levels in metabolic systems [23] and to compare structural variants of glycolysis by the cost of ATP production [4] . However , to make such optimization schemes generally applicable , some open problems need to be addressed . First , our knowledge of the kinetic rate laws and parameters contains large gaps for the vast majority of enzymes [25] , and combining rate constants from different sources may lead to inconsistent models [26 , 27] . Second , the optimization problem may be computationally challenging for large networks and realistic rate laws . To turn enzyme cost minimization into a generally applicable method , we address a number of questions: ( i ) When setting up models for enzyme cost prediction , how can we deal with missing , uncertain , or conflicting data on rate constants ? Are there approximations , for example based on thermodynamics , that yield good predictions with fewer input parameters ? ( ii ) How do factors such as the kcat , driving force , or rate law affect enzyme demand , and how do they shape the optimal metabolic state ? ( iii ) How can enzyme optimization be formulated as a numerically tractable optimality problem ? Existing approaches for flux and enzyme prediction have focused on different aspects ( stationary state , energetics [8 , 28] , kinetics [23 , 29] , molecular crowding [19 , 30] , as well as enzyme cost [31 , 32] , metabolite cost [24] , or flux cost [17] ) . The new approach , which uses a modular kinetic rate law to translate fluxes into enzyme demand , shows how these approaches are logically related , and how heuristic assumptions by other methods , e . g . an avoidance of small driving forces , follow from enzyme economy as a general principle ( S1 Text section 4 ) . We show that enzyme cost minimization is closely related to cost-benefit approaches , which treat cell fitness as a function of enzyme levels [31 , 33–37] . Some general results of these approaches , e . g . , relationships between enzyme costs and metabolic control coefficients , can be reproduced . Given a pathway flux profile and a kinetic model of the pathway , one can predict the enzyme demand by assuming that cells minimize the enzyme cost in that pathway . A reaction rate v = E ⋅ r ( c ) depends on enzyme level E and metabolite concentrations ci through the enzymatic rate law , r ( c ) . If the metabolite levels were known , we could directly compute enzyme demands E = v/r ( c ) from fluxes , and similarly calculate the flux-specific enzyme demand E/v = 1/r ( c ) . However , metabolite levels are often unknown and vary between experimental conditions . Therefore , there can be many solutions for E and c realizing one flux distribution . To select one of them , we employ an optimality principle: we define an enzyme cost function ( for instance , total enzyme mass ) and choose the enzyme profile with the lowest cost while restricting the metabolite levels to physiological ranges and imposing thermodynamic constraints . As we shall see below , the optimal solution is in many cases unique . Let us demonstrate this with a simple example ( Fig 2a ) . In the pathway X ⇌ A ⇌ B ⇌ Y , the external metabolite levels [X] and [Y] are fixed and given , while the intermediate levels [A] and [B] need to be found . As rate laws for all three reactions , we use reversible Michaelis-Menten ( MM ) kinetics v ( s , p , E ) = E k cat + s / K S - k cat - p / K P 1 + s / K S + p / K P ( 1 ) with enzyme level E , substrate and product levels s and p , turnover rates k cat + and k cat - , and Michaelis constants KS and KP . In kinetic modeling , steady-state concentrations would usually be obtained from given enzyme levels and initial conditions through numerical integration . Here , instead , we fix a desired pathway flux v and compute the enzyme demand as a function of metabolite levels: E ( s , p , v ) = v 1 + s / K S + p / K P k cat + s / K S - k cat - p / K P . ( 2 ) Fig 2 shows how the enzyme demand in each reaction depends on the logarithmic reactant concentrations . To obtain a positive flux , substrate levels s and product levels p must be restricted: for example , to allow for a positive flux in reaction 2 , the rate law numerator k cat + [ A ] / K S - k cat - [ B ] / K P must be positive . This implies that [B]/[A] < Keq where the reaction’s equilibrium constant Keq is determined by the Haldane relationship , K eq = ( k cat + / k cat - ) · ( K P / K S ) . With all model parameters set to 1 , we obtain the constraint [B]/[A] < 1 , i . e . , ln[B] − ln[A] < 0 , putting a linear boundary on the feasible region ( Fig 2 ( c ) ) . Close to chemical equilibrium ( [B]/[A] ≈ Keq ) , the enzyme demand E2 approaches infinity . Beyond the boundary ( [B]/[A] > Keq ) no positive flux can be achieved ( grey region ) . Such a threshold exists for each reaction ( see Fig 2b–2d ) . The remaining feasible metabolite profiles form a triangle in log-concentration space , which we call metabolite polytope P ( Fig 2e ) , and Eq ( 2 ) yields the total enzyme demand Etot = E1 + E2 + E3 , as a function on the metabolite polytope . The demand increases steeply towards the edges and becomes minimal in the center . The minimum point marks the optimal metabolite profile , and via Eq ( 2 ) we obtain the resulting optimal enzyme profile . The metabolite polytope and the large enzyme demand at its boundaries follow directly from thermodynamics . To see this , we consider the unitless thermodynamic driving force Θ = −ΔrG′/RT [38] derived from the reaction Gibbs free energy ΔrG′ . For a given mass-action ratio Q = [B]/[A] , the thermodynamic force can also be written as Θ = ln ( Keq/Q ) , i . e . , the driving force is positive whenever Q < Keq , and it vanishes if Q = Keq . How is this force related to enzyme cost ? A reaction’s net flux is given by the difference v = v+ − v− of forward and backward fluxes , and the ratio v+/v− depends on the driving force as v+/v− = eΘ . Thus , only a fraction v/v+ = 1 − e−Θ of the forward flux acts as a net flux , while the remaining forward flux is canceled by the backward flux ( Figure A in S1 Text ) . Close to chemical equilibrium , where the mass-action ratio approaches the equilibrium constant , i . e . Q → Keq , the driving force goes to zero , the reaction’s backward flux increases , and the flux per unit enzyme level drops . This is what happens at the triangle edges in Fig 2 . Exactly on the edge , the driving force vanishes and no enzyme level , no matter how large , can support a positive flux . The quantitative cost depends on model parameters: for example , lowering a kcat value increases the cost of each enzyme unit , making the polytope boundary steeper and thus the optimum is shifted away from the boundary ( see Fig 2f and Figure B in S1 Text ) . The prediction of optimal metabolite and enzyme levels can be extended to models with general rate laws and complex network structures . In general , enzyme demand depends not only on driving forces and kcat values , but also on the kinetic rate law , which includes Michaelis-Menten constants ( KM ) and allosteric regulation . Thus , one must model these factors using the available kinetic information [39 , 40] , or approximate them when the information is not available . For some of these parameters , genome-scale prediction methods exist [41 , 42] . The rate of a reaction depends on enzyme level E , forward catalytic constant k cat + ( i . e . the maximal possible forward rate per unit of enzyme , in s−1 ) , driving force ( i . e . , the ratio of forward and backward fluxes ) , and on kinetic effects such as substrate saturation or allosteric regulation . If all active fluxes are positive , reversible rate laws like the Michaelis-Menten kinetics in Eq ( 1 ) can be factorized as [22] v = E · k cat + · η rev · η kin . ( 3 ) Negative fluxes , which would complicate this formula , can be avoided by orienting all reactions in the direction of fluxes . The reversible Michaelis-Menten rate law Eq ( 1 ) , for example , can be written in this separable form [22]: v = E k cat + s / K S 1 - k cat - k cat + p / K P s / K S 1 + s / K S + p / K P = E k cat + 1 - k cat - k cat + p / K P s / K S ︸ η rev s / K S 1 + s / K S + p / K P ︸ η kin , ( 4 ) and similar factorizations exist for reactions of any stoichiometry ( see S1 Text section 2 . 2 ) . The term E · k cat + describes the maximal reaction velocity , which is reduced , depending on metabolite levels , by condition-specific factors ηrev and ηkin ( see Fig 1b ) , accounting for backward fluxes , incomplete substrate saturation , or saturation with product ( see Table 1 for a summary of all mathematical symbols used throughout this paper ) . The reversibility factor ηrev can be expressed in terms of the driving force Θ ≡ −ΔrG′/RT by the general formula ηrev = 1 − e−Θ , which also applies to reactions with multiple substrates and products [22] . The factor ηkin depends on the rate law and thus on the enzyme mechanism considered ( see S1 Text section 2 . 2 ) . In some cases , it could be convenient to subdivide Eq ( 3 ) even further: the k cat + value can be decomposed into a product k cat + = k cat ∞ · η cat , where k cat ∞ denotes the catalytic constant of a hypothetical , infinitely fast enzyme whose rate is only limited by substrate diffusion . The enzyme-specific , constant factor η cat = k cat + / k cat ∞ is a unitless number between 0 and 1 . A realistic value of k cat ∞ = 10 8 s−1 can be obtained by considering a very fast enzymatic reaction , the breakdown of water structure around a polymer [43] . Furthermore , with some rate laws , ηkin can be further decomposed into ηkin = ηsat ⋅ ηreg , where ηreg refers to certain types of allosteric regulation ( see example in Methods ) . The factorization in Eq 3 , and any finer subdivision into factors , will lead to a subdivision of enzyme demands . Enzyme demand can be quantified as a concentration ( e . g . , enzyme molecules per volume ) or mass concentration ( where enzyme molecules are weighted by their molecular weights ) . If rate laws , fluxes , and metabolite levels are known , the enzyme demand of a single reaction l follows from Eq 3 as E l ( c , v l ) = v l · 1 k cat , l + · 1 η l rev ( Θ ( c ) ) · 1 η l sat ( c ) · 1 η l reg ( c ) . ( 5 ) To determine the enzyme demand of an entire pathway , we sum over all reactions: Etot = ∑l El . Based on its enzyme demands El , we can associate each metabolic flux with an enzyme cost q = ∑ l h E l E l , describing the effort of maintaining the enzymes . The burdens h E l of different enzymes represent , e . g . , differences in molecular mass , post-translation modifications , enzyme maintenance , overhead costs for ribosomes , as well as effects of misfolding and non-specific catalysis . The enzyme burdens h E l can be chosen heuristically , for example , depending on enzyme sizes , amino acid composition , and lifetimes ( see S1 Text section 2 . 1 ) . Setting h E l = m l ( protein mass in Daltons ) , q will be in mg protein per liter . Considering the specific amino acid composition of enzymes , we can also assign specific costs to the different amino acids . Alternatively , an empirical cost per protein molecule can be established by the level of growth impairment that an artificial induction of protein would cause [44 , 45] . Thus , each reaction flux vl is associated with an enzyme cost ql , which can be written as a function q l ( c , v l ) ≡ h E l E l ( c , v l ) of flux and metabolite concentrations . From now on , we refer to log-scale metabolite concentrations xi = ln ci in order to obtain simple optimality problems below . From the separable rate law Eq 5 , we obtain the enzyme cost function q ( x , v ) ≡ ∑ l h E l E l ( x , v l ) = ∑ l h E l · v l · 1 k cat , l + · 1 η l rev ( x ) · 1 η l sat ( x ) · 1 η reg ( x ) ( 6 ) for a given pathway flux v . If the fluxes are fixed and given , our enzyme cost becomes , at least formally , a function of the metabolite levels . We call it enzyme-based metabolic cost ( EMC ) to emphasize this fact . The cost function is defined on the metabolite polytope P , a convex polytope in log-concentration space containing the feasible metabolite profiles . Like the triangle in Fig 2 , the polytope is defined by physiological and thermodynamic constraints . It can be bounded by two types of faces: On “E-faces” , one reaction is in equilibrium , and enzyme cost goes to infinity; “P-faces” stem from physiological metabolite bounds . The shape of the cost function depends on rate laws , rate constants , and enzyme burdens , and its minimum points can be inside the polytope or on a P-face ( see Fig 2f ) . The cost function q ( x , v ) reflects a trade-off between fluxes to be realized and enzyme expression to be minimized , where the relation between fluxes and enzyme levels is not fixed , but depends on metabolite log-concentrations x . Wherever trade-offs exists in biology , it is common to assume that evolution converges to Pareto-optimal solutions [1] , i . e . cases where there are no other solution with both a higher flux and a lower cost . Therefore , we can now use this principle to predict metabolite and enzyme concentrations in cells . As with our simple model in Fig 1 , the metabolite profile that minimizes the enzyme cost for a given flux , and the corresponding enzyme profile ( computed using Eq 5 ) could be good predictions for the abundance of metabolites and enzymes in naturally evolved organisms . The resulting method , which we call enzyme cost minimization ( ECM ) , is a convex optimization problem and can be solved with local optimizers . Enzyme demand and enzyme cost functions , for single reactions or pathways , are differentiable , convex functions on the metabolite polytope . This convexity holds for a variety of rate laws , including rate laws describing polymerization reactions [46] , and even for the more complicated problem of preemptive enzyme expression , i . e . , a cost-optimal choice of enzyme levels that allows the cell to deal with a number of future conditions ( see S1 Text section 3 . 7 ) . If a model contains non-enzymatic reactions , this changes the shape of the metabolite polytope , but not the enzyme cost function , and the polytope remains convex , e . g . , if the non-enzymatic reactions are irreversible with mass-action rate laws ( see Methods ) . Obviously , metabolite and enzyme levels may be subject to various other constraints that are not reflected by our pathway model . To assess how easily the metabolic state can be adapted to external requirements , we can study the cost of deviations from the optimal metabolite levels . If the cost function q ( x ) has a broad optimum as in Fig 2 , cells may flexibly realize metabolite profiles around the optimal point , and the choice of metabolite levels may vary from cell to cell . We can quantify the tolerable variations by relaxing the optimality assumptions and computing a tolerance range for each metabolite level ( see Methods ) . To apply ECM in practice , we developed a workflow in which a kinetic model is constructed , all necessary enzyme parameters are determined by a method called parameter balancing , and optimal metabolite and enzyme levels are predicted along with their tolerance ranges . In parameter balancing [47 , 48] , a complete , consistent set of enzyme parameters is determined from measured values by employing prior distributions , parameter dependencies arising from thermodynamic laws , and Bayesian statistics ( for details , see Methods ) . Different kinds of EMC functions and constraints ( e . g . , defining concentration ranges for specific metabolites ) can be chosen . Missing data ( e . g . , KM values ) , can thus be handled in two ways: either , by using a simplified EMC function that does not require this parameter , or by relying on parameter values chosen by the workflow . Beyond minimizing the total enzyme cost , one can also use ECM to analyze the individual enzyme demands . When the metabolite levels are known , the demand can be directly calculated and each efficiency factor in Eq ( 6 ) reflects a different part of the cost ( see Methods ) . Alternatively , by omitting some factors or replacing them with constant numbers 0 < η ≤ 1 , simplified enzyme cost functions with fewer parameters can be obtained . For example , ηrev = 1 would imply an infinite driving force Θ → ∞ and a vanishing backward flux , ηkin = 1 implies full substrate saturation , as well as full allosteric activation and no allosteric inhibition ( or no allosteric regulation at all ) . In these limiting cases , enzyme activity will not be reduced , and enzyme demand will be given by the capacity-based estimate v / k cat + , a lower bound on the actual demand . Such simplifications are practical if rate constants are unknown . Depending on the data available ( e . g . , kcat values , equilibrium constants , or even KM values ) , one may choose between different cost functions with different data requirements: EMC0 ( “sum-of-fluxes-based” same prefactors for all enzymes ) , EMC1 ( “capacity-based” , setting all η = 1 and thus replacing reaction rates by the maximal velocities ) , EMC2 ( “reversibility-based”; considering driving forces , and setting ηkin = 1 ) , EMC3 ( “saturation-based” , assuming simple rate laws depending on products of substrate or product concentrations , and including the driving forces ) , and EMC4 functions ( “kinetics-based”; with dependence on individual metabolite levels ) . Details of the simplified EMC functions are given in Table 2 and Table A in S1 Text . Each EMC function is a lower bound on the subsequent functions; i . e . , even if only a simplified cost function can be used , it will always yield a lower bound on the cost computed using the full EMC4 model . Let us consider the various simplifications in more detail . If fluxes are the only data available , we may assign identical catalytic constants and burdens to all enzymes and assume that all reactions run at their maximal velocities . Then , enzyme levels and fluxes will be proportional for all reactions , the cost function in Eq ( 6 ) will be EMC0 , and the cost will be proportional to the sum of fluxes . However , catalytic constants span many orders of magnitude [25] , as do molecular masses of enzymes , suggesting that EMC0 is an oversimplification . If individual k cat + and h E l values are known , we can define an individual flux burden avlcat=hEl/kcatl+ for each enzyme , independent of metabolite levels . Then we obtain an EMC1 cost function ∑ l a v l cat v l , which is the same as the cost weights used in FBA with flux minimization [17] or molecular crowding [19] . When kcat values are unknown , they can be estimated [42] , replaced by “typical” values [25] , or bounded by the value k cat ∞ = 10 8 1/s for a very fast , but diffusion-limited enzyme . The enzyme burdens hE can include factors like protein size , protein lifetime , covalent modifications , or space restrictions ( see [20] and S1 Text section 2 . 1 ) . However , by assuming that enzymes work at their maximal rate and setting ηrev = ηkin = 1 , we may obtain unrealistic results . First , the simplifying assumption ηrev = ηkin = 1 implies uncontrollable metabolic states . In a kinetic model with completely irreversible and substrate-saturated enzymes , the reaction rates would be independent of metabolite levels and the steady-state fluxes and metabolite levels would depend on finely tuned enzyme levels [15] . Random variation in enzyme levels would lead to non-steady states , with fast accumulation or depletion of intermediate metabolites . Such states are extremely fragile and thus uncontrollable . When assuming efficiencies ηrev or ηkin smaller than 1 , we accept an increased cost and thereby acknowledge that control must be paid for by enzyme investments . Second , EMC1 functions underestimate all enzyme costs , and for reactions close to chemical equilibrium the errors may be quite large . For a reaction Gibbs energy of ΔrG′ = −0 . 1RT , the efficiency of the catalyzing enzyme is reduced by a factor of ηrev = 1 − e0 . 1 ≈ 0 . 1 , and the demand for enzyme increases by a factor of 1/ηrev ≈ 10 . To account for this decreased efficiency , we can use EMC2 functions , which include the reversibility factor η l rev = 1 - e - Θ l ( x ) . The driving forces are expressed in terms of metabolite log-concentrations Θl ( x ) and equilibrium constants , which need to be known . This factor approaches infinity as reactions reach equilibrium ( i . e . where Θl → 0 ) , which is what forces reactions away from equilibrium during cost minimization ( see , for example , Fig 2 ) . The advantage of reversibility-based cost functions ( EMC2 ) is that they are based on kcat and equilibrium constants only . Several in-silico methods exist to estimate Keq for virtually any biochemical reaction [41 , 49] and the values can be easily obtained at http://equilibrator . weizmann . ac . il/ [50] . As in the case of EMC1 , kcat values can be estimated or set to a default constant value . Methods like MDF [15] and mTOW [23] have been developed to address exactly this situation , where detailed kinetic information is hard to obtain . We discuss the relation between EMC2 and MDF in section 4 of the S1 Text . Aside from the EMC2 function , there are other reversibility-based estimates of the enzyme cost . For instance , the enzyme demand in Fig 2 ( an EMC3-function with kinetic constants , fluxes , and enzyme burdens set to 1 ) has the reversibility-based cost a v pw = ∑ l [ 1 - e - Θ ( c ) ] - 1 as a lower bound . Since 1 − e−x ≤ x for all positive x , an even lower estimate is ∑l Θ ( c ) −1 ( Figure B and Figure C in S1 Text ) . Some variants of FBA relate fluxes to metabolite profiles , which are then required to be thermodynamically feasible , i . e . , within the metabolite polytope . ECM constrains the metabolite profiles even further: as shown in Fig 2 , profiles close to an E-face are very costly and can never be optimal . This holds for EMC2 functions and for the more realistic enzyme costs , which will even be higher . Thus , regions close to E-faces can be excluded from the polytope . At P-faces , defined by physiological bounds , there will be no such increase , so the optimum may lie on a P-face ( see Fig 2f ) . To exclude regions near E-faces , we simply define lower bounds for all driving forces ( see S1 Text section 7 . 1 ) . These bounds can be used both in ECM or in thermodynamic FBA to reduce the search space . The next logical step is to relax the assumption that ηkin = 1 . Just like the reversibility factor ηrev , the kinetic factors ηsat and ηreg can be used to define tighter constraints on metabolite levels . However , unlike ηrev , the kinetic terms may take various forms and contain many kinetic parameters . To obtain simple , but reasonable formulae in EMC3 , we first consider rate laws in which enzyme molecules exist only in three possible states: unbound , bound to all substrate molecules , or bound to all product molecules . Metabolites affect the rate only through the mass-action terms S = ∏i ( si/KMi ) ( for substrates ) and P = ∏j pi/KMj ( for products ) , and the degree of saturation is determined by ηsat = S/ ( 1 + S + P ) , where the formula effectively has two Michaelis-Menten constants: one for substrates and one for products ( which are equivalent to the product of all KMi and all KMj values ) . EMC3 represents a balance between complexity and requirement for kinetic parameters , and is a practical cost function if simple , realistic rate laws are desired . The EMC4 functions , finally , represent general rate laws and ηkin can take many different forms depending on mechanism and order of enzyme-substrate binding . Again , for simplicity , we resort to analyzing only a small set of relatively general templates for EMC4 , known as convenience kinetics [51] or modular rate laws [21] . Nevertheless , our formalism allows a much wider range of rate laws , and we consider EMC4 a wild-card cost function that covers almost any reasonable rate law ( see S1 Text section 2 . 2 for more details ) . To benchmark our optimality-based prediction of metabolite , we applied ECM to a model of E . coli central metabolism , containing three major pathways: glycolysis , the pentose phosphate pathway , and the TCA cycle ( see Fig 3a , and Methods for modeling details ) . Fig 3b–3d compares predicted enzyme profiles to measured protein levels [53] . The absolute values of predicted enzyme levels arise directly from the model , using the fluxes reported in [52] , while cellular protein concentrations were obtained from proteomics data ( measured in similar conditions [53] ) and assuming an average cell volume of ~ 1 fL ( 10−15 liters ) [54] . EMC4 predicts values that are of the right order of magnitude and reflect differences in enzyme levels along the pathways . The prediction error of 0 . 42 for enzyme levels ( RMSE: root mean square error on a log10-scale ) corresponds to a typical fold error of 10RMSE = 2 . 6 . In line with the measured protein levels , the predicted enzyme levels tend to be larger in glycolysis than in TCA and pentose phosphate pathway , reflecting the larger fluxes and less-favorable thermodynamics . All predictions including metabolite concentrations , thermodynamic forces and c/KM ratios can be found online at the accompanying website www . metabolic-economics . de/enzyme-cost-minimization/ . We note that predicted enzyme levels become more accurate as more complex cost functions are used , with a prediction error decreasing monotonically from 1 . 35 with EMC0 to 0 . 42 with EMC4 . The capacity-based enzyme cost ( EMC1 ) assumes that enzymes operate at full capacity ( v = E k cat + ) and therefore underestimates all enzyme levels ( Fig 3b ) . In reality , many reactions in central metabolism are reversible and many substrates do not reach saturating concentrations . When taking these effects into account , predictions come closer to measured enzyme levels ( Fig 3c–3e ) . For instance , FUM ( fumarase , fumA ) and MDH ( malate dehydrogenase ) have a much higher predicted level in EMC2-4 than in EMC1 as the reversibility-based costs account for their low driving forces . Similarly , the predicted levels of two pentose-phosphate enzymes ( ribulose-5-phosphate epimerase RPE and ribose phosphate isomerase RPI ) are much higher in EMC3 and EMC4 because of their low affinity for the substrate ribulose-5-phosphate ( Ru5P ) . In some cases , however , the more complex EMC4 fails to improve the prediction over the simpler methods . For instance , the 6-phosphogluconolactonase ( PGL ) and phosphoglycerate kinase ( PGK ) reactions are underestimated by all EMC functions , perhaps due to regulation mechanisms that reduce activity such as allosteric inhibition . In very few cases , EMC4 overestimates the level of an enzyme that has a more precise prediction in EMC1-3 , e . g . phosphofructokinase ( PFK ) . Overall , the EMC4 function performs substantially better on average than the simpler cost functions even though it relies on a larger set of parameters , many of which are known with low certainty . Moreover , EMC4 predicts well the total of all enzyme levels ( 0 . 64 mM , compared to the measured value—0 . 62 mM ) , while the other EMC function underestimate this value ( 0 . 17 , 0 . 24 and 0 . 43 mM for EMC1 , EMC2 and EMC3 respectively ) . To test the sensitivity of our results to the choice of parameters , we performed random sampling of kinetic constants , fluxes and fixed metabolite levels , and analyzed the effect on the enzyme level predictions ( see Methods and Fig 4a ) . We further tested the sensitivity to our choice of proteomic data , by repeating the entire analysis using measured enzyme concentrations from [55] and reached essentially the same findings ( see Figure F in S1 Text ) . Finally , we tested whether our kinetic model can also predict enzyme levels without the assumption of cost optimality: to do so , we randomly sampled feasible metabolite profiles from the metabolite polytope , computed the resulting enzyme profiles , and compared them to proteomic data . It turned out that the cost-optimal metabolite profile , or similar profiles , yielded significantly better predictions than metabolite profiles sampled from a broader range ( see Methods and Fig 4b ) . This supports the hypothesis that cost-optimality shapes the metabolic state in E . coli . Although ECM puts enzymes on a pedestal due to their relatively high cost , the metabolite concentrations are key to minimizing that cost . One would thus expect to find good correspondence between the predicted metabolite profile and concentrations measured in vivo , especially when predictions of enzyme levels are good . Since some EMC functions leave metabolite levels underdetermined , we penalized very high or low metabolite concentrations by adding a second , concentration-dependent objective to the optimization problem . In particular for EMC0 and EMC1 , this regularization term is the only term—aside from global constraints—that determines the metabolite concentrations as they do not affect enzyme cost whatsoever . In all other cases , the term mostly influences metabolites that have a minimal effect on the cost . Comparing the EMC metabolite prediction with in-vivo experimental data , as shown in Figure E in S1 Text , the predicted metabolite levels are in the correct scale . Similar to enzyme level predictions , EMC4cm has the smallest prediction error—about 0 . 62 ( corresponding to a typical fold error of 4 . 1 ) . We can now use EMC analysis to rationalize cellular enzyme levels . Fig 5 ( like the scheme in Fig 1b ) shows the specific contributions to enzyme demand for each reaction . The reversibility cost terms provided by EMC2s ( purple bars in Fig 5a ) improve the enzyme demand predictions in most cases , compared to the basic capacity-based costs . However , the EMC4cm predictions show that saturation-based costs ( orange bars in Fig 5b ) are often larger than the reversibility costs , and they improve the predictions even more . For practical cost estimates , for example when computing flux burdens for FBA , we can conclude that multiplying the experimentally determined kcat values by reversibility factors will likely improve the fidelity of FBA predictions . For more details , see S1 Text section 4 . When applying mathematical models to learn about biology , one typically faces a conflict between model accuracy and the amount of available data . Metabolic systems are known to abide to several physical and physiological considerations , all of which are mathematically well-described ( e . g . flux balance , thermodynamics , kinetics , and cost-benefit optimality ) . Taking all of these aspects into account would create very detailed models but at the price of considerably increasing the demand for data . Here , we obtained a flexible modeling method by combining the two main modeling approaches , constraint-based and kinetic modeling , in a new way: with fixed metabolic fluxes , kinetic models are used to determine a cost-optimal state . The tiered approach in ECM allows for different levels of detail , which can easily be matched to the amount of existing data . The minimal requirement for running ECM is to have a metabolic network with given steady-state fluxes , while the maximal requirement would be a fully parameterized kinetic model . The method applies to individual metabolic pathways and , theoretically , entire metabolic networks . No matter if we model exponentially growing cells , microbial cells in stationary phase , or non-growing eukaryotic cells , the sum of enzyme costs per unit flux is a meaningful objective for pathways used by the cell . Although similar approaches exist in dynamic modeling [48 , 56] and enzyme optimization [4 , 15 , 23] , ECM extends these ideas to the most general kinetic rate laws and cost functions , while proving that the emerging optimization problem is convex and thus easily ( albeit numerically ) solvable . ECM advances metabolic modeling in six different ways: 1 . Solving the enzyme optimality problem in metabolite space One way of modeling the cost and benefit of enzymes is to study kinetic models and to treat enzyme levels as free variables to be optimized . However , this calculation can be hard because enzyme profiles may lead to one , several , or no steady states , and the resulting optimality problem can be non-convex . By fixing fluxes and using metabolite concentrations as our primary variables , we drastically simplify this optimization problem . Flux directions and the second law of thermodynamics impose constraints that define a set of feasible metabolite profiles , the metabolite polytope . This polytope is used here as a space for screening , sampling , and optimizing metabolic states; accordingly bounds on metabolite concentrations or driving forces can be easily formulated as linear constraints . Using log-concentrations as free variables , and given a ( steady and non-steady ) flux distribution , we can parametrize the set of metabolic states very easily: we simply consider all feasible metabolite profiles and compute , for each of them , the corresponding enzyme profile by taking the inverse rate laws . With enzyme levels as free variables , parameterizing the set of metabolic states would be much more complicated . 2 . Convexity The metabolite polytope not only provides a good search space , but it also facilitates optimization because enzyme cost is a convex function of the metabolite log-concentrations ( see S1 Text section 3 . 2 ) . Convexity makes the optimization tractable and scalable—unlike a direct optimization in enzyme space . Simple convexity holds for a wide range of rate laws and for extended versions of the problem , e . g . , including bounds on the sum of ( non-logarithmic ) metabolite levels or bounds on weighted sums of enzyme fractions . By using specific rate laws ( e . g . , the ECM4cm rate law , as shown by our colleague Joost Hulshof—personal communication ) or by adding a regularization term , representing additional biological objectives , we can even ensure strict convexity , and thus the existence of a unique optimum that can be efficiently found . It is important to distinguish this computational scalability , which is facilitated by convexity , from other pragmatic issues that arise when increasing the scale of a model , in particular the scarcity of kinetic data . Standard kinetic modeling is difficult to apply to whole-cell metabolic networks due to both scalability problems . Therefore , even if network-wide kcat and KM values were to become available ( e . g . by estimation methods that rely on high-throughput data [42] ) , it would still be impractical to exhaustively search the parameter space . ECM—due to its convexity—is solvable even on a genomic scale . 3 . Separable rate laws disentangle individual enzyme cost effects To assess how different physical factors shape metabolic states , we focused on separable rate laws , which lead to a series of easily interpretable , convex cost functions . The terms in these functions represent specific physical factors and require different kinetic and thermodynamic data for their calculation . By neglecting some of the terms , one obtains different approximations of the true enzyme cost . The more terms are considered , the more precise our predictions about metabolic states becomes ( see Methods and S1 Text section 2 ) . By comparing the different scores , we can estimate the enzyme cost that cells “pay” for running reactions at small driving forces ( to save Gibbs free energy ) or for keeping enzymes beneath substrate-saturation ( e . g . , to dampen fluctuations in metabolite levels ) . Of course , it is often important to keep models simple and the number of parameters small , and therefore the stripped-down versions of ECM can be useful in practice . For example , in some conditions such as batch-fed E . coli , a simple enzyme economy might still be a realistic approximation . Our results in Fig 3 indicate that indeed one can predict enzyme levels quite well even with relatively simple enzyme cost objectives . Finally , in conditions where ECM’s predictions are far from the measured enzyme levels , we can focus on specific enzymes or pathways that deviate the most , which may therefore display optimization or adaptations beyond simple resource allocation . 4 . Relationship to other optimality approaches Beyond the practical advantages of using factorized enzyme cost functions , they also allow us to easily compare our methods to earlier approaches . These approaches typically focused on only one or two of the factors that are taken into account in ECM , and many of them can be reformulated as approximations of ECM ( as we have shown for MDF [15] and , by proxy , earlier thermodynamic profiling methods [57 , 58] ) . For example , the optimization performed by FBA with flux minimization is equivalent to using EMC0 , while EMC1 is based on the same principles as FBA with molecular crowding [19] , pathway specific activities [2] , and Constrained Allocation Flux Balance Analysis ( CAFBA ) [59] . Thermodynamic profiling methods [15 , 57 , 58] which use driving forces as a proxy for the cost , can be compared to EMC2 ( where all kcat are assumed to be equal , see S1 Text section 4 ) . To our knowledge , ECM is the first method that accounts for substrate and product saturation ( as well as allosteric ) effects in the optimization process and guarantees a convex ( i . e . , relatively tractable ) optimality problem . Moreover , ECM highlights how different aspects of metabolism are linked: most importantly , thermodynamic feasibility [15] is generalized by the quantitative notion of thermodynamic efficiency , which then turns out to be a natural precondition for enzyme economy . 5 . Kinetics-based flux cost functions for flux balance analysis Accordingly , results from ECM can be used to improve flux analysis [13 , 23] by defining more realistic flux cost functions for FBA and by providing formulae for the pathway specific activity [2] ( see S1 Text section 2 . 3 ) . In practice , the cost weights used in FBA so far ( typically , defined by kcat values and enzyme sizes ) could be adjusted by dividing them by efficiency factors obtained from our workflow . In FBA ( specifically in variants with flux minimization or molecular crowding ) , flux cost or enzyme demand are linear functions of the fluxes . Enzyme Cost Minimization allows us to compute plausible prefactors for this formula from detailed knowledge of enzyme kinetics: by rearranging Eq ( 6 ) , we can write the enzyme cost as a linear function q = ∑ l a v l ⋅ v l with flux burdens a v l ( c ) = h E l · 1 k cat , l + · 1 η l rev ( c ) · 1 η l sat ( c ) · 1 η reg ( c ) . The flux burden has a lower bound a v l cat = h E l / k cat , l + , denoting the cost per flux under ideal conditions . Ignoring all dependencies on metabolite levels , a v l cat could be used as a cost weight to define flux cost functions for FBA . However , these values are further increased by the reciprocal values of the enzyme efficiency factors . A similar , flux-specific enzyme cost ( or , inversely , a flux per enzyme invested ) can also be defined for entire pathways . The Pathway Specific Activity ( PSA ) [2] is defined as the flux per enzyme mass ( in units of mmol/s per mg of enzyme ) and can be computed by treating enzyme mass as a cost function . Assuming that ηrev = ηkin = 1 and that cost is expressed in terms of protein mass in Daltons ( h E l = m l ) , we obtain the pathway specific activity using the formula Apw = vpw/q . 6 . Embedding ECM into flux analysis Furthermore , ECM could be “embedded” into FBA by screening a finite set of possible flux distributions , characterizing each of them by quantitative cost ( using ECM ) and choosing the most cost-favorable mode . Since we now know that any metabolic state that has maximal specific rate is an elementary flux mode [60] , it would be sufficient to scan only the elementary flux modes . This could be seen as a version of minimal-flux FBA , but one that uses kinetic knowledge instead of the various heuristic assumptions that go into FBA . Second , we can derive realistic bounds on thermodynamic forces based on kinetics and enzyme cost , or lower/upper bounds on substrates/products concentrations to avoid extreme saturation effects . All these constraints follow systematically from setting upper limits on the individual efficiency factors . By applying them in thermodynamics-based flux analysis , we shrink the metabolite polytope by excluding strips at its boundary where costs would be too high to allow for an optimal state . Similarly , by giving individual weights to thermodynamic driving forces , MDF could be used as a method to optimize some lower bound on the system’s enzyme cost ( see S1 Text section 4 ) . ECM is based on the central assumptions that the metabolic states of cells are cost-optimized and that cost arises from cellular protein levels . Both assumptions are of course debatable . There is ample evidence that cells assume apparently sub-optimal states in order to maintain robust homeostasis or to gain metabolic flexibility for addressing future challenges [1] . For example , an allosterically regulated enzyme will often not reach its maximal possible activity , so investment in enzyme production appears to be wasted . Nevertheless , cells pay this price in order to gain the ability to adjust quickly to changes ( i . e . within seconds rather than the minutes required for altering gene expression ) . One intriguing example is the bacterium Lactococcus lactis , which uses the exact same enzyme expression profile for completely different anaerobic growth modes [61]: slow growth / high yield acetate fermentation , and fast growth / low yield lactate fermentation . The reason that low-yield strategies achieve higher growth rates is typically attributed to much lower protein investments , but obviously , this is not the case in the Lactococcus lactis experiments . This stands in contrast to aerobic fermentation in E . coli , which seems to be explained well by predictable shifts in protein allocation [5] . As these examples show us , the importance that certain cells attribute to saving on protein costs is highly variable and , in some cases , can be negligible: for instance , when protein levels are already low or when protein demands change quickly and unpredictably . Moreover , random fluctuations in protein levels will be tolerable as long as the impact on fitness is not very high . Nevertheless , we think that a simple principle of cost optimality as in ECM can be a useful heuristics . On the one hand , it can reveal the minimal protein investment that would be required to support a certain metabolic state . In metabolic engineering , such predicted investments may be used to rule out potential , but uneconomical metabolic pathways . On the other hand , ECM can be used as a background model to be compared to more complicated optimality-based cell models . Such comparisons can allow us to quantify the impact of other fitness objectives in units of “protein cost” , to learn which objectives can best explain cellular behavior , and to describe non-optimality as a deviation from a presumable cost-optimal state . Furthermore , ECM can be extended to cover more realistic optimality scenarios . Some alternative objectives can be integrated into ECM by adding them to the objective function . We have tried to keep our method as general as possible to facilitate such objectives , e . g . by allowing for non-linear , convex enzyme costs ( h ( E ) ) . In particular , metabolite levels may be under additional constraints or optimality pressures because they appear in pathways outside our model , which may favor high or low levels of the metabolites . Also chemical molecule properties , such as hydrophobicity or charge , may affect the preferable metabolite levels in cells [62] . For example , if our model captures an ATP-producing pathway , low ATP levels will be energetically favorable , whereas other ATP-consuming pathways would favor higher ATP levels . To account for this trade-off , a requirement for sufficiently high ATP levels can be included in our ECM model by constraints or additional objectives b ( c ) ( x ) that penalize low ATP levels ( see Methods ) . If metabolite levels are kept far from their upper or lower physiological bounds , this will allow for more flexible adjustments in case of perturbation . If enzyme profiles were shaped by optimal resource allocation , as assumed in ECM , this would have consequences for the shapes of enzyme and metabolite profiles . Enzyme cost , thermodynamic forces , and an avoidance of low substrate levels would be tightly entangled , and the shapes of enzyme profiles would reflect the role of enzymes in metabolism , i . e . , the way in which they control metabolic concentrations and fluxes . Among other things , this would imply three general properties of enzyme profiles: 1 . Enzyme cost is related to thermodynamics In FBA , thermodynamic constraints and flux costs appear as completely unrelated aspects of metabolism . Thermodynamics is used to restrict flux directions , and to relate them to metabolite bounds , while flux costs are used to suppress unnecessary fluxes . In ECM , thermodynamics and flux cost appear as two sides of a coin . Like in FBA , flux profiles are thermodynamically feasible if they lead to a finite-sized metabolite polytope , allowing for positive forces in all reactions . However , the values of these forces also play a role in shaping the enzyme cost function on that polytope . Together , metabolite polytope and enzyme cost function ( as in Fig 2 ) summarize all relevant information about flux cost . 2 . Enzyme profiles reflect local metabolic necessities What are the factors that determine the levels of specific enzymes ? High levels are required whenever catalytic constants , driving forces , or substrate concentrations are low . Accordingly , an efficient use of enzymes requires metabolite profiles with sufficient driving forces ( for energetic efficiency ) and sufficient substrate levels ( for saturation efficiency ) . Trade-offs between these requirements , together with predefined bounds , will shape the optimal metabolite profiles [23]: in a linear pathway , a need for energetic efficiency will push substrate concentrations up and product concentrations down; the need for saturation efficiency has the same effect . However , since the product of one reaction is the substrate of another reaction , there will be trade-offs between efficiencies in different reactions . Therefore , where enzymes are costly or show low kcat values , we may expect a strong pressure on sufficient driving forces and substrate levels . 3 . Enzyme profiles reflect global effects of enzyme usage If enzyme profiles follow a cost-benefit principle , costly enzymes should provide large benefits . Such a correspondence has been predicted , for example , from kinetic models in which flux is maximized at a fixed total enzyme investment [63]: in optimal states , high-abundance enzymes exert a strong control on the flux , and enzymes with strong flux control are highly abundant . If this applies in reality , then high investment ( e . g . , large enzyme levels shown in Fig 1A ) could be seen as a sign of large benefit , in terms of flux control . Here , we studied a different optimality problem ( fixing the fluxes and optimizing enzyme levels under constraints on metabolite levels ) , and obtain a more general result . The optimal enzyme cost profile obtained by ECM is a linear combination of flux control coefficients and , possibly , control coefficients on metabolites that hit upper or lower bounds ( see S1 Text section 7 . 4 ) . In simple cases ( e . g . , the example in Fig 2 ) , where there is only one flux mode and none of the metabolites hits a bound , enzyme demands and flux control coefficients will be directly proportional . Beyond the analysis of central metabolism , ECM can be applied to select candidate pathways in metabolic engineering projects . A prediction of enzyme demands or specific activities ( S1 Text section 2 . 3 ) can be helpful at different stages of pathway design . The optimal expression profile for a pathway can be determined , critical steps in a pathway can be detected ( i . e . , steps where lowering the enzyme’s flux-specific cost a v l would be most important ) , and enzyme demand and cost can be compared between pathway structures . This type of application is not unique to ECM , and although several of the methods that we mention throughout this manuscript [2 , 4 , 23 , 32 , 64 , 65] have been used for this purpose in the past , we believe that ECM manages to bring them all under one umbrella . A metabolic network with given flux directions , equilibrium constants , and metabolite bounds defines the metabolite polytope . This convex polytope P in the space of log-concentrations xi = ln ci represents the set of feasible metabolite profiles . The flux profile used can be stationary ( e . g . determined by FBA or 13C MFA ) or non-stationary ( e . g . from dynamic 13C labeling experiments [66] ) . If the provided flux directions are thermodynamically infeasible , the metabolite polytope will be an empty set , P = ∅ . The faces of the metabolite polytope arise from two types of inequality constraints . First , the physical ranges x i min ≤ x i ≤ x i max of metabolite levels define a box-shaped polytope ( bounded by P-faces ) . Some metabolite levels may even be constrained to fixed values . Second , each reaction must dissipate Gibbs free energy , and to make this possible , driving forces and fluxes must have the same signs ( Θl ⋅ vl > 0 ) , and thus sign ( v l ) = sign ( Δ r G ′ l ∘ / R T + ∑ i n i l x i ) . The resulting constraints define E-faces of the metabolite polytope ( representing equilibrium states , Θl = 0 ) . Close to these faces , enzyme cost goes to infinity . According to Eq ( 3 ) , reversible rate laws can be factorized into four terms: the enzyme level E , its forward catalytic constants k cat + , and two efficiency factors [22] . In Fig 6 we add a non-competitive allosteric inhibitor x . While the enzyme level and k cat + are not directly affected by the concentration of metabolites ( although k cat + can vary with conditions such as pH , ionic strength , or molecular crowding in cells ) , the efficiency factors are concentration-dependent , unitless , and can vary between 0 and 1 . The reversibility factor ηrev depends on the driving force ( and thus , indirectly , on metabolite levels ) , and the equilibrium constant is required for its calculation . The saturation factor ηsat depends directly on metabolite levels and contains the KM values as parameters . Allosteric regulation yields additive or multiplicative terms in the rate law denominator , which in our example can be captured by a separate factor ηreg . In general , ηsat and ηreg can be combined into one kinetic factor ηkin , as depicted in Eq 6 . The second equation in Fig 6 describes the enzyme cost for a flux v , and contains the terms from the rate law in inverse form multiplied by the enzyme burden hE . The left-hand part of the equation , h E v / k cat + , defines a minimum enzyme cost , which is then increased by the following efficiency factors . Again , 1/ηkin can be split into 1/ηsat ⋅ 1/ηreg . By omitting some of these factors , one can construct simplified enzyme cost functions with higher specific rates , or lower enzyme demands ( compare Fig 1b ) . Since both rate and enzyme demand are a product of several terms , it is convenient to depict them as a sum on a logarithmic scale ( Fig 7 ) , where the simplified functions are seen as upper/lower bounds on the more complex rate/demand functions . Enzyme cost minimization ( ECM ) uses a metabolic network , a flux profile v , kinetic rate laws , enzyme burdens , and bounds on metabolite levels to predict optimal metabolite and enzyme concentrations . The enzyme cost of reactions or pathways is a convex function on the metabolite polytope ( proof in S1 Text section 3 . 2 ) , that is , a log-scale metabolite vector x , linearly interpolated between vectors xa and xb , cannot have a higher cost than the interpolated cost of xa and xb . Convexity also holds for cost functions h ( E ) that are non-linear , but convex over E . Some EMC functions are strictly convex ( i . e . , Eq . ( S18 ) holds with a < sign instead of ≤ ) , while others are not ( e . g . EMC2 ) . The most simplified EMC functions are actually constant ( as in EMC0 and EMC1 ) . To find an optimal state , we choose an EMC function and minimize the total enzyme cost within the metabolite polytope . Optimal metabolite profiles , enzyme profiles , and enzyme costs are obtained by solving the enzyme cost minimization ( ECM ) problem x opt ( v ) = argmin x ∈ P q ( x , v ) E opt ( v ) = E ( x opt ( v ) , v ) q opt ( v ) = q ( x opt , v ) . ( 7 ) The total cost q ( x , v ) ( defined in Eq ( 6 ) ) is the sum of enzyme costs given by EMC functions . Since q ( x ) and the metabolite polytope itself are convex , ECM is a convex optimization problem . The optimal enzyme levels depend on external conditions and have to be recalculated after any change in external metabolite levels . There are cases where q ( x ) is convex , but not strictly convex , and therefore Eq ( 7 ) will have a continuum of optimal metabolite solutions . This holds , in particular , for EMC1 scores , which are independent of metabolite levels , and for EMC2 scores , which only depend on reaction Gibbs free energies , i . e . , on some linear combinations of the logarithmic metabolite levels . In such cases , to enforce a unique solution one may add a strictly convex side objective that scores the log-metabolite levels , e . g . , a quadratic function favoring metabolite levels close to some typical concentration vector x ^ : min x ∈ P ( q ( x , v ) + λ \| \| x - x ^ \| \| ) , where λ is a small , heuristically chosen weighing factor ( see S1 Text section 3 . 3 ) . Such extra objectives can be justified biologically , e . g . by assuming that intermediate metabolite levels give cells more flexibility to adapt to perturbations . Strict convexity not only simplifies numerical calculations , but it also guarantees that the optimization problem has a unique solution . In fact , metabolite polytope and cost functions remain convex even under various modifications of the problem . When adding constraints on the total metabolite level , on weighted sums of metabolite levels , or on weighted sums of enzyme levels , the metabolite polytope is intersected with curved manifolds ( since we are dealing with concentrations in logarithmic scale ) but remains convex ( S1 Text section 3 . 4 ) . Finally , we can consider the more complicated problem of preemptive enzyme expression , where a fixed enzyme profile and allosteric inhibition must allow a cell to realize different flux distributions under different conditions . Also this problem is convex ( S1 Text section 3 . 7 ) . If a model contains non-enzymatic reactions ( or non-enzymatic processes such as metabolite diffusion out of the cell or dilution in growing cells ) , each such reaction leads to an extra constraint on the metabolite polytope ( S1 Text section 3 . 8 ) . A known flux in an irreversible diffusion or dilution reaction fixes the concentration of one metabolite . In the presence of irreversible non-enzymatic reactions with mass-action rate laws , the polytope is intersected by a subspace . In both cases , the resulting sub-polytope may be empty , i . e . , the given flux distribution will not be realizable . Flux balance analysis and kinetic models rely on the assumption that certain metabolites are mass-balanced: in FBA , this assumption , together with stationarity , defines the set of steady-state fluxes; in kinetic models , the mass-balanced metabolites are the ones whose dynamics is described by the system equations . ECM , in contrast , assumes fluxes to be given and makes no assumption about mass balances . If the fluxes in our pathway model lead to a mass imbalance in a metabolite , we may still assume that the entire cell is in stationary state , but that mass balances are reached with the help of other pathways that are not part of our model . Alternatively , we may assume that the metabolite is actually not mass-balanced and that we are describing a non-stationary , transient state . In both cases , ECM is fully applicable as long as metabolic fluxes are predefined and loop-less ( in order to be realizable by a thermodynamically consistent state [67] ) . A key point in ECM is the choice of metabolite levels on the model boundary . If we predefine all these metabolite levels , our pathway will be “isolated” from the rest of the network , and any information about the surrounding network can be safely ignored . In our E . coli model , ATP is one such important boundary metabolite: if we allowed for a lower ATP level , ATP could be produced at a lower enzyme cost because of the more favorable driving forces . If we do not fix the ATP concentration , but define an allowed range , ECM would choose the lowest possible ATP level; thus , if the allowed ATP range is too broad , no meaningful predictions can be expected . In a model of ATP-consuming biosynthesis pathways , the situation would be exactly the opposite: here , it is a high ATP level that would lead to higher driving forces and to lower enzyme requirements . Whole-cell models contain both types of pathways—ATP-producing and ATP-consuming ones . In such a model , ECM could predict some meaningful compromise , i . e . an intermediate ATP level that minimizes the enzyme cost of ATP production plus the enzyme cost of biosynthesis . Since in our central metabolism model there are only 3 reactions that produce or consume ATP , it is unlikely that so few reactions would be representative of the cost tradeoff between the dozens of enzymes that use ATP in the full metabolic network . Therefore , we chose to fix the ATP level to its measured value ( ∼3 mM ) . Our use of small-scale pathway models , which ignore most of the metabolic network , is therefore justified: as long as we predefine all metabolic fluxes and all metabolite levels at the pathway boundary , pathways can be modeled separately and the models can later be combined without any adjustment . This makes the ECM approach fully modular . All the input parameters ( kinetic parameters , fluxes and boundary concentrations ) are directly obtained from measurements without any further tuning . By setting all relevant fluxes and boundary metabolite concentrations to their measured values , we isolate our submodel from any effects that the surrounding network might have on predicted enzyme cost . Finally , there may be metabolites that are “free” in a model , but that affect enzymes that are not in the model ( e . g . pyruvate , which affects 30 other enzymes ) . By neglecting these enzymes , we ignore some of the complex compromises between them , and the predicted metabolite concentration may be wrong . In our specific case of central metabolism , whose enzymes comprise a very large fraction of E . coli’s proteome , this effect is probably not so severe . In any case , this problem can be easily fixed by imposing constraints or fixed concentrations for central metabolites such as pyruvate ( similar to how we deal with ATP and other co-factors ) . Evolution could tolerate non-optimal enzyme costs; this tolerance depends on population dynamics and can sometimes be quite significant , e . g . in small isolated communities . To compute realistic tolerance ranges for the ECM problem , we start from the optimum ( total cost q ) and choose a tolerable cost qtol ( e . g . , one percent higher than the optimal cost ) . This defines a tolerable region in P : P tol ≡ { x ∈ P \| q ( x ) ≤ q tol } . A tolerance range for each metabolite is defined by the minimal and maximal values the metabolite can show within P tol . Tolerance ranges for enzyme levels are defined in a similar way . Alternatively , tolerance ranges and nearly optimal solutions can be estimated from the Hessian matrix ( see S1 Text section 7 . 3 ) . The predicted enzyme and metabolite levels depend on the kinetic model chosen , and in particular on the kinetic constants ( kcat and KM values ) . Errors or uncertainties in these constants will cause errors or uncertainties in the predicted enzyme profiles . To estimate these uncertainties , we considered a joint distribution of all model parameters , describing both the uncertainties of individual parameters and the correlations between dependent parameters . This probability distribution was directly obtained from parameter balancing ( S1 Text section 5 . 2 ) . We sampled the kinetic parameters from this distribution , sampled metabolic fluxes according to their experimental mean values and standard deviations , and varied the fixed metabolite levels in a ± 5% range around their standard values . Then we applied ECM on each of the sampled parameter sets , and gathered statistics for the optimal enzyme and metabolite levels . Fig 4 ( a ) shows the distributions of the predicted enzyme levels . For narrow parameter distributions , the mean values , variances , and covariances of the predicted enzyme levels can even be computed , approximately , from the ECM solution with standard parameters ( see S1 Text section 3 . 5 ) . Enzyme uncertainties caused by parameter uncertainties should not be confused with the tolerance ranges described before . The tolerance ranges are always associated with sub-optimal solutions , i . e . , enzyme profiles with a higher total cost; enzyme variations caused by parameter variation , in contrast , can go both ways and may sometimes decrease the cost . Our enzyme level predictions rely on two main assumptions: a mechanistic model that defines a quantitative relation between metabolite levels , enzyme levels , and fluxes , and an optimality assumption stating that metabolite levels are optimized for a minimal total enzyme cost . To test whether such cost optimality holds in reality , we used the same mechanistic model and predicted enzyme levels based on feasible , randomly sampled metabolite profiles . We first sampled metabolite profiles around the ECM optimum by adding normally distributed random numbers ( standard deviation 0 . 05 , for metabolite levels on natural log scale ) ; then we sampled metabolite profiles in a much wider range , by sampling convex combinations of extreme points in the metabolite polytope ( i . e . , points realizing minimal or maximal values of individual metabolite concentrations ) . As shown in Fig 4 ( b ) , the metabolite profiles close to the ECM optimum yield significantly better enzyme level predictions than broadly sampled metabolite profiles . The fact that predictions from the same kinetic model , without the optimality assumption , become much worse provides strong support for cost-optimality as a principle in living cells . To predict enzyme and metabolite levels in metabolic pathways we developed an automated workflow ( Fig 8 ) . In a consistent model , all parameters must satisfy Wegscheider conditions for equilibrium constants [68] and Haldane relationships between equilibrium constants and rate constants [69] . The kinetic constants used in rate laws should represent effective parameters , which may differ from “ideal” parameters , e . g . , by crowding effects . However , since measured parameter values are usually incomplete and inconsistent , parameter balancing [47] is used to translate measured kinetic constants into consistent model parameters . Based on a network and given fluxes , the software extracts relevant data from a database ( thermodynamic constants , rate constants , fluxes , and protein sizes; metabolite and protein levels for validation ) , determines a consistent set of model parameters , builds a kinetic model , and optimizes enzyme and metabolite profiles for the EMC function chosen . To assess the effects of parameter variation , parameter sets can be sampled from the posterior distribution as described above . The workflow has been implemented in MATLAB and uses Systems Biology Markup Language ( SBML ) for model structures and the SBtab table format for numerical data [70] . The model shown in Fig 3 was built automatically from a list of chemical reactions in E . coli central metabolism ( for details , see S1 Text section 6 ) . Equilibrium constants were estimated using the component contribution method [41] , kinetic constants ( k cat + and KM values ) were obtained from the BRENDA database ( after which each value was curated manually ) , and a complete , globally consistent parameter set was determined by parameter balancing . During ECM , all metabolite levels were limited to predefined ranges , and the levels of cofactors and some other metabolites were fixed at experimentally known values . To compute tolerances for predicted metabolite and enzyme levels , we defined an acceptable enzyme cost , one percent higher than the minimal value , and determined ranges for metabolite levels that agree with this cost limit . The enzyme cost function accounts for protein composition , giving different costs to different amino acids . However , models with equal cost weights for all proteins , or with size-dependent protein costs yielded similar results ( results are provided on the website ) . Data , model , and MATLAB code for ECM can be obtained from www . metabolic-economics . de/enzyme-cost-minimization/ .	“Enzyme cost” , the amount of protein needed for a given metabolic flux , is crucial for the metabolic choices cells have to make . However , due to the technical limitations of linear optimization methods , this cost has traditionally been ignored by constraint-based metabolic models such as Flux Balance Analysis . On the other hand , more detailed kinetic models which use ordinary differential equations to simulate fluxes for different choices of enzyme allocation , are computationally demanding and not scalable enough . In this work , we developed a method which utilizes the full kinetic model to predict steady-state enzyme costs , using a scalable and robust algorithm based on convex optimization . We show that the minimization of enzyme cost is a meaningful optimality principle by comparing our predictions to measured enzyme and metabolite levels in exponentially growing E . coli . This method could be used to quantify the enzyme cost of many other pathways and explain why evolution has selected some low-yield metabolic strategies , including aerobic fermentation in yeast and cancer cells . Furthermore , future metabolic engineering projects could benefit from our method by choosing pathways that reduce the total amount of enzyme required for the synthesis of a value-added product .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "protein", "metabolism", "enzymes", "enzymology", "vertebrates", "animals", "mammals", "metabolites", "enzyme", "metabolism", "enzyme", "kinetics", "thermodynamics", "enzyme", "chemistry", "proteins", "metabolic", "pathways", "metabolism", "physics", "biochemistry", "cats",...	2016	The Protein Cost of Metabolic Fluxes: Prediction from Enzymatic Rate Laws and Cost Minimization
The skin secretion of many amphibians contains an arsenal of bioactive molecules , including hormone-like peptides ( HLPs ) acting as defense toxins against predators , and antimicrobial peptides ( AMPs ) providing protection against infectious microorganisms . Several amphibian taxa seem to have independently acquired the genes to produce skin-secreted peptide arsenals , but it remains unknown how these originated from a non-defensive ancestral gene and evolved diverse defense functions against predators and pathogens . We conducted transcriptome , genome , peptidome and phylogenetic analyses to chart the full gene repertoire underlying the defense peptide arsenal of the frog Silurana tropicalis and reconstruct its evolutionary history . Our study uncovers a cluster of 13 transcriptionally active genes , together encoding up to 19 peptides , including diverse HLP homologues and AMPs . This gene cluster arose from a duplicated gastrointestinal hormone gene that attained a HLP-like defense function after major remodeling of its promoter region . Instead , new defense functions , including antimicrobial activity , arose by mutation of the precursor proteins , resulting in the proteolytic processing of secondary peptides alongside the original ones . Although gene duplication did not trigger functional innovation , it may have subsequently facilitated the convergent loss of the original function in multiple gene lineages ( subfunctionalization ) , completing their transformation from HLP gene to AMP gene . The processing of multiple peptides from a single precursor entails a mechanism through which peptide-encoding genes may establish new functions without the need for gene duplication to avoid adaptive conflicts with older ones . In response to stress or injury , many amphibians release a viscous secretion through granular glands in their skin . In several frog families , the most abundant class of secreted molecules consists of peptides and proteins . Since the 1960s , several of these peptides have been identified as structural analogues of neurohormones that are evolutionarily conserved among vertebrates and play key roles in gastrointestinal functioning . These skin-secreted hormone-like peptides ( hereafter abbreviated as HLPs ) have been hypothesized to provide passive defense against predation , by disturbing gastrointestinal processes upon ingestion [1] , [2] , [3] . Other peptides however , were found to show little similarity to any vertebrate hormone , and their function was unclear at the time of their discovery [4] , [5] , [6] . In 1987 , two 23-AA-long peptides in the skin secretion of the African clawed frog Xenopus laevis were shown capable to kill a broad range of microorganisms [7] . Both peptides , called magainins , were cleaved from a larger precursor protein as predicted from cloned cDNA sequences [7] , [8] and are thus encoded by a single gene . The discovery of similar gene-encoded antimicrobial peptides ( AMPs ) in other amphibians fueled the perception that these animals possess a genetically controlled arsenal of antimicrobials in their skin that provides first-line protection against infectious microorganisms in their environment . AMPs are now considered key effectors of the innate immune system of many organisms , but amphibian skin secretions continue to be explored as promising sources of potential lead compounds for the development of new antibiotics . Amphibian species in which skin AMPS have been found typically secrete 5–20 different peptides [9] , [10] although in Odorrana species , the number may exceptionally exceed even 100 [11] . A recent study suggested that AMPs in distantly related anuran lineages represent independently evolved defense arsenals [12] . However , the genetic mechanisms and processes that underlay the evolutionary origin and functional diversification of any single defense arsenal remains unknown . Frogs of the family Pipidae ( including the genera Xenopus and Silurana ) possess some of the best-studied skin secretions of all amphibians [13]–[23] . The model species X . laevis for example , secretes the HLPs caerulein [24] , levitide [25] , and xenopsin [26] and apart from the two magainins , confirmed AMPs include PGLa [4] , and pGQ [27] , caerulein precursor factor ( CPF ) , and two xenopsin precursor factors ( XPF ) . cDNA sequences have shown that HLPs and AMPs in X . laevis are posttranslationally cleaved from strikingly similar precursor proteins [8] , [28] , and in some cases , both types of peptide are even processed from the very same precursor [e . g . the HLP xenopsin and the AMP xenopsin precursor fragment , the HLP caerulein and the AMP caerulein precursor fragment [5] , [25] , [29] . Consequently , these peptides are not only subject to the same gene-regulatory mechanisms , they also share interdependent evolutionary histories . Previoulsy , we identified a precursor gene in the species Silurana tropicalis with homology to the caerulein genes of X . laevis [30] . Phylogenetic analyses showed that they represented a single gene lineage , which evolved from the cholecystokinin ( cck ) hormone gene in a pipid ancestor . In the present study , we combined transcriptome , genome , and peptidome data to obtain a comprehensive overview of the entire defense peptide arsenal that evolved from this gene lineage , from the underlying genes to the active peptides . Besides identifying new peptides with potential therapeutic applications , our analyses disclose the full gene repertoire that encodes the AMP/HLP arsenal of S . tropicalis , elucidates the birth-and-death process by which it diversified , and illustrate how peptides with new functions arose in the expanding arsenal . Throughout this study , we apply the nomenclature introduced by Conlon et al . [13]–[23] to describe new peptides , and extend its use to their precursor proteins and genes . To chart the full repertoire of AMP genes in the S . tropicalis genome , we obtained new transcriptome data by preparing a cDNA library of skin samples of two S . tropicalis frogs . Comparative alignments and BLAST analyses of 384 randomly sequenced clones yielded 113 query sequences with homology to X . laevis mRNA transcripts and genes encoding known AMPs and HLPs ( Table 1 ) . This number corresponds to a remarkably high proportion ( ∼29% ) of the recovered skin transcriptome . Together , they represent nine different mRNA transcripts , one of which encodes the recently described Xt6LP precursor [30] ( Following the nomenclature of Conlon et al . , this gene is hereafter called cpf-St7 , where ‘cpf’ stands for ‘caerulein precursor fragment’ , ‘St’ stands for S . tropicalis and ‘7’ indicates that it encodes the seventh CPF peptide known for this species ) . BLAST-screening of the S . tropicalis genome using the recovered transcripts identifies a single cluster of 15 homologous genes spanning a ∼380-kb region over the scaffolds 665 and 811 ( Fig . 1A ) . They occur in both orientations , are composed of three ( cpf-St4 ) , four ( cpf-St5 , cpf-St6 , and cpf-St7 ) or five exons ( magainin-St1 , xpf-St4 , xpf-St5 , xpf-St6 , xpf-St7 , xpf-St8 , pgla-St2 , and pgla-St3 ) and show considerable variation in length , ranging from ∼3 . 6 kb ( cpf-St6 ) to over 23 . 5 kb ( xpf-St5 ) . One gene ( xpf-St8p ) shows signs of pseudogenization , with a degraded 5′-UTR and a premature stop-codon , while another ( xpf-St7p ) seems to lack a 5′-UTR and start codon altogether . The thirteen others are transcriptionally active , as evidenced by the nine query transcripts and by non-annotated EST sequences available in GenBank ( Table 1; Text S1 ) . One of these genes , cpf-St4 , lacks a terminal exon but 17 GenBank EST sequences show that it is involved in the production of alternative splicing variants , by combining the cpf-St4 exons with the terminal exon of the adjacent cpf-St5 ( Fig . 1A ) . No related genes were found on any other scaffold , suggesting that the entire AMP gene repertoire of S . tropicalis is organized in a single cluster . This cluster is flanked by the genes trak-1 , ulk-4 and ctnnb-1 on scaffold 665 , which in other vertebrates lie adjacent to cck ( Fig . 1B ) , delivering solid genomic support for the hypothesis that the pipid AMP/HLP gene family evolved from an ancestral cck gene [30] . All S . tropicalis AMP genes encode precursor proteins of 75 to 96 amino acids ( AA ) spanning exons 2–4 and containing an N-terminal signal peptide as predicted by the SignalP server [31] . Comparison of the inferred precursor proteins with previously isolated peptides from Silurana and Xenopus species allows us to predict the position and cleavage of functional peptides ( Fig . 2 ) . All precursor sequences share a region divided over exons 2 and 3 , with sequence homology to previously reported AMPs . In nearly all cases , this region is flanked by an N-terminal -RXXR- motif and a C-terminal -RXXR- , -KR- , or -RR- motif , corresponding to common cleavage sites in other vertebrate peptide precursors . In addition , several of the S . tropicalis genes , similar to those of X . laevis , seem to encode secondary peptides , some of which show homology to known HLPs . The C-terminal region of prepro-CPF-St6 shows sequence similarity to the HLP caerulein , as previously reported for prepro-CPF-St7 [30] . Prepro-XPF-St4 and prepro-XPF-St5 share a C-terminal region with similarity to the X . laevis peptide levitide [25] , and prepro-XPF-St7 , prepro-XPF-St7p and prepro-XPF-St8 show similarity to peptide phenylalanine-glutamine-amide ( pFQa ) , a 14-AA-long peptide of Silurana epitropicalis [16] . Nano-liquid chromatography tandem mass spectrometry data ( nanoLC-MS/MS ) from skin extracts confirm that S . tropicalis synthesizes a peptide arsenal that parallels the AMP/HLP diversity of X . laevis . Screening of these spectra against a database composed of AMP/HLP precursor proteins predicted from cDNA and gene sequences confirms the cleavage and posttranslational processing of 11 predicted peptides , while providing support ( albeit with insignificant identity scores ) for two more ( Table 2 , Text S2 ) . These include nine of the predicted AMP homologues , two levitide-related peptides ( Levitide-St1 and Levitide-St2 ) and a single homologue of pFQa ( pFQa-St1 ) . Our analysis additionally indicates the processing of a secondary peptide ( PLD-St1 ) from prepro-PGLa-St3 without any apparent homology to known peptides . Visual inspection of this precursor confirms the proximity of an N-terminal -KR- site ( Fig . 2 ) . This cleavage site is separated from PLD-St1 by the dipeptide Met-Ala , indicating that a cleaved 16-AA-long peptide may be further processed by exopeptidases to obtain PLD-St1 . Posttranslational modifications of individual residues are observed in six of the cleaved peptides , all of which undergo C-terminal amidation , and two of which undergo N-terminal pyroglutamate formation ( Table 2 ) . The majority of the predicted peptides show the typical structural features of amphibian AMPs ( Table 3 ) . Most of them have the potential to form an alpha-helix according to secondary structure predictions [32] , are cationic at a pH of 7 . 0 , and show an alternated sequence of hydrophylic/cationic and hydrophobic AAs , which in an alpha-helical configuration results in a strong amphipathic structure . Upon contact with cell membranes , the cationic side of the helix allows interaction with negatively charged phospholipid heads on the membrane surfaces , while the hydrophobic side facilitates alignment with , or intrusion into the intermembrane region , potentially inducing membrane pore formation and eventually , cell lysis [33] , [34] . It was recently argued that AMPs , besides being effectors of innate immune response , may play a role in antipredator defense through their cytolytic effects [35] . We therefore investigated the activity of 17 peptides by concentration-dependent assays in liquid media against both microorganisms and vertebrate cells . Target cells included two gram-negative bacteria ( Escherichia coli and Pseudomonas aeruginosa ) , two gram-positive bacteria ( Staphylococcus aureus and Micrococcus luteus ) , a fungus ( Saccharomyces cerevisiae ) , a protozoan parasite ( Trypanosoma brucei ) , mouse ( Mus musculus ) red blood cells , and mouse spleen cultures ( containing a mixture of T-lymphocytes , macrophages and myeloid cells ) . The variable nature of these cells required the use of different measures of peptide activity ( see Materials and Methods ) but combined , their results support a number of distinct patterns . First , 13 of the 17 synthesized peptides show antimicrobial activity as evidenced by minimum inhibitory concentrations ( MIC ) . These include 12 peptides homologous to known AMPs , and PFQa-St2 ( Table 4 ) . In all cases examined , subculturing from MIC solutions to agar media lacking peptides did not yield any colonies , confirming that growth of the population was inhibited by effectively killing the cells . Second , ten of the peptides are to some extent capable of inducing lysis of red blood cells . Observed hemolytic activity is generally relatively low , with peptide concentrations inducing 50% hemolysis ( HC50 ) ranging down to 64 µM . In addition , 12 peptides are capable to inhibit the Concanavalin A-stimulated proliferation of T-lymphocytes in spleen cell cultures . Peptide concentrations inducing 50% inhibition of T-cell proliferation ( IC50 ) range down to 2 µM ( CPF-St4 ) while higher concentrations typically eliminate any proliferation . Concanavalin A induces T-cell proliferation indirectly , by stimulating macrophages to activate T-cells . Consequently , the peptides may inhibit proliferation directly ( by suppression or lysis of T-cells ) , indirectly ( by inhibition or lysis of macrophages ) , or both . These results confirm that amphibian AMPs may indeed contribute to antipredator defense through cytolytic and/or immunosuppressive effects . Third , the tested peptides show notable variation in activity against any single cell type . For S . aureus for example , MIC values range from >512 µM ( e . g . , magainin-St1 ) down to 1 µM ( e . g . , CPF-St5 ) . For T . brucei , a single peptide may additionally show variation in activity across repeated assays . This pattern may be related to the use of pleomorphic parasite cultures , representing a variable mixture of two parasite forms ( “long-slender” and “short-stumpy” ) with potentially different susceptibility to membrane permeabilization . Fourth , our assays do not provide evidence of target cell specificity or complementary target spectra among different peptides as observed in other frogs ( e . g . , [36] ) . Although there is substantial variation in the sensitivity of cell types to the AMPs , ( e . g , the gram-positive M . luteus is highly sensitive to most peptides ) , the measured activities of peptides across cell types are correlated . In other words , peptides either seem effective against the broad range of cell types ( e . g . , CPF-St5 ) or hardly effective at all ( e . g . , magainin-St1 ) . Previous studies have demonstrated that relatively weak AMPs from the same amphibian species may yield an increased antimicrobial effect when applied in combination [11] , [36] , [37] , [38] . We searched for the existence of such synergistic effects among the S . tropicalis AMPs by MIC assays against S . aureus in which peptides were pairwise combined in a 1∶1 stochiometry . Our analyses reveal a single case of synergistic activity: combination of two of the weakest AMPs , magainin-St1 and PGLa-St1 , yields a combined MIC of 64 µM ( 32 µM for each peptide ) , while individually , magainin-St1 had no apparent effect on S . aureus ( MIC>512 µM ) and PGLa-St1 had a very weak effect ( MIC = 512 µM ) . The phylogeny of the pipid AMP/HLP arsenal is consistent with a birth-and-death model of gene evolution , similar to those implicated for vertebrate immune gene families [58] and reptile toxin genes [59] , [60] , in which some gene lineages survive for a prolonged time , and may give rise to additional genes through subsequent duplication , while others are eventually lost . Gene loss is confirmed by the presence of several incomplete genes in the S . tropicalis cluster , one of which shows clear signs of pseudogenisation ( cpf-St8b ) , and the apparent absence of orthologues of xpf-St1 and xpf-St6 in X . laevis . Comparative peptidome analyses of several Silurana and Xenopus species have recently indicated that ancient genome duplication events did not result in an increased AMP diversity in allopolyploid species [16] , [17] , [20] . This observation led to the conclusion that polyploidization events were compensated by subsequent gene losses , counterbalancing the expected doubling of the AMP arsenal . Our analyses confirm similar levels of AMP diversity in S . tropicalis and X . laevis but also reveal that the comparable arsenals in both taxa arose under contrasting patterns of gene evolution . The diploid S . tropicalis has the largest AMP/HLP gene repertoire ( due to higher rates of tandem gene duplication or lower rates of gene loss ) but each gene encodes a single AMP and/or a single HLP . Instead , X . laevis , despite being tetraploid , has fewer genes ( due to lower rates of gene duplication or increased gene loss ) , but several of them ( e . g . the magainin and caerulein genes ) are characterized by multiple tandem-repeated peptide–encoding exons . The difference between both patterns may have been maintained by the self-sustaining nature of tandem repeats ( whether genes or exons ) , because as their number increases , so does the probability of unequal crossing-over among neighboring copies , potentially creating additional duplicates of the same type . Exon duplication is considered an important mechanism of transcriptional economy by which peptide diversity can be increased , either through alternatives splicing [61] or by differential cleavage of tandem-encoded peptides [62] . However , duplicated exons in the caerulein and magainin genes show quasi-zero sequence divergence ( Fig . 4 ) , possibly reflecting very recent duplication events , extremely low evolutionary rates , or gene conversion . Alternative splicing therefore adds little to peptide diversity , but in the X . laevis genes encoding multiple tandem-repeated exons may provide a way to boost the synthesis of peptides at a low transcriptional cost . Our phylogenetic analyses allow us to formulate an evolutionary scenario for the timing of origin and loss of specific peptide types and their corresponding functions in light of major structural changes in the pipid AMP gene repertoire ( Fig . 6 ) . The first major step in the evolution of the peptide arsenal entails the transition of a neurohormone ( CCK ) to a basal defense peptide ( Figs . 6A , 6B ) . The HLP caerulein , encoded by three genes in X . laevis , contains the same bioactive site as CCK and shares its capacity to bind vertebrate CCK receptors , thereby inducing pancreatitis , vomiting , diarrhea , hypotension , and inhibition of exploratory and feeding behavior [63] . In addition , C-terminal sequences homologous to caerulein are retained in the S . tropicalis precursors prepro-CPF-St6 and prepro-CPF-St7 . Together , these Xenopus and Silurana genes represent an early-diverged lineage in the pipid AMP gene tree . It is therefore likely that caerulein retained the basal defense function of the cck-derived gene family . Its origin was accompanied by a gene duplication event ( gene duplication 1 in Fig . 6B ) and a shift in expression from the gastrointestinal tract and brain to the granular skin glands . The taxonomic distribution of newly identified promoter elements ( Fig . 3 ) provide an indication that at least part of the changes required for this expression shift may have happened long before the cck gene duplication , implying that the ancestral cck gene was “preconditioned” to acquire a skin-secretory function . A second major step involves the origin of a cationic alpha-helical AMP in the central region of the ancestral precursor protein , leading to a bifunctional AMP/HLP gene product with an architecture similar to , e . g . the present day prepro-CPF-St6 ( Figs . 6A , 6B ) . This step is marked by truncation of the N-terminal spacer region , and the origin of an additional exon , encoding the C-terminal half of pipid AMPs . In contrast , an evolutionary conserved -RXXR- site , inducing the N-terminal cleavage of the longest functional CCK-isoforms in vertebrates ( e . g . CCK-58 in humans ) has been preserved in the majority of pipid AMP precursors , where it induces cleavage of the functional AMPs . Third , after establishment of an antimicrobial function , the basal CCK-like defense function was lost twice independently ( Fig . 6A ) , yielding monofunctional AMP precursors with an architecture similar to e . g . , the present-day prepro-CPF-St5 precursor ( Figs . 6A , 6B ) . One loss occurred in the gene ancestral to cpf-St4 and cpf-St5 , while a second loss happened in the gene ancestral to magainin-St1 through pgla-St3 . In both cases , the loss of the CCK-like function followed gene duplication events and were accompanied by the 5′-truncation of the genes' last exon , eliminating the sequence encoding the CCK-like bioactive site . Fourth , after loss of the basal CCK-like defense function , new AMPs and HLP functions ( pFQa-like peptides , levitide-like peptides , PLD-St1 ) , arose alongside the antimicrobial function , again yielding a bifunctional gene . All of these secondary peptides are encoded by an extra exon ( exon 4 in xpf-St4 , -St5 , -St-6 , -St7 , and -St8 , and pgla-St2 , and -St3 ) , but due to their lack of structural similarity , it is unclear whether they originated independently , or share a common origin . The latter would imply extraordinary structural diversification of the peptides due to extremely high substitution rates in exon 4 . On at least two occasions , evolution of the pipid AMP genes gave rise to striking cases of evolutionary convergence in peptide sequence . One case is represented by caerulein , which independently evolved to an identical structure in Litoria frogs , starting from a different neurohormone ( gastrin ) [30] . A second example is represented by xenopsin , which according to our phylogenetic analyses evolved from a levitide-like peptide ( Figs . 4 , 6A , 6B ) . Apart from AA substitutions , this evolutionary process involved a 7-AA deletion in the peptide and loss of C-terminal amidation ( Fig . 6C ) . Xenopsin shows strong structural and functional similarities to the hormones neurotensin ( NT ) and xenopsin-related peptide 2 ( XRP-2 ) , two universal NT receptor agonists in vertebrates [26] , [64] , [65] . These hormones share five and six C-terminal AAs with xenopsin , respectively , all of which are involved in neurotensin receptor binding [65] . Effects of NT and XRP-2 include stimulation of exocrine pancreatic secretion , suppression of food intake , increase of vascular permeability ( enhancing inflammation responses and hypotension ) , and stimulation of histamin release by mast cells ( enhancing an allergenic reaction ) [66] , [67] . While NT is cleaved from its own precursor along with the hormone neuromedin N [68] , XRP-2 is cleaved from the N-terminal region of Coatomer Subunit alpha ( CopA ) , a 138-kDa intracellular protein [69] . Because xenopsin is unrelated to either , its evolution is likely to reflect adaptive convergence to neurotensin receptor binding . Toxin gene families showing complex functional diversity among their members in venomous animal taxa have been used as model systems to investigate the role of gene duplication and diversifying selection in functional innovation , and lended support to different theoretical models of neofunctionalisation [59] , [70] , [71] , [72] . The pipid AMP/HLP arsenal provides an opportunity to extend the same theoretical framework to amphibian defense peptide arsenals and presents an empiricial illustration of how peptide-encoding genes may acquire new functions independent of gene duplication . Current theoretical models invoke gene duplication as a key factor of functional innovation . Under the original model of neofunctionalisation ( “mutation during nonfunctionality” , [73] ) gene duplication delivers a functionally redundant gene , which , released of selective constraints to maintain the ancestral function , accumulates mutations that lead to a new function . Alternative models instead invoke subfunctionalisation following gene duplication to divide multiple functions of the parental gene over the duplicates in a nonadaptive way ( “duplication-degeneration-complementation” ; [74] ) , or to resolve adaptive conflicts between different ancestral functions ( “escape from adaptive conflict” [75] , [76] ) . Although gene duplication of the cck gene allowed the origin of a CCK-like defense function in a pipid ancestor , it did not trigger subsequent functional innovation . Instead , new functions recurrently arose by mutation of a precursor protein , resulting in the posttranslational processing of a secondary peptide ( an AMP ) alongside the original one ( e . g . a CCK-like HLP ) . The resulting bifunctional gene is consistent with the starting condition in several theoretical models invoking subsequent gene duplication [74] , [76] , [77] . However , pipid AMP/HLP genes deviate from these models ( or represent a special case ) by attaining multifunctionality through the production of multiple peptides , rather than a single peptide or protein with multiple functions . One implication of this mechanism is that genes in the AMP/HLP arsenal acquired new functions while being under purifying selection to retain an older one . This could have been possible by mutations in a gene region that is subject to only marginal purifying selection [10] . Selection on a gene to retain its original function is likely to affect only regions of the gene that are crucial for the proper expression , processing and functioning of the peptide [10] , [78] . In the case of the Pipid AMP/HLP gene repertoire , this is confirmed by purifying selection on gene regulatory elements , and exon regions encoding signal peptide , cleavage sites and the CCK-like bioactive site ( Figs . 3 and 5 ) . The evolution of a new functional peptide in precursor regions under relaxed selection thereby not avoids adaptive conflicts with the original peptide but may even rely on the gene regions under purifying selection to obtain an adaptive value . Another implication is that selection on a gene to retain its original function is likely to affect the probability of possible new functions to arise . In pipid frogs , purifying selection to retain a gene with all the necessary features to produce a skin-secretory peptide , increases the probability that any new function will be related to skin secretion as well . Therefore , while representing a red line through the evolutionary history of the pipid AMP gene repertoire , skin gland expression acted as a key determinant of functional innovation . Although gene duplication did not act a the direct trigger of functional innovation , it may have subsequently still led to subfunctionalisation [74] , [75] , [76] , by relaxing selective pressures to retain both functions in each daughter gene . Subfunctionalisation may explain the convergent loss of the original CCK-like defense function in multiple gene lineages after duplication , completing the final step in their functional transformation from HLP gene to AMP gene ( Fig . 6B ) . The pipid defense peptide arsenal illustrates an evolutionary mechansim independent of gene duplication that could explain the presence of precursor proteins yielding multiple peptides with distinct structures and bioactivities in the poison or venom glands of various animals . Transcripts of a single gene isolated from Bombina toad skins encodes both bradykinin peptides and a structurally unrelated bradykinin inhibitor [79] , while natriuretic peptide precursors produced in the venom glands of vipers and Heloderma lizards are further cleaved to obtain angiotensin-converting enzyme ( ACE ) inhibitors , and bradykinin receptor antagonists , respectively [80] , [81] . Furthermore , the evolution of multiple peptides within a single precursor , when followed by gene duplication and subfunctionalisation , provides an explanation for the existence of closely related genes encoding structurally and functionally distinct peptides . Homologous genes encoding fundamentally different peptides , like AMPs , bradykinin-like HLPs and opioid peptides , have been identified in ranid and hylid frogs [3] , [10] , [82] , [83] . A recent evolutionary study of snake venom metalloproteinases ( SVMPs ) showed how differential protein domain loss following gene duplication gave rise to proteins with distinct toxin functions [70] . This study implied that the ancestral SVMP toxin was composed of multiple domains , each with a distinct function , but did not explain how these domains were combined in a single protein . The experiments , maintenance and care of mice complied with the guidelines of the European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes ( CETS n° 123 ) . The experiments for this study were approved by the Ethical Committee for Animal Experiments of the Vrije Universiteit Brussel , VUB , Brussels , Belgium ( Permit Number: 08-220-8 ) . Freshly dissected skin tissue of S . tropicalis was snap-frozen in liquid nitrogen , ground to powder with mortar and pestle , and stored at −80°C . mRNA was extracted using the Qiagen Oligotex mRNA midi kit and cDNA libraries were made using the Clontech Creator SMART cDNA library construction kit . Plasmids with cDNA inserts were cloned in Oneshot Electrocompetent GeneHog E . coli cells ( Invitrogen Corp . ) and 384 randomly selected clones were sequenced by the Australian Genome Research Facility . Clones corresponding to AMP-encoding transcripts were identified by BLAST searches based on sequence homology to mRNA and gene sequences encoding known peptides in X . laevis . Subsequent online translation of these transcripts yielded the corresponding precursor sequences . Functional peptides in the precursors were predicted by comparison with previously described peptides from various Silurana and Xenopus species . The S . tropicalis AMP gene repertoire was charted by BLAST-searches against the Xenopus tropicalis 4 . 1 genome ( DOE Joint Genome Institute; http://genome . jgi-psf . org/Xentr4/Xentr4 . home . html; X . tropicalis is the previous name of S . tropicalis ) using the newly obtained transcripts as queries . Based on a large sequence overlap including pgla-St2 and pgla-St3 , we concatenated genomic scaffolds 665 and 811 . We assume that these scaffolds were kept separate during contig assembly due to sequence gaps whose lengths were improperly estimated , resulting in apparent sequence conflicts . Intron/exon boundaries of identified AMP genes were checked by verifying the presence of GT/AG intron borders flanking the inferred exons . Freshly dissected dorsal skin tissue of S . tropicalis frogs were placed in a 90∶9∶1 ( v∶v∶v ) CH3OH/H20/HCOOH solution and stored at −20°C . After sonication ( 5 times 2 min ) and centrifugation ( 10 min at 9000 g ) , the supernatant was lyophilized ( SpeedVac Concentrator ) . Dry peptide extracts were reconstituted in 100 µL of 0 . 1% aqueous HCOOH and lipids were removed by re-extraction with equal volumes of ethyl acetate and n-hexane respectively . The sample was filtered through a 0 . 22 µm spin-down filter ( Ultrafree-MC; Millipore ) and 1/10th of the sample was analyzed by nanoLC-MS/MS on an Ultimate 3000 Nano and Capillary LC System ( Dionex ) coupled to a microTOF-Q mass spectrometer ( Bruker Daltonics GmbH ) . The peptides were separated on a Pepmap C18 column ( 3 µm , 75 µm×150 mm , LC Packings ) using a 50 min linear gradient from 95% solvent A and 5% solvent B to 50% solvent A and 50% solvent B at a flow rate of 200 nl/min ( solvent A: deionised water containing 0 . 1% HCOOH; and solvent B: CH3CN containing 0 . 1% HCOOH ) . In the mass spectrometer , doubly or triply charged ions of sufficient abundance are selected for fragmentation by the software ( MS/MS ) . [84] . MS/MS peak list files were submitted to an in-house version of MASCOT server ( Matrix Science , USA ) and screened against a database of predicted AMP/HLP precursor protein sequences . This way , the actual cleavage and posttranslational processing of predicted peptides from the precursors was verified . The resulting spectra and corresponding tables are provided in Text S2 . Secondary structure of AMPs was predicted using a neural network algorithm as implemented on the Psipred server [32] . Peptides predicted to have an alpha-helical structure were projected in a wheel projection to visualize their amphipathic nature . Hydrophobic moments were calculated using the combined consensus scale implemented in HydroMCalc [85] . Seventeen predicted and recovered peptides were de novo synthesized using solid-phase technology by CASLO Laboratory ApS ( Lyngby , Denmark ) and delivered as HPLC-purified ( >95% ) trifluoracetate ( TFA ) salts . The peptide salts were stored as 5 . 12 mM stock solutions in 0 . 01% ( V/V ) acetic acid/0 . 2% ( m/V ) BSA . Peptides were tested against Escherichia coli ( ATCC 25922 ) and Pseudomonas aeruginosa ( ATCC 15692 ) , Staphylococcus aureus ( ATCC 25923 ) , Micrococcus luteus , Saccharomyces cerevisiae , a protozoan parasite ( Trypanosoma brucei brucei ) , mouse ( Mus musculus ) red blood cells and mouse spleen tissue cultures ( containing 50–60% B-cells , ∼25–30% T-cells , and 5% myeloid cells ) . All assays were performed in duplicate or triplicate . Activity against the bacteria and fungus was measured by assessing the lowest peptide concentration in a series of twofold dilutions at which no growth was detected ( known as the minimum inhibitory concentration , MIC ) . Bacterial and yeast cultures ( 5×105 colony forming units/ml ) were prepared in Müller-Hinton ( MH ) broth and Luria-Bertani ( LB ) broth respectively and transferred to serial dilutions of peptides ranging from 512 or 256 µM down to 1 or 0 . 5 µM in 96-well polypropylene plates . After incubation at 37°C ( bacteria ) or 30°C ( S . cerevisiae ) for 18 hours , growth of the cultures was checked by eye . In addition to MIC , we determined minimum microbicidal concentrations ( MMC ) by inoculating samples of the MIC assays on agar plates lacking peptides and checking for overnight growth . In all cases , MIC and MMC were identical . Activity against T . brucei parasites was assessed as the minimum concentration in a series of twofold dilutions required to kill 95% of parasites in 30 minutes ( LC95 ) . Frozen stabilates of Trypanosoma brucei brucei AnTat1 . 1 bloodstream parasites were expanded by infection of C57Black/6 mice ( Janvier ) . Mice with systemic parasitaemia ( typically 4–5 days post infection ) were exsanguinated and parasites were purified from heparinized blood by DEAE-cellulose ( DE52 , Whatman ) chromatography . Collected parasites were washed twice with Phosphate-Saline-Glucose buffer ( PSG-buffer ) and enumerated microscopically using a Bürker hematocytometer . The DEAE52-purified parasites ( stock: 106 parasites/ml PSG buffer containing 5% FCS ) were added to polypropylene 96-well plates containing the serial peptide dilutions to obtain total volumes of 200 µl . After incubation for 30 minutes , trypanolytic activity was assessed using a light microscope by calculating the percentage of dead parasites in a minimum of 200 counted cells . Activity against mouse red blood cells was examined by assessing the lowest peptide concentration in a series of twofold dilutions causing at least 50% hemolysis ( HC50 ) . Briefly , heparinized blood was obtained by cardiac puncture of mice euthanized via CO2 . The blood was diluted 1/100 in RPMI and added to polypropylene 96-well plates containing the serial peptide dilutions to obtain total volumes of 100 µl . The plates were incubated at 37°C in a humidified atmosphere for 30 minutes and centrifuged at 1400 rpm for 2 minutes to pellet intact red blood cells allowing visual observation of hemolysis . Supernatants were subsequently transferred to 96-well flat-bottom plates to allow spectrophotometry using an ELISA reader ( OD measured at 550 nm ) . The percentage of hemolyis for each peptide concentration was calculated as 100× ( ODobs−OD0% ) / ( OD100%−OD0% ) , where ODobs is the OD measured for the peptide concentration , OD0% is the average OD in the absence of peptides ( 0% hemolysis ) , and OD100% is the average OD in the presence of 1% Tween-20 ( 100% hemolysis ) . Activity against T-lymphocytes was assessed as the lowest peptide concentration in a series of twofold dilutions causing at least 50% inhibition of Concanavalin A-induced T-cell proliferation ( IC50 ) . Solutions containin 2×105 naive C57Black/6 splenocytes in the presence of 2 . 5 µg Concanavalin A were added to polypropylene 96-well plates containing the serial peptide dilutions to obtain total volumes of 100 µl . After incubation for 24 hours at 37°C in a humidified atmosphere , the cells were pulsed with [3H]thymidine and incubated for an additional 18 hours . The amount of [3H]thymidine incorporation was assessed by a β-counter as a measure of cell proliferation and the percentage of proliferation inhibition for each peptide concentration was calculated as 100× ( 1− ( Tobs/T100% ) ) , where Tobs is the [3H]thymidine level measured for the peptide concentration , and T100% is the [3H]thymidine level measured for Concanavalin-A-induced cell proliferation in the absence of peptide ( representing 100% proliferation ) . Intergenic regions in the cluster were screened for the presence of evolutionary conserved elements ( phylogenetic footprints ) using the program Tracker [86] . Tracker identified the gene promoter regions as only conserved non-repeat regions in between adjacent AMP genes . The 500-bp upstream regions of all S . tropicalis AMP genes , four X . laevis AMP genes and the cck genes of five vertebrates were aligned using the EINSI algorithm in MAFFT 6 . 704 [87] . Candidate regulatory elements in the gene promoter regions were identified by: ( 1 ) determining aligned sites with more than 75% sequence conservation across all AMP genes , and ( 2 ) by searching for significantly overrepresented sequence motifs in the promoter regions using MEME 4 . 8 . 0 [42] . AMP gene sequences retrieved from the S . tropicalis genome were combined in a single dataset with related gene and mRNA sequences of X . laevis retrieved from GenBank . Five amphibian cck gene or mRNA sequences were added to serve as outgroups . Phylogenetic relationships were estimated using the conventional ‘two-step’ approach , involving sequence alignment and tree reconstruction as separate steps of phylogeny inference , and a ‘direct optimization’ approach that accounts for alignment uncertainty by integrating sequence alignment and tree construction in a single algorithm . For the two-step approach , sequences were aligned using MAFFT . To avoid the erroneous alignment of nonhomologous sequences , exons were aligned separately and subsequently concatenated . Phylogenetic analyses were performed with MrBayes 3 . 1 . 2 [49] using a GTR+G+I model of DNA substitution . Two parallel runs of four incrementally heated ( temperature parameter = 0 . 2 ) Markov chain Monte Carlo ( MCMC ) chains were performed , with a length of 10 , 000 , 000 generations , a sampling frequency of 1 per 1 , 000 generations , and a burn-in corresponding to the first 2 , 000 , 000 generations . Convergence of the parallel runs was confirmed by split frequency standard deviations ( <0 . 01 ) and potential scale reduction factors ( approximating 1 . 0 ) for all model parameters , as reported by MrBayes . Adequate posterior sampling was verified using Tracer 1 . 5 [88] , by checking if the runs had reached effective sampling sizes >200 for all model parameters . Bayesian analyses under direct optimization were conducted with BAli-Phy 2 . 0 . 2 [50] . Alignment constraints were imposed to maintain gene alignments that respect exon boundaries . The data set was analyzed under a GTR+G+I model of DNA substitution combined with a RS07 model of DNA insertion/deletion . Four independent analyses of a single MCMC chain each were run for ten million generations , and alignments and trees were sampled every 1000 generations . Again , convergence of the runs , and effective sampling size of the log-likelihood values and model parameters was checked using Tracer . We investigated whether diversifying ( positive ) or purifying ( negative ) selection affected the evolution of coding sequences in the expanding gene cluster using the random effects likelihood ( REL ) method , as implemented in the HYPHY software package [89] Kosakovsky Pond et al . , 2005] . This method is considered the most suitable for datasets <50 sequences , draws sitewise synonymous and nonsynonymous codon substitution rates from separate rate heterogeneity distributions , and implements an empirical Bayes approach to test for significant selection at any specific site [90] . As such , REL represents an extension of the methods implementend in the benchmark program PAML [91] . Analyses were conducted using a MG94 codon substitution model ‘crossed’ with a GTR nucleotide substitution model , and codon sites were identified as subject to significant diversifying or purifying selection at Bayes factors >50 . Due to the gain , loss and truncation of exons in the history of pipid AMP genes , the selection analyses were conducted on separate data sets for exon 2 ( combining CCK and AMP gene sequences ) , exon 3 ( composed of AMP gene sequences only ) and the last exon ( composed of CCK gene sequences and the homologous sequences of four AMPs that encode a CCK-like bioactive site ) .	Many amphibians defend themselves against predation and infections by secreting a mixture of gene-encoded toxins and antimicrobials . How does such an integrated defense weapon arise and how does it diversify to gain distinct antipredatory and antimicrobial functions ? We took advantage of the availability of a sequenced genome for the African clawed frog Silurana tropicalis to provide the first comprehensive overview of an amphibian peptide defense arsenal , from its underlying genes to its bioactive components . A reconstruction of the evolutionary history of this gene repertoire allows us to elucidate the timing and mode of evolution of distinct defense functions . Our study shows that the basal transition from a gastrointestinal hormone function to a skin-secretory defense function was accompanied by major restructuring of regulatory sequences in the ancestral gene . Instead , subsequently diversifying defense genes underwent functional shifts by entering a bifunctional stage ( by cleavage of two distinct defense peptides from a single precursor protein ) and occasionally losing the original defense function ( by loss of the original defense peptide ) . This pattern provides an evolutionary explanation for the processing of structurally or functionally unrelated toxins from the same or closely related precursor proteins in other poisonous and venomous animals .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "evolutionary", "biology", "zoology", "biology", "genomics" ]	2013	Origin and Functional Diversification of an Amphibian Defense Peptide Arsenal
Myelin is required for proper nervous system function . Schwann cells in developing nerves depend on extrinsic signals from the axon and from the extracellular matrix to first sort and ensheathe a single axon and then myelinate it . Neuregulin 1 type III ( Nrg1III ) and laminin α2β1γ1 ( Lm211 ) are the key axonal and matrix signals , respectively , but how their signaling is integrated and if each molecule controls both axonal sorting and myelination is unclear . Here , we use a series of epistasis experiments to show that Lm211 modulates neuregulin signaling to ensure the correct timing and amount of myelination . Lm211 can inhibit Nrg1III by limiting protein kinase A ( PKA ) activation , which is required to initiate myelination . We provide evidence that excessive PKA activation amplifies promyelinating signals downstream of neuregulin , including direct activation of the neuregulin receptor ErbB2 and its effector Grb2-Associated Binder-1 ( Gab1 ) , thereby elevating the expression of the key transcription factors Oct6 and early growth response protein 2 ( Egr2 ) . The inhibitory effect of Lm211 is seen only in fibers of small caliber . These data may explain why hereditary neuropathies associated with decreased laminin function are characterized by focally thick and redundant myelin . Myelin is essential for rapid impulse propagation and the proper function of the nervous system . Schwann cells ( SCs ) form myelin in peripheral nerves in 2 subsequent steps , radial sorting of axons and myelination . During radial sorting , immature SCs segregate axons with a diameter larger than 1 μm to the edge of embryonic axon bundles and then acquire a 1:1 relationship with these axons and differentiate into promyelinating SCs . Immature SCs express the transcription factor Oct6 ( Pou3f1 ) that is later downregulated [1 , 2] , while promyelinating SCs express the transcription factor early growth response protein 2 ( Egr2 or Krox20 ) that is necessary to transition into wrapping and myelination [3] . A signaling pathway consisting of 3′-5′-cyclic adenosine monophosphate ( cAMP ) and protein kinase A ( PKA ) , possibly via regulation of Nfkb and Oct6 , is required to achieve full Egr2 activation and myelination [2 , 4–7] . Egr2 , in turn , is the master regulator of myelin protein and lipid genes [reviewed in 8] . Following Egr2 activation , myelin-forming SCs start to elaborate a myelin sheath around axons . Myelin thickness depends on the number of myelin wraps that a SC makes around an axon and is correlated to axon diameter [9] . Signaling molecules that include Disc Large MAGUK scaffold protein 1 ( DLG1 ) and Phosphatase and Tensin homolog ( PTEN ) are then required to terminate wrapping [10–12] . Finally , groups of small axons that are not radially sorted into a 1:1 ratio remain ensheathed by nonmyelinating SCs , which organize the associated axons into a Remak bundle . These developmental steps are regulated by the axonal growth factor neuregulin 1 type III ( Nrg1III ) and by the extracellular matrix ( ECM ) component laminin α2β1γ1 ( Lm211 ) . However , how these 2 signals are integrated is unknown . Laminins in the basal lamina are required for radial sorting by enabling cytoskeletal rearrangements needed for changes in SC morphology [13 , 14] . The major SC laminin is composed of α2 , β1 , and γ1 chains , encoded by Lama2 , Lamb1 , and Lamc1 genes , respectively . Mutations or targeted inactivation of Lama2 result in radial sorting defects [13 , 15 , 16] , as does inactivation of the genes encoding the Lm211 receptors a6β1 , a7β1 integrins , and dystroglycan [17–19] . Whether Lm211 also controls the initiation of myelin wrapping after radial sorting is unclear . Experiments in vitro indicate that Lm211 promotes the initiation of myelination [20 , 21] , but this could simply represent Lm211 enabling the prerequisite step of radial sorting . Lama2 deletion in vivo does not prevent myelination , but this could be explained by compensation by other laminins that do not contain the α2 chain [16 , 22] . In addition , it has been reported that Lm211 positively regulates myelin thickness [23] , but in contrast with this finding , laminin signaling via the α6β4 integrin receptor was recently shown to inhibit myelination via serum and glucocorticoid-induced kinase 1 [24] , and human biopsies from patients with LAMA2 mutations show thickened and irregularly folded myelin sheaths [25 , 26] . Thus , the role of Lm211 in myelination remains unresolved . On the other hand , it is well established that the level of axonal Nrg1III is a key instructive signal for myelination [27 , 28] and regulates myelin thickness [27–29] . Nrg1III promotes myelination by engaging the ErbB2/ErbB3 receptor tyrosine kinase , which in turn stimulates several signaling pathways , including PI3K/Akt , Calcineurin and MAPK/ERK , which are thought to converge on Oct6 and Egr2 activation . Although Nrg1III has a clear role in the initiation of myelination , its role during axonal sorting has not been proven directly . A role for Nrg1III in radial sorting is suggested by the observation that SCs fail to ensheathe Nrg1III-deficient axons in vitro and by the fact that Nrg1III haploinsufficiency causes defects in axon ensheathment in Remak bundles [27 , 30] . Thus , the role that Nrg1III plays during axonal sorting is unclear . Finally , how Nrg1III signaling from axons is integrated with Lm211 signaling from the basal lamina is poorly understood . Here , using genetic epistasis experiments in vivo and in vitro , we show that Lm211 and Nrg1III interact functionally . By modulating the expression of Nrg1III and Lm211 , we show that in early development and in small fibers , Lm211 inhibits Nrg1III-driven myelination , but this effect is only revealed when the amount of Nrg1III is also altered . How does Lm211 inhibit Nrg1III ? cAMP and PKA are required in parallel to Nrg1III to initiate myelination , by functioning as a switch that triggers Egr2 expression , but cAMP/PKA are not required for Egr2 maintenance [4 , 5 , 31] . Here , we show that Lm211 inhibits PKA activity and various signaling steps downstream of Nrg1III . We propose that Lm211 limits Nrg1III signaling via PKA to prevent precocious myelination during radial sorting , inappropriate myelination of small fibers that normally do not become myelinated , and excessive myelin wrapping of small myelinated fibers . It is known that laminins in the SC basal lamina are necessary for radial sorting [32–35] , that Lm211 promotes myelination in vitro [21] , and can even replace ascorbic acid to induce myelination [20] . While this has been interpreted as Lm211 promoting wrapping and myelination , Lm211 may instead only enable SCs to arrive into a 1:1 relationship with axons ( promyelinating stage ) , which then leads to myelination upon Nrg1 activation independently of laminins . To begin to make this distinction , we first asked if Lm211 could lead to myelination without Nrg1III . We took advantage of the ability of Lm211 to replace ascorbic acid in inducing myelination in vitro and asked if wild-type SCs treated with Lm211 could myelinate Nrg1III-deficient neurons . As reported [20] , SCs myelinate wild-type neurons when exogenous laminin is added to the media ( Fig 1D ) , while no myelination was present without the addition of Lm211 ( Fig 1A ) . However , Lm211 cannot induce myelination in the absence of Nrg1III ( Fig 1J ) . Fewer SCs were observed in culture with Nrg1III deficient axons , but Lm211 increased the number of SCs ( Fig 1M ) . This was not accompanied by an increase in SC proliferation ( Fig 1N ) but could be due to improved association of SC with Nrg1III-deficient neurons ( Fig 1O–1R ) . This experiment shows that Lm211 signals need to converge with those activated by Nrg1III for myelination to proceed in vitro . Myelination is preceded by radial sorting . While it is evident from the literature that Lm211 mutants have impaired radial sorting , a role for Nrg1III in this process is unclear . To test if Lm211 and Nrg1III cooperate to regulate radial sorting , we performed a genetic interaction experiment and asked if partial loss of Nrg1III ( Nrg1III heterozygous null mice , as constitutive nulls are embryonic lethal ) worsen the radial sorting defects of Lama2−/− mice . We quantified radial sorting defects in sciatic nerves at postnatal day 16 ( P16 ) because Lama2−/− mice in the C57/BL6 background die around P21 . At P16 , radial sorting is completed in normal sciatic nerves , all axons larger than 1 μm have been myelinated , and small bundles , hardly visible by semithin sections ( Fig 2A ) , contain small caliber axons that are beginning to differentiate into mature , nonmyelinated Remak fibers . By electron microscopy ( EM ) , the majority of axonal bundles in wild-type nerves contained fewer than 50 axons that were usually smaller than 1 μm ( Fig 2B , 2C and 2D ) . As previously reported , Lama2−/− sciatic nerve presents sizable bundles of axons visible by semithin sections ( Fig 2A , arrow ) , which contain large numbers of naked axons including those with diameters >1 μm ( Fig 2B , asterisks , Fig 2C and 2D ) , a sign of impaired axonal sorting . Nrg1III+/− nerves contained abnormally ensheathed , small bundles of axons ( Fig 2B , red asterisk ) , as previously described [27] , but no significant increase in the number of naked axons per bundle ( Fig 2B and 2C ) or in the number of axons with diameter >1 μm in bundles ( Fig 2D ) . Importantly , Nrg1III haploinsufficiency did not increase the percentage of unsorted axons in Nrg1III+/−//Lama2−/− nerves ( Fig 2 ) . Finally , the levels of laminin α2 protein were normal in Nrg1III+/− sciatic nerve at P3 and P16 ( Fig 2E ) , indicating that Nrg1III does not regulate Lm211 expression . We conclude that the partial loss of Nrg1III in vivo does not cause significant defects in radial sorting nor does it further aggravate radial sorting phenotypes due to loss of Lm211 . We next asked if Lm211 and Nrg1III interacted genetically at later timepoints , i . e . , during myelination , using the same animals . Axonal Nrg1III levels regulate myelin thickness: Nrg1III haploinsufficiency causes hypomyelination ( thin myelin ) while Nrg1III overexpression causes hypermyelination ( thick myelin ) [28] . In contrast , the role of laminin in regulating the onset or extent of myelination is unclear . Based on our initial hypothesis that Lm211 and Nrg1III are both promyelinating signals , we asked if loss of Lama2 further reduced myelination in Nrg1III+/− mice . As reported , sciatic nerves from Nrg1III+/− mice showed thinner myelin ( Fig 3A and 3B ) . In contrast , sciatic nerves from Lama2−/− mice displayed no significant changes in myelin thickness . To our surprise , nerves from Nrg1III+/−//Lama2−/− mice showed a return of myelin thickness close to wild-type levels ( Fig 3A and 3B ) . When the average g-ratio was plotted against the diameter of the fibers , it became clear that the rescue was mainly due to an effect on axons smaller than 2 μm ( Fig 3C and 3D ) . We confirmed that the distribution of axon diameters was not altered in the different genotypes ( Fig 3E ) . This experiment shows that loss of Lama2 suppresses the hypomyelination phenotype of Nrg1III+/− mice in small fibers . Thus , in a genetic context in which Nrg1III-induced myelination is reduced , the role of Lm211 on myelination becomes evident . These data suggest a repressive role for Lm211 in the myelination of small caliber axons . So far , we showed that Lm211 may inhibit Nrg1III-induced myelination ( Fig 3 ) , although we could not reveal an interaction between the 2 molecules during radial sorting ( Fig 2 ) . To further explore these results , we crossed mice that overexpress Nrg1III ( Nrg1IIItg ) [36] with Lama2−/− mice and analyzed sciatic nerve morphology at P16 by EM . Nrg1IIItg nerves had normal Remak bundles ( arrow in B ) and no abnormal unsorted bundles of axons ( Fig 4A , 4B and 4C–4C” ) . Unexpectedly , Nrg1IIItg//Lama2−/− sciatic nerves had more severe radial sorting defects than Lama2−/− mice ( Fig 4A ) , with a 3-fold increase in the number of unsorted axon bundles per sciatic nerve cross section ( Fig 4C ) . This difference was present already at early stages in P5 sciatic nerves ( S1A and S1B Fig ) . Thus , Nrg1III overexpression enhances the Lama2−/− phenotype . Phenotypic enhancement suggests that the overexpressed gene , in our case Nrg1III , stimulates a pathway that is inhibited by the loss-of-function gene , in our case Lama2 [37 , 38] . Thus , a possible explanation is that Lm211 limits Nrg1III signaling pathways , thereby preventing precocious myelination of small-diameter fibers prior to the completion of axonal sorting and formation of an appropriate 1:1 relationship . In agreement with this hypothesis , unsorted bundles in sciatic nerves of Nrg1IIItg//Lama2−/− mice often contained axons that were myelinated by SCs before reaching the promyelinating 1:1 stage ( Fig 4D , arrows ) . Perturbed SC number can cause radial sorting defects [39] . Since both Lm211 and Nrg1III influence SC proliferation and survival [16 , 40] , we asked if these parameters were synergistically altered in the double mutants and could explain the severe radial sorting phenotype . During radial sorting at P5 , double mutants showed no increase in the percentage of TUNEL-positive nuclei or statistically significant decrease in phosphorylated-histone3 ( P-H3 ) positive nuclei , and cell density was not changed ( S1C Fig ) . At P16 , there was a trend for increased apoptosis and decreased proliferation in the double mutants , but the changes were minimal ( less than 0 . 4% apoptotic cells ) and did not reach statistical significance . At P16 , we also detected a decrease in cell density in Nrg1IIItg mice , probably due to thicker myelin , and an increase in cell density in Nrg1IIItg/Lama2−/− , probably due to the reduction of myelinated fibers ( S1D Fig ) . Overall , differences in SC number do not appear to be a major cause for the increased radial sorting defects observed in Nrg1IIItg/Lama2−/− mice . Despite the severe radial sorting defects described in Nrg1IIItg//Lama2−/− nerves , some axons were myelinated , giving us the opportunity to measure myelin thickness . As before , Lama2−/− nerves at P16 had normal g-ratios , and , as reported , Nrg1IIItg sciatic nerves had decreased g-ratios due to increased myelin thickness ( Fig 5A and 5B ) . In the double Nrg1IIItg//Lama2−/− nerves , the overall average g-ratio was intermediate between wild-type and Nrg1IIItg ( Fig 5A and 5B ) , but plotting the g-ratio as function of the axon diameter revealed that removal of Lm211 further decreased the g-ratio of small fibers while progressively restoring to normal values the g-ratio of larger fibers ( Fig 5C and 5D ) . This indicates that , as seen in Nrg1III+/− mice ( Fig 3 ) , inhibition of Nrg1-induced myelination by Lm211 is predominant in small fibers . Notably , axons much smaller than 1 μm , which should not be myelinated , were often surrounded by a thick myelin sheath in double mutant nerves ( Fig 5F ) . In Nrg1IIItg//Lama2−/− nerves , even large fibers with normal or thin myelin sheaths often displayed abnormal and redundant myelin , with infolding and signs of myelin degeneration ( Fig 5G ) . These dysmyelinating features were also occasionally observed in Nrg1IIItg animals and are characteristic of certain forms of hereditary neuropathies , including those associated with deficiency of Lm211 [26] . Taken together , these data further substantiate the notion that Lm211 inhibits Nrg1III to prevent inappropriate myelination of small , unmyelinated axons and to limit myelin thickness in small caliber axons and the formation of redundant myelin in general . The phenotypic enhancement shown by the genetic experiments described above suggests that Lm211 inhibits a pathway or a substrate that is normally stimulated by Nrg1III to promote myelination . In an attempt to find the pathway or substrate that it is inhibited by Lm211 , we first reevaluated published work and performed experimental analysis that showed that Lm211 does not inhibit ERK or Akt ( S2 Fig ) . We next turned to PKA , which is required in parallel to Nrg1III to achieve full Egr2 activation and myelination [4–6] . PKA is a good candidate molecule because its hyperactivity causes a phenotype similar to that observed in Nrg1IIItg/Lama2−/− mice: an arrest in radial sorting with some promyelinating SCs undergoing premature myelination [41] . In addition , PKA may be activated by Nrg1 [42 , 43] and by Gpr126 , a g-coupled protein receptor that binds various ligands , including Lm211[44 , 45] . We hypothesized that PKA , or one of its substrates , may be normally inhibited by Lm211 and that the phenotype of Nrg1IIItg/Lama2−/− mice may be due to excessive Nrg1III-driven promyelinating signals , plus disinhibited PKA signaling ( Fig 6A ) . To test this idea , we first evaluated the amount of substrates phosphorylated by PKA in sciatic nerves using an antibody that recognizes the PKA-phosphorylated consensus motif RxxS/T ( p-Sub antibody ) . This revealed a discrete number of bands , many of which were upregulated in nerves deficient in Lm211 but not in those with Nrg1III overexpression ( Fig 6A ) . We also measured PKA activity directly in sciatic nerves at P5 and P16 and confirmed that PKA was hyperactive in the absence of Lm211 at both timepoints ( Fig 6B and S2 Fig ) . In contrast PKA activity was normal in Nrg1IIItg at P16 , higher at P5 , and normal in Nrg1III+/− nerves , suggesting that Nrg1III may not regulate PKA in SCs in vivo . PKA is regulated by levels of cAMP or by lipids and peptides in a cAMP-independent fashion [46–49] . To determine if the hyperactivity of PKA in Lm211 null SCs was caused by an increase in cAMP , we measured cAMP concentration in sciatic nerves of mutant mice . Interestingly , the levels of cAMP at P16 and P5 were low in Lama2−/− nerves and normal in Nrg1IIItg ( S2 Fig ) . Overall , these results indicate that Lm211 inhibits PKA activation , possibly by a cAMP-independent mechanism . We next investigated which steps of the Nrg1III signaling cascade were influenced by the Lm211 and PKA axis . In vitro , PKA phosphorylates ErbB2 [50] , thus , one PKA substrate that accumulates in Lama2−/− nerves could be the Nrg1 receptor ErbB2/ErbB3 itself . Interestingly , sustained treatment of cultured SCs with a PKA-selective agonist increased phospho-ErbB2 in a dose-dependent manner , in the absence of Nrg1 in the culture media ( Fig 6C ) . In contrast , an agonist of exchange protein directly activated by cAMP ( EPAC ) did not increase phospho-ErbB2 . Higher doses and longer treatment were required to activate ErbB2 , similar to the conditions required to promote SC differentiation and Egr2 expression [5 , 51 , 52] , while short treatments did not activate ErbB2 , as previously reported [50] . The PKA-selective agonist also induced phosphorylation of Grb2-Associated Binder-1 ( Gab1 ) , an adaptor protein that is phosphorylated upon Nrg1III/ErbB signaling in SCs and is required for myelination [53] ( Fig 6C ) . These data suggest that PKA can directly transactivate the ErbB2-Gab1 axis independently of Nrg1III . It is known that cAMP-PKA is also required to amplify Nrg1 signals in SCs [4–6] . To confirm this , we exposed primary SCs to either Nrg1 alone or Nrg1 and dbcAMP and showed that the phosphorylation of ErbB2 and Gab1 were enhanced if the SCs were exposed to both Nrg1 and dbcAMP ( Fig 6D ) . To confirm that Gab1 phosphorylation was downstream of Nrg1-ErbB signaling , we pretreated SCs with the ErbB2 inhibitor PKI166 [54] and showed that Gab1 phosphorylation was inhibited . In contrast , treatment with the Src-kinase inhibitor PP2 did not have any effect ( Fig 6D ) . Thus , PKA can directly activate ErbB2 , and cAMP sensitizes the response of SCs to the Nrg1III-ErbB2-Gab1 pathway , at least in vitro . To ask if ErbB2 and Gab1 were modulated by Lm211 in vivo , we next measured their phosphorylation status in mutant sciatic nerves . Strikingly , phosphorylation of ErbB2 and Gab1 was not increased in nerves of Nrg1IIItg , probably due to the presence of an intact Lm211 “brake;” however , deleting Lm211 in the context of Nrg1III overexpression ( Nrg1IIItg/Lama2−/− ) significantly increased ErbB2 and Gab1 phosphorylation ( Fig 6E and 6F ) . A similar trend was observed at P5 ( S4 Fig ) . Overall , these data support the view that Lm211 , via inhibition of PKA , reduces the output of Nrg1 signaling in SCs in vivo . We next tested if the Oct6 and Egr2 transcription factors , downstream of PKA and Nrg1 , were modulated in our system . By western blot ( WB ) , the levels of both Oct6 and Egr2 were increased in Lama2−/− and double mutant nerves at P16 ( Fig 7A and 7B ) , but not in Nrg1IIItg nerves , suggesting as before that Lm211 inhibition has to be released to drive excessive Nrg1III-induced SC differentiation . A similar trend was observed for Egr2 in double mutants earlier in development , corroborating that SCs may initiate premature differentiation ( S4 Fig ) . The number of Oct6 and Egr2 positive nuclei were also increased in Lm211-deficient nerves ( Fig 7C–7E ) , but the number of Egr2 positive nuclei was decreased in double mutants , likely due to the arrested development with a reduced number of SCs reaching the promyelinating stage ( see Fig 4 ) . Both cAMP and Nrg1 are required to induce sustained expression of Egr2 in SCs in culture [5 , 6 , 55] , and cAMP-PKA increases ErbB2 phosphorylation both in response to Nrg1 [50] and independently of Nrg1 ( Fig 6C ) . Therefore , we hypothesized that PKA could also induce Oct6 and Egr2 expression independently of Nrg1 . To test this , we asked if treatment of SCs with a PKA agonist induced Oct6 and Egr2 protein levels . Indeed , exposure of SCs to the specific PKA agonist induced both transcription factors ( Fig 7F ) . This induction was present in the absence of Nrg1 , and was not seen with an EPAC-specific agonist . To test if ErbB2 phosphorylation was required for the induction of Oct6 and Egr2 , we treated rat SCs with dbcAMP and analyzed Oct6/Egr2 expression after inhibition of ErbB2 with a specific inhibitor , PKI166 . ErbB2 inhibition caused the expected block in dbcAMP-induced Gab1 phosphorylation , but did not alter Oct6 or Egr2 levels ( Fig 7G ) . In contrast , PKA inhibition with H89 suppressed the induction of both transcription factors by dbcAMP ( Fig 7H ) . These results suggest that in cultured SCs , PKA may activate Oct6 and Egr2 independently of Nrg1 . We showed that in Nrg1IIItg//Lama2−/−nerves , ErbB2 and Gab1 are more active , and this is associated with an increase in PKA activation due to Lm211-deficiency . To determine if the effect of Lm211 loss is indeed mediated by PKA activation ( Fig 8A ) , we inhibited PKA activity in vitro and in vivo using the selective PKA antagonists H89 and KT5720 and asked if this was sufficient to decrease ErbB2 and Gab1 phosphorylation . In cultured SCs , H89 , dose-dependently suppressed the activation of ErbB2 and Gab1 ( Fig 8B ) . In vivo , we injected H89 and KT5720 beneath the gluteus superficialis and biceps femoris muscles , in which the sciatic nerve resides , every day from P3 to P6 and sampled the nerves at P7 . The contralateral side was injected with DMSO and used as control . This procedure has been shown to effectively deliver pharmacological treatment within sciatic nerves [56] and indeed we could observe a reduction of PKA substrate phosphorylation in nerves treated with the inhibitors ( Fig 8C ) . Strikingly , the inhibitors significantly decreased ErbB2 and Gab1 activation in Nrg1IIItg//Lama2−/− mice , indicating that PKA directly contributes to the promyelinating signals initiated by Nrg1III in Schwann cells ( Fig 8D and 8E ) . The expression of Oct6 and Egr2 instead could not be consistently modulated by this short pharmacological treatment ( S5 Fig ) . Taken together , our results strongly suggest that Lm211 , through inhibition of PKA , limits the activation of promyelinating signaling molecules such as ErbB2 and Gab1 in SCs . Overall , based on our data , we conclude that Lm211 inhibits Nrg1III via PKA in several instances: during radial sorting , to prevent premature SC differentiation; at the onset of myelination , to prevent myelination of fibers smaller than 1 μm; and during myelination , to limit myelin thickness in small fibers ( Fig 9 ) . Our data clarify for the first time that Lm211 modulates and can even inhibit , rather than promote , myelination in vivo . SC development depends on a discrete number of extrinsic signals originating from the axon and the ECM [58 , 59] . How SCs coordinate these different signals to achieve myelination is poorly understood . Here , we focused on 2 of the major extrinsic signals: the axonal molecule Nrg1III and the ECM component Lm211 . Although these molecules have been known for years to be important for SC development , it was not known how these signals on 2 opposite surfaces of the SC collaborate to achieve myelination . We show that Lm211 has an inhibitory role on several downstream effectors of the Nrg1 pathway . We show that this effect is mediated by inhibition of PKA , that PKA is hyperactive in the absence of Lm211 , and that this leads to overactivation of ErbB2 and Gab1 when combined with Nrg1III overexpression , resulting in an arrest of radial sorting and in premature myelination . Taken together , our results strongly indicate that Lm211 limits PKA activation and blocks this parallel pathway that needs to converge with Nrg1III to initiate myelination . Our data begin to clarify how Nrg1III can regulate such different SC responses: proliferation , survival , and myelination , using the single receptor ErbB2/ErbB3 . We propose that deposition of Lm211 in the basal lamina modulates the SC response to axonal Nrg1III to favor proliferation , survival , and axonal ensheathment during radial sorting while inhibiting myelination , effectively modulating the Nrg1 response during development from a proliferative to a myelinating signal . Based on our data , Lm211 limits the response of SC to Nrg1III at multiple steps of development . In immature SCs , Lm211 promotes radial sorting and inhibits Nrg1-driven premature myelination . In promyelinating SCs , Lm211 prevents Nrg1-driven inappropriate myelination of axons smaller than 1 μm and limits myelin thickness on small axons ( Fig 9 ) . Thus , although Lm211 function predominates during radial sorting and Nrg1 predominates during myelination , both are required to fine-tune the onset and extent of each developmental step . During radial sorting , the levels of Nrg1III on axons are probably read by ErbB2/3 receptors as a binary fate choice ( myelination versus nonmyelination ) : above a threshold of Nrg1III SCs start myelination , as increasing the expression of Nrg1III can switch the fate of nonmyelinated axons to myelinated [27] . However , our data indicate that Lm211 increases the Nrg1III threshold required for myelination . In contrast , expressing a dominant-negative form of integrin β1 in oligodendrocytes increases the threshold for myelination , but it is unclear if this is mediated by the Lm211 ligand and if this function is linked to a modulation of Nrg1III signaling also in the central nervous system [60] . As summarized above , we provide evidence that Lm211 and Nrg1III have reciprocal roles during radial sorting and myelination . This could suggest that there is a reciprocal inhibition also from Nrg1III to Lm211 , and that this inhibition may be required to terminate radial sorting driven by laminin . Nrg1III could conceivably suppress the sorting behavior of SCs also via PKA ( dotted-arrow in S3B Fig ) . This could explain the drastic impairment in radial sorting observed in Nrg1IIItg//Lama2−/− mice , and there is evidence in the literature indicating that Nrg1 activates PKA [42 , 43] . However , our results in vivo were conflicting , with Nrg1III overexpression increasing PKA activity at P5 but not at P16 , and Nrg1III haploinsufficency not reducing PKA activation . Thus , our data did not allow us to conclusively confirm this idea . Our data indicate that myelin thickness is also modulated by Lm211 , but interestingly , the effect is opposite in small- and large-caliber fibers . This could be potentially be explained by the fact that Lm211 uses different receptors ( α6β1 and α6β4 integrin , dystroglycan , Gpr126 ) that may exert positive and negative effects , which when removed together by virtue of removing the common Lm211 ligand , do not affect myelin thickness . It is also interesting to note that the axonal sorting defects in Lm211 mutants are more severe in motor than sensory roots , suggesting a higher dependence of motor fibers on Lm211 . It follows that it will be interesting to explore if the different effects of Lm211 on myelin thickness of large and small fibers may coincide with a different effect of laminin–Nrg1III interactions on motor versus sensory nerve fibers . Finally , the regulation of myelin thickness by Nrg1III is also more evident in small fibers [28] . Up to now , the role of Lm211 in the control of myelin thickness was controversial , and this can be explained by the fact that Lm211 has different effects in small and large fibers and that the modulating role of Lm211 cannot be revealed when the Nrg1III axis is intact . Why were laminins considered only as a promoter of myelination throughout the years ? Multiple experiments were interpreted to show that laminin is necessary for SC to myelinate in vitro because formation of myelin was used as an endpoint [20 , 21 , 32] , rather than considering radial sorting and myelination as 2 distinctive steps in development . Indeed , when radial sorting was examined by EM in these studies , SCs were blocked at the immature and not at the promyelinating stage . Similarly , in vivo , it was reported that ErbB2 phosphorylation was decreased in the nerves of mice lacking all laminins in SCs [35] . SCs in these mutant mice are arrested at the immature stage , and they are more undifferentiated than their wild-type counterparts , likely explaining why ErbB2 phosphorylation appeared to be decreased . Finally , it was previously reported that loss of Lm211 leads to a reduction in myelin thickness in the same Lama2−/− animal model that we used [23] . This discrepancy can be explained by the fact that myelin thickness , rather than g-ratio was measured , using an automated program and light microscopy . In our hands , only the analysis of measurement of g-ratio using EM could reliably and consisistently reveal the changes in g-ratio in small caliber fibres . Also , in previous studies , myelin thickness was not evaluated as a function of axonal diameter , potentially confusing the results , based on the differences that we have observed between small- and large-caliber fibers . This , together with the fact that normal myelin thickness was also reported in other Lm211 mutants generated in the past [15 , 16 , 61] , make us confident of our conclusion that Lm211 deletion alone is not sufficient to influence myelin thickness . The molecular mechanisms through which Lm211 inhibits PKA and PKA promotes myelination are only partially understood . Lm211 binds Gpr126 , a G protein-coupled receptor that increases cAMP signaling and is required for peripheral myelination [44 , 45] . One of the downstream effectors of GPR126-cAMP is PKA , and together they are required to activate Egr2 expression and initiate myelination in a Nrg1III-dependent manner [4] . We originally postulated that Lm211 decreases PKA activity by regulating Gpr126 and cAMP . Gpr126 binds Lm211 to regulate the release of an inhibitory fragment with context-dependent effects on the levels of cAMP[45] . However , cAMP levels were decreased in mutants lacking Lm211 , suggesting that the net effect of Lm211 on Gpr126 is stimulatory for cAMP production . Lm211 may indirectly favor binding of Gpr126 to the activating ligands collagen IV [57] and cellular prion protein on axons [62] . Collagen IV binding may depend on Lm211 because laminins favor basal lamina polymerization [63 , 64] , and prion binding may depend on Lm211 for proper radial sorting and contact between SCs and axons [15] . On the other hand , our finding that cAMP was low in Lm211-deficient nerves also suggests that the increased activation of PKA is cAMP-independent . There are several examples of cAMP-independent PKA activation in other cell types [46–49] , and Lm211 receptors such as α6β4 integrins and dystroglycan could potentially be involved [24 , 65] . The mechanism by which PKA regulates myelination has been the subject of recent work , and it is only partially clarified . PKA in SCs regulates the cytoskeleton [66] , signaling molecules , and transcription factors , such as members of the CREB family and Egr2 [4 , 5] . How PKA induces Egr2 is unclear . One possibility is by activating NfkB and inducing Oct6 . A cytoplasmic pool of PKA phosphorylates the p65 NfkB subunit on Serine 276 in SC [7] , and , interestingly , this regulation may be cAMP-independent [67 , 68] . NF-kb is required for axonal ensheathment and activation of Oct6 and binds the chromatin remodeler Brg1 , which is essential for myelination [69 , 70] . Oct6 , in turn , activates Egr2 expression [71 , 72] . Therefore , it is tempting to speculate that Gpr126 , cAMP , PKA , NfkB , and Oct6 are all part of a transient , linear pathway that switch-on Nrg1III-driven myelination , and that Lm211 , collagen IV , and prion proteins modulate it . Our data have implications for human diseases . Loss-of-function mutations in LAMA2 causes Congenital Muscular Dystrophy 1A , which include demyelinating peripheral neuropathies characterized by heterogeneous myelin thickness with focal hypermyelination , loss of nerve fibers , short internodes , and wide nodes of Ranvier [26 , 73] . While the mechanisms of the short internodes and wide nodes of Ranvier have been clarified [74–76] , the molecular basis of the focal hypermyelination in these patients was unclear . Our finding that Lm211 inhibits Nrg1III signaling in small fibers could explain the effect of Lm211 deficiency in a neuropathic nerve , in which Nrg1 signals may be secondarily imbalanced . Similar alterations in the balance between Nrg1III and Lm211 signaling during myelination could explain other human neuropathies , such as Charcot-Marie-Tooth 4F and leprosy . The former is due to recessive mutations in periaxin , an interactor of the dystrophin-complex linked to dystroglycan in SCs . Charcot-Marie-Tooth 4F is also linked to hypermyelination and demyelination , possibly explained by the interrupted connection between dystroglycan and its ligand , Lm211 , in the SC basal lamina . Similarly , the leprosy mycobacterium infects peripheral nerves by binding Lm211 in SCs and activates ErbB2 [77 , 78] . The resulting dedifferentiation and demyelination could be explained by concomitant hyperactivation of ErbB2 and inhibition of Lm211 , which , as we have shown here , is deleterious for myelin-forming SCs . All experiments involving animals followed experimental protocols approved by the San Raffaele Scientific Institute Animal Care and Use Committee and Roswell Park Institute Animal Care and Use Committee . The approved protocols at San Raffaele ( n . 363 ) and at the University of Buffalo/Roswell Park ( UB1188M , UB1194M , UB1196R ) adhered to the guidelines set forth by the “Guide For The Use of Laboratory Animals , ” National Research Council , National Academy Press , Washington D . C . , 1996 . Nrg1III+/− mice were characterized in [27] and were a gift from Drs . Talmage and Role at SUNY Stony Brook; Nrg1IIItg mice were characterized in [36]; Lama2−/− mice were characterized in [23] and were a gift from Dr . Takeda , National Center of Neurology and Psychiatry , Tokyo . All animals used in this work were congenic into the C57/BL6N background . Genotyping of mutant mice was performed by PCR on tail genomic DNA , as described in [27 , 36] . For Lama2−/− , we used the following primers: 5′-CCCGTGATATTGCTGAAG-3′; 5′-CCTCTCCATTTTCTAAAG-3′; 5′-CAGGTGTTCCAGATTGCC-3′ . PCR was carried out at 95°C for 45 s , 50°C for 45 s , followed by extension at 72°C for 60 s , for 30 cycles . The expected 246 nt product for the wild-type allele and 450 nt product for the mutant allele were separated on a 2% agarose gel . Mutant and control littermates were sacrificed at P5 and P16 , and sciatic nerves were dissected . Semithin sections and EM analyses of sciatic nerves were performed as previously described [79] . The quantification of the number and the diameter of the axons in the unsorted bundles , the determination of g-ratios ( axon diameter/fiber diameter ) and the quantification of Remak bundles were performed on ultrathin sections . At least 3 animals per genotype were analyzed . Unsorted bundles were defined as groups of “naked” axons with no SC cytoplasm among them , and they contained some axons larger than 1 μM . All of the unsorted bundles in a section were counted . In contrast , Remak bundles differed from unsorted bundles because they contained ensheathed axons , all smaller than 1 μM . G-ratio were determined for at least 150 fibers chosen randomly . EM analyses on SC–DRG neurons cocultures from WT and CRD KO embryos following Lm211 treatment were performed as described [80] . Dorsal root ganglia ( DRG ) neurons were generated as described in [27] . Rat SCs ( 200 , 000 cells/coverslip ) were added to established cocultures , and myelination was initiated by supplementing the media ( Fetal Calf Serum [FCS] 10% , L-glutamine 2 mM , D-glucose 4 g/l , Nerve Growth Factor [NGF; Harlan , Bioproducts for Science] 50 ng/ml in MEM medium [Invitrogen] ) with 50 μg/ml of recombinant Lm211 purified as described in [81 , 82] or obtained from ( Biolamina ) and dyalized . Primary SCs were isolated from the sciatic nerves of 4-day-old Sprague-Dawley pups according to [83] . To expand the SC population , cells were kept in growing medium: DMEM containing 1% FBS , Nrg1 30 ng/ml ( human NRG1-β1 extracellular domain , R&D Minneapolis , MN ) , and forskolin 5 μM ( Calbiochem , Gibbstown , NJ ) for 2 to 4 generations . More than 95% of SC purity was verified based on their morphology and S100 immunoreactivity [54] . For experiments , cells were subcultured into 12 well dishes in growing medium , and after the cell density reached 70% confluency , they were kept for 3 days in differentiation medium: DMEM containing 1% FBS and dbcAMP ( Dibutyryladenosine 3′ , 5′-cyclic monophosphate sodium salt , Bremen , Germany ) without NRG1 . PKA inhibitor H89 ( Novartis , Basel , Switzerland ) and ErbB2 inhibitor PKI166 ( Novartis , Basel , Switzerland ) were added after 24h of dbcAMP treatment and then left for 2 days in combination with dbcAMP ( Figs 6 and 8 ) . For the sensitization experiment ( Fig 6 ) , cells were first treated with cAMP for 3 days , then Nrg was added for the indicated time in the presence of cAMP . When indicated , PKI166 and PP2 ( Calbiochem , Gibbstown , NJ ) , were added 30 min before Nrg1 treatment . 6-Bnz-cAMP and 8-pCPT-2-O-Me-cAMP were obtained from Biolog ( Bremen , Germany ) . All other undesignated reagents were purchased from Sigma . All antibodies used were previously validated for the applications used . Antibodies against ErbB2 ( sc-7301 ) , pErbB2Y1248 ( sc-12352-R ) for western blot ( WB ) were from Santa Cruz Biotechnology . Anti-ErbB2 ( 4290 , 1:1000 for WB ) , Gab1 ( 3232 , 1:1000 for WB ) , p-GabY627 ( 3231 , 1:1000 for WB ) , and Phospho-PKA Substrate ( RRXS/T ) ( 9621 , 1:1000 for WB ) were from Cell Signaling . Anti-Oct6 were either from Santa Cruz ( sc-11661 ) or from ( D . Meijer , University of Edinburgh , United Kingdom , 1:1000 for IHC and WB ) . Anti-Egr2 was either from Covance ( PRB-236P ) or from D . Meijer , University of Edinburgh , UK , 1:1000 for IHC and WB ) . p-Histone H3 ( 06–570 , 1:200 for IHC ) was from EMD Millipore . Anti-neurofilament M was from Covance ( PKC-593P , 1:1000 for IHC ) and Anti-MBP from V . Lee , University of Pennsylvania , USA ( 1:6 for IHC ) . Anti-calnexin ( C4731 , 1:4000 for WB ) , tubulin ( T4026 , 1:2000 for WB ) , GAPDH ( G9545 , 1:5000 for WB ) and β-actin ( A5441 , 1:1000 for WB ) were from Sigma-Aldrich . Secondary antibodies: goat antichicken DyLight 550 , ( Abcam , ab96948 , 1:700 for IHC ) , goat antirabbit Alexa Fluor 488 ( 111-545-003 , 1:700 for IHC ) and goat antirabbit HRP ( 111-035-003 , 1:5 . 000 for WB ) goat antimouse HRP from Jackson ImmunoResearch; ( SIGMA , A2554 , 1:10000 for WB ) . Infrared secondary antibodies for quantitative WB analyses were obtained from LI-COR Biosciences and used 1:10 , 000 ( goat antimouse IRDye 680 926–68070; goat antimouse IRDye 800 926–68070; goat antirabbit IRDye 680 926–68021; goat antirabbit IRDye 800 926–32211 ) . Frozen sciatic nerves dissected from P5 and P16 mice were pulverized and resuspended in lysis buffer ( 95 mM NaCl , 25 mM Tris-HCl , pH 7 . 4 , 10 mM EDTA , 10mM EGTA , 2% SDS , 1 mM Na3VO4 , 1 mM NaF , 1% Protease Inhibitor Cocktail [Sigma-Aldrich] ) , 1% Phosphatase inhibitor cocktail C2 and C3 ( Sigma-Aldrich ) , boiled for 5 min , and centrifuged 10 min at 17 , 000 g at 16°C . The protein concentration in supernatants was determined by BCA protein assay ( Thermo Scientific ) according to the manufacturer's instructions . Equal amounts of homogenates were loaded with reducing sample buffer . SC were lysed and boiled in 2X SDS dye lysis buffer ( 1 M Tris-HCl pH 6 . 8 , 10% Sodium Dodecyl Sulfate , Glycerol , 1% Dichloro-Diphenyl-Tichloroethane , 1% Bromophonol blue ) and centrifuged at 17 , 000 g at 4°C , the supernatants were denatured , resolved on SDS-polyacrylamide gel and electroblotted onto PVDF or nitrocellulose membrane ( Millipore ) . Blots were then blocked with BSA 5% in PBS or Odyssey buffer ( LI-COR Biosciences ) and incubated with the appropriate antibody . Blots were developed with ECL or ECL prime ( GE healthcare ) , and band intensity was quantified from films using ImageJ software . Alternatively , for quantitative WB analyses , filters were analyzed using the Odyssey Infrared Imaging System ( LI-COR Biosciences ) according to manufacturer's instructions . DRG cocultures were fixed with 4% PFA for 20 min , washed , permeabilized with cold methanol for 20 min , incubated in blocking solution ( 20% FCS , 2% bovine serum albumin , and 0 . 1% Triton in PBS ) for at least 1 h , and then incubated overnight with antineurofilament , MBP , or PH3 antibodies in blocking solution . Explants were then incubated with secondary antibodies and counterstained with Dapi . 6 ( for MBP ) or 5 ( for PH3 ) images from each DRG were acquired by epifluorescence on a Leica DM5500B or DM6000 microscope with a 10X or 20X objective . This analysis was performed on at least 3 coverslips per embryo and on 2 ( for PH3 ) or 3 ( for MBP ) embryos per genotype for at least of 6 ( for PH3 ) or 9 ( for MBP ) coverslips per condition . Sciatic nerves were dissected from P5 or P16 mice and fixed 1 h in 4% PFA at 4°C , cryo-protected in 20% sucrose ( Sigma-Aldrich ) , embedded in OCT ( Miles ) , and snap-frozen in liquid nitrogen . Alternatively , unfixed nerves were directly embedded in OCT and snap frozen . Staining was performed on 10-μm longitudinal sections on unfixed tissue for Egr2 and fixed tissue for Oct6 . Sections were permeabilized in cold methanol for 5 min and blocked with 20% FBS , 2% BSA , and 0 . 1% triton and incubated overnight with the primary antibody . Images from 3 different sections per animal were acquired with a 20X objective . 3 different mice per for each genotype were analyzed . These assays were performed as described [84] . PKA activity was assessed in vivo using the SigaTect cAMP-dependent PKA assay system ( Promega , V4780 ) . Sciatic nerves at P5 and P16 were sampled and immediately used for the assay according to the manufacturer instructions . To measure cAMP concentration in P5 and P16 sciatic nerves , a cAMP assay kit ( Enzo Life Sciences ) was used as described [85] . Treatments with the PKA signaling inhibitors H-89 ( EMD Millipore ) and KT5720 from ( Enzo Life Sciences ) were performed through injections beneath the gluteus superficialis and biceps femoris muscles . Animals were injected on 4 consecutive days from P3 to P6 with 10 μl of inhibitor solution ( 10 μM H-89 DMSO , 1 μM KT5720 , 0 . 1% DMSO diluted PBS ) per day . Untreated nerves were injected with 10 μl of 0 . 1% DMSO diluted in PBS . Sciatic nerves were sampled at P7 . Data were collected randomly and assessed blindly . The data distribution was assumed to be normal , although we did not formally test it . All statistical analyses were performed on at least 3 independent experiments . Statistical detailed analyses are reported in each figure legend and all assays ( 1-way ANOVA multiple comparison test , 2-sided moderate t test and 2-tailed unpaired t test ) were performed using the Prism Software package ( GraphPad , San Diego , CA ) . The numerical data used in all figures are included in S1 Data .	Myelin is formed by the wrapping of glial cell membranes around axons and is required for the fast conduction of nerve impulses and to support axons . In the peripheral nervous system , myelin is produced by Schwann cells . Peripheral myelin defects cause debilitating diseases , whose molecular pathogeneses are only partially understood . Here , we reveal for the first time how 2 crucial extracellular modulators of myelin formation , neuregulin 1 type III ( Nrg1III ) and laminin α2β1γ1 ( Lm211 ) , work together in the peripheral nervous system . Although Lm211 was believed to promote myelination , we show that it can also inhibit myelin formation by suppressing the activity of Nrg1III , limiting the activation of its downstream signaling cascade . These results help to explain why certain inherited neuropathies are characterized by hypermyelination and redundant myelin sheaths .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "phosphorylation", "medicine", "and", "health", "sciences", "nervous", "system", "neuroscience", "signal", "inhibition", "nerve", "fibers", "protein", "kinase", "signaling", "cascade", "animal", "cells", "axons", "proteins", "biochemistry", "signal", "transduction", "ce...	2017	Laminin 211 inhibits protein kinase A in Schwann cells to modulate neuregulin 1 type III-driven myelination
Methodology to estimate malaria incidence rates from a commonly occurring form of interval-censored longitudinal parasitological data—specifically , 2-wave panel data—was first proposed 40 years ago based on the theory of continuous-time homogeneous Markov Chains . Assumptions of the methodology were suitable for settings with high malaria transmission in the absence of control measures , but are violated in areas experiencing fast decline or that have achieved very low transmission . No further developments that can accommodate such violations have been put forth since then . We extend previous work and propose a new methodology to estimate malaria incidence rates from 2-wave panel data , utilizing the class of 2-component mixtures of continuous-time Markov chains , representing two sub-populations with distinct behavior/attitude towards malaria prevention and treatment . Model identification , or even partial identification , requires context-specific a priori constraints on parameters . The method can be applied to scenarios of any transmission intensity . We provide an application utilizing data from Dar es Salaam , an area that experienced steady decline in malaria over almost five years after a larviciding intervention . We conducted sensitivity analysis to account for possible sampling variation in input data and model assumptions/parameters , and we considered differences in estimates due to submicroscopic infections . Results showed that , assuming defensible a priori constraints on model parameters , most of the uncertainty in the estimated incidence rates was due to sampling variation , not to partial identifiability of the mixture model for the case at hand . Differences between microscopy- and PCR-based rates depend on the transmission intensity . Leveraging on a method to estimate incidence rates from 2-wave panel data under any transmission intensity , and from the increasing availability of such data , there is an opportunity to foster further methodological developments , particularly focused on partial identifiability and the diversity of a priori parameter constraints associated with different human-ecosystem interfaces . As a consequence there can be more nuanced planning and evaluation of malaria control programs than heretofore . Estimation of incidence , and in some cases recovery , rates for malaria infection is a central objective of ongoing community surveillance programs [1] . By incidence rate we mean the number of new infections acquired in a small interval of time per person at risk ( i . e . uninfected ) at the beginning of the interval . Analogously , a recovery rate is the number of terminations of infection in a small interval of time per person at risk ( i . e . infected ) at the beginning of the interval . It is important to observe that incidence rate , as defined above , is not the same as incidence , as commonly used in the contemporary malaria literature [e . g . , see reference 2] . In the latter case , incidence is defined as [Number of positive species-specific clinical cases observed during the duration of a survey]/[ ( Number of people observed over the survey duration ) * ( duration of survey ) ] . We quantitatively compare and contrast these notions in the Discussion section . Ideally one would like to have continuous histories of infection status on designated populations so that incidence rates could be directly ascertained over short intervals starting at any designated time . This would facilitate showing the impact of local-in-time weather events , seasonal variation , and the impact of intervention strategies on these rates at arbitrary times of interest to health service workers , government personnel setting malaria control policy , and research investigators . Continuous infection status histories are virtually never available at a community level . The most common longitudinal data collection plan is a time series of 2-wave panel data sets ( or observations of infection status on the same individuals taken at two time points separated by an interval of length Δ ) , with spacing between waves varying from a few weeks [3 , 4] to several months [5] . Thus , there are unobserved transitions between states ( uninfected and infected ) that create a challenge for estimation of incidence and recovery rates . To address this challenge , we require specification of a class of continuous-time 2-state stochastic process models of infection status dynamics that must be shown to be consistent with observed data , and within which estimation of incidence and recovery rates is feasible . The first attempt to carry out this program was by Bekessy et al . [6] using data from the Garki malaria surveys [3] in Kano State , Nigeria from 1970–1975 . They introduced the 2-state continuous time Markov chains as candidates to represent the unobserved infection dynamics . The empirical question associated with this choice was whether or not a transition matrix P ( Δ ) generated by a member of this class of models could represent a transition matrix P^ ( Δ ) arising from field data . Here Δ is the time interval between observations collected at a survey date T1 and a later survey date T2 . P ( Δ ) is a 2x2 transition matrix associated with a continuous time Markov chain with entries pi , j ( Δ ) = conditional probability that an individual has infection status j at the end of a time interval of length Δ given that his/her infection status at the beginning of the interval is i . Here , i and j can take on the values 1 = uninfected or 2 = infected . P^ ( Δ ) is a 2x2 stochastic matrix with entries ni , j/ ( ni , 1 + ni , 2 ) , where ni , j counts the number of individuals observed to be in state i at the beginning of interval Δ and in state j at the end of the interval . The entries are interpreted as conditional frequencies of observing a transition i → j between the consecutive survey dates . For the Garki surveys , Δ ≈ 10 weeks . The transition matrices P ( Δ ) have a representation in terms of incidence and recovery rates given by: P ( Δ ) =exp ( ΔQ ) Q= ( −q1q1q2−q2 ) with qi≥0 for i=1 , 2 ( 1 ) Here q1 is the incidence rate at any time t in the interval Δ , and q2 is the corresponding recovery rate . Bekessy et al . [6] showed that a 2x2 stochastic matrix , P* , has a representation of the form Eq ( 1 ) if and only if trace P* > 1 ( 2 ) Thus , the empirically determined stochastic matrices P^ ( Δ ) can be generated by observations taken at two points in time on a continuous time Markov chain provided trace P^ ( Δ ) >1 . Formal statistical tests of this hypothesis were put forth by Singer & Cohen [7] . Interestingly , Bekessy et al . [6] , Singer & Cohen [7] , and Molineaux & Gramiccia [3] found that all pairs of consecutive surveys during the baseline period of data collection in the Garki project satisfied eq ( 2 ) . However , there were pairs of consecutive surveys conducted during the intervention phase of the project where trace P^ ( Δ ) <1 . In these instances , Molineaux & Gramiccia [3] claimed that incidence and recovery rates were not estimable . While this assertion is correct for the class of model eq ( 1 ) , the unanswered question as of 1980 was: what alternative and substantively defensible models could generate transition matrices satisfying trace P^ ( Δ ) <1 and within which incidence and recovery rates could be estimated ? Surprisingly , to the best of our knowledge , this question has not been taken up in the past 40 years . Nevertheless , its importance stems from the fact that 2-wave panel data in a diversity of malaria surveys/surveillance projects have a majority of their transition matrices , P^ ( Δ ) , satisfying trace P^ ( Δ ) <1 . In addition , incidence and recovery rates are important quantities for evaluation of malaria intervention programs . The purposes of this paper are to: ( i ) present a class of models with associated 2x2 transition matrices , P* , within which those satisfying eq ( 1 ) are nested , some members of which satisfy trace P* < 1 , and for all of which incidence and recovery rates can be calculated; ( ii ) exhibit identifiability and/or partial identifiability criteria arising from specific malaria contexts that ensure uniqueness , or highly constrained non-uniqueness of incidence and recovery rates; and ( iii ) apply the models and methods in ( i ) and ( ii ) to a time series of 2-wave panel data sets from Dar es Salaam , Tanzania as an example of the applicability of the proposed method [5 , 8 , 9] . To facilitate dissemination and utilization of the methodology by the malaria community , we developed a code to calculate incidence rates from longitudinal data utilizing the R package v . 2 . 15 . 1 [10] , which we make available as Supporting Information in a documented version ( S1 Code ) and as a R file ( S2 Code ) . To allow replication of our results , we provide the Dar es Salaam data utilized in this paper ( Tables 1 and 2 ) . An important feature of our considerations is allowance for non-identifiability in situations where subject-matter-based constraints are too weak to ensure unique parameter identification from input information . Depending upon the particular setting , the extent of non-identifiability—or partial identifiability—may be sufficiently limited that the unidentified parameters are restricted to a narrow range of values , with accompanying incidence rates also varying over a small range . This is , indeed , the situation in Dar es Salaam , the site of our empirical data . As a consequence , we show how to explicitly incorporate variability due to a small degree of under-identification together with sampling variability to produce composite uncertainty intervals for incidence rates . Useful discussions of partial identifiability , including in the setting of mixture models , are given by Manski [11] , Gustafson [12] , and Henry et al . [13] . Ethical approval to use the UMCP data was provided by the Harvard T . H . Chan School of Public Health Institutional Review Board ( Protocol # 20323–101 ) . We begin with the observation that every 2x2 stochastic matrix , P , with non-zero entries can be a transition matrix for at least one 2-component mixture of continuous-time Markov chains . Thus , P has a representation of the form P=SU+ ( I−S ) V ( 3 ) where S is a diagonal matrix with entries si , i = 1 , 2 in the unit interval , [0 , 1] . U and V are each stochastic matrices having representations of the form eq ( 1 ) . Hence they satisfy the condition eq ( 2 ) : trace U > 1 and trace V > 1 . It will be convenient to represent the matrices P , U , and V as points in the unit square; namely p = ( p1 , p2 ) , u = ( u1 , u2 ) , and v = ( v1 , v2 ) . The coordinates pi , ui , and vi , i = 1 , 2 are the diagonal entries in P , U , and V respectively . Depending on the empirical setting , we will also have occasion to consider data of the form p = ( p1 , 0 ) and p = ( 0 , p2 ) ; i . e . points on the boundary of the unit square . In the first of these conditions , P has a representation of the form eq ( 3 ) , but with u = ( 1 , 0 ) ( hence , trace U = 1 ) and trace V > 1 . For p = ( 0 , p2 ) , the representation eq ( 3 ) also holds but now with u = ( 0 , 1 ) and trace V > 1 . Data of the form p = ( p1 , 0 ) occur frequently in the Dar es Salaam data , as discussed later . In these boundary cases , U still has a representation of the form eq ( 1 ) but with q = ( 0 , −∞ ) when p = ( p1 , 0 ) , and q = ( −∞ , 0 ) when p = ( 0 , p2 ) . In the first case , the interpretation of q2 = ∞ is that there is an infinitely fast recovery rate . This would be associated with a population where everyone is on prophylaxis , or where effective anti-malarial drugs are administered immediately following a diagnosis of infection . In the second case , q1 = ∞ corresponds to an infinitely fast incidence rate . This would be a situation where there is instantaneous new infection of any exposed individual . For data corresponding to p in the interior of the unit square—i . e . P^ ( Δ ) with non-zero entries—we calculate the incidence and recovery rates for the mixture eq ( 3 ) via: r1=s1q1 ( U ) + ( 1−s1 ) q1 ( V ) ( incidence rate ) r2=s2q2 ( U ) + ( 1−s2 ) q2 ( V ) ( recovery rate ) ( 4 ) In terms of U and V , the rates qi ( U ) and qi ( V ) , i = 1 , 2 can be expressed as ( note that these formulas are entries in matrix logarithms of U=eQ ( U ) Δ and V=eQ ( V ) Δ – see e . g . [7] ) : qi ( U ) =log ( trace U−1 ) trace U−2 ( 1−ui ) Δqi ( V ) =log ( trace V−1 ) trace V−2 ( 1−vi ) Δ When p = ( p1 , 0 ) , we have r1= ( 1−s1 ) q1 ( V ) = ( 1−s1 ) log ( trace V−1 ) trace V−2 ( 1−v1 ) Δ =log ( trace V−1 ) trace V−2 ( 1−p1 ) Δ , since s1=p1−v11−v1 ( 5 ) The recovery rate is ∞ for every s2 > 0 . When s2 = 0 , we have r2=log ( trace V−1 ) trace V−2 ( 1−v2 ) Δ An analogous argument yields r1 and r2 when p = ( 0 , p2 ) . We first re-express Eq ( 3 ) via the system of equations p1=s1u1+ ( 1−s1 ) v1p2=s2u2+ ( 1−s2 ) v2 ( 6 ) where ( si , ui , vi ) ∈ [0 , 1]6 for i = 1 , 2 and u1+u2>1 , v1+v2>1 . Given p = ( p1 , p2 ) , eq ( 6 ) is an under-identified system . Additional subject-matter motivated constraints must be imposed to either identify ( s , u , v ) uniquely or restrict this vector to a small set in [0 , 1]6∩ ( ( u , v ) :u1+u2>1 , v1+v2>1 ) . In the context of malaria in Dar es Salaam , we impose the constraints: u1 ≤ 0 . 2 , u2 = 1 , 0 . 9<v1 < 1 , v2 < 0 . 5 , s1 < 0 . 5 , and v1 − v2 ‘large’ . A full rationale for the above restrictions will be given later when we present the Dar es Salaam case study . However , the central point here is that a system of constraints such as these is essential for parameter identification or partial identification . The conditions that v1 − v2 be ‘large’ and v1 < 1 require additional comment . First , it is a matter of judgment about what is a high probability of being observed uninfected at consecutive surveys of the V-process , while still not being a sure thing—i . e . v1 = 1 . In identifying parameters , we first select v1 ∈[0 . 9 , 1 ) and secondarily choose v2 as small as possible consistent with the other constraints . Two examples will serve to illustrate the issues . When p = ( p1 , p2 ) is in the interior of the unit square , we generate 1 , 000 tables by doing binomial sampling for row 1 with probability p1 and for row 2 with probability p2 with sample sizes n11 + n12 and n21 + n22 , respectively . If , for a particular table , the system of eq ( 6 ) has a unique solution ( s , u , v ) , subject to context-specific constraints , then we compute an incidence rate , r1 , for that table . If there is a zone of non-identifiability , as previously exemplified by the equation 0 . 95s1 − s1u1 − 0 . 25 = 0 in Example 1 , then we compute r1 for each of 100 values u1 ( which then determines s1 ) subject to the a priori constraints on s1 and u1 . This yields a set of incidence rates that reflect variation due to non-identification . We used the minimum , median , and maximum values of r1 from each such set of 100 values and viewed them as the summary rates for the particular table . Finally we take the summary rates , for tables where non-identifiability is an issue , and the unique rate for tables when the system eq ( 6 ) is identified , and rank this full set of rates . We designate the 2 . 5th percentile and the 97 . 5th percentile of the ranked list as the upper and lower bounds on a 95% variation interval for the incidence rate of the original table . This takes both sampling variability and variation due to non-identifiability into account . When p = ( p1 , 0 ) , we treat the 0 as a structural zero—in the case of Dar es Salaam—and only do binomial sampling on the first row to generate 1 , 000 tables having this same structure . We then describe the variation in r1 in the same manner as indicated above . In applications where p = ( p1 , 0 ) does not have a structural zero , we perturb the second coordinate to a small value—e . g . 10−5 or less—and do binomial sampling for the second row with this value . Then we proceed as in the above paragraph to calculate a confidence interval for r1 . Light microscopy has limitations as a technology for diagnosing Plasmodium falciparum infections , particularly in low transmission settings [14–16] . In a recent study of Okell et al . [16] , the supplementary information for the paper contains an especially interesting and useful table comparing prevalence estimates using microscopy and PCR on the same blood samples . The data come from a wide variety of settings , and exhibit considerable variation in prevalence rates as ascertained via microscopy . The prevalence ratio , p = [prevalence rate from microscopy]/[prevalence rate from PCR] provides a basis for adjusting empirical microscopy rates to what you would expect to find if PCR had actually been done on the same blood samples . This calculation will , of course , only yield adjusted prevalence rates . For our longitudinal data , it would have been desirable to have microscopy and PCR based estimates of p12 , p21 and p22 , from which we could directly recover p11 . However , the lack of identifiability of transition probabilities from prevalence rates can still be dealt with in particular settings , such as Dar es Salaam , by invoking an additional , and obviously context dependent , constraint . We exhibit the methodology on p = ( 0 . 7 , 0 . 2 ) –the transition rates in example 1–augmented by a table of counts with entries nij consistent with these values . Using an adjusted table of counts , nij* , and thereby an adjusted vector , p= ( p1 , p2* ) , we calculate the incidence rate , r1* , that represents what we might have expected to find if PCR had been done on the blood samples in Dar es Salaam . We introduce the table of counts {nij , 1 ≤ i , j ≤ 2} with n11 = 100 , n12 = 43 , n21 = 90 , and n22 = 23 . For this table , p = ( 0 . 7 , 0 . 2 ) . It is one of a myriad of tables that could have been selected to illustrate our points about submicroscopic infection . However , it is comparable in size to many of the sub-ward tables in the Dar es Salaam data , and thus especially useful for illustrating an adjustment methodology . The prevalences at the initial and final rounds of data collection for the above table are: at initial survey = [90 + 23]/256 = 0 . 4414 , and at final survey = [23 + 43]/256 = 0 . 2578 . From Table S1 in Okell et al . [16] , we find the prevalence from microscopy that is closest to the prevalence at initial survey in Dar es Salaam given by 0 . 4414 . This is the prevalence of 0 . 481 based on data from Guinea Bissau . The corresponding prevalence ratio is p = 0 . 551 . Thus our estimate for a corresponding PCR-based prevalence rate at the initial survey is 0 . 4414/0 . 551 = 0 . 8011 . In contrast to the initial survey situation , there are four nearby microscopy-based prevalence rates to associate with the prevalence rate for the final survey given above by 0 . 2578 . These values , their associated prevalence ratios , and our estimate for the corresponding PCR-based prevalence rates are shown in Table 4 . Each PCR rate is equal to 0 . 2578/p . For our analysis , we use the average of these PCR rates , namely 0 . 5173 . To obtain an associated table of counts nij* , we first observe that n11+n12+n21+n22=256= total count from the microscopy-based table with entries nij . From PCR prevalence at initial survey = 0 . 8011 , we obtain n21+n22=205 . From PCR prevalence at final survey = 0 . 5173 , we obtain n12+n22=132 . Adding and subtracting n22* to the equation for total count , we can rewrite it as ( n11−n22 ) +n12+n22+n21+n22=256 . Then we have that n11−n22=−81 . To be consistent with the microscopy-based vector , p = ( 0 . 7 , 0 . 2 ) , where obviously p1 > p2 , we choose n11* to ensure that p1>p2 . Table 5 shows some choices for estimated PCR-based tables . Here we use as an example p1=0 . 784 and p2=0 . 590 ( third line of Table 5 ) for subsequent calculations . To obtain what is interpreted as a PCR-based incidence rate , we proceed as in the previously described microscopy-based rate calculations . First set v1=0 . 95 . Then from p1=s1u1+ ( 1−s1* ) v1* , we obtain 0 . 95s1−s1u1−0 . 166=0 . Using v2=[0 . 590− s2]/ ( 1−s2 ) >1−v1=0 . 05 , we set s2=0 . 5 . Then v2=0 . 18 , and v1+v2=1+ε=1 . 13 . Thus q1 ( V ) =logεε−1 ( 1−v1* ) Δ=0 . 00173 , with ε = 0 . 13 and Δ = 40 . Along the curve of ( s1* , u1* ) values given by 0 . 95s1−s1u1−0 . 166=0 , we obtain values of q1 ( U ) and r1* as indicated in Table 5 . The continuous-time Markov chains all have exponentially distributed waiting time distributions in each state , which implies that their hazard rates are constant . Two-wave panel data , assumed to be generated by some 2-state continuous-time stochastic process , is not sufficiently rich to provide a basis for testing this hypothesis . However , several basic facts about malaria in diverse ecosystems make this assumption untenable . For example , in the Garki study [3] , persons who survive repeated episodes of malaria in infancy and childhood have antibody titers that ensure an increasing hazard rate in the infected state—i . e . the longer an individual has detectable parasites , the more likely he/she is to clear parasites free of any intervention , and return to the uninfected state . For uninfected individuals at the end of a dry season , the hazard rate for onset of a new infection is also increasing , corresponding to the propensity for rain and , thereby , standing water . One of many possible parameterizations of these qualitative ideas is given by the waiting time distributions F ( i ) ( t ) , t > 0 , where i = 1 , 2 designate the states uninfected ( i = 1 ) and infected ( i = 2 ) , and having probability density functions fi ( t ) =βiαitαi−1e−βitΓ ( αi ) , αi , βi , t>0 for i=1 , 2 ( 7 ) and hazard rates hi ( t ) =fi ( t ) 1−F ( i ) ( t ) , i=1 , 2 . Here hi ( t ) is: increasing if αi > 1 , constant if αi = 1 , and decreasing if αi < 1 . The family of Gamma densities eq ( 7 ) are the basis for obtaining an expression for P ( Δ ) within the class of semi-Markov models by solving the backward integral equation system [30] . pij ( 0 , t ) =δij[1−Fi ( t ) ]+∑k=1r∫0tfi ( s ) mikpkj ( 0 , t−s ) ds ( 8 ) where δij = 1if i = j , δij = 0 if i ≠ j , and 1 ≤ i , j ≤ r with M = ‖mik‖ an r x r stochastic matrix having mii = 0 for 1 ≤ i , j ≤ r . Specification eq ( 8 ) holds for general r-state process and waiting time densities fi ( t ) , 1 ≤ i ≤ r . For our purposes , r = 2 , m12 = m21 = 1 , and we focus on the 2-parameter family eq ( 7 ) . Here we set t = Δ and identify pij ( 0 , Δ ) with the ( i , j ) entry in P ( Δ ) . The equation system eq ( 8 ) is amenable to the following interpretation: ( i ) when i ≠ j , pij ( 0 , t ) consists of the sum of products of three factors: the probability of a first departure from state i at time s , the probability of a transition from state i to state k at that instant , and the probability of transferring to state j by some combination of moves in the interval t−s The summation is over all intermediate states k and all time divisions s in ( 0 , t ) ; ( ii ) when i = j , in addition to the preceding term , there is the probability of not transferring out of state i during ( 0 , t ) . This is given by the first term . With empirical data , solving the system of equations p^i , j ( Δ ) =pij ( 0 , Δ ) , i=1 , 2 for parameter values as in the Gamma specification above , requires a priori context-dependent constraints on the parameters—to secure identification or partial identification—and numerical inversion calculations in eq ( 8 ) . Going back to the empirical analyses in the Garki baseline surveys [3] , the disconcerting issue that now arises is that the entire set of 2-wave panel data sets shown in Singer & Cohen [7] , with incidence and recovery rates computed within the class of continuous time Markov chain models , could just as well have been used to estimate incidence and recovery rates within the class of 2-state semi-Markov models with Gamma distributed waiting times . The same can , of course , be said for the rates computed in the prior section for Dar es Salaam now using , in some of the trace P < 1 cases , 2-component mixtures of semi-Markov models with analogous inequality constraints facilitating identification or partial identification of parameters . If we use a crude incidence rate given by r1 ( crude ) =p^12 ( Δ ) Δ , this essentially assumes that there is at most one unobserved transition in the inter-survey interval , Δ . For the Garki surveys where Δ = 10 weeks , we find , not surprisingly , that r1 ( crude ) <r1 ( Markov ) in the baseline surveys . In the lower endemicity setting of Dar es Salaam , the same inequality holds when trace P^>1 , but now Δ ≈ 40 weeks . If Δ had been approximately 10 weeks for Dar es Salaam , we would anticipate very little difference between r1 ( crude ) and r1 ( Markov ) . We also find that r1 ( crude ) < r1 ( Mixture ) when trace P^<1 in Dar es Salaam , but this is decidedly influenced by the long inter-survey intervals . As two examples , consider the ward Buguruni , for survey intervals R23 ( with Δ = 42 . 2 weeks ) and R56 ( with Δ = 49 . 2 weeks ) . For R23 , r1 ( crude ) = 0 . 0031 and r1 ( Mixture ) = 0 . 0081 . For R56 , r1 ( crude ) = 0 . 0007 and r1 ( Mixture ) = 0 . 0039 . Thus , during the last survey interval , R56 , when the larvicide intervention was operating effectively , we still have r1 ( crude ) and r1 ( Mixture ) differing by a factor of 5 . 6 . Taking formal account of the possibility of multiple unobserved transitions clearly makes a difference , and is preferable to the crude incidence rate , generally . For the 2-wave panel data in Dar es Salaam , we have: Inc=n22+n21+n12N*Δ ( 9 ) where N = n11+n12+n21+n22 . Using eq ( 9 ) we compare Inc with r1 ( Mixture ) for two wards , Buguruni and Kurasini , early in the larviciding program , R23 , and at the end of it , R56 ( Table 8 ) . The general lesson here is that Inc < r1 ( Mixture ) except for survey rounds where there are almost no infected cases . Indeed , for Kurasini at R56 we have a raw table of counts given by ( n11n12n21n22 ) = ( 63010 ) . Here there is no apparent transmission between survey rounds 5 and 6 . The infected individual at survey is , in accordance with the study protocol , treated with anti-malarial drugs . The general inequality Inc < r1 ( Mixture ) is basically a consequence of the fact that unobserved transitions are not taken into account in Inc . In moderate to high transmission areas , we anticipate that Inc will be substantially downward biased as a result of lack of formal consideration of unobserved transitions . Different methods have been proposed to generate incidence ( as defined in this paper , or the rate of occurrence of an event ) from prevalence data [31–34] . As more applications of the methodology here introduced are made , analysis that would produce both incidence and prevalence rates could provide a unique opportunity to evaluate the best approach to obtain incidence rate estimates from prevalence rates—currently , it is unclear what is the best strategy . Such an exercise could lead to clear recommendations that would have both a wide applicability in malaria endemic countries and a crucial importance for National Malaria Control Programs ( e . g . , planning and evaluation of the cost-effectiveness of interventions ) . In conclusion , this paper introduced new methodology for estimating incidence rates from 2-wave panel data with interval censoring , which is applicable in the many cases where the extant Markovian models are inapplicable . The methodology is suitable to settings with any malaria transmission level , given the availability of longitudinal data . In addition , we present a strategy for quantifying the uncertainty in estimation of incidence rates . It is hereby distributed with a well-documented programming code that allows the use of the method in R software .	Incidence rates measure the transitions between the states of noninfected to infected per unit of time and per person at risk . Usually calculated from longitudinal observations , they provide an indication of how rapidly a disease develops in a population over time . In the context of malaria , longitudinal data on infection status are obtained through consecutive survey rounds , separated by a certain time interval . Depending on the length of the interval , some changes of infection status may be missed , and thus only uncensored information would be available . Methodology to calculate incidence rates from this type of data was first proposed in 1976 , but its assumptions were not applicable to low transmission settings , particularly in the presence of control measures . No alternative methodology has been proposed in the past 40 years , limiting attempts to obtain estimates of incidence rates in the current scenario of declining malaria transmission worldwide . In this paper we address this gap and introduce new methodology to estimate malaria incidence rates from longitudinal data that can be applied to settings with any transmission level . We provide a complete example of the method , including sensitivity analysis , and an assessment of possible differences between data based on microscopy vs . PCR diagnostics . To facilitate replication and wide use of the method , we make available a programming code in R language and the example dataset .	[ "Abstract", "Introduction", "Methods", "Discussion" ]	[ "invertebrates", "medicine", "and", "health", "sciences", "markov", "models", "atmospheric", "science", "drugs", "tropical", "diseases", "social", "sciences", "parasitic", "diseases", "parasitic", "protozoans", "animals", "antimalarials", "mathematics", "protozoans", "pha...	2016	Malaria Incidence Rates from Time Series of 2-Wave Panel Surveys
Translesion synthesis ( TLS ) enables DNA replication through damaged bases , increases cellular DNA damage tolerance , and maintains genomic stability . The sliding clamp PCNA and the adaptor polymerase Rev1 coordinate polymerase switching during TLS . The polymerases Pol η , ι , and κ insert nucleotides opposite damaged bases . Pol ζ , consisting of the catalytic subunit Rev3 and the regulatory subunit Rev7 , then extends DNA synthesis past the lesion . Here , we show that Rev7 binds to the transcription factor TFII-I in human cells . TFII-I is required for TLS and DNA damage tolerance . The TLS function of TFII-I appears to be independent of its role in transcription , but requires homodimerization and binding to PCNA . We propose that TFII-I bridges PCNA and Pol ζ to promote TLS . Our findings extend the general principle of component sharing among divergent nuclear processes and implicate TLS deficiency as a possible contributing factor in Williams-Beuren syndrome . DNA bases experience many types of damage caused by both endogenous and exogenous factors . DNA repair pathways , such as the global genomic nucleotide excision repair ( GG-NER ) pathway , actively remove damaged bases [1] . In addition , when damaged bases are not completely removed , DNA translesion synthesis ( TLS ) allows replication past these lesions , thus increasing DNA damage tolerance and maintaining genomic integrity [2] . TLS requires a set of specialized DNA polymerases , including the Y family polymerases , Rev1 , Pol η , ι , and κ , and the B family polymerase Pol ζ containing the Rev3 catalytic subunit and the Rev7 regulatory subunit [2] . Certain TLS polymerases , including Pol ζ , are involved in somatic hypermutation of immunoglobulin genes [3]–[5] . Recent advances have established that multiple polymerase-switching events occur during TLS , and have begun to elucidate the elaborate molecular mechanisms that regulate these steps . When replicative polymerases encounter damaged DNA bases , such as those crosslinked by UV , the sliding clamp PCNA is ubiquitinated [6] . Ubiquitinated PCNA recruits Pol η , ι , or κ through the adaptor polymerase Rev1 [7]–[13] . Pol η , ι , or κ inserts nucleotides directly opposite to the DNA lesion [14]–[16] . In a poorly understood second switch , Pol ζ is employed to extend DNA replication past the lesion . Rev1 can simultaneously bind to Pol ζ and one of the Y family polymerases , Pol η , ι , or κ , suggesting that this polymerase switching step might occur through simple repositioning of a large , multi-polymerase assembly on the DNA template [17] . After TLS is completed , replicative DNA polymerases re-engage with PCNA and resume high-fidelity replication . TFII-I was first identified as a general transcription factor that bound to a pyrimidine-rich Initiator ( Inr ) sequence at the transcription start site and supported transcription in an in vitro reconstituted system [18] . TFII-I contains an N-terminal dimerization domain , six repeated domains ( called R1-R6 ) , and four AlkB homologue 2 PCNA-interacting motifs ( APIM ) motifs , among other features [19] , [20] . Recent studies have suggested that TFII-I is not a general transcription factor required for all Inr-dependent transcription [21] . Instead , it has signal- and context-dependent regulatory roles in the transcription of specific genes . Interestingly , TFII-I is one of 26–28 genes affected by a hemizygous deletion of the chromosome 7q11 . 23 region in the rare human disorder , Williams-Beuren syndrome ( WBS ) [22] . WBS patients exhibit a wide spectrum of phenotypes , including distinctive craniofacial features , cardiovascular abnormalities , and mental retardation . Heterozygous mutant mice with the N-terminal 140 residues of TFII-I deleted show WBS-like craniofacial and neurobehavioral alterations , linking this region of TFII-I to a subset of WBS phenotypes [23] . In this study , we show that TFII-I physically interacts with the Pol ζ subunit Rev7 ( also known as Mad2B ) . Functional studies reveal that TFII-I is indeed required for TLS and DNA damage tolerance in human cells . Depletion of TFII-I affects the transcription of a small number of genes , none of which are known to be involved in TLS , suggesting that the TLS function of TFII-I is independent of its role in transcription . Instead , both PCNA binding and homodimerization of TFII-I are required for TLS . We propose that TFII-I connects Pol ζ to PCNA and facilities TLS . Because a TFII-I mutant lacking its N-terminal dimerization domain is defective in TLS , our findings also implicate TLS deficiency as a potential contributing factor of WBS . Rev7 shares sequence similarity with the spindle checkpoint protein Mad2 [24] . Both contain a HORMA ( Hop1-Rev7-Mad2 ) domain that mediates protein–protein interactions [25] . Because of our long-standing interest in the structure and function of Mad2 [26]–[28] , we examined the function and regulation of Rev7 . We created 293T cell lines stably expressing human Rev7 fused at its N-terminus with a tandem affinity purification ( TAP ) tag and purified TAP-Rev7 complexes from these cells with or without UV irradiation ( 10 J/m2 ) . TAP-Rev7 preparations from both samples contained prominent doublet bands at 140 kDa ( Figure 1A ) . Mass spectrometry analysis of the unirradiated sample revealed that these bands belonged to the transcription factor , TFII-I , which was known to have multiple alternative splicing variants ( Table S1 ) . The sequence coverage of TFII-I was 52 . 4% . In addition to TFII-I , we also identified the known Rev7-binding protein , ZNF828/CAMP [29] , with a sequence coverage of 13 . 6% . Another potential Rev7-binding protein was CAD ( Carbamoyl-phosphate synthetase 2 , Aspartate transcarbamylase , and Dihydroorotase ) , a key multifunctional enzyme in the pyrimidine biosynthetic pathway , suggesting a possible link between TLS and pyrimidine biosynthesis . Rev3L was not detected in Rev7 preparations , presumably due to its low abundance in cells . Because TFII-I was the most abundant Rev7-binding protein in the TAP-Rev7 samples , we focused on the Rev7–TFII-I interaction in this study . Endogenous Rev7 and TFII-I proteins interacted with each other in human cells ( Figure 1B ) . TFII-I did not interact with Pol ι or Pol η ( Figure S1A ) . Recombinant GST-Rev7 bound to in vitro translated TFII-I ( Figure 1C ) . A minimal Rev7-binding domain of TFII-I was mapped to its middle region containing R2-R4 repeats ( Figure 1D and Figure S1 ) . This minimal TFII-I domain , however , bound more weakly to Rev7 than the full-length TFII-I did , suggesting that additional regions of TFII-I might contribute to Rev7 binding . These results suggested that Rev7 physically interacted with TFII-I . As a regulatory subunit of Pol ζ , Rev7 simultaneously binds to a small Rev7-binding motif ( RBM ) in Rev3L and the C-terminal domain ( CTD ) of Rev1 , thus bridging an interaction between Rev1 and Rev3L ( Figure S2A ) [17] . When bound to Rev3L , Rev7 adopts the closed conformation and traps Rev3L RBM with a topological embrace through its “seat belt” [30] . The Rev1 CTD binds Rev7 at a site opposite of the Rev3L-binding site [17] , [31] . We next tested whether TFII-I binding to Rev7 was compatible with Rev3L–Rev7 or Rev1–Rev7 interactions . A recombinant purified TFII-I fragment ( residues 350–667 ) containing R2-R4 co-fractionated with Rev7 bound to Rev3L RBM ( residues 1847–1898 ) with Rev1 CTD ( residues 1140–1251 ) ( Figure 1E and 1F ) . Based on gel filtration , the native molecular mass of this miniature TFII-I–Rev7–Rev3L–Rev1 complex was 78 kDa , which was consistent with the formation of a 1∶1∶1∶1 heterotetramer with an expected molecular mass of 81 kDa . With the small amount of each protein loaded , the Rev3L fragment was not visible by Coomassie blue staining . This fragment could only be visualized with large amounts of proteins loaded ( Figure S2B ) . Thus , our results suggest that TFII-I can bind to the Rev3L–Rev7–Rev1 complex in vitro . Inactivation of either subunit of Pol ζ , Rev3L or Rev7 , reduces colony formation of mammalian cells treated with UV and cisplatin , presumably because they are required to bypass DNA damage induced by these agents [32]–[34] . We first confirmed that human 293T and U2OS cells depleted of Rev3L or Rev7 with small interfering RNA ( siRNA ) were indeed sensitive to UV or cisplatin using colony formation assays ( Figure 2 ) . Because antibodies that could detect endogenous Rev3L were unavailable , the efficiency of Rev3L depletion was indirectly inferred from the reduction of Rev3L mRNA as measured by quantitative PCR ( Figure 2B ) . Depletion of TFII-I similarly resulted in UV and cisplatin sensitivity ( Figure 2 ) . Importantly , depletion of both TFII-I and Rev7 did not cause more severe phenotypes than depletion of either one alone did , suggesting that TFII-I might be required for DNA damage tolerance . Efficient depletion of Rev7 and TFII-I was confirmed by Western blots . There were no discernable cell cycle defects in cells depleted of TFII-I , Rev3L , or Rev7 in the absence of UV ( Figure S3 and data not shown ) . To obtain additional evidence for a role of TFII-I in DNA damage tolerance , we stained control and TFII-I RNAi cells for γ-H2AX , a DNA double-strand break ( DSB ) marker , at different times following the treatment of low-dose UV , and performed flow cytometry analysis . UV irradiation induced DNA damage in all samples . UV-induced DNA damage is expected to stall replication forks in S phase and indirectly produce DSBs . About 40% of all groups of cells were in S phase and positive for γ-H2AX staining at 2 hrs following UV treatment ( Figure 3A and S3 ) . At 12 hrs , the majority of these cells were γ-H2AX-positive and blocked in S phase ( Figure S3 ) . At 24 hrs after UV irradiation , few siControl cells were γ-H2AX-positive , indicating that they had progressed through S phase and effectively repaired their damaged DNA ( Figure 3A and S3 ) . In contrast , the majority of cells depleted of Rev3L or Rev7 remained blocked in S phase , and were γ-H2AX-positive ( Figure S3 ) , consistent with a known role of Pol ζ in DNA damage tolerance . Cells depleted of TFII-I were also less efficient in passing through S phase and in repairing DNA damage , as about 40% of TFII-I RNAi cells had positive γ-H2AX staining at 24 hrs after UV irradiation ( Figure 3A , B ) . Most cells positive for γ-H2AX had DNA contents between 2C and 4C , indicating that they were blocked in S phase . Importantly , ectopic expression of siRNA-resistant Myc-TFII-I transgene at levels comparable to that of the endogenous TFII-I largely rescued the S phase block and DNA damage of TFII-I RNAi cells ( Figure 3A-D ) , based on both flow cytometry and γ-H2AX immunostaining . These results indicate that , like Pol ζ , TFII-I is required for DNA damage tolerance , S phase progression , and genomic stability . Pol ζ has recently been suggested to play a direct role in DSB repair through homologous recombination [35] . Our results cannot distinguish between a direct role for TFII-I and Pol ζ in DSB repair and an indirect role for them in DNA repair through supporting TLS and DNA damage tolerance . In the future , it will be interesting to test whether TFII-I is directly involved in DSB repair . We directly tested whether TFII-I was required for translesion synthesis . To do so , we performed a mutation frequency assay on the UV-irradiated SupF shuttle vector plasmid pSP189 [36] . As expected , depletion of Rev3L or Rev7 greatly reduced the mutation frequency of the SupF region in the UV-damaged shuttle vector ( Figure 4A ) , consistent with their known roles in TLS . Depletion of TFII-I with multiple siRNAs also reduced the mutation frequency of SupF . Importantly , depletion of both TFII-I and Rev7 did not produce a stronger phenotype as did the depletion of either protein alone ( Figure 4B ) , suggesting that TFII-I might work in the same pathway as Pol ζ . DNA sequencing of the mutated SupF clones revealed that inactivation of TFII-I or Pol ζ did not alter the mutation spectrum ( Figure 4C and Figure S4 ) . In all samples , the majority of mutations were C∶G to T∶A transitions . Thus , TFII-I depletion reduces TLS efficiency in human cells . We note that there might be subtle differences in the mutation hotspots among different samples ( Figure S4 ) . In addition , mutations involving large deletions appeared to be absent in the siTFII-I cells . The significance and the underlying reasons for these apparent differences are unclear at present . Because TFII-I has known functions in transcription , we tested whether the TLS defects of TFII-I RNAi cells were indirectly caused by a defect in the transcription of TLS genes . Using quantitative PCR , we first showed that TFII-I depletion did not substantially alter the mRNA levels of Rev1 , Rev7 , Rev3L , and Rad18 , genes known to be involved in TLS ( Figure 5A ) . Next , we performed gene expression profiling of HeLa Tet-On cells transfected with siControl , siTFII-I , or siRev7 . Depletion of TFII-I reduced by two-fold the mRNA levels of only 48 genes ( Figure 5B ) . None of these genes are known to be involved in TLS . Therefore , the TLS deficiency caused by TFII-I RNAi is not an indirect consequence of gross transcriptional defects , although we cannot rule out the possibility that subtle transcriptional defects of multiple genes cumulatively impact TLS . The fact that TFII-I depletion only affects the transcription of so few genes in HeLa cells is not surprising , as many TFII-I target genes are involved in neuronal functions or immune response [37] . Furthermore , two other TFII-I related genes , GTF2IRD1 and GTF2IRD2 , might have compensated for the partial loss of TFII-I . Likewise , depletion of Rev7 only decreased the mRNA levels of about 50 genes ( Figure 5B ) . Moreover , only 12 genes were commonly suppressed in both siTFII-I and siRev7 cells . Therefore , Rev7 does not appear to have a major role in transcription . The primary function of the Rev7–TFII-I interaction is unlikely to be transcriptional regulation in HeLa cells . We next explored the mechanism by which TFII-I contributed to TLS . TFII-I contains four APIM motifs [20] , [38] , which mediates its binding to PCNA ( Figure 6A ) . Because PCNA has critical roles in mediating polymerase switching during TLS , we tested whether PCNA binding by TFII-I was required for TLS . We created a TFII-I mutant with all four APIM motifs mutated to alanine ( TFII-I mAPIM ) . The endogenous TFII-I interacted with PCNA ( Figure 6B ) . Myc-TFII-I wild type ( WT ) , but not the Myc-TFII-I mAPIM mutant protein , interacted with PCNA in cells depleted of endogenous TFII-I ( Figure 6B ) . Moreover , consistent with an earlier report [20] , GFP-TFII-I WT , but not GFP-TFII-I mAPIM , was recruited to laser-induced DNA damage sites in U2OS cells , along with DsRed-PCNA ( Figure 6C ) . This result confirmed that TFII-I mAPIM lost its functional interaction with PCNA . Importantly , Myc-TFII-I mAPIM still interacted with Rev7 ( Figure 6B ) . Compared to Myc-TFII-I WT , the Myc-TFII-I mAPIM mutant protein was significantly less efficient in rescuing TLS defects caused by TFII-I RNAi ( Figure 6D ) . Thus , the PCNA-binding activity of TFII-I is required for its function in TLS . The simplest model to explain the involvement of TFII-I in TLS is that TFII-I binds simultaneously to both PCNA and Rev7 , bridging an interaction between the two proteins and contributing to the recruitment of Pol ζ to DNA lesions . Unfortunately , we could not detect the recruitment of GFP-Rev7 to laser-induced DNA damage sites , barring us from testing this notion using cytological methods . We , therefore , tested this hypothesis using IP-Western methods . Interactions among PCNA , TFII-I , and Rev7 were detectable in unirradiated cell lysates ( Figure S5A ) . These interactions were enhanced following UV irradiation . More importantly , depletion of TFII-I abolished the interaction between PCNA and Rev7 in both cases ( Figure S5A ) . Expression of Myc-TFII-I WT , but not mAPIM , restored the interaction between PCNA and Rev7 ( Figure 6B ) . These results suggest that TFII-I bridges an interaction between PCNA and Rev7 in human cells , and that this function of TFII-I requires its APIM motifs . We next checked whether the recombinant purified TFII-I330–667 fragment could form a ternary complex with Rev7 and PCNA using gel filtration . To our surprise , we found that TFII-I330–667 , Rev7 , and PCNA did not form a ternary complex ( Figure 7A ) . Addition of PCNA to the pre-formed TFII-I330–667–Rev7 complex produced a TFII-I330–667–PCNA binary complex and free Rev7 . Thus , binding of PCNA and binding of Rev7 to a monomeric fragment of TFII-I are mutually exclusive . On the other hand , TFII-I is known to homodimerize , and contains an N-terminal dimerization domain [19] . Indeed , a TFII-I fragment containing residues 1–667 fractionated with an apparent molecular mass of about 160 kD on gel filtration columns , which was consistent with it forming a homodimer ( Figure 7B ) . By contrast , the TFII-I270–667 fragment that lacked the N-terminal dimerization domain fractionated as a monomer by gel filtration . Moreover , differentially tagged TFII-I , but not a TFII-I mutant protein lacking the first 90 residues ( TFII-I Δ90 ) , interacted with each other in human cells ( Figure 7C ) , confirming that TFII-I could oligomerize in vivo . Finally , the monomeric TFII-I Δ90 mutant protein was defective in supporting TLS ( see Figure 6D above ) . Consistently , this mutant could not restore the PCNA-Rev7 interaction in TFII-I-depleted cells ( Figure 6B ) . Therefore , homodimerization of TFII-I is required for its function in TLS and for bridging the PCNA–Rev7 interaction . We propose that one monomer of a TFII-I dimer can bind to PCNA while the other can bind to Rev7 ( Figure 7D ) . In this way , the TFII-I dimer bridges the interaction between PCNA and Rev7 , and contributes to the recruitment of Pol ζ to DNA lesions during TLS . Because Rev7 also interacts with the C-terminal domain ( CTD ) of Rev1 , we tested whether recruitment of Rev1 to DNA damage sites was dependent on TFII-I or Rev7 . GFP-Rev1 was recruited to laser-induced DNA damage sites in human cells ( Figure S5B ) . Depletion of TFII-I or Rev7 did not alter this recruitment . Thus , Rev1 is recruited to DNA damage sites independently of TFII-I and Rev7 . Taken together , our results suggest that TFII-I and Rev1 collaborate to recruit Pol ζ to DNA damage sites through TFII-I–Rev7 and Rev1-CTD–Rev7 interactions ( Figure 7D ) . Pol ζ plays critical roles in DNA translesion synthesis ( TLS ) , cellular DNA damage tolerance , and the maintenance of genomic stability . In this study , we have discovered the transcription factor TFII-I as a new , functionally important interactor of Pol ζ in human cells . We found that PCNA binding and dimerization of TFII-I are required for efficient TLS . Our study thus provides key insights into the mechanism and regulation of Pol ζ in human cells . We propose the following model to explain the involvement of TFII-I in TLS ( Figure 7D ) . In this model , Rev1 and the TFII-I homodimer are independently recruited to ubiquitinated PCNA at DNA damage sites . This complex then simultaneously engages Rev7 and recruits Pol ζ to these lesions . Rev1 also anchors Pol η , ι , or κ to PCNA . After these Y-family polymerases insert nucleotides directly opposite to the DNA lesion , Pol ζ extends DNA synthesis past the lesion . Because TFII-I specifically interacts with the Pol ζ subunit Rev7 , but not with Pol η or ι , we speculate that TFII-I might also mediate polymerase switching from Pol η/ι/κ to Pol ζ . In support of a role of TFII-I in recruiting Pol ζ to DNA lesions , we showed that TFII-I bridges an interaction between PCNA and Rev7 in UV-irradiated human cells , using IP-Western experiments . We could not reconstitute a complex containing TFII-I , PCNA , Rev7 , Rev3 , and Rev1 in vitro using purified recombinant proteins , due to the difficulty of expressing full-length TFII-I and larger fragments of Rev3 and Rev1 . Complex formation might also require DNA or additional accessory subunits of Pol ζ [39]-[41] . Rev3L has been reported to contain a putative APIM motif [20] . In addition , PolD3 ( p66 ) , an accessory subunits of Pol ζ , contains a functional PCNA-binding PIP motif [41] , [42] . Furthermore , in addition to its ability to bind ubiquitin on ubiquitinated PCNA , Rev1 has been implicated in direct binding to unmodified PCNA [8] , [43] , [44] . Therefore , along with our finding that TFII-I binds to PCNA and Rev7 , it is clear that the TLS machinery makes multiple contacts with PCNA . A cell-free system that can support PCNA- and Pol ζ-dependent TLS is needed to definitively establish the role of TFII-I in this process and dissect the relative contributions of the multiple PCNA-binding mechanisms . Finally , there are no known TFII-I orthologs in the budding yeast . It is possible that yeast Pol ζ uses distinct mechanisms to interact with PCNA . We were unable to directly test whether TFII-I is required for Pol ζ recruitment to DNA damage sites , as we could not detect the enrichment of either endogenous Rev7 at UV-induced nuclear foci using immunofluorescence or the recruitment of GFP-Rev7 to laser-induced DNA damage sites . The underlying reason for the lack of Rev7 enrichment at DNA damage sites is unclear , but could be due to the transient nature of the TFII-I/Rev1-bridged interactions between PCNA and Pol ζ . Alternatively , the Rev1–Rev7 and TFII-I–Rev7 interactions are required , but are not sufficient , to recruit Rev7 to the site of DNA damage . Only the intact , functional Pol ζ ( i . e . the Rev3L–Rev7 complex ) can be efficiently recruited . Because Rev3L is a low-abundance protein in human cells , recruitment of Pol ζ to DNA damage sites might be below the detection limits of our cytological assays . Two lines of evidence suggest that the TLS function of TFII-I is independent of its roles in transcription . First , depletion of TFII-I causes only a mild transcription defect in human cells . Of the few genes whose expression was down-regulated by TFII-I depletion , none had known roles in TLS . Second , the PCNA-binding APIM motifs of TFII-I are critical for TLS . These motifs do not have expected roles in transcription . Williams-Beuren syndrome ( WBS ) is a rare genetic disorder caused by deletion of one copy of the chromosome 7q11 . 23 region , which contains TFII-I and about 25 other genes [22] . WBS patients have multiple symptoms , including distinctive craniofacial features , mild mental retardation , and cardiovascular defects . Different phenotypes have been linked to different genes in the 7q11 . 23 region . Mice with a heterozygous deletion of N-terminal 140 residues of TFII-I exhibit craniofacial and neurobehavioral alterations [23] , implicating this region of TFII-I in WBS pathophysiology . In this study , we showed that the N-terminal region of TFII-I is critical for TLS , raising the intriguing possibility that defective TLS might underlie a subset of symptoms in WBS . It will be interesting to test whether cells derived from WBS patients exhibit sensitivity to UV irradiation and are defective in TLS , and may have defects in components of this TLS complex ( Figure 7D ) . In addition to TLS , Pol ζ is involved in somatic hypermutation , DNA interstrand crosslink repair , and DSB repair through homologous recombination [35] , [45] . Future experiments are needed to test whether TFII-I also contributes to the functions of Pol ζ in these processes . Furthermore , inactivation of Pol ζ sensitizes human cancer cells to killing by the chemotherapeutic drug , cisplatin [32] . Chemical compounds targeting Pol ζ may enhance the efficacy of cisplatin . Our discovery of TFII-I as a novel Pol ζ regulatory factor presents new opportunities for the chemical inhibition of this important polymerase complex . Finally , the general transcription factor TFIIH has a well-established role in nucleotide excision repair [1] . Our findings linking TFII-I to TLS further strengthen the general principle of component sharing in diverse nuclear processes . HeLa Tet-On , 293T , and U2OS cells were grown in Dulbecco's modified Eagle's medium ( DMEM; Invitrogen ) supplemented with 10% fetal bovine serum ( FBS ) . Plasmid and siRNA transfections were performed with the Effectene reagent ( Qiagen ) , Lipofectamine 2000 , and Lipofectamine RNAiMAX ( Invitrogen ) for 48 hrs in the indicated cell lines before the desired analysis unless otherwise noted . To establish the TAP-Rev7 cell line , 293T cells were transfected with the pIRES-Puro-TAP ( Clontech ) or pIRES-Puro-TAP-Rev7 vectors , and selected with 2 µg/ml puromycin . Individual clones were isolated for further analysis . The following siRNAs were chemically synthesized at or purchased from Dharmacon: siControl ( 5′-GACCGUUAGGUACAGAAGAUU-3′ ) , siLuc , ( 5′-UCAUUCCGGAUACUGCGAU-3′ ) , siRev7-1 ( 5′-CGGACAUUUUAAAGAUGCA-3 ) , siRev7-2 ( 5′UGCAUCUUUAAAAUGUCCG-3′ ) , and siGENOME Smartpools against human Rev3L and TFII-I . For UV-C ( 254 nm ) treatment , the growth medium was removed from the cells and reserved . Cells were washed twice with PBS . The plates ( without PBS ) were transferred to a UV cross-linker ( Stratagene ) and irradiated with the indicated UV doses . The UV-C dose delivered to the cells was confirmed with a UV radiometer ( UVP , Inc . ) . The reserved medium was added back to cells . The cells were returned to the incubator . For tandem affinity purification of TAP-Rev7 , ten 150-mm dishes of 293T cells stably expressing TAP-Rev7 were harvested in the TAP lysis buffer ( 50 mM HEPES pH 7 . 5 , 100 mM KCl , 2 mM EDTA , 10% glycerol , 0 . 1% NP-40 , 10 mM NaF , 0 . 25 mM Na3VO4 , 50 mM β-glycerolphosphate , 2 mM DTT , and 1X protease inhibitor cocktail ) . Cleared lysates were bound to IgG-Sepharose beads ( GE Amersham ) for 4 hrs at 4°C . Beads were subsequently washed three times with the lysis buffer and once with the TEV buffer ( 10 mM HEPES pH 8 . 0 , 150 mM NaCl , 0 . 1% NP-40 , 0 . 5 mM EDTA , 1 mM DTT , and 1X protease inhibitor cocktail ) . Protein complexes were cleaved off the beads by 70 µg TEV protease in TEV buffer overnight at 4°C . Supernatants were diluted in calmodulin-binding buffer ( 10 mM HEPES pH 8 . 0 , 150 mM NaCl , 1 mM magnesium acetate , 1 mM imidazole , 0 . 1% NP-40 , 6 mM CaCl2 , 10 mM 2- mercaptoethanol ) and incubated with calmodulin-sepharose beads ( GE Amersham ) for 90 minutes at 4°C . Captured protein complexes were washed three times with the calmodulin-binding buffer and the calmodulin rinse buffer ( 50 mM NH4HCO3 pH 8 . 0 , 75 mM NaCl , 1 mM magnesium acetate , 1 mM imidazole , 2 mM CaCl2 ) . Proteins were eluted in SDS sample buffer , boiled for 10 min , concentrated in microcon concentrators ( Millipore ) , and subjected to SDS-PAGE . Gels were stained with colloidal Coomassie blue stain ( Pierce ) according to manufacturer's protocols . Unique bands were excised and in-gel proteolysis was performed using modified porcine trypsin digestion overnight . The resulting peptide mixture was dissolved and subjected to nano-LC/MS/MS analysis on a ThermoFinnigan LTQ instrument , coupled with an Agilent 1100 Series HPLC system . Peptide sequences were identified using the Mascot search engine ( Matrix science ) . Those proteins identified in the TAP-REV7 purification with multiple peptides and not identified in the TAP-vector control pull-downs were considered hits . The antibodies used in this study are: α-Myc ( Roche ) , α-Rev7 ( BD Transduction ) , α-TFII-I ( Bethyl , A301-330A ) , α-Pol ι ( Bethyl , A301-303A ) , α-Pol η ( Abcam , ab17725 ) , α-tubulin ( Sigma ) , α-γH2AX ( Millipore , 05-636 ) , and α-PCNA ( Santa Cruz , PC10 ) . For immunoblotting and immunofluorescence , the antibodies were used at a final concentration of 1 µg/ml . For immunoblotting , cells were lysed in SDS sample buffer , sonicated , boiled , separated by SDS–PAGE , and blotted with the indicated antibodies . Horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse IgG ( Amersham Biosciences ) was used as the secondary antibodies . Immunoblots were developed using the ECL reagent ( Amersham Biosciences ) according to the manufacturer's protocols and exposed to film . For immunoprecipitation , cells were lysed with the lysis buffer ( 50 mM Tris-HCl , pH 8 . 0 , 250 mM NaCl , 5 mM MgCl2 , 5 mM EDTA , 0 . 5% Triton X-100 , 10 mM NaF , 80 mM β-glycerophosphate , 10% glycerol , 1 mM DTT , and protease inhibitor cocktail ) . The lysates were cleared by centrifugation for 30 min at 4°C at top speed in a microcentrifuge . Control IgG ( Sigma ) or α-TFII-I antibodies were covalently coupled to Affi-Prep protein A beads ( Bio-Rad ) . The supernatants were incubated the antibody-coupled beads . The beads were washed with the lysis buffer . Proteins bound to the beads were dissolved in SDS sample buffer , boiled , separated by SDS-PAGE , and blotted with α-Rev7 and α-TFII-I antibodies . For the immunoprecipitation of the PCNA complex , U2OS cells were fixed in PBS containing 0 . 25% formaldehyde for 10 min at room temperature , and the reaction was stopped by the addition of glycine to a final concentration of 0 . 125 M . After being washed twice with PBS , cells were resuspended in Lysis Buffer 1 ( 10 mM HEPES pH 6 . 5 , 10 mM EDTA , 0 . 5 mM EGTA , 0 . 25% Triton X-100 , 1X protease inhibitor ) and kept on ice for 10 min . Following centrifugation at 1700 g for 10 min at 4°C , pellets were washed with Lysis Buffer 2 ( 10 mM HEPES pH 6 . 5 , 200 mM NaCl , 10 mM EDTA , 0 . 5 mM EGTA , 1X protease inhibitor ) and again pelleted at 1700 g for 5 min at 4°C . Pellets were then resuspended in Lysis Buffer 3 ( 25 mM Tris-HCl pH 6 . 5 , 100 mM NaCl , 2 mM MgCl2 , 1X protease inhibitor , 10% glycerol , 1 mM DTT , 10 mM BGP , 5 mM NaF , 3 mM NaVO4 , Turbo nuclease ) , incubated on ice for 10 min , and sonicated . Lysates were then centrifuged at 14 , 000 rpm for 15 min at 4°C . The supernatant was incubated with Affi-Prep Protein A beads coupled to α-PCNA for 3 h at 4°C . Beads were washed five times with Lysis Buffer 3 . Protein crosslinks were reversed by incubating the beads in SDS buffer at 95°C for 30 min . Proteins bound to beads were analyzed by SDS-PAGE and immunoblotting . For immunofluorescence , HeLa Tet-On cells transfected with the indicated siRNAs were plated in four-well chamber slides ( LabTek ) , treated with 10 J/m2 UV or left untreated , and fixed with 4% paraformaldehyde in 250 mM HEPES pH 7 . 4 , 0 . 1% Triton X-100 at 4°C for 20 min . After 3–5 washes over 20 min in PBS , cells were permeabilized in PBS containing 0 . 5% Triton X-100 for 20 min , and then washed with PBS . The cells were blocked in PBS containing 5% BSA followed by a 2-h incubation with the primary antibodies . After 3–5 washes over 20 min with PBS , cells were incubated with fluorescent secondary antibodies ( Alexa Fluor 488 or 647 , Molecular Probes ) for 30 min at room temperature . After incubation , cells were washed with PBS , and their nuclei were stained with DAPI ( 1 µg/ml ) . Slides were mounted and viewed with a 100X objective on a DeltaVision microscope . All images were taken at 0 . 2 µm intervals , deconvolved , and stacked . The images were further processed in ImageJ . For GST pulldown assays , Myc-TFII-I or its fragments were in vitro translated in rabbit reticulocyte lysate in the presence of 35S-methionine and incubated with bacterially expressed GST or GST-Rev7 in the binding buffer ( 25 mM Tris-HCl pH 8 . 0 , 2 . 7 mM KCl , 137 mM NaCl , 0 . 05% Tween-20 ) for 1 h at room temperature . Protein complexes were then bound to Glutathione-Sepharose beads for 30 min at room temperature . After 5 washes with the binding buffer , the proteins were eluted with SDS sample buffer , boiled , and subjected to SDS-PAGE . The bound proteins were analyzed with a phosphor imager ( Fujifilm ) to visualize 35S-labeled TFII-I and Coomassie staining to visualize GST and GST-Rev7 . Human His6-Rev7 R124A mutant bound to the Rev7-binding region of human Rev3L ( residues 1847–1898 ) and untagged human PCNA were prepared as previously described [46] , [47] . ( Rev7 forms dimers in vitro , but in vivo function of this dimerization event is unclear . The R124A mutation disrupts Rev7 dimerization . ) Human TFII-I fragments and the C-terminal domain ( CTD ) of Rev1 ( residues 1140–1251 ) were expressed as GST-fusion proteins in bacteria and purified with the glutathione-Sepharose 4B resin ( GE Healthcare ) . The eluted proteins were digested with the PreScission protease ( GE Healthcare ) , and further purified with anion exchange and size exclusion chromatography . To assay for complex formation and to determine the apparent molecular weight of the complexes , the gel filtration standard ( Bio-Rad , 151-1901 ) , Rev7–Rev3L–TFII-I , PCNA–TFII-I , Rev7–Rev3L–Rev1–TFII-I , and Rev7–Rev3L–TFII-I in the presence of PCNA were fractionated on a Superdex 200 10/300 GL column ( GE Healthcare ) . 293T cells were transfected twice with the indicated siRNAs in a 24 h period , and replated into six-well plates at 60 h after the first siRNA treatment , with 500 , 2000 , 10 , 000 , and 40 , 000 cells per well . After another 24 h , cells were exposed to varying doses of UV ( 0 , 4 , 8 , 12 , and 16 J/m2 ) . Twelve days later , colonies were fixed and stained in a solution containing 3∶1 methanol and glacial acetic acid plus 1% trypan blue ( Sigma ) . Colonies containing 50 or more cells were counted . The surviving fractions for each group represent the plating efficiency for each treatment divided by the plating efficiency of the corresponding untreated control samples . 293T cells were transfected with the appropriate siRNAs . At 24 h after siRNA transfection , pSP189 plasmids were irradiated with UV ( 1000 J/m2 ) and transfected into the cells using Lipofectamine 2000 ( Invitrogen ) . Cells were harvested 48 h later for plasmid purification using the DNA miniprep kit ( QIAGEN ) . The purified plasmids were digested with DpnI and transformed into the bacterial strain MBM7070 by electroporation . Bacterial cells with wild-type SupF tRNA expressed functional β-galactosidase and formed blue colonies on X-gal plates , whereas bacteria with mutated SupF formed white colonies . The mutation frequency in the SupF gene was analyzed by counting the ratio between blue ( wild-type ) and white ( mutant ) colonies . Mutations in the SupF gene were confirmed by DNA sequencing . HeLa Tet-On cells transfected with the appropriate siRNAs were irradiated with 10 J/m2 UV . Samples were taken at the indicated timepoints , fixed with 70% ice-cold ethanol , blocked with PBS containing 5% BSA and 0 . 25% Triton-X100 , and stained with the anti-γ-H2AX monoclonal antibody . Cells were washed , incubated with the Alexa Fluor 488 donkey anti-mouse secondary antibody ( Invitrogen ) , and counterstained with propidium iodide in PBS containing RNase A . Cells were analyzed with a BD FACSCalibur flow cytometer by using the CellQuest software . Data were processed with FlowJo ( FloJo , Ashland , OR ) . Total RNA was harvested from untreated and siRNA-treated HeLa Tet-On cells at 48 h after siRNA transfection using the RNeasy Kit ( Qiagen ) . cDNA was synthesized from the total RNA , purified , and hybridized to a HumanHT-12 v4 BeadChip array at the UTSW Microarray Core facility . The arrays were then washed , stained , and scanned according to the manufacturer's protocol . For quantitative PCR ( qPCR ) , cells were lysed in Trizol ( Invitrogen ) . Total RNA was extracted by chloroform extraction and isopropanol precipitation . About 1–2 µg of total RNA was reverse transcribed with the high-capacity cDNA reverse transcription kit ( Applied Biosystems ) according to the manufacturer's instructions . Taqman probes for human Rev3L ( Hs01076848_m1 ) , Rev1 ( Hs01019771_m1 ) , Rev7 ( Hs01057448_m1 ) , and Rad18 ( Hs00892551_m1 ) , and GAPDH ( Applied Biosystems ) were used for qPCR in TaqMan master mix ( Applied Biosystems ) according to the manufacturer's protocol . Samples were run in triplicates with the appropriate negative controls . U2OS cells were transfected with DsRed-PCNA and GFP-TFII-I or GFP-Rev1 along with the appropriate siRNAs . DSBs were introduced in the nuclei of cultured cells by microirradiation with a pulsed nitrogen laser ( Spectra-Physics; 365 nm , 10 Hz pulse ) [48] . The laser system was directly coupled ( Micropoint Ablation Laser System; Photonic Instruments , Inc . ) to the epifluorescence path of an Axiovert 200 M microscope ( Carl Zeiss MicroImaging , Inc . ) for immunostaining imaging or time-lapse imaging and focused through a Plan-Apochromat 63×/NA 1 . 40 oil immersion objective ( Carl Zeiss MicroImaging , Inc . ) . The output of the laser power was set at 60% of the maximum . Time-lapse images were taken with an AxioCam HRm ( Carl Zeiss MicroImaging , Inc . ) . During microirradiation , imaging , or analysis , the cells were maintained at 37°C in 35-mm glass-bottom culture dishes ( MatTek Cultureware ) . The growth medium was replaced by CO2-independent medium ( Invitrogen ) before analysis . The images were further processed by ImageJ and Photoshop .	DNA translesion synthesis ( TLS ) allows the DNA replication machinery to replicate past damaged bases , thus increasing cellular tolerance for DNA damage and maintaining genomic stability . Suppression of TLS is expected to enhance the efficacy of the anti-cancer drug , cisplatin . TLS employs a special set of DNA polymerases , including Pol ζ . The TLS polymerases are also involved in somatic hypermutation and proper immune response in mammals . Thus , it is critical to understand the underlying mechanisms of TLS . In this study , we have discovered the transcription factor TFII-I as a new Pol ζ-binding protein in human cells . We show that TFII-I is indeed required for TLS and DNA damage tolerance . We further delineate the mechanism by which TFII-I contributes to TLS . Our study significantly advances the molecular understanding of TLS , and provides a fascinating example of component sharing among disparate nuclear processes . Finally , because one copy of the TFII-I gene is deleted in Williams-Beuren syndrome ( WBS ) , our work implicates TLS deficiency as a potential causal factor of this human genetic disorder .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "genetics", "biochemistry", "biology", "and", "life", "sciences" ]	2014	The Transcription Factor TFII-I Promotes DNA Translesion Synthesis and Genomic Stability
Enterohemorrhagic Escherichia coli ( EHEC ) is one major type of contagious and foodborne pathogens . The type VI secretion system ( T6SS ) has been shown to be involved in the bacterial pathogenicity and bacteria-bacteria competition . Here , we show that EHEC could secrete a novel effector KatN , a Mn-containing catalase , in a T6SS-dependent manner . Expression of katN is promoted by RpoS and OxyR and repressed by H-NS , and katN contributes to bacterial growth under oxidative stress in vitro . KatN could be secreted into host cell cytosol after EHEC is phagocytized by macrophage , which leads to decreased level of intracellular reactive oxygen species ( ROS ) and facilitates the intramacrophage survival of EHEC . Finally , animal model results show that the deletion mutant of T6SS was attenuated in virulence compared with the wild type strain , while the deletion mutant of katN had comparable virulence to the wild type strain . Taken together , our findings suggest that EHEC could sense oxidative stress in phagosome and decrease the host cell ROS by secreting catalase KatN to facilitate its survival in the host cells . Enterohemorrhagic Escherichia coli ( EHEC ) is a globally important zoonotic pathogen capable of causing diarrhea , hemorrhagic colitis and hemolytic–uremic syndrome ( HUS ) [1] . A prominent feature of EHEC infection is its low infectious dose . A dose of 50–100 colony forming units ( CFUs ) of EHEC is sufficient to cause disease in healthy individuals [2] . EHEC consists of multiple serotypes , among which O157:H7 is the one most commonly linked to epidemic and sporadic diseases in humans throughout China , Japan , North America , and Europe . For efficient colonization in human intestines , EHEC needs adhere to the follicle-associated epitheliums , which results in the rapid contact of EHEC with underlying macrophage cells [3 , 4] . Macrophages are important components of the host innate immune system , and the lamina propria of the large intestine is rich in macrophages . When the intestinal epithelial barrier is damaged and microorganisms cross the basement membrane , the macrophages may be the first line of host defense system . Macrophage cells produce and release reactive oxygen species ( ROS ) in response to phagocytosis or various stimuli [5] . ROS are short-lived molecules derived from incomplete reduction of oxygen metabolites that at high levels have a bactericidal function by damaging DNA , lipid , proteins , and membrane [6] . To survive , replicate , and disseminate throughout the body , bacterial pathogens , especially these intracellular bacteria must overcome the antimicrobial oxidative burst produced by macrophages . General ROS degradation enzymes including catalases , peroxidases , and superoxide dismutases are used by most bacteria to survive under oxidative stress [7] . E . coli possesses multiple distinct catalases to defend itself against oxidative stress , including hydroperoxidase I ( HPI ) , KatG and hydroperoxidase II ( HPII ) , KatE [8 , 9] . The alkyl hydroperoxide reductase complex , consisting of AhpC and AhpF , could catalyze the reduction of hydrogen peroxide in an NADH-dependent manner in E . coli [10] . Another plasmid-borne catalase-peroxidase , KatP , has been found to contribute to the complex gene network protecting EHEC from peroxide-mediated oxidative damage [11] . The intracellular survival of bacterial pathogens relies on the specialized secretory systems , which can inject bacterial effectors into the cytosol of host cells . The type VI secretion system ( T6SS ) is widely spread in both pathogenic [12] and non-pathogenic Gram-negative bacteria [13] , and contributes to competition in bacterial communities by delivering bacteriolytic toxins to target cells [14] . Besides participating bacterial competition , T6SS is also involved in the pathogenicity of several Gram-negative pathogens [15–20] . More than 10 orthologs of known T6SS components are uncovered in genome-sequenced pathogenic E . coli strains by in silico analysis [21] . It has been shown that the T6SSs in both avian pathogenic E . coli ( APEC ) [22–24] , enteroaggregative E . coli ( EAEC ) [16] contribute to bacterial virulence . In the previous study , we showed that T6SS is functional in neonatal meningitis-causing E . coli K1 ( NMEC ) , and two Hcp family proteins participate in different steps of bacterial interaction with human brain microvascular endothelial cells ( HBMEC ) in a coordinate manner , e . g . , binding to and invasion of HBMEC , the cytokine and chemokine release followed by cytoskeleton rearrangement , and apoptosis [19] . Our preliminary data showed that both EHEC strains EDL933 and Sakai contain the T6SS gene cluster . We thus question whether the T6SS gene cluster plays roles in EHEC infection of host cells and/or competition with other bacteria . In this study , we confirmed that the T6SS of EHEC strain EDL933 is functional , and a novel T6SS effector protein KatN was identified to be the catalase secreted by EHEC to antagonize ROS of eukaryotic host cells to evade host immune killing . Genome sequence analysis showed that a 37-kb DNA fragment encoding a putative T6SS existed in the genome of EHEC strain EDL933 . This fragment possessed several typical features of a composite pathogenicity island: it is associated with a tRNA-encoding gene aspV and harbors virulence genes . Annotation showed that this gene cluster contained z0250 , z0254 , z0255 , hcp-1 , hcp-2 , hcp-3 , and z0267 , which are homologous to vasK ( icmF-like ) , vasH ( clpV homolog ) , vasF ( dotU homolog ) , hcp1 , hcp2 , hcp3 , and vgrG in Vibrio cholerae , respectively ( Fig 1 ) . These genes represented the core and conserved accessory components of the T6SS , suggesting the T6SS is functional in EHEC . The detailed annotation of EHEC T6SS ORFs was listed in the S1 Table . In silico analysis also showed that EHEC had another two vgr genes , vgrG-2 ( z0707 ) and vgrE ( z2262 ) located in the other regions of genome . We then used our integrated database SecReT6 ( http://db-mml . sjtu . edu . cn/SecReT6/ ) to compare the T6SS sequences [25] , and found that the components and gene organization of the T6SS gene clusters were highly conserved in pathogenic E . coli strains compared with other bacteria including V . cholerae and Salmonella Typhimurium ( S1 Fig ) . We reanalyzed our previous RNA-seq data and found that the expression of T6SS in vitro was relatively low compared with other genes in EHEC genome [26] . The average reads per kilobase per million ( RPKM ) of T6SS gene cluster ( z0243 to z0275 ) was 5 . 7 , while the average RPKM of EHEC whole genome was 198 . 1 ( S2 Fig ) . We then tested different culture conditions , including salt , pH , media and growth temperature , to stimulate the expression of T6SS in vitro . However , the qPCR results showed that these conditions could not induce the expression of T6SS ( S3 Fig ) , suggesting T6SS may not be required for EHEC growth in vitro . Consistent with this speculation , the growth rate of the deletion mutant of T6SS was comparable to that of the parental strain ( S4 Fig ) . Because the highly conserved histone-like global repressor H-NS was found to repress the expression of T6SS in V . cholerae [27] , Edwardsiella tarda [28] and S . Typhimurium [29] , we speculate that H-NS may be a repressor of the T6SS in EHEC . We thus constructed the deletion mutant of hns in EHEC by replacing hns gene with chloramphenicol resistance cassette using λ Red recombination system [30] . Then , the bacterial total RNA was isolated , and the transcription level of T6SS genes was measured by quantitative real-time polymerase chain reaction ( qPCR ) . The results showed that the expression of all the tested T6SS genes were upregulated in the deletion mutant of hns significantly ( Fig 2A ) . Hcp-2 ( Z0264 ) is a homologue of Hcp ( Hemolysin co-regulated protein ) , which is an essential member of T6SS and involved in the assembly of the cogwheel-like protein complex in V . cholerae [31] . As a hallmark of T6SS , the transcription of hcp-2 in the deletion mutant of hns increased by 60-fold compared with that of the parental strain . To exclude the potential polar effect of hns deletion , we used pACYC184 to deliver a copy of hns to the deletion mutant of hns . The qPCR result showed that the transcription of T6SS genes in the deletion mutant of hns was inhibited by hns complementation ( Fig 2B ) . The secretion of the T6SS hallmark effector protein Hcp is considered a reliable indicator of a functional T6SS [18 , 19] . We then used Western blot to investigate the secretion of Hcp to confirm the activity of T6SS and the repression of H-NS on the T6SS in EHEC . The deletion mutant of T6SS was constructed by replacing the 37 kb T6SS gene cluster with chloramphenicol resistance cassette . The mutant was verified by PCR and sequencing before further experiments . The Hcp-2 ( Z0264 ) was fused by a His-tag sequence at C-terminus in the plasmid ( pQE80-z0264 ) , and the plasmid was transferred to the deletion mutants of T6SS or hns individually . After IPTG induction , the supernatants from these two strains were isolated and separated in SDS-PAGE . Western blot using anti-His antibody showed that Hcp-2 was expressed in the cytosol in both strains , but was only detected in the supernatant of the deletion mutant of hns ( Fig 2C ) , suggesting that the T6SS was functional and could secrete hallmark effector protein Hcp-2 in EHEC . Since the transcription of T6SS genes was upregulated in the deletion mutant of hns , we wonder whether the assembly of T6SS apparatus was also increased in this mutant . The assembly of an active T6SS can be visualized by fluorescent tagged ClpV , which is an AAA+ family protein that has been postulated to couple ATP hydrolysis to T6SS effector translocation . A functional T6SS apparatus results in a focused localization of ClpV , instead of diffused distribution [32] . We then fused GFP ( green fluorescent protein ) to the C-terminus of ClpV ( Z0254 ) to visualize T6SS by co-incubating EHEC strains with E . coli K-12 strain MG1655 at a ratio of 10:1 . The results showed that more hns mutant cells ( 47 . 0% ) had GFP foci compared with the wild type strain ( 17 . 5% ) , indicating the deletion of hns could derepress the assembly of T6SS apparatus ( Fig 2D ) . Interestingly , most GFP foci showed apparent polar localization in the deletion mutant of hns , which was contradictory to the central location of T6SS apparatus with bactericidal activity [33] . EHEC infection specifically results in the pathological attaching and effacing lesions of the follicle-associated epithelial layer [34] , followed by rapid contact with underlying human macrophage cells . To test the role of T6SS in EHEC interaction with macrophages , we used RAW264 . 7 murine macrophage-like cells to study the intracellular survival of the wild type EHEC strain and the deletion mutant of T6SS . After 20 h incubation , the survival rate of the wild type strain in RAW264 . 7 cells was about 38 . 0% , while the survival rate of the deletion mutant of T6SS decreased dramatically to 24 . 8% ( Fig 2E ) , suggesting that T6SS was involved in the intracellular survival of EHEC . Recently , several findings reveal that T6SSs in some bacterial species have antagonistic bactericidal activity towards heterologous bacterial species [35–38] . The bactericidal activity of T6SS provides a strong competitive advantage against other bacteria in the host or the natural environments . As an intestinal pathogen , EHEC can colonize in the human intestine where a dense and diverse intestinal microbiota comprising a large amount of various bacteria exists [39] . It has been reported that P . aeruginosa could effectively kill Acinetobacter baylyi T6SS+ cells [33] , so these two strains were included in the competition assay as the positive control . When mixed with P . aeruginosa ΔretS , a strain showing a dramatic increase in T6SS activity due to the derepression of T6SS by retS deletion , ~ 1000-fold fewer A . baylyi were recovered compared to that mixed with P . aeruginosa ΔppkA , a T6SS defective strain ( Fig 3A ) . However , P . aeruginosa ΔretS could not kill the wild type EHEC strain ( Fig 3B ) or the deletion mutant of T6SS , and vice versa ( Fig 3C and 3D ) . It has been shown that certain T6SSs can kill T6SS negative ( T6SS- ) species [33] . We wonder whether EHEC could kill T6SS- bacteria ( e . g . E . coli K12 strain MG1655 ) . As shown in Fig 3E , none of EHEC , Δhns , ΔT6SS or the double deletion mutant ( Δhns/T6SS ) could kill E . coli K12 strain , suggesting that EHEC did not have detectable bactericidal activity in vitro . Since T6SS was critical to EHEC survival in macrophages ( Fig 2E ) , we propose that T6SS may utilize unknown effector ( s ) to achieve its roles in vivo . The supernatants from the deletion mutant of T6SS and the WT cultures were isolated and purified by combining ultrafiltration membrane package and centrifugal filter [19] , and separated in SDS-PAGE ( S5A Fig ) . The quality of the supernatants was confirmed by Western blot using the antibody of RNA polymerase subunit alpha ( RpoA ) , a cytosol marker of E . coli ( S5B Fig ) . The supernatants were applied to liquid chromatography-tandem mass spectrometry ( LC-MS/MS ) , and proteins that were only present in the supernatant of the wild type strain but not that of the T6SS deletion mutant were considered as potential T6SS effector candidates in EHEC . We repeated the above LC-MS/MS analysis using two sets of biological samples and found 46 and 31 T6SS effector candidates , respectively ( S2 Table ) . Further analysis showed that three candidates , Z0873 , Z1921 and Z5583 were detected by both LC-MS/MS performances , suggesting these three proteins were likely to be T6SS effectors in EHEC . We next used His-tag fusion strategy to verify the secretion pathway of the three candidates . The plasmids harboring z0873 , z1921 or z5583 with a His-tag sequence fusion at their C-termini were constructed and transformed to the T6SS deletion mutant or the wild type strain individually . The culture supernatants from these strains were isolated by trichloroacetic acid ( TCA ) precipitation [40] and separated in SDS-PAGE followed by Western blot using anti-His tag antibody . The results showed that all the three proteins were expressed in the bacterial cytosols , and only z0873 and z1921 could be secreted into supernatants . Because Z0873 was detected in the supernatants of both the WT and T6SS deletion mutant , we assumed that this protein was secreted in a T6SS-independent manner in EHEC and excluded for further analysis . Interestingly , Z1921 was only detected in the supernatant of the WT , but not in that of T6SS deletion mutant ( Fig 4A ) , suggesting that the secretion of Z1921 was dependent on the T6SS in EHEC . Amino acids sequence analysis showed that Z1921 of EHEC shared 84% identity and 93% similarity with Mn-containing catalase KatN of Salmonella enterica [41] , therefore Z1921 was named as KatN ( S6 Fig ) . We then transformed the plasmid harboring KatN-His tag fusion into the deletion mutant of z0254 ( clpV ) . The supernatants of strains Δz0254 , ΔT6SS and the WT bearing KatN-His tag fusion plasmid were isolated and applied to Western blot to check the secretion pathway of KatN . The result showed that the loss of z0254 alone or T6SS as a whole caused the failure of KatN secretion in EHEC ( Fig 4B ) . To consolidate these findings , we transformed pCX-katN ( KatN-bla fusion was made by the insertion of the bla gene at the C-terminus of katN ) into the WT and Δz0254 or ΔT6SS individually and examined the β-lactamase activity in the bacterial culture supernatants as described previously [19] . In this condition , the secretion of β-lactamase is dependent on the secretion of KatN . As expected , significant color change of nitrocefin was detected in the culture supernatant of WT-pCX-katN , while only marginal color alteration was observed in these of ΔT6SS-pCX-katN or Δz0254-pCX-katN , which may be due to the leaky expression of β-lactamase ( Fig 4C ) . In addition , we also examined the translocation of the KatN fusions in infected RAW264 . 7 cells by using TEM-1 β-lactamase as a fluorescence-based reporter [42] . EHEC strains expressing the above-mentioned fusion proteins were used to infect RAW264 . 7 cells , and the KatN-bla fusion protein levels at the WT and ΔT6SS were comparable ( S7 Fig ) . After 3 h incubation , the cells were washed and incubated for an additional 60 min with the β-lactamase substrate CCF2 . The cells were then analyzed by a confocal microcopy with the simultaneous observation of the green fluorescence emitted by the CCF2-AM substrate and the blue fluorescence emitted by the cleaved CCF2-AM . Cells infected with ΔT6SS harboring pCX-katN fusions appeared dominantly green , indicating the absence of β-lactamase activity in these cells . In contrast , cells infected with EHEC expressing KatN fused to β-lactamase showed strong blue fluoresce signals ( Fig 4D ) . The percentage of blues cells was about 76% in RAW264 . 7 cells infected by the WT EHEC bearing pCX-katN , while the percentages in control groups ( WT-pCX340 and ΔT6SS-pCX-katN ) were only about 33% ( S8 Fig ) , indicating that KatN was efficiently translocated into the cells . To avoid any effects that might result from overexpression of a plasmid-encoded gene , we then prepared polyclonal antibody of KatN and used this antibody to check the secretion of KatN in the supernatants of the deletion mutants of z0254 , T6SS or the WT . The results again confirmed that KatN could only be detected in the supernatants of the WT , but not in these of the deletion mutants of z0254 or T6SS ( Fig 4E ) . Furthermore , our data showed that the secretion deficiency of KatN due to the loss of z0254 could be restored by complementation ( S9 Fig ) . Since T3SS of EHEC is responsible for the secretion of multiple effectors [43] , we then tested the relationship between the secretion of KatN and T3SS , and results showed that the absence of a key T3SS gene , escN , did not disturb the secretion of KatN , indicating that T3SS was not involved in the secretion of KatN in EHEC ( S10 Fig ) . Taken together , these data clearly demonstrated that KatN was secreted via T6SS in EHEC . Since KatN of EHEC may be a Mn-containing catalase , we then overexpressed and purified KatN to apparent homogeneity ( S11A Fig ) to check its catalase activity . The result showed that the specific activity of KatN is 268 . 3 U/mg protein ( S11B Fig ) . Since KatN has a high catalase activity , we wonder whether it contributes to the EHEC response to oxidative stress . Thus , we used different concentrations of hydrogen peroxide to treat the WT , ΔkatN and ΔkatN-c , and monitored the bacterial growth . The absence of katN did not affect the bacterial growth in regular LB medium ( S4 Fig ) , but caused the growth defect of EHEC under hydrogen peroxide concentrations from 1 mM to 2 mM ( Fig 5A ) . This growth retardation could be restored by complementation of a copy of katN , indicating that katN contributed to EHEC response to oxidative stress in vitro . It has been shown that OxyR is a principal regulator for hydrogen peroxide detoxification , and RpoS is a general stress response regulator at the stationary phase [44] . Since the expression of catalases KatE and KatG was regulated by OxyR and RpoS [45] , we speculate that expression of katN might be regulated by these two regulators . The deletion mutants of oxyR , rpoS as well as the WT and ΔT6SS were treated by hydrogen peroxide , and qPCR was employed to determine the transcriptional levels of katN compared to mock treatment in these strains . The results of qPCR showed that both OxyR and RpoS were involved in the activation of katN regardless of the presence or absence of hydrogen peroxide . Specifically , RpoS was essential for katN transcription , and OxyR promoted the transcription of katN ( Fig 5B ) . Morgan et al . demonstrates that KatG and AhpC are induced by hydrogen peroxide in S . Typhimurium [46] , while our data showed that hydrogen peroxide did not induce the transcription of katN in EHEC , which is similar to the case of HPII KatE [47] . Moreover , the deletion of T6SS gene cluster did not affect the transcription of katN . The protein levels of KatN in the above conditions were determined by Western blot using anti-KatN antibody , showing that KatN protein levels were well correlated with the mRNA levels of katN in both log phase and stationary phase ( Fig 5C ) . As a global regulator , H-NS could inhibit the transcription of T6SS in EHEC . We then tested whether the transcription of katN and other catalase genes , katG and katE is also regulated by H-NS . The qPCR results showed that transcriptional level of katN increased by over 450-fold in the deletion mutant of hns compared with that of the WT ( Fig 5D ) . Another catalase gene katE was upregulated in the deletion mutant of hns , however , transcription of katG was barely promoted in the deletion mutant of hns . The repression of H-NS on the expression of katN was further confirmed by Western blot using anti-KatN antibody ( Fig 5E ) . EHEC has multiple catalases to resist oxidative stress [11 , 48 , 49] . Given the previous evidence that KatN was secreted by T6SS , we further wonder whether EHEC could secrete other catalases through the same secretory apparatus for better intracellular survival . We then used His-tag strategy to determine the secretion pattern of all the catalases KatE , KatG , KatP and AhpC in EHEC . The results showed that all the catalases were expressed in the cytosols of the WT , ΔT6SS and Δz0254 , but none of these catalases could be secreted ( S12 Fig ) , suggesting KatN was the sole secreted catalase in EHEC up to date . EHEC infection specifically results in the pathological attaching and effacing lesions of the follicle-associated epithelium [34] , followed by its rapid contact with underlying human macrophage cells . To test the role of katN in EHEC interaction with macrophages , we used murine macrophage cell line RAW264 . 7 to study the survival of the wild type EHEC strain and its derived mutants . After 20 h incubation , the survival rate of the wild type strain in RAW264 . 7 cells is about 38 . 0% , while the survival rate of the deletion mutant of katN decreased dramatically to 16 . 1% . The survival defect of the deletion mutant of katN could be restored to by introducing a copy of katN ( Fig 6A ) . This observation was confirmed by using primary peritoneal macrophage cells ( Fig 6B ) , suggesting that katN contributed to the survival of EHEC in macrophage cells . Since T6SS played an important role in the survival of EHEC in macrophage cells ( Fig 2E ) , while the expression of T6SS was relatively low in vitro ( S2 Fig ) , we speculate that the expression of T6SS may be induced when EHEC was phagocytized . To test this hypothesis , we infected RAW264 . 7 cells with the wild type EHEC strain and isolated the intracellular bacteria , followed by total RNA purification and qPCR analysis . The result showed that the expression of T6SS was upregulated dramatically . The transcriptional levels of all the tested genes including z0254 , z0264 , z0266 and z0267 were increased by 40 , 8 , 7 and 8 fold compared with these of the genes in vitro , respectively ( Fig 6C ) . We also studied the transcriptional levels of all the catalases genes in EHEC including katN , katE , katG , katP , and ahpC and found that all catalases genes were upregulated by 4 fold to 67 fold in the intracellular bacteria except ahpC ( S13 Fig ) . These data suggested that after phagocytosis by macrophage , EHEC induced the expression of T6SS and catalases for a better intramacrophage survival . Next , we want to give direct evidence that KatN is secreted when EHEC is phagocytized by macrophage . To address this question , we used the wild type EHEC strain , the deletion mutant of T6SS , the deletion mutant of katN , and katN complementation strain by introducing KatN-expression plasmid in the katN deletion background ( ΔkatN-c ) to infect RAW264 . 7 cells . Cells were then stained by DAPI ( 4’ , 6’-diamidino-2-phenylindole ) , anti-EHEC LPS antibody or anti-KatN antibody , followed by corresponding fluorescent-dye conjugated secondary antibodies detection . The anti-EHEC LPS antibody indicated that the infection rates of RAW264 . 7 cells by different EHEC strains were similar . As we expected , the WT EHEC-infected RAW264 . 7 cells showed strong KatN fluorescence signal in the cytosol , while neither ΔT6SS- nor ΔkatN-infected RAW264 . 7 cells showed KatN signal . When a copy of katN was introduced into ΔkatN , the signal of KatN reappeared in the cytosol of RAW264 . 7 cells ( Fig 6D ) . As innate immune cell species , macrophages can generate ROS to kill some types of engulfed bacteria [5] . Since EHEC could secrete catalase KatN into host cell cytosol via T6SS , we speculate that EHEC may utilize the secreted KatN to antagonize host cell ROS to escape from killing by phagocytic cells . Therefore , we used HyPer-3 , a genetically encoded fluorescent indicator to determine intracellular ROS level in macrophages . Specifically , ROS can change the excitation spectrum of HyPer-3 with an excitation maximum at 405 nm for the reduced and 488 nm for the oxidized state , and the ratio of signals at 488 nm and 405 nm presents the intracellular ROS level [50] . In other words , higher ratio of signals at 488/405nm indicates a more oxidized intracellular environment , or high ROS stress . After 1 h incubation , the ratio between both channels ( 488 nm/405 nm ) in uninfected cells was 1 . 17±0 . 03 , while the ratio in the wild type EHEC-infected cells increased to 1 . 41±0 . 01 . As we expected , the ratio of 488 nm/405 nm of RAW264 . 7 cells infected by the deletion mutant of T6SS increased to 1 . 52±0 . 05 ( vs that of WT-infected cells: P<0 . 01 ) ( Fig 6E ) . Interestingly , the ratio of 488 nm/405 nm in the deletion mutant of katN-infected cells reached 1 . 80±0 . 03 ( vs that of WT-infected cells: P<0 . 001 ) , while introduction of katN in the deletion mutant of katN ( ΔkatN-c ) decreased the ratio of 488 nm/405 nm to 1 . 28±0 . 12 ( vs that of ΔkatN-infected cells: P<0 . 001 ) . To confirm this observation , we used DCFH-DA ( 2 , 7-dichlorodi hydrofluorescein diacetate ) to directly monitor the production of intracellular ROS [51] . We infected RAW264 . 7 cells with the deletion mutant of hns ( Δhns ) , double deletion mutant of hns and katN ( Δhns/ΔkatN ) , and katN complementation strain ( Δhns/ΔkatN-c ) . After incubation and washing , DCFH-DA was applied to the cells followed by flow cytometry analysis . As we expected , the absence of katN increased the ROS level of RAW264 . 7 cells , which could be reverted by introducing a copy of katN ( S14 Fig ) . These data showed that EHEC could utilize T6SS to secrete catalase KatN into host cells to decrease the ROS level and facilitate bacterial intracellular survival . To determine the contribution of T6SS to EHEC virulence in mouse model , the WT EHEC , ΔT6SS or ΔkatN strains were used to infect mice , respectively . Groups of 8 BALB/c mice were infected by oral gavage with 7×1010 CFU bacteria or Phosphate-buffered saline ( PBS ) , and monitored for survival over a 16-day period . Mice infected with the WT or ΔkatN appeared ruffled fur immediately after administration and began dying both at day 3 post-infection , and were all dead within 5 days . However , 40% of ΔT6SS-infected mice were still alive ( vs WT: P<0 . 001 , Log-rank analysis ) and showed no infection-associated morbidity such as ruffling of fur and wasting ( Fig 7A ) . We also monitored the mice body weight during the whole experiment period , and the result showed that the WT- and ΔkatN-infected mice lost body weight dramatically after administration ( vs ΔT6SS: P<0 . 001 , at day 3 , Student’s t test ) . Instead , ΔT6SS-infected mice lost body weight slightly compared with PBS control , and started to gain body weight from day 5 post-infection as PBS control ( Fig 7B ) . The bacterial load in feces was determined from day 1 to day 15 post-infection . The result showed that the colonization of all the three EHEC strains in the feces kept decreasing from day 1 post-infection . The bacterial burden in feces of ΔT6SS-infected mice was 7 . 6×107 CFU , while the bacterial burden in feces of the WT- or ΔkatN-infected mice were 8 . 8×108 CFU and 9 . 0×108 CFU ( WT or ΔkatN vs ΔT6SS: P<0 . 001 at day 3 ) , respectively on day 3 post-infection ( Fig 7C ) . After 10 days post-infection , EHEC strains could not be detected in the feces of ΔT6SS infected mice , indicating that the bacteria were completely cleared from the gut . T6SS has been shown to be widely distributed in up to 25% of Gram-negative bacteria [21] . Although most of these bacteria possess only one single T6SS gene cluster , some species harbor multiple distinct T6SS copies . For example , five phylogenetically distinct T6SS loci have been identified in S . enterica [52] , Burkholderia pseudomallei and Burkholderia thailandensis display six and five T6SSs , respectively [53] , and P . aeruginosa possesses three T6SS clusters [54] . The distribution and copy number of T6SS in E . coli are various . APEC and EAEC contain up to 3 phylogenetically distinct T6SS clusters [55] , while UPEC strain CFT037 only harbors one set of T6SS locus . There is one T6SS gene cluster in EHEC strains of Sakai and EDL933 . Although the number of T6SS gene cluster is identical in UPEC and EHEC , the sequence and organization of their T6SS genes are distinct , suggesting the origin of T6SS is highly complicated . T6SS has been extensively explored in bacterial antagonism in various environments . T6SS-dependent bactericidal activity was fulfilled by injecting an array of toxins to hydrolyze the prey cell wall . In this study , we found that both the wild type EHEC strain and the deletion mutant of hns ( a T6SS activating strain ) failed to kill T6SS+ or T6SS- bacteria ( Fig 3D and 3E ) , suggesting that T6SS in EHEC does not have antibacterial activity in vitro . At this stage , we can not rule out the possibility that EHEC may have bactericidal activity in vivo . EHEC is a zoonotic pathogen that thrives in the rumen of cattle [56] , we then used bile salt supplement to mimic the human gastrointestinal tract environment , and found that the transcription levels of T6SS genes were dramatically induced by bile salt supplement ( S15A Fig ) . Moreover , we found that T6SS contributed to the resistance of EHEC to bile salt ( S15B Fig ) . Considering the expression level difference of T6SS between in vitro and in vivo ( Fig 6C ) , one might postulate that with an enhanced T6SS expression , EHEC may replicate and target host microbiota in a T6SS-mediated manner for better survival and competition with other bacteria and protozoans in the cattle rumen . Since T6SS was initially identified as a virulence factor in E . tarda in 2004 [57] , it has been associated with bacterial virulence-related phenotypes , including adhesion , invasion , cytoskeletal alteration , intracellular survival , cytotoxicity , and host response , suggesting T6SS is indeed crucial for pathogen virulence [15–20] . As a protein secretion machine , T6SS in pathogenic bacteria is supposed to secrete an array of effectors to mediate bacterial interaction with eukaryotic cells . The involvement of EHEC T6SS in virulence was proved in this study ( Figs 2E and 7A ) . Although Shiga toxin is considered to be the major virulence factor of EHEC [56] , we found that deletion of T6SS did not disturb the expression and secretion of Shiga toxin ( S16 Fig ) , indicating that T6SS contributed to the pathogenesis of EHEC independent of Shiga toxin activity . The T6SS-dependent effectors remain elusive . There are few virulence-related effectors identified in the past decade except canonical Hcp and Vgr family effectors , which were discovered more than ten years ago [18] . Up to now , only four non-canonical T6SS effectors were reported , including EvpP from E . tarda [58] , VasX from V . cholerae [59 , 60] , PldB from P . aeruginosa [61] , and TecA in B . cenocepacia [62] . In this study , we identified a catalase KatN as a novel T6SS effector in EHEC , and this effector could decrease the ROS level of host cell to facilitate EHEC survival in phagocytic cells . EHEC possesses multiple catalases and peroxidases for defense against oxidative stress , including KatE , KatG , KatP , and alkyl hydroperoxide reductase AhpC [8 , 9 , 11] . AhpC and KatG are regulated by both OxyR and RpoS [46 , 63 , 64] . KatE is only induced by RpoS during the switch from exponential to stationary growth [47] . We found that KatN is also regulated by both OxyR and RpoS ( Fig 5B ) , suggesting a complicated regulatory network in response to oxidative stress in EHEC . Although EHEC has several catalases , it seems these catalases are not redundant and play protective roles under different circumstances . Both KatG and KatE scavenge hydrogen peroxide in E . coli , and KatG is the major protective enzyme when AhpC is saturated by high level of hydrogen peroxide [65] . Our previous RNA-Seq data showed that RPKMs of ahpC and katG are 3471 . 8 and 618 . 9 , respectively , which are much higher than the average RPKM of whole genome [26] . Considering the RPKM of katN is 4 . 0 under in vitro condition , and the transcriptional level of katN in macrophage cells was increased by more than 10 fold ( S13 Fig ) , we postulate katN is not critical for EHEC response to oxidative stress in vitro but rather plays an important role in the interaction between EHEC and the host phagocytic cells . The survival defect of katN deletion strain in macrophage clearly supported this hypothesis . The similar phenotype of ΔkatN as wild type strain in mouse model suggests other catalase ( s ) may play redundant roles in vivo . We analyzed the distribution of katN in sequenced bacterial genomes and found that most pathogenic E . coli strains contain katN , and non-pathogenic E . coli strains do not possess katN . Moreover , katN and z1922 , z1923 and z1924 consist of an operon in the genome of EHEC strain EDL933 , which is highly homologous to the yciGFEkatN operon in S . Typhimurium [41] . When we extended the searching range from E . coli to other species , the results showed that most katN genes were linked to the operon yciGFE and located on a cryptic prophage ( CP-933X ) , suggesting katN was acquired via horizontal gene transfer in EHEC . Further analysis showed that about 130 sequenced bacterial genomes contain katN , while most of them ( 100 of 130 ) were distributed in the T6SS containing bacterial species ( S17 Fig ) , suggesting a co-existence of T6SS and katN in the pathogenic bacterial species . We further analyzed the distribution of katE , katG , ahpC and katP , and found that katE , katG and ahpC were highly conserved and wildly distributed in the bacterial genomes , while katP was only present in 47 species , and the distributions of the above genes were not correlated with these of T6SS clusters . It has been reported that some bacteria species in human gut can induce rapid , physiological generation of ROS to regulate host immune function , intracellular signalling , and cytoskeletal dynamics [66–70] . For example , commensal bacteria , Lactobacillus can stimulate NADPH oxidase I to promote ROS generation in the gut [71] . On the other hand , intestinal pathogens must overcome the intestinal ROS barrier to colonize in the gastrointestinal tract to cause infection and disease . For instance , OxyR and two catalases are critical for scavenging environmental ROS to facilitate V . cholerae growth and zebrafish intestinal colonization [72] . The present work showed that EHEC could secrete catalase KatN in both cell contact-independent and contact-dependent manners . Therefore , we speculate that once entered into the intestine , EHEC might use T6SS to secrete catalase KatN to hydrolyze ROS around bacterial cell to form a low ROS level niche and facilitate EHEC population growth , further infection and development of disease . The bacterial strains and plasmids used in this work were listed in S3 Table . E . coli O157:H7 strain EDL933 was used as the wild type EHEC strain . All mutants derived from strain EDL933 were constructed by λ Red recombinase system . E . coli BL21 ( DE3 ) was used to overexpress KatN . E . coli DH5α was used for the cloning procedures . Mouse macrophage-like RAW264 . 7 cells obtained from Cell Resource Center of Shanghai Academy of Sciences , Chinese Academy of Sciences were used in this study . Bacteria were cultured routinely at 37°C in Luria Bertani ( LB ) medium unless noted otherwise . Cell lines were cultured in DMEM ( 10% FBS ) at 37°C under 5% of CO2 . Antibiotics were used at the following concentrations: ampicillin 50 μg/ml; gentamicin 100 μg/ml; chloramphenicol , 30 μg/ml; kanamycin , 50 μg/ml; streptomycin 200 mg/ml; tetracycline , 10 μg/ml . The core of T6SS is composed of 13 proteins . Genes encoding these proteins were chosen as baits in sequential BLASTN , BLASTX and BLASTP to identify homologues in the genome of EHEC strain EDL933 ( e-value less than 10−5 ) . The genomic DNA of E . coli was extracted using Easy-DNA kit ( Invitrogen , Carlsbad , CA ) . The genes mentioned in this manuscript were amplified with the corresponding primers listed in S4 Table using the genomic DNA of EHEC strain EDL933 as template . The PCR products were digested with the corresponding restriction enzymes and then cloned into the appropriate vectors predigested by same restriction enzymes . The deletion mutant strains were constructed by the method described by Datsenko and Wanner [30] . Briefly , the EHEC strain EDL933 was transformed with the pKD46 , which contains genes coding for the arabinose-induced λ Red recombinase system that promotes recombination between linear pieces of DNA ( PCR products ) and the host chromosome . The recombination is based on short stretches of homology ( 50 nucleotides ) on the linear DNA to the site of recombination . The PCR products for knocking out the target genes were amplified , gel extracted , and electroporated into competent strain EDL933 containing pKD46 prepared with the presence of arabinose . The deletion mutants were screened by the antibiotics and were verified by PCR using primers adjacent to the gene region and followed by sequencing . The secreted proteins were isolated by a previously method with some modifications [19] . The overnight inoculums of strain EDL933 and its derived mutants were diluted 1:100 into 1 L M9 minimal medium supplemented with 44 mM NaHCO3 , 8 mM MgSO4 , 0 . 4% glucose , and 0 . 1% Casamino Acids for further growth to an OD600 of 0 . 8 , at 37°C . The supernatant was collected by centrifugation twice at 20 , 000 g for 15 min and then filtered through a 0 . 2 μm filter . The filtered supernatant was condensed 100 fold via Vivaflow 50 system , according to the manufacturer’s instruments ( Sartorius ) . The condensed supernatant was centrifuged at 25 , 000 g for 15 min to remove insoluble salts . The cleaned supernatant was further condensed by Amicon Ultra system at a cutoff of 3 kDa ( Millipore ) . The resulting sample with a volume of about 100 μl were mixed with 30 μl of 5× SDS-PAGE Sample buffer and boiled for 5 min . The samples were applied to SDS-PAGE , Western blot or LC-MS/MS analysis . Western blot was carried out according to standard procedures . Briefly , proteins were resolved on 12% SDS-PAGE , transferred to PVDF membranes . 50 mM Tris-HCl ( pH 7 . 5 ) with 150 mM NaCl , 0 . 5% ( V/V ) Tween-20 ( TBST ) and 5% skim milk was used for blocking the PVDF membranes and diluting the antibodies . After blocking , membranes were probed with the anti-KatN ( rabbit-derived ) antibody or anti-RpoA ( mouse-derived ) antibody or anti-His tag ( mouse-derived ) antibody over night at 4°C . Membranes were washed with 1× TBST 3 times for 10 min each , and probed with appropriate secondary antibody . The β-Lactamase assay was performed as previously described [19] , with some modifications . Briefly , EHEC strain EDL933 and its derived mutants bearing pCX340 or pCX-katN were grown overnight in LB medium with 10 μg/ml tetracycline and 0 . 25 mM IPTG at 37°C . The cultures were centrifuged at 12 , 000 g for 15 min , and the supernatants were collected . Five microliters of nitrocefin ( Calbiochem ) stock solution ( 1 mM ) was added to 95 μl supernatant of each sample , then the mixture was incubated at room temperature for up to 15 min to allow red color to develop . Spectrophotometric assays for β-lactamase were carried out by measuring changes in absorbance at 486 nm . For purification of KatN , pQE80YX1-katN was constructed and verified by sequencing . The plasmid was introduced into E . coli strain BL21 ( DE3 ) , and the cells were grown in LB medium containing 100 μg/ml ampicillin at 37°C with shaking . IPTG was added to a final concentration of 0 . 1 mM at the time point when the absorbance of the culture at OD600 reached 0 . 6–0 . 8 . The culture was continuously incubated for another 3–4 h . The cells were harvested by centrifugation and disrupted by ultrasonication ( Sonics , USA ) . The supernatant was collected by centrifugation at 20 , 000 x g for 40 min at 4°C , and supplemented with imidazole at a final concentration of 10 mM . KatN was purified by the ÄKTA system ( GE Health , USA ) . His-tag resin beans were washed with 10× volume binding buffer A ( 20 mM sodium phosphate , 500 mM NaCl and 20 mM imidazole , pH 7 . 5 ) , recombinant protein was eluted using 10× volume of buffer B ( 20 mM sodium phosphate , 500 mM NaCl and 500 mM imidazole , pH 7 . 5 ) . The protein was checked by SDS-PAGE before shock-frozen storage at -80°C . KatN concentration was determined by the Bradford protein assay ( Bio-Rad , USA ) . Bovine serum albumin ( BSA ) was used as the protein concentration standard . The KatN protein tagged with hexa-histidine ( 6× His ) was overexpressed and purified by nickel affinity chromatography . The purified KatN was used to immunize rabbits for the production of antibodies . Catalase activity was measured by using the Catalase Assay Kit ( Beyotime ) according to the manufacturer’s instruction . The bovine liver catalase ( Sigma Aldrich , USA ) was used as a positive control . The overnight cultures of strain EDL933 and its derived mutants were diluted and transferred into fresh LB medium to reach OD600 value of 1 . 0 . To test the resistance to H2O2 , bacteria cultures were diluted 100 times in 5 ml LB broth supplemented with 0 mM , 1 mM , 2 mM , 4 mM , 6 mM or 100 mM H2O2 . After incubation at 37°C , 220 rpm for 7 h , the resistance to H2O2 was measured by determining the values of OD600 . Briefly , the total RNA was isolated by RiboPure-Bacteria kit ( Ambion , USA ) , and the concentration was determined by measuring the A260 . Three microgram of total RNA was reverse-transcribed into cDNA by using Super Script III First-Strand Synthesis System for RT-PCR ( Invitrogen , USA ) . Primers for qPCR were listed in S4 Table . Samples were run in triplicates and amplified using SYBR Premix Ex Taq II ( TAKARA ) in the 7500 fast Real-Time PCR System . The relative transcriptional level was determined by the methods of 2−ΔΔCt [73] . 16S rRNA was used as a reference gene . Bacterial competition assays were performed according to the method previously described , with minor modifications [33] . The gentamicin-resistant plasmid was transformed into strain EDL933 or P . aeruginosa PAO1 for bacterial competition assays; A . baylyi ADP1 was spontaneous streptomycin resistant mutant , E . coli K12 MG1655 was kanamycin resistant derivative mutant . Overnight cultures of bacteria ( EHEC EDL933 , P . aeruginosa PAO1 or A . baylyi ADP1 ) were washed by LB and diluted 100 times into fresh LB broth and cultivated to OD600 ~ 0 . 8–1 . 0 . Cells were washed by 1× PBS and enriched to OD600 ~ 10 . Then , cells were mixed in 20:1 ratio , and 5 μl of the mixture was spotted on LB agar plate . After incubated at 37°C for 2 . 5 h , bacterial spots were cut out and the cells were resuspended in 1 ml 1× PBS . The suspensions were diluted serially in 1× PBS , and 5 μl of the suspensions was spotted on selective LB agar plates ( gentamicin for strain EDL933 and P . aeruginosa PAO1 , streptomycin for A . baylyi ADP1 ) , followed by 16 h incubation at 30°C . Antibiotic concentrations were gentamicin , 15 μg/ml; streptomycin , 100 μg/ml; kanamycin , 10 μg/ml . At least three biological replicates were performed . The RAW264 . 7 cells were maintained in DMEM supplemented with 10% fetal bovine serum ( FBS ) for survival experiment [73 , 74] . The RAW264 . 7 cells were seeded at 2×105 cells per well in a 24-well tissue culture plate and cultured at 37°C under 5% of CO2 overnight . Then , the RAW264 . 7 cells were infected with 5×106 cells of EHEC strain EDL933 and its derived mutants ( MOI = 10 ) . The plate was centrifuged briefly ( 400 g , 5 min ) to synchronize the infection and incubated for 30 min ( 0 h ) at 37°C under 5% of CO2 . The cells were washed three times with PBS and fresh DMEM-10% FBS containing 100 μg/ml gentamicin was added to kill extracellular bacteria . After incubation for 2 h at 37°C under 5% of CO2 , the cells were washed with PBS three times and lysed in 0 . 025% SDS , and then diluted with PBS for CFU counting on LB agar plates . To assess intracellular growth , the DMEM-10% FBS containing 100 μg/ml gentamicin was replaced with DMEM-10% FBS containing 15 μg/ml gentamicin and parallel cell cultures were analyzed for viable bacteria after 24 h incubation at 37°C under 5% of CO2 . The transcription of T6SS genes of EHEC in RAW264 . 7 cells was analyzed by qPCR . In brief , 2×105 RAW 264 . 7 cells were infected with 1×107 EHEC strain EDL933 ( MOI = 50 ) at 37°C under 5% of CO2 . After incubation for 1 h , excess bacteria were washed off with PBS , and the infected cells were incubated with DMEM ( 10% FBS ) at 37°C under 5% of CO2 for 8 h . The EHEC strain EDL933 was incubated individually with DMEM ( 10% FBS ) at 37°C under 5% of CO2 for 8 h as an uninfected control . After 8 h incubation , the supernatants were removed , and the infected cells were washed three times with PBS . Bacteria were released from infected macrophages by treatment with 1% Triton X-100 for 10 min . The total bacterial RNA was isolated , and the mRNA levels of z0254 , z0264 , z0266 , z0267 , katG , katE , katP , ahpC and katN were quantified by qPCR . Thioglycolate-elicited primary peritoneal macrophages ( Peritoneal MΦ ) were harvested as described before [74] . Briefly , BALB/c mice were intraperitoneally injected with 4% Brewer’s thioglycolate medium ( Sigma ) . After 3 days , mice were sacrificed by cervical dislocation , and cells were isolated by flushing the peritoneal cavity with 50 ml PBS per mouse . Cells were seeded in 24-well plates , and non-adherent cells were removed by washing with DMEM . The adherent peritoneal macrophages were used for subsequent experiments . The animal procedures were approved by Shanghai Jiao Tong University School of Medicine , and this study was carried out in strict accordance with the National Research Council Guide for Care and Use of Laboratory Animals [SYXK ( Shanghai 2007–0025 ) ] . All surgery was performed under sodium pentobarbital anesthesia , and all efforts were made to minimize suffering . Five-week-old BALB/c mice were used for all animal infection experiments . Two days before infection , mice ( 8/group ) were gavaged with 200 μl PBS ( 200 mg/ml streptomycin ) to disrupt the microbiota in intestinal track . Mice were not fed 8 h before intragastric inoculation . EHEC strain EDL933 and its derived mutants were grown to exponential phase ( OD600 = 1 ) followed by washing and resuspension at a concentration of 1×1010 CFU in 250 μl of PBS . The body weight and survival of mice were recorded every day from day 0 to day 15 after administration . The extent of bacterial colonization was monitored daily for 8 days by quantitation of the strain EDL933 and its derived mutants shed into fecal pellets . For collecting feces , mice of every group were placed into clean , empty cages , allowed to defecate , and the feces were collected and weighed . The fecal material was diluted 1:10 by weight into sterile PBS then homogenized by vortex . Large debris was pelleted by brief centrifugation at 1 , 750 g , and the supernatants were diluted and plated onto Sorbitol MacConkey Agar ( SMAC ) plates to determine CFU/g feces . The immune-fluorescence assays were performed as previously described with minor modifications [38] . RAW264 . 7 cells were grown in a 24-well tissue culture plate on glass cover-slips at 37°C under 5% of CO2 . Strain EDL933 and its derived mutants were cultured overnight at 37°C in LB broth . Then , RAW264 . 7 cells were infected by EDL933 and its mutants ( MOI = 10 ) . After incubation for 40 min , the cells were fixed with 4% para-formaldehyde ( PFA ) in PBS for 30 min at room temperature and then washed three times with PBS . The cells were blocked with 3% BSA-PBS for 30 min at 37°C . The antibodies were diluted in 3% BSA-PBS at the indicated dilutions: mouse-anti-O157 antibody ( 1:200 ) ; rabbit-anti-KatN antibody ( 1:6000 ) ; goat-anti- mouse FITC conjugate ( 1:300 ) ; goat-anti-rabbit rhodanmine red-X conjugate ( 1:300 ) . The cover-slips were incubated with anti-O157 and anti-KatN antibodies simultaneously overnight at 4°C . After incubation the cover-slips were washed three times with PBS and were incubated with goat-anti-mouse FITC conjugated and goat-anti-rabbit rhodanmine red-X conjugated antibodies away from light at 37°C for 1 h . Cell nuclei were counterstained with 2 . 5 μg/ml nuclear stain DAPI for 10 min . The cover-slips were washed three times with PBS and then were mounted on glass slides and sealed with Entellan ( Merck ) . Samples were analyzed using a laser-scanning confocal microscope ( Leica TCS-NT ) as described previously [13] . The Hyper-3 probe-based reactive oxygen species test were performed as previously described with some modifications [50] . HyPer-3-transfected RAW264 . 7 cells were plated on glass coverslips ( Fisher ) overnight at a cell density of 3×105/ml . Cells were washed three times with PBS and then infected with strain EDL933 and its derived mutants ( MOI = 10 ) in DMEM without Phenol Red and FBS at 37°C under 5% of CO2 . The uninfected cells were used as a negative control . After incubation for 50 min , the DMEM was removed and cells were washed three times with PBS to remove the extracellular bacteria . Cells were analyzed by a laser-scanning confocal microscope ( Leica TCS-NT ) at excitation wavelengths of 405 and 488 nm . For every sample , the average 488 nm/405 nm ratios of at least ten cells were analyzed . The intracellular ROS were estimated using a fluorescent probe DCFH-DA by flow cytometry [51] . Briefly , RAW264 . 7 cells were grown overnight in a 24-well tissue culture plate at a cell density of 3×105/ml . Cells were washed three times with PBS and then infected with mutant strains ( MOI = 10 ) in DMEM without Phenol Red and FBS at 37°C under 5% of CO2 . The uninfected cells were used as a negative control . After incubation for 50 min , the DMEM was removed and cells were washed three times with PBS . DCFH-DA ( 900 μl , diluted in DMEM without Phenol Red and FBS at a final concentration of 10 μM ) was added separately to wells and cells were incubated at 37°C under 5% of CO2 for 45 min . After removing DCFH-DA solution , cells were washed three times with PBS and were suspended in 200 μl PBS with propidium iodide ( PI ) ( 5 μg/ml ) . PI was used as a counter stain dye for DCFH . Cells were analyzed by the flow cytometer FACScan ( Becton Dickinson ) . The mean fluorescence of 10000 PI-negative cells was calculated using the flow cytometer software FlowJo , version 6 . 4 . 2 ( FlowJo , USA ) . The translocation assay was performed as previously described with some modifications [42] . RAW264 . 7 cells were seeded at 3×104 cells/ml on glass coverslips ( Fisher ) overnight in minimal essential medium ( MEM ) with 10% fetal bovine serum ( FBS ) . EHEC strain EDL933 ( WT ) and ΔT6SS bearing pCX340 or pCX-katN were grown overnight in LB broth with 10 μg/ml tetracycline at 37°C , subcultured 1:100 in LB broth with 10 μg/ml tetracycline and 0 . 8 mM IPTG , followed by growing at 37°C for 2 h . RAW264 . 7 cells were washed three times with Hanks’ balanced salt solution ( HBSS ) and infected with 3×106 bacteria ( MOI = 100 ) in MEM medium containing 1% FBS , 10 μg/ml tetracycline and 0 . 8 mM IPTG for 3 h . Cells were washed three times with HBSS and then incubated with 2 μM CCF2-AM ( Invitrogen ) for 60 min . Cells were washed three times with HBSS to remove CCF2-AM and then analyzed by a laser-scanning confocal microscope ( Leica TCS-NT ) at excitation wavelengths of 409 and 488 nm ( X20 magnification ) . EHEC survival in bovine bile salts assay was performed as previously described with modifications [56] . Cells from a single clone were grown overnight in LB broth at 37°C , subcultured 1:100 in 6 ml LB broth . Cells were grown to an OD600 of 0 . 3 , at which time cultures were supplemented with 0 mg/ml , 25 mg/ml or 50 mg/ml bovine bile salts ( Sigma Aldrich ) and incubated at 37°C . At 0 h , 3 h , and 6 h post-bile salt treatment , samples were collected for OD600 determination and total RNA isolation . Descriptive statistics and statistical comparison were performed resourcing to the GraphPad Prism , version 5 . 00 ( GraphPad Software , San Diego , CA ) . Statistical comparison of mean values between two specific groups was carried out using the Student’s t-test , and ANOVA analysis was applied to compare mean values of more than two groups . * , P<0 . 05; , P<0 . 01; * , P<0 . 001; ns , Not significant .	The type VI secretion system ( T6SS ) is a specific macromolecular protein export apparatus , and widely distributed in Gram-negative bacteria . Generally , T6SS has been shown to play an important role in anti-bacterial competition and virulence to eukaryotic hosts . Enterohemorrhagic Escherichia coli ( EHEC ) can cause severe foodborne disease , including abdominal cramps and diarrhea that may progress to bloody diarrhea and hemolytic uremic syndrome . In the current study , we show that the T6SS of EHEC is involved in its intracellular survival and virulence in mice . Specifically , the novel effector KatN , a Mn-catalase identified in this work updates the general role of the T6SS in the pathogenesis of EHEC , in the context of oxidative stress in host cytoplasm . Combined with the biased distribution of katN in the T6SS-containing bacteria , our data suggest that KatN , as a new T6SS effector , is a key virulence factor in the pathogenesis of T6SS+ bacteria .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "bacteriology", "blood", "cells", "medicine", "and", "health", "sciences", "immune", "cells", "pathology", "and", "laboratory", "medicine", "intracellular", "pathogens", "enzymes", "pathogens", "immunology", "enzymology", "microbiology", "physiological", "processes", "bac...	2017	Type VI secretion system contributes to Enterohemorrhagic Escherichia coli virulence by secreting catalase against host reactive oxygen species (ROS)
HIV-1 is present in anatomical compartments and bodily fluids . Most transmissions occur through sexual acts , making virus in semen the proximal source in male donors . We find three distinct relationships in comparing viral RNA populations between blood and semen in men with chronic HIV-1 infection , and we propose that the viral populations in semen arise by multiple mechanisms including: direct import of virus , oligoclonal amplification within the seminal tract , or compartmentalization . In addition , we find significant enrichment of six out of nineteen cytokines and chemokines in semen of both HIV-infected and uninfected men , and another seven further enriched in infected individuals . The enrichment of cytokines involved in innate immunity in the seminal tract , complemented with chemokines in infected men , creates an environment conducive to T cell activation and viral replication . These studies define different relationships between virus in blood and semen that can significantly alter the composition of the viral population at the source that is most proximal to the transmitted virus . Sexual transmission of the human immunodeficiency virus type 1 ( HIV-1 ) is the most common mode of transmission worldwide . During sexual transmission , genital secretions are the most proximal source of the transmitted virus . Thus , an understanding of the virus at these sites is central to understanding the transmission event and the nature of the transmitted virus . In this study we have explored the nature of viral populations in seminal plasma . Virus enters the male genital tract during primary infection [1]–[5] . Initially , the virus found in the semen is similar , if not identical , to that found in the blood [6] , [7] . During primary infection the viral RNA load is elevated in both the blood and the semen [1] , [3] . The probability of transmission is related to the level of virus in the blood of the donor [8]–[11] and , based on a small cohort , to the level of virus in the semen [12] . Factors that induce inflammation in the seminal tract , such as sexually transmitted infections ( STI ) , can raise the level of virus in semen [13] , and this may contribute to the transmission of HIV-1 by the sexual route [14] . In addition , the endogenous semen-derived enhancer of virus infection ( SEVI ) , a fragment of prostatic acid phosphatase , has been shown to increase infectious viral titers in vitro by several orders of magnitude [15] . The presence of virus in semen raises the possibility that virus found in semen could be the product of replication within the seminal tract . CD4+ T cells are found in semen indicating the presence of target cells that could support replication [16] , [17] . SIV-infected macaques have infected cells within the tissues of the seminal tract [18] , [19] , supporting the possibility for local viral replication . Several studies have examined the relationship between viral populations found in blood and semen and noted differences ( i . e . compartmentalization ) using discordant drug resistance markers [20]–[24] , differences in population markers [25] , [26] , or phylogenetic analysis [27]–[32] . In this study we have carried out a detailed examination of the viral populations in semen , comparing the env gene in blood plasma and seminal plasma . The men were therapy-naive and chronically infected with subtype C ( n = 12 ) or subtype B ( n = 4 ) HIV-1 . We found a varied and complex relationship between these two compartments which suggests multiple types of biological phenomena . There is evidence for the direct import of virus from the blood to the semen , evidence for clonal amplification of a subset of genotypes within the seminal tract , and evidence for sustained replication and distinct evolution of virus within the seminal tract resulting in compartmentalization . The latter two of these phenomena result in seminal plasma viral populations that are distinct from those found in the blood , and thus distinct at the site proximal to transmission . Furthermore , semen is enriched in cytokines which may increase the potential for independent viral replication within the seminal tract . In this study , we examined viral populations and cytokine/chemokine relationships in paired blood and semen samples collected from men with chronic HIV-1 infection in Lilongwe , Malawi ( n = 12 ) [33] , or from the CHAVI 001 clinical study ( n = 4 ) . We utilized a cohort of men without urethritis to minimize any potential confounders on viral loads , viral populations , or cytokine profiles . The clinical parameters of their HIV-1 infection are shown in Table 1 . There was no evidence of urethritis in the dermatology clinic subjects , although the diagnosed cases of syphilis ( 1 ) and trichomonas ( 4 ) were treated with appropriate antibiotics ( Table 1 ) . In addition , blood and semen samples were obtained from twelve HIV-1-negative men without STIs from North Carolina and from 6 men from Malawi to serve as a control for the cytokine and chemokines analyses . We used viral RNA extracted from blood plasma and seminal plasma to generate cDNAs to use as a template in the single genome amplification ( SGA ) protocol of the viral env gene [34]–[36] . The use of viral RNA allows for the examination of contemporaneously replicating virus . A mean of 27 amplicons were analyzed per sample , with a range of 15 to 34 in the blood and 14 to 44 amplicons in the seminal plasma . Sampling this number of genomes provides a 95% chance to detect a subpopulation in the range of 10–15% , and provides reasonable power to estimate the relative proportions of the major variants in the population [35] . The sequence of the entire env gene was determined for each amplicon . The use of SGA precludes PCR recombination as a source of confusion about the relationship between the viral genomes present in each sample [36] . Figure 1 depicts the phylogenetic analysis of viral sequences in the blood and semen for subjects C011 and C111 . There was a diverse population in the blood and this diversity was fully represented in the semen . Furthermore , the complexity of the sequences in the blood , where no two sequences were identical , was also represented in the semen . We conclude that in these subjects there was no compartmentalization in the seminal tract . If there were local replication of virus in the seminal tract of these subjects it must have represented the full complexity of the virus in the blood . Alternatively , this virus did not replicate in the seminal tract but rather was imported from the blood . In either scenario , the viral populations in the blood and semen were essentially identical , representing well-equilibrated populations . A different phylogenetic pattern was detected in subjects C007 , C009 , C012 , C019 , C070 , and C109 ( Fig . 2 and Fig . S1 ) . Similar to the subjects where blood and semen populations were well-equilibrated , these subjects had viral populations in the semen that represented the full diversity of the virus in the blood . In addition , the blood populations were highly complex and consistent with a diverse viral population , with no sampling of identical blood sequences with the exception of patient C109 . However , there was an additional feature of the viral populations in the semen of these subjects that distinguished them from the virus in the blood . In these subjects sampling of the viral population in semen resulted in examples where identical or nearly identical sequences were observed ( Fig . 2 , Fig . S1 , Fig . S3 , and Fig . S4-S15 ) . Patient C109 had a clade of identical/nearly identical sequences that comprised nearly 75% of the entire semen viral population . Similarly , patient C009 had three duplicated viral variants that each comprised <10% of the semen population , indicating a broad range in the amount of sequence duplication that can exist within semen . We term this phenomenon clonal amplification , and because of the nature of the SGA strategy , this cannot be the result of PCR resampling since each amplicon was generated from a single template . In four of the 12 subjects with subtype C HIV-1 ( C047 , C083 , C018 , and C113 ) , a third relationship between the viral populations in the blood and semen was seen . For these subjects there was a deep branch point with high bootstrap support in the phylogenetic tree separating sequences found in the blood from sequences found in the semen ( Fig . 3 and Supplemental Fig . S2 ) . In addition to visual inspection of the phylogenetic trees to identify semen clades with long branch lengths with high bootstrap support , the presence of compartmentalized sequences was confirmed with the Slatkin-Maddison statistical [37] and correlation coefficient tests [38] available through Hypothesis testing through Phylogenies ( HyPhy ) [39] . Previous analyses have revealed that there is no gold standard from the variety of statistical measures available for detecting compartmentalization; therefore , multiple tests are recommended to determine the existence of compartmentalization [40] . Compartmentalization tests were performed with all viral sequences , and after removal of duplicated sequences since amplified variants in the semen can increase both the frequency of compartmentalization calls and the statistical support for those calls ( Supplemental Table S2 ) . Thus , compartmentalization of these viral populations was observed in subjects C047 , C083 , C018 , and C113 and indicates an autonomously replicating subpopulation in the seminal tract that followed a distinct evolutionary pathway . As a result of this compartmentalized subpopulation , the virus in the semen was genetically distinct from the virus in the blood . Two subjects ( C083 and C018 ) had compartmentalization of semen-derived sequences without clonal amplification ( Fig . 3 and Fig . S2 ) . In addition , two subjects ( C047 and C113 ) had both clonal amplification and compartmentalization of semen-derived sequences ( Fig . 3 , Fig . S2 , Fig . S8 , and Fig . S11 ) . Similar to the previous subjects with semen clonal amplification ( with the exception of C109 as previously mentioned ) , there were no duplicated blood sequences . Thus , these data indicate that the male genital tract is capable of supporting complex viral populations , and that compartmentalization and amplification can occur independently . In addition to the 12 men with HIV-1 subtype C infection , we analyzed blood and semen plasma viral RNA populations from 4 men with subtype B infection ( Fig . 4 , Fig . S3 , Fig . S12-15 ) . Each of the four had identical sequences ( clonal amplification ) in the seminal plasma that ranged from <10% to one-third of the semen viral population . In contrast , none of the patients had identical sequences in the blood plasma . In addition , three of the men had equilibrated blood and seminal plasma sequences: 700010333 , 700010501 , and 700011145; whereas , one of the 4 subtype B infected men ( 700010380 ) had significant compartmentalization of semen-derived sequences . Thus , clonal amplification and compartmentalization within the seminal plasma is a common feature of HIV-1 of different subtypes . We carried out an analysis for each subject using Bayesian Evolutionary Analysis by Sampling Trees ( BEAST ) [41] to estimate the time to most recent common ancestor ( TMRCA ) of the amplified variants , and/or the TMRCA of compartmentalized variants using maximum likelihood trees . Of note , the topologies of the neighbor-joining and maximum likelihood trees were very similar ( data not shown ) , indicating that these two different phylogenetic methods produced concordant results in their evolutionary models . As a control to compare the BEAST estimates to known values obtained from previously published sequence data sets , a separate analysis was performed using a subset of published longitudinal env sequences ( Fig . S16 ) [42] . From this data set , we calculated the time of divergence using C2-V5 env sequences obtained from longitudinal plasma samples at 3 , 29 , 42 , 58 , 70 , and 100 months post-seroconversion; BEAST estimates of 10 , 34 , 49 , 147 , 144 , and 204 months , respectively , were observed with a high coefficient of determination ( R2 = 0 . 9155 ) . Thus , in the setting of chronic HIV-1 infection , the observed BEAST estimates were similar to the expected values for periods up to several years , but there is a trend to overestimate time periods greater than four years by approximately two-fold . Next , we determined the TMRCA of amplified and/or compartmentalized variants within seminal plasma . The TMRCA for the oligoclonal amplifications within the seminal compartment ranged from 1 to 375 days , with a mean of 57 days , indicating recent divergence . In contrast to the short evolutionary times observed with the semen variants displaying oligoclonal amplification , the subjects with significant semen compartmentalization had divergence estimates from 1 . 5 to 9 . 7 years , with a mean of 5 . 2 years . If clonally-amplified sequences were used only once , there was negligible effect on the TMRCA of the entire tree , or the TMRCA of amplified or compartmentalized variants ( data not shown ) . Thus , the TMRCA of the clonally amplified variants tends to be relatively short in contrast to compartmentalized variants , which represent more distant divergence . However , we do not know if the rate of evolution in the semen is comparable to the blood adding additional uncertainty to the accuracy of the absolute values generated with the BEAST analysis . To determine if populations were evolving randomly under neutral evolution , a Tajima's neutrality test was performed using DnaSP [43] . Fifteen of the 16 patients showed no evidence of selection ( P values >0 . 10 ) ; however , C019 had a Tajima's D of -2 . 1 ( P value <0 . 05 ) implying either population size expansion , or positive selection . Thus , a coalescent model of viral evolution as assumed by BEAST remains valid for the majority of patients . In the case of C019 , the violation of a coalescent model was most likely due to the blood compartment ( Tajima's D of -1 . 82 , P value <0 . 05 ) vs . the semen compartment ( Tajima's D of -1 . 60 , P value >0 . 05 ) . Taken together , these data suggest that BEAST is a robust tool to compare the TMRCA of amplified and compartmentalized variants for the majority of the patients that were analyzed . To determine if the seminal plasma has a distinct immunologic profile relative to blood plasma , we measured the levels of nineteen cytokines and chemokines in the paired blood and semen samples from 12 of the men with chronic HIV-1 subtype C infection . As a control , we measured cytokines and chemokines from paired blood and semen samples in 12 uninfected men from the US and 6 uninfected men from Malawi without STIs . There were two features of the patterns of cytokines and chemokines ( Fig . 5 ) that are noteworthy . First , a subset of cytokines and chemokines ( IL-5 , IL-7 , IL-8 , MIG , IP-10 , and MCP-1 ) were concentrated in the semen of uninfected men with median levels that were 5 to approximately 1000 fold greater than in the blood; none of the remaining cytokines or chemokines was as high as five-fold concentrated in the semen ( Fig . 5 ) . Second , for seven of the cytokines and chemokines ( IL-1b , IL-4 , IL6 , IL-7 , IL-8 , GM-CSF , and MCP-1 ) there was a significant increase in the semen:blood ratio of HIV-infected subjects compared to the uninfected subjects; conversely , MIG was significantly decreased in the infected subjects . Although our small sample size prevented a robust analysis , there were no cytokine correlates with amplification or compartmentalization of HIV-1 sequences in the semen . Moreover , there were no correlates with cytokine levels , and HIV-1 viral loads , amplification , compartmentalization , or the presence of asymptomatic STIs that were detected in five of the HIV-infected men ( data not shown , although the small sample size and the intersubject variability precludes an assessment beyond more general trends ) . The seminal compartment is the source of the transmitted virus in a majority of the transmission events for HIV-1 . Thus , an understanding of the biology of HIV-1 in the seminal tract is integral to understanding the biology of transmission , and a comparison of blood and seminal sequences is critical to increase our knowledge of viral dynamics . We have used viral sequence populations to examine the dynamic relationship between virus and host in the seminal tract , and identify multiple mechanisms by which HIV-1 populations exist in the male genital tract . A significant limitation of this study is that it is cross-sectional , involving a single time point . Another limitation of the current work is that there are no proviral sequences from semen cells to define the source of the amplified or compartmentalized variants . Previous work has identified paired blood and semen samples where the viral populations were discordant . In some cases this involved a comparison of viral RNA in blood plasma and seminal plasma , or a comparison of the sequences in viral DNA in blood cells and seminal cells [21]–[26] , [28] , [29] , [32] . While these studies clearly established the potential for the virus to become compartmentalized , in most cases there were two potential limitations intrinsic to the experimental approach: the possibility of recombination of viral sequences during PCR which would introduce artifacts into the phylogenetic analysis , and the analysis of a fairly small number of viral genomes in each population precluding a comparison of the population structure . As a result the phenomenon of compartmentalization has been described as a dichotomous state , i . e . the presence or absence of compartmentalized viral populations . However , in subjects where there is equilibration in the seminal tract over the entire range of complexity in the blood compartment , virus in the semen is most easily explained by the direct import of virus into the seminal plasma from blood , perhaps with no local replication of this population . The enrichment of cytokines and chemokines in the seminal tract ( Fig . 5 ) likely contributes to an environment that is supportive of HIV-1 replication . Our data , as well as others [44] , [45] , show that in the absence of HIV-1 infection several cytokines and chemokines are enriched , suggesting that the seminal tract maintains a constitutive state of innate immune activation . This state is exacerbated with HIV-1 infection where the concentration of a broader array of cytokines and chemokines indicates both innate and adaptive responses shaping the environment [46] . Thus , target CD4+ T cells and macrophages are likely to be in an activated state in this environment , enhancing their ability to support viral replication . In several subjects ( C109 and 701010380 ) there is some evidence for clonal amplification of sequences in the blood , with this being more pronounced in C109 . However , we do not know if the mechanism causing selective outgrowth in the blood is the same as that in the seminal tract , and in these subjects it is rare in the blood compared to the detection of clonal amplification in the semen in 12 of 16 men . The detection of clonal amplification within the seminal compartment raises several important questions . First , does amplification represent an initial stage of immunodeficiency ? We have detected an example of clonal amplification during primary infection ( data not shown ) suggesting clonal amplification can occur at any stage of infection . Given that clonal amplification was detected in equilibrated and compartmentalized populations , this also suggests that clonal amplification is not determined by the overall state of immunodeficiency . Second , what is the cellular source where this amplification occurs ? At one extreme clonally amplified sequences could be the product of a single cell . This seems unlikely since the seminal tract can support very complex populations in the compartmentalized state consistent with many available target cells , and some of the clonally amplified populations have some population structure suggesting they are the result of multiple rounds of replication ( supported by the longer BEAST estimates of TMRCA for some of these populations ) . The alternative is that clonal amplification occurs in a population of cells that are not infected by diverse viral genotypes . We suggest that either uninfected CD4+ T cells concentrate in specific sites , or are seeded by a single cell that then expands , until the focus of cells becomes infected with a single virus that spreads through this isolated population until the target cells are depleted . This would explain the self-limiting nature of the clonal amplification and explain how several clonal amplifications can occur concurrently . Finally , this process could be at work during compartmentalized virus replication , and thus account for the clonal amplification process also appearing during the replication of a complex compartmentalized population . A corollary of the isolation of the clonally amplified population is that the complex , compartmentalized population must be sustained by a distinct mechanism . There is likely continued import of virus from blood; however , the amount of locally replicating virus must obscure detection of this imported population . Based on these inferences we propose a model ( Fig . 6 ) to account for virus in semen . An assumption of this model is that viral populations within blood and semen are turning over similarly , and this is supported by a recent report in the literature showing similar decay kinetics of HIV-1 populations in blood and semen in men who initiate antiviral therapy [47] . In addition , our model is distinct from the semen being a viral reservoir , which is associated with reduced levels of viral replication [48] . We suggest that virus in the semen is derived from multiple sources . First , there is direct import of virus from the blood compartment , potentially without replication in the seminal tract , accounting for virus that is fully equilibrated between the blood and seminal tract compartments . Second , there is infiltration of individual infected CD4+ cells or virions into pockets of uninfected target cells that generate local foci of infection in the seminal tract , giving rise to clonal amplification of virus in this compartment . Third , ongoing local immune activation provides an environment that can support sustained , autonomous virus replication giving rise to compartmentalized virus . We estimate that this distinct population can replicate independently for a significant period of time , although lack of information about the rate of evolution in the compartment precludes a detailed analysis of the age of the population . An alternative interpretation of the appearance of compartmentalization is that there is delayed equilibration between blood and the seminal tract . In this circumstance a change in the population in the blood would not immediately be reflected in the semen , giving the transient appearance of compartmentalization . We do not favor this interpretation since the complexity of the virus in the semen can be quite high giving TMRCA values of months to years . However , the analysis of longitudinal samples in subjects displaying compartmentalization will be required to resolve this issue . An important unanswered question is the site within the seminal tract where virus undergoes independent replication . A relevant observation in this regard is that vasectomy does not preclude the presence of virus in semen [49] , [50] , suggesting that production of significant amounts of virus occurs outside of the testis , and implicating the seminal vesicles and prostate . Moreover , distal genitourinary sources other than the prostate have been implicated as the major source of seminal HIV-1 in men without urethritis or prostatitis [51] . In the setting of the blood compartment , disease progression is associated with higher levels of immune cell activation [52] . This may reflect an increasing trend to fail to control viral replication but with a continued response to the presence of viral antigen . We suggest a similar process may occur in the seminal tract and perhaps in other peripheral sites of viral replication . Recent literature reports the existence of clonal amplification of HIV-1 sequence in other compartments , including the CSF [53] , breast milk [54] , and cervicovaginal lavage fluid [55] . Thus , an influx of activated immune cells into areas where virus suppression is incomplete could lead to sustained viral replication and a distinct evolutionary pathway . In this regard , the presence of activated immune cell infiltrates that have been observed in the seminal tract of SIV-infected macaques [18] provides the likely sites where viral replication could occur in the male genital tract . A result of independent replication in the seminal tract , both clonal amplification and sustained replication , is to alter the composition of the viral population in the semen relative to that in the blood plasma . These differences can make blood a suboptimal surrogate for the seminal compartment in assessing the relationship of virus in the donor and recipient of a sexual transmission event . Several studies have noted differences between the transmitted virus and the virus in donor blood for subtypes A , C , and D [56]–[60] but not subtype B [35] , [58] , with differences in either glycosylation patterns , variable loop lengths , or susceptibility to neutralizing antibodies . It will be important to determine if the distinctive features of virus in semen play a role in transmission and/or in defining the nature of the transmitted virus . The patients infected with HIV-1 subtype C ( n = 12 ) were enrolled through the Kamuzu Central Hospital in Lilongwe , Malawi , between January and March , 1996 [33] . The protocol was approved by the University of North Carolina Committee on the Protection of Human Rights and the Malawi Health Sciences Research Committee . All study participants gave written informed consent and were offered a small payment for their participation . The original study design was a prospective , sequential comparison of two cohorts: HIV-1-infected men with urethritis who had urethral discharge on physical exam and at least five white blood cells per high-power field from a urethral swab , selected from the STI clinic , and HIV-infected men without urethritis on physical exam , selected from the dermatology clinic [33] . Blood and semen samples used in the current study were collected from men attending the dermatology clinic in Lilongwe , Malawi as described previously [33] . For both the STI and dermatology clinics , screening for gonorrhea , trichomonas , syphilis , and chlamydia was performed . In addition , blood and semen samples were obtained from participants with HIV-1 subtype B from the US ( n = 4 ) who were enrolled through the CHAVI 001 clinical core , a multi-center , prospective , observational cohort study of acute HIV-1 infection . IRB approval was awarded by each participating center as well as the Division of AIDS . All study participants gave written informed consent and were offered a small payment for their participation . None of the subtype B infected participants had urethritis on physical exam , and were negative for gonorrhea , Chlamydia , syphilis , and trichomonas infection . Consistent with established infection , all HIV-1 infected Malawi and US patients were confirmed EIA and Western Blot positive at study enrollment . Paired blood and seminal plasma samples from HIV-1 uninfected males for cytokine/chemokine analyses were obtained from the CHAVI 001 clinical core sites in the US ( n = 12 ) and Africa ( n = 6 ) . Cell-free blood plasma and seminal plasma were isolated and frozen as previously described [1] . HIV-1 viral loads from blood and seminal plasma from the Malawi men were determined by quantitative nucleic-acid sequence-based-analysis ( NASBA , Organon-Teknika ) [33] , and by Roche Amplicor vRNA or Abbott RealTime HIV-1 assays for the US men . Virus in the seminal plasma was pelleted by centrifugation prior to RNA isolation to remove the seminal plasma . The blood plasma or the resuspended virus pellet from the seminal plasma was extracted to isolate viral RNA using the QIAMP Viral RNA Mini Kit ( Qiagen ) . For each sample , approximately 10 , 000 viral RNA copies based on viral load were extracted and eluted . cDNA synthesis was performed using Superscript III Reverse Transcriptase ( Invitrogen ) with an oligo-d ( T ) primer as previously described [34]–[36] . To confirm that proviral DNA was not the source of SGA env-derived amplicons from cell-free viral RNA , RT-minus blood ( n = 11 ) and seminal ( n = 11 ) plasma samples were subjected to the SGA protocol; the remaining samples had insufficient volume remaining for the RT-minus control experiment . To preclude PCR recombination and Taq-induced errors , single genome amplification ( SGA ) of the env gene was performed using limiting dilution [34]–[36] , [61]–[63] . PCR amplicons were bidirectionally sequenced . To ensure that sequences arose from single DNA molecules , chromatograms with double peaks , indicating amplification from more than one cDNA template , were excluded . SGA-derived env amplicons with frameshift mutations that resulted in premature stop codons were also excluded . GenBank accession numbers are HM638460 to HM639260 . DNA sequence alignments were performed using clustal W [64] . Phylogenetic trees were generated using a neighbor-joining method ( MEGA 4 . 0 ) [65] . Pairwise DNA distances were computed using MEGA 4 . 0 . Highlighter plots were generated to visualize sequence differences ( www . hiv . lanl . gov ) . A Tajima's D test for neutrality was performed for each patient using DnaSP [43] . Compartmentalization of viral sequences was assessed by using the Slatkin-Maddison test [37] and correlation coefficient [38] available through HyPhy [39] . Gene flow was determined by the number of migration events compared between semen and blood after 10 , 000 permutations for the Slatkin-Maddison test . Compartmentalization was defined when P values <0 . 01 were obtained with the Slatkin-Maddison test using all sequences except the clonally-amplified sequences , of which only one was included , and when concordant results were obtained with the correlation coefficient test . More extreme P values were obtained when all of the clonally amplified sequences were included ( Supplemental Table S2 ) . Nineteen cytokines and chemokines were analyzed by luminex from paired blood and seminal plasma from 12 HIV-1 infected and uninfected subjects as previously described [66] . Concentrations of IL-1b , IL-2 , IL-4 , IL-5 , IL-6 , IL-7 , IL-8 , IL-10 , IL-12 ( p70 ) , IL-13 , IFN-g , TNF-a , and GM-CSF were measured using LINCOplex Luminex high-sensitivity 13-plex kits ( Millipore ) according to the manufacturer's instructions . Concentrations of MIP-1a , MIP-1b , RANTES , MCP-1 , MIG , and IP-10 were measured using custom standard sensitivity 6-plex kits ( Bio-Rad ) according to the manufacturer's instructions . Each sample was assayed in duplicate , and cytokine standards supplied by the manufacturer were run in parallel . Data were collected using the Bio-plex Suspension Array Reader ( Bio-Rad ) and a regression formula was used to calculate sample concentrations from standard curves . Values that were below the lower limit of detection were reported as the mid-point between the lower level of detection and zero . Semen:blood analyte ratios were calculated for each subject; however , data were excluded if both compartments were below the level of detection . Saturated values were reported as the upper limit of detection . Sensitivity values were adjusted for samples where the volume was limited and had to be diluted before the measurement . Statistical tests comparing analyte levels were non-parametric Mann-Whitney-U test for two groups ( compartmentalization vs . equilibration ) , and Kruskal-Wallis test for three groups ( high amplification , low amplification , no amplification ) . The non-parametric Mann-Whitney test was used to compare semen:blood cytokine ratios between HIV-1 infected and uninfected subjects . For each subject , blood and semen sequences were aligned using ClustalW 2 . 0 . 7 [67] . Markov Chain Monte Carlo Simulation ( MCMC ) using Bayesian inference was used to resolve a phylogenetic tree with the highest posterior probability to estimate the time of divergence from the most recent common ancestor ( MRCA ) implemented in BEAST ( Bayesian Evolutionary Analysis by Sampling Trees v . 1 . 4 . 8 ) [41] . Each independent run had a chain length of 30 , 000 , 000 with a sample frequency of 1000 . A general time-reversible substitution model was used , with site heterogeneity using a gamma distribution with a proportion of invariant sites and sampling across four categories . Analyses were performed using an HIV-1 generation time of 1 . 6 days [68] . Rate heterogeneity across codon positions was unlinked , and the mean fixed substitution rate was 2 . 16×10−5 under a relaxed uncorrelated exponential molecular clock . A coalescent piecewise-constant Bayesian Skyline model with ten groups was used as the tree prior . The MCMC log output of each run was examined in Tracer 1 . 4 to verify adequate chain mixing and estimated sample sizes of greater than 200 for parameters of interest , and log and tree files with a minimum of two independent runs were combined with a 10% burn-in using LogCombiner 1 . 4 . 8 . The target tree for each patient was summarized using TreeAnnotater 1 . 4 . 8 , and visualized in FigTree1 . 1 . 2 .	The work described in this report is directed at how HIV-1 viral RNA populations differ between the blood plasma and male genital tract in established infection . This site is of special interest since it is the proximal source of most transmissions of HIV-1 . Thus , lessons learned about HIV-1 in the seminal tract are directly relevant to the mechanism of HIV-1 transmission . We have used single genome amplification to generate viral sequences from paired blood and semen samples in men with chronic HIV-1 infection . When compared to viral populations in blood plasma , we observe that virus in the seminal plasma can be equilibrated , clonally-amplified , or compartmentalized . We have also performed a characterization of the cytokine and chemokine milieu in these two compartments . We report a dramatic concentration of immune modulators in the seminal plasma relative to the blood , and these likely enhance the potential for viral replication in this compartment by creating an environment where target cells are kept in an activated state . These data define new and distinct features of virus:host interactions and represent a significant advance in our understanding of HIV-1 replication in the male genital tract .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "infectious", "diseases/hiv", "infection", "and", "aids", "molecular", "biology", "virology/vaccines" ]	2010	HIV-1 Populations in Semen Arise through Multiple Mechanisms
Autosomal dominant lateral temporal epilepsy ( ADTLE ) is a focal epilepsy syndrome caused by mutations in the LGI1 gene , which encodes a secreted protein . Most ADLTE-causing mutations inhibit LGI1 protein secretion , and only a few secretion-positive missense mutations have been reported . Here we describe the effects of four disease-causing nonsynonymous LGI1 mutations , T380A , R407C , S473L , and R474Q , on protein secretion and extracellular interactions . Expression of LGI1 mutant proteins in cultured cells shows that these mutations do not inhibit protein secretion . This finding likely results from the lack of effects of these mutations on LGI1 protein folding , as suggested by 3D protein modelling . In addition , immunofluorescence and co-immunoprecipitation experiments reveal that all four mutations significantly impair interaction of LGI1 with the ADAM22 and ADAM23 receptors on the cell surface . These results support the existence of a second mechanism , alternative to inhibition of protein secretion , by which ADLTE-causing LGI1 mutations exert their loss-of-function effect extracellularly , and suggest that interactions of LGI1 with both ADAM22 and ADAM23 play an important role in the molecular mechanisms leading to ADLTE . Mutations in the leucine-rich , glioma-inactivated 1 ( LGI1 ) gene cause autosomal dominant lateral temporal epilepsy ( ADTLE ) [1 , 2] , a genetic epilepsy syndrome characterized by focal seizures with prominent auditory or aphasic symptoms , normal magnetic resonance imaging , and usually benign evolution [3–5] . ADLTE is inherited in autosomal dominant fashion with reduced penetrance [6 , 7] , and LGI1 mutations are found in about 30% of families with this syndrome [7] . To date , more than 30 ADLTE-causing mutations have been detected throughout the protein-coding region of LGI1 , resulting in either protein truncation or single amino acid substitutions [8 , 9] . LGI1 is mainly expressed in neurons [1 , 10 , 11] and shows no similarity to known ion channels . The predicted structure of the LGI1 protein comprises a signal peptide , four leucine-rich repeats ( LRRs ) [12] , and seven repeats named EPTP [13] or EAR [14] likely forming a beta-propeller structural domain [15] . Both LRR and beta-propeller domains mediate protein-protein interactions [15 , 16] . The LGI1 protein is secreted [10 , 17 , 18] , and most ADLTE-causing LGI1 mutations inhibit protein secretion [10 , 17 , 19–21] , consistent with a loss-of-function effect of mutations . We recently reported the first disease-causing LGI1 mutation ( R407C ) with no inhibitory effect on LGI1 secretion [22] . LGI1 has been implicated in various functions , some of which are mediated by interactions with two ADAM ( A Disintegrin And Metalloprotease domain ) receptors . LGI1 has been shown to bind to the postsynaptic receptor ADAM22 and this ligand-receptor complex participates in the control of synaptic strength at excitatory synapses [23] . It also binds to ADAM23 to stimulate neurite outgrowth both in vitro and in vivo [24] and may act as a trans-synaptic protein connecting the pre-synaptic ADAM23 with the post-synaptic ADAM22 receptors [25] . Though different in nature , each of these functions may potentially be related to epilepsy if impaired by mutations of LGI1 that prevent or disturb interactions with ADAM22 and ADAM23 receptors . Recent work has shown that serum LGI1 autoantibodies from patients with limbic encephalitis ( LE ) , which is characterized by cognitive dysfunction and seizures [26 , 27] , prevent interaction of LGI1 with ADAM22 [28] . It has also been shown that some ADLTE-related mutations allowing secretion of LGI1 impair its binding to ADAM22 but not to ADAM23 [29] . In this paper we show that secretion-positive LGI1 mutations impair extracellular binding to both ADAM22 and ADAM23 receptors , providing further evidence for the importance of the LGI1-ADAM22/23 protein complex in the molecular mechanisms underlying ADLTE . Previous 3D modelling of the LGI1 protein [30] predicted that some ADLTE-causing , nonsynonymous mutations in the C-terminal EPTP ( beta-propeller ) domain could allow secretion of the mutant protein and exert their pathogenic effects extracellularly , suggesting a distinction between correct protein folding and physiological function . In the present study , we used the 3D model of the LGI1 EPTP domain to envisage the possible effects of four disease-causing mutations , T380A , R407C , S473L and R474Q , on the structure or function of this protein region . The genetic features of these mutations are summarized in Table 1 . The clinical features associated with three of them have been described [22 , 31 , 32] , whereas the phenotype caused by the T380A mutation will be described elsewhere . Fig 1 shows the LGI1 EPTP domain with the T380A , S473L , R474Q , and R407C mutated residue positions highlighted and additional information on conservation and electrostatics on the protein surface . All four mutated residues lie on the surface of the beta-propeller domain within or very close to a ring of conserved residues located on the top surface , which has been predicted to be crucial for the protein interactions mediated by the EPTP domain [30] . Amino acid substitutions perturbing the organization of this area may affect LGI1 function by altering the interaction surface selectively while leaving the protein fold virtually unaffected . This situation is remarkably different from that predicted for mutations suppressing LGI1 secretion , which likely alter the folding of either LRR or EPTP domain [see ref . 30] , and may provide valuable clues into LGI1 function arising through protein-protein interactions . The inability of the R407C substitution to suppress LGI1 secretion was demonstrated previously [22] . To ascertain the consequences of the LGI1 c . 1138A>G ( T380A ) , c . 1418C>T ( S473L ) , and c . 1421G>A ( R474Q ) mutations on secretion of LGI1 , we transfected expression constructs containing the wild type and mutated LGI1 cDNAs into human embryonic kidney 293T ( HEK293T ) cells and analyzed cell lysates and concentrated conditioned media by western blot using an anti-LGI1 antibody . All three mutant proteins as well as the wild type control were detected in the conditioned media of transfected cells and , though in variable amounts , in the cell lysates ( Fig 2 ) . We consistently observed in three independent experiments a low amount of LGI1-T380A in the medium ( 31% of the secreted wild type protein; S1 Fig ) , suggesting that secretion of this mutant protein was partially hampered ( see Discussion ) . In contrast , the LGI1 mutant protein carrying the pathogenic I122K substitution [21] was not secreted from transfected cells ( Fig 2 ) , as previously shown for many other ADLTE-causing single amino acid substitutions in both LRR and EPTP regions [10 , 17 , 19–21] . Thus , according to this secretion assay , the S473L and R474Q mutations do not inhibit secretion of LGI1 , as it was the case for R407C , whereas T380A allows secretion of part of the mutated protein . For this reason all four study mutations are collectively named here sec+ mutations , whereas mutations inhibiting LGI1 secretion are termed sec- mutations . In ADLTE patients carrying LGI1 heterozygous mutations , allelic wild type and mutant proteins are synthesized in the same neuronal cells . To test whether mutant proteins could perturb secretion of co-expressed wild type LGI1 , we co-transfected HEK293T cells with wild type LGI1 tagged with green fluorescent protein ( GFP ) and either LGI1-R407C ( sec+ ) or LGI1-I122K ( sec- ) flagged constructs and determined the amounts of secreted LGI1-GFP by western blot . Fig 3 shows that neither the sec+ nor the sec- mutant proteins altered the amount of secreted LGI1-GFP . Repeated experiments yielded consistent results , indicating that both types of mutant proteins do not interfere with the secretion process of wild type LGI1 synthesized in the same cells . To investigate whether the four sec+ mutations affected the interaction of LGI1 with ADAM22/23 receptors , we overexpressed the wild type and mutant LGI1-Flag cDNAs together with HA-fused ADAM22 or ADAM23 constructs in COS7 cells . Thirty-six hours after co-transfection , cells were stained with anti-Flag and anti-HA antibodies and double-immunofluorescence analysis was carried out using confocal imaging . As exemplified in Fig 4 , the wild type LGI1 protein , when co-expressed with either ADAM22 or ADAM23 , mostly co-localized with either receptor on the cell membrane , whereas LGI1 mutant molecules failed to interact or interacted poorly with ADAM22 and ADAM23 receptors . This was consistently observed in three different experiments . Overall , the percentage of sec+ mutant proteins bound to either ADAM receptor on the cell membrane ( 0–33% ) was significantly lower than that of wild-type LGI1 ( 76–88%; Chi-square test , p < 0 . 0001 ) ( Table 2 ) . Blood sera from patients with LE containing high titre of LGI1 autoantibodies [26 , 27] have been shown to neutralize the LGI1-ADAM22 interaction [28] . To show that co-expressed LGI1 and ADAM22 interact on the cell surface , we co-transfected wild type LGI1-Flag and HA-ADAM22 cDNAs into COS7 cells that were cultured in conditioned medium containing serum from an LE patient . We consistently found in three different experiments that increasing concentrations of LE serum in the conditioned media progressively reduced binding of LGI1 to membrane-bound ADAM22 ( total cell counts in Table 3 ) . This result demonstrates that secreted LGI1 binds to the extracellular domain of ADAM22 , and that this interaction is disturbed by extracellular LGI1 autoantibodies . To independently assess whether LGI1 mutations affect binding to ADAM22 and ADAM23 , we performed co-immunoprecipitation assays from HEK293T cells overexpressing HA-tagged ADAM22 or HA-ADAM23 and LGI1-Flag . As shown in Fig 5A , HA-ADAM22 , which is present in both immature ( 110 kDa ) and mature ( 90 kDa ) forms as previously observed ( PMID:20156119 ) , efficiently co-precipitated wild type LGI1 . Instead , all LGI1 sec+ mutations affected LGI1-ADAM22 interactions: the T380A mutation completely disrupted the interaction , the S473L mutation almost abolished it , whereas R407C and R474Q reduced the affinity for ADAM22 . We also assessed the effect of sec+ mutations on LGI1-ADAM23 interaction using co-immunoprecipitation . As shown in Fig 5B , only wild type LGI1 was clearly immunoprecipitated by ADAM23 , consistent with the results obtained with ADAM22 , while pathological LGI1 mutations displayed reduced affinity for ADAM23 . We obtained consistent results from two experiments . In both cases expression of HA-ADAM23 was found to be lower than that of HA-ADAM22 , resulting in a lower efficiency of LGI1 immunoprecipitation . We show in this paper that several ADLTE-causing LGI1 mutations affecting amino acids of the C-terminal EPTP domain allow protein secretion and reduce affinity of secreted LGI1 for the ADAM22 and ADAM23 neuronal receptors . We also show that secreted LGI1 binds to ADAM22 on the cell surface , which was assumed but not demonstrated in previous works , and that secretion of wild type LGI1 is not influenced by LGI1 mutant proteins co-expressed in the same cells , ruling out any dominant-negative effect of both sec- and sec+ mutant proteins on trafficking and secretion of wild type LGI1 . Inhibition of secretion has long been the sole demonstrated mechanism by which LGI1 mutations cause loss of protein function [8] . Recently we and others reported ADLTE-causative LGI1 mutations that do not inhibit protein secretion [22 , 29] , indicating that LGI1 mutations can be either secretion-defective ( sec- ) or secretion-competent ( sec+ ) . Both types of mutations have loss-of-function effects , but sec+ mutations appear to exert their effect extracellularly by decreasing molecular affinity for both ADAM22 and ADAM23 . Overall , out of the 22 LGI1 missense mutations tested for secretion [10 , 17 , 19–22 , 29 , 33 , and present work] , four ( 18% ) are sec+ , all giving rise to an amino acid substitution in the EPTP domain . The presence of a comparatively large fraction of ADLTE-causing mutations that do not affect LGI1 secretion suggests that they allow the protein fold to be maintained while disrupting LGI1 interactions with other proteins . The surface residues replaced by sec+ mutations might either be directly responsible for the LGI1-ADAM22/23 interactions , or perturb , when mutated , the capability of nearby residues to interact with critical amino acids on the surface of ADAM22 and ADAM23 . Particularly , any mutation disturbing the conserved residues forming a ring on the EPTP top surface ( see Fig 1 ) might weaken the interaction or even abolish it entirely . In a recent work , Yakoi et al . [29] showed that the sec+ mutation S473L decreased binding of LGI1 to ADAM22 , but interaction with ADAM23 remained unchanged . They concluded that the pathogenic mechanisms underlying ADLTE in the case of sec+ mutations might result mostly , if not completely , from a dysfunction of the LGI1-ADAM22 complex . In contrast , we found that sec+ mutant proteins exhibit reduced binding affinity to both ADAM22 and ADAM23 . This discrepancy may be explained by the different assays employed . Compared with the tandem-affinity purification of LGI1 from brain tissue of transgenic mice expressing mutant proteins used by Yakoi and co-workers , our optimized cell-based assays ( 36 h incubation after co-transfection; see Methods ) may be more sensitive in detecting the decreased binding ability of sec+ mutations . Thus , our results suggest that both ADAM receptors likely play a role in the pathogenesis of ADLTE through their interactions with LGI1 . Genetic disorders that are caused by defects in a ligand for a particular receptor are frequently mirrored by disorders in which that receptor is deficient . However , a causative role for ADAM22 and ADAM23 in ADLTE has been questioned in genetic studies of families without LGI1 mutations: direct sequencing of ADAM22 exons revealed no disease-causing mutations [34 , 35] , and linkage analysis with microsatellite markers within or near the ADAM23 gene failed to reveal any significant linkage peak [36] . The absence of causative mutations in ADAM22 and ADAM23 , however , is not in contrast with their involvement in the molecular pathway underlying ADLTE , and this apparent contradiction can be explained in different ways . It is possible that mutations in these genes have not been detected in the limited set of ADLTE families tested because they occur at low frequency , a hypothesis supported by recent studies suggesting a relatively high genetic heterogeneity in ADLTE families free of LGI1 mutations [7] . Alternatively , LGI1-associated ADLTE may reflect a partial loss of function of the LGI1 ligand at both receptor proteins ADAM22 and ADAM23 , whereas heterozygous mutations in only one of these receptors may not be sufficient to cause the syndrome . The recent identification of ADAM22 compound heterozygous mutations in a patient with severe encephalopathy and epilepsy , who inherited the mutations by healthy parents , supports the hypothesis that ADAM22 may primarily be a recessive disease gene [37] . Also , protein models suggesting that LGI1 is able to interact with ADAM22 and ADAM23 simultaneously [25 , 30] provide support to this interpretation , which might also apply to other as yet unknown proteins interacting with LGI1 . Further support to the importance in epileptic disorders of the ligand-receptor interaction between LGI1 and ADAM22/23 comes from a recent study showing that these interactions are specifically impaired by LGI1 autoantibodies from the sera of patients with limbic encephalitis , which is characterized by amnesia and seizures [28] . Reduced binding of LGI1 to ADAM22/23 may therefore be a pathogenic mechanism for both genetically inherited and acquired epilepsy disorders . Recent work has shown that haploinsufficiency caused by sec- mutations does not result from lack of protein secretion but from intracellular degradation of misfolded mutant proteins by the endoplasmic reticulum protein quality-control mechanisms [29] . This is in agreement with the predictions of our 3D model for sec- mutations , which likely destabilize protein folding [30] , and also provides a possible explanation for the low amount of secreted LGI1-T380A we observed . This mutant protein was found virtually absent in the cell culture medium in a previous work [29] , in which the T380A substitution was therefore regarded as a sec- mutation . Instead , we found a low but consistently detectable amount of LGI1-T380A in the cell medium ( S1 Fig ) . Although we regarded it as a sec+ mutation for simplicity , the T380A substitution is not readily classifiable because only a relatively low proportion of the mutant protein appears to be secreted , whereas the rest of the mutant molecules might be degraded in the endoplasmic reticulum , possibly due to a mild alteration of the EPTP domain folding resulting from this mutation . In any case , the secreted LGI1-T380A protein failed to bind to both ADAM receptors under our experimental conditions , suggesting a loss-of-function effect of this genetic defect . Despite its possible mixed effects on protein stability and secretion , this mutation leads to the same clinical phenotype caused by LGI1 sec+ as well as sec- mutations . The discovery of sec+ mutations provides new opportunities for investigating LGI1 functions and understanding their relevance to ADLTE . For example , the study of sec+ mutant proteins in in vitro cell systems may help clarify whether some functions attributed to LGI1 , such as control of dendrite growth and cell adhesion , are related to ADLTE; also , animal models incorporating these types of mutations may provide new insights into the role of LGI1 in processes such a synaptic maturation and transmission . These studies will improve our comprehension of the pathogenic mechanisms of ADLTE as a paradigm of non-ion channel idiopathic epilepsy . Moreover , new therapeutic strategies that make use of chemical correctors to restore protein folding might be effective with LGI1 sec- mutations [29] but not with sec+ mutations , which have little or no effect on protein folding . Therefore , a correct classification of LGI1 mutations based on their effects of protein folding and secretion might have in the future important therapeutic implications . Both the LGI1 beta-propeller structure model as recently published [30] and the crystal structure of ADAM22 , PDB identifier 3G5C [38] , were visualized using PyMOL ( DeLano Scientific , URL: http://pymol . sourceforge . net/ ) . The degree of conservation on the protein surface was mapped with ConSurf [39] , using all known LGI1 through LGI4 homologs and the ADAM22 orthologs from OMAbrowser [40] respectively . The electrostatic surface was calculated with BLUUES [41] and mapped on the protein structures . The pathogenicity of four LGI1 mutations has been predicted also using the two widely used computational methods , SIFT ( http://sift . bii . a-star . edu . sg/ ) and Polyphen-2 ( http://genetics . bwh . harvard . edu/pph2/ ) . The SIFT score indicates the degree of conservation derived from a sequence alignment of closely related proteins . Polyphen-2 is based on both sequence and structural information . It uses the PSIC ( Position-Specific Independent Count ) software to calculate a profile matrix for each position in the sequence alignment . The difference between profile scores of allelic variants indicates the probability of the substitution to be observed in the protein family at that position . Rabbit anti-LGI1 antibodies were from Abcam ( Cambridge , UK; catalog No . ab30868 ) ; mouse anti-HA . 11 , clone No . 16B12 , from Covance ( Princeton , NJ , USA ) ; rabbit anti-Flag from Sigma Aldrich ( St . Louis , MO , USA; catalog No . F7425 ) . Secondary antibodies Alexa-Fluor 488-conjugated goat anti-mouse IgG were from Life Technology ( Grand Island , NY , USA; catalog No . A11001 ) ; and Cy3-conjugated goat anti-rabbit IgG from Jackson ImmunoResearch Laboratories ( West Grove , PA , USA; catalog No . 111-165-003 ) . Horseradish peroxidase-conjugated mouse-specific IgG , catalog No . P0260 , and horseradish peroxidase-conjugated rabbit-specific IgG , catalog No . P0448 , were from DakoCytomation ( Denmark A/S ) . Cell-based LGI1 secretion assay was performed as described in detail previously [18] . Briefly , expression constructs containing the LGI1 wild type or mutant cDNAs were transfected into HEK293T cells using Lipofectamine 2000 ( Life Technology ) , following the manufacturer instructions . Twenty-four hours after transfection , cells were washed twice and then re-fed with serum-free medium Opti-MEM ( Life Technology ) . After 16–20 hours , the medium was collected and centrifuged to pellet cell debris , and the supernatant was concentrated about 20x using Centricon YM30 concentrators ( Merk-Millipore , Billerica , MA , USA ) . Cells were lysed in Triton lysis buffer ( 25 mM Tris pH 7 . 4 , 150 mM NaCl , 1% ( vol/vol ) Triton , 10% ( vol/vol ) Glycerol , 1mM EDTA ) supplemented with proteases and phosphatase inhibitors . Cell proteins ( 14 μg/lane ) and concentrated media were separated on 12% NuPAGE ( Life Technology ) and then electroblotted onto nitrocellulose membrane . The integrity of the Western blot was analysed by Red Ponceau staining . Destained membranes were blocked with 10% ( w/v ) skimmed milk in tris-buffered saline ( TBS ) for 1 hr and then incubated with primary antibody in TBS containing 2% ( vol/vol ) skimmed milk for 2 hrs at room temperature . Proteins immunostained with anti-LGI1 antibody were detected with a horseradish peroxidase-labelled secondary antibody and enhanced chemiluminescence reagent and visualized by autoradiography . To examine the effect of LGI1 mutant proteins on secretion of wild type LGI1 , HEK293T cells were co-transfected with wild type LGI1-GFP and a Flag-containing LGI1 mutant , either c . 1219C>T ( R407C ) or c . 365C>A ( I122K ) , constructs ( 1 . 5 ug each ) . Following cell incubation , serum-free conditioned medium was concentrated and analyzed by western blot as described above . As gel loading control , a fixed quantity of IgG ( 14 ng ) was added to 6 ml cell medium before concentration ( 60x ) so as to load approximately 2 ng of IgG in each gel lane . The IgG were detected with a horseradish peroxidase-labelled secondary antibody . This experiment was repeated three times to confirm reliability of results . COS7 cells were seeded on sterile glass coverslips and co-transfected with wild-type or mutant LGI1-Flag and HA-tagged ADAM22 or ADAM23 cDNAs ( 3 ug of total DNA ) . Thirty-six hours after transfection , cells were fixed with 4% paraformaldehyde ( PFA ) at room temperature for 10 min and blocked with PBS containing 2 mg/ml bovine serum albumin ( BSA ) ( Sigma-Aldrich ) for 40 min . Fixed cells were stained with rabbit anti-Flag antibody ( 1:300 ) , followed by Cy3-conjugated anti-rabbit secondary antibody ( 1 . 300 ) . Then , the cells were permeabilized with 0 . 1% Triton X-100 for 10 min , blocked with PBS containing 2mg/ml BSA and stained with anti-HA monoclonal antibody ( 1:500 ) , followed by Alexa-Fluor 488-conjugated anti-mouse secondary antibody ( 1:500 ) . All the antibodies were diluted in 1% BSA/PBS; this was followed by washes with PBS . Finally , they were fixed with mounting medium containing DAPI ( Vectashield; Vector Laboratories ) . Two coverslips were made for every transfection experiment , and twenty random fields were taken ( magnification 400X ) . For every field the number of co-immunostained cells ( LGI1 and ADAM 22 or ADAM 23 ) was counted and the percentage of cells with both signals on the cell surface was estimated . In total , three independent experiments were performed . For cell counting , slides were analysed using a Leica-DM 5000B Epifluorescence microscope . Confocal images were acquired with a Radiance 2000 confocal microscope ( BioRad ) . The GraphPad software program ( http://www . graphpad . com/quickcalcs/ ) was used to calculate Chi-square and the two-tailed P-value to compare the frequency of membrane-bound LGI1 proteins between cells carrying the wild type LGI1 protein and cells expressing LGI1 mutated proteins . Blood serum from a patient with LGI1-positive autoimmune LE was used to investigate weather interaction of LGI1 with ADAM22 occurred on the outer surface of the cell membrane . COS7 cells were seeded on sterile glass coverslips and co-transfected with wild type LGI1-Flag and HA-tagged ADAM22 cDNAs as described above . Twenty hours after transfection the medium was replaced with serum-free medium additioned with 2 , 5% , 5% and 10% LE patient serum . After twenty-eight hours cells were processed and immunofluorescence analysis was performed as described above . Statistical analysis was performed as described above . LGI1-Flag wild-type or clinical mutants were co-transfected with HA-ADAM22 or HA-ADAM23 into HEK293T cells seeded on a six wells plate . Thirty-six hours after transfection , cells were lysed in 200 μl of lysis buffer ( 50 mM Tris-HCl , pH 7 . 5 , 1 mM EDTA , 1 mM sodium orthovanadate , 10 mM sodium β-glycerophosphate , 5 mM sodium pyrophosphate , 0 . 27 M sucrose and 1% ( w/v ) Tween 20 in the presence of lx protease inhibitor cocktail ( Sigma Aldrich ) and incubated on ice for 20 minutes . Lysates were subsequently clarified by centrifugation at 14000 g for 20 minutes and 0 . 5 mg of total proteins in 500 μl of volume were incubated with 0 . 5 μg of mouse monoclonal anti-HA antibody ( Roche ) . After 2 hours of incubation at 4°C with gentle rocking , 20 μl of Protein G sepharose ( GE Healthcare Life Sciences ) were supplemented and incubated 1 hour at 4°C . After 5 washes in washing buffer ( 50 mM Tris/HCl , pH 7 . 5 , 1 mM EDTA , 1 mM sodium orthovanadate , 10 mM sodium β-glycerophosphate , 5 mM sodium pyrophosphate , 0 . 27 M sucrose , 1% ( v/v ) tween 20 and 250 mM NaCl ) , proteins were eluted in 2x sample buffer ( Invitrogen ) , separated in NuPAGE 4–12% ( Life Technology ) and electro-blotted on polyvinylidene fluoride membrane ( PVDF ) ( Merk-Millipore ) . PVDF membranes were saturated with 10% skimmed milk in TBS supplemented with 0 . 05% tween 20 and proteins detected using mouse primary anti-HA ( 1:1000 ) and rabbit anti-LGI1 ( 1:1000 ) antibodies and secondary antibodies conjugated with horseradish peroxidase mouse- specific IgG and rabbit-specific IgG ( both diluted 1:2000 ) .	Temporal lobe epilepsy is the most common form of focal epilepsy . It is frequently associated with structural brain abnormalities , but genetic forms caused by mutations in major genes have also been described . Autosomal dominant lateral temporal epilepsy ( ADLTE ) is a familial condition characterized by focal seizures with prominent auditory symptoms . ADLTE-causing mutations are found in the LGI1 gene in about 30% of affected families . LGI1 encodes a protein , LGI1 , that is secreted by neurons . Most LGI1 mutations suppress protein secretion , thereby preventing protein function in the extracellular environment . In this paper , we examine the effects of four LGI1 mutations and show that they do not inhibit secretion of the LGI1 protein but impair its interaction with the neuronal receptors ADAM22 and ADAM23 . In agreement with these findings , a three- dimensional model of the protein predicts that these mutations have no impact on LGI1 structure but instead may affect amino acids that are critical for interactions with ADAM receptors . Our results provide novel evidence for an extracellular mechanism through which mutant LGI1 proteins cause ADLTE and strengthen the importance of LGI1-ADAM22/23 protein complex in the mechanisms underlying ADLTE .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "protein", "transport", "medicine", "and", "health", "sciences", "protein", "interactions", "cell", "processes", "physiological", "processes", "mutation", "substitution", "mutation", "immunoprecipitation", "co-immunoprecipitation", "cellular", "structures", "and", "organelles...	2016	Secretion-Positive LGI1 Mutations Linked to Lateral Temporal Epilepsy Impair Binding to ADAM22 and ADAM23 Receptors
The procambium and cambium are meristematic tissues from which vascular tissue is derived . Vascular initials differentiate into phloem towards the outside of the stem and xylem towards the inside . A small peptide derived from CLV-3/ESR1-LIKE 41 ( CLE41 ) is thought to promote cell divisions in vascular meristems by signalling through the PHLOEM INTERCALLATED WITH XYLEM ( PXY ) receptor kinase . pxy mutants , however , display only small reductions in vascular cell number , suggesting a mechanism exists that allows plants to compensate for the absence of PXY . Consistent with this idea , we identify a large number of genes specifically upregulated in pxy mutants , including several AP2/ERF transcription factors . These transcription factors are required for normal cell division in the cambium and procambium . These same transcription factors are also upregulated by ethylene and in ethylene-overproducing eto1 mutants . eto1 mutants also exhibit an increase in vascular cell division that is dependent upon the function of at least 2 of these ERF genes . Furthermore , blocking ethylene signalling using a variety of ethylene insensitive mutants such as ein2 enhances the cell division defect of pxy . Our results suggest that these factors define a novel pathway that acts in parallel to PXY/CLE41 to regulate cell division in developing vascular tissue . We propose a model whereby vascular cell division is regulated both by PXY signalling and ethylene/ERF signalling . Under normal circumstances , however , PXY signalling acts to repress the ethylene/ERF pathway . Organised cell division and differentiation are required throughout nature for development of ordered body plans . The annual rings of trees which result from seasonal differences in radial growth are a widely recognisable example of the highly regulated nature of this process . Radial growth is achieved by generation of new vascular tissue that occurs via ordered cell divisions in the vascular meristem known as the cambium . Divisions in the cambium result in displacement of older cells to its periphery where they subsequently differentiate into xylem towards the inside of the stem or phloem towards the outside . Cambial cells divide in a highly ordered manner along their long axis giving rise to files of cells in a process that is most apparent in the growth rings of trees but also apparent in most higher plants such as Arabidopsis [1] . The ordered nature of this cell division is required for vascular tissue organisation and consequently is essential for both primary and secondary vascular development [2] . The receptor kinase PHLOEM INTERCALATED WITH XYLEM ( PXY ) was identified as being essential for ordered , coordinated cell divisions in the procambium [2] and has been shown to bind a peptide derived from CLV-3/ESR1-LIKE 41 ( CLE41 ) and CLE44 [3] , which was originally identified as TDIF , a peptide that represses tracheary element formation in transdifferentiation assays [4] . CLE41 , and related CLE42 [5] , [6] also function through the PXY receptor to provide positional information required for orientation of the cell division plane in the procambium [7] . CLE41 and CLE42 over-expression lines have more cells in vascular bundles than those of wild type counterparts [7] and an increased diameter of the hypocotyl vascular cylinder [3] , [8] . These increases in vascular cell number and hypocotyl diameter are completely abolished in pxy 35S::CLE41 and pxy 35S::CLE42 lines [7] . Consequently , CLE41/42 induced vascular cell divisions occur in a PXY dependent manner demonstrating that PXY signalling , in addition to setting the division plane , also promotes the divisions themselves [3] , [7] . A downstream target of PXY , the WUSCHEL-RELATED HOMEOBOX ( WOX ) gene , WOX4 is thought to be required for the promotion of these divisions [9] and wox4 mutants have been shown to have defects in vascular proliferation [10] , [11] . Given that PXY signalling promotes vascular cell division , it might be expected that pxy mutants demonstrate a reduction in cell division , however in inflorescence stems of 5 week old plants no defects in the rate of cell division were reported [2] . Furthermore , pxy mutant hypocotyls exhibit only a small reduction in diameter at senescence suggesting only a small decrease in the total number of vascular cell divisions [7] . One explanation for this apparent contradiction is that a compensatory pathway exists that may be activated in the absence of pxy . The gaseous hormone ethylene , has been shown to promote radial growth in several tree species [12] , [13] , [14] , and more recently , radial growth and increased cambial cell division in tension wood of poplar was shown to be ethylene-induced [15] . Here we demonstrate that pxy and wox4 work together with several ETHYLENE RESONSE FACTOR ( ERF ) transcription factors and ethylene signalling to regulate cell divisions during Arabidopsis vascular development . We propose that in pxy mutants , cell numbers are maintained by the up-regulation of an ethylene pathway that increases expression of these ERFs . We present evidence for a model whereby vascular cell division is promoted by an interaction between PXY and ethylene signalling . Consequently , in addition to its role in mediating stress responses including the development of tension wood [15] , our results suggest a more general role for ethylene in regulation of vascular cell division . There are apparent contradictory observations with regard to the role of PXY/CLE41 in the regulation of the rate of vascular cell division . While CLE41 overexpression results in more cells [7] , loss of PXY has little effect on vascular cell number [2] . One possible explanation is that an alternative pathway that also promotes vascular cell division is upregulated in pxy mutant plants . To test this hypothesis , we generated microarray expression data for the central part of pxy-3 mutant inflorescence stems and compared it to comparable data from wild type ( Experiment E-MEXP-2420 , http://www . ebi . ac . uk/arrayexpress ) . Intriguingly 12 members of the AP2/ERF family of transcription factors , predominantly from classes VIII-X [16] were found to be expressed at higher levels in pxy than wild type ( Table S1 ) . ERF109 ( At4g34410; also known as RRTF [17] ) , ERF11 ( At1g28370 ) , ERF104 ( At5g61600 ) , and ERF018 ( At1g74930; also known as ORA47 [18] ) were increased 4 . 3 , 3 . 0 , 2 . 8 and 2 . 8 , -fold , respectively ( Table S1 ) . Four further AP2/ERF family members AtERF1 ( At4g17500 ) , ERF2 ( At5g47220 ) , ERF5 ( At5g47230 ) , and ERF6 ( At4g17490 ) demonstrated between 1 . 5 and 2-fold increases in expression . To confirm that the expression changes identified in array experiments were robust , we used qRT-PCR to retest expression levels of ERF018 , ERF109 and AtERF1 in wild type and pxy-3 plants using RNA isolated from similar tissue to that used in microarrays . We observed similar fold changes in qRT-PCR to those previously identified in arrays when relative expression levels were normalised to that of ACT2 or 18s rRNA ( Figure S1 ) . Arabidopsis inflorescence stems represent a developmental series as vascular tissue at the top of stems is newly initiated in contrast to more mature vasculature at the base of stems . To further investigate the expression pattern of genes differentially expressed in pxy , we assayed expression of four of the most upregulated ERF's ( AtERF1 , ERF11 , ERF109 and ERF018 ) at both the top ( 2–4 cm below the shoot apex ) and the base ( 1–3 cm above the rosette ) of inflorescence stems from 5 week old plants using qRT-PCR ( Figure 1A ) . The base of stems demonstrated larger fold changes in gene expression in pxy than was observed in the middle of stems ( Table S1; Figure S1 ) as ERF109 , ERF11 , AtERF1 and ERF018 expression was increased 20 , 7 , 7 and 3-fold , respectively . In contrast , at the top of stems significant changes were only observed for AtERF1 and ERF11 suggesting that expression of these genes is upregulated in newly formed pxy mutant stems and this upregulation is progressively increased as vascular tissue matures ( Figure 1A ) . Similar increases in ERF expression were also observed in pxy hypocotyls compared to wild type counterparts ( Figure 1A ) . WOX4 has been placed in a pathway downstream of the PXY receptor kinase [9] so we hypothesised that these genes up-regulated in pxy should also be up-regulated in wox4 mutants . qRT-PCR analysis of expression of the same ERF's upregulated in pxy mutants also demonstrated increases in expression in wox4 ( Figure 1B ) . These observations suggest that ERF expression is suppressed by the pxy signalling pathway and that repression of ERF expression occurs downstream of WOX4 . We tested for vascular gene expression of two ERF transcription factors , ERF109 and ERF018 , using in situ hybridization on sections of inflorescence stem from 5 week old plants 4 cm below the shoot apex ( Figure 2 ) and found that Digoxigenin labelled antisense probes labelled many cell types . However , ERF109 and ERF018 expression was strongest in vascular bundles . Notably , in wild type , expression for both genes was most prominent in the procambium ( arrows in Figure 2A , 2B ) but absent from the phloem . In pxy mutant vascular tissue , ERF109 and ERF018 expression also appeared most prominently in the procambium and xylem ( Figure 2 ) . Sense negative controls for both genes did not label tissue above background levels but an antisense CLE41 positive control specifically labelled phloem tissue ( Figure 2C ) as previously reported [7] . Quantitative data from microarrays and qRT-PCR , combined with prominent vascular expression of these genes consequently suggests a role for ERF109 and ERF018 in vascular tissue . To determine the functional relevance of the gene expression changes observed in ERF's , we identified erf018 and erf109 loss-of-function mutants in publicly available T-DNA insertion libraries as these genes demonstrated relatively large increases in expression in pxy mutants . A confirmed T-DNA insertion within the coding sequence of ERF109 ( Salk_150614 ) was renamed erf109-1 , however , no insertion mutant was available that disrupted the coding sequence of ERF018 . Salk_109440 line ( erf018-1 ) was found to harbour a T-DNA insertion 142 base pairs upstream of the transcriptional start site and 249 base pairs upstream of the ATG . qPCR was used to analyse the expression of ERF018 in these lines and we found that expression was reduced to 60% of wild type levels ( Figure S2 ) indicating that erf018-1 is a weak allele . Gross morphology of erf018 , erf109 single and erf109 erf018 double mutants appeared identical to wild type counterparts ( Figure 3 ) and the number of cells in erf018 and erf109 mutant vascular bundles was unchanged from wild type in 10 week old inflorescence stems ( Figure 4A , Figure 5A ) . In contrast , erf109 erf018 double mutants demonstrated a small but significant reduction in the number of cells per vascular bundle ( 78% of wild type; Figure 4A , Figure 5A ) suggesting that ERF109 and ERF018 act redundantly in promoting cell division in vascular bundles . The role of ERF109 and ERF018 in secondary growth was addressed in Arabidopsis hypocotyls . Several authors have used hypocotyl diameter as a measure of cell division during secondary growth [3] , [7] , [8] , [19] , [20] , [21] . Consistent with our observation that erf109 erf018 lines had reductions in vascular cell division in stems , hypocotyl diameter was also reduced in erf109 erf018 double mutants to 83% of wild type diameter ( Figure 4B , Figure 5B ) . Consequently , ERF109 and ERF018 are required for promotion of vascular cell division during both primary and secondary growth . It was clear from our analysis that ERF109 and ERF018 are required for promoting vascular cell divisions . Since these transcription factors are upregulated in pxy mutants it may be hypothesised that they represent a mechanism by which vascular cell division is maintained in the absence of PXY . pxy erf109 erf018 triple mutants were therefore generated with the expectation that if ERF transcription factors do compensate for loss of pxy then pxy erf109 erf018 lines would demonstrate a significant reduction in cell number when compared to pxy , erf109 erf018 or wild type . pxy mutant vascular bundles have been previously characterised with intercalated xylem and phloem , however , in inflorescence stems of 5 week old plants no differences in vascular cell number were observed [2] . We reasoned that differences in the number of cells in pxy vasculature may be observed in 10 week old tissue , particularly in hypocotyls which undergo continuous radial expansion , as subtle differences in the rate of cell division would have time to accumulate . All experiments on 10 week old tissue in this manuscript ( see below ) demonstrated a trend towards a reduction in cell number in pxy mutant vascular bundles compared to wild type ( see below ) , however , in this instance , differences proved not to be statistically significant ( Figure 4A , Figure 5A ) . Consistent with our hypothesis , vascular bundles of pxy erf109 mutants demonstrated a 27% reduction in cell number compared to wild type in contrast with pxy and erf109 which showed no significant difference ( Figure 4A , Figure 5A ) . Consequently , clear defects in pxy mutant vascular cell number only became apparent when pxy was combined with an erf109 mutant . pxy erf018 and pxy erf018 erf109 were also generated to determine whether erf018 demonstrated a similar interaction with pxy . pxy erf018 double mutant inflorescence stem vascular tissue did not differ from parental lines ( Figure 4A , Figure 5A ) , however , pxy erf018 erf109 lines demonstrated a 44% reduction in cells/vascular bundle demonstrating a significant enhancement of the pxy erf109 phenotype ( Figure 4A , Figure 5A ) . When analysing secondary growth in pxy erf109 double mutant hypocotyls , we found that the relationship between erf109 and pxy was similar to that observed in vascular bundles . pxy erf109 hypocotyls had the characteristic altered orientation of cell division associated with pxy mutants [7] but the hypocotyl diameters were narrower than either pxy or erf109 single mutants ( Figure 4B ) . The decrease in hypocotyl diameter was most dramatic in pxy erf109 erf018 mutants where hypocotyl diameters were only 63% of that observed in wild type . These observations are consistent with fewer cell divisions having occurred in the triple mutant than the respective doubles , single mutants and wild type ( Figure 4B , Figure 5B ) . As with our observation in vascular bundles , defects in vascular cell number are greatly enhanced when pxy mutants are combined with mutations in the ERF transcription factors erf018 and erf109 . We further examined ERF function in PXY signalling by analysing the function of AtERF1 ( At1g17500; upregulated in both pxy and wox4 mutants; Figure 1B , Table S1 ) . A T-DNA mutant ( Salk_036267 ) was isolated and used to test whether AtERF1 acted similarly to ERF018 and ERF109 . erf1 single and erf1 erf109 double mutants were indistinguishable from wild type , and although erf1 pxy double mutants were suggestive of a reduction in the size of vascular bundles compared to pxy single mutants , differences proved not significant ( Figure 5C ) . In contrast , pxy erf1 erf109 triple mutants demonstrated a dramatic decrease in the number of cells/vascular bundle ( 48% of that observed in wild type ) , a significant reduction when compared to respective single and double mutants when assayed at the base of the inflorescence stems of 10 week old plants ( Figure 5C ) . Similarly , pxy erf109 erf1 lines demonstrated reduced hypocotyl diameter ( 52% of wild type ) when compared to control lines ( ≤80% of wild type; Figure 5D ) . erf1 therefore enhances vascular cell division defects of erf109 pxy mutants in both inflorescence and hypocotyl . These data are consistent with a role for AtERF1 , ERF109 and ERF018 in promoting vascular cell division in the absence of PXY . Five of the ERF genes upregulated in pxy; AtERF1 , ERF2 , ERF5 , ERF003/Atg525190 and ERF11 have previously been shown to be induced by ethylene [22] , [23] , [24] , [25] . Furthermore , an enzyme responsible for catalysing the rate-limiting step of ethylene biosynthesis , ACS6 ( At4g11280 ) was upregulated 2 . 5 fold in pxy mutants ( Table S1; Figure 1B ) . Consequently , we hypothesised that the increase in expression of ERF transcription factors in pxy and wox4 mutants may be the result of an increase in ethylene signalling . To determine whether these genes also demonstrated elevated expression in stems of plants with higher levels of ethylene than wild type , their response to ethylene exposure was tested . We subjected five week old wild type Arabidopsis plants to ethylene stimuli of 3 hours , 16 hours and also made use of ethylene overproducer1 ( eto1 ) mutants which produce more ethylene than wild type [26] . Expression levels of ERF's were compared in inflorescence stems using qRT-PCR ( Figure 6 ) . Expression was increased in response to exogenous ethylene treatment as in plants exposed to ethylene for 3 hours , AtERF1 and ERF11 underwent a 3-fold induction ( Figure 6A ) and following a 16 hour treatment 2- and 5-fold inductions were observed ( Figure 6B ) . Increased ERF109 and ERF018 expression of approximately 3-fold was observed in eto1 . Consequently ERF109 , ERF018 , AtERF1 and ERF11 are ethylene responsive; however , the dynamics of induction varies in inflorescence stems . ERF1 and ERF11 demonstrated an early ethylene response and ERF018 and ERF109 expression was increased in response to a constitutive ethylene production ( Figure 6 ) . To confirm the relationship between ERF expression and ethylene in stems , we carried out the converse experiment . ERF levels were determined in ethylene insensitive 2 ( ein2 ) plants in which the ethylene signal transduction pathway is thought to be entirely abolished [27] . Consistent with ERF109 , ERF018 , AtERF1 and ERF11 acting downstream of the ethylene response in inflorescence stems , expression of the genes tested was reduced by half ( Figure 6D ) . It is notable that ein2 mutants do not demonstrate reductions in vascular cell number ( see below ) . Consequently , differences in ERF expression cannot be explained by phenotypic differences in vascular tissue and are likely the result of reduced ethylene signalling . Reports in poplar have demonstrated that ethylene promotes vascular cell division during secondary growth [15] , so in order to determine whether ethylene , and therefore ERF's , function similarly in Arabidopsis we analysed the inflorescence stems of eto1 mutants at six weeks and found that they exhibited an increase in the number of procambial cells ( Figure 7A–7B ) . eto1 mutants also demonstrated early onset of secondary growth as vascular cell divisions were observed in the interfascicular region prior to any divisions in wild type plants at an equivalent stage of development ( Figure 7A–7B ) . This phenotype was particularly evident when eto1 mutant vascular sections were subjected to in situ hybridization with ERF109 antisense probes . In wild-type , labelling was absent from interfascicular tissue but present in eto1 ( Figure 2 ) . The phenotypic consequences of constitutive ethylene production were confirmed by analysis of eto2 mutants [28] . At ten weeks , as with eto1 mutants , eto2 plants had larger vascular bundles and cell divisions between vascular bundles indicative of secondary growth which were absent in wild type ( Figure S3 ) . To confirm that the observed differences in eto1 and eto2 were significant , the number of cells in vascular bundles of 10 week inflorescence stems and hypocotyl diameters were determined . Increases in vascular cell number of 21% for eto1 and 34% for eto2 in inflorescence stems and increases in hypocotyl diameter of 19% and 31% , respectively were apparent ( Figure 7C–7D ) . Our data are therefore consistent with the idea that elevated levels of ethylene result in both an increase in vascular cell division and increased expression of ERF transcription factors which we have shown are required to promote vascular cell division in the absence of PXY . To confirm that ERF transcription factors upregulated in pxy mutants were required for vascular eto phenotypes and therefore ethylene mediated vascular expansion , we generated eto1 erf109 erf1 triple mutants . eto1 erf109 and eto1 erf1 double mutant lines were indistinguishable from eto1 single mutants , but in eto1 erf109 erf1 lines , vascular cell number was significantly smaller than that observed in eto1 ( Figure 8 ) . Furthermore interfascicular cell divisions that were sometimes present in eto1 lines were not observed in eto1 erf109 erf1 counterparts ( Figure S4 ) , and eto1 stems demonstrated an increase in diameter compared to those of eto1 erf109 erf1 ( Figure 3 ) , consistent with a requirement for ERF's in eto secondary growth phenotypes . Consequently , we have demonstrated that the ERF transcription factors that demonstrate increased expression in pxy mutants are upregulated in response to ethylene , their expression is reduced in ethylene signalling mutants and they are required for the phenotypic consequences of ethylene over-production in vascular tissue . To directly address the relationship between PXY and ethylene signalling , we crossed mutants that are unable to respond to ethylene to pxy . ein2 encodes an integral membrane of unknown function that is essential for ethylene signal transduction [27] and is the only single mutant thought to entirely abolish ethylene signalling [27] . pxy ein2 double mutants developed normal rosettes and inflorescence stems were initiated normally , however , the plants senesced early so analysis of ten week plants , consistent with quantitative phenotypic analysis elsewhere in this manuscript was not possible . Analysis was carried out on six week old plants but at this developmental stage , wild type plants had similar numbers of cells in vascular bundles as present at ten weeks suggesting that vascular proliferation in the stem was complete ( Figure 9A , 9C ) . Wild-type and ein2 vasculature in inflorescence stems were indistinguishable , with no significant difference in vascular cell number . pxy ein2 mutant vascular tissue demonstrated a dramatic reduction in vascular cell number ( 55% of wild type ) , having significantly fewer cells than pxy or ein2 single mutants ( Figure 9A , 9C ) and clearly demonstrating that ein2 is required for maintenance of vascular tissue in pxy mutants . Similar results were observed in the hypocotyl ( Figure 9B , 9D ) with pxy ein2 lines significantly smaller than ein2 or pxy single mutants . To confirm the results obtained with ein2 , two further mutants in the ethylene signal transduction pathway were analysed . ethylene receptor1 ( etr1 ) and ethylene insensitive5 ( ein5 ) encode an ethylene receptor [29] , and an exoribonuclease involved in ethylene signalling [30] , [31] , respectively . Neither etr1-3d nor ein5-1 exhibit the triple response and are partially ethylene insensitive [32] . In primary vascular tissue in inflorescence stems , in common with ein2 mutants , ein5 and etr1-3d were indistinguishable from wild type ( Figures S5 , S6 ) but in both cases a dramatic enhancement of the reduction in vascular cell number observed in pxy mutants ( see above ) was observed when pxy etr1-3d double mutants were analysed ( Figures S5 , S6 ) . Analysis of the role of etr1-3d and ein5 in hypocotyl secondary growth was also carried out . Hypocotyl diameters were measured at 10 weeks and ein5 was found not to differ from wild type , however etr1-3d demonstrated a small reduction ( Figure S6 ) . Although this differed from observations in ein2 and ein5 , this is likely due to the age of plants tested with respect to ein2 and differences in the level of reduction of ethylene signalling with respect to ein5 . In common with ein2 , etr1-3d and ein5 both strengthened the pxy phenotype as double mutants were smaller than respective singles ( Figures S5 , S6 ) . If ERF109 , ERF018 and ERF1 are targets of an ethylene-induced signalling mechanism that is upregulated in the absence of pxy , then pxy erf mutants should appear similar to those of pxy ein2 , pxy etr1-3d and pxy ein5 . As such , pxy erf109 erf018 and pxy erf109 erf1 vascular tissue was similar to that of pxy ein5 , pxy etr1-3d and pxy ein2 as in all instances , the pxy cell division phenotype was enhanced . It is notable that ethylene signalling does not appear to greatly influence PXY signalling . Expression levels of CLE41 , CLE42 , PXY and WOX4 were unchanged in ein2 mutants or erf109 erf018 plants ( Figure S7 ) . CLE41 , CLE42 and WOX4 expression was also unchanged in plants exposed to an ethylene stimulus , however , ethylene did promote PXY expression ( Figure S7 ) suggesting that PXY is to some extent ethylene responsive . Up regulation of the PXY signal transduction pathway by over expression of the CLE41 ligand results in massively increased vascular cell divisions , however , pxy mutants exhibit only limited reductions in cell division . We have identified a group of 12 ERF transcription factors that are upregulated in pxy mutants ( Table S1; Figure 1 ) . Loss of function analysis of three of these genes; ERF109 , ERF018 and AtERF1 resulted in plants with inflorescence stems that were characterised by reduced numbers of vascular cells suggesting that these genes promote cell division in vascular meristems ( Figure 4 , Figure 5 ) . This data suggests that these ERF transcription factors form part of a mechanism that is up-regulated in response to loss of pxy . Previous authors have demonstrated that five of the genes identified have increased expression in response to ethylene in seedlings [22] , [23] , [24] , [25] . We have demonstrated that several of the family members are upregulated in stems of ethylene overproducing eto1 mutants or in plants subjected to ethylene treatment ( Figure 6 ) . Furthermore these ERF's are required for the increased vascular tissue observed in eto1 plants ( Figure 7 , Figure 8 ) . An involvement of ethylene in vascular cell division in pxy plants is supported by analysis of ein2 , ein5 and etr1-3d mutants . EIN2 , EIN5 and ETR1 are required for normal ethylene signal transduction [32] and pxy ein2 , pxy ein5 and pxy etr1-3d plants had significant reductions in the number of vascular cells compared to single mutants or wild type ( Figure 9; Figures S5 , S6 ) . Taken together , our results demonstrate that ERF transcription factors promote vascular cell division , that their expression is influenced by PXY-repression of ethylene signalling , and consequently , these signalling pathways interact to control the rate of cell division in plant vascular tissue ( Figure 10 ) . 1-aminocyclopropane-1-carboxylic acid synthase 6 ( AtACS6 ) is also upregulated in pxy ( Table S1; Figure 1B ) . ACS enzymes catalyse the rate-limiting step of ethylene biosynthesis [33] , [34] , i . e . conversion of S-adenosylmethionine to ACC [35] . Ethylene has previously been shown to promote cell division in the organising centre of Arabidopsis roots [36] , and in the cambium of hybrid poplar [15] . In tree species , ethylene is produced in association with physical stress [13] , and is known to have a role in promoting development of tension wood [15] . Our results suggest that it may have a more general role in regulating the rate of cell division in the cambium ( Figure 10 ) , particularly as etr1-3d mutants demonstrated significant reductions in radial growth compared to wild type in hypocotyls ( Figure S6 ) . These conclusions are also consistent with earlier studies that demonstrate that in trees ethylene levels are higher at the height of the growing season than towards the end [13] . It may be case that ERF109 , ERF018 , AtERF1 and other ERF's upregulated in pxy are essential components in the vascular developmental programme and their expression can be modulated by ethylene and other external factors . ERF transcription factors affect a variety of processes [17] , , but strikingly , ERF018 and AtERF1 have also been described as being up-regulated by jasmonic acid [18] , [25] . Jasmonic acid has recently emerged as a key modulator of cell division in the cambium [39] and consequently ERF018 and AtERF1 may be key integration points for vascular development . In support of this hypothesis ( Figure 10 ) , phenotypes were observed in erf018 mutants ( in combination with erf109 ) despite the weak nature of the erf018 allele identified in this study . Our observations , and those of previous authors are consistent with an emerging picture that many of these transcription factors form part of a network [40] and how the plant responds to them is very much context dependent [41] . The ERF transcription factors analysed in this study have been suggested as having roles unrelated to vascular development [17] , [18] , [42] , [43] and consequently , ERF109 and ERF018 have broad expression patterns ( but are nevertheless expressed in vascular tissue ) , however , single mutants and neither erf109 erf1 nor erf109 erf018 double mutants demonstrated visible architectural defects , such as a change in height ( Figure 3 ) , suggesting that reductions in vascular cell number in these lines is not the consequence of a general disruption to plant growth . The phytohormones brassinosteroid [44] , cytokinin [45] , [46] , [47] , strigolactone [48] and auxin [49] , [50] have also been shown to have roles in Arabidopsis vascular development , and have been shown to be regulate each other's biosynthetic pathways [25] . Brassinosteroid upregulates genes required for ethylene biosynthesis and auxin up regulates genes involved in cytokinin biosynthesis . More complex interactions occur between ethylene and auxin , brassinosteroid and auxin as well as cytokinin and brassinosteroid , where phytohormone biosynthesis genes are both induced and repressed in response to respective phytohormone treatments [25] . A proper understanding of vascular cell division and differentiation will need to take into account interactions between these signalling pathways and their downstream targets . One phytohormone in addition to ethylene that has been placed in a network with PXY signalling is auxin . WOX4 is positively regulated by auxin which has led to the suggestion that auxin lies upstream of PXY signalling in a hypothesis that is supported by the observation that part of the WOX4 response to auxin is PXY-dependent [10] . In support of this hypothesis , ERF genes upregulated in pxy mutants also demonstrated increased expression in wox4 lines ( Figure 1B ) . However , WOX4 regulation may not depend entirely on PXY signalling and could also be regulated by an auxin-dependent , PXY-independent mechanism [51] because despite the observation that PXY is required for the WOX4 auxin response , WOX4 expression nevertheless increases in pxy mutants subjected to a 1 day auxin induction [10] . Further evidence for a WOX4 , PXY-independent response comes from the observation that wox4 enhances pxy mutants [9] . If WOX4 expression was entirely controlled by PXY signalling wox4 pxy double mutants and respective single mutants would have identical phenotypes . Experiments presented here , and by previous authors therefore place PXY signalling in a network containing auxin , ethylene and JA signalling ( Figure 10 ) . The procambium and cambium are meristematic tissues and as such , demonstrate similarities with the shoot apical meristem ( SAM ) and root apical meristem ( RAM ) . All three structures require CLAVATA ( CLV ) -like , and phytohormone signalling mechanisms for their regulation . In the SAM , CLE41-related CLV3 is secreted from stem cells and binds to the PXY-related CLV1 receptor [52] , . This precipitates a signal which culminates in negative regulation of WUSCHEL ( WUS ) , a homeobox gene which promotes stem cell fate – and therefore cells that secrete CLV3 . This feedback loop enables the plant to dynamically regulate the size of its apical stem cell population thus balancing organ generation with maintenance of its stem cells [56] , [57] . WUS expression is also controlled by cytokinin signalling , which is thought to add robustness to the feedback mechanism [58] . It is tempting to speculate that the relationship between PXY and ethylene signalling acts similarly . In this case , interaction between the PXY and ethylene pathways culminates in appropriate regulation of downstream transcription factors required to regulate the rate of cell division and recruitment of daughter cells into xylem and phloem . Plant growth conditions , and pxy alleles have been described previously [2] . T-DNA insertion lines in ERF018 ( salk_109440 ) , ERF109 ( salk_150614 ) , AtERF1 ( salk_036267 ) and wox4-1 ( GABI-Kat_462G01 ) were identified using the TAIR database [59] and confirmed using PCR . Insertion lines and eto1-1 , ein5-1 , and etr1-3d were obtained from NASC . erf109 erf018 , pxy erf109 , pxy erf018 , pxy erf018 erf109 , erf1 erf109 , pxy erf1 and pxy erf1 erf109 lines were identified in segregating F2 populations . Primers for pxy genotyping have been described previously [7] . Oligos SALK_ERF109LB and SALK_ERF109RB ( CGCGATGCTTTGTAGGAGTAG and TGTCAGGGTTTTTCCAGTGAC ) , SALK_ERF018LB and SALK_ERF018RB ( TTCATGCTCATGATGATGAGC and ATCGACGGTGGATTATTAGGG ) and salk-ERF1-F and salk-ERF1-R ( CGTTCCTAACCAAACCCTAGC and TCCTACTCTTCTCCCTGCTCC ) were used for the identification of erf109 , erf018 and erf1 mutants . pxy ein2 , pxy ein5 and pxy etr1-3d doubles were selected in the F3 generation from families that were ethylene insensitive as determined using a triple response screen [26] . Ethylene treatments of plants prior to measurement of ERF expression in inflorescence stems was carried out by placing Arabidopsis plants in a sealed container and generating ethylene gas to a maximal concentration of 500 µl l−1 of ethylene gas as described previously [60] . For comparison of wild type and pxy transcriptomes , Col-0 and pxy-3 lines were used . For each replicate , plants were germinated on MS agar plates prior to transfer to soil ( 6 plants per 10 cm pot ) , where they were grown on for 5 weeks under long day ( 16/8 h light/dark ) conditions at which point the inflorescence stem had 4–6 expanded siliques . Pots were randomised and rotated daily . For each replicate , the 6 primary inflorescence stems were taken from all the plants in a pot . Cauline leaves and side branches were removed . Stems were divided into 4 sections of equal size and RNA was isolated from the third section from the top using TRIzol Reagent ( Invitrogen ) . RNA was sent to the University of Manchester Genomic Technologies Facility ( http://www . ls . manchester . ac . uk/research/facilities/microarray/ ) where it was assessed for quality . ATH1 Affymetrix GeneChip oligonucleotide arrays were used to analyse the gene expression from each sample . Biotinylated cDNA samples from three biological replicates of pxy and wild type stems were synthesised and hybridized to Arabidopsis ATH1 genome oligonucleotide arrays . Technical quality control was performed with dChip ( V2005; www . dchip . org ) , using the default settings [61] . Background correction , quantile normalization , and gene expression analysis were performed using RMA in Bioconductor [62] . Differential expression analysis was performed using Limma using the functions lmFit and eBayes [63] . Microarray data has been submitted in a MIAME compliant standard to the Array Express database ( Experiment E-MEXP-2420 , http://www . ebi . ac . uk/arrayexpress ) . For RT-PCR , RNA was isolated using Trizol ( Invitrogen ) . cDNA synthesis , following DNase treatment , was performed using Superscript III reverse transcriptase ( Invitrogen ) . All samples were measured in technical triplicates on biological triplicates . The qRT-PCR reaction was performed using SYBR Green JumpStart Taq ReadyMix ( Sigma ) using an ABI Prism 7000 machine ( Applied Biosystems ) with the standard sybr green detection programme . A melting curve was produced at the end of every experiment to ensure that only single products were formed . Gene expression was determined using a version of the comparative threshold cycle ( Ct ) method . The average amplification efficiency of each target was determined using LinReg [64] and samples were normalised to 18S rRNA or ACT2 . Primers for qRT-PCR are described in Table S2 . ERF109 and ERF018 probe templates for Digoxigenin-labelling of mRNA were generated by PCR amplification and subsequent cloning of products into pENTR-D-topo using primers ( caccaacagagtcgcaaga and catgctttcttgttcttgttc for ERF109; caccaattcaaccaaaccgaat and ccagatttctccatgactcca for ERF018 ) . The resulting plasmids were used with M13 forward and reverse primers to generate a template for antisense probes , and sense probe control templates were PCR amplified with a forward primer containing a T7 promoter site ( taatacgactcactatagggatgcattatcctaac for ERF109; taatacgactcactatagggatggtgaagcaagcg for ERF018 ) . Reverse primers were as above . CLE41 positive control and methods for probe labelling and in situ hybridization were as used in [7] , and based on the method described in [65] . Analysis of vasculature tissue in thin sections , was carried out as described previously [66] . For hand cut sections , tissue was stained with either aqueous 0 . 02% Toluidine Blue or 0 . 05M Aniline blue in 100 mM Phosphate buffer ( pH 7 . 2 ) . Cell counts were carried out using thin sections of 10 week old stems on ≥10 biological replicates . Cells included were protoxylem ( marking the inner part of the bundle ) , phloem cap cells ( marking the outer part of the bundle ) and all vascular cell types between . An area of secondary growth is reported to be present up to 2 . 4 mm above the upper rosette leaf [39] . Consequently , sections were taken 10 mm above the upper rosette leaf to avoid the secondary growth region . Statistical analysis ( ANOVA ) was carried using SPSS statistical analysis software using an LSD post-hoc test . AGI accession numbers for the genes used in this study are as follows: At3g24770 ( CLE41 ) , At5g61480 ( PXY ) , At4g34410 ( ERF109 ) , At1g28370 ( ERF11 ) , At5g61600 ( ERF104 ) , and At1g74930 ( ERF018 ) , At4g17500 ( AtERF1 ) , At5g47220 ( ERF2 ) , At5g47230 ( ERF5 ) , At4g17490 ( ERF6 ) , At4g11280 ( ACS6 ) , At1g54490 ( EIN5 ) , At3g15770 ( ETO1 ) , At5g65800 ( ETO2 ) , At1g66340 ( ETR1 ) , At5g03280 ( EIN2 ) , At1g46480 ( WOX4 ) .	Plants transport water and nutrients throughout their bodies using a specialised vascular system . Vascular tissue is also responsible for providing structural support to plants; for example , wood is made up of specialised vascular cells . Consequently , the vascular system constitutes the majority of plant biomass . Chemicals from plant biomass could be used to make the next generation of biofuels in order to reduce dependence on fossil fuels . Vascular tissue is derived from a group of dividing cells present in a structure called the procambium , but mechanisms controlling cell division in this structure remain poorly understood . Understanding the events that occur in the procambium may help us to understand how we can best utilise plants for increased plant biomass , for example , for biofuel and wood production . We have identified a number of genes that regulate cell division in the procambium that are controlled by the gaseous plant hormone ethylene . We show that ethylene signalling , in turn , interacts with PXY , a gene encoding a signalling component that also controls vascular cell division . Our results demonstrate that the interaction between ethylene and PXY signalling is responsible for maintaining the plant vascular system .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "plant", "science", "genetic", "mutation", "plant", "growth", "and", "development", "gene", "expression", "genetics", "plant", "genetics", "plant", "biology", "biology", "genetics", "and", "genomics", "gene", "function" ]	2012	Plant Vascular Cell Division Is Maintained by an Interaction between PXY and Ethylene Signalling
During the replication cycle of double-stranded ( ds ) RNA viruses , the viral RNA-dependent RNA polymerase ( RdRP ) replicates and transcribes the viral genome from within the viral capsid . How the RdRP molecules are packaged within the virion and how they function within the confines of an intact capsid are intriguing questions with answers that most likely vary across the different dsRNA virus families . In this study , we have determined a 2 . 4 Å resolution structure of an RdRP from the human picobirnavirus ( hPBV ) . In addition to the conserved polymerase fold , the hPBV RdRP possesses a highly flexible 24 amino acid loop structure located near the C-terminus of the protein that is inserted into its active site . In vitro RNA polymerization assays and site-directed mutagenesis showed that: ( 1 ) the hPBV RdRP is fully active using both ssRNA and dsRNA templates; ( 2 ) the insertion loop likely functions as an assembly platform for the priming nucleotide to allow de novo initiation; ( 3 ) RNA transcription by the hPBV RdRP proceeds in a semi-conservative manner; and ( 4 ) the preference of virus-specific RNA during transcription is dictated by the lower melting temperature associated with the terminal sequences . Co-expression of the hPBV RdRP and the capsid protein ( CP ) indicated that , under the conditions used , the RdRP could not be incorporated into the recombinant capsids in the absence of the viral genome . Additionally , the hPBV RdRP exhibited higher affinity towards the conserved 5’-terminal sequence of the viral RNA , suggesting that the RdRP molecules may be encapsidated through their specific binding to the viral RNAs during assembly . Double-stranded ( ds ) RNA viruses are a diverse group of viruses that vary widely in host range ( humans , animals , plants , fungi , and bacteria ) , genome segment number ( one to twelve ) , and in the number of capsid layers , with many of them considered important pathogens of either agriculture or human health . A common feature of dsRNA viruses is that their capsid associated polymerase performs both of its functions , namely replicating as well as transcribing the viral genome , from within the confines of the virus capsid . This sequestration of the polymerase and the dsRNA genome prevents the activation of the host’s RNA induced antiviral response [1] . During the viral replication cycle , dsRNA viruses have been shown to encapsidate up to twelve RNA-dependent RNA polymerase ( RdRP ) molecules in each virus particle [2–5] To date , several different mechanisms of incorporating the RdRP molecules into the capsid have been identified . Those that possess multi-layered capsids , such as the bacteriophage ϕ6 , rotavirus , and reovirus , as well as the single-layered capsids of cypoviruses have been shown to attach their polymerase molecules to the inner surface of the capsid through direct protein-protein interactions [6–10] , suggesting that non-covalent protein-protein interaction plays an important role in RdRP incorporation . A number of single-shelled dsRNA viruses , such as the yeast L-A virus , express their polymerase as a capsid protein ( CP ) -RdRP ( gag-pol ) fusion protein , which is then incorporated into viral particles as a minor structural component during capsid assembly [11 , 12] . With few exceptions , most notably the polymerase of bacteriophage ϕ6 [13] , polymerases from dsRNA viruses are not fully active when their respective capsid proteins are not present [14–16] . It has been proposed that the dependence of polymerase activity on the presence of capsid proteins may help to ensure that dsRNA products are preferentially produced only within a capsid enclosure [17] . The crystal structures of the RdRPs from several dsRNA viruses ( i . e . ϕ6 , reovirus , rotavirus , and birnavirus ) have been determined , and all have been found to contain a core polymerase domain with a right-hand shape [18–21] . In reovirus and rotavirus polymerases , which catalyze conservative RNA transcription , possess elaborate N- and C-terminal domains that interact with the core polymerase domain , thus creating a cage-like structure with four channels leading in and out of the active site at the center of the molecule [22] . These polymerases also possess an mRNA cap binding site that may facilitate the initiation of viral RNA transcription [19 , 21] . In contrast , the ϕ6 and birnavirus polymerases , which produce RNA transcripts in a semi-conservative manner , are relatively smaller in size with a structure containing only three active site channels . Distinct structural features have been identified in the RdRPs of ϕ6 , reovirus and rotavirus that function as a structural platform for the binding of a single priming nucleotide to allow for de novo initiation of RNA synthesis [18 , 19 , 21 , 23 , 24] . Picobirnaviruses ( PBV ) are small , non-enveloped dsRNA viruses infecting a wide range of mammalian and avian species [25–29] . Human PBV ( hPBV ) has been identified on almost every continent , and has been associated with acute gastroenteritis primarily in children and people that are immunocompromised [30 , 31] . It has a bi-segmented genome with the genome segment two ( PBV2 ) encoding the viral RdRP and the genome segment one ( PBV1 ) encoding the CP and a protein of unknown function [32] . The crystal structure of a rabbit PBV virus-like particle ( VLP ) shows that the capsid possesses T = 1 icosahedral symmetry in which the asymmetric unit is a dimer . Such an organization is sometimes referred to as “T = 2” structure and is unique for dsRNA viruses [33] . However , the overall capsid organization of PBV appears to be somewhat different from the organization of the larger dsRNA viruses like reoviruses [34] . Instead of having side-by-side CP dimers clustering around the 5-fold symmetry axes ( i . e . CP decamers ) , the PBV capsid is made of diamond-shaped dimers of dimers ( i . e . CP tetramers ) . Such an unusual capsid organization has also been observed in fungal infecting partitiviruses , another family of dsRNA viruses with a bi-segmented genome [35 , 36] . The CPs for PBV and partitiviruses are small in size , with a structural fold that somewhat differs from those found in larger dsRNA viruses such as reoviruses and rotaviruses . Interestingly , the major capsid protein P1 of cystoviruses likely also forms tetramers , but the geometric shape of such tetramers and the structural fold of the P1 are somewhat different from those of PBV and partivirus CPs [37 , 38] . To elucidate the mechanisms of RNA replication , transcription , and RdRP encapsidation by this group of largely uncharacterized , small dsRNA viruses , we have determined the structure of the hPBV RdRP and systematically characterized its biochemical and enzymatic activities . The hPBV RdRP possesses a canonical polymerase fold with a 24 amino acids ( aa ) long C-terminal insertion loop structure that partially occupies the active site of the polymerase . A hPBV RdRP lacking this insertion loop , ΔLOOP , was subsequently generated to determine the functional role of this structure . Both the wild-type ( WT ) and ΔLOOP RdRPs are capable of RNA synthesis using both homologous and heterologous , single- and double-stranded RNA templates in the absence of the CP . However , while the WT RdRP utilizes a de novo initiation mechanism for RNA synthesis , the ΔLOOP could only initiate RNA replication through back-priming , suggesting that the insertion loop serves as a platform for initiation . For transcription the hPBV polymerase uses a semi-conservative mechanism in which the positive-strand of the template dsRNA is dislodged from the duplex on the RdRP surface as the negative-strands enters into the template tunnel and the newly produced transcript forms a duplex with the negative-sense RNA strand . We also demonstrated terminal nucleotidyl transferase ( TNTase ) activity for hPBV polymerase being the second polymerase among dsRNA viruses reported with TNTase activity . Co-expression of the hPBV RdRP and CP resulted in the formation of ~35 nm VLPs that were incapable of sequestering the RdRP molecules . Results from gel shift assays indicate that the hPBV RdRP has a strong preference for the 5’-terminal untranslated region of the positive-sense genomic RNA ( i . e . 5’ ( + ) UTR ) . Our results thus suggest that PBV most likely has its RdRP molecules incorporated into viral particles through direct interactions with the genomic RNAs , which are selectively packaged through specific interactions with the viral CP . Recombinant hPBV RdRP ( strain Hy005102 ) was overexpressed in Escherichia coli as a soluble protein . The purified hPBV RdRP ( 534 aa , ~62 kDa with a N-terminal 6xHis tag ) exists in solution as a monomer , based on its gel filtration chromatography elution profile ( S1 Fig ) . hPBV RdRP was crystallized in the space group of P21 with two molecules in each asymmetric unit . The crystal structure was solved to 2 . 4 Å resolution by single-wavelength anomalous dispersion ( SAD ) using Se-Met derivatized crystals ( Table 1 ) . In the final model of the hPBV RdRP , one molecule is comprised of residues 1–494 , 500–511 , and 518–534 ( Fig 1A ) , while the other molecule contains residues 1–495 and 520–534 . The hPBV RdRP has an overall oval shape and is ~50 x 60 x 60 Å3 in size ( Figs 1 and 2 ) . There are a total of 24 α-helices and 14 β-strands in each molecule . The polypeptide can be divided into three domains based on their function: an N-terminal domain ( aa 1–84 ) , a core polymerase domain ( aa 85–470 ) , and a C-terminal domain ( aa 471–534 ) ( Fig 1A and 1C; S2 Fig ) . The polymerase domain , which is structurally highly conserved amongst RNA viruses [39] , has a right-hand configuration with three subdomains: the fingers ( aa 85–230 , 268–324 ) , palm ( aa 231–267 , 325–414 ) , and thumb ( aa 415–470 ) ( Fig 1A and 1C ) . The palm subdomain hosts the three key aspartic acid residues , D261 , D359 , and D360 , of the active site ( Fig 1A ) . The hPBV core polymerase domain is rather compact with only 386 aa in total ( Fig 1B ) . The palm subdomain , which hosts the catalytic active site , is composed of five α-helices and six β-strands ( i . e . α11–13 , 16 , 17 and β6 , 9–13 ) and contains the polymerase motifs A-E that are conserved among all RdRPs ( Fig 1B and 1C; red ) [40] . The most noticeable structural feature of the palm is a central , four-stranded β-sheet consisting of a β-hairpin ( i . e . β9 , β10 ) and two anti-parallel β-strands ( i . e . β6 , β11 ) . The polymerase motif C , which contains the highly conserved “-GDD-” sequence , is mapped to the β-hairpin . The two other β-strands ( β6 , β11 ) of the central β-sheet contain motifs A and D , respectively . The motif A has the conserved sequence of “DXXXXD” whereas motif D is more variable in sequence . The first aspartate from motif A ( i . e . D261 ) and the two aspartates from motif C ( i . e . D359 and D360 ) constitute the active site as they help to coordinate two divalent metal ions for charge relay and intermediate stabilization during catalysis . The motif D mediates the binding of the incoming nucleotide substrate and plays an important role in determining the efficiency and fidelity of nucleotide addition [41] . Motif B , which folds into a strand-turn-helix structure at the interface between the fingers and palm subdomains , has been found to interact with the RNA template to guide it into the active site of the polymerase [42] . Motif E folds into a β-hairpin at the interface between the palm and the thumb subdomains , and forms a part of the “primer grip” as discussed below [40] . Close inspection shows that the fingers subdomain of the hPBV RdRP is composed of eight α-helices and seven β-strands ( Fig 1A and 1C; blue ) . At the top of the fingers subdomain is a twisted , four-stranded β-sheet ( β4 , β5 , β7 , and β8 ) that forms the fingertip structure with an extended loop , which contains the rNTP binding sequence ( residues 182–199 ) denoted as the RdRP motif F ( Fig 1B and 1C ) [40] . Structural studies on ϕ6 and hepatitis C virus ( HCV ) RdRPs have revealed that the basic residues in the rNTP binding loop interact with the phosphates of the incoming nucleotide [18 , 43] . The fingers subdomain also contains the RdRP motif G [44] that is located near a three-stranded antiparallel β-sheet ( β1 , β2 , and β3; Fig 1B and 1C ) . The structures of the ϕ6 and reovirus RdRPs have revealed that the residues of motif G interact with the entering RNA template [18 , 21] . Situated at the other side of the polymerase palm across from the fingers subdomain is the thumb subdomain , which is comprised of one β-strand ( β14 ) followed by three α-helices ( α18 - α20; Fig 1A and 1C; green ) . The initial β14-strand forms a part of the “primer grip” along with a β-hairpin ( β12 and β13 ) from the palm subdomain ( Fig 1A ) . The “primer grip” motif is commonly observed in viral RdRPs and has been shown to interact with the nascent/primer strand during RNA synthesis [40] . The hPBV N-terminal domain is made of the first 84 residues of the polypeptide and contains four α-helices ( α1–4 ) . With an overall L-shape , the N-terminal domain wraps around the fingers and thumb subdomains with its long and short arms , respectively ( Fig 1A; yellow ) . This interaction allows the RdRP to maintain its active site in a closed conformation despite that there is very little direct contact between the hPBV fingertip and the thumb subdomain . The N-terminal domain in the RdRPs from infectious bursal disease virus , reovirus , and the rabbit hemorrhagic disease viruses also helps to encircle the polymerase active site although the size of these N-terminal domains can vary substantially [20 , 21 , 45] . The C-terminal domain of hPBV is rather short with four α-helices . It lies adjacent to the thumb subdomain at the front end of the polymerase palm ( Fig 1A; magenta ) . The hPBV RdRP contains three channels leading to the active site of the protein that are believed to allow for dsRNA product export , template entry and NTP uptake ( Fig 2 ) . The dsRNA product channel , which locates in the front of the molecule , is the largest of all three with a diameter of ~18–20 Å , comparable to the diameter of a dsRNA helix ( Fig 2A–2C ) . Both the template and NTP channels are located near the interface between the fingers and thumb subdomains , and are heavily lined with positively charged residues ( Fig 2B , 2C and 2D ) . The distance from the surface of the RdRP to the active site along the putative template entry channel could be spanned by 5 to 6 nucleotides as observed in the ϕ6 RdRP [46] . A patch of positively charged residues is found near the putative template entry channel ( Fig 2C , dotted oval ) . A positively charged plough has been previously noted in the structure of ϕ6 RdRP , and is believed to play a role in separating the two strands of a dsRNA molecule allowing the template RNA to enter the template entry channel , while the non-template RNA slides over the positively charged patch and is directed away from the RdRP [18] . Based upon a pairwise comparison using the program Dali [47] , the structure of the hPBV RdRP closely resembles that of the RdRPs from the members of the Caliciviridae ( i . e . Z = 26 . 7 for the rabbit hemorrhagic disease virus and Z = 26 . 5 for the Norwalk virus ) [45 , 48] , Flaviviridae ( i . e . Z = 26 . 4 for the HCV and Z = 24 . 2 for the bovine viral diarrhea virus ( BVDV ) ) [43 , 49–51] and Picornaviridae ( i . e . Z = 25 . 3 for the poliovirus ) [52] families , suggesting a potential evolutionary relationship between the RdRP of PBV and the RdRPs of these three viral families having positive-sense ssRNA genomes . By contrast , the correlation of the hPBV polymerase to the RdRPs from other dsRNA virus families appears to be more distant , with a Z = 18 . 8 for phage ϕ6 [18] , Z = 12 . 8 for rotavirus [19] , Z = 12 . 1 for infectious bursal disease virus [20] , and Z = 9 . 3 for reovirus [21] . The hPBV RdRP possesses a highly flexible , 24-aa insertion loop structure ( aa 495–518 ) formed by an internal sequence from the C-terminal domain . This loop structure , which is associated with higher than average temperature factor values , extends from the C-terminal domain towards the catalytic site ( Fig 1A ) . The location of the insertion loop resembles the structure that functions as the initiation platform of the bacteriophage ϕ6 RdRP [23 , 24] . The structure of the ϕ6 RdRP in complex with an oligonucleotide template ( PDB ID 1HI0 ) was superimposed onto the hPBV RdRP structure in order to model RNA binding by the hPBV RdRP . In this model , the 3’-end of the template collides with the hPBV insertion loop structure , suggesting that the insertion loop in its current position would sterically prevent the template from binding to the active site of the RdRP as expected in the assembly of a productive initiation complex ( Fig 1A , right ) . Therefore , the hPBV insertion loop structure must undergo a significant conformational change in order for RNA binding and replication to take place . Similar structural rearrangement is probably required for dsRNA egress during both transcription and genome replication . Interestingly , such structural arrangements were indeed observed in the β-hairpin structure that serves as an initiation platform for the HCV RdRP following RNA binding [53] . In the ϕ6 polymerase , the loop that serves as the initiation platform remains in place during the assembly of the initiation complex , but undergoes a conformational change at the onset of the elongation process [54] . Based on our structural modeling and comparison to both the ϕ6 and HCV polymerases , we speculate that the insertion loop of the hPBV RdRP may have two possible functions . First , the loop may play an important role during the de novo initiation of RNA synthesis . Like in the case of the ϕ6 , HCV and several other flavivirus polymerases , the loop could function as a platform to support the assembly of an initiation complex using a single nucleotide as a primer . Another attractive hypothesis is that the insertion loop may function as a gatekeeper to regulate RNA synthesis during the PBV virion assembly . The insertion loop in its observed structural form would inhibit an RNA template from binding , but upon the RdRP binding with the CP ( i . e . in an assembled capsid ) , the insertion loop would adopt an alternative conformation to allow efficient RNA synthesis in the fully/partially assembled hPBV particles . To determine the exact biological function of the insertion loop structure , a hPBV RdRP lacking the loop structure ( i . e . ΔLOOP ) was synthesized and expressed . Gel filtration chromatography showed that ΔLOOP was eluted at a similar position as the WT protein ( S1 Fig ) . The ΔLOOP RdRP was crystallized and its structure solved by molecular replacement using the WT RdRP structure as the phasing model ( Fig 3A , Table 1 ) . The overall structure of ΔLOOP appears to be essentially the same as the WT RdRP ( root-mean-square deviation in distance = 0 . 4 Å for 3275 common atoms ) , except for the deleted insertion loop which became unstructured in the ΔLOOP RdRP . Given this structural information , we are confident that the removal of the insertion loop should not affect the overall folding of the polymerase , and that the ΔLOOP RdRP should provide an excellent tool to study the function of the insertion loop structure using in vitro assays . Polymerase activity assays were performed to determine if the WT and ΔLOOP RdRPs could synthesize dsRNA from an ssRNA template . Both RdRPs were found to replicate the positive- and negative-strands of the PBV genome segment 2 ( PBV2+ and PBV2- , respectively ) as well as ϕ6-specific ssRNA template s+ ( i . e . the positive-strand of the small , S , genome segment ) with similar efficiency ( Fig 3B ) . However , the polymerase encountered some processivity issues , as it produced also dsRNA products which were shorter than the expected full length dsRNAs ( Fig 3B ) . Further enzymatic assays showed that the WT RdRP is able to use dsRNA templates to carry out transcription but the transcription activity of the ΔLOOP RdRP was compromised ( Fig 3B ) . Transcription activity was observed whenever homologous ( i . e . PBV2 ) or heterologous dsRNAs ( i . e . ϕ6 genomic RNA composed of segments S , M and L ) were used as templates ( Fig 3B , S1 Table ) . This indicates that the hPBV RdRP is enzymatically active , can efficiently replicate and transcribe both homologous and heterologous templates , and does not require the presence of the viral CP as a cofactor for RNA synthesis . Our findings thus rule out the scenario where the loop structure functions as a regulatory element to prevent premature dsRNA synthesis by the capsid-free polymerase . Terminal nucleotidyl transferase ( TNTase ) activity was observed for both the WT and ΔLOOP RdRPs ( Fig 3C and 3D ) . TNTase activity involves the addition of one or several nucleotide ( s ) to the 3’-end of a nucleic acid molecule ( Fig 3C ) . While there was no clear preference for the hPBV-specific RNA substrate , both the WT and ΔLOOP RdRPs showed a strong preference for ssRNA over dsRNA ( Fig 3D ) . Additionally , it was noted that the removal of the insertion loop significantly increased the TNTase activity of the RdRP ( Fig 3D ) . Given our TNTase activity data and the fact that the insertion loop is located near the C-terminus of the hPBV RdRP , we propose that the removal of the insertion loop structure leaves the dsRNA exit channel of the protein permanently open ( Fig 2B ) , thus allowing the 3’-end of the RNA molecules to reach the active site for nucleotidyl addition in an orientation that is compatible for nucleotide addition . To rule out any histidine-tag ( His-tag ) associated artifacts , three versions of the hPBV RdRP , an N-terminally and C-terminally His-tagged as well as a non-tagged RdRP , were tested for RNA synthesis activity using the PBV2+ and PBV2- ssRNA templates as well as the PBV2 dsRNA template ( S3 Fig ) . Our results indicated that there was no detectable difference in the enzymatic activity between these three versions of the polymerase . Therefore , all of the polymerase activity assays described in this paper has been performed using the N-terminally His-tagged protein . To evaluate the potential role of the insertion loop in the initiation of dsRNA synthesis ( i . e . replication ) , aliquots of the dsRNA products were heat denatured before being analyzed by electrophoresis in a native agarose gel ( Fig 4A and 4B ) . Due to the heat denaturation , the dsRNA products generated by the WT RdRP were converted to ssRNA , while the dsRNA products of the ΔLOOP RdRP retained the same mobility as the original dsRNA ( Fig 4A and 4B ) . This result indicates that back-priming was taking place during dsRNA synthesis by the ΔLOOP RdRP , thus producing a product that was covalently linked to the ssRNA template ( Fig 4A , right panel and b ) . Additionally , isotope incorporation from ( γ-32P ) GTP into the dsRNA product was only detected for reactions containing the WT RdRP , further indicating that this protein utilizes the de novo initiation mechanism ( Fig 4C ) . Taken together , these results show that the insertion loop structure of the hPBV RdRP can effectively block template back-priming and facilitates initiation via a primer-independent mechanism , possibly by providing a docking site for the 3’-end of the RNA template and a binding site for the priming nucleotide . This finding is consistent with the results obtained for the ϕ6 RdRP whenever its equivalent stabilizing platform was removed [23 , 24] . There are two major ways for the transcription of the dsRNA genome: conservative and semi-conservative transcription mechanism ( Fig 5A ) . The isotope labeling of the dsRNA molecules as an outcome of the transcription reaction indicated that the hPBV polymerase uses a semi-conservative transcription mechanism ( Figs 3B , 5A and 5B ) . The ability of the hPBV RdRP to incorporate radioactivity from [γ-32P]-labeled GTP into the dsRNA product also confirmed that the observation of radiolabeled dsRNAs was due to de novo RNA synthesis instead of terminal nucleotidyl transfer catalyzed by the polymerase ( Fig 5B , right panel ) . To get further support for the semi-conservative mechanism a time-course study using the ϕ6 genomic dsRNA as a template was performed ( S4 Fig ) . Indeed , 33P-label was first incorporated in dsRNA molecules before any labeled ssRNA could be observed ( S4 Fig ) , confirming that semi-conservative transcription was taken place ( not replication of ssRNA transcripts produced by conservative mechanism ) . The ϕ6 L segment was transcribed at the highest efficiency , resulting in a strong dsRNA band that started to appear at around 30 minutes with increasing intensity till 120 minutes after the experiment began ( S4 Fig ) . Considering that the L segment of ϕ6 is ~6 . 4kb long , we estimate that the rate of transcription by the hPBV RdRP is ~210 bases per minute . The ( - ) strands of the two hPBV genome segments both start with 3’-CAU ( S1 Table ) , so we set forth to test the transcription specificity of the hPBV RdRP using dsRNA template with different terminal sequences . Coincidentally , the ( - ) strand of the ϕ6 L segment also starts with 3’-CAU , while the ( - ) strands of the ϕ6 S and M segments begin with 3’-CCU ( S1 Table ) . We found that , under the applied reaction conditions , the hPBV RdRP has a preference for dsRNA templates in which the ( - ) strand starts with 3’-CAU over those starting with 3’-CCU ( S4 Fig , Figs 3B and 5B ) . The preference for the ϕ6 L segment over the M and S segments was lost when the terminal sequence of L was mutated to 3’-CCU ( S5 Fig ) . On the other hand , when an opposite change was introduced in the ( - ) strands of the ϕ6 genome segments M and S ( 3’-CCU to 3’-CAU ) , the amount of product synthesized by the hPBV RdRP increased considerably ( S5 Fig ) . The template preference of the hPBV RdRP was also evident when the usage of the PBV2 and ϕ6 S or ϕ6 L dsRNAs was compared in a competition experiment or side-by-side ( S6B Fig ) . We then tested the ability of hPBV RdRP to transcribe full-length PBV2 and its two deletion mutants , one with the first 33 nucleotides removed and the other with the first 645 nucleotides removed ( Fig 5C ) . As a result of the deletions , the 3’-ends of the ( - ) strands of the two truncation mutants , Δ1–33 and Δ1–645 , became 3’-CCU and 3’-CCC , respectively . Transcription assays showed that out of the three dsRNA templates , only the WT PBV2 , which has a 3’-CAU termini at the ( - ) strand , was efficiently transcribed ( Fig 5C ) . Taking together the results with the PBV2- and ϕ6-specific dsRNA constructs ( S5 Fig and Fig 5C ) sharing identical 3’-terminal sequences in ( + ) strands ( S1 Table ) but variable in ( - ) strands , it appears that PBV RdRP has strong preference on dsRNA templates having 3’-CAU . This data also suggest that the RdRP predominantly produces ( + ) strands of the PBV2 and ϕ6 genome segments during transcription reaction . There are two possible explanations for the stronger transcription activity exhibited by the hPBV RdRP towards native PBV2 ( - ) strand sequence ending with 3’-CAU . First , the lower melting temperature associated with the 3’-CAU sequence likely facilitates the initiation of RNA transcription . Second , it is possible that the 3’-CAU sequence makes specific interaction with the hPBV RdRP , which would facilitate the binding of the viral RNA template and thus helps to enhance its transcription activity . We have not observed any preference of the hPBV RdRP towards ssRNA template ( Figs 3B and S6A ) . Indeed , the fact that virus-specific ssRNA was not replicated to a higher level when mixed with non-specific ssRNA in a competition experiment argues against the second hypothesis . Therefore , we believe that the preference for virus-specific dsRNA templates by the hPBV RdRP during transcription is largely due to the lower melting temperature of the template terminal sequence . Upon finding that the hPBV RdRP was capable of synthesizing dsRNA and ssRNA in the absence of the CP ( Fig 3B ) , the hPBV RdRP and CP were co-expressed to determine if the encapsidation of the RdRP is mediated through direct protein-protein interactions with the CP , as has been previously observed for the majority of dsRNA viruses characterized so far [6 , 8 , 9] . Overexpression of the hPBV CP in E . coli resulted in the spontaneous formation of VLPs . Such VLPs could be purified by ultracentrifugation using a CsCl gradient and have a density of ~1 . 3 g/ml according to their migration behavior . Transmission electron microscopy ( TEM ) images of the hPBV VLPs determined that they have a diameter of ~35 nm ( Fig 6A ) , similar to the size of the rabbit PBV VLPs previously reported [34] . While the N-terminus of the rabbit PBV CP is proteolytically removed by self-cleavage after particle assembly , this does not seem to be the case for the hPBV CP . We found that the hPBV CP without the N-terminal peptide ( Δ45 ) was also capable of self-assembly , and that the N-terminally truncated CP migrated faster than the full-length protein in a reducing sodium dodecyl sulfate polyacrylamide gel electrophoresis ( SDS-PAGE ) ( Fig 6A and 6B ) . The hPBV RdRP and the CP were co-expressed in E . coli using two expression vectors with different antibiotic selectors . Any unpackaged His-tagged RdRP molecules that may have been present in the clarified cell lysate were removed by Ni-NTA affinity chromatography . After ultracentrifugation , the VLP fraction was collected , denatured , and a Western blot was performed to test for the presence of the RdRP utilizing an antibody directed against the 6xHis-tag located at the N-terminus of the protein ( Fig 6C ) . While the RdRP could be clearly detected in the clarified lysate , no RdRP could be detected in the sample of the isolated VLPs . Therefore , our results suggest that direct interactions between the hPBV RdRP and CP are weak . We hypothesize that viral genomic material must be present for RdRP incorporation to occur and that the interplay between the CP , RdRP , and the viral genome is needed for the encapsidation of the RdRP during virion assembly . The fact that the constructs used to express the CP and RdRP contained only the protein-coding sequence of the two open reading frames suggests that the region of the hPBV genome required for RdRP encapsidation is potentially located in the untranslated regions ( UTRs ) of the genome segments . Gel shift assays were conducted to further examine the interaction between the RdRP and the hPBV genome . Three 20-nt long RNA oligonucleotides were probed , one containing a non-specific CA-repeat and the other two bearing the terminal sequences of the 5’- and 3’-UTRs of the ( + ) strand of the hPBV genome segment 2 ( i . e . 5’- ( + ) UTR and 3’- ( + ) UTR ) . It was determined that the WT RdRP has approximately a 10-fold higher affinity for the first 20 nucleotides of the 5’- ( + ) UTR of the hPBV genome as compared to the last 20 nucleotides of the 3’- ( + ) UTR or to a nonsensical CA-repeat ( Fig 7A ) . Additionally , deleting the insertion loop appeared to have a minimal effect on the overall RNA binding of the RdRP ( Fig 7B ) . Notably , both segments of the hPBV genome have an AU-rich ( ~80% for nucleotides 1 to 20 ) sequence at the 5’-end of their ( + ) strands that begins with a conserved 5 nucleotide motif 5’-GUAAA- . The AU-rich terminal sequence is predicted to form an RNA stem-loop structure ( S7 Fig ) according to the program ViennaRNA Package 2 . 0 [55] . To test the binding affinity of the hPBV RdRP towards authentic viral RNAs , we synthesized by in vitro transcription the full-length PBV2+ ssRNA and a truncated version PBV2+ ( Δ1–645 ) ssRNA with the first 645 nucleotides removed . Results from our gel shift assays show that the binding affinity of hPBV RdRP for full-length PBV2+ is ~10 times stronger than PBV2+ ( Δ1–645 ) , thus confirming that the terminal sequence at the 5’-end of the PBV2+ binds specifically to the polymerase ( Fig 7C ) . The same trend was observed for the ΔLOOP RdRP , suggesting that the insertion loop is not directly involved in the 5’- ( + ) UTR binding ( Fig 7D ) . The high level of sequence conservation among RdRPs from various mammalian and avian PBV strains indicates similar three-dimensional structures ( S2A Fig ) . Five of the six variable regions are located in either the N-terminal domain or the fingers subdomain , and all are mapped to the surface of the protein ( S2B Fig ) . The core polymerase domain of hPBV RdRP closely resembles those from the members of the Caliciviridae , Flaviviridae , and Picornaviridae families all having ( + ) sense ssRNA genomes . An interesting structural feature of this RdRP is the presence of a 24-aa loop structure that extends from near the C-terminus of the protein to just above the active site ( Fig 1A ) . Through site-directed mutagenesis , we determined that the loop structure most likely functions as a priming platform to support the binding of a single priming nucleotide ( Fig 4 ) . Loop structures that perform a similar function have been observed in the viral RdRPs of ϕ6 , reovirus , rotavirus , and HCV [18 , 19 , 21 , 23 , 24 , 53] . In four-tunnel RdRPs such as those from reovirus and rotavirus , the internal priming loops are formed by sequences located between the fingers and palm domain . However , in three-tunnel polymerases that catalyze semi-conservative RNA synthesis ( i . e . ϕ6 and HCV ) , the priming loops appear to be extended structures from the C-terminal domain . The position of the insertion loop in the hPBV apo structure would prevent a template from reaching the active site due to steric hindrance ( Fig 1A , right panel ) . Therefore , we expect the insertion loop to undergo a significant conformational change in order to accommodate an RNA template , similar to the conformational change observed for the HCV priming loop upon template binding [53] . While the first half of the insertion loop sequence is highly conserved , the other half is somewhat variable ( S2B Fig ) . We speculate that the conserved region , including a strictly conserved tyrosine , may interact with the priming nucleotide and/or template to support initiation . Structural elements that support priming by a single nucleotide are also known to prevent back-priming by spatially restricting access to the active site . Back-priming occurs during RNA synthesis when the 3’-end of the template strand loops back to form a hairpin like structure that is then extended by the RdRP [23] ( Fig 4A , right panel ) . This results in the daughter strand being covalently linked to the initial template preventing further replication of the back-primed RNA . This phenomenon has been observed in vitro for HCV , BVDV , and a ϕ6 RdRP lacking the priming loop [23 , 24 , 56–59] . Likewise , our results indicate that the insertion loop structure from the hPBV RdRP can effectively prevent back-priming during dsRNA synthesis ( Fig 4B ) and to support the de novo initiation ( Fig 4C ) , consistent with the expected functionality for the priming loop based on previous observations in other RdRPs . In this paper we have systematically characterized the replication and transcription activity of the hPBV RdRP as a paradigm for the Picobirnaviridae family . In vitro , the hPBV RdRP is able to catalyze RNA synthesis using both ssRNA and dsRNA templates in the absence of the viral CP ( Figs 3 and 5 ) . The enhanced transcription activity observed for the WT protein using hPBV-specific dsRNA or templates harboring homologous 5’ terminal sequences ( Fig 5C and S6 Fig ) can be explained by the lower melting temperature associated with the terminal sequences . Alternatively , the enhanced transcriptional activity of the hPBV specific dsRNA may be explained by base-specific interaction between the template and the RdRP itself , but we consider it unlikely because enhanced replication activities were not observed for hPBV-specific ssRNA ( Fig 3B ) . Interestingly , the identity of the second nucleotide of the template RNA also regulates the transcription activity of the phage ϕ6 RdRP [60] . hPBV RdRP appears to transcribe dsRNA templates in a semi-conservative fashion ( Fig 5 ) . Results from our time-course study show that nucleotides labeled with α-33P were first incorporated into dsRNA , indicating that the newly synthesized RNA formed a duplex RNA with its template RNA ( S4 Fig; Fig 5A , right panel ) . Except for the RdRPs of the members of Reoviridae family , semi-conservative transcription is reported for most of the dsRNA virus polymerases characterized to-date , including RdRPs from the members of Partitiviridae [61] , Birnaviridae [62] , and Cystoviridae [63] families . The rate of transcription by hPBV RdRP is ~210 bases per minute . Although we have not experimentally confirmed that the ( - ) strand is used as a template for transcription , our results with the terminally mutated PBV- and ϕ6-specific dsRNAs strongly favor a scenario in which the sequence at the 3’-end of the ( - ) strand rather than ( + ) strand determines the transcription efficiency ( Figs 5C and S5 , S1 Table ) . This likely reflects the substantially lower melting temperature associated with the 3’-end of the ( - ) strands compared to that of the ( + ) strands ( S1 Table ) . We also observed TNTase activity for both the WT and ΔLOOP hPBV RdRPs ( Fig 3C and 3D ) . TNTase activity involves the addition of one or several , non-templated nucleotide ( s ) to the 3’-end of an RNA molecule , and has previously been observed for a number of RdRPs including those from HCV , BVDV , norovirus , poliovirus , and ϕ6 [64–67] . For many viral RdRPs the biological implication of the TNTase activity is not yet clear . Template-independent TNTase activity is probably used by the RdRP as a mechanism to terminate the synthesis of nascent RNA strands , which would acquire one or more extra nucleotides at the 3’-end [65] . Alternatively , TNTase activity may function to repair the 3’-ends of the viral genomes that have been partially degraded [66] . The results of our experiments support the notion that the RNA substrate for the TNTase activity potentially enters the active site through the product exit channel , as removing the insertion loop would leave the product exit channel of the protein permanently open , thus explaining why the ΔLOOP RdRP displayed higher TNTase activity than the WT enzyme , especially for the dsRNA substrates ( Fig 3D ) . All dsRNA viruses enclose RdRP molecules within their infectious particles . Our results indicate that the hPBV RdRP and CP do not directly interact during the capsid assembly and that the viral RdRP cannot be incorporated into the viral capsid in the absence of the viral genome ( Fig 6C ) . This finding is surprising because several other dsRNA viruses have been found to package their RdRP molecules through direct protein-protein interactions or by expressing a CP-RdRP fusion protein that is then incorporated into the viral particles as a minor structural component [6 , 8 , 9 , 11 , 12] . We propose that the PBV RdRP molecules get encapsidated as a complex with the viral genomic RNA , as we found that the hPBV polymerase preferably interacts with the 5’-end of the ( + ) strand . It is likely that the co-packaging mechanism applies not only to PBV but also to partitiviruses , another family of dsRNA viruses with a small capsid that is arranged in a manner similar to PBV [35 , 36] . Such a co-packaging model is consistent with the observation that in small dsRNA viruses , such as partitiviruses , the number of RdRP molecules packaged during assembly is similar to the number of packaged genome segments [68 , 69] . Interestingly , recent studies on cypoviruses , which are members of the family Reoviridae , show that only ten instead of twelve polymerase complexes are visible in each particle [7 , 10] , indicating that protein-RNA interactions also play an important role in genome packaging in these multi-layered dsRNA viruses . A secondary structure prediction of both segments of the hPBV genome by the program ViennaRNA Package 2 . 0 [55] predicted that both PBV ( + ) strand RNA segments begin with an AU-rich ( ~80% ) , RNA stem-loop structure ( S7 Fig ) . The low melting temperature associated with the AU-rich 5’- ( + ) UTR should facilitate the denaturation of the dsRNA genome into an ssRNA template during the initiation of the transcription . We speculate that the specific binding of the hPBV polymerase to the 5’- ( + ) UTR may play two critical roles during the viral life cycle . First , the 5’- ( + ) UTR may function as a recognition element to ensure the specific packaging of the viral RNAs during virus assembly . Second , in the mature virion , this binding could help to direct the 3'-end of the genomic ( - ) strand RNA into the template entry channel of the RdRP to initiate the transcription ( Fig 8 ) . It has been reported that polymerases from members of the Reoviridae family utilize a 5’-cap binding activity to gain access to the 3’-end of the ( - ) RNA template during transcription initiation [21] . PBV does not encode a capping enzyme , and its genomic ( + ) strand RNA is not expected to have a cap . Nevertheless , the 5’- ( + ) UTR sequence of the PBV genome could play a similar role as the reovirus mRNA cap to facilitate transcription initiation and also to ensure that only the ( - ) strand RNA template is selectively used as a template for viral mRNA production . It is also worthwhile to point out that our recombinant hPBV capsid does not undergo proteolytic cleavage as was previously reported for the rabbit PBV capsid , in which the first 65 residues of the capsid protein were removed from the assembled particles presumably due to self-cleavage [34] . Expression of the hPBV CP without the structurally flexible N-terminal peptide ( corresponding to the first 65 residues in the rabbit PBV CP ) produced recombinant capsids that were indistinguishable from the WT capsid ( Fig 6A ) . As the cleavage site residues are conserved in the human PBV CPs , it is unclear whether the lack of proteolytic cleavage is due to strain difference or perhaps the use of a different expression host ( i . e . E . coli vs . insect cells ) . Our findings lead us to propose a model for PBV assembly and genome replication in which the RdRP first binds to one of the PBV positive-sense RNAs utilizing the AU-rich sequence located at the 5'- ( + ) UTR ( Fig 8 ) . The same genome segment is also acted upon by the PBV CP that is in the process of forming a virus capsid , most likely utilizing the highly flexible N-terminal region of the protein [34] . The capsid then assembles around the two genome segments which are bound to two separate RdRP molecules . Conversion of ssRNA to dsRNA occurs either in partially or fully assembled capsids resulting in transcriptionally active particles . Within the capsid the 5’- ( + ) UTR binding to the RdRP facilitates template strand selection and recognition during repeated transcription events . Transcription of the PBV genome segments proceeds in a semi-conservative manner where the parental ( + ) strand is replaced by the newly synthesized strand ( Fig 5A , right ) . To fill in the gaps in our model , further experiments are needed to elucidate how the RdRP recognizes the 5’- ( + ) UTR and to map the molecular signals responsible for the specific packaging of the PBV genome by the viral CP . We expect that future studies of the PBV will help to define a new paradigm for the assembly and replication of small dsRNA viruses in which the viral RdRP most likely functions independently of the viral capsid . The gene encoding the RdRP of the hPBV Hy005102 strain ( 534 aa , ~62 kDa with either N-terminal or C-terminal 6xHis tag and without His-tag ) was cloned into a pET28b ( + ) -vector ( Novagen ) . The corresponding UTRs were not included in the cloned sequence . The resulting constructs were transformed into a Rosetta 2 strain of E . coli ( Novagen ) and expressed by isopropyl β-D-thiogalactopyranoside ( IPTG ) induction . The cells where then pelleted before being resuspended and sonicated in a lysis buffer composed of 50 mM Tris-HCl pH 7 . 5 , 500 mM NaCl , 10% glycerol , 5 mM imidazole , 5 mM β-mercaptoethanol , 20 μg/ml phenylmethylsulfonyl fluoride ( PMSF ) or 1 mM Pefablock , 50 μg/ml ribonuclease A , and 10 μg/ml DNase 1 . When the RdRP was purified for activity assays glycerol , β-mercaptoethanol and ribonuclease A were omitted and lysozyme was included . The lysate was clarified by centrifugation at 20 , 000 ×g for 60 minutes . The hPBV histidine tagged RdRPs were purified using a nickel-NTA ( Thermo Fischer Scientific ) , HiTrap Heparin HP , and a Superdex 200 size exclusion column ( GE Healthcare Life Sciences ) . The RdRP was eluted from the Superdex column in a buffer composed of 50 mM Tris-HCl pH 7 . 5 , 300 mM NaCl , 10% v/v glycerol , 1 mM EDTA , 1 mM NaN3 , and 5 mM β-mercaptoethanol . The purified RdRP was concentrated to 10 mg/ml for crystallization . When the RdRP was purified for the activity assays the Superdex column was omitted and the polymerase was stored in a buffer composed of 50% ( v/v ) glycerol , 50 mM Tris-HCl pH 8 . 0 , 0 . 1 mM EDTA , 0 . 1% Triton X-100 , ~150 mM NaCl at -20°C . The ΔLOOP RdRP was expressed and purified using the same protocol as the WT protein . For the untagged RdRP used for activity assays , a similar protocol was adopted except that the Ni-NTA-column was not used and HiTrap Q HP column purification step was added before the Heparin HP column purification ( GE Healthcare ) . For structure determination , the selenomethionine ( SeMet ) labeled RdRP was obtained by expressing the protein in M9 minimal media containing SeMet and a mixture of six other amino acids to prevent methionine synthesis [70] . The SeMet labeled protein was expressed and purified using the same protocol as the native protein . Crystals of the N-terminally histidine tagged hPBV RdRP were obtained by the hanging drop vapor diffusion method . The crystallization drop contained 1 . 5 μl of the RdRP solution and 0 . 5 μl of the mother liquor solution composed of 200 mM sodium acetate , 100 mM sodium cacodylate pH 6 . 5 , and 30% ( w/v ) polyethylene glycol ( PEG ) 8000 . Crystals appeared after incubation at 20°C for about 3 days and grew to full size ( 80 x 200 x 70 μm3 ) in approximately a week . The crystals were then transferred into a cryoprotectant solution ( 30% ( v/v ) glycerol , 200 mM sodium acetate , 100 mM sodium cacodylate pH 6 . 5 , 30% ( w/v ) PEG 8000 ) and flash frozen in liquid nitrogen . The crystals were sent for data collection at the Beamline 4 . 2 . 2 at the Advanced Light Source , Lawrence Berkeley National Laboratory . The data were reduced and scaled using HKL2000 ( Table 1 ) [71] . The heavy atom sites and experimental phases were determined by the AutoSol Wizard in the PHENIX software suite [72] . The protein model was built using PHENIX Autobuild , manually adjusted using COOT [73] , and refined with phenix . refine . All of the structures presented were prepared using the program PYMOL unless otherwise specified ( The PyMOL Molecular Graphics System , Version 1 . 2r3pre , Schrödinger , LLC ) . The coordinates have been deposited at the Protein Data Bank ( PDB ID 5I61 and 5I62 for the full-length and ΔLoop RdRP , respectively ) . The plasmids used for ssRNA and dsRNA production are presented in S2 Table . The full-length hPBV genome segment 2 ( hPBV2 ) in a pMA-RQ-plasmid was synthesized by Life Technologies . The ϕ6 and hPBV specific ssRNAs were prepared by in vitro transcription using the T7 RNA polymerase and complementary DNA templates amplified by polymerase chain reaction ( PCR ) . The primers used for hPBV2-specific cDNA production were hPBV2_T7_Forward ( 5’ CGCGTAATACGACTCACTATAGTAAAATTTTCGAATTTTATAATAATTAAG ) and hPBV2_Reverse ( 5’ GCAGTTGGGACTGTTAGTCCCAATG ) as well as hPBV2_Forward ( 5’ GTAAAATTTTCGAATTTTATAATAATTAAG ) and hPBV2_T7_Reverse ( 5’ CGCGTAATACGACTCACTATAGCAGTTGGGACTGTTAGTCCCAATG ) for ( + ) strand and ( - ) strand ssRNA production , respectively . For ϕ6 specific cDNA production primers T7-1 and 3’end [74] were used . For the production of the truncated PBV ( + ) strand ssRNAs the hPBV2_T7_Forward primer was replaced with PBV2_T7_5'_34 ( 5’ CGCGTAATACGACTCACTATAGGAGTTTAATAGTTTATCACAACTTAAAAGTG ) or PBV2_T7_5'_646 ( 5’ CGCGTAATACGACTCACTATAGGGTGGCGAGGCCAGGAG ) for Δ1–33 and Δ1–645 , respectively . The produced ssRNAs were purified using chloroform extraction and successive precipitations with 4 M LiCl and 0 . 3 M sodium acetate , pH 6 . 5 . The ϕ6 and hPBV specific ssRNAs were converted to dsRNA using the ϕ6 RdRP as described in [74] . The reaction mixtures were incubated for 1–3 hours at 30°C and the dsRNA was purified from the ssRNA by stepwise precipitation with 2 M and 4 M LiCl as described previously [75] . The ϕ6 genomic dsRNA was purified using Trizol/chloroform ( 5:1 ) extraction followed by successive precipitations with 44% ( v/v ) isopropanol , 4 M LiCl , and 0 . 3 M sodium acetate pH 6 . 3 . The purified RNA was washed with cold 70% ( v/v ) ethanol and dissolved to sterile Milli-Q water . The replication , transcription and TNTase activities of the hPBV RdRP were assayed in 6% ( w/v ) PEG 4000 , 20 mM NH4Ac , 0 . 1 mM EDTA , 2 mM MgCl2 , 0 . 1% ( v/v ) Triton X-100 , 50 mM HEPES-KOH pH 7 . 5 and 0 . 4 U/μl RNase inhibitor RiboLock ( Thermo Scientific ) , typically in 10 μl reaction volumes , using 55 nM concentrations of the WT or ΔLOOP polymerase and equimolar amounts of the RNA strands . The conditions were not stringent since PBV RdRP exhibited replication and transcription activity also in different divalent cation conditions ( MgCl2 and MnCl2 ) and NTP concentrations . When back-priming and transcription time course reactions were assayed the RdRP amount was increased eight fold ( without changing the molarity of the template RNA ) . A final concentration of 0 . 2 mM NTPs was used in the replication and transcription reactions . The TNTase activity was assayed in the presence of 0 . 03 μM UTP . For the identification of newly synthetized RNAs the reactions were supplemented with ( α-33P ) labeled UTP ( 0 . 1 μCi/μl reaction , Perkin-Elmer , 3000 Ci/mmol ) or ( γ-32P ) labeled GTP ( 0 . 2–0 . 3 μCi/μl reaction , Perkin-Elmer , 6000 Ci/mmol ) . The reactions were incubated for 1 hour at 37°C and stopped with the addition of 2×U loading buffer ( 8 M Urea , 10 mM EDTA , 0 . 2% ( w/v ) SDS , 6% ( v/v ) glycerol , 0 . 05% ( w/v ) bromophenol blue , 0 . 05% ( w/v ) xylene cyanol ) . For the evaluation of the back-priming activity , the reaction products were boiled for 3 minutes to denature the sample . The reaction mixtures were analyzed by gel electrophoresis in 0 . 8 or 1 . 2% ( w/v ) agarose . The gels were dried and the signals were collected on imaging plates ( Fujifilm ) , which were subsequently scanned using Typhoon TRIO Imager ( GE Healthcare ) . Quantification was done using AIDA Image Analyzer . The gene encoding the CP of the hPBV Hy005102 strain ( 552 aa , 62 kDa ) was cloned into a pET19b ( + ) vector without the corresponding UTRs . This construct was co-transformed along with the hPBV RdRP gene in a pET28b ( + ) vector into a Rosetta 2 strain of E . coli . The expression of both proteins ( i . e . untagged CP and His-tagged RdRP ) was then induced by the addition of IPTG . The cells were pelleted , resuspended , sonicated , and clarified in as described for the WT RdRP . After clarification , the supernatant was run through a nickel-NTA column to remove any free RdRP . The VLPs were subsequently purified using density gradient ultracentrifugation in a CsCl gradient of 1 . 1–1 . 4 g/cm3 . The ultracentrifugation was performed using a SW41 Ti rotor ( Beckman Coulter ) at 35 , 000 rpm for 6 hours , after which the light-scattering VLP-zone was collected . Western blots were carried out using an anti-hPBV CP antibody obtained from Pacific Immunology ( Ramona , CA , US ) and an anti-6xHis antibody obtained from ThermoFisher ( Houston , TX , US ) . VLP formation was confirmed using TEM ( JEOL 2010 , Japan ) as previously described [76] . The RNA oligos for the gel shift assays were commercially purchased from Sigma-Aldrich . The three RNA oligonucleotides used for these experiments were the PBV 5’- ( + ) UTR ( 5’-GUAAAAUUUUCGAAUUUUAU-3’ ) , PBV 3’- ( + ) UTR ( 5’-GGACUAACAGUCCCAACUGC-3’ ) , and a nonsensical CA repeat ( 5’-CACACACACACACACACACA-3’ ) . The two PBV2+ ssRNA molecules , including the full-length PBV2+ and a deleted version PBV2+ ( Δ1–645 ) , were made by in vitro transcription as described above . RNA oligos and PBV2+ ssRNA molecules were labeled with γ-32P-ATP using T4 polynucleotide kinase ( New England Biolabs ) at 37°C for 30 minutes . The labeling was halted by the addition of 5 μl of 500 mM EDTA . The labeled oligonucleotides were then purified and desalted using a Sephadex G-50 Nick column ( GE Life Sciences ) . The gel shift assays were conducted in 50 mM Tris-HCl pH 7 . 5 , 50 mM KCl , 1 mM NaN3 , 5 mM β-mercaptoethanol , and 10% ( v/v ) glycerol . A 1 nM concentration of 32P-labeled RNA was incubated with increasing concentrations of either the WT or the ΔLOOP hPBV RdRP for 30 min at room temperature . The samples were then loaded onto a 15% ( w/v ) native polyacrylamide gel and run at 50 V for 4 hours . The polyacrylamide gel setup was placed in an ice bath during the experiments to minimize heating . The radioactive RNA was visualized by phosphorimaging using the FujiFilm FLA-5000 imager and quantified using the program ImageGuage v4 . 0 . The fraction of the RNA bound was then calculated as the amount of the bound RNA divided by the sum of the total RNA and plotted versus the corresponding concentration of the protein .	Viral polymerases replicate the genome of the virus , which is essential for the synthesis of the progeny . All double-stranded ( dsRNA ) viruses have virion-associated polymerases that catalyze RNA synthesis within an intact capsid . Picobirnavirus ( PBV ) is a small dsRNA virus , and it has been shown that the capsid of PBV possesses an unusual architecture suggesting an uncommon assembly pathway that it could potentially share with another group of small dsRNA viruses called partitiviruses . We have performed both structural and functional studies to look at how the PBV polymerase performs its function and how these molecules are placed within the capsid . The PBV RdRP structure was solved and revealed that the RdRP has an interesting loop structure in its interior . Additionally , activity assays showed that the RdRP possesses catalytic activity in the absence of other viral proteins . Removing the loop structure was found to change the way that the RdRP initiates RNA replication . Further experiments also showed that the RdRP did not interact with the viral capsid protein ( CP ) but had a strong affinity for a conserved terminal sequence of the PBV genome . This suggests that in PBV the RdRP may be encapsidated based upon both the RdRP and CP co-interacting with the viral genome .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "sequencing", "techniques", "nucleic", "acid", "synthesis", "dna-binding", "proteins", "microbiology", "viral", "structure", "viruses", "polymerases", "rna", "viruses", "sequence", "motif", "analysis", "molecular", "biology", "techniques", "rna", "synthesis", "microbial",...	2016	Initiation of RNA Polymerization and Polymerase Encapsidation by a Small dsRNA Virus
Biological systems can share and collectively process information to yield emergent effects , despite inherent noise in communication . While man-made systems often employ intricate structural solutions to overcome noise , the structure of many biological systems is more amorphous . It is not well understood how communication noise may affect the computational repertoire of such groups . To approach this question we consider the basic collective task of rumor spreading , in which information from few knowledgeable sources must reliably flow into the rest of the population . We study the effect of communication noise on the ability of groups that lack stable structures to efficiently solve this task . We present an impossibility result which strongly restricts reliable rumor spreading in such groups . Namely , we prove that , in the presence of even moderate levels of noise that affect all facets of the communication , no scheme can significantly outperform the trivial one in which agents have to wait until directly interacting with the sources—a process which requires linear time in the population size . Our results imply that in order to achieve efficient rumor spread a system must exhibit either some degree of structural stability or , alternatively , some facet of the communication which is immune to noise . We then corroborate this claim by providing new analyses of experimental data regarding recruitment in Cataglyphis niger desert ants . Finally , in light of our theoretical results , we discuss strategies to overcome noise in other biological systems . An intuitive description of the model follows . For more precise definitions , see , Section The models in the Supplementary Information . Consider a population of n agents . Thought of as computing entities , assume that each agent has a discrete internal state , and can execute randomized algorithms—by internally flipping coins . In addition , each agent has an opinion , which we assume for simplicity to be binary , i . e . , either 0 or 1 . A small number , s , of agents play the role of sources . Source agents are aware of their role and share the same opinion , referred to as the correct opinion . The goal of all agents is to have their opinion coincide with the correct opinion . To achieve this goal , each agent continuously displays one of several messages taken from some finite alphabet Σ . Agents interact according to a random pattern , termed as the parallel-PULL model: In each round t ∈ N + , each agent u observes the message currently displayed by another agent v , chosen independently and uniformly at random from all agents . Importantly , communication is noisy , hence the message observed by u may differ from that displayed by v . More precisely , for any m , m′ ∈ Σ , let Pm , m′ be the probability that , any time some agent u observes an agent v holding some message m ∈ Σ , u actually receives message m′ . The probabilities Pm , m′ define the entries of the noise-matrix P [21] , which does not depend on time . The noise is characterized by a noise parameter δ > 0 . Our model encapsulates a large family of noise distributions , making our bounds highly general . Specifically , the noise distribution can take any form , as long as it satisfies the following criterion . Definition 1 ( δ-uniform noise ) We say that the noise is δ-uniform if Pm , m′ ≥ δ for any m , m′ ∈ Σ . When messages are noiseless , it is easy to see that the number of rounds that are required to guarantee that all agents hold the correct opinion with high probability is O ( log n ) [16] . In what follows , we aim to show that when the δ-uniform noise criterion is satisfied , the number of rounds required until even one non-source agent can be moderately certain about the value of the correct opinion is very large . Specifically , thinking of δ and s as constants independent of the population size n , this number of rounds is at least Ω ( n ) . To prove the lower bound , we will bestow the agents with capabilities that far surpass those that are reasonable for biological entities . These include: We show that even given this extra computational power , fast convergence cannot be achieved . All the more so , fast convergence is impossible under more realistic assumptions . In all the statements that follow we consider the parallel-PULL model satisfying the δ-uniform noise criterion , with cs/n < δ ≤ 1/\|Σ\| for some sufficiently large constant c , where the upper bound follows from the criterion given in Definition 1 . Hence , the previous lower bound on δ implies a restriction on the alphabet size , specifically , \|Σ\| < n/ ( cs ) . Theorem 1 . 1 Any rumor spreading protocol cannot converge in less than Ω ( n δ s 2 ( 1 - δ \| Σ \| ) 2 ) rounds . Observe that the lower bound we present loses relevance when s is of order greater than n , as our proof technique becomes uninformative in presence of a large number of sources ( see Remark 2 in the Supplementary Information ) . Recall also that we assume that a source is aware that it is a source , but if it wishes to identify itself as such to agents that observe it , it must encode this information in a message , which is , in turn , subject to noise . We also consider the case in which an agent can reliably identify a source when it observes one ( that is , this information is not noisy ) . For this case , the following lower bound , which is weaker than the previous one but still polynomial , apply ( see also the S1 Text , Detectable sources ) : Corollary 1 . 1 Assume that sources are reliably detectable . There is no rumor spreading protocol that converges in less than Ω ( ( n δ s 2 ( 1 - δ \| Σ \| ) 2 ) 1 / 3 ) rounds . Our results suggest that , in contrast to systems that enjoy stable connectivity , structureless systems are highly sensitive to communication noise ( see Fig 1 ) . More concretely , the two crucial assumptions that make our lower bounds applicable are: 1 ) stochastic interactions , and 2 ) δ-uniform noise ( Fig 1 , right hand panel ) . When agents can stabilize their interactions the first assumption is violated . In such cases , agents can overcome noise by employing simple error-correction techniques , e . g . , using redundant messaging or waiting for acknowledgment before proceeding to the next action . As demonstrated in Fig 1 , ( left hand panel ) , when the noise is not uniform , it might be possible to overcome it with simple techniques based on using default neutral messages , and employing exceptional distinguishable signals only when necessary . Our theoretical results assert that efficient rumor spreading in large groups could not be achieved without some degree of communication reliability . An example of a biological system whose communication reliability appears to be deficient in all of its components is recruitment in Cataglyphis niger desert ants . In this species , when a forager locates an oversized food item , she returns to the nest to recruit other ants to help in its retrieval [31 , 32] . In our experimental setup , summarized in Fig 2 , recruitment occurs in the small area of the nest’s entrance chamber ( Fig 2a ) . We find that within this confined area , the interactions between ants follow a near uniformly random meeting pattern [33] . In other words , ants seem to have no control over which of their nest mates they will meet next ( Fig 2b ) . This random meeting pattern approximates the first main assumption of our model . Another of the model’s assumptions is that ants interact in parallel . This implies that the interaction rate per ant be constant and independent of group size . Indeed , the empirical rate of interaction during the recruitment process was measured to be 0 . 82 ± 0 . 07 ( mean ± sem , N = 44 ) interactions per minute per ant and induces a small increase with group size: 0 . 62 ± 0 . 13 for two ants ( N = 8 ) and 1 ± 0 . 2 for a group sizes of 9-10 ( N = 5 ) . It has been shown that recruitment in Cataglyphis niger ants relies on rudimentary alerting interactions [34 , 35] which are subject to high levels of noise [32] . Moreover , the information an ant passes in an interaction can be attributed solely to her speed before the interaction [32] . Binning ant speeds into three discrete messages and measuring the responses of stationary ants to these messages , we can estimate the probabilities of one message to be mistakenly perceived as another one ( see Estimating δ in the Methods ) . We find that this communication is extremely noisy which complies with the uniform-noise assumption with a δ of approximately 0 . 3 ( Fig 2c ) . While artificially dividing the continuous speed signals into a large number of discrete messages ( thus creating a larger alphabet ) would inevitably decrease δ , this is not supported by our empirical data ( see Section Methods ) . Finally , the interaction scheme , as exhibited by the ants , can be viewed somewhere in-between the noisy-push and the noisy-pull models . Moving ants tend to initiate more interaction [32] and this may resemble , at first glance , a noisy-push interaction scheme . However , the ants’ interactions actually share characteristics with noisy-pull communication . Mainly , ants cannot reliably distinguish an ant that attempts to transmit information from any other non-communicating individual [32] . The fact that a receiver ant cannot be certain that a message was indeed communicated to her coincides with the lack of reliability in information transmission in line with our theoretical assumptions ( see more details on this point in the Section Separation between PUSH and PULL ) . Given the coincidence between the communication patterns in this ant system and the requirements of our lower bound we expect long delays before any uninformed ant can be relatively certain that a recruitment process is occurring . We therefore measured the time it takes an ant , that has been at the food source , to recruit the help of two nest-mates for different total group size . One might have expected this time to be independent of the group size or even to decrease as two ants constitute a smaller fraction of larger groups . To the contrary , we find that the time until the second ant is recruited increases with group size ( p < 0 . 05 Kolmogorov-Smirnov test over N = 24 experiments , see Fig 2d ) . Our theoretical results set a lower bound on the minimal time it takes uninformed ants to be recruited . Note that our lower bounds actually correspond to the time until any individual can be sure with more than 2/3 probability of the rumor . In the context of the ant recruitment experiment this means that if an ant goes out of the nest only if she is sure with some probability that there is a reason to exit , then the lower bounds correspond to the time until the first , and similarly the second ( see Fig 2d ) , ants exit the nest . Our lower bound is linear in the group size ( Theorem 1 . 1 ) . Note that this does not imply that the ants’ biological algorithm matches the lower bound and must be linear as well . Rather , our theoretical results qualitatively predict that as group size grows , recruitment times must eventually grow as well . This stands in agreement with Fig 2d . Thus , in this system , inherently noisy interactions on the microscopic level have direct implications on group level performance . Here , we provide the intuition for our main theoretical result , Theorem 1 . 1 . For a formal proof please refer to the S1 Text , The lower bounds . The proof can be broken into three parts and , below , we refer to each of them separately . Several of the assumptions discussed earlier for the parallel-PULL model were made for the sake of simplicity of presentation . In fact , our results can be shown to hold under more general conditions , that include: 1 ) different rate for sampling a source , and 2 ) a more relaxed noise criterion . In addition , our theorems were stated with respect to the parallel-PULL model . In this model , at every round , each agent samples a single agent u . a . r . In fact , for any integer k , our analysis can be applied to the model in which , at every round , each agent observes k agents chosen u . a . r . In this case , the lower bound would simply reduce by a factor of k . Our analysis can also apply to a sequential variant , in which in each time step , two agents u and v are chosen u . a . r from the population and u observes v . In this case , our lower bounds would multiply by a factor of n , yielding , for example , a lower bound of Ω ( n2 ) in the case where δ and s are constants . Observe that the latter increase is not surprising as each round in the parallel-PULL model consists of n observations , while the sequential model consists of only one observation in each time step . See more details in the Supplementary Information . Our lower bounds on the parallel-PULL model ( where agents observe other agents ) should be contrasted with known results in the parallel-PUSH model ( this is the push equivalent to parallel-PULL model , where in each round each agent may or may not actively push a message to another agent chosen u . a . r . ) . Although never proved , and although their combination is known to achieve more power than each of them separately [16] , researchers often view the parallel-PULL and parallel-PUSH models as very similar on complete communication topologies . Our lower bound result , however , undermines this belief , proving that in the context of noisy communication , there is an exponential separation between the two models . Indeed , when the noise level is constant for instance , convergence ( and in fact , a much stronger convergence than we consider here ) can be achieved in the parallel-PUSH using only logarithmic number of rounds [20 , 21] , by a simple strategy composed of two stages . The first stage consists of providing all agents with a guess about the source’s opinion , in such a way that ensures a non-negligible bias toward the correct guess . The second stage then boosts this bias by progressively amplifying it . A crucial aspect in the first stage is that agents remain silent until a certain point in time that they start sending out messages . This prevents agents from starting to spread information before they have sufficiently reliable knowledge and allows for a balanced control of the rumor spread . More specifically , marking an edge corresponding to a message received for the first time by an agent , the set of marked edges forms a spanning tree of low depth , rooted at the source . The depth of such tree can be interpreted as the deterioration of the message’s reliability . On the other hand , as shown here , in the parallel-PULL model , even with the synchronization assumption , rumor spreading cannot be achieved in less than a linear number of rounds . Perhaps the main reason why these two models are often considered similar is that with an extra bit in the message , a PUSH protocol can be approximated in the PULL model , by letting this bit indicate whether the agent in the PUSH model was aiming to push its message . However , for such a strategy to work , this extra bit has to be reliable . Yet , in the noisy PULL model , no bit is safe from noise , and hence , as we show , such an approximation cannot work . In this sense , the extra power that the noisy PUSH model gains over the noisy PULL model , is that the very fact that one node attempts to communicate with another is reliable . This , seemingly minor , difference carries significant consequences . Communication in man-made computer networks is often based on reliable signals which are typically transferred over highly defined structures . These allow for ultra-fast and highly reliable calculations . Biological networks are very different from this and often lack reliable messaging , well defined connectivity patterns or both . Our theoretical results seem to suggest that , under such circumstances , efficient spread of information would not be possible . Nevertheless , many biological groups disseminate and share information , and , often , do so reliably . Next , we discuss information sharing in biological systems within the general framework of our lower-bounds . The correctness of the lower bounds relies on two major assumptions: 1 ) stochastic interactions , and 2 ) uniform noise . Communication during desert ant recruitment complies with both these assumptions ( see Fig 2b and 2c ) and indeed the speed at which messages travel through the group ( see Fig 2d ) is low . Below , we discuss several biological examples where efficient rumor spreading is achieved . We expect that , in these examples , at least one of the assumptions mentioned above should break adding some degree of reliability to the overall communication . The group can then utilize this reliability and follow one of the strategies mentioned in Section Theoretical results , in order to yield reliable collective performance . We begin by discussing examples that violate the first assumption , namely , that of stochastic interactions , and then discuss examples that violate the second assumption , namely , uniform noise . To estimate the noise parameter δ we used interactions between ants moving at three different speed ranges ( measured in cm/sec ) , namely , ‘a’: 1-10 , ‘b’: 10-20 , and ‘c’: over 20 and “receiver” ants . Only interactions in which the receiver ant was initially stationary were used as to ensure that the state of these ants before the interaction is as similar as possible . The message alphabet is then assumed to be Σ = {a , b , c} . The response of a stationary ant v to the interaction was quantified in terms of her speed after the interaction . An alphabet of three messages was used since the average responses of v to any two messages were significantly different ( all p-values smaller than 0 . 01 ) justifying the fact that these are not artificial divisions of a continuous speed signal into a large number of overlapping messages . On the other hand , dividing the bins further ( say , each bin divided into 2 equal bins ) yielded statistically indistinguishable responses from the receiver ( all p-values larger than 0 . 11 ) . Therefore , our current data best supports a three letter alphabet . Assuming equal priors to all messages in Σ , and given specific speed of the receiver ant , v , the probability that it was the result of a specific message i ∈ Σ was calculated as pi ( v ) = p ( v ∣ i ) /∑k∈Σ p ( v ∣ k ) , where p ( v ∣ j ) is the probability of responding in speed v after “observing” j . The probability δ ( i , j ) that message i was perceived as message j was then estimated as the weighted sum over the entire probability distribution measured as a response to j: δ ( i , j ) = ∑v p ( v ∣ j ) ⋅ pi ( v ) . The parameter δ can then be calculated using δ = min{δ ( i , j ) ∣ i , j ∈ Σ} .	Biological systems must function despite inherent noise in their communication . Systems that enjoy structural stability , such as biological neural networks , could potentially overcome noise using simple redundancy-based procedures . However , when individuals have little control over who they interact with , it is unclear what conditions would prevent runaway error accumulation . This paper takes a general stance to investigate this problem , concentrating on the basic information-dissemination task of rumor spreading . Drawing on a theoretical model , we prove that fast rumor spreading can only be achieved if some part of the communication setting is either stable or reliable . We then provide empirical support for this claim by conducting new analyses of data from experiments on recruitment in desert ants .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "collective", "animal", "behavior", "invertebrates", "ecology", "and", "environmental", "sciences", "sociology", "social", "sciences", "animals", "habits", "animal", "behavior", "deserts", "social", "communication", "zoology", "nesting", "habits", "computer", "and", "in...	2018	Limits on reliable information flows through stochastic populations
A gene expression atlas is an essential resource to quantify and understand the multiscale processes of embryogenesis in time and space . The automated reconstruction of a prototypic 4D atlas for vertebrate early embryos , using multicolor fluorescence in situ hybridization with nuclear counterstain , requires dedicated computational strategies . To this goal , we designed an original methodological framework implemented in a software tool called Match-IT . With only minimal human supervision , our system is able to gather gene expression patterns observed in different analyzed embryos with phenotypic variability and map them onto a series of common 3D templates over time , creating a 4D atlas . This framework was used to construct an atlas composed of 6 gene expression templates from a cohort of zebrafish early embryos spanning 6 developmental stages from 4 to 6 . 3 hpf ( hours post fertilization ) . They included 53 specimens , 181 , 415 detected cell nuclei and the segmentation of 98 gene expression patterns observed in 3D for 9 different genes . In addition , an interactive visualization software , Atlas-IT , was developed to inspect , supervise and analyze the atlas . Match-IT and Atlas-IT , including user manuals , representative datasets and video tutorials , are publicly and freely available online . We also propose computational methods and tools for the quantitative assessment of the gene expression templates at the cellular scale , with the identification , visualization and analysis of coexpression patterns , synexpression groups and their dynamics through developmental stages . Deciphering and integrating the genetic and cellular dynamics underlying morphogenesis and homeostasis in living systems is a major challenge of the post-genomic era . Although full genome sequencing is available for a number of animal model organisms [1] , quantitative data for the spatial and temporal expression of genes is still lacking [2] . Remarkable advances in photonic microscopy imaging [3] , [4] , [5] and labeling techniques [6] allowed gathering data at all levels of a multicellular system's organization with adequate spatial and temporal resolutions . Fluorescent in situ hybridization techniques [7] , immunocytochemistry and transgenesis , combined with 3D optical sectioning , make it now possible to assess the dynamics of gene expression throughout animal development with precision at the single-cell level . However , moving forward from databases of gene expression that contain average values at low spatiotemporal resolutions—such as those obtained from DNA microarrays available for most model organisms—to a dynamic , cell-based 4D atlas is a major paradigm shift that requires the development of appropriate methods and tools . In this context , the design and implementation of automated image analysis strategies to build a gene expression atlas with resolution at the cellular scale is an important methodological bottleneck towards greater biological insights [8] , [9] . The task of assembling imaging data from cohorts of individuals , or analyzed embryos , onto a series of 3D prototypes , or templates ( one per developmental stage ) , can be approached by finding a spatial correspondence between individuals based on registration methods , a technique used in medical imaging [10] . Yet , gathering and consolidating into a single prototype multimodal and multiscale features from different specimens that exhibit phenotypic variability remains a difficult challenge . Recent studies on different model organisms have explored computational strategies for building atlases either by measuring cell positions to create prototypic specimens [11] , [12] or by gathering gene expression patterns observed in cohorts of specimens [13] , [14] , [15] , [16] . Yet , very few frameworks have combined both features . Long et al . [11] collected data from 15 C . elegans specimens at the earliest larval stage ( L1 with 357 cells ) to build a statistical 3D atlas of nuclear center positions . C . elegans presents a number of advantages facilitating the reconstruction process . The entire organism can be imaged with resolution at the single-cell level and its cell lineage tree is stereotyped enough to allow spatiotemporal matching of different individuals at this level . The same features allowed the reconstruction of a prototypic lineage for a cohort containing six specimens of Danio rerio ( zebrafish ) embryos throughout their first 10 cell division cycles [12] . Peng et al . [15] achieved the spatial matching of 2 , 945 adult Drosophila brains to collect the expression patterns of 470 different genes . Similarly , Lein et al . [13] constructed a comprehensive atlas of the adult mouse brain containing about 20 , 000 gene patterns . The first gene expression atlas with resolution at the cellular scale was produced by Fowlkes et al . [14] . They integrated 95 gene expression patterns observed at 6 different developmental stages in a total of 1 , 822 different Drosophila embryos within a common 3D stencil . Applying this approach to vertebrate model organisms is more difficult because of higher cell lineage variability and heterogeneous levels of gene expression within highly dynamic patterns . In addition , the reconstruction of 3D gene expression templates at cellular scale for vertebrate species is likely to require the acquisition of partial volumes recorded at high resolution [15] from single specimens , and their precise mapping onto in toto reference specimens . The zebrafish , a vertebrate model organism increasingly used for its relevance to biomedical applications [17] , cumulates good properties for investigating the reconstruction of the multiscale dynamics of early embryogenesis . The gene regulatory network ( GRN ) architecture of the zebrafish early embryonic development is under construction [18] and the embryo is easily accessible and amenable to transgenesis , multiple in situ staining and 3D+time imaging . The spatiotemporal data offered by a 4D atlas of gene expression with resolution at the cellular level is expected to provide the necessary measurements for further modeling of the GRN dynamics and possible integration of the genetic and cellular levels of organization [19] . Such data would make the zebrafish the first vertebrate model amenable to a systemic study . However , building 3D templates of gene expression for the zebrafish blastula and gastrula stages is especially problematic due to the lack of morphological landmarks required for the registration of patterns [20] , [21] . We provide a methodology to construct , visualize and analyze a gene expression atlas composed of templates at various stages of vertebrate early development . We designed , implemented and now deliver two computational frameworks , Match-IT and Atlas-IT , to support the automatic mapping of 3D gene expression patterns from different individuals ( the analyzed embryos ) onto common reference specimens ( the templates ) with resolution at the cellular scale . This “virtual multiplexing” procedure [14] overcomes the limited number of gene products that can be jointly stained and measured in a single specimen . Match-IT was used to produce the prototypic cartography of 9 gene expression patterns imaged from 3D double fluorescent in situ hybridization at 6 developmental stages ( Table S1 , Movie S1 , Figs . S1 , S2 , S3 , S4 , S5 , S6 , S7 ) . Atlas-IT was designed to interactively visualize gene coexpression patterns and their dynamics . We validated our 4D atlas construction methodology by an automated quantitative assessment of gene patterns' similarity and overlap through time . Analytical tools , such as clustering , were designed to identify morphogenetic domains and gene synexpression groups , i . e . groups of genes sharing the same spatiotemporal expression patterns . The proposed spatiotemporal atlas of zebrafish blastula and early gastrula preserves the information of the cell as the gene expressing unit , providing means for the integration of genetic and cellular data unavailable so far . We designed a computational framework ( Fig . 1 ) , going from image acquisition to image data analysis , to perform the mapping of different stained gene expression patterns onto a common prototypic model at each developmental stage ( Fig . S8 ) , thus creating a series of 3D templates of gene expression with resolution at the cellular scale . The processing workflow consisted of embryo staining , image data acquisition ( Materials and Methods ) , nuclear center detection , gene pattern segmentation , mapping of the analyzed embryos onto a template at each stage , and selection of template cells positive for the expression of specific genes . This methodology was designed to document at a sufficient spatial and temporal resolution the gene expression dynamics underlying the formation of the Spemann organizer and the embryonic axis of zebrafish early embryos . To this end , we imaged the dorsal side of fluorescently stained embryos with cellular resolution from fixed specimens about every 30 min from 4 to 6 . 3 hpf . The resulting 6 templates comprised a stencil of in toto 3D images of the template specimens ( Fig . 2a ) at different stages , and mappings of the partial 3D views of the analyzed embryos ( Fig . 2b ) . In order to integrate 3D data into one template , our novel Match-IT tool ( Software S1 and User Guide S1 ) performed the segmentation of gene expression domains , the mapping of analyzed embryos onto a common reference specimen and the identification of positive cells ( Fig . 1 and Movie S2 ) , eventually delivering a 3D database that summarized the genetic profile of single cells . Analysis of the 3D templates produced by Match-IT required dedicated visualization tools to test hypotheses and derive biological insights . The available software kits did not fulfill our requirements , either because they were too specific for a given model organism ( such as PointCloudXplore [26] for Drosophila ) or because they were too generic as visualization and processing tools ( such as Icy [27] , Vaa3D [28] , or CellProfiler [29] ) and did not allow displaying selections of individual cellular positions or querying a template for coexpression domains with resolution at the cellular scale . For these reasons , we designed , developed and deliver here the Atlas-IT interactive visualization interface ( Fig . 4a and Software S2 and User Guide S2 ) to explore 4D atlas resources . With this tool , we can interact with the complete atlas data , in particular superimpose raw images ( either as 3D volumes or orthoslices ) , segmented patterns , and the whole set of detected template nuclei or selected positive nuclei at any time point ( Movie S3 ) . Atlas-IT can be used to assess the dynamics of gene coexpression domains or the variability of gene expression patterns . We used Match-IT and Atlas-IT together to reconstruct a 4D atlas of zebrafish early embryogenesis , which is now released . It comprises 6 developmental stages and 9 gene expression patterns chosen to study a specific embryological question , namely the genetic dynamics underlying the formation of the Spemann organizer at the dorsal midline [1] ( a region in the zebrafish containing precursors of the segregation between the prechordal plate and the notochord [30] ) . The 9 genes are: gsc , sox32 , tbx16 , oep , snai1a , foxa2 , ntla , flh , and egfp , where the latter was was detected in a custom-made transgenic line Tg ( −4gsc:egfp ) isc3 . These genes appear as nodes in the axial mesendoderm GRN proposed by Chan et al . [18] . In addition , egfp allowed us to validate the transgenic line as a faithful reporter of early gsc gene expression ( Fig . S15 ) . The time series of 3D templates was chosen to explore gene expression dynamics from the onset of zygotic activation at 3 hpf until early gastrulation , and encompasses the following developmental stages: sphere ( 4 hpf ) , dome ( 4 . 3 hpf ) , 30% epiboly ( 4 . 7 hpf ) , 50% epiboly ( 5 . 3 hpf ) , shield ( 6 hpf ) and late shield ( 6 . 3 hpf ) according to the staging defined at . For each new gene expression to be mapped , a cohort of individuals was processed for double in situ hybridization and 3 of them were imaged . The atlas construction methodology was established by using one specimen of each cohort ( Table S1 ) . The atlas was constructed to be able to compare gene expression patterns from different stained specimens . Establishing spatial relationships between gene patterns required assessing gene expression variability and calculating mean expression domains ( Materials and Methods ) . The expectation was that the spatial relationships observed between two genes stained in the same embryo should be maintained between their mean expression domains in a template . The expression of gsc was revealed in 9 different specimens , which comprised 8 analyzed embryos and one template , at each developmental stage . It provided a paradigmatic case to calculate a mean expression domain and assess gene variability ( Fig . S16 ) . At any given stage , we quantified the mean distance from the complete outer surface of each individual gsc domain ( ) to the closest boundary point of the mean domain ( ) , following a leave-one-out protocol ( Fig . 4b ) . The measured distance , which reflected both the accuracy of our mapping scheme and the inter-individual variability between the boundaries of the gsc expression domains , was on average less than , i . e . approximately one cell diameter ( Fig . S17 ) . This accuracy error remained within the same range independently from the thickness of the three main embryo planes ( Fig . S18 ) . Additionally , more than 80% of all the individual border points were less than one cell row away from the border , indicating that there were no large distance discrepancies along the contours ( Fig . S19 ) . To demonstrate that this level of accuracy was maintained in regions far from the gsc expression domains , we replicated the same quality measure with another gene , tbx16 , which spread across a much larger area than gsc . With Match-IT , we added two new tbx16 datasets , tbx16b and tbx16c , to the already existing tbx16a expression in the atlas at 6 . 3 hpf ( Fig . S20a ) . The mean distance from the complete outer surface of each individual tbx16j domain to the closest boundary point of tbx16 remained under one cell diameter ( Fig . S20b ) . Moreover , the histogram of distances between border points of tbx16j and tbx16 confirmed that most of the expression contours lay within two cell rows from each other ( Fig . S20c ) . Note that this quality measure was an upper bound of the registration quality reflecting both the mapping variability and the intrinsic inter-embryo variability . Additionally , we confirmed that the spatial relationships between every gene and the patterns in the analyzed embryos were the same in each template with respect to the domain . In particular , this was the case for the oep-gsc pair illustrated in Fig . 4c , d . Various analysis tools for the quantitative analysis of a spatiotemporal atlas of gene expression were also developed ( see Materials and Methods ) . We performed an automated identification of gene coexpression pattern dynamics in space and time , explored clustering strategies at the cellular level to automatically identify morphogenetic domains or spatiotemporal gene synexpression groups , and introduced an “entropy” analysis for gene expression . Our choice to work with a hybrid automated/supervised method of nuclear center detection proved to be suitable for quantifying certain features of gene expression pattern dynamics at the cellular level . This opens the possibility to discuss , in terms of cell number , the overlap between gene expression patterns and their evolution in time . It also allows studying whether cell proliferation alone is enough to account for the expansion of gene expression patterns , by correlating internuclear distance and cell division , which , in zebrafish early development , happens at constant global cell volume ( Fig . S26 ) . On the other hand , the resolution of the atlas at the cellular scale is a requirement to exploit the correlation between gene expression dynamics and cell lineage . Cellular resolution enables further mapping of the atlas onto digital specimens reconstructed from live in toto imaging , starting with our transgenic line . Working at the cellular resolution was also intended to tackle the problem of gene expression quantification . Current strategies for in situ hybridization could at best provide relative measurements suitable for quantifying graded patterns and fuzzy borders within each analyzed embryo . Such a relative quantification would be readily available from our atlas ( Fig . S10 ) . We expect future developments of the programmable in situ amplification technique [7] to help achieve quantification of gene expression comparable among different analyzed embryos at the cellular level . The relevance of the atlas relies on its ability to represent and integrate the same information as would be obtained by inspecting different patterns in the same specimen . This depends on the accuracy of the registration strategies but most importantly on how the atlas construction scheme deals with individual variability . Every step of the mapping strategy has to cope with individual variations in terms of shape , cell number , cell density , and variability of the reference gene pattern . In this context , the choice of the template is crucial . The template should be closest to the mean of the population , based on geometric parameters and gene expression . Ideally , a multiscale model of individual variability should drive the choice of the atlas template as well as representative reference patterns or features to guide the mapping . In our case , the gsc pattern served as a guide for the registration step , based on the hypothesis that its expression is symmetric with respect to the bilateral plane . Although this is a reasonable assumption , it is an approximation that might be confronted to other features such as other reference gene patterns or additional morphological traits . The templates used in this paper were visually chosen to be the closest to the mean . Although this choice may not be fully representative of the average morphology , the concept of average is also not completely relevant for the released proof-of-principle atlas that comprises 9 specimens per developmental stage . The tools released here open the way for a broader population that could ideally produce a more representative template . In this context , we calculated a mean gsc expression pattern after registering the domains from 9 different specimens . The resulting domain could be subsequently used as a new reference to refine the global mappings . Moreover , all the genes gathered in the atlas could be averaged , thus preventing potentially misleading conclusions based on single specimens that might be outliers . The increase in size of the cohorts will allow exploring the possible convergence of the averaging strategy toward a single or multiple prototypical specimens . Atlas resources will only be fully exploited with the development and use of automated analysis methods and dedicated visualization tools . Toward this objective , we designed Atlas-IT to provide a number of functionalities not available in any of the visualization tools that we examined: augment/visualize/analyze raw data and segmented data , calculate mean gene expression domains , gene coexpression patterns , synexpression groups , and morphogenetic domains by cell clustering . Interactive visualization and data display are essential to reveal biologically relevant information . The exploration of analytical methods to highlight spatial and temporal correlations is also a major endeavor . Typically , clustering methods have been used to establish the gene expression profiles of cells and tissues from microarray data , and more recently to group anatomical regions according to their gene expression profile [13] , [32] , [33] . Although clustering of spatial gene expression patterns has been described elsewhere [34] , it is the first time that this method is applied to gene expression profiles at the cellular level , , providing the means to reveal morphogenetic domains and synexpression groups . Additionally , whereas the Shannon entropy has been used to measure gene expression complexity [35] , it is also the first time that this measure is applied to spatially mapped data . Introducing the concept of “genetic entropy” in the analysis of atlas data offers a new systematic way to assess cell diversification and its underlying genetic complexity . This analysis proved to be robust against the noise due to errors in the segmentation and/or spatial mapping . Although a relatively high proportion ( 100 out of 512 ) of all possible gene expression profiles were found in the atlas , only 30 of them ( i . e . ) produced 75% of all the atlas genetic information ( Fig . S24 ) . Making a gene expression atlas is a necessary step toward the integration of multiscale and multimodal data , which should be organized , displayed and annotated to provide and share as much relevant information as possible . Developmental biology remains far behind the biomedical field in the construction and sharing of this type of resources . Thus , before reaching a consensus and establishing standards in the field , a lot remains to be explored in terms of different schemes , their flexibility , their potential and limitations . The atlas construction process presented here allowed us to address some of the most difficult biological questions linked to individual variability , its components and characteristic scales . A gene expression atlas often comprises hundreds or even thousands of genes [36] . On the other hand , resources can grow and diffuse only if deployed together with appropriate algorithms and analytical tools . Our novel construction and manipulation methods , which led to the first release of the zebrafish blastula and early gastrula atlas , are meant as a contribution toward the complete reconstruction of the zebrafish embryonic physiome ( or “embryome” ) under different genetic and environmental conditions . In vitro fertilization was used to synchronize the spawn from wild type ( wt ) or transgenic crosses from the custom made fish line Tg ( −4gsc:egfp ) isc3 . Embryos , staged according to Kimmel et al . [37] , were fixed 24 h at in PFA 4% then rinsed 3 times in PBS 0 . 1% Tween and stored at in ethanol . Double fluorescent in situ hybridization ( FISH ) was carried out as described in Brend et al . [38] using antisense RNA probes labeled with fluorescein or digoxygenin . Probes were detected with an anti-digoxigenin-POD Fab fragment and anti-fluorescein-POD Fab fragment ( Roche ) used at 1∶250 in a blocking reagent solution ( Roche ) . Probe detection was done with Cy3 or Cy5 mono NHS ester ( Amersham ) or NHSFluoresceine ( Pierce ) tyramides as POD substrates . Nuclei were stained in DAPI ( Invitrogen D3571 ) . As an input , our methodology used 3D images acquired by confocal laser scanning microscopy from fixed zebrafish embryos with fluorescent staining of gene expression patterns and DAPI counterstain to highlight cell nuclei . Image acquisition was performed with a Leica SP2 two-photon ( for DAPI ) and confocal laser scanning upright microscope with a Leica objective HCX APO 20X/0 , 5W U-V-I or HCX APO 10X/0 , 3 . Embryos were mounted in a teflon mold at the bottom of a 3 cm Petri dish filled with 1×PBS , 01% twin 20 , and maintained properly oriented with 1% agarose . The nuclei and gsc expression domains were systematically revealed in all the analyzed embryos and templates , and used to compute the gene expression mappings . In addition to the reference gene , gsc , each analyzed embryo was stained for the expression of another gene of interest . The template data was obtained by imaging the whole embryo with a 10× objective while the analyzed specimens were imaged with a 20× objective providing a 3D view limited to the dorsal side of the embryo with a better spatial resolution ( Fig . 2a , b and Movie S1 ) . The fluorescent in situ hybridization used a state-of-the-art protocol [38] and reproduced standard data ( zfin . orgzfin . org ) . More details about data acquisition parameters and specimen features can be found in Table S1 and Fig . S1 , S2 , S3 , S4 , S5 , S6 , S7 . The Match-IT custom-made code was implemented in ITK and Matlab , including the MathWorks package “geom3D” redistributed under a BSD license . A public release of this software , together with sample datasets and a user guide , accompanies the publication of this article , http://bioemergences . iscpif . fr/documents/MatchIT . zip . The segmentation of the gene expression patterns in each analyzed embryo was carried out by a thresholding operation supervised by a biologist to best define the domain features . This operation was followed by “morphological closing” [39] , a mathematical transformation based on a spherical structuring element the size of a typical cell diameter ( i . e . internuclear distance ) . Finally , a converse “morphological opening” operation left only the largest connected pattern . The common referential extraction started by applying a spherical fit to the outer cell nuclei in all analyzed embryos and templates . The blastoderm margin was identified with a plane , , fitted to the 5% southernmost nuclei . The bilateral symmetry plane , , was found by connecting the spherical model center and the center of mass of the gsc segmented domain perpendicular to the blastoderm margin . The origin of the triplet was placed at the latitude of the blastoderm margin , and the longitude was defined by the center of mass of the gsc domain . The registration ( [10] , [40] ) of the analyzed embryo images on the template employed the ITK registration toolkit to optimize the cross-correlation metric between the embryo shape of the template and that of the analyzed embryos according to a step gradient optimizer . The embryo shapes were weighted by the inverse distance function to the external blastoderm contour ( i . e . half the average internuclear distance away from the outermost nuclear layer ) . The Atlas-IT custom-made visualization platform was implemented in Processing . A public release of this software , together with sample datasets and a user guide , accompanies the publication of this article , http://bioemergences . iscpif . fr/documents/AtlasIT . zip At each developmental time point , a total of 9 different analyzed embryos with staining were mapped onto the template where expression was also revealed . Consequently , every nucleus , , in the template was assigned a value , , ranging from 0 to 9 , depending on the number of analyzed patterns that led to its selection as positive for the expression of . We used a Voronoi diagram to model the cell around each nucleus and assigned these cells their corresponding value . In order to measure the variability of the resulting mean expression , we studied the profile of across 3 cutting lines centered on the mean centroid and following the specimen anatomy along the lateral , radial and sagittal directions respectively ( Fig . S16 ) . We demonstrated that the proposed clustering and entropy schemes are robust against changes in the thresholds employed to segment the gene expression patterns in the atlas . In particular , we chose two gene expressions in the atlas at 6 . 3 hpf: , which co-expresses with , and , which spreads through a much larger area than . For the expression of these two genes , we modified by the thresholds chosen by the biologist expert , computed the new segmented patterns and modified accordingly the number of positive cells found in the atlas . The entropy and clustering resulting from these modified atlases were compared to the original atlas and showed to be robust against these threshold changes ( Fig . S25 ) . To compare the modified vs . the original clustering ( Fig . 5b–c ) we used two metrics previously employed in literature: a ) the correlation between the distance matrix that generate the modified and the original clustering hierarchical trees [42] , b ) the cophenetic correlation , a measure of how faithfully a hierarchical tree preserves the pairwise distances between the original data points [43] , [44] . In this later case , the cophenetic coefficient was extracted by comparing the original hierarchical tree to the new pairwise distances generated by the modified atlases . To compare the modified vs . the original entropy we computed the difference in number of bits . The biggest difference between all the modified and the original atlas was 0 . 15 bits in entropy and a 0 . 03 decrease for both the cophenetic coefficient and the correlation between distance matrix ( Fig . S25 ) . To put these values in perspective , the minimal possible variation to the atlas ( changing the value of one gene expression for one cell only ) had an impact of 0 . 0003 bits of entropy and 0 . 005 in cophenetic correlation , whereas a substantial variation to the atlas ( e . g . changing one third of the atlas values or substituting it by a random atlas ) had an impact of 0 . 83 and 3 . 6 bits of entropy and 0 . 52 and 0 . 79 in cophenetic correlation respectively .	We propose a workflow to map the expression domains of multiple genes onto a series of 3D templates , or “atlas” , during early embryogenesis . It was applied to the zebrafish at different stages between 4 and 6 . 3 hpf , generating 6 templates . Our system overcomes the lack of significant morphological landmarks in early development by relying on the expression of a reference gene ( goosecoid , gsc ) and nuclear staining to guide the registration of the analyzed genes . The proposed method also successfully maps gene domains from partially imaged embryos , thus allowing greater microscope magnification and cellular resolution . By using the workflow to construct a spatiotemporal database of zebrafish , we opened the way to a systematic analysis of vertebrate embryogenesis . The atlas database , together with the mapping software ( Match-IT ) , a custom-made visualization platform ( Atlas-IT ) , and step-by-step user guides are available from the Supplementary Material . We expect that this will encourage other laboratories to generate , map , visualize and analyze new gene expression datasets .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "vertebrates", "osteichthyes", "developmental", "biology", "zebrafish", "fishes", "gene", "mapping", "engineering", "and", "technology", "image", "processing", "signal", "processing", "biology", "and", "life", "sciences", "molecular", "biology", "techniques", "computation...	2014	A Digital Framework to Build, Visualize and Analyze a Gene Expression Atlas with Cellular Resolution in Zebrafish Early Embryogenesis
Drosophila melanogaster is a valuable invertebrate model for viral infection and antiviral immunity , and is a focus for studies of insect-virus coevolution . Here we use a metagenomic approach to identify more than 20 previously undetected RNA viruses and a DNA virus associated with wild D . melanogaster . These viruses not only include distant relatives of known insect pathogens but also novel groups of insect-infecting viruses . By sequencing virus-derived small RNAs , we show that the viruses represent active infections of Drosophila . We find that the RNA viruses differ in the number and properties of their small RNAs , and we detect both siRNAs and a novel miRNA from the DNA virus . Analysis of small RNAs also allows us to identify putative viral sequences that lack detectable sequence similarity to known viruses . By surveying >2 , 000 individually collected wild adult Drosophila we show that more than 30% of D . melanogaster carry a detectable virus , and more than 6% carry multiple viruses . However , despite a high prevalence of the Wolbachia endosymbiont—which is known to be protective against virus infections in Drosophila—we were unable to detect any relationship between the presence of Wolbachia and the presence of any virus . Using publicly available RNA-seq datasets , we show that the community of viruses in Drosophila laboratories is very different from that seen in the wild , but that some of the newly discovered viruses are nevertheless widespread in laboratory lines and are ubiquitous in cell culture . By sequencing viruses from individual wild-collected flies we show that some viruses are shared between D . melanogaster and D . simulans . Our results provide an essential evolutionary and ecological context for host–virus interaction in Drosophila , and the newly reported viral sequences will help develop D . melanogaster further as a model for molecular and evolutionary virus research . Viral infections are universal , and virus-mediated selection may play a unique role in evolution [1] . Viruses are also responsible for highly pathogenic diseases , and the detection , treatment , and prevention of viral disease are important research goals . The model fly , Drosophila melanogaster , provides a valuable tool to understand the biology of viral infection [2 , 3] and antiviral immune responses in invertebrates [4 , 5] , as well as the interaction between viruses and their vectors [6] . Drosophila has also helped elucidate the role of RNA interference ( RNAi ) as an antiviral defence [7 , 8] and has shown that endosymbiotic Wolbachia can protect against viruses [9 , 10] . Recently , the Drosophilidae have been used to address important questions in virus evolution , including determinants of host-range and disease emergence [11–13] . However , although Drosophila virus research has a long history , few D . melanogaster viruses are known in the wild [4 , 14] , and experiments using non-natural Drosophila pathogens may bias our understanding of immune function and its evolution [12] . Following the discovery of Sigma Virus in D . melanogaster ( DMelSV , Rhabdoviridae; reviewed in [15] ) , classical virology surveys in the 1960s and 1970s uncovered Drosophila C Virus ( DCV , Dicistroviridae ) , Drosophila A Virus ( DAV , related to Permutotetraviridae ) , Drosophila X Virus ( DXV , Birnaviridae ) , DFV ( Reoviridae ) , DPV , and DGV ( unclassified ) in D . melanogaster [4 , 14] . Subsequent transcriptomic studies of D . melanogaster identified D . melanogaster Nora Virus ( unclassified Picornavirales; [16] ) , and analyses of small RNAs from D . melanogaster cell culture [17] identified Drosophila Totivirus , American Nodavirus ( closely related to Flockhouse Virus ) , and Drosophila Birnavirus ( closely related to DXV ) . However , only four of these viruses have been isolated from wild flies , have genome sequences available , and are available for experimental study . These include DCV [18] , DMelSV [19] , DAV [20] , and Nora Virus [16] , while DXV is reported to be a cell culture contaminant [14 , 21] . Of these four , only DMelSV has been widely studied in the field [15 , 22 , 23] . Our limited knowledge of D . melanogaster’s natural viruses reflects a historically tight research focus on viruses with direct medical and economic impact . While high-throughput “metagenomic” sequencing has broadened our knowledge of viral diversity in general [24 , 25] , most studies focus on vertebrate faeces [26 , 27] , potential reservoirs of human and livestock disease [28–30] , or crop plants [31] . Relatively few studies have performed metagenomic virus discovery in invertebrates ( for a review , see [32] ) . Recently , a large survey identified an exceptional and unsuspected diversity of negative sense RNA viruses associated with arthropods , suggesting that we may only have been scratching the surface of viral diversity [33] . However , aside from some lepidopteran pests [34] and hymenopteran pollinators [35] , we still know little about the biology of most invertebrate viruses . Although we have much to learn about Drosophila viruses , experiments using both natural and non-natural Drosophila pathogens have given us a better understanding of viral infection and immunity in Drosophila than in any other invertebrate [2 , 5] . DCV , DXV , and Nora Virus have all been used to study the molecular biology of host–virus interaction [7 , 36–39] , and classical genetic approaches have elucidated the basis of host resistance to DCV and DMelSV [40–45] . Many insect viruses—notably Cricket Paralysis Virus , Flock House Virus ( from beetles ) , Sindbis Virus ( mosquito-vectored ) , Vesicular Stomatitis Virus ( mosquito-vectored ) , and Invertebrate Iridovirus 6 ( from mosquitoes ) —have helped characterise the roles of the RNAi , IMD , Toll , autophagy , and Jak-Stat pathways in antiviral immunity ( see [5] for a review ) . These studies show that Drosophila has a sophisticated and effective antiviral immune response , and both molecular [12] and population genetic [46–48] studies suggest that this immune system may be locked into an evolutionary arms race with viruses . However , until we understand the diversity , distribution , or prevalence of viral infection in D . melanogaster , it is hard to put these results into their evolutionary or ecological context . Here we use a metagenomic approach to identify more than 20 novel viruses associated with D . melanogaster , including the first DNA virus to be identified in D . melanogaster . Based on the presence of virus-derived 21 nucleotide ( nt ) small RNAs ( which are characteristic of an antiviral RNAi response in Drosophila [49] ) , we argue that these sequences represent active viral infections . Using a survey of individual wild-collected flies , we give the first quantitative estimates of prevalence for 15 different viruses in D . melanogaster and its close relative D . simulans , and rates of co-occurrence with the Wolbachia bacterial endosymbiont . In addition , by examining publicly available RNA datasets , we catalogue the presence of these viruses in D . melanogaster stock lines and cell culture . Our results provide an unprecedented insight into the virus community of Drosophila , and thereby provide the evolutionary and ecological context needed to develop Drosophila as a model for virus research . We used metagenomic sequencing of ribosome-depleted total RNA to identify virus-like sequences in five large collections of wild-caught adult Drosophilidae . Three collections ( denoted E , K , and I ) were sampled from fruit baits in Kilifi ( Kenya ) and Ithaca ( New York , United States ) . The aliquots pooled for total RNA sequencing represented around 2 , 000 individuals of D . melanogaster , D . ananassae , D . malerkotliana , and Scaptodrosophila latifasciaeformis . Two collections ( S and T ) were sampled from fruit baits in southern England , and aliquots pooled for total RNA sequencing represented around 3 , 000 D . melanogaster ( <1% other Drosophila ) . In total 0 . 5% of all reads mapped to DAV , DCV , DMelSV , and Nora Virus ( S1 Fig and S1 Table ) . Viral read numbers varied dramatically between samples , with DCV absent from pool EIK and DAV absent from pool ST ( verified by qPCR; S2 Fig ) . Only three reads mapped to DXV , and no reads mapped to Drosophila Totivirus , Drosophila Birnavirus , or American Nodavirus ( S1 Fig and S1 Table ) . As DXV is reported to be a cell culture contaminant [14] , and the other unmapped viruses were described from cell culture [17] , this suggests that these viruses may be rare in wild flies . To discover novel viruses , we assembled all RNA-seq reads de novo using Trinity [50] and Oases [51] , and used the Basic Local Alignment Search Tool ( BLAST ) [52] against the Genbank protein reference sequences [53] to identify virus-like sequences . Raw de novo metagenomic contigs are provided in S2 Data . Virus-like contigs were supplemented with PCR and targeted Sanger sequencing to improve completeness , and in total we identified more than 20 partial viral genomes . Those sequences that could be unambiguously associated with D . melanogaster rather than other Drosophila species were provisionally named according to collection locations . Based on sequence similarity , these “BLAST-candidate” viruses included two Reoviruses ( “Bloomfield Virus” and “Torrey Pines Virus” ) , three Flaviviruses ( “Charvil Virus” , and two others ) , a Permutotetravirus ( “Newfield Virus” ) , a Nodavirus ( “Craigie’s Hill Virus” ) , a Negevirus , a Bunyavirus , two Iflaviruses ( “La Jolla Virus” and “Twyford Virus” ) , two Picorna-like viruses ( “Thika Virus” and “Kilifi Virus” ) , a virus related to Chronic Bee Paralysis Virus ( “Dansoman Virus” ) , a virus related to Sobemoviruses and Poleroviruses ( “Motts Mill Virus” ) , six Partitiviruses , and a Nudivirus ( “Kallithea Virus” ) . These novel viruses constituted a further 3 . 3% of RNA-seq reads , taking the viral total to 3 . 8% of all reads ( S1 Fig ) . Further details of the new viruses are given in Table 1 and S2 Table , and virus sequences have been submitted to Genbank as KP714070-KP714108 and KP757922- KP757936 . We note that fragments of Bloomfield Virus and Thika Virus were simultaneously identified in [54] , there denoted DRV and DUV , respectively . To place the virus sequences in a phylogenetic context , we subjected conserved regions to phylogenetic analysis along with known viruses and uncurated viral sequences from the NCBI Transcriptome Shotgun Assemblies [55] . Kallithea Virus , a DNA virus , was closely related to Nudiviruses from Drosophila innubila [56] and the beetle Oryctes rhinoceros [57] . Most of the new RNA viruses were relatives of known or suspected insect pathogens ( Table 1 and S3 Fig ) . For example , based on polymerase sequences , Torrey Pines Virus is distantly related to Aedes pseudoscutellaris reovirus and several lepidopteran Cypoviruses , while Dansoman Virus is related to Chronic Bee Paralysis Virus . Others were distantly related to arthropod viruses , but were close to uncurated transcriptome sequences ( Fig 1 ) . For example , Bloomfield Virus is closely related to transcriptome sequences from the flies Delia antiqua and Teleopsis dalmanni ( 63% AA identity in the replicase ) , and distantly related to Nilaparvata lugens reovirus . Similarly , Motts Mill Virus is related to the recently described Ixodes scapularis ( tick ) associated viruses 1 and 2 [58] , but is closer to uncurated transcriptomic sequences from three bees and the bobtail squid Euprymna scolopes . Together these sequences appear to represent a novel clade related to plant Sobemoviruses and Poleroviruses ( Fig 1 ) . Strikingly , the six unnamed Partitivirus-like polymerases were not related to any known insect pathogens ( known Partitiviruses are pathogens of plants and fungi ) , but were related to uncurated sequences from arthropod transcriptomes—again suggesting a novel lineage of insect viruses ( Fig 1 ) . The close relationship between the newly identified virus-like sequences and known arthropod viruses suggests that they are likely to be Drosophila pathogens , and some may be previously described viruses that lack sequence data . For example , Kilifi Virus or Thika Virus ( related to the Picornavirales; S3 Fig ) may correspond to DPV—which was reported to have a 25–30 nm particle and infect gut tissues [59] . Similarly , either Torrey Pines Virus or Bloomfield Virus could correspond to Drosophila F Virus [14] , Drosophila K Virus or other reoviruses [21] . However , as those viruses were described only from capsid morphology , density , and serology , it would be challenging to conclusively link them with the novel sequences presented here . Although phylogenetic analyses show that these sequences are viral in origin , they may represent Endogenous Viral Elements integrated into the Drosophila genome ( “fossil” viruses , or EVEs [60] ) , or they may derive from gut contents or surface contamination rather than active infections . To exclude the possibility that these sequences represent EVEs segregating in D . melanogaster , we mapped the raw genomic reads from 527 distinct D . melanogaster genomes [61] to our set of BLAST-candidate viruses and confirmed that no genome mapped at a rate high enough to be consistent with a genomic copy of any virus in that individual . As this test for EVE status was not possible for other Drosophila species , only sequences associated with D . melanogaster were named as viruses , and others ( which could potentially represent EVEs in other taxa ) were denoted “virus-like . ” To test if these sequences derive from active viral infections , we additionally sequenced all 17–29 nt small-RNAs from the EIK and ST pools , reasoning that virus-like sequences will only be processed into 21 nt siRNAs ( viRNAs , derived from replicative intermediates by Dicer-2 ) if they represent active infections within host cells [49 , 62] . In total ca . 7% of all 17–29 nt small-RNAs derived from DAV , DCV , DMelSV , and Nora Virus , and ca . 9% derived from the new “BLAST-candidate” viruses . As expected , for most viruses the viRNA size distribution was tightly centred on 21 nt reads and included reads from both the genomic and complementary strands , consistent with active viral infections processed by antiviral RNAi in Drosophila ( Figs 2 , S4 and S5 ) . Even if viral sequences do not represent EVEs or inactive contaminants , they could instead be active infections of Drosophila-associated microbiota rather than Drosophila . However , the 21 nt viRNAs observed for the majority of these viruses are inconsistent with viral infection of likely parasites or parasitoids such as hymenoptera , chelicerata , or nematodes , which have predominantly 22 nt viRNAs [63 , 64] . And , while 21 nt viRNAs could derive from viral infections of some fungi [65] or from eukaryotes with uncharacterised antiviral RNAi , in most cases the phylogenetic position of these viruses , high read numbers , and/or their appearance in laboratory fly stocks and cell culture ( below ) argue in favour of Drosophila as the host . In addition to the viral candidates identified by BLAST , we reasoned that contigs which lack BLAST similarity to reference sequences , but which display a signature of Dcr-2 processing ( high levels of 21–23 nt siRNAs and low levels of 25–29 nt piRNAs ) , may also be viral in origin ( Fig 3; This approach was very recently advocated in [54] , but see also [17 , 65] ) . Using these small RNA criteria we identified a list of “siRNA-candidate” contigs that are potentially viral in origin , but could not be placed within a phylogeny of known viruses . Several siRNA-candidate viruses identified in this way were subsequently attributable as fragments of the BLAST-candidate viruses by other means ( below ) . Of those that could not be attributed to viral genomes , two were provisionally named ( Chaq Virus and Galbut Virus ) and the remaining 57 contigs were submitted to Genbank as uncultured environmental virus sequences ( KP757937–KP757993 ) . These unnamed siRNA-candidate viruses contribute 0 . 2% of all RNAseq reads , and 1% of 17–29 nt small RNAs in the EIK and ST pools ( S1 Fig and S1 Table ) . As with the BLAST-candidate viruses , these siRNA-candidate viruses were absent from D . melanogaster genomic reads and their siRNAs were predominantly 21 nt in length and derived from both strands ( S6 Fig ) , again suggesting that they represent active viral infections of Drosophila . Although the siRNA-candidate viruses displayed no strong BLAST similarity to known viruses , Galbut Virus and siRNA-candidate 24 display weak similarity to Nilaparvata lugens Commensal X Virus ( a satellite virus with unknown helper [66] ) , and Galbut and Chaq viruses appear to be related to uncurated sequences present in a diverse set of arthropod transcriptome shotgun datasets ( Fig 1 and Table 1; S2 Table and S3 Fig ) . To test whether viruses vary in their small RNA profile , we analysed small RNAs from the EIK and ST pools , and from two libraries for each of the five collections ( E , I , K , S , and T ) , with and without “High Definition” ligation adaptors designed to reduce ligation bias [67] . Overall , the number of viRNA reads per RNAseq read varied substantially among viruses ( “viRNA ratio;” Figs 3 and S7 ) , with Twyford Virus and Motts Mill Virus giving rise to more than a 1 , 000-fold additional 21–23 nt viRNAs than DCV , and DMelSV giving rise to nearly 7 , 000-fold more ( S7 Fig ) . For viruses present in both EIK and ST , the viRNA ratios were highly correlated between pools ( rank correlation ρ > 0 . 99; S7 Fig ) , suggesting that they are repeatable . This may reflect differences in the proportion of non-replicating viral genomes ( e . g . , encapsidated viruses in the gut lumen ) which can contribute to RNAseq but are not actively processed by Dcr-2 . Alternatively , differences could result from the action of viral suppressors of RNAi ( VSRs ) , such as those encoded by DCV and Nora Virus [7 , 38] . The sizes and strand bias of small RNAs also varied substantially among viruses , although small RNA reads from the majority of RNA viruses were biased toward the positive strand ( 50%–70% ) and to 21 nt in length ( Fig 2 ) , as expected for virus-derived small RNAs in Drosophila [49] . Exceptions included Nora Virus , DCV , Kilifi Virus , and Thika Virus , which showed a stronger positive-strand bias ( 85%–95% of reads ) and a broad size range of positive-sense viRNAs peaking at 21 nt , with a wide “shoulder” from 23 to 27 nt ( Fig 2 ) . This size distribution is not seen in most DCV infections of cell culture [49] , and although 26–30 nt viral piRNAs have been reported in OSS cell culture [17] , the 25–28 nt reads identified here did not display the 5′ U bias expected of piRNAs ( Fig 2 ) [68] . As these four picorna-like viruses also displayed low numbers of viRNAs ( Fig 3; S1 Table and S4 Fig ) , this is consistent with a difference in the way that the viral genome and/or viRNAs are processed—perhaps reflecting a higher fraction of non-specific degradation products for these viruses . However , because we found that viRNA properties were reproducible across sequencing libraries ( S4 Fig ) , and because viRNAs were sequenced from large pooled samples composed of mixed infections , these profiles cannot result from idiosyncrasies of RNA extraction or library preparation . In contrast to the other RNA viruses , Twyford Virus displayed unusual viRNAs . Although Twyford Virus is an Iflavirus ( S3 Fig ) and thus has a positive sense ssRNA genome , the strand bias was strongly negative and the viRNAs peaked sharply at 22–23 nt ( cf . 21 nt for other RNA viruses ) . In addition , although other viruses displayed no strong 5′ base-composition bias except for a weak bias against 5′ G , most Twyford Virus viRNAs were 5′ U , as is seen for piRNAs ( but lacking the A at position 10 expected of Drosophila piRNAs; S5 Fig ) . This bias is not due to small sample size or low sequence diversity , as we saw >9 , 000 unique sequences , and 3 , 500 of those were seen more than once . Comparison with the virus genome showed the 5′ U bias was driven by differential production or retention of viRNAs , rather than subsequent editing or addition . Interestingly , the 3′ position was also slightly enriched for U ( S7 Fig ) , and a substantial fraction of 3′ U were non-templated , indicative of 3′ uridylation as seen in D . melanogaster and other species in the absence of the Hen-1 methytransferase [69] . These observations suggest that Twyford Virus may not be processed by the Drosophila Ago2-Dcr2 pathway , and could instead represent viral infection of an unknown eukaryotic commensal . Nevertheless , potential arthropod parasites such as chelicerata have not been reported to display this pattern of viRNAs , and although the 5′ U is reminiscent of the 21U piRNA of C . elegans and related nematodes [70] , neither Rhabditid nor Tylenchid nematodes could be detected by PCR in individual wild-caught D . melanogaster carrying Twyford Virus . Similarly , while 22 nt 5′-U small RNAs are known from the filamentous fungus Neurospora [71] , those are derived from the host genome and are not associated with viral infection . Thus , although speculative , if these 22–23U viRNAs cannot be explained by a non drosophilid host , then it is possible that they reflect a previously unrecognised tissue-specific phenomenon in Drosophila , or one associated with the action of a novel suppressor of RNAi ( e . g . , suppression of Hen-1 ) . The siRNA-Candidate 14 ( KP757950 ) shows a similar pattern ( 22–23 nt peak in viRNAs , 5′ U ) , suggesting it forms part of the same virus and/or is processed in the same way ( S6 Fig ) . We also identified many 21 nt viRNAs widely dispersed around the Kallithea Virus genome . This is consistent with an antiviral RNAi response against this DNA Nudivirus , as has been previously reported for Invertebrate Iridovirus 6 artificially infecting D . melanogaster [72] . As DNA viruses often encode miRNAs [73 , 74] , and miRNAs have been implicated in the establishment of latency in Heliothis zea Nudivirus [75] , we screened small RNAs from Kallithea Virus for potential virus-encoded miRNAs . One 22 nt RNA sequence was highly abundant ( S5 Fig ) , and is predicted to represent the 5′ miRNA from a pre-miRNA-like hairpin ( miRDeep2 [76]; S1 Text ) . The predicted mature 5′ miRNA ( AUAGUUGUAGUGGCAUUAAUUG ) represented >35% of all small RNAs derived from Kallithea Virus , while the 3′ RNA “star” sequence represented 0 . 3% of reads . This sequence was less highly represented , relative to 21nt viRNAs from Kallithea Virus , in an oxidised library , consistent with the absence of 2′O-methylation at the 3′ end , as expected for miRNAs in Drosophila ( S4 Fig ) [77] . The seed region displays no obvious similarity to known miRNAs ( although positions 5–17 are similar to dme-miR-33-5p ) , but a survey of potential binding sites in D . melanogaster using miRanda 3 . 3a [78] identified 522 genes with at least one potential binding site in the 3′-UTR ( miRanda score ≥150 ) . These were highly enriched for Gene Ontology terms that might be associated with viral function ( including , amongst others , Regulation of Gene Expression , Cell development , mRNA binding , and Plasma Membrane; S3 Table ) . This miRNA could alternatively regulate virus gene expression [75] , and 21 potential binding sites were identified in the Kallithea Virus genome . While predicted miRNA target sites include many false positives [79] , and experimental work would be required to confirm a biological role , it is interesting to note that Kallithea Virus’ closest relative ( Oryctes rhinoceros Nudivirus ) does not encode a detectable homolog of the miRNA and contains only four predicted binding sites ( miRanda score ≥150 ) . To detect the BLAST-candidate and siRNA-candidate viruses in previously published Drosophila studies , we mapped up to 2 million reads from each of 9 , 656 publicly available fly and cell culture RNAseq and small-RNA datasets to the new and previously described Drosophila viruses . Around 33% of “run” datasets contained viral reads above a threshold of 100 reads per million ( rpm ) , representing 39% of “samples” and 58% of submitted “projects” ( Figs 4 and S8 and S5 Data ) . The proportion of positive samples varied with log-threshold , so that 53% had at least one virus at ≥10 rpm , but 17% of runs had at least one virus at ≥1 , 000 rpm . These rates are slightly lower than , but not dissimilar to , previous estimates from serial passage of fly stocks , which found around 40% of fly stocks were infected [14] . BLAST-candidate and siRNA candidate viruses were both found , and by noting their co-occurrence , we were able to identify several siRNA-candidates as component parts of other virus genomes ( e . g . , segments of Bloomfield Virus and the second segment of Craigie’s Hill Virus were initially identified as siRNA Candidate Viruses ) . The presence of BLAST-candidate and siRNA-candidate viruses in laboratory cultures supports these as bona fide infections of Drosophila and demonstrates the utility of the viRNA signature as a marker for viruses . Based on the 100 rpm threshold , DAV ( 1 , 025 of 9 , 656 datasets ) , DCV ( 979 datasets ) , Nora Virus ( 629 datasets ) , Newfield Virus ( 483 datasets ) , FHV , Drosophila Totivirus , American Nodavirus , Drosophila Birnavirus , DMelSV , Thika Virus , Kilifi Virus , La Jolla Virus , Craigie’s Hill Virus , Bloomfield Virus , Chaq Virus and Galbut Virus were all present in public datasets ( Figs 4 and S8 and S5 Data ) . However , some viruses ( Kilifi Virus , Craigie’s Hill Virus , Chaq Virus , Galbut Virus , DMelSV ) were extremely rare , appearing in 12 or fewer datasets . It is widely known that Drosophila cell culture harbours many viruses [21] , and we identified multiple viruses and occasionally high numbers of viral reads in cell culture datasets . For example , reads mapping to nine different viruses were found in datasets SRR770283 and SRR770284 ( Piwi CLIP-Seq in OSS cells ) [80] and ≥70% of reads from SRR609669-SRR609671 were viral in origin ( total RNA from piRNA-pathway knock-downs ) [81] . Virus presence/absence for widely used Drosophila Cell cultures [82] is presented in S9 Fig . RNAseq datasets were also available for species other than D . melanogaster , and these , too , included virus-like sequences . We detected DAV in D . pseudoobscura , D . virilis , D . bipectinata , D . ercepeae and D . willistoni , DCV in D . simulans , D . ananassae and D . mojavensis , Nora Virus in D . simulans , D . ananassae , and D . mojavensis , Thika Virus in D . virilis and D . ficusphila , Kilifi Virus in D . bipectinata , La Jolla Virus in D . simulans , and Bloomfield Virus in D . virilis . Given the presence of viruses in public RNAseq and siRNA datasets , we selected a subset of 2 , 188 datasets to perform viral discovery by de novo assembly . In adult D . melanogaster this identified a novel Picorna-like virus ( present in three datasets among 9 , 656 ) and a novel Negevirus ( present in 10 datasets among 9 , 656 ) . We have provisionally named these Berkeley Virus and Brandeis Virus , respectively ( Table 1 , S2 Table and S6 Data ) . The survey also identified a novel Totivirus and several ( possibly fragmentary or non-coding ) Reovirus segments in cell culture , at least one of which is widespread ( 214 datasets; Figs 4 and S8 and S5 Data ) . The small number of novel viruses we identified in fly stocks over and above those described previously , may suggest that few further Drosophila viruses remain to be found regularly infecting laboratory stocks . To infer viral prevalence and distribution in wild flies we used reverse transcription PCR ( RT-PCR ) to assay for the presence of 16 different viruses in a total of 1 , 635 D . melanogaster and 658 D . simulans adults sampled from 17 locations across the world ( Fig 5 and S4 Table ) . Excluding siRNA-candidate viruses , the most prevalent virus in large samples of D . melanogaster was La Jolla Virus ( 12/16 locations , 8 . 6% of flies averaged across locations ) and the rarest was Twyford Virus ( 1/16 locations , average 0 . 3% of flies ) . Of those detected in large samples of D . simulans , the most prevalent was Thika Virus ( 4/7 locations , average 4 . 5% of flies ) while the rarest was Motts Mill Virus ( 1/7 locations , average 0 . 2% of flies ) . Despite their high prevalence in the lab and presence in the metagenomic pools , DCV and Newfield Virus were not detected at any of the 17 locations , and Twyford Virus , DMelSV , Dansoman Virus , and Craigie’s Hill Virus were not detected in D . simulans ( Fig 5 and S4 Table ) . Although sampling locations varied substantially in overall viral prevalence ( in Athens GA >80% of D . melanogaster carried a virus; in Marrakesh less than 10% ) there was no clear geographic structure in viral prevalence ( Figs 5 and S10 ) , and for most viruses prevalence was not correlated between D . melanogaster and D . simulans from the same location . Excluding siRNA-candidate viruses , around 30% of D . melanogaster individuals and 13% of D . simulans individuals carried at least one virus , and over 6% of D . melanogaster individuals carried more than one virus ( S11 Fig ) . We are unable to explain the unusually high viral prevalence in D . melanogaster sampled from Athens ( Georgia , US ) —flies were separated to individual vials within a few hours , and D . simulans and D . melanogaster were netted together , but D . simulans did not show unusually high virus prevalence ( S10 Fig ) . In contrast to the other viruses , the novel siRNA-candidate viruses Galbut Virus and Chaq Virus often displayed extremely high prevalence . In D . melanogaster , Galbut Virus ranged from 13% in Plettenberg to 100% in Accra , and Chaq Virus from <5% in Edinburgh to 35% in Porto . In D . simulans , Galbut Virus ranged from 57% ( Athens ) to 76% ( Torquay , Australia ) , although Chaq Virus was not highly prevalent ( only 2/7 locations , at low prevalence ) . These rates are much higher than for the other viruses , and the inclusion of siRNA candidate viruses in overall infection rates brings many populations to ≥70% of flies carrying at least one virus ( Figs 5 and S11 ) . The high prevalence of these “siRNA-candidate” viruses is surprising and could perhaps imply an alternative ( non-virus ) origin for the sequences . However , we believe a viral origin is supported by a combination of the viral-like siRNA signature , their absence from D . melanogaster genomic reads , their close relationship to unclassified insect transcriptome sequences ( Figs 1 and S3 ) , and their occasionally low prevalence ( S10 Fig ) . For DMelSV , which is the only virus previously surveyed on a large scale [15] , our data agree closely with earlier estimates based on other assays: here 4 . 6% of D . melanogaster infected versus 2 . 8% in [22] and 5 . 0% in [23] ( DMelSV is absent from D . simulans ) . Nevertheless , our estimates more generally should be treated with some caution as RT-PCR assays are unlikely to be reliable for all virus genotypes , leading to PCR failure for divergent haplotypes and thus potentially underestimation of prevalence . Virus prevalence was strikingly different between publicly available RNA datasets and our field survey ( Fig 4 ) . For example , we identified Newfield Virus and DCV in many fly stocks and cell cultures but very rarely in the field ( Fig 4; also compare Figs 5 and S8 ) , whereas Kallithea Virus and Motts Mill Virus were common in the wild ( 4 . 6% and 6 . 7% global average prevalence in D . melanogaster; Fig 5 ) but absent from DNA and RNA public datasets . The case of DCV is particularly striking as early surveys assayed by serial passage in laboratory cultures suggested DCV may be common in the field [14] , as did PCR surveys of recently established laboratory stocks [83] . In the case of Newfield Virus and DCV , a downward collection bias ( for example , if high titre flies do not get collected ) may explain the result . However , for viruses that have higher prevalence in the field than the lab such as Kallithea Virus and Motts Mill Virus , and also the siRNA candidates Galbut Virus and Chaq Virus , it seems likely that differences in ( e . g . ) transmission ecology between the lab and field may explain the disparity . Infection with the bacterial endosymbiont Wolbachia pipientis has previously been shown to confer protection against secondary infection by some RNA [9 , 10] but not DNA [84] viruses in insects . We therefore surveyed Wolbachia prevalence by PCR in the wild flies , and tested whether Wolbachia was correlated with virus presence . We found Wolbachia at detectable levels in all populations , ranging from 1 . 6% of individuals ( Accra ) to 98% ( Edinburgh ) in D . melanogaster and from 82% ( Plettenberg ) to 100% ( Marseille ) in D . simulans ( Figs 5 and S10 ) . As expected [85] , overall Wolbachia prevalence was higher in D . simulans ( approximately 90% ) than in D . melanogaster ( approximately 50% ) . We could not detect any consistent correlation across populations between the prevalence of Wolbachia and that of any virus ( S12 Fig ) , nor could we detect any association between Wolbachia and viral infection status within populations ( contingency table tests with p-values combined across populations using Fisher’s method; all p > 0 . 05 ) . While this may indicate that Wolbachia-mediated antiviral protection is not an important determinant of viral infection in the field , our relatively small sample sizes ( median n = 63 flies per location ) mean that our power to detect an interaction is low . In addition , it is unclear whether a positive or negative association should be expected , as some Wolbachia-host combinations appear to protect via increased tolerance rather than increased resistance [86] , not all strains are highly protective [87] , virus titre may correlate with Wolbachia titre even if there is no presence–absence relationship , and for at least one DNA virus , resistance is decreased [84] . To provide a preliminary examination of the evolutionary and ecological dynamics of previously described and novel viruses we used a time-sampled Bayesian phylogenetic approach to infer dates of common ancestry and substitution rates [88 , 89] . To allow for differences in data coverage across ( partial ) viral genomes , and to allow for potential recombination , we divided alignments into blocks . These blocks were permitted to vary in mutation-rate ( substitution-rate ) and date parameters . For Nora Virus and DAV , we were able to use early genomic samples [20 , 90] to aid time calibration , and we estimated the all-sites mutation rate for these viruses at approximately 4 to 8 × 10−4 nt-1 yr-1 ( Nora Virus; range across blocks , posterior medians ) and 2 to 6 × 10−4 nt-1 yr-1 ( DAV ) . However , short sampling timescales means that these rates are estimated with low precision , resulting in a spread of 95% highest posterior density ( HPD ) credibility intervals across blocks of 2 to 12 × 10−4 nt-1 yr-1 and 1 to 9 × 10−4 nt-1 yr-1 , respectively . This range falls within the expected range for single-stranded RNA viruses [91] . These mutation rates imply dates for the most recent common ancestors of DAV and Nora Virus ( Most Recent Common Ancestor [MRCA] for known extant lineages ) of approximately 100–200 years ago ( range across blocks; range of credibility intervals across blocks 70–320 years ) and 50–300 years ago ( range across blocks; range of credibility intervals across blocks 33–630 years ) , respectively . This rapid movement of viruses is consistent with the rapid global movement that has been seen in transposable elements [92] , Wolbachia [93] , and even Drosophila genomes [94] . Nevertheless , inferred dates of a few hundred years in the past should be treated with caution , as weak constraint can lead to substantial underestimates of the true time to common ancestry [95] . For the other nine RNA viruses analysed , sample sizes were smaller and dated samples were only available from a 5 year window , making unconstrained estimates of mutation rate and MRCA dates difficult so that quantitative estimates for these viruses should be treated with caution ( results inferred using a strong prior on mutation-rate are presented in S13 Fig and S8 Data ) . Viruses detected in both D . melanogaster and D . simulans might either be reciprocally monophyletic , with substantial divergence between host-specific lineages , or the hosts may be distributed across the virus phylogeny , if between-host movement is common . By incorporating information on host species ( D . melanogaster versus D . simulans ) in the phylogenetic analysis , we were able detect past movement of RNA viruses between these host species ( e . g . , Fig 6 ) and obtain rough estimates of host switching rate [88 , 89] ( S13 Fig and S8 Data ) . These rate estimates have very low precision , associated with the relatively broad posterior estimates for substitution rate , and should be treated with caution . Nevertheless , there appeared to be substantial variation in host-switching rate amongst the seven RNA viruses that were detected in both hosts , with Chaq Virus showing the lowest rate and La Jolla Virus the highest ( Fig 6; S13 Fig and S8 Data ) . A similar analysis of inter-continental geographical movement suggests substantial genetic structuring between geographic regions may be present in some viruses ( S13 Fig and S8 Data ) , although again further time-sampled data are required for the analyses to provide useful precision . To quantify patterns of selection acting on protein-coding sequences , we applied a phylogenetic approach to infer relative rates of synonymous ( dS ) and non-synonymous ( dN ) substitution [96] in 11 of the viruses ( quantified as dN-dS or dN/dS; S13 Fig , S9 Data , and S14 Fig ) . For most viruses , sample sizes were insufficient to obtain robust inferences of selection; however , in general the protein sequences were strongly constrained ( S13 Fig ) . DAV and Nora Virus , which had the most comprehensive sampling , showed among the lowest mean dN/dS ratios ( 0 . 25 and 0 . 24 , respectively ) , although only Nora Virus displayed codons with strong evidence of positive selection . There were no clear patterns of dN/dS variation across genes within DAV or Nora virus genomes ( S14 Fig ) , although three of the four positively selected codons were in the Nora Virus capsid [97] , perhaps indicative of host-mediated selection . There was no evidence for positive selection or an elevated dN/dS ratio in the viral suppressor of RNAi encoded by Nora Virus [38] , despite apparently rapid host-specialisation in this gene [12] and the rapid adaptive evolution seen in its antagonist , Drosophila Ago2 [47] . Nevertheless , within-species sampling may be unlikely to detect species-wide selective pressure [98] and the disparity in evolutionary timescales of host and virus may mean that even strong reciprocal arms races are not reflected in elevated viral dN/dS ratios ( i . e . , the time of common ancestry for the viruses is too recent for the host to have driven multiple substitutions within the sampling timeframe [99] ) . We have identified around 20 new viruses associated with D . melanogaster in the wild and shown that some of them are common in the laboratory and the field . A substantial fraction of the virus lineages described here are newly ( or only recently [33 , 58] ) reported to infect arthropods . These include the novel “siRNA candidate” viruses with uncertain affiliation , and new lineages related to Partitiviruses , Sobemoviruses and Poleroviruses , and Picornavirales . The presence of these viruses in cell culture and laboratory stocks , their absence from fly genomes , and the presence of virus-derived 21 nt small RNAs all support the majority as bona fide Drosophila infections . Drosophila-associated viruses include many with positive sense RNA genomes ( Dicistroviridae , Permutotetraviridae , Flaviviridae , Iflaviridae , Nodaviridae , Negeviruses , and others ) , negative sense RNA genomes ( Rhabdoviridae; relatives of Bunyaviridae ) , and double-stranded RNA genomes ( Birnaviridae , Totiviridae , Partitiviridae , Reoviridae ) , but few with DNA genomes ( only the Kallithea and Drosophila innubila Nudiviruses [56] to date ) and no retroviruses ( cf . “Errantivirus” endogenous retroviruses [100]; retro-elements or retrotransposons that are usually transmitted as genomic integrations ) . It therefore seems increasingly unlikely that the apparent wealth of RNA viruses and paucity of retroviruses and DNA viruses in Drosophila represents a sampling artefact , and this may instead reflect underlying Drosophila ecology or immune function . This newly discovered diversity of Drosophila viruses will prove valuable for the use of D . melanogaster as a model for viral infection and antiviral immunity . First , these data will facilitate the curation and management of laboratory stocks and cell cultures . Second , they will facilitate the isolation and culture of viruses for future experimental work , including natural pathogens of Drosophila that are closely related to medically or economically important viruses such as Flaviviruses and Bunyaviruses , or Chronic Bee Paralysis Virus . Third , our analyses of virus prevalence , distribution , and dynamics will provide an informed evolutionary and ecological context to develop D . melanogaster and its relatives as a model system for viral epidemiology and host-switching . In combination , we hope this work will provide a key reference point for future studies of Drosophila-virus interaction . For metagenomic viral discovery , five large collections of Drosophila melanogaster and other sympatric drosophilid species were made in 2010 by netting flies from fruit bait . For United Kingdom samples ( denoted S and T ) , flies were morphologically screened to exclude species other than D . melanogaster . For non-UK samples ( denoted E , I , and K ) , flies were screened to bear a superficial similarity ( small , pale , stripes ) to D . melanogaster . Samples E and K , each of ca . 700 adult pale-bodied Drosophilidae , were collected in Kilifi ( Kenya ) by JA in July 2010 . Sample I of ca . 650 adult pale-bodied Drosophilidae were collected in Ithaca ( New York , US ) by BPL in August 2010 . Sample S of ca . 1 , 250 adult D . melanogaster were collected in Sussex ( UK ) by DJO and EHB in August 2010 . Sample T of ca . 1 , 700 adult D . melanogaster were collected in Twyford ( UK ) by DJO and Alasdair Hood in July 2010 . In each case , flies were maintained in groups of up to 200 for 5–10 d on a hard-agar/sugar medium before maceration under liquid nitrogen . RNA extraction was performed separately on each of the five samples . For analyses of prevalence and sequence evolution , collections of D . melanogaster and D . simulans were made at the following locations between November 2008 and October 2012: Thika ( Kenya ) by John Pool in January 2009; San Diego ( California , US ) by DJO in November 2008; Edinburgh ( UK ) by DJO in August 2009; Athens ( Greece ) by Natasa Fytrou in June 2009; Marseille ( France ) by Nicolas Gompel in July 2009; Ithaca ( New York , US ) by BPL in September 2009; Athens ( Georgia , US ) by PRH in September 2009; Sussex ( UK ) by DJO in August 2009; Two locations in Accra ( Ghana ) by CLW and JA in January 2010; Plettenburg ( South Africa ) by Francis Jiggins in January 2010; Edinburgh ( UK ) by CLW in September 2009; Marrakech ( Morocco ) by CLW in September 2010; Montpellier ( France ) by PRH in September 2010; Torquay ( Australia ) by BL in February 2011; Lisbon ( Portugal ) by DJO in October 2012; Porto ( Portugal ) by DJO in October 2012 ( S4 Table for details ) . In each case flies were aspirated or netted from fruit bait at intervals of 24 hours or less , and maintained individually in isolation for up to 10 days on a sugar/hard-agar medium prior to freezing at -80°C . In addition , a small number of laboratory-stock flies were provided by David Finnegan ( University of Edinburgh ) in February 2008 and February 2010 . All raw reads have been submitted to the Sequence Read Archive under project accession SRP056120 . RNA was extracted using Trizol ( Life Technologies ) and DNAse treated ( Life Technologies ) according to the manufacturer’s instructions . For metagenomic RNAseq and small-RNA sequencing , aliquots of the samples E , I , K , S , and T were initially mixed into two pools: Dmel/UK ( denoted ST ) and mixed-species/non-UK ( EIK ) . RNAseq was performed by Edinburgh Genomics ( Edinburgh ) . Following ribosome-depletion using two cycles of RiboMinus ( Life Technologies ) , around 40 million high-quality paired-end 100 nt Illumina reads were generated from a single sequencing lane of each library ( accessions SRR1914484 and SRR1914527 ) . Small-RNA sequencing from the same two pools was performed with a periodate oxidation step to reduce the relative ligation efficiency of miRNAs , which do not carry 3′-Ribose 2′O-methylation [77] , and sequenced by Edinburgh Genomics ( Edinburgh ) . An equivalent library , without periodate oxidation , was constructed and sequenced by BGI ( Hong Kong ) These libraries generated between 12 and 20 million small-RNA reads each ( SRR1914671 , SRR1914716 , SRR1914775 , SRR1914792 ) . After preliminary analysis , RNA aliquots of all five pools were mixed ( mix denoted EIKST ) and RNAseq was repeated by BGI ( Hong Kong ) following either double-stranded nuclease normalization ( around 65 million high-quality paired-end 90 nt Illumina reads; SRR1914412 ) , or ribosome depletion using Ribo-Zero ( Illumina; around 65 million high-quality paired-end 90 nt Illumina reads; SRR1914445 ) . Finally , individual small-RNA libraries were constructed for each of the five samples E , I , K , S , and T , without periodate oxidation using both NEBnext adaptors ( New England Biolabs ) ( datasets SRR1914946 , SRR1914952 , SRR1914955 , SRR1914957 , SRR1914959 ) , and equivalent “High Definition” adaptors ( datasets SRR1914958 , SRR1914956 , SRR1914954 , SRR1914948 , SRR1914945 ) , which seek to reduce ligation bias by incorporating random tetramers into the ligation adaptor [67] . These ten libraries were sequenced by Edinburgh Genomics ( Edinburgh ) . Any small RNAs produced by synthesis ( e . g . , nematode 22G small RNAs ) carry a 5′ triphosphate and are unlikely to be ligated by this protocol . To estimate species-composition of the UK ( ST ) and non-UK ( EIK ) pools , RNAseq datasets were quality trimmed using ConDeTri2 . 0 and mapped using Bowtie2 [101] to a 359 nt fragment of COI ( position 1781–2139 in D . melanogaster reference mtDNA ) drawn from multiple species . This identified the EIK RNA sample to have been a mixture of D . simulans ( 0 . 8% ) , D . hamatofila ( 1 . 2% ) , D . ananassae ( 12 . 4% ) , Scaptodrosophila latifasciaeformis ( 16% ) , D . melanogaster ( 24 . 1% ) , and D . malerkotliana ( 45 . 5% ) . It also identified a small amount of cross-species contamination in the D . melanogaster ST sample ( 0 . 4% D . simulans , 0 . 4% D . immigrans ) . To estimate the contribution of each of several potential sources of RNA , reads were also mapped to known Drosophila viruses , the D . melanogaster , D simulans , and D ananassae genomes , Wolbachia from D . melanogaster , D . simulans , D . willistoni , D . suzukii , and D . ananassae , and the bacteria Pseudomonas entomophila , Providencia sneebia , P . rettgeri , P . burhodogranariea , P . alcalifaciens , Gluconobacter morbifer , Enterococcus faecalis , Commensalibacter intestini , Acetobacter pomorum , A . tropicalis , and A . malorum ( S1 Table ) . The percentage of mapping reads was quantified relative to total high-quality reads . All paired-end RNAseq datasets were combined and de novo assembled using three different approaches . First , all reads were assembled using Trinity ( r2013-02-25; [50] ) . Second , data were digitally normalised using the “normalize-by-median-pct . py” script from the Khmer package , and then assembled using Trinity . Finally , this normalised dataset was also assembled using the Oases/Velvet pipeline ( version . 0 . 2 . 08; [51] ) . Contigs are provided in S2 Data . Using the longest inferred contig for each putative locus and a minimum contig length of 200 nt , contigs were compared to the Genbank protein reference sequence database using BLASTx [52] with default search parameters and an e-value threshold of 1 × 10−10 to identify the top 10 hits . Contigs with no BLAST hit at this threshold were translated to identify the longest open reading frame in the contig , and the resulting amino acid sequence compared to Genbank protein reference sequence using BLASTp with default parameters and an e-value threshold of 1 × 10−5 to identify the top 10 hits . For each contig with BLAST hits , the phylogenetic hierarchy of the best hits was traversed upward to identify the lowest taxonomic classification displaying a 75% majority taxonomic agreement , and this was used to guide subsequent cross-assembly and manual curation using the SeqManPro assembler ( DNAstar ) . The resulting BLAST-candidate virus sequences were then used as targets for ( RT- ) PCR and Sanger sequencing confirmation ( both on discovery samples EIKST , and on individual flies ) , with primer design guided by assuming synteny with close relatives . Final sequences for phylogenetic analysis and submission to NCBI were generated either by PCR and Sanger sequencing , or from a majority consensus derived by re-mapping RNA-seq reads to the combined de novo/PCR contig ( sequences have been submitted to Genbank as KP714070-KP714108 and KP757922-KP757936 ) . The Drosophila Dcr2-Ago2 RNAi pathway processes double-stranded RNA derived from viruses and transposable elements into 21 nt siRNAs [7 , 8 , 102] . In the case of viruses , Dcr2 may process both replicative intermediates and fold-back structures . However the presence of replicative strand reads ( i . e . , negative strand reads in positive sense RNA viruses ) demonstrates at that replicative intermediates are a major substrate for Dcr2 . Thus , the presence of virus-derived 21 nt siRNA sequences from both strands is diagnostic of an antiviral response against replicating viruses . The piRNA pathway also generates small RNAs ( 25–29 nt piRNAs from transposable elements [68] ) , however , although viral piRNAs are found in Drosophila OSS cell culture expressing Piwi [17] , these have not been detected in whole flies . Therefore the presence of large numbers of 21 nt small RNAs , relative to total RNA or to 25–29 nt small RNAs , provides corroborating evidence that virus-like sequences are recognised by the Drosophila immune system . In addition , we reasoned that this signature of 21 nt siRNAs could also be used to identify viral sequences amongst the substantial fraction of contigs that displayed no BLAST similarity to any sequence in the NCBI reference protein database . To quantify the siRNA profile of the de novo contigs we mapped the 19–29 nt small RNAs from the ST and EIK datasets to all de novo contigs , recording all potential mapping locations . As a proxy for total RNA , we also re-mapped combined EIKST RNAseq data using the dataset generated by BGI ( Hong Kong ) with Ribo-Zero rRNA depletion . For each contig this resulted in a count per unit length of the number of 21–23 nt siRNAs , the number of 25–29 nt putative piRNAs , and the number of RNAseq reads . Using the well-studied Drosophila viruses DCV , DAV , Nora virus , and DMelSV as threshold a guide , we then used a high ratio of siRNA:RNAseq reads and a high ratio of siRNA:piRNA reads to corroborate the BLAST-candidate viruses outlined above , and to propose novel siRNA-candidate viruses amongst contigs lacking BLAST hits ( Fig 3 ) . Final siRNA-candidate sequences for phylogenetic analysis and submission to Genbank were generated either by RT-PCR and Sanger sequencing , or from a majority consensus derived by re-mapping RNA-seq reads to the each contig ( Submitted to Genbank as KP757937–KP757993 ) . We used translations from the novel viruses to perform a BLASTp similarity search of the Genbank protein reference database to identify relatives and suitably conserved genomic fragments for phylogenetic inference . In addition , we searched the Genbank transcriptome shotgun assemblies ( TSA: database tsa_nt ) using tBLASTn [52] to identify potential virus sequences ( i . e . , sequences currently unannotated as viral ) in recently generated high-throughput transcriptomes . Protein sequences from known viruses , novel viruses , and TSA candidate viruses were then aligned using T-Coffee “psicoffee” [103] , and some poorly aligned regions at the 5′ and 3′ ends of the selected fragment were manually identified and trimmed . Phylogenetic trees were inferred using MrBayes [104] under a protein substitution model that allowed model jumping between widely advocated amino acid substitution models , and gamma-distributed rate variation among sites . Two independent MCMC chains were run , sampling every 100th step until the posterior sample of tree topologies reached stationarity ( i . e . , standard deviation in split frequencies between chains dropping to <0 . 01 ) and the effective sample size of all parameters was >300 ( after 25% burn-in ) . Maximum clade-credibility consensus trees were prepared by combining both MrBayes chains using TreeAnnotator from the BEAST package [89] . For most of the novel RNA viruses , the phylogeny was inferred from the RNA polymerase , which tends to be highly conserved , but for Kallithea Virus ( a DNA Nudivirus ) we selected five loci that have previously been used in Nudivirus phylogenetics [105] and the phylogeny was inferred from their concatenated alignments . To quantify the relative number of small RNAs produced from each virus , RNAseq reads and small RNAs from the EIK and ST pools were re-mapped using Bowtie2 [101] . To quantify the length distribution and 5′ base-composition , the small RNAs from 14 different sequencing runs were mapped to known miRNAs ( mirBase [106] ) , known transposable elements ( Flybase [107] ) , all viral genomes , and other parts of the D . melanogaster reference genome ( Flybase ) . The 14 siRNA datasets comprised: each of E , I , K , S , and T ligated using the NEBnext protocol according to manufacturer’s instructions; each ligated using the NEBnext protocol replacing oligos with “High Definition” equivalents that incorporate four additional random bases [67]; and the EIK and ST pools sequenced by Edinburgh Genomics with and without periodate oxidation ( BGI ) to reduce miRNA ligation efficiency [77] . To identify potential miRNA sequences in the large dsDNA genome of Kallithea Virus , all small-RNAs were combined and mapped to the Kallithea Virus genome using Bowtie2 , and novel miRNAs were inferred from read numbers and predicted hairpin-folding using miRDeep2 with default parameters [76] . Several viruses are known to be endemic in laboratory stocks and cell cultures of D . melanogaster [14 , 21] , and we expected that many of the novel viruses identified here may also be detectable in those datasets . We therefore searched the European Nucleotide Archive for Illumina , SOLiD , and Roche454 sequencing “run” datasets that derive from the Drosophilidae and subordinate taxa . This resulted in 14 , 123 DNA sequencing “run” datasets and 9 , 656 RNAseq and siRNA sequencing “run” datasets ( as of 9 May 2015 ) , from each of which we were able to download and map the first 2 million forward-orientation reads to new and previously known RNA viruses using Bowtie2 [101] ( Bowtie for SOLiD colour-space reads ) . To allow for mismapping , an arbitrary threshold of 100 reads per million ( rpm ) was chosen as a detection limit for viruses to be recorded as present . This threshold suggests at least one virus was present in 33% of runs , but the proportion of positive samples decreased with increasing log ( threshold ) , suggesting that true infection rates may be higher . Across the RNAseq and siRNA datasets the co-occurrence of BLAST-candidate viruses with siRNA-candidate viruses allowed us to tentatively assign some of the siRNA-candidate viruses as components of BLAST-candidate virus genomes ( e . g . , Bloomfield Virus and Craigie’s Hill Virus ) . To detect additional novel RNA viruses present in these public datasets , but absent from the literature or from our metagenomic sequencing of wild flies , we further selected 2 , 188 Illumina RNAseq datasets for viral discovery . The first 20 million read pairs ( or as many as were available , if fewer ) from each dataset were mapped to the D . melanogaster genome ( Bowtie2 , local alignment ) , and unmapped reads digitally normalised ( Khmer ) and de novo assembled using Trinity [50] as above . Resulting contigs of >3 . 5 kbp in length were compared to known viral proteins using BLASTx , and the top hits manually curated to create a second group of BLAST-candidate viruses ( such sequences cannot be accepted by Genbank , and are instead provided in S6 Data ) . We reasoned that the absence of virus-like sequences from D . melanogaster genomes can provide evidence that these sequences do not represent “fossilised” endogenous viral elements ( EVEs ) . However , if only a minority of individuals carry a particular EVE , and if genomes are derived primarily by mapping to a reference , the EVEs may not appear in assembled datasets . We therefore mapped all reads ( forward and reverse reads mapped separately ) from 779 Drosophila NEXUS genomic read datasets ( representing 527 D . melanogaster genomes [61] ) to known Drosophila viruses , BLAST-candidate viruses , and siRNA-candidate viruses . Because a small number of reads may mismap to viral sequences , potentially leading to false inference of an EVE , read numbers were quantified relative to target sequence length , and normalised to a 7 kbp single-copy region of the D . melanogaster genome ( as a segregating EVE present in a genome should generate reads at a rate approaching that of other loci in that genome ) . We used RT-PCR to survey wild populations for the presence of four previously published Drosophila viruses ( DCV , DAV , DMelSV , and Nora Virus ) , ten novel BLAST-candidate viruses ( Bloomfield Virus , Craigie's Hill Virus , Dansoman Virus , Kallithea Virus , La Jolla Virus , Motts Mill Virus , Newfield Virus , Thika Virus , Torrey Pines Virus , and Twyford Virus ) , and two novel siRNA-candidate viruses ( Chaq Virus and Galbut Virus ) . In addition , we surveyed the same RNA extractions by RT-PCR for the presence ( but not titre ) of Wolbachia pipientis , using primers for the Wolbachia surface protein ( wsp_F1 GTCCAATARSTGATGARGAAAC and wsp_R1 CYGCACCAAYAGYRCTRTAAA ) . This survey comprised a total of 1 , 635 D . melanogaster individuals and 658 D . simulans individuals sampled across 17 locations ( see “Sample Collections” above ) , with between 2 and 236 individuals assayed for each species at each location ( median n = 63 ) . RNA extractions and random hexamer reverse transcription was performed on single flies ( males and females ) or small bulks of 2–20 flies ( usually ≤10 males , with species initially assigned based on morphology ) . Fly species was confirmed using a competitive 3-primer PCR for the single-copy nuclear gene Ago2 , which gives a 300 bp product in D . simulans and a 400 bp product in D . melanogaster ( Primers: Mel_SimF CCCTAAACCGCAAGGATGGAG , Dsim303R GTCCACTTGTGCGCCACATT , Dmel394R CCTTGCCTTGGGATATTATTAGGTT ) . For each of these 16 viruses we designed up to four PCR assays based on conserved positions in the metagenomic RNAseq datasets , and these were tested on the E , I , K , S , and T metagenomic discovery RNA pools and on a selected subset of wild-collected individual flies . Those primer sets that consistently gave repeatable bands in agreement with other assays for the same virus , and which could be sequenced to confirm band identity , were selected for use as assays for the wider geographic sampling ( some designed as multiplex assays; details of primers and conditions are given in S2 Text ) . To allow for the possibility of PCR failure caused by primer-site variants not detected in virus discovery or preliminary sequencing , we designated flies as nominally positive for a virus if at least one assay for that virus resulted in an RT-PCR product . Nevertheless , the potential for unknown sequence variants to prevent PCR amplification means that our estimates of viral prevalence could be considered lower bounds . As only one fly needs to carry a virus for the bulk to test positive for that virus , the proportion of “positive” single-fly and/or small-bulk assays does not directly estimate the frequency of virus carriers . We therefore used a maximum likelihood approach to infer the underlying prevalence ( and its log-likelihood confidence intervals ) . Briefly , we assumed that for each species/location combination there was a single underlying prevalence ( i . e . , fraction of flies carrying detectable virus ) and that the number of flies testing positive or negative by PCR assay would be binomially distributed given this prevalence . Then , given the number of flies included in each small bulk , and the number of small bulks or single flies testing positive , we searched across the range of possible prevalence values ( zero to one ) to identify the prevalence which maximised the likelihood of the observed set of positive/negative PCR assays . It is possible that differences in the sex ratio between individual-fly assays ( majority female ) and small-bulk assays ( majority male ) could make this approach misleading if prevalence differs between males and females . However , based on likelihood ratio tests ( identical prevalence versus small bulks differ from single flies ) we were unable to identify any widespread significant differences between the sampling regimes . Specifically 4 . 4% of tests were nominally “significant” at α = 5% ( as expected ) , and only two were “significant” using the Benjamini-Hochberg approach to limit the false discovery rate to 5% ( Galbut Virus in Accra-1 , Wolbachia in Plettenberg ) . Eleven of the viruses with high prevalence ( DAV , Nora Virus , DMelSV , Craigie’s Hill Virus , Dansoman Virus , La Jolla Virus , Motts Mill Virus , Thika Virus , Torrey Pines Virus , Chaq Virus , and Galbut Virus ) were selected for further sequence analyses . PCR products amplified during the prevalence survey were supplemented with additional amplicons and sequenced in both directions using BigDye V3 . 1 ( Life Technologies ) . Reads were assembled into partial or near-complete genomes using SeqManPro ( DNAstar ) and all variable sites within or between chromatograms were examined by eye . Virus sequences that displayed substantial within-sample ( within-fly or small bulk ) heterozygosity for any amplicon , and therefore deemed likely to represent mixed infections , were discarded . Sequences were submitted to Genbank under accessions KP969454-KP970120 . Sequence alignments were divided into blocks , minimising missing data within blocks , and these alignment blocks were tested for evidence of recombination using GARD [108] from the HyPhy package under a GTR model of nucleotide substitution with inferred base frequencies and 4-category gamma-distributed rate variation between sites . Those blocks that displayed any evidence of recombination were further subdivided at the inferred recombination break points , to result in multiple alignment blocks for each virus . Alignment blocks ( containing no detectable recombination ) were subjected to phylogenetic analysis using BEAST [89] under a strict-clock model with HKY substitution using inferred base frequencies , and 4-category gamma-distributed rate variation between sites . Separate substitution models with relative clock rates were used for 1st , 2nd , and 3rd codon and for non-coding positions , with substitution models linked across alignment blocks . Time calibration information was incorporated using a uniform step prior on tip sample-dates based on the value and precision of sample freezing date , and molecular clock rate was unlinked between alignment blocks . As the limited sampling timespan meant that little rate information was available for most viruses , we chose to implement a strong prior on nucleotide clock-rate , based on published virus estimates [91] . For RNA viruses this was a log-normal distribution with mean 4 × 10−4 site-1 year-1 and 95% of the prior density between 1 × 10−4 and 11 × 10−4 . When greater time-sampling depth was available ( DAV and Nora Virus ) we ran models using a hierarchical prior for global nucleotide clock rate , in which clock rates for individual segments are drawn from a shared log-normal distribution . To infer rates of host-switching between D . melanogaster and D simulans , and rates of movement among broadly-defined sampling locations ( Europe and North Africa , sub-Saharan Africa , North America , Australia , Laboratory Stocks ) , tips were labelled by location and host species and treated as discrete traits [88 , 109] . All analyses assumed a standard neutral coalescent in a constant-size population , and other parameters took default values and priors . Two analyses were run for each model for up to 108 steps sampling every 50 , 000 steps , and convergence was inferred for all parameter values both visually from stationary distributions , and from effective sample sizes of >1 , 000 . In one case ( Thika virus ) , time and date parameters mixed poorly , and the analyses were re-run using a hard upper limit to the nucleotide clock rate for one alignment block ( 2 . 5 × 10−3 site-1 yr-1 ) . Posterior parameter estimates and maximum clade credibility trees were inferred from the two independent parallel runs combined after discarding 10% of steps of each as burn-in . Patterns of synonymous ( dS ) and nonsynonymous ( dN ) substitution , and evidence for positive selection ( dN-dS > 0 ) were inferred using the FUBAR [96] analysis from the HyPhy package . The in-frame viral coding sequence alignment blocks ( as used above in phylogenetic reconstruction ) were analysed based on maximum clade-credibility tree topologies inferred using BEAST , with substitution parameters and branch lengths inferred by HyPhy . FUBAR provides a fast approximate approach to infer dN-dS and sites under selection by pre-computing a grid of conditional likelihoods for dS and dN ( where the expectation of dS is 1 ) and reusing the precomputed values to infer the posterior distribution of dN-dS without recalculating likelihoods . The analysis used a 20 × 20 grid with ten independent MCMC chains each providing 1 , 000 subsamples from the posterior ( each 5 × 108 steps after 5 × 108 burn-in steps ) . Sites were considered to display evidence of positive selection if more than 95% of the posterior sample for that site supported dN>dS .	The fruit fly Drosophila melanogaster is extensively used as a model species for molecular biology and genetics . It is also widely studied for its evolutionary history , helping us understand how natural selection has shaped the genome . Drosophila research has been particularly valuable in determining how the insect immune system interacts with viruses and how co-evolution between hosts and viruses can shape the host immune system . Understanding insect–virus coevolution is important because some viruses—such as those which cause dengue and yellow fever in humans—also infect their insect vectors , and because the viruses of bees and other pollinators are implicated in pollinator decline . Although we have an increasingly good idea of how flies recognise and combat viral pathogens , we still have much to learn about the viruses they encounter and interact with in the wild . In this paper , we sequence all of the genetic material from a large collection of wild fruit flies and use it to identify more than 20 new viruses . We then survey individual wild flies and laboratory stocks to find out which viruses are common , which are rare , and which species of fruit fly they infect . Our results provide valuable tools and an evolutionary and ecological perspective that will help to improve Drosophila as a model for host–virus biology in the future .	[ "Abstract", "Introduction", "Results", "and", "Discussion", "Materials", "and", "Methods" ]	[]	2015	The Discovery, Distribution, and Evolution of Viruses Associated with Drosophila melanogaster
The level of available iron in the mammalian host is extremely low , and pathogenic microbes must compete with host proteins such as transferrin for iron . Iron regulation of gene expression , including genes encoding iron uptake functions and virulence factors , is critical for the pathogenesis of the fungus Cryptococcus neoformans . In this study , we characterized the roles of the CFT1 and CFT2 genes that encode C . neoformans orthologs of the Saccharomyces cerevisiae high-affinity iron permease FTR1 . Deletion of CFT1 reduced growth and iron uptake with ferric chloride and holo-transferrin as the in vitro iron sources , and the cft1 mutant was attenuated for virulence in a mouse model of infection . A reduction in the fungal burden in the brains of mice infected with the cft1 mutant was observed , thus suggesting a requirement for reductive iron acquisition during cryptococcal meningitis . CFT2 played no apparent role in iron acquisition but did influence virulence . The expression of both CFT1 and CFT2 was influenced by cAMP-dependent protein kinase , and the iron-regulatory transcription factor Cir1 positively regulated CFT1 and negatively regulated CFT2 . Overall , these results indicate that C . neoformans utilizes iron sources within the host ( e . g . , holo-transferrin ) that require Cft1 and a reductive iron uptake system . Pathogenic microbes such as the fungus Cryptococcus neoformans face a major challenge in acquiring iron during infection of vertebrate hosts . Free iron in tissues and fluids is maintained at extremely low levels due to the binding properties of the host proteins transferrin and lactoferrin . Moreover , vertebrates use iron deprivation as an important natural defense strategy against microbial pathogens [1] . For example , transferrin , which accounts for ∼1% of the total iron in the human body , is maintained at ∼33% saturation with iron in serum and effectively scavenges free iron [2] . Lactoferrin is similar to transferrin in structure and function but this protein retains iron in acidic conditions , such as at sites of inflammation , whereas transferrin binds iron at neutral pH [3] . Iron bound to heme is abundant in mammalian hosts but its availability during fungal pathogenesis is not yet clear because most of the heme is present in hemoglobin within erythrocytes . The data presented herein suggest that heme and transferrin may both be important iron sources for C . neoformans because each can support the growth of the fungus in culture . The mechanisms by which microbes acquire iron during infection are of considerable interest and are best characterized in bacterial pathogens [4] . For example , many pathogenic bacteria produce siderophores that bind to ferric iron with high affinity , and many are able to utilize ferritin , transferrin , lactoferrin , heme and heme-containing proteins . In many species , the mechanisms of iron acquisition have been elucidated in detail and preferences for specific iron sources during infection are being identified . For example , Staphylococcus aureus preferentially uses iron from heme rather than from transferrin during infection [5] . Mechanisms of iron acquisition are less well studied in pathogenic fungi . However , iron transport pathways have been well characterized in the model fungus Saccharomyces cerevisiae , which has at least two distinct high-affinity uptake systems . One is a reductive pathway in which ferric iron is reduced to ferrous iron by cell surface reductase activity with subsequent transport across the plasma membrane by the high-affinity iron permease ( Ftr1 ) –multicopper ferroxidase ( Fet3 ) complex [6–8] . The second high-affinity iron transport pathway uses siderophores from other organisms and transports iron bound to these molecules via cell surface transporters encoded by the ARN gene family [9–11] . Similar iron transport pathways exist in fungal pathogens of humans , although these have been less well studied . For example , both the reductive and siderophore iron uptake systems are found in Candida albicans and Aspergillus fumigatus . Two orthologous genes of the high-affinity iron permease FTR1 were identified in C . albicans and one of them , CaFTR1 , was shown to be required for systemic infection [12] . CaFTR1 also mediates iron acquisition from transferrin [13] . C . albicans is not able to synthesize siderophores but has siderophore transporters such as CaArn1/CaSit1 . However , the role of CaArn1/CaSit1 in an infected host may be minimal because a mutant only showed defects in epithelial invasion [14] . C . albicans does have hemolytic activity and utilizes heme and hemoglobin as iron sources [15–17] . Furthermore , cell surface proteins that bind heme and hemoglobin have been identified in C . albicans [18] . Additionally , this yeast uses hemoglobin as a signaling molecule to alter gene expression and to induce adhesion to host cells , and also to trigger the yeast to hyphae transition that is required for pathogenesis [19] . However , the question of whether heme or hemoglobin utilization plays a role in virulence in vivo is still unclear because none of previous studies showed virulence effects . Another animal pathogen , A . fumigatus also possesses the high-affinity iron permease FtrA , but this enzyme is not required for virulence . In contrast , the A . fumigatus SidA protein , which is responsible for siderophore synthesis , is essential for virulence [20 , 21] . Regulatory mechanisms that govern expression of the high-affinity iron permease have also been investigated . In S . cerevisiae , the global transcriptional activator Aft1 activates expression of FTR1 and other genes of the iron regulon [22 , 23] . Interestingly , orthologs of S . cerevisiae FTR1 in other fungi are negatively regulated by a conserved GATA-type zinc finger protein . Thus , the iron permease gene FIP1 is regulated by Fep1 in Schizosaccharomyces pombe and the permease gene FER2 is regulated by Urbs1 in Ustilago maydis [24 , 25] . We recently identified and characterized a global transcriptional regulator , Cir1 , in C . neoformans that shows sequence and functional similarities to Fep1 and Urbs1 . We found that Cir1 regulates many genes for iron acquisition including genes for putative high-affinity iron permeases , as well as genes involved in virulence in C . neoformans [26] . It has been suggested that the cAMP pathway influences iron uptake by controlling expression of the high-affinity iron permease in fungi . In S . cerevisiae , the catalytic subunit of cAMP-dependent protein kinase ( PKA ) , Tpk2 , negatively regulates expression of FTR1 and FET3 , and Tpk2 may indirectly control respiratory growth by negative regulation of iron uptake [27] . This connection between respiration and iron uptake is supported by the finding that the aft1 mutant fails to grow in the presence of non-fermentable carbon sources [28 , 29] . A connection between cAMP and iron uptake also exists in U . maydis because the expression of FER2 is positively regulated by the cAMP pathway [24] . Similar regulatory connections exist in C . neoformans because transcriptome studies demonstrated that genes for reductive iron uptake are differentially expressed in mutants lacking components of the cAMP pathway [30] . C . neoformans utilizes several transport systems to acquire iron from the environment and both high and low affinity iron uptake activities mediated by cell surface reductases have been detected [31] . Nonenzymatic reduction of ferric iron by a secreted reductant , 3-hydroxyanthranilic acid , and by melanin in the cell wall , may also contribute to iron acquisition [32] . C . neoformans reportedly does not produce siderophores but is capable of utilizing iron bound to siderophores secreted from other microorganisms [31] . The SIT1 gene , which encodes a siderophore transporter , was found to mediate siderophore utilization , but the gene is not required for virulence [33] . In this study , we identified and functionally characterized two candidate iron transporters in C . neoformans , CFT1 and CFT2 ( Cryptococcus Fe Transporter ) , which are orthologs of S . cerevisiae FTR1 . Mutants lacking CFT1 and/or CFT2 were constructed and characterized for their ability to use host iron sources and to cause disease . We found that CFT1 is involved in a reductive iron uptake pathway that is required for utilization of transferrin . CFT1 is also required for full virulence thus indicating that C . neoformans may preferentially utilize transferrin in a tissue specific manner , especially in the brain . CFT2 does not appear to play a role in iron acquisition under the conditions we tested but did contribute to virulence . We also demonstrate that CFT1 and CFT2 are differentially regulated by Cir1 and that transcript levels of both genes are influenced by the cAMP pathway . We initially searched the genome of the highly virulent strain H99 ( serotype A ) of C . neoformans to identify orthologs of the S . cerevisiae high-affinity iron permease Ftr1 [8 , 34] . Two highly conserved paralogous candidate genes were identified and the gene on chromosome 12 was designated CFT1 ( Cryptococcus Fe Transporter 1 ) while the gene on chromosome 3 was named CFT2 . We had previously identified CFT1 ( but not CFT2 ) as a candidate iron permease gene in transcriptional profiling experiments using serial analysis of gene expression ( SAGE ) and microarrays to examine the response to iron levels , cAMP signaling and experimental meningitis [30 , 35 , 36] . In the study of Lian et al . , ( 2005 ) , disruption of CFT1 in a serotype D strain background yielded a mutant with poor growth on low iron medium . Interestingly , CFT2 was also found among genes with induced expression upon phagocytosis [37] . A comparison of the predicted amino acid sequences showed 36% identity and 54% similarity between Ftr1 of S . cerevisiae and Cft1 , 36% identity and 53% similarity between Ftr1 and Cft2 , and 53% identity and 66% similarity between Cft1 and Cft2 . Similarities in genome arrangements exist for CFT1 and CFT2 in that CFT1 was paired with an adjacent putative ferroxidase gene designated CFO1 ( Cryptococcus Ferroxidase 1 ) and CFT2 was paired with the putative ferroxidase gene CFO2 . Both ferroxidases showed high similarity to the Fet3 protein of S . cerevisiae ( data not shown ) . CFT1 and CFO1 are transcribed bi-directionally as are CFT2 and CFO2 , with 791 bp and 2103 bp promoter regions , respectively ( Figure 1A ) . Measurements of basal transcript levels of the genes under low-iron conditions ( by RT-PCR ) showed that the transcripts of CFT1 and CFO1 were readily detected whereas CFO2 was less abundant and CFT2 was undetectable ( Figure 1B ) . We focused on the roles and regulation of CFT1 and CFT2 for this study . Severance et al . ( 2004 ) showed that S . cerevisiae Ftr1 has seven transmembrane domains with an orientation of N-terminal outside and C-terminal inside the cell [38] . Furthermore , their study suggested that two motifs , REXLE and DASE , are essential for iron transport and are strongly conserved among other fungal Ftr1 homologs . In this context , we analyzed the transmembrane ( TM ) helix number and TM topology of Cft1 and Cft2 in silico with the protein localization prediction program Localizome and found that both proteins have seven predicted TM domains and the same predicted topology as Ftr1 in S . cerevisiae [39] . Amino acid alignments with the S . cerevisiae proteins Ftr1 and the vacuolar iron transporter Fth1 , and also with the C . albicans Ftr1 and Ftr2 proteins , showed that both Cft1 and Cft2 possess the highly conserved motifs thought to be essential for iron transport ( Figure 1C ) . Transcript levels of the high-affinity iron permease genes are influenced by iron levels in S . cerevisiae , C . albicans and S . pombe [12 , 22 , 25] . Therefore , we tested whether CFT1 and CFT2 are regulated in a similar manner in C . neoformans . Wild-type cells were cultured in different concentrations of iron ( 0 , 10 , 100 and 1000 μM ) , and expression of CFT1 or CFT2 was measured by real-time RT-PCR . Transcript levels of both CFT1 and CFT2 were reduced as iron levels increased in cultures of the wild-type strain ( Figure 2 ) . Therefore , it appears that C . neoformans responds to iron deprivation by increasing transcription of these candidate iron permease genes . Note that although the transcript levels for CFT2 were influenced by iron , the significance for CFT2 function is unclear because the basal transcriptional level of CFT2 in low iron media was 100-fold lower than that of CFT1 ( see also Figure 1B ) . It may be that Cft2 has a minor or redundant function in iron uptake relative to Cft1 , that the gene plays a role in other growth conditions , or that Cft2 functions to transport iron from stores in the vacuole . Additionally , CFT1 and CFT2 transcript levels were reduced 10-fold and 3-fold respectively in the wild-type strain during growth in 1000 μM of iron , and this regulatory response may also suggest a minor or different role for CFT2 in iron uptake . We previously showed that iron permease genes are downstream targets of the iron regulatory transcription factor Cir1 in a C . neoformans serotype D strain by microarray analysis . Northern analysis also suggested that CFT2 transcript levels were influenced by Cir1 in a serotype A strain [26] . To further examine the regulation by Cir1 , we performed transcriptional analysis to determine whether both CFT1 and CFT2 are downstream of Cir1 in the serotype A strain . Transcription of CFT1 and CFT2 was monitored by real-time RT-PCR in the cir1 mutant grown in media with different concentrations of iron . Our results showed that CFT1 transcript levels were reduced in the cir1 mutant , indicating positive regulation by Cir1 ( Figure 2 ) . Furthermore , the CFT2 transcript was higher in the cir1 mutant indicating negative regulation by Cir1 ( Figure 2 ) and this result is consistent with our previous observations [26] . In addition , it appears that when Cir1 is deleted , the transcript levels of CFT1 and CFT2 are no longer responsive to iron concentrations compared with the wild-type strain . To characterize the functions of CFT1 and CFT2 , we generated mutants lacking each of the genes . Double mutants lacking both genes were also constructed to potentially unmask phenotypes hidden by redundancy ( see Materials and Methods ) . Reconstituted strains with a reintroduced wild-type copy of CFT1 or CFT2 at the original locus were also constructed and analyzed . Initial tests indicated that several independently generated cft1 and cft2 mutants showed similar growth rates in YPD medium at 30°C . These tests also revealed that the mutants did not differ from wild type with regard to capsule formation in low iron medium , melanin synthesis and growth at 37°C . The cft1 cft2 double mutants also did not display changes in capsule or melanin production , but did display a reduced growth rate in YPD medium ( Figure S1 ) . The wild-type strain , each single mutant and the reconstituted strains were tested for utilization of different iron sources in vitro . Strains were first grown in low-iron medium to reduce intracellular iron stores , and then were transferred to fresh low-iron medium and low-iron medium supplemented with the inorganic iron salt FeCl3 , apo-transferrin , holo-transferrin , heme or siderophores . Transferrin and heme were of particular interest because of their abundance as iron sources in mammals . The different concentrations of iron sources were prepared by serial dilution to ensure that any growth phenotypes observed could be correlated to a dependence on the iron source . The wild-type strain grew well with all iron sources and showed particularly robust growth in the presence of the iron-loaded siderophore , feroxamine ( Figures 3 and 4 ) . As expected , little or no growth was observed with apo-transferrin and the siderophore deferoxamine that lacks iron . The cft2 mutants behaved like the wild-type strain in all conditions ( Figures 3B and 4B ) . In contrast , the cft1 mutants showed reduced growth in the presence of inorganic iron ( FeCl3 ) or holo-transferrin , but not heme or feroxamine ( Figures 3A and 4A ) . Moreover , the cft1 mutant displayed growth defects for all concentrations of FeCl3 or holo-transferrin , a result consistent with the idea that Cft1 may be a high-affinity iron permease in C . neoformans . An analysis of the time course of growth for the strains also confirmed the growth defect of the cft1 mutant with FeCl3 as the iron source , and demonstrated that the cft1 cft2 double mutant behaved like the cft1 mutant in this assay ( Figure S2A ) . These findings suggest that CFT1 is required for the reductive iron uptake pathway in C . neoformans because ferric iron and iron from transferrin are believed to be transported via this pathway . Given that transferrin is a major iron carrier in the mammalian host , CFT1 may play a key role for iron acquisition during infection . These results also indicate that uptake of siderophore-bound iron is independent of CFT1 and CFT2 . The ability of the wild-type and mutant strains to take up iron was directly compared by assaying accumulation of iron from 55FeCl3 and from 55Fe-loaded transferrin ( Figure 5 ) . In the assay with 55FeCl3 , iron uptake by the cft1 mutant and the cft1 cft2 double mutant occurred at only 27% and 11% of the level found for the wild-type strain , respectively , suggesting that Cft1 plays a primary role in iron uptake in C . neoformans ( Figure 5A ) . Although Cft2 appeared to make a contribution based on the lower uptake of the double mutant , the cft2 mutant did not show a statistically significant reduction in uptake compared to the wild-type strain . Cft1 also played a major role in the acquisition of 55Fe from transferrin because negligible uptake was detected compared with the wild-type or reconstituted strains ( Figure 5B ) . The cft1 cft2 double mutant again behaved like the cft1 mutant and no influence on uptake was seen in the strain lacking CFT2 . Overall , these results highlight the role of Cft1 in iron uptake and are consistent with the poor growth of the cft1 mutant on iron sources that require reductive uptake . Loss of the high affinity iron uptake system would be expected to cause lower intracellular iron availability and to potentially influence the expression of iron-responsive genes . In addition , studies in S . cerevisiae indicate that modifying the expression of Fet4 , which is responsible for low-affinity iron uptake , can modulate expression of components of the high-affinity iron uptake system . Specifically , disruption of FET4 increases the activity of the high affinity uptake system and overexpression of FET4 decreases the activity [40] . In light of these observations , we investigated transcript levels for CFT1 in the cft2 mutant and for CFT2 in the cft1 mutant . We also analyzed levels of the SIT1 transcript in the cft1 and cft2 mutants to determine whether alteration of the reductive iron uptake pathway influenced the non-reductive siderophore uptake pathway . The wild-type strain and the cft1 and cft2 mutants were grown in media containing different concentrations of iron and the transcript levels of CFT1 , CFT2 and SIT1 were measured by real-time RT-PCR . We found that the transcript levels of CFT1 displayed similar and expected expression patterns in both the wild-type strain and the cft2 mutant; specifically , mRNA levels of CFT1 decreased as the iron concentration increased ( Figure 6A ) . In contrast , CFT2 transcript levels were elevated in the cft1 mutant , especially in the presence of 10 μM and 100 μM iron , compared to the reduced levels seen in the wild-type strain in response to iron ( Figure 6B ) . This result supports the conclusion that loss of Cft1 leads to reduced intracellular iron levels . It is possible that the elevated CFT2 transcript levels resulting from loss of CFT1 could potentially result in more Cft2 product and influence the iron acquisition capabilities of the cells . For the SIT1 gene , both the wild-type strain and the cft2 mutant showed an iron-dependent reduction of SIT1 transcript levels ( Figure 6C ) . However , SIT1 levels were higher in the cft1 mutant compared to other strains , and the levels were no longer entirely responsive to environmental iron concentration . These results again imply that the cft1 mutant is debilitated in its ability to import iron and that intracellular iron levels may become constitutively low . This situation would potentially account for the unresponsive expression of SIT1 even in the presence of 1000 μM of iron . This is in contrast to the decrease in the CFT2 transcript at the same iron concentration and this difference may reflect distinct mechanisms of iron regulation for the two genes . As mentioned above , CFT2 is negatively regulated by the iron-responsive transcription factor Cir1 and we have shown previously that SIT1 is positively regulated [26] . Overall , our results suggest that disruption of CFT1 influences expression of CFT2 and SIT1 , a result that is consistent with a reduction in intracellular iron levels in the cft1 mutant . The cAMP pathway controls the expression of high-affinity iron permeases in S . cerevisiae and U . maydis . However , these two fungi display opposite regulatory patterns . In S . cerevisiae , expression of the gene for the high-affinity iron permease , FTR1 , is negatively regulated whereas in U . maydis , expression of the orthologous gene , FER2 , is positively regulated [24 , 27] . These observations led us to investigate regulatory mechanisms of the high-affinity iron permease genes CFT1 and CFT2 in relation to the cAMP pathway in C . neoformans . For this analysis , strains lacking the genes encoding the catalytic subunits ( PKA1 , PKA2 ) or the regulatory subunit ( PKR1 ) of PKA were grown in media containing different concentrations of iron , and transcript levels of CFT1 and CFT2 were compared by real-time RT-PCR . The results revealed a similar expression pattern for CFT1 in the wild-type strain and the pka1 and pka2 mutants such that the transcript level was dependent on iron concentration . However , CFT1 expression was found to be reduced by 2- to 3-fold in the pkr1 mutant compared to the wild-type strain ( Figure 7A ) . Conversely , the transcript levels of CFT2 were found to be elevated upon deletion of PKA1 ( Figure 7B ) . These results indicate that the cAMP pathway negatively regulates the transcript level of CFT2 via Pka1 in C . neoformans and suggest that PKA also negatively influences CFT1 expression but to a lesser extent . As mentioned earlier , the basal transcript level for CFT2 is much lower than that of CFT1 and this may influence both the detection and the biological relevance of PKA regulation of these genes . It is interesting , however , that the loss of different subunits of PKA influences transcript levels for the two genes , and that reciprocal patterns of regulation for loss of the catalytic and regulatory subunits were not observed . These results suggest that the influence of the cAMP pathway may be exerted by different mechanisms for each of the genes . A relationship between iron levels and susceptibility to antifungal drugs that act at the level of ergosterol biosynthesis has been reported [41] . Specifically , the availability of iron influences susceptibility of C . albicans to antifungal drugs , and strains lacking the iron permease CaFtr1 display increased sensitivity to fluconazole due to alteration of expression of genes in the ergosterol synthesis pathway [41] . A similar connection also exists in S . cerevisiae because cytosolic iron levels influence C-4 methyl sterol oxidase ( Erg25 ) , an enzyme that is essential for ergosterol synthesis [42] . In this context , we tested the C . neoformans iron permease mutants to investigate whether a deficiency in iron uptake alters susceptibility to antifungal drugs . The cft1 mutant displayed increased sensitivity to the azole , miconazole , suggesting that a deficiency in iron uptake may also influence ergosterol synthesis in C . neoformans ( Figure 8 ) . This phenotype is also supported by the observation that cft1 mutants showed an increase in sensitivity to the antifungal drug amphotericin B , compared to the wild-type strain ( Figure 8 ) . These findings suggest that inhibitors of Cft1 activity could provide a synergistic effect when used in combination with existing antifungal drugs , and that Cft1 could be a novel drug target for treatment of cryptococcosis . The deficiency of the cft1 mutant in growth and iron acquisition with transferrin prompted an examination of the virulence of the cft1 and cft2 mutants in the mouse inhalation model of cryptococcosis . We found that mice infected with the wild-type strain , the cft2 mutant , and the CFT1 or the CFT2 reconstituted strains , showed 100% mortality by ∼20 days . However , mice infected with the cft1 mutant survived to ∼43 days indicating that deletion of CFT1 caused a significant attenuation of virulence ( Figure 9A ) . These results support the conclusion that Cft1 plays a role in iron acquisition in vivo . Furthermore , because the cft1 mutants showed defects in transferrin but not in heme and siderophore utilization in vitro , we hypothesize that transferrin is an important iron source for C . neoformans during growth in mammalian hosts . The virulence attributes of the cft1 single and the cft1 cft2 double mutants were also compared in a separate experiment . We found that mice infected with the double mutant survived to ∼54 days , a time significantly longer than mice infected with the cft1 single mutants ( Figure 9B ) . These results suggest that Cft2 makes a contribution during infection that is evident in the absence of CFT1 . The distribution of fungal cells in the infected mice was also assessed to determine the ability of the cft1 mutant to colonize host tissue . Organs from mice infected with the wild-type strain , the cft1 mutant or the CFT1 reconstituted strain were collected and fungal burden was measured by determining colony-forming units ( CFU ) . We harvested the organs at days 19 and 34 because mice infected with the wild-type and reconstituted strains succumbed to infection on or around day 19 and mice infected with the cft1 mutant started to reach the endpoint at approximately day 34 . These time points for the inhalation model of cryptococcosis would therefore represent the maximum disease progression in terms of fungal dissemination and proliferation . As shown in Figure 10A , wild-type and reconstituted cells were widely disseminated throughout the host at day 19 . However , in mice infected with the cft1 mutant , the number of the fungal cells was much lower at day 19 suggesting a defect in dissemination and/or colonization of organs ( Figure 10A ) . We hypothesize that the failure of the cft1 mutant to disseminate or colonize during infection is due , at least in part , to an inability to use transferrin . The cft1 mutant eventually disseminated from the lung to the spleen and the liver in infected mice by day 34 ( Figure 10A ) , implying that the cft1 mutants can presumably utilize host iron sources other than transferrin ( e . g . , heme ) . In particular , the fungal burdens for the cft1 mutant were higher at day 34 in the lung , spleen and liver ( 30- , 17- and 21-fold , respectively ) compared to day 19 . However , the number of fungal cells in the brains of mice infected with the cft1 mutant remained low and was only two-fold higher at day 34 compared to day 19 ( Figure 10A ) . This observation implies that transferrin or similar iron sources transported by CFT1 via the reductive iron uptake pathway may be primary iron sources for C . neoformans in the brain . It is likely that the cft1 and the cft1 cft2 mutants can use other iron sources in infected mice , and heme is a likely candidate because these mutants grow well with heme as the sole iron source in vitro ( Figures 4A and S2 ) . We also tested whether a different route of inoculation of fungal cells by tail vein injection could influence the behavior of the cft1 mutant in vivo . Mice infected with the wild-type and the CFT1 reconstituted strains succumbed within 7 days . In contrast , mice infected with the cft1 mutant survived to day ∼29 ( data not shown ) thus revealing that the cft1 mutant was also attenuated for virulence in this model of cryptococcosis . In general , approximately three orders of magnitude more fungal cells were found in the brains of the mice infected via tail vein injection than via inhalation . For this experiment , we chose to compare fungal burden at the time of illness for the mice infected with the different strains . Specifically , organs from mice infected with the wild-type strain or the CFT1 reconstituted strain were collected at day 7 , and organs from mice infected with the cft1 mutant were harvested at day 29 . Similar to results from inhalation model , lower numbers of the cft1 mutant cells were found in the brain , the spleen and the liver when the mice became ill at day 29 compared to infection with the wild-type strain at day 7 ( Figure 10B ) . The result for the lung was a notable exception in that higher numbers of cft1 mutant cells were found in mice infected with the cft1 mutant at day 29 ( Figure 10B ) , a phenomenon that was also observed in the inhalation model . We further confirmed the reduced ability of the cft1 mutant to colonize the brain or lungs compared to the wild-type strain in a three-day time course after tail vein injection ( Figure 10C ) . Overall , these data suggested that CFT1 makes a contribution to virulence during cryptococcosis , and that transferrin or other iron sources dependent on Cft1 could potentially be primary sources for C . neoformans within the host , especially in the brain . Iron uptake and homeostasis functions are known to be important for the pathogenesis of C . neoformans because iron overload exacerbates experimental cryptococcosis and iron levels influence the expression of virulence factors [26 , 36 , 43] . Pioneering physiological studies indicate that C . neoformans possesses at least two iron uptake systems [31 , 32 , 44] . The first is the utilization of siderophores although the fungus does not appear to produce these molecules [31] . The other is a reductive iron uptake system in which ferric iron is reduced to the ferrous form by cell surface reductase activity , by secreted reductants and by melanin in the cell wall [32 , 44] . Jacobson et al . ( 1998 ) detected activities of both high- and low-affinity iron uptake systems from cell cultures of C . neoformans , estimated the Km of the high-affinity uptake to be 0 . 6 μM and found that the activity of the high iron uptake system is down-regulated by 15 μM of iron [31] . The low-affinity uptake system has not been characterized . In the context of these findings , we determined the roles of the CFT1 and CFT2 genes that show sequence similarity to genes for the iron permeases responsible for high-affinity iron uptake in other fungi such as S . cerevisiae and C . albicans . Our analysis of mutants defective in CFT1 , CFT2 or both genes revealed that CFT1 plays the major role in iron acquisition in C . neoformans . The cft1 mutant displayed growth defects in the presence of iron sources ( FeCl3 and transferrin ) that are known to be transported by reductive iron uptake system in C . albicans [13] . These results , together with our expression data showing that CFT1 is down-regulated at higher iron concentrations , suggest that Cft1 functions as a high-affinity iron permease responsible for iron uptake in C . neoformans . However , Cft1 is not involved in the heme or siderophore uptake systems in C . neoformans because we found that the cft1 mutant displays normal growth with these iron sources . The role of Cft2 in C . neoformans is less clear because of the low transcript level for the gene and the lack of robust phenotypes for the cft2 mutant . Several lines of evidence suggest that Cft2 may play a minor role and may function redundantly with Cft1 . First , the cft1 cft2 double mutants displayed a growth defect in the rich medium YPD ( an iron-replete condition ) , whereas the cft1 and cft2 single mutants did not show growth defects in this medium . Second , the expression of CFT2 is up-regulated in the cft1 mutants and in response to iron limitation . Finally , the cft1 cft2 mutant showed attenuated virulence compared with the single cft1 mutant . It is possible that CFT2 encodes a low affinity iron permease or , alternatively , that it encodes a vacuolar permease that functions to transport stored iron to the cytoplasm . It is also possible that the expression of CFT2 may be different ( e . g . , higher ) , and that the gene may make a contribution , under conditions not tested in our study . Further biochemical analyses and localization studies will be needed to better define the role of Cft2 . The presence of two candidate iron permease genes , CFT1 and CFT2 , in C . neoformans is , at first glance , similar to the situation in C . albicans where the iron permease genes CaFTR1 and CaFTR2 have been characterized [12] . As was found with CaFTR1 , CFT1 appears to play a major role in iron uptake , growth on specific iron sources and virulence . A mutant lacking CaFTR2 is similar to the cft2 mutant in that neither showed growth phenotypes and neither is required for virulence . However , the regulation of CaFTR2 is quite different because it has a low transcript level in low iron conditions and an elevated transcript at higher levels . Ramanan and Wang [12] suggest that the differential expression of CaFTR1 and CaFTR2 may reflect their functions in different environments . Similarly , CFT2 may be expressed under specific conditions and , in this context , it is interesting that Fan et al . [37] found that CFT2 and CFO2 transcript levels were induced upon phagocytosis . Thus , these genes may make a contribution during infection consistent with the observed additional attenuation in virulence found for the cft1 cft2 mutant compared with the cft1 mutant . Previously , we described the role of the GATA type zinc finger transcription factor Cir1 in iron-related regulation in C . neoformans [26] . Microarray analysis revealed that Cir1 regulates expression of genes required for iron uptake in both a negative and a positive manner [26] . In that report , we found that a putative iron permease ( XP_569788 ) is a negatively regulated , downstream target of Cir1 in the serotype D strain B3501A . The current study revealed that this gene is equivalent to CFT2 in the serotype A strain H99 . We also identified another iron permease gene ( XP_568258 ) from the microarray experiments that we now know corresponds to CFT1 . We should note that values representing differential expression of the CFT1 ortholog ( XP_568258 ) were not statistically significant in the microarray analysis in the serotype D strain background . The current study using a serotype A strain revealed that CFT1 is a positively-regulated downstream target gene of Cir1 , which is opposite to the regulation found for CFT2 . This confirms our previous findings of a dual mode of Cir1 regulation in C . neoformans [26] . The SIT1 gene , which encodes the siderophore transporter [33] , also belongs in the regulatory network of genes positively controlled by Cir1 . As mentioned , the cAMP pathway negatively regulates genes involved in the high affinity iron uptake through the protein kinase A subunit Tpk2 in S . cerevisiae . U . maydis also possesses similar connections but the pathway positively regulates genes involved in the high affinity uptake [24] . In our study , we observed that the cAMP pathway regulates transcriptional levels of iron permease genes in C . neoformans . An ∼2-fold reduction in CFT1 transcript levels was found in the pkr1 mutant and a marked elevation of the CFT2 transcript was found in the pka1 mutant ( ∼10-fold ) . These data suggest that regulation of CFT1 by the cAMP pathway may be indirect and of minor significance under the conditions we tested . Also , the pka1 mutant did not show a reciprocal pattern of differential expression compared with the pkr1 mutant in terms of CFT1 transcript levels , again suggesting that the influence may be indirect . The more substantial regulation of CFT2 by Pka1 further supports the idea that there may be specific conditions where Cft2 makes a contribution to iron acquisition or other functions . The reason why Pka1 controls CFT2 , but not CFT1 , remains to be investigated . Finally , the regulation by cAMP extends to other genes involved in iron acquisition in C . neoformans because we previously found elevated transcripts for the SIT1 gene ( siderophore transporter ) in the pka1 mutant [33] . The expression of the CFT1 and CFO1 genes was recently shown to be influenced by deletion of the regulatory gene SRE1 , which functions in an oxygen sensing pathway in C . neoformans [45] . Based on predictions from work on Sre1 in S . pombe , Chang et al . [45] found that Sre1 in a serotype D strain of C . neoformans is activated by inhibition of sterol synthesis and low oxygen levels , and that the gene is required for wild-type levels of growth under hypoxic and low iron conditions . A microarray experiment to compare the transcriptomes of wild type and sre1 mutant cells grown in a low oxygen environment revealed that Sre1 positively influences the expression of ∼100 genes . These genes encoded enzymes for ergosterol biosynthesis as well as iron transport functions such as Cft1 and Sit1 [33] , and the copper transporter Ctr4 [46] . Deletion of SRE1 negatively influenced the expression of another 414 genes and many of these encoded stress-related functions . The role of the SRE1 gene has also been characterized in a serotype A strain of C . neoformans [47] . In this strain background , the sre1 mutant displays a hypoxia-sensitive phenotype , slight defects in capsule and melanin formation , and a reduced ability to proliferate and cause disease in a mouse model of infection . Interesting , Chun et al . [47] did not observe the changes in transcript levels for the CFT1 and CFO1 genes as a result of hypoxia or loss of SRE1 that were seen by Chang et al . [45] . This may reflect differences in the regulation of iron uptake functions between strains of the A and D serotypes . The connections between oxygen , iron and sterol biosynthesis established by Chang et al . [45] are interesting in light of our finding that the cft1 mutant displayed increased sensitivity to an inhibitor of ergosterol biosynthesis . This observation further supports the idea that C . neoformans , as a strict aerobe , must balance iron availability with oxygen levels and ergosterol synthesis [26 , 45] . Similar connections have been described in S . cerevisiae , although this fungus is capable of anaerobic growth . In yeast , anaerobiosis results in reduced heme synthesis , a lower rate of synthesis of respiratory proteins and loss of the ability to synthesize sterols because of the iron dependence of enzymes in the pathway [48] . SRE1 in C . neoformans is also required for the establishment and growth of the fungus in the brains of infected mice thus indicating that this tissue site may be limited for oxygen , and perhaps for iron . Our finding that the cft1 mutant has reduced colonization of the brain further suggests that this tissue is limited for iron . Additional evidence for integration of iron , oxygen and sterol biosynthesis comes from the finding that a 2 . 2-fold reduction was seen for SRE1 transcript levels under the iron limited condition in the cir1 mutant [26 , 45] . Thus , SRE1 may be a direct or indirect target of Cir1 regulation and the two regulators could also potentially interact at the promoters of genes such as CFT1 . Additionally , these factors both influence the transcript levels of the copper transporter Ctr4 . Waterman et al . [46] showed that this gene is activated by the transcription factor Cuf1 and , in parallel with the findings for the sre1 and cft1 mutants , cuf1 mutants show reduced CNS colonization in a murine model of infection . Overall , these results provide the first glimpses of the integration of oxygen , sterol , copper and iron sensing regulatory schemes that influence virulence and CNS colonization in C . neoformans . Our analysis revealed that the cft1 mutant ( but not the cft2 mutant ) showed reduced virulence in the mouse model of the cryptococcosis , a result consistent with a role for Cft1 in iron acquisition in vivo . This is consistent with SAGE experiments on C . neoformans cells from a rabbit model of cryptococcal meningitis and from a mouse pulmonary infection which show that CFT1 transcript was abundant during growth in the host [35] ( Hu and Kronstad , manuscript in preparation ) . CFT2 appears to make a minor contribution because the cft1 cft2 double mutant showed a further reduction in virulence compared to the cft1 mutant . Two models of cryptococcosis ( inhalation and tail vein injection ) revealed reduced numbers of fungal cells in the brains , spleens and livers of mice infected with the cft1 mutant compared to mice infected with the wild-type and reconstituted strains . Overall , it appeared that the mutant exhibited a generalized growth defect that could account for the slower disease progression . In particular , we noted a delayed appearance of the fungal cells in brains of mice infected with the cft1 mutants by either route . The cft1 mutant cells did eventually reach high levels in the lungs near the time that the mice succumbed to infection . The higher burden was particularly striking compared with the low numbers in the brain at the same time . This observation may indicate that lung tissue provides a growth environment in which the requirement for Cft1 is not as stringent as in the brain . For example , tissue differences in oxygen availability may influence the requirement of the fungus for iron [49] . The delayed disease progression for the cft1 and the cft1 cft2 mutants supports our hypothesis that C . neoformans is partially dependent on an iron source that requires a reductive uptake system during infection . Transferrin is likely the source because of its abundance and because the cft1 mutant showed defects in the utilization and uptake of iron from transferrin in vitro . We hypothesize that transferrin may be particularly important for the growth of C . neoformans in the brain . Transferrin may be the primary iron source in the CNS because it is the only iron carrier protein that can be transported through the blood-brain barrier ( BBB ) [50] . Although a reduction in fungal burden in mice infected with the cft1 mutant was apparent , the mutant eventually caused mortality . We propose that the cft1 mutant ( and the cft1 cft2 mutant ) was able to grow in the host because it could also utilize iron sources such as heme via a non-reductive mechanism . This idea is consistent with our finding that wild-type cells showed robust growth in the presence of heme or siderophores . The mechanisms of heme utilization have not been investigated in C . neoformans . C . neoformans must breach the BBB and invade the CNS to cause meningoencephalitis . Only a few pathogenic microbes are able to cross the BBB and the process is best understood in bacteria . For example , the bacterium Listeria monocytogenes infects leukocytes and is then transported into the CNS during leukocyte migration through the BBB ( i . e . , the “Trojan horse” mechanism ) [51] . The encapsulated bacteria Streptococcus pneumoniae and Neisseria . meningitis use transcytosis to cross a monolayer of brain endothelial cells [52 , 53] . Evidence to date indicates that C . neoformans is also able to adhere to and transcytose across human brain microvascular endothelial cells [54] . In this context , we speculate that CFT1 may play an iron acquisition role while the fungus is within the endothelial cell during transcytosis and/or while the fungus is within CNS . Interestingly , holo-transferrin also enters the endothelial cells of the BBB by transcytosis so this iron source would potentially be available to the fungus [55] . The mechanisms of iron acquisition by C . neoformans during phagocytosis by macrophages also remains to be investigated . As mentioned , CFT2 transcripts are elevated upon phagocytosis but we note that recent studies show that phagosome extrusion of the fungal cells occurs as early as 2 hours after uptake [56 , 57] . Overall , our results provide insights into the role of iron acquisition functions in cryptococcal disease , reveal iron source preferences , and suggest possible targets for antifungal therapy , especially in the context of treating fungal meningitis . The strains used in this study ( Table S1 ) were routinely grown in yeast extract , bacto-peptone medium with 2 . 0% glucose ( YPD , Difco ) or yeast nitrogen base ( YNB , Difco ) with 2 . 0% glucose . Defined low iron medium was prepared as described [58] and we have determined that this medium contains approximately 1 . 3 μM iron ( data not shown ) . This medium was used for the experiments described in Figures 2–4 , 6 , and 7 , and 0 μM in the figure labels indicates that no additional iron was added . Iron-replete medium was prepared by adding the iron sources FeCl3 , holo-transferrin , heme or feroxamine into low-iron medium at the final concentrations indicated in the text . To assess growth , cells were first grown in low-iron medium for two days at 30°C to deplete intracellular iron stores and to fully induce the high affinity iron uptake system in C . neoformans , as suggested by Jacobson et al . [31] . The number of cells was determined using a haemocytometer and 2 . 0 × 104 cells were transferred to the wells of a 96-well plate containing low-iron medium as a control or low-iron medium containing different iron sources . Iron sources were diluted by serial two-fold dilutions in a total volume of 200 μl . The plates were incubated at 30°C for three days , and the optical density of each well was read with a microtitre plate reader at 595 nm . Additional experiments to examine growth rate were performed in 5 ml cultures containing low iron medium supplemented with the same iron sources . For antifungal sensitivity tests on plates , 10-fold serial dilutions of cells were spotted onto YPD plates containing miconazole or amphotericin B . Plates were incubated at 30°C for two days . The locus numbers for CFT1 and CFT2 in the C . neoformans serotype A genome are CNAG_06242 . 1 and CNAG_02959 . 1 , respectively ( http://www . broad . mit . edu/annotation/genome/cryptococcus_neoformans ) . The sequences for these genes were used for mutant construction . All primers used for the experiments are listed in Table S2 . To construct a cft1 mutant , the genomic region of 1740 bp that corresponds to the entire coding sequence of CFT1 was replaced by a disruption cassette containing the nourseothricin acetyltransferase gene ( NAT ) using 5' and 3' flanking sequences of CFT1 . The disruption cassette was constructed by an overlap PCR method using primers TL2061 , TL2062 , TL2063 , TL2064 , TL2065 and TL2066 , along with strain H99 genomic DNA and the plasmid pCH233 as templates [59 , 60] . The construct was biolistically transformed into the wild-type strain as described previously [61] . Positive transformants were identified by PCR , confirmed by Southern blot analysis and named TLF1–9 ( Figure S3A ) . To reconstitute the cft1 mutant , primers H9F1RCF and H9F1RCR were used to amplify the wild type CFT1 gene from genomic DNA . The PCR fragments were digested with SacI and SpeI , and cloned into pJAF to construct pWH045 containing the neomycin resistant marker ( NEO ) . The plasmid pWH045 was digested with NdeI and transformed into the cft1 mutant . Positive transformants containing the wild-type CFT1 at its authentic locus were identified by PCR and named CFT1R9–4 . To construct the cft2 mutant , the genomic region of 2010 bp that corresponds to the entire coding sequence of CFT2 was replaced by a disruption cassette containing the NEO marker using 5' and 3' flanking sequences of CFT2 . Primers TL2071 , TL2072 , TL2073 , TL2074 , TL2075 and TL2076 were used for construction of the disruption cassette by an overlap PCR method and H99 genomic DNA and the plasmid pJAF were used as templates [59 , 60] . The construct was transformed as described above and positive transformants were identified by PCR , confirmed by Southern blot analysis and named TLF2–9 ( Figure S3B ) . To reconstitute the cft2 mutant , primers H9F2RCFBglII and H9F2RCRBglII were used to amplify the wild type CFT2 gene . The PCR fragments were digested with BglII and cloned into BamHI digested pCH233 to construct pWH056 containing the NAT marker . The plasmid pWH056 was digested with NdeI and transformed into the cft2 mutant . Positive transformants containing the wild-type CFT2 at its authentic locus were identified by PCR and named CFT2R9–2 . The cft1 cft2 double mutants were constructed by transforming the CFT2 disruption cassette into the cft1 mutant . The positive transformants were selected by PCR , confirmed by Southern blot analysis and named cft1Δ cft2Δ ( Figure S3C ) . Iron uptake assays were performed as described previously with minor modifications [62 , 63] . Briefly , cells were grown in YNB medium overnight , transferred to YNB medium containing 1 mM ascorbic acid and 1mM ferrozine and incubated at 30°C for another 12 h . For iron uptake from FeCl3 , 2 . 0 × 107 cells were withdrawn from each culture , centrifuged and washed once with uptake buffer ( 0 . 2 M 3- ( N-Morpholino ) -propanesulfonic acid ( MOPS ) , 2% glucose , 20 mM Na-citrate pH 6 . 8 ) . Cells were resuspended in 0 . 5 ml of uptake buffer and kept at room temperature for 15 min for equilibration . Uptake buffer ( 0 . 5 ml ) containing 20 μM of 55FeCl3was added to each sample , and the cells were incubated for 30 min at room temperature . After incubation , 5 ml of quenching buffer ( 0 . 375 M Succinic acid , 0 . 625 M Tris , 0 . 128 M EDTA , pH 6 . 0 ) was added , and each sample was immediately vacuum-aspirated through a GF/A filter ( Whatman ) . The filter was washed with 20 ml of quenching buffer , and radioactivity was measured by liquid scintillation counting . The uptake of iron from transferrin was performed as described previously [13] . Briefly , human apo-transferrin was purchased from Sigma and 55Fe labelled transferrin was prepared by the addition of a threefold molar excess of 55FeCl3 to 10 μM apo-transferrin in transferrin loading buffer ( 0 . 1 M HEPES , pH 7 . 5 , 0 . 15 M sodium chloride , and 10 mM sodium bicarbonate ) . The reaction mix was incubated for 30 min at 22°C and 30 min on ice , and the labelled transferrin was separated from unbound 55Fe using Sephadex G-25 spin columns . Cells were grown to mid log phase in the presence of 300 μM Ferrozine . They were then washed three times with citrate-glucose buffer ( 0 . 1 M Morpholinethanesulfonic acid ( MES ) buffer pH 6 . 0 , 20 mM Na-citrate , 2% glucose ) and resuspended in the same buffer followed by incubation in a 30°C water bath for 15 min prior to the addition of 2 μM 55Fe-transferrin . After 1 h , samples were removed and quenched using 3 ml of quenching buffer . The free iron was removed by washing and the 55Fe radioactivity associated with the cells was measured by liquid scintillation counting on Wallac 1409 liquid scintillation counter . Primers for real-time RT-PCR analysis were designed using Primer Express software 3 . 0 ( Applied Biosystems ) and are listed in Table S3 . Cell cultures were prepared by growth in low-iron medium as described above , and then transferred to the same medium containing different concentrations of iron sources as indicated in the text . Total RNA was purified with the RNeasy kit ( Qiagen ) , treated with DNAse ( Qiagen ) and cDNA was generated using the SuperScript First-Strand Synthesis System ( Invitrogen ) . PCR reactions were monitored as described previously [33] , and relative gene expression was quantified using the SDS software 1 . 3 . 1 ( Applied Biosystems ) based on the 2-ΔΔCT method [64] . ACT1 was used as a control for normalization . To examine virulence in an inhalation model of cryptococcosis , strains were cultured in 5 ml YPD overnight at 30°C . The overnight cultures were washed twice with PBS , and the fungal cells were resuspended in PBS . 4–6 week old female A/Jcr mice were anesthetized intraperitoneally with ketamine ( 80 mg/kg ) and xylazine ( 5 . 5 mg/kg ) in saline , and suspended on a silk thread by the superior incisors . A cell suspension of 5 × 104 cells in 50 μl was slowly dripped into the nares of the anesthetized mice , and the mice were suspended for 10 min on the thread . For the tail vein injection model , strains were directly injected in the lateral tail veins of mice at a density of 5 × 104 cells in 200 μl of PBS . The status of the mice was monitored daily post-inoculation . Mice reaching the humane endpoint were euthanized by CO2 anoxia . To assess fungal burden of organs from mice infected by inhalation or tail vein injection , three mice were used for each strain at each time point . Mice were euthanized with CO2 , four organs including the brain , the lungs , the liver and the spleen , were aseptically removed and soaked in 0 . 5 ml PBS overnight at 4°C . Organs were homogenized manually with a sterile plastic pestle . Cell strainers ( BD , 70 μm Nylon ) were used to remove tissue debris in the samples . The samples were serially diluted and 200 μl of the diluted samples were plated and spread with sterile 4 mm glass beads on YPD plates containing chloramphenicol ( 35 μg/ml ) . CFUs were determined after three days of incubation at 30°C . The following method was used to examine the timing of dissemination to the brain and lungs following tail vein injection . Cells were cultured in 5 ml of YPD overnight at 30°C , washed twice with PBS , and resuspended in PBS . Cell counts were performed with a haemocytometer . A total of 28 6-week-old female A/Jcr mice were inoculated with 200 μl of the wild-type strain H99 or the cft1 mutant ( 2 . 5 × 105 cells/ml in PBS ) . The brain and the lungs of the mice were harvested aseptically on days 1 , 2 and 3 post-inoculation . The organs were manually homogenized in 1 ml PBS with sterile plastic pestles . Cell strainers ( BD , 70 μm nylon ) were used to remove tissue debris in the samples . The samples were serially diluted and 200 μl of the dilutions were plated followed by incubation at room temperature for 3 days . The protocols for the virulence assays ( protocol A99–0252 ) conformed to regulatory standards of and was approved by the University of British Columbia Committee on Animal Care .	Opportunistic fungal pathogens and other invading microbes must overcome extreme iron limitation to proliferate in the mammalian host . It is not yet known which iron sources are preferred by fungal pathogens of mammals , although the mechanisms of acquisition are beginning to be explored . Some fungi produce iron-chelating siderophores to capture iron from host proteins , while others appear to require a membrane-bound iron permease–ferroxidase system . We describe the ability of the encapsulated yeast Cryptococcus neoformans to use host iron sources including transferrin and heme , and we identify an iron permease that is required for full disease progression in experimental mouse models . The permease is required for iron utilization from transferrin but not heme during growth in laboratory culture . This result when combined with the observed slow growth of the permease mutant during the experimental infections implicates transferrin as an important iron source in the host . However , we find that mutants lacking the permease eventually do cause disease , thus revealing that additional iron sources such as heme and other uptake mechanisms are available to C . neoformans . Finally , we noted that the permease mutant showed particularly poor growth in the brains of infected animals , suggesting that transferrin may be an especially important iron source in this tissue .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "yeast", "and", "fungi", "cell", "biology", "infectious", "diseases", "microbiology" ]	2008	Iron Source Preference and Regulation of Iron Uptake in Cryptococcus neoformans
The utilization of symbols such as words and numbers as mental tools endows humans with unrivalled cognitive flexibility . In the number domain , a fundamental first step for the acquisition of numerical symbols is the semantic association of signs with cardinalities . We explored the primitives of such a semantic mapping process by recording single-cell activity in the monkey prefrontal and parietal cortices , brain structures critically involved in numerical cognition . Monkeys were trained to associate visual shapes with varying numbers of items in a matching task . After this long-term learning process , we found that the responses of many prefrontal neurons to the visual shapes reflected the associated numerical value in a behaviorally relevant way . In contrast , such association neurons were rarely found in the parietal lobe . These findings suggest a cardinal role of the prefrontal cortex in establishing semantic associations between signs and abstract categories , a cognitive precursor that may ultimately give rise to symbolic thinking in linguistic humans . Humans and animals share an evolutionarily old quantity representation system that allows the estimation of set size or number of events [1] . The assessment of numerical information is advantageous for the individual's fitness . This is particularly evident in social interactions ( fight or flight decisions in contests ) [2] , foraging ( exploiting the richer food source ) [3] , and parenting ( discrimination of offspring ) [4] . Quantity representations arise spontaneously without training as has been shown numerous times in monkeys [3] and human infants [5 , 6] , supporting the idea that numerical competence is an ontogenetically and phylogenetically early faculty . Nonverbal numerical cognition , however , is limited to approximate quantity representations [1 , 7] and rudimentary arithmetic operations [5 , 6 , 8 , 9]; precise number representations and exact calculation are beyond its reach . In contrast , humans familiar with number symbols are able to grasp exact cardinalities and to execute even the most abstract calculations . Humans learn to use number symbols as mental tools during childhood . Prior to the utilization of signs as numerical symbols [10] , long-term associations between initially meaningless shapes ( that become numerals ) and inherently semantic numerical categories must inevitably be established [11 , 12] . Associations between shapes and quantities , a necessary first step towards the utilization of number symbols in linguistic humans , can even be mastered by animals [13–16] . Several studies in humans point to the prefrontal cortex ( PFC ) and the intraparietal sulcus ( IPS ) as key structures for both non-symbolic [17 , 18] and symbolic quantity information [18–20] . In monkeys , it has been shown that potentially homolog brain areas are involved in processing non-symbolic numerosity [21–26] . These studies support the hypothesis of a phylogenetic precursor system in monkeys on which higher , verbal-based numerical abilities in adult humans build up [27] . If the precursor hypothesis holds true , the same network that is involved in quantity estimation in nonhuman primates should also be engaged in the association of visual shapes with numerical categories . Here , we test this prediction by investigating whether single cells associate approximate numerosity representations with symbolic-like representations , and if so , what the respective contributions of the prefrontal and parietal cortices in this mapping process could be . To that aim , we trained monkeys to assign visual shapes to numerical categories and recorded from single cells in both candidate regions . We report that many neurons in the PFC encoded the learned numerical value of a visual shape . In contrast , such association neurons were rarely found in the parietal lobe . Overall , the results suggest that the PFC is the prime source for the linking of signs to numerical categories in monkeys and may serve as a neuronal precursor for number symbol encoding . We trained two rhesus monkeys in a delayed match-to-sample protocol to discriminate small numerosities ( one to four ) in multiple-dot patterns ( Figure 1A; dot protocol ) . The monkeys had to judge whether two successive task periods ( first sample , then test ) separated by a 1-s delay showed the same numerosity . If so , the animals had to release a lever . In a second step , the monkeys learned over months to associate visual shapes ( Arabic numerals ) with the numerosity in multiple-dot displays , i . e . , Arabic numeral 1 was associated with one dot , numeral 2 was associated with two dots , and so on ( Figure 1B; shape protocol ) . Finally , both protocols were presented in a randomly alternating fashion within a given session . We ensured that non-numerical parameters in the dot protocol could not be used by the monkeys to solve the task by varying and controlling low-level visual features . For each session , 100 different images per numerosity were generated with pseudo-randomly varied visual properties . Sample and test stimuli were never identical . All four quantities were presented in each session with one standard and one control condition . Different control conditions were applied day by day . Controls in the dot protocol included dot displays with constant circumference , linear configuration , and constant density across all presented quantities ( see Figure 1C ) . To force the monkeys to generalize to the overall sign characteristics in the shape protocol , the numeral shapes were varied in size , position , and font . The font “Arial” was used for the standard condition; fonts “Times New Roman , ” “Souvenir BT , ” and “Lithograph Light , ” were used in control conditions ( see Figure 1D ) . The test stimulus for the shape protocol consisted of sets of black dots , equivalent to the dot protocol . Trials of the standard and control conditions as well as the dot and shape protocols were pseudo-randomly intermingled and appeared with equal probabilities in each session . Both monkeys learned reliably to associate numerical values with the visual shape of numerals . Average performance in the dot protocol ( Figure 2A and 2B ) and the shape protocol ( Figure 2C and 2D ) was comparable ( 87% and 88% , respectively ) and significantly better than chance for all tested quantities ( p < 0 . 0001 , binomial test ) . The numerical size and distance effect [22] could be observed in both protocols , irrespective of whether the standard or control condition was applied ( see Figure S1 ) . This suggests that the monkeys were indeed judging the direct and associated numerical values . We recorded 692 randomly selected neurons from the lateral PFC of the monkeys while they performed the tasks . Intermingled presentation of both protocols during each session allowed us to investigate individual neurons' responses to both dot and shape protocols . Many neurons were selective to numerical category and discharged strongest to specific ( direct or associated ) numerical values , irrespective of the protocol . Neuron 1 , in Figure 3A–3E , for example , showed a maximum response to numerosity two ( the neuron's preferred numerosity ) in the early sample phase , and a progressive drop-off with increasing numerical distance from the preferred numerosity in the dot protocol ( Figure 3A ) . The same neuron preferred the same ( associated ) numerical value ( i . e . , two ) in the shape protocol ( Figure 3B ) and had an equivalent tuning function ( Figure 3C ) . Neuron 2 , in Figure 3F–3J , preferred numerosity four in both the sample and delay phase in the dot protocol ( Figure 3F ) . The same neuron exhibited a remarkably similar temporal discharge pattern to the signs associated to specific numerical values in the shape protocol ( Figure 3G ) . For both protocols , the neuron showed monotonically increasing tuning functions ( Figure 3H ) . Neuron 3 , in Figure 3K–3O , showed strikingly similar responses during the memory period in both the dot ( Figure 3K ) and shape protocol ( Figure 3L ) , with a preferred numerical value two; the tuning functions obtained with the dot and the shape protocols were almost identical ( Figure 3M ) . For a quantitative analysis of the neurons' selectivity to numerical values , we first calculated a two-way analysis of variance ( ANOVA ) ( with factors numerical value [i . e . , 1 , 2 , 3 , 4] × stimulus condition [i . e . , standard versus control] , p < 0 . 05 ) separately for the dot and shape protocols . During the sample period , 263 ( 263/692 , or 38% ) neurons were selective for shapes and 229 ( 229/692 , or 33% ) for the number of dots irrespective of whether standard or control conditions were used ( significance only for factor “numerical value”; no other significant effects ) . During the delay period , 297 ( 297/692 , or 43% ) and 300 ( 300/692 , or 43% ) neurons were significantly tuned to shapes and the number of dots , respectively . We found 210 neurons during the sample and/or delay phase that were selective only to factor “numerical value” in both protocols , irrespective of the displays' visuospatial properties . For all quantities from one to four , we found neurons with the same preferred numerosities and associated numerical values . The observed frequency of those neurons was significantly higher compared to chance occurrence ( p < 0 . 001 , binomial test; Figure S5; see Materials and Methods for a description of chance level calculation ) . More precisely , more neurons exhibited the same preferred numerical value in the dot and shape protocols than expected assuming independence between the encoding of the two stimulus protocols . Neurons that were ANOVA-selective in both protocols ( especially those with the identical preferred numerical value in both protocols ) constitute a potential neural substrate for long-term numerical associations . In addition to mere selectivity in the dot and shape protocols , however , neurons should have similar tuning functions for the ( direct and associated ) numerical values in both protocols . To test this hypothesis , and to investigate the time course of association , we performed a sliding cross-correlation analysis between each neuron's tuning functions in the shape and dot protocols for all 210 ANOVA-selective cells and derived the cross-correlation coefficients ( CCs; see Figures S2–S4 for details ) . The significance of the CCs was evaluated by using a sliding receiver operating characteristic ( ROC ) analysis . For each neuron , we derived the ROC values of the difference between CCs and the shuffle predictors ( SPs , which constitute chance CCs ) in 25-ms time steps [28] ( see Materials and Methods ) . Based on this analysis , 157 cells ( 157/692 , or 23% ) were significantly correlated and classified as “association neurons . ” For instance , neuron 1 associated between visual shapes and numerical values during the sample onset phase , i . e . , 175 ms after stimulus onset and , taking its response latency of 120 ms into account , 55 ms after its earliest visual response ( Figure 3D and 3E ) . The associative neuronal responses of neuron 2 ( Figure 3I and 3J ) ranged from 250 ms ( latency-corrected: 11 ms ) after stimulus onset to 50 ms before the end of the delay period . As an example of a late-associating cell , neuron 3 associated throughout the entire memory phase ( see Figure 3N and 3O ) . The time course of association shown in Figure 4A for the entire sample of association neurons revealed many neurons that associated the numerical values of shapes and dots early after sample onset . While individual cells coded the ( direct and associated ) numerical values during specific time phases in the trial ( represented by the black bars in Figure 4A ) , the neuronal population represented the numerical association throughout the entire trial . When corrected for response latency , about half of the association neurons started to associate numerical values within the first 200 ms after neuronal response onset . One hundred and thirteen neurons began to associate during the sample phase , and 44 neurons during the delay phase ( Figure 4B ) . Interestingly , the tuning functions of association neurons showed a distance effect [22] for both protocols , i . e . , a drop-off of activation with increasing numerical distance from the preferred numerical value ( numerical distance 1 versus 3 , dot protocol , p < 0 . 001 , n = 104; shape protocol , p < 0 . 01 , n = 91; Wilcoxon signed-rank test , two-tailed; see single-cell examples in Figure 3C , 3H , and 3M , and population analysis in Figures 4C and S6 ) . The distance effect found in the shape protocol indicates that association neurons responded as a function of numerical value rather than visual shape per se . However , the neuronal response drop-off between the preferred and second-preferred numerical values was larger in the shape protocol ( 50% ) than in the dot protocol ( 39 . 8% ) ( p = 0 . 016 , n = 157 , Wilcoxon signed-rank test , two-tailed ) . This might indicate a more precise encoding of numerical values represented by signs than by sets of dots . Is the association of numerical values by single PFC neurons really relevant for the monkeys' behavior ? If association neurons constitute a neuronal correlate for the monkeys' ability to link signs with numerosities , the tuning correlations for both protocols should be weakened whenever the monkeys failed to associate visual shapes with their corresponding numerosities in error trials . To address this issue , we calculated the CCs of association neurons between correct trials in the dot protocol and error trials in the shape protocol . Because of the monkeys' low overall error rates , error trials were only available for a subset of numerical values ( e . g . , 2 , 3 , and 4 ) for many neurons . Only neurons recorded during errors to two or more numerical values were included into the error trial analysis . This criterion was fulfilled by 153 out of the 157 association neurons . As shown in Figure 5A and 5B , the correlation patterns for individual neurons were disturbed in error trials , and the mean population CCs were significantly decreased in error trials during and after cue presentation ( p < 0 . 001 , n = 153 , Wilcoxon signed-rank test , two-tailed ) . As expected , baseline correlation during the fixation period was unaffected ( p = 0 . 44 ) . These findings strongly argue for association neurons as a neuronal substrate of the semantic mapping processes between signs and categories . During PFC recordings , we simultaneously recorded from 437 neurons in the fundus of the IPS ( see Figure 6 ) and analyzed the neurons' responses in the same manner ( i . e . , two-factor ANOVA and cross-correlation analysis ) . In the IPS , we found many neurons encoding either the visual shapes or the numbers of dots separately ( 67/437 , or 15% , and 62/437 , or 14% , respectively , during the sample period and 58/437 , or 13% , and 83/437 , or 19% , respectively , during the delay period; see Figure 6B for a summary of sample and delay ) . The proportion of neurons showing stimulus condition and/or interaction effects in the dot and shape protocols was significantly higher in the IPS ( 118/437 , or 27% , and 107/437 , or 24% ) than in the PFC ( 119/692 , or 17% , and 133/692 , or 19% ) ( p < 0 . 001 and p < 0 . 05 , respectively; Chi-square test ) . This argues for a more abstract encoding of numerical values in the PFC and a more sensory-driven activity in the IPS . In contrast to the abundance of significantly tuned IPS neurons for the shape and dot protocols , only very few IPS neurons were selectively tuned to both protocols ( n = 19 ) ; even fewer turned out to have significant correlations ( 8/437; Figure 6B ) . Compared to the PFC , for the IPS , the proportion of association cells from the pool of all selective cells was significantly lower ( p < 0 . 001 , Chi-square test; Figure 6D ) . Nevertheless , the proportion of neurons with identical preferred numerical values in both protocols was slightly higher than expected by chance ( p < 0 . 001 , binomial test ) ( see Figure S5 and Materials and Methods for calculation of chance level and the distribution of preferred numerical values ) . Correlation time course and correlation strength ( as measured by the ROC values ) were fundamentally different between PFC and IPS neurons ( Figure 7 ) . In the PFC , ROC values showed a sharp increase right after sample onset and remained elevated throughout the entire trial ( Figure 7A ) . In the IPS , however , neuronal association was weak and occurred much later during the trial ( Figure 7B ) ; ROC values showed an increase around the end of the sample and delay period , but in contrast to values for the PFC , the IPS values were low during both periods . In summary , only PFC neurons seemed to be crucially involved in associating shapes with numerical magnitudes . Previous studies showed that neurons in the PFC encode learned associations between two purely sensory stimuli without intrinsic meaning ( e . g . , the association of a certain color with a specific sound , or pairs of pictures ) [29–31] . In the anterior inferotemporal cortex , Miyashita and co-workers found “pair-coding neurons” that responded to arbitrary pairs of images monkeys learned to match in a pair-association task [32] , and evidence that the PFC is important for active retrieval of these associative representations [33] . Here we show , to our knowledge for the first time , that neurons in the PFC represent semantic long-term associations not only between pairs of pictures , but between arbitrary shapes and systematically arranged categories with inherent meaning ( i . e . , the ordered cardinalities of sets ) . Our results suggest that the PFC may not only control the retrieval of long-term associations , but may in fact constitute a cardinal processing stage for abstract semantic associations . The prefrontal region is strategically situated for such associations [34]; it receives input from both the anterior inferotemporal cortex , which encodes shape information [35] , and the posterior parietal cortex , which contains numerosity-selective neurons [23 , 24] . The described association neurons and their response characteristics suggest such cells as neuronal correlates of semantic association . We observed that many neurons associated visual shapes with numerical values transiently , and not until the end of the delay period ( Figure 4A ) , whereas prospective activity typically dominates near the end of the delay [29] . More importantly , a high proportion of neurons associated numerical values in the shape and dot displays right after sample onset ( see Figure 4A and 4B ) . This argues for a direct involvement of these neurons in linking numerical values to shapes , rather than encoding upcoming match stimuli in a prospective manner . Finally , an analysis of error trials ( see Figure 5 ) revealed that tuning correlation between both protocols was weakened whenever the monkeys failed to associate visual shapes with their corresponding numerosities . This again provides evidence that association neurons constitute a neuronal correlate for the monkeys' ability to link signs with numerosities . While quantity representations are spontaneously developed [3 , 6] , associations between visual shapes and numerical categories clearly have to be learned by mapping shape representations onto numerical categories . This neuronal learning could start with two classes of PFC cells: one class encoding visual characteristics of shapes ( input possibly via inferotemporal cortex [35] ) , the other class representing numerical information most likely received from the IPS [23 , 24] . According to the Hebbian learning rule [36] , the connections may be strengthened between these two classes of neurons so that cells encoding matching pairs ( e . g . , Arabic numeral 3 and three dots ) are interconnected and become associative . This learning behavior could potentially be modeled via a recurrent neuronal network as has been done for pair-association encoding in inferotemporal neurons [37] or for somatosensory parametric working memory in PFC [38] . Even though numerosity-selective neurons in IPS are relatively abundant and encode numerical information earlier than PFC neurons [23] , association neurons were surprisingly rare in the parietal lobe . Moreover , IPS neurons differentiated to a larger extent between the sensory features of the visual displays; they responded less abstractly than PFC neurons , which generalized across visual properties . At first glance , the sparseness of association IPS neurons in the nonhuman primate seems to be at odds with the well-known role of the posterior parietal cortex in adult humans for both non-symbolic [17 , 18] and symbolic numerical cognition [18–20] . Beyond possible species-specific differences between humans and monkeys , this difference might also be the consequence of training duration; our monkeys were trained for few months to match numerosities with visual shapes , whereas humans acquire symbols over years . Because of the monkeys' inferior proficiency , it is likely that the shape–numerosity association was not automatically executed in the monkey brain , but required a strong involvement of the PFC in order to manage the high cognitive demands [34] . Support for this assumption comes from recent functional magnetic resonance imaging studies with human children . In contrast to adults , preschoolers lacking proficiency with number symbols show elevated PFC activity when dealing with symbolic cardinalities [39–41] . Only with age and proficiency does the activation seem to shift to parietal areas . This frontal-to-parietal shift has been interpreted as being a result of increasing automaticity in number tasks . This shift of symbolic associations to the parietal lobe could release the limited cognitive resources of PFC for new demanding tasks [34] . The PFC could , thus , be ontogenetically and phylogenetically the first cortical area establishing semantic associations , which might be relocated to the parietal cortex in human adolescents [27 , 42] in parallel with the maturing language capabilities [43] that endow our species with a sophisticated symbolic system [42] . During cultural evolution , humans invented number symbols as mental tools . Number symbols endow our species with an exact understanding of cardinality and the ability to execute the most complicated calculations . Given that the first ancient number symbols have been dated back to only a couple of thousand years ago [44] , it is impossible that the human brain has developed areas with distinct , culturally dependent number symbol functions [27] . It is more parsimonious to assume that existing brain structures , originally evolved for other purposes , are reused and built upon in the course of continuing evolutionary development ( by a process called “exaptation” [45] ) , an idea captured by the “redeployment hypothesis” [46] ( also termed “recycling hypothesis” [27] ) . According to this hypothesis , already existing simpler networks are largely preserved , extended , and combined as networks become more complex , instead of there being a de novo creation of intricate structures [47] . In the number domain , evidence suggests that existing neuronal components ( located in PFC and IPS ) —originally developed to serve nonverbal quantity representations—are used for the new purpose of number symbol encoding , without disrupting their participation in existing cognitive processes [18] . While monkeys use the PFC and IPS for non-symbolic quantity representations [23] , only the prefrontal part of this network is engaged in semantic shape–number associations . Interestingly , this pattern of brain area use seems to be preserved in human children [39–41] . Moreover , we found that numerical values represented by signs were encoded more selectively as than analog set sizes . This sharpening of the tuning functions for signs was predicted by a recent network model [48] and might indicate the advent of a digital representation via symbol-like signs in the primate brain . We speculate that our data in the monkey provide a first glimpse of redeployment of the PFC for symbolic-like learning , thus paving the way for the neuronal quantity network to encode real number symbols in language-endowed humans . We trained two monkeys to match either a set of dots with another set of dots ( delayed match-to-sample task , or dot protocol; see Figure 1A ) or a visual shape with a set of dots ( delayed association task , or shape protocol; see Figure 1B ) . Stimuli were sets of black dots or black Arabic numerals pseudo-randomly varying in size and position and displayed on a gray background . A trial started when the monkey grasped a lever and fixated ( ± 1 . 75° of visual angle , monitored with an infrared eye tracking system ) on a central target . After a monkey fixated for 500 ms , the sample appeared for 800 ms ( multiple-dot display in the dot protocol; Arabic numeral in the shape protocol ) . The monkey then had to maintain fixation until the end of a 1 , 000-ms delay period , after which the test stimulus was presented ( always a multiple-dot pattern ) . In 50% of cases the test stimulus was a match , i . e . , it showed the same number of dots as cued during the sample period by a multiple-dot pattern or a shape . In the other 50% of cases the first test stimulus was a nonmatch , which showed—with equal probabilities—either a higher or lower numerosity than the sample display . After a nonmatch test stimulus , a second test stimulus appeared that was always a match . To receive a fluid reward , monkeys were required to release the lever as soon as a match appeared . Trials were pseudo-randomized and balanced across all relevant features ( e . g . , match versus nonmatch , dot versus shape protocol , standard versus control , etc . ) . The stimuli for the dot protocol were randomly arranged black dots displayed on a gray background ( diameter 6° of visual angle ) . For each session , 100 different images per numerosity were generated with pseudo-randomly varied visual features: the diameter of the dots ranged from 0 . 5 to 0 . 9° of visual angle , and their positions were restricted only by the border of the gray background circle and the fact that they were not allowed to overlap each other . Sample and test stimuli were never identical . All four quantities were presented in each session with one standard and one control condition . Controls in the dot protocol included dot displays with constant circumference ( the summed circumference of the dots was constant , such that dot size decreased as dot number increased , as opposed to in the standard condition ) , linear configuration ( i . e . , all dots were linearly arranged ) , and constant density ( i . e . , constant mean distance between dots ) across all presented quantities ( see Figure 1C ) . These measures prevented the monkeys from memorizing visual patterns instead of using the numerical information to solve the task . For the shape protocol , a sample stimulus consisted of a black Arabic numeral on a gray background circle . Font size ( range 26 to 42 points ) and position of the shapes were varied pseudo-randomly from trial to trial . The font “Arial” was used for standard trials; “Times New Roman , ” “Souvenir BT , ” and “Lithograph Light” were control fonts ( see Figure 1D ) . The test stimulus for the shape protocol consisted of sets of black dots in the style of the dot protocol . Standard and control trials as well as trials from the dot and shape protocols were pseudo-randomly intermingled and appeared with equal probabilities in each session . These measures ensured that the monkeys generalized to the overall shape characteristics instead of memorizing local features . Recordings were made from one left and one right hemisphere of the ventral convexity of the lateral PFC and in the fundus of the IPS of two rhesus monkeys ( Macaca mulatta ) in accordance with the guidelines for animal experimentation approved by the Regierungspräsidium Tübingen , Germany . These areas were chosen because in preceding studies [21–26] they were shown to contain visual numerosity-selective cells and , from human studies , are known to be activated during numerosity-related tasks [17–20 , 39–41] . Single-cell recordings were made with arrays of tungsten electrodes ( 1–2 MOhm impedance ) . Recording sites were localized using stereotaxic reconstructions from magnetic resonance images . Recordings in the IPS were done exclusively at depths from 9 to 13 mm below the cortical surface ( Horsley-Clark coordinates , anterior/posterior , −5 mm or 0 mm ) [24] . No attempts were made to preselect neurons . Off-line sorting was routinely applied to separate single units . As of the publication of this article , both monkeys are still engaged in discrimination tasks . To determine the neuronal response latencies , averaged spike density histograms were derived with a 1-ms resolution , smoothed by a sliding window with a kernel bin width of 10 ms for all sample stimuli . A 200-ms time window before stimulus onset was used as baseline . If five consecutive time bins after stimulus onset reached a value higher than the maximum of the baseline period , response latency was defined by the first of these time bins . A default latency of 100 ms was used if no value could be calculated . Putative association neurons were preselected based on a two-factor ANOVA . To account for different temporal response phases , spike rates were tested in four adjacent , nonoverlapping time windows . The first window ( 400 ms ) started at the beginning of the sample period and was shifted by the neurons' response latencies . The second window ( 400 ms ) followed right after the first one and covered the rest of the sample period . The subsequent two windows ( 450 ms each ) covered the first and second part of the delay period . Selectivity for numerical values was calculated based on these discharge rates separately for the dot and shape protocols using a two-way ANOVA with main factors “numerical value” ( one to four ) and “stimulus condition” ( standard and control ) . Cells were considered to be numerosity-selective only if they showed a significant main effect to “numerosity” in one of the four analysis windows , but no significant “stimulus condition” or interaction effect . To derive averaged numerosity-filter functions , the tuning functions of individual neurons were normalized by dividing all spike rates of the tuning functions by the maximum activity , thus setting the activity at the preferred numerical value to 100% . Pooling the resulting normalized tuning curves across the entire population of association cells resulted in averaged numerosity-filter functions ( see Figures 4C , S6A , and S6B ) . The population tuning functions were calculated for the time windows during which association neurons were significantly tuned to numerosity as tested by the two-way ANOVA . If neurons were significantly tuned in more than one window the analysis was restricted to the window with the smallest p-value . The correlation analysis aimed to extract tuning similarities of individual neurons to numerical values in the shape and dot protocols . Figure S2 describes the application flow of the analysis . For each protocol ( Figure S2A and S2B ) , eight trials per numerical value were chosen in a random manner ( Figure S2C ) . Tuning functions were built with the averaged spike rates of these trials ( Figure S2D and S2E ) . Next , the CCs between these tuning functions were calculated . The same subset of trials was shuffled so that the relation between neural activity and numerical value was abolished ( Figure S2F ) ; with this shuffled dataset , we calculated dummy tuning curves ( Figure S2G and S2H ) and computed the CCs ( termed SPs ) between them . This procedure was repeated 1 , 000 times , always using a new random subset consisting of eight trials to create two distributions of CCs and SPs . We quantified the discriminability between these distributions by ROC analysis . This analysis was accomplished for each of the sliding windows separately ( one exemplary window is shown by the shaded bars in Figure S2A and S2B ) . Each separate analysis step is described in more detail below . Out of the set of all trials ( Figure S2A and S2B ) , we randomly drew eight trials per numerosity and protocol ( i . e . , in total four numerosities × two protocols × eight trials = 64 trials per turn; Figure S2C ) . This was done 1 , 000 times with replacement . We took care that no trial combination occurred more than once . The CCs and the SPs were calculated for each turn of the bootstrapping algorithm . This method filters robust effects across trials and provides reliable distributions . The tuning functions tshape and tdot were composed of the spike rates of a given neuron obtained in the shape and dot protocols , respectively . Spike rates were obtained by averaging across the raw spike trains for 100 ms ( see shaded windows in Figure S2A and S2B ) . Each tuning function consisted of four spike rates ( corresponding to the neuron's responses to numerical value n = 1 , 2 , 3 , and 4 during the identical time window ) . The spike rates were combined into one tuning function by sorting them in ascending numerical order ( Figure S2D and S2E ) . The CCs provided a measure to quantify the similarity between tuning to the shape and dot protocols . The rationale behind this was the following . A neuron that was ANOVA-selective in both protocols constituted a potential neuronal association substrate between shapes and numerical values . In addition to the mere selectivity in the dot and shape protocols , however , neurons should have similar tuning functions for the ( direct and associated ) numerical values in both protocols . Neurons showing different tunings to the numerical values in the two protocols cannot be regarded as association neurons and should be excluded . The normalized cross-correlation is an appropriate method for filtering for these criteria . The cross-correlation takes a neuron's entire tuning functions tshape ( n ) and tdot ( n ) for the numerical values n ∈ [1 , 2 , 3 , 4] for dot and shape protocols , respectively , into account , rather than just comparing the preferred numerosities . We calculated the cross-correlation between these tuning functions for the shape and dot protocols . It is scale-invariant , since the means t̄shape and t̄dot are subtracted from each spike rate , and has the advantage of normalization , which allows comparison across all cells . The normalized CC was calculated as follows: The SP is supposed to represent the chance correlation level , irrespective of numerical values . For its calculation we abolished the relationship between neural activity and numerical value by randomly assigning each neural response a numerical value ( Figure S2F ) . Based on the tuning functions of this shuffled dataset ( Figure S2G and S2H ) , we calculated CCs . We termed the distribution of these CCs the SP . Since the SP was calculated within the bootstrapping algorithm ( 1 , 000 repetitions ) , it provides a robust estimate of non-numerical-related fluctuations . In other words , the SP takes accidental correlations into account ( e . g . , those occurring at phasic “on” responses ) and can thus be regarded as baseline correlation irrespective of influences by the presented numerical values . To determine whether a given cell in a given time bin responded more similarly to shape and dot stimuli than expected by chance , we performed a ROC analysis [28] that provided a measure of how well the distributions of CCs and SPs were separated . The SPs were taken as the reference distribution . ROC values greater than 0 . 5 indicated that the CCs of a given cell were higher for the original dataset , arguing for correlated responses in the two protocols . We determined a significance threshold based on the ROC values obtained during the fixation period , during which only random correlations might occur . A neuron was termed an “association neuron” if it reached an ROC value after stimulus onset that was higher than the mean ROC value during the fixation period plus three standard deviations [49] . It needs to be emphasized that significant correlations are not caused by similar overall response modulations in the dot and shape protocols without being related to numerical value . Figure S4A and S4B shows an example neuron that responded very similarly to both protocols . Nevertheless , the CCs were close to zero ( see red line in Figure S4C ) , because this neuron did not show any tuning to numerical value . The SP was also characterized by values fluctuating around zero ( see blue line in Figure S4C ) . Consequently , the ROC analysis did not reveal any significant deviations from chance level ( Figure S4D ) . In contrast , the neurons in Figure 3 showed strong modulations of firing rates with numerical value . As a consequence , the CCs reached high values up to one ( see red lines in Figure 3D , 3I , and 3N ) . At the same time , however , the SP hovered around zero ( see blue line in Figure 3D , 3I , and 3N ) . Thus , the ROC analysis correctly detected the periods of meaningful correlations ( see Figure 3E , 3J , and 3O ) . We calculated the CCs , the SP , and the area under the ROC curve ( AUROC ) in sliding windows ( 100-ms duration , shifted by 25 ms; see shaded area in Figure S2A and S2B ) . This procedure allows a detailed analysis of correlation development over time ( Figure 7A and 7B ) and reveals the different temporal correlation patterns of individual neurons ( Figure 4A ) . We obtained almost identical proportions of association neurons when the analysis was based on nonoverlapping windows of 100-ms duration ( n = 167; values exceeding threshold in at least one window to reach significance ) . We evaluated the link between neuronal responses and behavior by analyzing the influence of erroneous judgments on the neuronal association . To that aim , we calculated CCs between the neuronal tuning functions based on error trials in the shape protocol and neuronal tuning functions obtained from correct trials in the dot protocol . Since the monkeys made very few errors , we often did not collect error trials for all tested numerical values . In these cases we restricted the analysis to the numerical values for which we obtained neuronal data during error trials ( at least two numerical values ) . We compared these error-related CCs with CCs based on correct trials ( again restricted to the same numerical values ) . Was the proportion of neurons tuned to the same numerical value in both the dot and shape protocols higher than expected by chance ? Since some neurons were tuned to numerosity in the dot protocol while others were encoding numerical information in the shape protocol , neurons encoding both formats may simply emerge by chance . We therefore compared the actual frequency of neurons with identical preferred numerical values in both protocols to chance occurrence based on probability calculations . To that aim , we considered the following three events: a cell is shape-selective , a cell is dot-selective , and a cell is selective for shapes and dots , formally written as Based on our dataset , we calculated the probabilities that a cell encodes a specific preferred numerical value n in one of the protocols alone , given that the cell was ANOVA-selective to any numerical value in both protocols ( P ( shape = n\|sig in both ) for the shape protocol and P ( dot = n\|sig in both ) for the dot protocol ) . To obtain the probability that a cell is encoding the preferred numerical value n in both protocols , given that the cell is selective to any numerical value in both protocols ( P ( ( shape = n ∧ dot = n ) \|sig in both ) ) , the two obtained probabilities were multiplied . This approach was legitimate , because the two probabilities P ( shape = n\|sig in both ) and P ( dot = n\|sig in both ) are independent because of the pseudo-randomized presentation protocol . Thus , we can phrase the probability that a cell by chance encodes a specific shape and a specific number of dots simultaneously given that the cell is significant in both formats as In total , the overall probability that a cell encodes one of the n shapes and the respective associated number of dots by chance , given that the cell is significant in both protocols , is the sum of the probabilities for all n: The predicted chance probability Ppred was compared to the observed probability calculated as the percentage of cells with the same preferred quantity in both protocols in the pool of cells that were ANOVA-selective in both the dot and shape protocols . We calculated binomial tests with Ppred as test proportion . The observed fractions in the PFC differed significantly from the test proportions during sample and delay period ( p < 0 . 001 , n = 93 , Ppred = 0 . 30 , and p < 0 . 001 , n = 139 , Ppred = 0 . 31 , respectively ) . The fraction of neurons in the IPS with the same preferred numerical value in both protocols was very small but differed significantly from the predicted frequency during the sample and delay period ( p < 0 . 001 , n = 5 , Ppred = 0 . 32 , and p < 0 . 001 , n = 16 , Ppred = 0 . 25 , respectively ) . The results are depicted as fractions of the entire sample of recorded neurons ( both selective and unselective ) in Figure S5E . This analysis represents a parallel argumentation line to the cross-correlation analysis . It shows on a stochastic basis that associations of visual signs and numerical values is not a coincidence .	We use symbols , such as numbers , as mental tools for abstract and precise representations . Humans share with animals a language-independent system for representing numerical quantity , but number symbols are learned during childhood . A first step in the acquisition of number symbols constitutes an association of signs with specific numerical values of sets . To investigate the single-neuron mechanisms of semantic association , we simulated such a mapping process in rhesus monkeys by training them to associate the visual shapes of Arabic numerals with the numerosity of multiple-dot displays . We found that many individual neurons in the prefrontal cortex , but only a few in the posterior parietal cortex , responded in a tuned fashion to the same numerical values of dot sets and associated shapes . We called these neurons association neurons since they establish an associational link between shapes and numerical categories . The distribution of these association neurons across prefrontal and parietal areas resembles activation patterns in children and suggests a precursor of our symbol system in monkeys .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "neuroscience" ]	2007	Semantic Associations between Signs and Numerical Categories in the Prefrontal Cortex
Paracoccidiodomycosis ( PCM ) is a clinically important fungal disease that can acquire serious systemic forms and is caused by the thermodimorphic fungal Paracoccidioides spp . PCM is a tropical disease that is endemic in Latin America , where up to ten million people are infected; 80% of reported cases occur in Brazil , followed by Colombia and Venezuela . To enable genomic studies and to better characterize the pathogenesis of this dimorphic fungus , two reference strains of P . brasiliensis ( Pb03 , Pb18 ) and one strain of P . lutzii ( Pb01 ) were sequenced [1] . While the initial draft assemblies were accurate in large scale structure and had high overall base quality , the sequences had frequent small scale defects such as poor quality stretches , unknown bases ( N's ) , and artifactual deletions or nucleotide duplications , all of which caused larger scale errors in predicted gene structures . Since assembly consensus errors can now be addressed using next generation sequencing ( NGS ) in combination with recent methods allowing systematic assembly improvement , we re-sequenced the three reference strains of Paracoccidioides spp . using Illumina technology . We utilized the high sequencing depth to re-evaluate and improve the original assemblies generated from Sanger sequence reads , and obtained more complete and accurate reference assemblies . The new assemblies led to improved transcript predictions for the vast majority of genes of these reference strains , and often substantially corrected gene structures . These include several genes that are central to virulence or expressed during the pathogenic yeast stage in Paracoccidioides and other fungi , such as HSP90 , RYP1-3 , BAD1 , catalase B , alpha-1 , 3-glucan synthase and the beta glucan synthase target gene FKS1 . The improvement and validation of these reference sequences will now allow more accurate genome-based analyses . To our knowledge , this is one of the first reports of a fully automated and quality-assessed upgrade of a genome assembly and annotation for a non-model fungus . Paracoccidioides spp . is a thermally dimorphic pathogenic fungus that causes paracoccidioidomycosis ( PCM ) , a neglected health-threatening human systemic mycosis endemic to Latin America where up to ten million people are infected . Disease can progress slowly , with roughly five new cases of disease per million infected individuals per year , with a male to female ratio of 13 to 1 . About 80% of PCM cases occur in Brazil , followed by Colombia and Venezuela [2] . Within the Paracoccidioides genus , the three characterized phylogenetic lineages of P . brasiliensis ( PS2 , PS3 , S1 ) and the one characterized lineage of P . lutzii ( Pb01-like ) can infect humans , and these groups can vary in virulence and induce different immune responses by the host [3] , [4] . To better understand the pathogenesis and to enable genomics-based studies , the genomes of Paracoccidioides spp . were sequenced , analyzed and made publicly available in 2011 [1] . The Broad Institute of MIT and Harvard in partnership with the Paracoccidioides research community selected three reference isolates for sequencing and genomic analysis; assembly size for these strains varied between 29 . 1 and 32 . 9 Mb , and between 7 , 875 and 9 , 132 genes were identified in each strain [1] . These included two strains of P . brasiliensis ( Pb18 representing the S1 lineage and Pb03 representing the PS2 lineage ) and one strain of P . lutzii ( Pb01 ) [1] . These sequenced isolates are extensively referenced in molecular biology and experimental mycology laboratories working with Paracoccidioides spp . and also other pathogenic fungi , including those working with yeast phase specific genes expressed during host infection . These sequences also serve as a reference to analyze high-throughput data increasingly generated by genomic , metagenomic , transcriptomic and proteomic approaches . Additionally , accurate sequences are critical for evolutionary analyses , e . g . , to identify positively selected genes , as well as to provide new targets for the design of diagnostic assays . The P . brasiliensis Pb18 and Pb03 strains and the P . lutzii Pb01 strain were sequenced using the sequencing technology and computational methods available at the time , which produced high quality draft assemblies . However , the assemblies included a large number of gaps and uncertain or low quality nucleotides in the final consensus sequences . Also , the annotation pipelines flagged only the most extreme annotation errors for curation and did not address the larger number of smaller scale errors in the gene models and underlying sequence [5] . Correction of such errors requires re-evaluation of the assembly consensus sequence and associated annotation . Assembly errors that could not be detected in previous data and passed standard quality control criteria at that time can now be corrected using next generation sequencing ( NGS ) for systematic assembly improvement . These include errors in gene-containing regions of the original genomic assembly affected by poor quality sequence or ambiguities , which can cause incorrect gene structure predictions . Since predicted genes of reference genomes are now frequently used for homology-based inference or confirmation of gene structures in closely related species , errors in the reference sequence may be propagated to other genomes [6] , [7] . Therefore , systematic improvement of a genome assembly and annotation can impact not just the understanding of that particular species , but also that of other related species for which it is used as a reference for comparison . Here , we present an update of the three Paracoccidioides reference genome sequences achieved using Illumina re-sequencing to correct assembly errors and document the improvements obtained . The improved and updated reference genome assemblies and annotations of this important human fungal pathogen now allow more accurate SNP analyses , genome-wide evolutionary ( e . g . , selection ) analyses that depend on high-quality sequences , phylogenetic footprinting studies of regulatory regions , or primer and probe design for diagnostic assays . Three reference isolates of Paracoccidioides spp . ( Pb01 , Pb03 and Pb18 ) , representing two species , were previously sequenced . The isolate of P . lutzii ( Pb01 ) was a clinical isolate originating from an acute form of paracoccidioidomycosis ( PCM ) in an adult male . The two P . brasiliensis isolates were from individuals presenting chronic PCM; Pb03 represents the PS2 phylogenetic group and Pb18 the S1 group [1] , [4] . In a partnership between the Broad Institute and the Paracoccidioides research community , these genomes were previously sequenced using multiple whole genome shotgun libraries constructed from genomic DNA for each strain; paired-end sequences were generated for each with Sanger technology and assembled using Arachne [1] ( assembly v1; S1 Table ) . The reference strains Pb01 ( previously sequenced DNA sample ) and Pb03 and Pb18 ( newly extracted DNA samples ) were re-sequenced using Illumina technology . For library construction , 100 ng of genomic DNA was sheared to ∼250 bp using a Covaris LE instrument and prepared for sequencing as previously described [8] . A library for each of the three samples was used to generate 101 base paired-end reads on the Illumina HiSeq2000 platform , producing an average genome coverage of 165X . To improve the genome sequence of Paracoccidioides spp . strains Pb01 , Pb03 and Pb18 , Illumina paired-end reads were aligned to the draft reference assemblies ( assembly v1 ) using BWA version 0 . 5 . 9 with default settings [9] . The assembly consensus sequence was re-evaluated by providing these alignments as input to the automated assembly improvement program Pilon ( version 1 . 4 , default parameters , www . broadinstitute . org/software/pilon/ ) . Pilon uses the Illumina read alignments for multiple classes of assembly correction . First , Pilon scans the read alignments for positions where the sequencing data disagree with the input genome ( assembly v1 ) and corrects small errors such as single nucleotide differences and small insertion/deletion events . Second , Pilon looks for coverage and alignment discrepancies to identify potential mis-assemblies and larger variants . Finally , Pilon uses reads anchored adjacent to discrepant regions and gaps in the input genome to reassemble the region , attempting to fill in the true sequence including large insertions . As output , Pilon provides the sequence of this improved genome assembly ( assembly v2; Fig . 1 ) along with files summarizing the changes and quality measures used in the assessment . Protein-coding genes were predicted in the improved assemblies ( assembly v2 ) using a combination of gene models from the prediction programs Augustus [10] , Genemark-ES [11] , GlimmerHMM [12] , Genewise [13] , and Snap [14] , as well as automated revision based on EST data ( e . g . , from [15] ) and manual gene revision of flagged calls . The predicted gene sets were then provided as input to EvidenceModeler ( EVM ) [16] to obtain the best consensus model for a given locus . The consistency of the gene models was evaluated by examining alignments of protein orthology groups identified using OrthoMCL [17] . EVMLite was used to rescue orphan genes not captured in EVM; only those genes with additional evidence such as overlap to Genewise or non-repeat HMMER3 PFAM domains were rescued , as well as non-redundant genes overlapping the OrthoMCL genes in clusters containing 2 or more genomes . Lastly spurious gene models matching repetitive or low-complexity sequences were removed . For each Paracoccidioides genome , we compared the original annotation ( v1 ) with the updated annotation ( v2 ) to evaluate the changes in the new gene sets . To precisely characterize the types of changes across the v1 and v2 annotations , we first mapped the corresponding gene between the two assemblies . The v1 and v2 assemblies were aligned using nucmer [18] , and the alignment coordinates were used to assign gene correspondence between the initial annotation v1 and the new annotation v2 . This mapping also allowed us to preserve locus numbers in the updated gene set . Each annotated gene was assigned a locus number , keeping where appropriate the previous locus number of the form PAAG_##### ( Pb01 ) , PABG_##### ( Pb03 ) or PADG_##### ( Pb18 ) , which serves as a unique identifier within each genome and across assemblies . New genes , merged genes , and genes with large structure and sequence changes in transcripts were assigned new and unique locus numbers following the last locus number of annotation v1 . Locus numbers of deleted genes do not appear in the final gene sets . To evaluate whether the changes in gene sequence and structure produced a more accurate gene set , the gene sequences of annotation v1 and of the updated annotation v2 were compared via sequence similarity and orthology analysis . To evaluate the consistency of gene structures for orthologs as well as their conservation between species , OrthoMCL version 1 . 4 with a Markov inflation index of 1 . 5 and a maximum e-value of 1e-5 was used to identify orthologous clusters across the six total protein sets corresponding to annotations v1 and v2 of each Paracoccidioides strain . For each cluster group representing putative orthologs , we compared the maximum length difference among the three Paracoccidioides genes in annotation v2 to that in the annotation v1 . To compare the functional content of the v1 and v2 gene sets , we evaluated both protein domain families ( PFAM ) and pathway information ( KEGG ) . Using HMMER3 [19] , we mapped v27 of the PFAM domain database [20] to both the v1 and v2 gene sets . KEGG domains [21] from release 65 were also mapped to both gene sets using BLAST . To evaluate changes in the gene structure , the corresponding transcripts from annotation v1 and v2 were identified as described above . We also aligned gene sets v1 and v2 using BLASTn [22] version 2 . 2 . 28+ with default parameters , using an in house Perl script to determine the types of modification for each gene , which included changes in gene length , gene coverage and percent nucleotide identity . We manually checked a random gene sample of each type of change ( up to 10 genes ) from both gene correspondence and BLAST analyses to verify that changes in gene set v2 were actually gene improvements . To evaluate changes in the coding regions of genes of high interest to the community , we selected known specific yeast-phase genes or virulence factors of Paracoccidioides spp . , as well as other genes that are generally considered relevant for research on Paracoccidioides or related dimorphic pathogens , for manual review . The sequences of these genes' coding regions were aligned at the protein level with CLUSTALW [23] version 2 . 1 , using both the v1 and v2 annotations . The coverage of Core Eukaryotic Genes defined by CEGMA [24] was evaluated using the CoreAlyze tool ( http://sourceforge . net/projects/corealyze/ ) to summarize results for all the v1 and v2 gene sets . BLASTp version 2 . 2 . 28+ was run with default settings using protein sets from annotations v1 and v2 as the database , with Saccharomyces cerevisiae and Schizosaccharomyces pombe CEGMA proteins as the query . We also included the protein gene sets of two close relatives of Paracoccidioides , the dimorphic fungal pathogens Blastomyces dermatitidis and Histoplasma capsulatum . In order to obtain a detailed picture of the changes where gene annotations were modified but not completely overridden , we compared protein sequences between the two versions , excluding proteins that were added or deleted from the final gene set v2 , as well as proteins for which the new annotation was for a completely different transcript at the same locus . For a hit to be counted , the protein needed to match a protein in the reference set with at least 75% identity for the v1 and v2 annotations . This percent identity cutoff was determined empirically to eliminate spurious low similarity alignments . The percent identity and the bit score between the query protein and each version of the Paracoccidioides annotations ( v1 and v2 ) were compared . The strains of the genomes of Paracoccidioides spp . previously sequenced [1] , Pb18 and Pb03 and Pb01 , were re-sequenced using Illumina 101 bp paired-end reads . This sequencing generated 93 . 6 million reads for Pb18 with an average coverage of 198X , 124 . 2 million reads for Pb03 with an average coverage of 150X and 110 . 0 million reads for Pb01 with an average coverage of 148X . This high coverage sequence data was then used to refine the consensus sequence of the original assembly by assessing differences between the new sequence and the previous assemblies . This can target a wide range of improvements , including correcting base calls , resolving ambiguous bases and closing gaps within scaffolds . Fig . 1 shows a simplified overview of the workflow of genome improvement . The new Illumina data were used to systematically improve the three Paracoccidioides spp . assemblies using Pilon ( http://www . broadinstitute . org/software/pilon/ ) . Pilon bases its improvement calls on an alignment of the reference genome and the sequenced reads . The aligned bases and depth at each sequenced position provides evidence for the reference base or for an alternative; where changes are supported they can result in single base differences , insertion or deletion of single bases or larger regions , identification of collapsed regions and more complex changes and gap filling based on local reassembly . Reads of each of the genomes of Pb18 , Pb03 and Pb01 were aligned to the corresponding reference assembly using BWA [9] and the resulting bam file was used as input for Pilon . In each of the Paracoccidioides assemblies , Pilon identified and fixed base errors in the consensus sequence . The statistical improvements for the assemblies v2 of Paracoccidioides spp . are summarized in Table 1 . The most frequent class of changes was single base substitutions , identified as single nucleotide polymorphisms ( SNPs ) between the assembly and reads . Between 3 , 018 and 3 , 290 single base errors were corrected in each assembly . Small insertions and deletions were also incorporated into each assembly . The major classes of changes can be attributed to bases added and removed in reassembly fixes , collapsed bases in the new assembly and the closing of gaps ( Table 1 ) . Regions of misassembly identified and fixed by Pilon resulted in bases added or removed but no new gaps opened . Across all the assemblies , 20% of all initial gaps were closed by Pilon; the number of gaps closed were 113 , 56 and 212 , for Pb18 , Pb03 and Pb01 , respectively . Overall , the assembly improvement process led to an increase of contig N50 for all strains . About ∼99% of low quality nucleotides in assemblies v1 were well supported or fixed with high flag coverage in assemblies v2 . Overall , the P . lutzii Pb01 genome assembly was most substantially improved , based on comparing statistics for all v1 and v2 assemblies ( S1 Table ) . The contig N50 for Pb01 v2 increased by 29 . 1 kb; more bases were added and removed after re-assembly fixes and more gaps were closed than in the other two genomes . The genome size and number of scaffolds of Pb18 and Pb03 were essentially unchanged . The Pb01 genome size decreased slightly from 32 . 94 to 32 . 93 Mb; the updated assembly contains one scaffold fewer , as two scaffolds were merged by closing the gap between them . The number of contigs was reduced in all three strains , which considering the increase in the contig N50 indicates that the assemblies v2 for Pb18 , Pb03 and Pb01 were less fragmented . All these changes indicate that the genome assemblies v2 after Pilon improvements were more contiguous , contained more bases with high quality , and had fewer gaps and errors . The gene annotations of the reference strains Pb18 , Pb03 and Pb01 were updated using a pipeline to transfer and revise gene structures ( Methods ) . The implemented annotation pipeline was an updated and improved version of the previous protocol used to annotate Paracoccidioides spp . assemblies v1 . The current pipeline includes an updated set of gene prediction programs , including the EVM caller used to select the best call for each locus . Databases used for training these gene prediction programs are also more comprehensive , with more sequences available from the dimorphic fungi group for comparison . Also , the databases used for homology inference and functional annotation were updated since the previous annotation . In addition , we identified orthologs to evaluate the gene calls for consistency ( see below ) . The incorporation of these new methods and data improved the evidence supporting gene prediction . The updated gene sets are more consistent across the three Paracoccidioides genomes ( S1 Table ) . The total gene count for the two P . brasiliensis genomes now only differs by 37 in v2 whereas the v1 gene counts differed by 866; overall the update removed 351 genes from Pb18 and added 552 genes to Pb03 . P . lutzii ( Pb01 ) also has a more similar gene count , due to 306 fewer genes in the v2 compared to v1 . A more detailed view of the gene structure changes by major categories is provided in Table 2; these statistics were calculated by mapping the transcripts from the previous annotation to the corresponding locus on assembly v2 . Notably , this analysis helped recover a large number of genes missed by the original annotation in each genome; the total genes newly added to a region was 840 in Pb18 , 933 in Pb03 and 936 in Pb01 . In addition , dubious genes were removed from each genome; the number of genes no longer present at the same locus was 1187 in Pb18 , 490 in Pb03 and 1265 in Pb01 . Other changes include extending or truncating transcripts , merging or splitting transcripts , changes to splice sites , and changes to UTRs ( Table 2 ) . Only 23% of genes in the v2 annotations were unchanged from v1; the primary transcripts were identical for 1816 genes in Pb18 , 2599 in genes Pb03 and 1581 in genes Pb01 . Genes with any type of change in their coding sequences represent a smaller subset , in total 5734 ( 68% ) in Pb18 , 4895 ( 58% ) in Pb03 and 6309 ( 71% ) in Pb01 . Both sequence addition ( gap filling and local reassembly ) and small changes in the genome assemblies ( single-nucleotide substitutions or insertion/deletion events ( indels ) ) , contributed to the improvement of the gene annotation in the update v2 . Two examples of how indel correction fixed gene structures are shown in Fig . 2 . In the first case ( left panel ) , an extra C was inserted at a polyC tract in PABG_00129 of Pb03; correction of this position resulted in extending the coding DNA sequence ( CDS ) of this gene by 423 bases . In the second case ( right panel ) , an A was deleted at a polyA tract in PABG_00790; correction of this position also corrected the reading frame , allowing for removal of a false intron that was needed to step over a stop codon and extension of the CDS of this gene by 252 bases . While these are small changes to the underlying assembly , both have had larger impact on correcting these gene structures . The annotation improvements were also analyzed by comparing the alignments of orthologs for all three Paracoccidioides genomes , identified by OrthoMCL ( Methods ) . For orthologs identified either from the v1 or v2 assemblies , maximum and minimum gene length was computed for each ortholog cluster . In comparing these gene lengths ( Fig . 3A ) , the v1 gene annotations ( red points in scatterplot ) exhibited a higher variation among Pb18 , Pb03 and Pb01 orthologs compared to annotation v2 ( blue points ) . The positions that are closer to the diagonal correspond to smaller differences in gene length between orthologs; as expected for an improved annotation , the v2 points are closer to the diagonal than v1 . These differences between maximum and minimum length of the genes within each orthologous cluster group were also plotted on a logarithmic scale ( Fig . 3B ) , based on sorting cluster differences from smallest to largest . The v2 annotation differences ( blue curve ) were lower and well separated from the v1 annotation ( red curve ) , providing additional support of the increased length concordance in the v2 annotation . Further analysis of gene conservation also supported the greater consistency among the Paracoccidioides spp . genomes in the v2 annotation . The number of genes found in all three genomes increased , whereas the number of unique genes specific to only one genome decreased; this has produced a more uniform set of protein coding genes ( Figure S2 ) . The improved structural annotation also led to improvements in functional annotation . The v2 annotation had more genes with assigned protein domain families ( PFAM ) and pathway information ( KEGG ) , using the same version of these databases for the v1 and v2 gene sets ( Figure S2 ) . This supports the higher functional content of the revised gene sets , despite the lower total gene counts in two of the genomes . We also manually reviewed and curated the predicted structures of a number of protein-coding genes that are of importance to the Paracoccidioides research community , including well-characterized yeast-phase specific genes and other virulence factors . This introduces changes to the transcript sequence of 27 of these genes ( Table 3 ) . The improvements to the assemblies resulted in updated transcript predictions for the vast majority of genes of the three reference strains , with substantially corrected gene structures for several virulence-associated or yeast-phase specific genes of central importance in Paracoccidioides or other dimorphic fungi , including HSP90 [25] , PbGP43 [26] , PbP27 [27] , RYP1-3 [28] , BAD1 [29] , catalase B , alpha 1 , 3 glucan synthase and the beta glucan synthase target gene FKS1 [30] , [31] . An extreme example is the HSP90 gene , where corrections were made to the sequence of each of the three Paracoccidioides genomes ( Fig . 4 ) . This example illustrates the annotation errors in v1 of all Paracoccidioides reference strains that were fixed in v2 after Pilon improvement and re-annotation . In this case one or more single-nucleotide errors , unknown single nucleotides ( N's ) , and/or single nucleotides that were erroneously reported as absent or duplicated by Sanger sequencing resulted in radically different predicted gene structures ( intron/exon and/or gene boundary errors ) . This is shown in detail for a cluster of errors present at the end of HSP90 in Pb03 ( Fig . 4B ) , which included alteration of the proper stop codon , resulting in premature truncation of this gene . Another example of coding sequence updates to multiple genomes is shown for FKS1 , where different regions of the Pb03 and Pb18 proteins were restored in the updated assemblies and annotations ( Figure S1 ) . The improvements in the annotation v2 were also analyzed for completeness by comparing to a set of highly conserved fungal genes defined by CEGMA and to protein sets of related dimorphic human pathogenic fungi . Genes in the v2 annotation showed a higher coverage of both the CEGMA and related dimorphic fungal data sets in comparison with annotation v1 , suggesting these v2 genes are more complete ( Figure S3 ) . Furthermore , we examined the level of conservation to other fungi , by analyzing the difference of the BLASTp score between the v1 and v2 protein sets compared to those of B . dermatitidis , H . capsulatum and the CEGMA genes of S . cerevisiae and S . pombe . We observed that in all cases the v2 annotation had more hits greater than the minimum-similarity cutoff , and that the vast majority of genes of the v2 annotation had higher BLAST score values than their counterparts from v1 annotation ( Figure S4 ) . The initial draft genomes of three isolates of Paracoccidioides ( P . brasiliensis isolates Pb03 and Pb18 , and P . lutzii isolate Pb01 ) served as the first complete genome references for this fungal species [1] . Although these assemblies were obtained using the best technology available at that time , they included gaps and low quality sequence in genic and intergenic regions , which in turn resulted in a number of suboptimal gene structures , coding sequences and predicted protein sequences . This work has revised these reference genomes , providing the Paracoccidioides community with more complete and accurate sequences; this provides a more accurate foundation for future genome-based , molecular biological or genetic research on paracoccidioidomycosis and the fungal strains that cause it . The strategy we have followed will more widely be useful also for other groups wishing to update fungal and other microbial genomes in future . The updating of a reference genome , in particular of the underlying assembly and annotation , can be thought of as a largely computational form of deep sequence curation . The success of the update we present here shows that next-generation sequencing ( NGS ) together with publicly available software tools can markedly enhance the quality of a eukaryotic genome resource . Indeed , the availability of affordable NGS sequencing opportunities makes such endeavors accessible to small bioinformatics groups . The massively computer-assisted component of such an update , which can include tabular and graphical views for monitoring improvements and performing quality control , can be complemented by choosing and following a few ‘guide genes’ to evaluate the process . This focused analysis provides tangible examples of how the update affected predicted properties of important genes , such as gene structure or encoded proteins . The accuracy of a genome sequence and associated annotations are critically important for many types of analysis; therefore validating and improving the accuracy of the sequence and annotation can have wide impact , especially for methods highly sensitive to sequence errors . One example involves examining a genome sequence for evidence of genes and genic regions likely to be under positive selection . Such genes and genic regions , which are believed to be relatively rare in many eukaryotic genomes ( see , e . g . , [32] ) , are sometimes associated in pathogenic organisms with virulence or rapid adaptation to host conditions , including resistance to defense by the host or avoidance of the host immune system . An example of such adaptation has been found for surface proteins of diverse pathogens [32] and in fungi of the proline-rich antigen gene in Coccidioides spp . [33] . Positive selection can also occur in response to antimicrobial drugs , as in chloroquine resistance in Plasmodium falciparum [34] . Candidate regions under positive selection are commonly identified as sections of coding regions having unusually high rates of nonsynonymous ( amino-acid changing ) substitutions . Precisely because such regions are quite rare , a coding region of low sequence quality having several sequencing errors could be categorized as a region under positive selection , and if there are several such regions in a genome , an automated genome-wide screen will report a high percentage of false positives . Conversely , assembly and annotation improvements such as we describe here can effectively evaluate and fix such regions of a genome so that even error-sensitive evolutionary analyses become realistic . Similar considerations also apply to analyses that have more obvious clinical relevance . For example , improving a DNA sequence's accuracy can bring it closer to being ‘clinical grade’ or ‘diagnostic grade’ . Indeed , identification of a clinical sample of a human pathogenic fungus isolated in a hospital using sequence comparison requires certainty that any nucleotide differences ( e . g . , resulting from single nucleotide polymorphisms/SNPs ) observed between the sequenced sample and trusted reference strain ( s ) or isolate ( s ) are not simply errors in the reference . For fungi encountered in clinical contexts , only one or a few traditionally used loci are typically represented by reliable reference sequences , which are often from the ribosomal DNA , or from one or two protein-coding genes known beforehand to be diagnostically informative . Reference diagnostics , as well as diagnostic PCR assays ( e . g . , primer/probe design in real-time PCR assays ) , depend on such regions that have been reliably characterized at the molecular level in a fair number of related species or strains that could be present in clinical settings . Whole-genome gene sets offer , however , new perspectives; if their sequence quality is high , one could then systematically and exhaustively screen alignments of the full gene sets for diagnostically promising genes and genic regions that are likely to be informative for the identification task at hand . Such genome-wide screens should be able to identify new , candidate target loci , and molecular assays could then be developed for them and validated . Genome sequences also allow for metagenomic or metatranscriptomic analysis , where reference genomes enable identification of the pool random sequence from the population of microbes in a sample . Such wide applications will be better powered by efforts such as this to improve the set of reference genomes that form a fundamental basis of comparison and analysis . By re-sequencing three reference strains of Paracoccidioides spp . , using deep sequencing depth of Illumina paired-end reads , we have been able to substantially improve the assemblies and annotations for this important human fungal pathogen . Here we have presented the updated and improved annotated genome sequences , which constitute new references that can be used in diverse future molecular projects by those working in the field of medical mycology . Since the process leading to the new sequences is largely automated using publicly available programs and the NGS technology used is cost-effective , the success of our strategy represents a proof of concept that may stimulate similar updates of other genomes in future .	The fungal genus Paracoccidioides is the causal agent of paracoccidioidomycosis ( PCM ) , a neglected tropical disease that is endemic in several countries of South America . Paracoccidioides is a pathogenic dimorphic fungus that is capable of converting to a virulent yeast form after inhalation by the host . Therefore the molecular biology of the switch to the yeast phase is of particular interest for understanding the virulence of this and other human pathogenic fungi , and ultimately for reducing the morbidity and mortality caused by such fungal infections . We here present the strategy and methods we used to update and improve accuracy of three reference genome sequences of Paracoccidioides spp . utilizing state-of-the-art Illumina re-sequencing , assembly improvement , re-annotation , and quality assessment . The resulting improved genome resource should be of wide use not solely for advancing research on the genetics and molecular biology of Paracoccidioides and the closely related pathogenic species Histoplasma and Blastomyces , but also for fungal diagnostics based on sequencing or molecular assays , characterizing rapidly changing proteins that may be involved in virulence , SNP-based population analyses and other tasks that require high sequence accuracy . The genome update and underlying strategy and methods also serve as a proof of principle that could encourage similar improvements of other draft genomes .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "fungal", "genomics", "mycology", "bioinformatics", "genome", "analysis", "fungal", "genetics", "database", "and", "informatics", "methods", "genetics", "biology", "and", "life", "sciences", "computational", "biology", "software-aided", "sequence", "analysis", "research",...	2014	Genome Update of the Dimorphic Human Pathogenic Fungi Causing Paracoccidioidomycosis
Small RNAs ( sRNAs ) are becoming increasingly recognized as important regulators in bacteria . To investigate the contribution of sRNA mediated regulation to virulence in Vibrio cholerae , we performed high throughput sequencing of cDNA generated from sRNA transcripts isolated from a strain ectopically expressing ToxT , the major transcriptional regulator within the virulence gene regulon . We compared this data set with ToxT binding sites determined by pulldown and deep sequencing to identify sRNA promoters directly controlled by ToxT . Analysis of the resulting transcripts with ToxT binding sites in cis revealed two sRNAs within the Vibrio Pathogenicity Island . When deletions of these sRNAs were made and the resulting strains were competed against the parental strain in the infant mouse model of V . cholerae colonization , one , TarB , displayed a variable colonization phenotype dependent on its physiological state at the time of inoculation . We identified a target of TarB as the mRNA for the secreted colonization factor , TcpF . We verified negative regulation of TcpF expression by TarB and , using point mutations that disrupted interaction between TarB and tpcF mRNA , showed that loss of this negative regulation was primarily responsible for the colonization phenotype observed in the TarB deletion mutant . Vibrio cholerae is the causative agent of cholera [1] , a disease characterized by voluminous secretory diarrhea that is frequently fatal in the absence of treatment [2] . Cholera is endemic in parts of South Asia and Africa and is capable of causing massive epidemics whenever clean drinking water is lacking . While the precise in vivo signals that lead to expression of the pathogenesis program in V . cholerae have not yet been determined , the regulatory events leading to expression of the primary virulence factors , cholera toxin ( CTX ) and the toxin co-regulated pilus ( TCP ) , have been well studied and the major protein factors in the cascade have been identified [3] , [4] . Central to transcription of the major virulence factors is production of the AraC family transcriptional activator ToxT [5] , [6] . ToxT activates production of CTX and TCP by binding to sequences known as toxboxes upstream of the −10 and −35 promoter elements in those operons and stimulating transcription [7] , [8] . ToxT has also been shown to inhibit expression of the mannose-sensitive hemagglutinin ( MSH ) pilus , which is an anti-colonization factor , both by stimulating its degradation and inhibiting its transcription [9] . Expression of these and other factors during infection is dynamic [10]–[12] presumably due to rapidly changing conditions within the small intestine as the infection proceeds . We hypothesized that some steps in this dynamic expression may be controlled by ToxT-regulated small non-coding RNAs ( sRNAs ) . Such regulators would have the advantage of being fast acting since an sRNA need only be transcribed in order to function . sRNAs influence a variety of processes in bacteria , mostly at the post-transcriptional level through sRNA-mRNA interactions [13] , [14] . Processes impacted by sRNA regulators include the DNA damage ( SOS ) response [15] , [16] , sugar uptake [16] , quorum sensing [17] , [18] , expression of outer membrane proteins [19] , [20] and many others . Recent investigation into the sRNA transcriptome of bacteria has indicated much greater complexity than was previously appreciated [16] , [21]–[23] . Given that sRNAs are such ubiquitous regulators of gene expression , we were interested in investigating whether they contributed to virulence factor regulation in V . cholerae . There are several pieces of evidence that suggest the existence of sRNA regulators of virulence in V . cholerae . The major sRNA chaperone Hfq , a protein which many sRNAs act in conjunction with , is required for V . cholerae pathogenesis [24] . In addition , two sRNAs that contribute to virulence were recently discovered . The first regulates the porin OmpA and outer membrane vesicle formation [20] but is not under the control of the virulence regulon , while the second regulates glucose uptake and is a member of the ToxR regulon as it is transcriptionally activated by ToxT downstream of ToxR [25] . To conduct a thorough survey of the possible ToxT-regulated sRNAs , we took a genome-wide approach to discover sRNAs involved in virulence gene regulation by direct cloning and sequencing of sRNA transcripts and by identifying genomic sites bound by purified ToxT . We used direct cloning and deep sequencing of RNA transcripts 50–250 nucleotides in length [16] to compare a culture in which ToxT or an inactive version missing the helix-loop-helix DNA binding domain ( ΔHLH ) [11] was expressed from an arabinose inducible promoter on a plasmid ( pToxT or pToxTΔHLH ) . The highly abundant 5S rRNA and tRNAs present in this size range were depleted prior to sequencing as described [15] . After sequencing we removed residual tRNA and rRNA reads and aligned the remaining reads to the V . cholerae genome . The number of reads of each unique transcript in each library was normalized to the number of reads of MtlS , an abundant sRNA [16] that does not vary between the conditions tested here ( data not shown ) . A total of 14 , 578 unique sequences were identified between the two libraries , of which 13 , 309 were present in only one library or the other . Many sequences not shared between the libraries were very low in abundance and may represent products of random RNA degradation either in vivo or during preparation of the libraries . The positions of all reads aligned to the N16916 genome and their relative abundances in the two libraries is shown in ( Table S5 ) . The short sequencing reads were organized into clusters to provide an approximation of each putative sRNA sequence . Many of the 1 , 269 clusters shared between the libraries had large variations in abundance between the libraries . While this may reflect the true difference in the sRNA transcriptome between these two strains , to help us narrow the list of potential sRNAs we sought a method to determine which sRNAs were directly regulated by ToxT . Because sRNA promoters share many characteristics with open reading frame promoters , it seemed reasonable that any sRNA directly controlled by ToxT would have a ToxT binding site in cis . To investigate this we undertook a genome-wide ToxT pulldown of genomic DNA fragments 200–500 bp in length that were modified to allow for subsequent deep sequencing ( figure 1A and 1B ) , similar to an approach taken with the transcription factor CodY from Staphylococcus aureus [26] . Using a cut off of 3-fold enrichment in pulldown libraries over input libraries , we identified 199 putative binding sites of which 67 overlapped between technical replicates and likely represented the most specific sites ( table s6 ) . A DNA binding motif generated from the 67 enriched sites was a close , though not identical , match to the canonical toxbox [27] ( figure 1 panel C ) . Of the overall 199 putative binding sites , 64 mapped to the Vibrio Pathogenicity Island ( VPI ) , which is consistent with the fact that this locus contains the majority of ToxT-regulated genes . Most , but not all previously described ToxT binding sites were present in the pulldown library , notably absent are sites within the tcpA promoter [27] and sites within the MSH pilus operon [9] . Cross-referencing the putative ToxT binding sites with ToxT-regulated sRNA sequencing data yielded a collection of 18 potential sRNAs transcribed from intergenic regions with cis ToxT binding sites . The locations of these pulldown sites , sRNA transcripts and relative abundance between ToxT and ToxTΔHLH expressing strain libraries are shown in table 1 . This analysis revealed two putative sRNAs within intergenic regions in the VPI . To investigate whether these two sRNAs represented genuine transcripts , we probed for each by northern blot using total RNA from cultures expressing ToxT or ToxTΔHLH . Both of these sRNAs are dramatically upregulated upon expression of ToxT and both are present at the expected size predicted by the sRNA deep sequencing experiment ( figure 2 ) . One of these sRNAs was discovered independently by another group and was named TarA [25] ( for ToxT activated RNA A ) . The other , to the best of our knowledge , remains uncharacterized . Since it also showed dramatic up regulation upon expression of ToxT , and given its role in virulence ( described below ) , we named it TarB . Having now determined that at least two ToxT-regulated sRNAs were present in the VPI , we set out to determine whether they played detectable roles in the virulence of V . cholerae . Deletion of each sRNA was constructed in the genome and the mutants were competed against the fully virulent parental strain carrying a ΔlacZ marker . No significant difference in virulence was observed for the ΔtarA strain either when competed against the parental strain or a strain harboring tarA ( promoter and toxboxes included ) on a high-copy vector ( figure 3A ) . It was previously reported that a ΔtarA mutant had a decreased fitness relative to its parental strain [25] , however , those experiments were performed with a classical biotype strain of V . cholerae , and hence regulation by TarA may be less critical or perhaps is masked in the current pandemic El Tor biotype tested here . In contrast the ΔtarB strain outcompetes the parental strain by a small but statistically significant factor of 1 . 6 ( figure 3A ) suggesting TarB is a negative regulator of virulence . The ΔtarB and complemented strains show no change in growth rate or cell yield in Luria-Bertani ( LB ) broth or in a minimal medium , nor a change in survival in pond water ( figure S1 ) . To see if the negative effect on virulence could be complemented in trans , we competed a ΔtarB strain containing the sRNA with its own promoter cloned onto a low copy plasmid ( ptarB ) against a ΔtarB strain carrying empty vector ( pMMB ) . The ΔtarB strain out-competed the complemented strain to an extent that exceeds out competition of the parental strain ( figure 3A ) , which may be due to overexpression of TarB from ptarB . If expression of TarB is detrimental to colonization , as these data indicate , the plasmid carrying TarB may be selected against during the infection . To investigate this , small intestine homogenates were plated on LB agar and colonies were replica plated onto medium containing ampicillin , which selects for colonies containing the plasmid . Consistent with our hypothesis , the plasmid carrying TarB was lost more frequently than the empty plasmid ( figure 3B ) . This was not the case during growth in LB in the absence of antibiotic selection ( data not shown ) . For further confirmation of the hypercolonization phenotype of the ΔtarB mutant , we performed single strain infections with the ΔtarB and wildtype strains ( Figure 3C ) . Total colonization in these two strains indicated that , as seen in competition experiments , the ΔtarB mutant showed significant hypercolonization reflected by increased CFUs in the output . The out-competition phenotype of the ΔtarB strain in infant mice and more drastic attenuated phenotype of the complemented ΔtarB strain suggest that TarB is deleterious to colonization of the small intestine . The model that TarB is positively regulated by the master virulence gene activator ToxT , yet functions as a negative regulator of virulence , is counterintuitive . To investigate this model further we performed competitions after incubation of the competing strains for varying times in filter sterilized pond water in an attempt to test the strains in a scenario more similar to a natural infection . After 4 hours ( h ) of incubation in pond water , the ΔtarB mutant retained its ability to outcompete the parental strain , but this phenotype was lost after 6 h of incubation in the pond ( figure 3D ) . After 24 h of pond incubation , the parental now had a statistically significant advantage over the ΔtarB mutant when competed in vivo , but not when competed for in vitro growth in LB . To further investigate the ability of ToxT to control TarB expression , we measured expression of TarB under an in vitro virulence factor inducing condition , which is growth for 4 h static in AKI broth containing sodium bicarbonate followed by 4 h with aeration [28] . Expression of TarB is induced during the initial static phase of growth , but returns to background level after 4 h of growth with aeration ( figure 4A , top panel ) . The initial induction is dependent on toxT as well as toxR and tcpP/H ( figure 4A , bottom panel ) , which are genes upstream in the ToxR regulon that induce ToxT expression [3] , [4] , [29] . We also noted that TarB is overexpressed between 7–10 fold in the complemented strain , which is consistent with its in vivo phenotype being more dramatic then the parental strain in competitions with the ΔtarB mutant . We also investigated the role of the RNA chaperone Hfq in TarB stability and action as many sRNAs that act in conjunction with Hfq are destabilized in its absence [30] , [31] . To investigate expression from the TarB promoter we constructed a transcriptional fusion of a reduced half-life allele of GFP ( GFP-ASV ) [10] , [32] to the TarB promoter . The fusion was used to measure activity of the TarB promoter during induction of ToxT from the pToxT plasmid in both Hfq+ and Hfq− strains . In these same strains , steady state levels of TarB from a native copy of the gene were measured by northern blot . The results of these experiments are summarized in Figure S2 and indicate that Hfq likely does not play a role in stabilizing TarB or in its interaction with its target ( described later ) . Some basal expression of TarB is seen during culture in LB , which is greatly enhanced at the transition to stationary phase , however this increase is independent of ToxT ( figure 4B ) . Enhanced expression of TarB during late exponential and stationary phase growth in LB broth and in the static phase of AKI growth ( see above ) may be related to oxygen tension in solution . To investigate the contribution of oxygen tension during AKI static growth to TarB expression we measured expression of ToxT , TcpF and CadC by qRT-PCR and TarB via the TarB-GFP fusion over the static growth period of AKI . The transcription factor CadC is activated by the LysR homologue AphB under low oxygen and low pH conditions [33] , and its measurement is used here as a method of determining when the culture is undergoing those conditions . Additionally , AphB has been shown to be critical for activation of TcpP/H [34] , [35] , which in turn activates ToxT expression . The results of this experiment are summarized in figure S3 . As measured against expression after two hours of static growth , expression of ToxT and TcpF have more or less reached maximum by 3 hours of static culture ( Figure S3 A ) , though expression of TarB-GFP and CadC continue to rise , suggesting additional activation of the TarB and CadC promoters . Western blot for TcpF in the TcpF-FLAG fusion strain grown under the same conditions independently confirms this finding for TcpF ( Figure S3 A ) , though OmpU could not be used as a loading control for this blot [36] , so we did not carry out quantification of this blot . To investigate the contribution of anaerobiosis to expression of TarB , we prepared cultures of wildtype and ΔtoxT strains in phosphate buffered LB media to prevent large alterations in pH and with glucose to support anaerobic growth [37] . These cultures were prepared in an anaerobic chamber and then grown either aerated in 2 mL or in sealed 10 mL cultures to approximately the same optical density , RNA extracted from these cultures was used in northern blots for TarB ( Figure S3 B ) . The results show that anaerobic conditions do stimulate TarB expression independent of ToxT , though increases in expression of TarB in the wildtype culture were approximately twice as great when adjusting for loading , indicating that under anaerobic conditions , ToxT does drive some expression of TarB . Taken together these results suggest that anaerobic conditions activate TarB expression possibly through the action of AphB . The sequence upstream of the predicted TarB start site was investigated and revealed putative −10 and −35 sequences , as well as a direct repeat of putative toxboxes ( figure 5A ) . The 3′ end of TarB determined by deep sequencing corresponded to the poly-U tract of a Rho independent terminator . The toxboxes upstream of tarB are arranged in similar fashion to those upstream of the virulence gene tcpA [7] . To confirm binding of ToxT to this site , a DNA probe consisting of basepairs −100 to +1 relative to the predicted transcription start site was assayed for ToxT binding by gel shift assay . ToxT bound to this region with an affinity within the range of other reported toxboxes [38] , but not to a non-specific probe of similar length consisting of a PCR product of the 4 . 5S RNA sequence ( figure 5B ) . We next wanted to determine the target ( s ) of TarB that were responsible for the observed negative role of TarB in virulence . Nineteen putative mRNA targets were identified using the program targetRNA [39] , which searches for complementarity between the query sRNA and the 5′ untranslated region ( UTR ) of mRNAs of annotated ORFs within a given genome . To validate putative targets , we looked for changes in the steady-state level of the candidate mRNAs using quantitative reverse transcription PCR ( qRT-PCR ) on total RNA from TarB+ and ΔtarB strains both over-expressing ToxT . Of the six putative targets we selected for further analysis only two , tcpF and VC2506 , had any detectable expression under the conditions tested . When levels of the potential target transcripts were normalized to toxT transcript levels , a significant difference between the TarB+ and ΔtarB strains was revealed for the tcpF mRNA but not for VC2506 ( figure 6A ) . The observed increase in tcpF mRNA in the ΔtarB background suggests that TarB negatively regulates tcpF , which would be consistent with the negative role of TarB in virulence . To determine if TarB similarly affects TcpF protein expression level , we generated a C-terminal FLAG tag fusion to TcpF in the genome to measure expression by western blot after AKI induction . We also generated two sets of three point mutations each within the predicted region of complementarity between TarB and the 5′ UTR of tcpF , yielding tcpF* and tarB* alleles . These mutations are underlined in figure 6B . Because the tcpF and tcpE ORFs are very close together , there is some overlap between the coding sequence of tcpE and the 5′ UTR of tcpF , however the substitutions made do not affect the amino acid coding sequence of the upstream gene tcpE nor do they alter the Shine-Dalgarno sequence of tcpF . Moreover , the mutations were designed to preserve GC content of the region altered . Either set of mutations present alone ( tarB* or tcpF* ) would be predicted to disrupt the interaction between TarB and the tcpF 5′ UTR while the presence of both is compensatory and would be predicted to restore the interaction . A strain deleted for tarB was then used as the parent strain to construct derivatives having either the tcpF-FLAG or tcpF-FLAG allele . These two derivatives were then complemented with either ptarB , ptarB or empty vector ( pMMB ) . These six strains along with the wild type strain carrying the TcpF-FLAG fusion were grown through the static culture phase of an AKI induction and were western blotted to measure TcpF-FLAG expression . The blots were then stripped and probed for OmpU , which is not regulated by ToxT [40] , to serve as a loading control . Compared to the wild type strain ( figure 6C , first column ) the ΔtarB and ΔtarB tcpF* strains carrying the empty vector showed elevated TcpF levels . When the ΔtarB and ΔtarB tcpF* strains were complemented with ptarB* and ptarB , respectively , levels of TcpF remain largely unchanged , indicating that when either the tcpF mRNA or tarB sRNA are mutated , no interaction can take place and these strains show expression of TcpF similar to the ΔtarB mutant . However , when the ΔtarB and ΔtarB tcpF* strains were complemented with ptarB and ptarB* , respectively , to observe affects of the wild type or compensatory interaction when the sRNA is overexpressed , the levels of TcpF drop substantially . Six replicates of this experiment were performed and reveal that statistically significant drops in expression of TcpF occur only in strains containing either the wildtype TcpF target sequence complemented with wildtype TarB or strains in which the target sequence and sRNA have compensatory mutations ( Figure S4 ) . When these strains were blotted after the aeration growth phase of AKI induction , no differences in TcpF expression were visible ( data not shown ) , which would be expected given the up regulation of tarB during the static phase but return to basal level of expression during the aeration phase of AKI induction . To determine if the interaction of TarB with the 5′ UTR of tcpF was responsible for the phenotype in mice , competitions were carried out using tcpF* strain derivatives . Competition of the ΔtarB tcpF* ( ptarB* ) strain against the same strain carrying empty vector yielded the expected result of out-competition by the latter strain , which lacks tarB* ( figure 6D ) . Competition of the ΔtarB tcpF* ( ptarB ) strain against the same strain with vector alone yielded a competitive index that was significantly closer to one , which is expected since neither strain should have an interaction between sRNA and target . The difference between the two competitive indices was highly significant ( p<0 . 003 ) . To determine if the pond water-incubation phenotype of the ΔtarB mutant was related to expression of TcpF or TarB in this environment we carried out experiments to measure TarB and TcpF levels over the course of pond water incubation , the results of these experiments are summarized in Figure S5 . TcpF expression was followed through the course of pond water incubation via the C-terminal FLAG fusion in both the wildtype and ΔtarB backgrounds by anti-FLAG western blot . The results indicate that the wildtype and ΔtarB mutant show similar levels of TcpF expression initially , however , over the course of pond incubation , TcpF levels drop in the wildtype strain , but not the ΔtarB strain . Transcription of TarB , as measured by production of GFP from the TarB promoter-GFP fusion indicates that levels of TarB expression do not change dramatically over the course of pond water incubation . Northern blots for TarB expression over the course of pond water incubation suggest that TarB steady state levels drop ( data not shown ) , but this may be due to the observed wholesale degradation of RNAafter increasing time of incubation in pond water , such that accurate measurements of TarB expression via northern blot are not possible . These results indicate that while TarB expression levels do not vary dramatically over the course of pond water incubation , TcpF protein levels do drop , and this drop was absent in the ΔtarB mutant . This enhanced TcpF expression in the ΔtarB mutant may contribute to the phenotype of the ΔtarB mutant in vivo after pond water incubation , as over expression of TcpF in pond water would contribute to metabolic drain prior to infection . Deep sequencing has allowed the interrogation of processes in bacteria with unprecedented detail . Here we used two complementary approaches , deep sequencing of cloned sRNAs and ToxT-bound DNA fragments , to identify ToxT-regulated sRNAs . The number of previously estimated ToxT binding sites in the V . cholerae genome was between 17 and 20 [9] , [27] . We have now uncovered what may be a greatly expanded set of targets for ToxT to coordinate expression of protein coding genes as well as sRNAs . The results of the pulldown experiment returned regions of a few hundred basepairs in length that were enriched and many predicted sites are overlapping , which is due to the size range of the fragments used in the pulldown and the automated analysis of the pulldown data . Although many of these sites remain to be validated we are confident in proposing that the ToxR regulon encompasses many more transcripts , both protein coding and otherwise , than was previously thought . The results of the sRNA deep sequencing reveal the method to be exquisitely sensitive . Because of our exclusion of larger RNA transcripts and depletion of tRNA and 5S RNA in the sRNA size range and the use of Illumina massively parallel sequencing technology we have achieved tremendous depth of coverage of potential sRNA genes in V . cholerae [16] . Transcripts represented by ∼40 or more reads could be detected by northern blot ( this study and data not shown ) . However , transcripts represented by fewer than ∼40 reads , which may represent low abundance sRNAs , are difficult or impossible to detect by northern blot and other methods such as qRT-PCR are needed for independent validation . Of the 18 candidate ToxT-regulated sRNAs we report here , 11 ( including tarB ) were not identified as putative sRNAs in previous sequencing experiments or bioinformatics-based approaches to sRNA discovery [16] , [41] , displaying the depth of information that can be gained with high throughput sequencing technologies and the conditional expression of sRNAs . In comparison to other methods of sRNA discovery , our approach has the advantage of being targeted in its search for ToxT-regulated sRNAs but unbiased in its identification of sRNAs . Approaches utilizing RNA binding proteins such as Hfq [42] , [43] , are not exhaustive as the sRNA we report here likely does not interact with Hfq , though those methods do have the potential to identify mRNA targets as well as sRNAs . Additionally , this approach benefits from the vast strides made in high throughput sequencing recently which generates far more depth of data then microarray based methods [44] , including exact 3′ and 5′ ends and unbiased coverage of positive and negative strand sRNAs . Keeping the latter in mind , this approach can also identify many potential sense and anti-sense sRNAs [16] overlapping with protein coding genes although these potential sRNAs are not discussed here . In this study we identified a new sRNA member of the ToxR regulon that fine-tunes expression of a virulence factor also within the ToxR regulon , thus adding a new facet to the elaborate virulence gene regulation program in V . cholerae . However , when placed in the larger context of V . cholerae pathogenesis , it is not entirely clear why a repressor of an essential virulence factor would be produced at the same time as the virulence factor it negatively regulates . The answer may lie in the biphasic nature of V . cholerae gene expression during intestinal colonization [10] , [11] . The initial induction of virulence factors requires ToxR/S- and TcpP/H-dependent ToxT expression in the intestinal lumen . This is followed by a more robust activation of the TCP and CTX operons closer to the epithelial surface of the small intestine , driven by a positive feedback loop in ToxT expression that is thought to activated in part by the presence of bicarbonate [28] , [45] . During AKI induction in vitro in the absence of bicarbonate , ToxT production is stimulated during static growth but the transition to aerated growth is required for CTX production [46] . All experiments reported here included bicarbonate in the medium over the course of the experiment , which is sufficient to cause CTX production even during static growth [28] , [47] . Research done on the contribution of anaerobiosis to virulence gene expression in V . cholerae El Tor isolates has shown stimulation of VPI gene products [48] , and that the AphB protein , which functions upstream of tcpP/H , is active primarily at low oxygen tension and low pH [33] . Since TarB expression is greatest during the static phase of AKI induction , but repressed during aerated growth even though bicarbonate had been added to induce CTX and TCP expression prior to aeration , it is tempting to speculate that TarB expression is enhanced in microaerobic conditions . The experiments we performed under anaerobic conditions also suggest that oxygen plays a role in TarB expression , though it may be only one of a host of signals , which act on TarB in vivo . TarB's function under low oxygen tension could be to repress TcpF expression prior to penetration of the mucous barrier of the small intestine . Upon reaching the epithelial surface , the higher oxygen tension would contribute to reduced TarB expression , allowing TcpF to be fully expressed . This would fit with the proposed role of TcpF in colonization of the epithelium [49] . The intestinal brush border is a highly vascular structure , commensurate with its role in absorbing nutrients , and it would not be unreasonable to speculate that the lumenal space adjacent to it would have greater oxygen tension then the luminal fluid . The actual oxygen tension of the small intestine may be quite low as oxygen requiring luciferase reporter systems in bacteria do not function in the small intestine [50] , [51] . However , to the best of our knowledge , oxygen measurements at the brush border have not been reported . Other possible factors responsible for controlling TarB expression could be entry into stationary phase , as increased TarB expression is observed in V . cholerae grown in LB broth to late exponential and stationary phase . Also , during AKI induction , 4 hours of growth in static culture corresponds with entry into stationary phase [46] . Stationary phase regulation of TarB may also occur via an alternative sigma factor as was observed for the sRNA VrrA [20] , or possibly via CRP-cAMP mediated repression as carbon sources become depleted [35] . Coordination of TcpF expression by TarB appears to have a positive effect on colonization if the bacteria are coming from a resource poor environment , such as contaminated pond water , and even then , the differences in colonization efficiency of the ΔtarB mutant are quite small . In contrast , if the bacteria are grown in a rich medium prior to infection , overexpression of TcpF in the ΔtarB mutant appears to be beneficial . The reasons for this may relate to the details of the experimental system used here , wherein immunologically naïve infant mice are used as a host . In contrast , in nature many hosts in endemic areas will have some level of pre-existing immunity , and may harbor anti-TcpF antibodies as TcpF is a known antigenic protein [49] . It is possible that tight repression of TcpF provides a more pronounced fitness advantage in nature under different conditions then those used here , which would explain TarB's presence among all sequenced isolates of toxigenic V . cholerae ( data not shown ) . Further insight into the functional role of TcpF in colonization may shed more light on the necessity of the TarB-mediated post-transcriptional regulation observed here . All animal experiments were done in accordance with NIH guidelines , the Animal Welfare Act and US federal law . The experimental protocol using animals was approved by Tuft University School of Medicine's Institutional Animal Care and Use Committee . All animals were housed in a centralized and AAALAC-accredited research animal facility that is fully staffed with trained husbandry , technical , and veterinary personnel . V . cholerae O1 serogroup El Tor biotype isolate E7946 and derivatives were grown at 37°C in LB broth with aeration . For AKI induction , strains were grown in AKI broth ( 1 . 5% peptone , 0 . 4% yeast extract , 0 . 5% NaCl , 0 . 3% NaCHO3 ) statically for 4 h at 37°C followed by aeration for 4 h 37°C . To induce expression of cloned genes on plasmids , arabinose was added to 0 . 04% upon reaching mid-exponential phase ( optical density at 600 nm [OD] = 0 . 3 ) . All DNA manipulations were done in E . coli DH5α or derivatives with plasmids maintained with the appropriate antibiotics . All PCR reactions were carried out with EasyA polymerase according to the manufacturer's specifications using the indicated primers , the sequences of which can be found in table S1 . The descriptions of all plasmids used in this study are included in table S2 . Plasmids pToxT and pToxT ΔHLH plasmids we constructed by PCR amplification of the toxT ORF including native RBS from gDNA from either wildtype V . cholera E6749 or an E6749 strain carrying an internal deletion of the helix-loop-helix DNA binding domain [11] using primers NcoI_ToxT_F and XbaI_ToxT_R . This PCR product was then cloned into the NcoI and XbaI sites of the pBAD24 plasmid [52] to allow expression of ToxT upon addition of L-arabinose . Unmarked deletions of chromosomal genes were constructed by SOE PCR introduced using a derivative of the pCVD442 allelic exchange vector , pCVD442-lac which contains the pUC19 LacZ gene and MCS , as described [53] . Point mutations in the tarB gene were generated by SOE PCR using primers xbaI_TarB comp_F , TarB_mut_R1 and TarB_mut_F2 and SacI_TarB_comp_R , using E6749 genomic DNA as template . PCR products were mixed in a one to one ratio , and added to a PCR reaction run for 25 cycles at an annealing temperature of 50°C without primers and the mutated sRNA sequence plus promoter were amplified with XbaI_TarB_comp_F and SacI_TarB_comp_R which contain SacI and XbaI restriction sites which were subsequently used for cloning into pMMB67EH to generate ptarB* . The wildtype complementation vector ptarB was generated by cloning a PCR product generated using XbaI_TarB_comp_F and SacI_TarB_comp_R primers and genomic DNA as a template . Point mutations in the tcpF 5′ UTR were also generated by SOE PCR using primers XbaI_TcpF_mut_F1 , TcpF_mut_R1 , TcpF_mut_R2 and XbaI_TcpF_mut_R2 using an identical procedure as above . The final ∼2 kb product containing the mutated tcpF 5′ UTR sequence which was subsequently cloned into the XbaI site of the pCVD442-lac vector which was then mated into strains of interest . Double crossovers were selected on 10% sucrose plates . Individual double crossovers were screened for the mutated sequences by sequencing with the TcpF seq primer and the XbaI_TarB_comp_F primer and confirming double crossover by streaking on 10% sucrose as well as ampicillin containing plates to ensure sucrose resistance and ampicillin sensitivity . C-terminal FLAG fusions to TcpF were generated by amplification of the C-terminal 346 bp using the TcpF_qt_F primer and the TcpF-FLAG_R primer to add the FLAG amino acid sequence [54] , this product was subcloned into Topo pCR2 . 1 ( Invitrogen ) . The resulting plasmid was cut using KpnI and EcoRV and the insert containing the C terminus of TcpF with the FLAG fusion was cloned into a modified pGP704 suicide vector [55] which contains a chloramphenicol resistance drug marker in place of an ampicillin marker ( pGP704cat ) . This construct was then mated into strains of interest and single crossovers were selected for on chloramphenicol plates at 2 µg/mL . Proper insertions were confirmed by PCR using the TcpF-FLAG reverse primer and TcpF seq forward primer . A merodiploid strain was constructed by plasmid integration resulting in the placement of GFP ( ASV ) under the control of one copy of the TarB promoter followed by the native TarB locus downstream of the integrated plasmid sequence . The plasmid borne fusion was generated by amplifying the +3 to −376 positions in the TarB promoter from E6749 genomic DNA using primers TarB_F and TarB_-300_R and subcloning the product into pCR2 . 1 yielding ptarB-300 . GFP was amplified from pGfpmut3 . 1 plasmid ( Clonetech ) using primers Fgfp2 and Rgfp2 which adds a ribosomal binding site and SacI site at the 5′ end and the destabilizing ( ASV ) [32]C terminal amino acids and a SmaI site at the 3′ end . The GFP ( ASV ) PCR product was cloned in a triple ligation with the SacI/EcoRV fragment from ptarB-300 into pGP704cat digested with SmaI to generate the transcriptional fusion . The resulting plasmid ( pTarB-GFP ) was mated into E6749 strains and single crossovers were selected on chloramphenicol and confirmed by PCR using primers Rgfp2 and XbaI_ΔTarB_R2 . Single colonies of strain AC3763 ( ΔtoxT ) transformed with either pToxT or pToxTΔHLH plasmids were picked and grown in LB broth containing streptomycin and ampicillin both at 100 µg/mL overnight . Strains were back diluted from overnight cultures to an OD of 0 . 03 in 200 mL LB supplemented with streptomycin and ampicillin both at 100 µg/mL and were grown with aeration at 37°C until the strains reached mid-exponential phase ( OD = 0 . 3 ) . Arabinose was then added to 0 . 04% to induce expression of toxT alleles from pToxT plasmids , and induction was allowed to proceed for 20 minutes prior to RNA extraction . Total RNA was purified from the bulk culture by phenol/chloroform extraction and isopropanol precipitation . Cloning and sequencing of sRNA was carried out as previously described [16] , sequences of the micro RNA cloning linkers ( IDT ) used are included in table S4 . In order to further decrease tRNA and 5S rRNA in the final sequenced products , the depletion step described in the previously published procedure was carried out twice with the addition of an oligo targeting the serGCC tRNA ( 5′-GCGGTGAGTGAGAGATTCGAACTCTC-3′ ) . The final cDNA products were prepared for Illumina Genome Analyzer II sequencing using Illumina primers 1a , 1b and 1c ( table S1 ) for the first 10 cycles of PCR , followed by gel purification and Illumina primers 2a and 2b ( table S1 ) for the final 4 cycles of PCR followed by PCR clean up ( Stratagene ) . Final products were run on a Bioanalyzer High-Sensitivity DNA chip ( Agilent ) prior to Illumina sequencing to normalize loading of the two samples and ensure quality of the libraries . The libraries were pooled and placed on one lane of an Illumina Genome Analyzer IIx paired-end sequencing run at Tufts University Core Facility . Briefly , a paired-end sequencing run sequences both the 5′ and 3′ end of every DNA molecule attached to the flowcell . The first read is downstream of linker 1 and the second read is downstream of linker 2 ( ToxT library ) or linker 3 ( ΔHLH library ) so that for every pair , the directionality of the original RNA molecule could be determined . Sequence reads were trimmed to remove linker sequences and filtered so that 100% of the sequenced bases in each read had a minimum quality score of 5 ( base call accuracy at least 68% ) . Reads were aligned to the O1 biovar N16961 genome ( NCBI Accession Nos . NC_002505 , NC_002506 ) using Bowtie ( http://bowtie-bio . sourceforge . net ) . Reads matching rRNA or tRNA regions were removed from the alignment , leaving 1 , 062 , 048 reads in the ToxTΔHLH library and 2 , 212 , 216 reads in the ToxT library . Unique transcripts totaled 6 , 815 for ToxTΔHLH and 27 , 787 for ToxT . The alignments were then processed to generate a library of clustered transcripts using the method previously described [16] . This resulted in 3 , 309 clusters for the ToxTΔHLH library and 12 , 534 clusters for ToxT library . Clustered reads were output into “gff” format and viewed using GenomeView ( http://genomeview . org ) . The number of reads in sRNA clusters were normalized by dividing the number of reads in that cluster by the ratio of MtlS reads in that library to total MtlS reads . For example normalized readsToxT = cluster readsToxT/ ( MtlSToxT/ ( MtlSToxT+MtlSToxTΔHLH ) ) . E . coli strain BL21 ( DE3 ) was transformed with the plasmid pMAL-TEV-His-thr-ToxT ( table s3 ) . The resulting strain was grown on LB agar plates containing ampicillin and a single colony was picked for growth of a 4 mL overnight culture . The overnight culture was used to inoculate 1 L LB broth containing ampicillin at 100 µg/mL and was grown with aeration at 37°C . Transcription was induced once the culture had reached exponential phase ( OD = 0 . 5–1 ) by addition of IPTG to 1 mM . Induction was allowed to proceed shaking at 20°C for 16 h , after which , cell pellets were collected by centrifugation and resuspended in 20 mL lysis buffer ( 20 mM Tris-HCl pH 8 , 2 mM DTT , 1 mM EDTA , 250 mM NaCl ) plus Complete protease inhibitors ( Roche ) . Cell pellets were lysed and the lysate was cleared by centrifugation at 18 , 000 rpm in a SS34 rotor . The cleared lysate was then applied to a 5 mL dextrin MBPtrap column ( GE Life sciences ) . The column was washed with lysis buffer followed by elution with MBP elution buffer ( as lysis buffer , +1 mM maltose ) . The elution fractions were subsequently diluted 10-fold with buffer QB1A ( 20 mM Tris-HCl pH 8 . 0 , 1 mM DTT ) and applied to an 8 mL Source15Q anion exchange column equilibrated in QB1A . The protein was eluted using a 0 to 20% gradient of QB1B ( 20 mM Tris-HCl pH 8 . 0 , 1 M NaCl , 1 mM DTT ) developed over 25 column volumes . The peak fractions were diluted 5-fold in SB1A buffer ( 25 mM phosphate buffer pH 6 . 0 , 1 mM DTT ) and applied to a 8 mL Source15S cation exchange column equilibrated in SB1A . The protein was eluted using a 15 to 35% gradient QB1B ( 25 mM phosphate buffer pH 6 . 0 , 1 mM DTT , 1M NaCl ) , which resulted in two peaks , the second peak was known to be a soluble aggregate and was discarded . The initial peak was split into two aliquots , one of which was applied to a Superose 12 gel filtration column in EMSA buffer ( 10 mM Tris-HCl pH 7 . 5 , 200 mM KCl , 10 mM βME ) for use in mobility shift assays , the other aliquot was cleaved with TEV protease overnight at 4°C and subsequently diluted 5-fold in SB1A and applied to a 2 mL Source15S cation exchange column to separate His-ToxT from the cleaved MBP fusion protein . His-ToxT was eluted from this column with a 35 to 100% gradient of SB1B developed over 12 column volumes . Finally , His-ToxT peak fractions were applied to a Superdex 75 gel filtration column in EMSA buffer . These final steps did leave a small amount of TEV protease in the final purified product . Genomic libraries were prepared by centrifuging 10 mL of overnight growth of wild type ( AC53 ) V . cholerae , washing 2× with TBS and resuspending in 5 mL TBS . To generate gDNA fragment sizes of 300 to 1 , 000 bp , the cell pellet was subjected to four 30 second sonication cycles on ice using a sonicator micro tip ( Branson ) ; each sonication cycle was separated by a 30 second incubation on ice . After sonication , RNAase A was added to a concentration of 2 µg/mL , the samples were incubated at 37°C for 20 min to allow for degradation of RNA . DNA was purified with 2 rounds of extraction with citrate buffered phenol∶chloroform ( Ambion ) followed by a final extraction with chloroform only and then concentrated by ethanol precipitation . Fragmented DNA was used to prepare three different bar-coded libraries using adapters BC1a/BC1b , BC2a/BC2b and BC3a/BC3b ( table S4 ) as described [26] . For the final amplification and purification of bar-coded libraries , ten PCR reactions were done using linkered and size selected gDNA as template using primers Olj 139 and 140 and EasyA polymerase ( Stratagene ) . PCR conditions were as follows , denaturation for 5 minutes at 95°C , annealing for 30 seconds at 65°C , elongation for 30 seconds at 72°C , cycling back to denaturation at 95°C for 30 seconds for 15 cycles after which reactions were pooled and incubated with 50 µL ExoSAP-IT ( USB ) at 37°C for 1 h . Final purification of libraries was carried out by phenol∶chloroform extraction and ethanol precipitation and resuspension of libraries in 100 µL deionized water . Binding reactions contained 15 µg bar-coded DNA library in a total volume of 250 µL with 200 nM purified His6-tagged ToxT purified as above or with His6-tagged TEV protease in EMSA buffer with 10 µg/mL sheared salmon sperm DNA , 0 . 3 mg/mL BSA and 10% glycerol . Reactions were allowed to incubate with gentle mixing at 37°C for 1 h , after which the reaction was added to a microcentrifuge spin column ( Pierce ) packed with a 50 µL bead volume column of HisPur cobalt resin ( Pierce ) that had been equilibrated in the above buffer . The reaction was allowed to bind to the column by mixing gently at 37°C for 1 h . Flow through was then collected by spinning the column in a microcentrifuge at 3 , 000× g for 1 minute . The column was washed 3× by gentle resuspension of the bead volume in 250 µL of EMSA buffer with the above additions , followed by centrifugation . The column was washed an additional 3× as above , but in EMSA buffer only . After the final wash , the bead volume was resuspended in 10 mM Tris-HCl pH 8 and boiled for 5 minutes and allowed to cool to room temperature , then incubated with proteinase K ( 5 µg/ml ) for 30 minutes at 65°C , followed by boiling for 5 minutes . After centrifugation for 1 min at 3 , 000× g , the resulting 100 µl of the supernatant fluid was purified by using a PCR purification kit ( Qiagen ) and then subjected to 10 cycles of PCR amplification with primers Olj139 and Olj140 , repurified , quantified on the Bioanalyzer high sensitivity DNA chip ( Agilent ) , and subjected to deep sequencing , along with aliquots of the input libraries prior to pulldown , using the Illumina Genome Analyzer II on the paired end setting . Reads from the Illumina libraries were aligned to the N16961 genome . Sequence alignment and assembly were performed as described above . After alignment , reads that did not match the genome were discarded and the sets were normalized so that each set contained the same number of reads . Alignment positions were shifted by half their insert length as determined by each mapped pair , giving the center position of each sequenced DNA molecule . These positions were then tabulated and used to generate a coverage map of the genome using a rolling average with window size of 35 bases . Coverage maps were generated for every sample . For each genomic DNA and corresponding pulldown sample , an enrichment map was created , which represented the ratio of the values from the pulldown sample over that of the genomic DNA sample . Enrichment maps were then scanned to identify regions that had more than 3× the average coverage for more than 100 consecutive positions . The false discovery rate ( FDR ) was then calculated by performing the same analysis with the control and pulldown samples switched . At 3× coverage , the FDR was 0 . 03 and 199 enriched sites were identified totaled between the libraries , of which 67 were observed in both replicates . Significance of each enriched region was assessed using two methods [56] . First , the number of reads in that region in the control sample was used to generate a Poisson distribution . This was then used to assess the probability of the same number of reads occurring in the pulldown sample . Using this method , all regions identified had a p-value of <1×10−98 . Second , a Z-score was found by comparing the proportion of tags in the control sample to that in the pulldown . All of the regions identified had a significant difference in the proportion of tags counted between the control and pulldown samples , with z-scores >7 . 7 . The nucleotide sequences from the overlapping set were used as a training set for finding motifs using MEME 4 . 1 . 0 . We allowed MEME to find motifs that occurred at least one time in each fragment . The motif reported in figure 1 panel C is the lowest E-value motif for the 67 sites combined in both libraries . Primers TarB promoter R and TarB promoter F were used to amplify the upstream 100 bp of TarB , predicted to contain promoter elements and ToxT binding sites to serve as a probe in the mobility shift assay . The PCR product was purified ( Stratagene ) and 3 . 3 pmoles was end-labeled using T4 Polynucleotide Kinase ( NEB ) and 32P γ-ATP according to the manufactures instructions , and then purified using a Performa DTR spin cartridge ( Edge Biosciences ) . A negative control probe of similar size consisting of 4 . 5S RNA sequence was prepared in parallel . The binding reaction occurred in 20 µL with 3 nM labeled probe and varying concentrations of purified MBP-his-thr-ToxT in EMSA with 10 mM 10 µg/mL sheared salmon sperm DNA , 6 µg/mL BSA , 10% glycerol and 0 . 002% Orange G dye added . Binding was allowed to occur for 30 minutes at 30°C followed by loading of the entire reaction onto a 5% TBE-Polyacrylamide gel , which was then run at 100 V for 60 minutes . The gel was then used to directly expose a phosphor screen and the image was read on a FLA-9000IR using the IP setting . For AKI induction experiments , strains were grown overnight with aeration at 37°C in LB broth containing streptomycin at 100 µg/mL , ampicillin at 50 µg/mL ( excluded in the case of the TcpF C-FLAG integration in the wild type background and TarB-GFP strains without plasmid ) and chloramphenicol at 2 µg/mL . Overnight cultures were then diluted into prewarmed AKI media [47] containing 0 . 3% NaHCO3 and ampicillin at 50 µg/mL ( again excluded for the wild type background strain and TarB-GFP fusions ) to an OD of 0 . 01 . Strains were grown statically in an incubator at 37C for the indicated times at which culture aliquots were removed for analysis . After 4 hours of static growth , cultures were split into 1 mL aliquots and grown shaking at 37C for 4 hours . For anaerobic growth experiments , overnight cultures were prepared by inoculation of strains into phosphate buffered LB media containing 60 mM K2HPO4 , 33 mM KH2PO4 , 0 . 5% glucose and 100 µg/mL streptomycin . These cultures were grown overnight in an anaerobic chamber and used to subsequently inoculate either 2 mL aerated cultures or 10 mL cultures in sealed tubes prepared in the anaerobic chamber to an OD of approximately 0 . 01 . Aerobic and anaerobic cultures were then grown in parallel in a shaking 37C incubator to approximately the same OD and snap frozen on liquid nitrogen and subsequently used for RNA extraction and northern blots . For each culture the pH of the media was measured after growth was recorded and ranged between 6 . 3 and 6 . 5 for anaerobic cultures and 6 . 7 to 6 . 8 for aerobically grown cultures . For pond water incubation experiments , strains were grown overnight on M9 minimal media+glucose plates containing the proper antibiotics . Overnight growth was resuspended in saline and washed twice . After the final wash , strains were resuspended in filter-sterilized pond water and inoculated into 2 mL culture tubes of filter sterilized pond water to an OD of 0 . 1 and incubated shaking at 37°C for the indicated times . At those times , culture aliquots were prepared either for western blot by centrifugation followed by resuspension in sample buffer and boiling or diluted to a density of 1×103/µL as measured by OD for mouse infections . Experiments involving induction of ToxT from the arabinose inducible plasmid were carried out similarly to those used in sRNA sequencing experiments . Overnight cultures of the indicated strains were grown at 37°C overnight in LB containing the appropriate antibiotics . Overnight cultures were then diluted to an OD of 0 . 03 in 25 mL of the same media and allowed to grow shaking at 37°C . Once cultures reached mid-exponential phase ( OD = 0 . 3 ) , arabinose was added to a final concentration of 0 . 04% and induction was allowed to proceed for 1 h with 2 mL aliquots of culture taken at the indicated times and either spun down for western blot analysis or snap frozen in liquid nitrogen for RNA extraction later . Between 2 . 5–10 µg of total RNA purified using the Ambion mirVana kit from the indicated cultures was run on 10% TBE-urea polyacrylamide gels . Prior to transfer , gels were stained with GelStar ( Invitrogen ) and scanned on the FLA-9000IR ( Fuji ) to assess total RNA loading in each well and to use for normalization during quantification . RNA was transferred to Hybond N+ membranes ( Amersham ) in 1× TBE using the Mini Trans-Blot Cell apparatus ( Bio-Rad ) according to the manufacturer's instructions . Blots were prehybridized in Ultrahyb ( Ambion ) prior to addition of probe . RNA probes were transcribed from PCR-derived templates with T7 promoters using 32P-UTP and T7 polymerase ( Promega ) according to the manufacturer's instructions . Ambion Decade ladder labeled with 32P-ATP was run alongside RNA samples to provide estimations for the sizes of RNA bands . Hybridzation was carried out at 65°C overnight followed by washing 3× with low stringency buffer ( 2× SSC+0 . 05% SDS ) wash at room temp , followed by washing 3× with high stringency buffer ( 0 . 2×SSC+0 . 05% SDS ) at 65°C . Blots were then exposed to phosphor storage screens ( Fuji ) overnight . The image was subsequently read on a FLA-9000IR scanner . When reporting quantification , measurements taken from the phosphor screen after exposure were divided by fluorescent measurements of the 5S rRNA taken prior to transfer to normalize signal for loading using the MultiGage software ( Fuji ) . Total RNA was purified from cultures grown under the indicated conditions using the mirVana RNA purification kit . Total RNA was treated with DNAase with the TURBO-DNAfree kit ( Ambion ) prior to reverse transcription . cDNA used as template was generated using iScript complete kit ( BioRad ) from 2 µg of total RNA using random hexamers . Quantitative PCR was run using Strategene Mv3005P equipment and MxPro qPCR software . Each sample was measured in technical triplicate . In all cases , controls lacking reverse transcriptase were included to assess DNA contamination , all results were either below the baseline of detection , or were subtracted from values obtained with those templates . For western blot analysis of TcpF and GFP expression , strains carrying the TcpF C-terminal FLAG allele or the TarB-GFP fusion were grown under the indicated conditions at which times 2 mL culture aliquots were removed . Culture aliquots were immediately centrifuged at 10 , 000× g for 5 minutes to pellet cells , and supernatants were removed . Cell pellets were boiled in 50 µL ( static timepoints and plasmid induction experiments ) or 100 µL ( 4 h aeration timepoint ) of SDS loading buffer ( 50 mM Tris-HCl , pH 6 . 8 , 2% SDS , 0 . 5% bromophenol blue , 10% glycerol , 100 mM βME ) . Samples were cooled and a volume adjusted for differences in OD was loaded on an SDS-polyacrylamide gel electrophoresis ( PAGE ) gel and run 90 minutes at 125 V . Proteins were transferred to a nitrocellulose membrane at 25 V for 1 h . Membranes were loaded onto the SNAP-ID Western blotting system ( Millipore ) and blocked with 1× NAP blocking agent ( G Biosciences ) diluted in PBS+0 . 01% Tween-20 . Primary antibody to the FLAG peptide ( Invitrogen ) or against GFP ( Abcam ) was added to the membrane 1∶600 or 1∶1200 respectively , diluted in 3 mL 1× NAP block for 10 minutes and the membrane was washed with 90 mL PBS+0 . 01% Tween-20 . Secondary antibody ( Invitrogen ) ( Cy5 conjugated goat anti mouse for anti-FLAG blots or Cy5 conjugated goat anti rabbit for GFP blots , ) was added to the membrane at 1∶600 and diluted in 3 mL 1× NAP block for 10 minutes and the membrane was washed with 90 mL PBS+0 . 01% Tween-20 . Bands were visualized using the Cy5 setting the FLA-9000IR . After visualization of TcpF-FLAG , blots were stripped by incubating in 20 mL acid stripping buffer ( 25 mM glycine pH 2 , 1% SDS ) shaking for 30 minutes followed by washing 2× with 20 mL PBS+0 . 01% Tween-20 . After stripping , blots were reprobed as above with primary anti-OmpU at 1∶600 in 1× NAP block and secondary Cy5 conjugated goat anti-rabbit ( Invitrogen ) again in 1× NAP block and scanned on Cy5 setting on the FLA-9000IR . Fluorescence measurements were quantified using MultiGage software ( Fuji ) . Measurements TcpF-FLAG bands , adjusted for area and background , were divided by fluorescence measurements of corresponding OmpU bands adjusted for area and background . Loading-adjusted fluorescence values were then standardized to wild type expression and reported as fold expression of TcpF relative to wild type expression . The experiment shown is representative of six biological replicates . Single strain infections and competition assays in infant mice , LB broth and filter sterilized pond water were performed with the TarB unmarked deletion strain ( AC3744 ) ( LacZ+ ) and wild type with a lacZ deletion ( AC3745 ) for 24 h as described [57] . Inputs for competition assays and single strain infections were prepared by growth overnight on LB plates containing the appropriate antibiotics followed by resuspension in LB to an approximate density of 1×103/µL as measured by OD , mixing of equal volumes of either culture ( for competition experiments ) then inoculation of infant mice by oral gavage . Samples from pond water incubations were prepared as described above , mixed in equal volumes and then used for innocualtion of infant mice . Immediately after inoculation , input ratios and total CFU were determined by plating on LB plates containing 5-bromo-4-chloro-3-indolyl-D-galactopyranoside ( X-gal ) . The target input dose for all experiments was 105 bacteria/mouse , although over the course of the experiments doses ranged between 104 and 106 . Results are shown by the competition index ( CI ) , which is the ratio of mutant CFU to wild type CFU normalized for the input ratio . To show complementation in trans in all assays in this study , ΔtarB derivatives ( LacZ+ ) were complemented with either ptarB or ptarB* and were competed against the respective isogenic strain ( LacZ− ) carrying the pMMB67EH plasmid alone . CIs for these experiments are expressed as the ratio of mutant to complemented CFU corrected for input . To assess plasmid loss frequency , output plates were replica plated onto LB agar plates containing streptomycin and ampicillin at 100 µg/mL and X-Gal at 40 µg/mL to determine plasmid containing CFUs , and LB agar plates containing streptomycin and X-Gal to determine total CFUs . Growth of strains was determined by measured OD using a Bio-Tek microplate reader . Cultures grown overnight in LB plus streptomycin and ( ampicillin at 50 µg/mL for complemented strains ) or M9 glucose plus streptomycin and ( ampicillin at 50 µg/mL for complemented strains ) were resuspended to an OD of 0 . 01 in the respective media and pipetted into a 96-well plate in triplicate . Each growth curve was performed in biological triplicate . Bacteria were grown with aeration for 17 h at 37°C in the microplate reader with the OD being read every 17 minutes .	Vibrio cholerae is the causative agent of the diarrheal disease cholera , which remains a significant public health issue in Africa , South Asia and recently Haiti . To better understand virulence gene regulation in V . cholerae we sought to investigate the contribution of small non-coding regulatory RNAs ( sRNAs ) to regulation of virulence in V . cholerae . We undertook a genome wide approach to sRNA discovery combining direct sequencing of sRNA transcript cDNA and genome wide binding studies of the master protein regulator of virulence , ToxT . This approach yielded one previously known and 17 new potential sRNAs under the control of ToxT . We investigated one of these new sRNAs and showed that it negatively regulates expression of the secreted colonization factor TcpF , adding a new facet to the complex gene regulatory network necessary for virulence in V . cholerae .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "systems", "biology", "gram", "negative", "microbial", "pathogens", "genetics", "and", "genomics", "biology", "microbiology", "bacterial", "pathogens", "pathogenesis" ]	2011	A Genome-Wide Approach to Discovery of Small RNAs Involved in Regulation of Virulence in Vibrio cholerae
Dengue virus NS5 protein plays multiple functions in the cytoplasm of infected cells , enabling viral RNA replication and counteracting host antiviral responses . Here , we demonstrate a novel function of NS5 in the nucleus where it interferes with cellular splicing . Using global proteomic analysis of infected cells together with functional studies , we found that NS5 binds spliceosome complexes and modulates endogenous splicing as well as minigene-derived alternative splicing patterns . In particular , we show that NS5 alone , or in the context of viral infection , interacts with core components of the U5 snRNP particle , CD2BP2 and DDX23 , alters the inclusion/exclusion ratio of alternative splicing events , and changes mRNA isoform abundance of known antiviral factors . Interestingly , a genome wide transcriptome analysis , using recently developed bioinformatics tools , revealed an increase of intron retention upon dengue virus infection , and viral replication was improved by silencing specific U5 components . Different mechanistic studies indicate that binding of NS5 to the spliceosome reduces the efficiency of pre-mRNA processing , independently of NS5 enzymatic activities . We propose that NS5 binding to U5 snRNP proteins hijacks the splicing machinery resulting in a less restrictive environment for viral replication . Dengue virus ( DENV ) is currently the most important human viral pathogen transmitted by insects . It is responsible for about 390 million infections worldwide every year [1] . In spite of this great burden , vaccines and specific antivirals remain elusive . In fact , a steady increase in the number of infections is being registered in the last years ( http://apps . who . int/iris/bitstream/10665/75303/1/9789241504034_eng . pdf ? ua=1 ) . DENV belongs to the Flavivirus genus in the Flaviviridae family , together with a large number of emerging and re-emerging human pathogens that cause fevers and encephalitis , such as West Nile virus , Japanese encephalitis virus and Zika virus [2 , 3] . Like in other RNA viruses , the DENV genome encodes a limited set of proteins , but relies on the host machinery for productive replication . During infection , viral components subvert cellular processes , remodeling intracellular membranes , changing host metabolic routes and blocking innate antiviral responses [4 , 5] . These changes in the cellular environment are the result of an intimate host-virus interaction and co-evolution . Although in the last decade a great deal has been learned about the DENV biology , the intricate network of viral-host interactions that provide the appropriate setting for viral replication is largely unknown . Global mapping of protein-protein interactions through systematic overexpression of viral proteins and proteomic studies have recently identified complete cellular pathways harnessed by HIV and HCV infection [6–8] . Moreover , the generation of recombinant viruses able to replicate expressing tagged viral proteins allowed identification of protein complexes in the context of measles , influenza and HIV infections [9–11] . A technical limitation for this kind of proteomic approach is the feasibility to design recombinant viruses that tolerate the addition of tags . Here , we developed a tool for proteomic studies by incorporating purification tags in fully functioning DENV , and generated affinity purification-mass spectrometry data from infected human cells focusing on the viral protein NS5 . NS5 is the largest viral protein , bearing multiple enzymatic activities and functions during infection . It bears the RNA dependent RNA polymerase and methyl transferase activities , which are fundamental for viral genome amplification [12–18] . In addition , NS5 interacts with different host proteins to counteract the IFN-α mediated antiviral response through STAT2 degradation [19–21] . Although viral RNA replication takes place in the cytoplasm of the infected cell , NS5 distributes between the cytoplasm and the nucleus . Intriguingly , for certain DENV serotypes , the majority of NS5 accumulates in the nucleus [22–26] . The importance of this NS5 subcellular distribution remains unsolved and could be related to an auxiliary function that has not yet been elucidated . In order to identify new functions of NS5 , we conducted an unbiased proteomic study and constructed a map of physical interactions between NS5 and host proteins in the context of a productive DENV infection . These studies revealed the presence of NS5 in complex with core components of the cellular splicing machinery . The process of precursor mRNA ( pre-mRNA ) splicing involves the removal of introns and the precise joining of exons , which results in mature eukaryotic mRNAs [27] . The cellular splicing machinery , the spliceosome , is composed of five uridine-rich small nuclear ribonucleoprotein particles ( U snRNPs ) known as U1 , U2 , U4 , U5 and U6 snRNPs and non-snRNP associated proteins . Each snRNP comprises a small RNA molecule ( snRNA ) , a common set of proteins ( Sm or SMN ) and a number of particle-specific proteins [28] . The exon and intron definition depends on information present in the pre-mRNA: 5’ and 3’ splice sites ( SS ) , branch sites ( BS ) as well as intronic or exonic enhancers and silencers that recruit regulatory proteins that modulate assembly and disassembly of spliceosomal complexes . The splicing machinery is dynamic and exceptionally flexible , allowing fine regulation of gene expression and protein function [27] . Relevant for our studies , modulation of splicing has been shown to be a regulatory mechanism for balancing the host antiviral response [29 , 30] . In this regard , a growing number of pathways that link viral infection , antiviral responses and splicing regulation are being considered as emerging antiviral evasion mechanisms [29–31] . Here , we identified a novel function of DENV NS5 in modulating cellular splicing . NS5 alone or in the context of viral infection binds to CD2BP2 and DDX23 , core components of U5 snRNP , and incorporates in active spliceosomes . Analysis of endogenous splicing events showed that DENV infection affects splicing efficiency and alters cellular splicing patterns . Mechanistic studies using in vitro splicing assays and reporter minigenes in cultured cells indicate that NS5 interferes with pre-mRNA processing . Importantly , depletion of CD2BP2 , DDX23 or EFTUD2 by RNA interference ( RNAi ) increases the efficiency of DENV infection , providing evidence for how NS5 works in promoting viral replication independently of its canonical enzymatic activities . The DENV NS5 protein plays multiple functions during viral infection , likely by interacting with different cellular components . In order to investigate physical interactions of NS5 with cellular proteins during infection , we undertook an unbiased proteomic approach . We used an affinity-purification mass spectroscopy ( AP-MS ) strategy in the context of an infectious virus carrying a tagged NS5 protein . To this end , we developed recombinant DENVs encoding modified NS5 proteins . In contrast to over-expressing the individual protein , this approach facilitates identification of protein-protein interactions in a cellular environment that has been conditioned by the viral infection . Using the 3D structure of NS5 [13] , four different sites were selected to incorporate a purification tag ( 2 X Strep ) in the DENV2 16681 full length clone . The replication capacities of recombinant viruses were compared with that of the WT by immunofluorescence ( IF ) assays after RNA transfection ( Fig 1A ) . Only one of these viruses ( r4-DV ) retained a high replication capacity . A comparative analysis of the WT and r4-DV by IF as a function of the time and growth curves indicated similar replication of the two viruses ( S1 Fig ) . Therefore , the r4-DV was employed as the tool to establish a protocol for NS5-host protein complex isolation and purification ( Fig 1B ) . The pull-down was conducted under mild ( non-denaturing ) conditions , in order to recover proteins that interact either directly or indirectly with NS5 . Each NS5 purification experiment included three different treatments with the corresponding cell extracts then processed in identical conditions: a ) mock infected , b ) DENV WT infected , and c ) DENV NS5-tag infected ( Fig 1B ) . This experiment was repeated sixteen times using two different cell lines , in the presence and absence of nucleases . Eluates were examined by western blotting , coomassie blue and silver staining and submitted to mass spectrometry analysis . It is worth noting that a reduction in the amount of purified NS5 binders was observed if comparing cell lysates treated or not with nucleases , suggesting the presence of nucleic acid-mediated binders ( Fig 1C ) . Based on a total of 48 samples analyzed , we identified 362 cellular proteins as potential NS5 interactors . This list was shortened to 53 candidate proteins after removing any candidates that i ) appeared in negative control experiments ( untagged virus or uninfected cells ) , ii ) were not reproducible in >50% of the experiments , and iii ) were promiscuous protein binders based on databases of over-expression studies [32] . In this regard , proteomic data from strep tagged viral proteins ( capsid , NS3 and NS4B ) were also used as controls to confirm specificity ( S1 Table ) . The set of 53 proteins exhibited an important enrichment in proteins involved in transcription and RNA metabolism . In particular , about 40% of these proteins were splicing related factors , with many of them known as core spliceosomal components ( Fig 1D and 1E ) . Among the splicing proteins detected as NS5 interactors , the most represented complex was the U5 snRNP , for which seven of the eight specific core components were present , suggesting a strong interaction with this particular snRNP ( Fig 1E ) . It is important to highlight the consistent appearance of two members of this particle CD2BP2 and the helicase DDX23 . In addition , previously identified partners of NS5 such as STAT2 and UBR4 , and the splicing proteins PRPF8 and SNRNP200 were also detected [21 , 33] . In summary , we constructed a comprehensive NS5-host protein interactome map in the context of a DENV infection in human cells . To further examine NS5 interaction with core U5 snRNP components , we evaluated interactions with CD2BP2 , DDX23 , and EFTUD2 . The mature tagged viral protein was individually expressed in cultured human cells and extracts were used for non-denaturing pull down assays and western blot analysis . As a control , the same experimental procedure was carried out with a GFP expression vector . NS5 and GFP were expressed and purified with similar efficiencies ( Fig 2A ) . In contrast , co-precipitation of CD2BP2 , DDX23 , and EFTUD2 was only observed with NS5 ( Fig 2A , left panels ) , while the input content of the splicing factors in cells expressing GFP or NS5 was comparable ( Fig 2A , right panels ) . To examine whether NS5 binds to assembled complexes , we performed RIP analysis and evaluated the presence of the RNA component ( snRNAs ) of different snRNPs , since the presence of these molecules is indicative of fully assembled ribonucleoprotein particles . RNA isolated from pull-down experiments obtained from cells expressing NS5 or GFP was used for quantification of different snRNAs by RT-qPCR . Significant enrichment of U1 , U2 , U4 , U5 , and U6 RNAs from 2 to 9 fold was observed in the complexes purified with NS5 in relation to that with GFP ( Fig 2B ) . We next explored whether NS5 binds to active spliceosomes by measuring the presence of immature transcripts ( pre-mRNAs ) that are being spliced out in complex with NS5 . To this end , we evaluated the association of different pre-mRNAs including housekeeping or highly expressed endogenous genes ( HSPCB , Akt , RPS9 , HPRT1 and TBP ) to either NS5 or GFP-containing complexes . We found a significant enrichment of these pre-mRNAs in complexes with the viral protein NS5 compared to GFP ( Fig 2C ) . To further confirm this observation , we analyzed the subcellular localization of NS5 . Considering that active spliceosomes are mainly associated with chromatin [34] , we carried out subcellular fractionation and evaluated the presence of NS5 in the cytoplasm , nucleoplasm and chromatin fractions . As mentioned above , the viral protein NS5 has been previously demonstrated to translocate into the nucleus [22–26] . In this analysis , an evident accumulation of NS5 in the insoluble chromatin fraction was observed together with histone H3 and the U5 snRNP component CD2BP2 ( Fig 2D ) . Taken together , these results provide evidence that NS5 interacts with splicing proteins while they are integrated into assembled snRNPs as part of active spliceosomes . After confirming that the viral protein NS5 interacts with components of the cellular splicing machinery , both by individual expression and in the context of viral infection , we hypothesize that NS5 modulates host splicing . To assess the functional impact of this interaction on cellular splicing patterns , we first evaluated a number of well-characterized endogenous alternative splicing events in infected or mock infected cells . For each particular splicing event , we analyzed the relative abundance of two mRNA isoforms ( lacking or containing an alternative exon cassette ) by RT-PCR . The radiolabeled PCR products were separated by gel electrophoresis , excised from the gel and quantified , as depicted in Fig 3A . The inclusion/exclusion ratios can either increase or decrease in different cellular conditions , also depending on the specific splicing event analyzed . Examples of these two situations in DENV infected cells are shown in Fig 3A and S2 Fig ( ZNF35 and Casp8 ) . In both cases , the relative abundance of the different isoforms was significantly modified by viral infection . This analysis was extended to a battery of alternative splicing events in two human cell lines ( Fig 3B ) . The fold change of isoforms in infected versus uninfected cells showed that around 40% of the analyzed events were substantially modified during infection . These results provide evidence of an impact of DENV2 infection on cellular splicing . Because the nuclear localization of NS5 differs among DENV serotypes [22 , 24] , we also examined a subset of splicing events upon DENV4 infection ( S3 Fig ) . The results suggest that the two viruses modulate splicing to a different extent , providing evidence of a possible serotype specific alteration of splicing . It has been previously reported that modulation of alternative splicing is a mechanism to regulate the balance of the cellular antiviral state [30 , 35–38] . For instance , alternative splicing of IKKε results in different exon-skipping isoforms with one variant that retains full antiviral activity ( v-wt ) and two variants ( v1 and v2 ) that yield a dominant negative or partially inactive protein [39] . To extend our observation , we examined splicing variants of IKKε and MxA in DENV infected or mock infected cells . In the case of IKKε , an increased abundance of sv1 and sv2 relative to the v-wt was observed , suggesting the expression of dominant negative forms of IKKε during DENV infection ( Fig 3C ) . Similarly , in the case of MxA , the appearance of exclusion isoforms was observed concurrently with a reduction in the amount of the canonical ( inclusion ) isoform ( Fig 3C ) . This resembles a shift in the splicing pattern previously described for herpes virus infection , in which an MxA isoform supports instead of restricting viral infection [40] . Together , these observations suggest a model in which DENV infection alters cellular splicing patterns leading to changes in isoform abundance of antiviral factors , which could facilitate viral replication . The viral protein NS5 physically interacts with the spliceosome when individually expressed in human cells , and DENV infection causes a pronounced alteration of endogenous cellular splicing . Thus , we speculate that NS5 is , at least in part , responsible for splicing modulation during viral infection . To examine this possibility , splicing was monitored by analyzing mRNA isoforms derived from splicing reporter minigenes co-transfected with expression vectors for different variants of NS5 or control proteins . The reporter minigenes are cloned gene fragments that include the splicing event of interest , in particular , an alternative exon and the flanking intronic regions , downstream of a constitutive promoter . Transcription from this promoter generates minigene-derived pre-mRNA molecules , which are canonically processed to render distinguishable mature products . These reporters have been widely used to characterize cis-acting elements and trans-acting splicing modulators [41] . In our experimental setting , we used three different splicing reporter minigenes ( CFTR , EDI and Bclx ) [42–44] and the proportion of mRNA isoforms derived from each of them was analyzed by RT-PCR and quantified as indicated above . Different concentrations of the NS5 expression construct were co-transfected with each minigene . A dose dependent modulation of minigene-derived mRNA isoforms was observed upon expression of NS5 protein ( Fig 4A ) . A quantitative analysis of PCR products corresponding to each isoform indicated 2 , 5 and 1 . 5 fold change in the inclusion/exclusion ratios for CFTR , EDI and Bclx minigenes , respectively , in the presence of the viral protein ( Fig 4B ) . The magnitude of the changes produced by NS5 on the analyzed splicing events is comparable to that observed when cellular levels of well-known splicing regulators were perturbed [45] , indicating that NS5 modulates cellular splicing patterns . NS5 exerts two enzymatic activities on RNA molecules , RNA polymerase and methyltransferase ( MTase ) . In addition to acting on specific viral RNA templates , NS5 has non-specific terminal transferase activity , which adds non-templated nucleotides at the 3’ end of an RNA and it is capable of internally methylating RNA molecules [46 , 47] . Any of these activities could presumably modulate splicing , by altering pre-mRNA or snRNA properties . Thus , we evaluated whether the enzymatic activities of NS5 were involved in splicing modulation . To this end , we designed NS5 constructs with mutations that specifically impair either the MTase ( Mut-MTase ) or the RNA dependent RNA polymerase ( Mut-RdRp ) activity . Each of these mutant proteins retained the other ( unmodified ) enzymatic activity [47–49] . Using the minigene splicing reporter assay , it was found that both NS5 mutants were capable of causing a significant alteration of minigene alternative splicing patterns as compared with GFP or a control empty vector . The magnitude of the change observed with the two mutants was similar to that with the wild type NS5 protein ( Fig 4C ) . These results suggest that NS5 binding to spliceosome components , but not its enzymatic activities , is required for splicing regulation . In addition , we examined the impact of each domain of NS5 separately , by expressing the MTase or the RdRp domain , and only the RdRp showed a reduction of the inclusion/exclusion ratio of reporter minigenes ( S4 Fig ) . To investigate mechanistic aspects of the alteration of splicing by DENV infection and its association to NS5 , we examined known properties of this viral protein . It has been previously reported that NS5 , as a proteolitically processed protein originated from an uncleaved precursor , has the ability to induce STAT2 degradation by gathering the ubiquitin ligase UBR4 and STAT2 [21] . Based on this well established activity of NS5 , we investigated whether DENV infection could alter the stability of splicing factors that were identified as NS5 binders . To examine this possibility , the amount of different spliceosomal proteins was analyzed in DENV infected and mock infected cells . Western blot analyses were performed at different times post infection to evaluate the levels of U5 snRNP specific proteins . As previously reported , extracts from DENV infected cells showed a drastic reduction of STAT2 levels , used as a control ( Fig 5A ) . In contrast , no significant changes were observed for CD2BP2 , DDX23 , EFTUD2 , SNRNP40 ( U5 snRNP components ) or SF3A2 ( U2 snRNP component ) levels up to 48 hours of infection as compared with those in uninfected cells ( Fig 5A ) . We conclude that DENV infection does not induce degradation of these splicing factors as it was previously observed for STAT2 . Regarding subcellular distribution of splicing factors , it has been previously reported that Poliovirus 2A protease affects the function of certain splicing factors and RNA-binding proteins by regulating their nuclear-cytoplasmic shuttling [50] [51] . Thus , we evaluated a possible alteration of the subcellular localization of splicing proteins during DENV infection . Using confocal microscopy , we observed that while DENV2 NS5 protein localizes in the nucleus of infected cells , as previously described , viral infection was not associated to changes in the nuclear localization of spliceosomal factors . In this regard , NS5 as well as the U5 snRNP components , CD2BP2 and DDX23 , were detected in the nucleus of infected cells ( Fig 5B ) . We then considered the possibility that NS5 binding to spliceosomal proteins could alter snRNPs organization and/or stability . It has been previously reported that a sensitive indicator for snRNP stability is the measurement of snRNAs abundance because these RNAs are unstable when freed from the RNP particles [52] . Thus , we measured the global relative amounts of U2 , U4 , U5 , and U6 snRNAs by RT-qPCR in infected and uninfected cells . The results show no significant changes in snRNAs abundance during DENV infection in two different human cell lines ( Fig 5C ) . Together these data indicate that viral infection does not change the levels or the subcellular localization of U5 snRNP protein components , nor the stability of the snRNP complexes . To better understand the mechanism by which NS5 modulates splicing , we investigated whether NS5 has a direct effect on the splicing reaction . To examine this possibility , we used well-stablished in vitro system employing a radiolabeled pre-mRNA substrate that undergoes efficient splicing when combined with splicing-competent nuclear extracts from HeLa cells . This is a reconstituted system in which the substrate has been optimized to follow the reaction in controlled conditions [53–55] . The progress of the reaction over time can be monitored by denaturing polyacrylamide gel electrophoresis that allows the detection of initial substrates , RNA intermediates , and final splicing products , as exemplified in Fig 6A . The reaction was evaluated in the presence of purified NS5 protein , either using a heat-denatured or a native NS5 sample ( Fig 6B ) . An unrelated protein was also used as control ( S5 Fig ) . This study showed inhibition of mature mRNA product accumulation as well as lower amounts of intermediate ( exon-lariat ) when the native NS5 protein was present , suggesting inhibition of early catalytic steps . This is in agreement with sustained levels of pre-mRNA observed along the reaction in the presence of NS5 ( Fig 6C ) . Also , addition of increasing concentrations of the viral protein to the reaction shows a dose dependent inhibitory effect on mature mRNA product accumulation ( Fig 6B , right panel ) . These studies indicate that NS5 directly alters the efficiency of splicing by interfering at early stages of the reaction . It has been recently reported that RIG-I pre-mRNA splicing efficiency is susceptible to alterations in the EFTUD2 protein levels [30] . Considering that DENV NS5 interacts with this spliceosomal protein as shown in Fig 2A and interferes with an in vitro splicing reaction as shown if Fig 6 , we analyzed the impact of NS5 on endogenous RIG-I pre-mRNA maturation . The RIG-I pre-mRNA is about 70 kb long and contains 18 exons . Based on this , we evaluated splicing efficiency by focusing at the excision of the intron between exons 10 and 11 ( Fig 7A ) . To examine the influence of the mature NS5 protein on RIG-I mRNA processing , the ratio between mature- and pre-mRNA was evaluated by RT-qPCR . Interestingly , NS5 expression led to a significant reduction in RIG I splicing efficiency as compared with controls ( p<0 . 001 ) ( Fig 7B ) . The efficiency of RIG-I pre-mRNA maturation was found to be very robust under different conditions . For instance , RIG-I induction by α-IFN increased the levels of immature and mature forms of the transcript , and the efficiency of processing remained constant ( Fig 7C ) , highlighting the effect observed with NS5 . It is also worth noting that DENV NS5 inhibits IFN signaling by interfering with STAT2 as described above ( Fig 4A ) . However , over-expression of mature NS5 is incapable of triggering STAT2 degradation [20] ( Fig 7D ) . Together these results further support an impact of the viral protein NS5 on modulating splicing efficiency in vitro and in cell culture . To examine the global impact of DENV infection on splicing , we used genome wide RNA sequencing . Transcriptome analysis of DENV infected and mock infected cells at 24 and 36 hours post infection were performed , and a recently developed R package for an integrative splicing analysis was used [56] . ASpli combines statistical information from exon , intron , and splice junction differential usage to calculate differences in the percentage of exon inclusion ( PSI metric ) and percentage of intron retention ( PIR metric ) ( Fig 8A and Materials and Methods ) . This method allows an unbiased evaluation of annotated and novel splicing events . Biological triplicates of DENV and mock infected cells were used to construct mRNA libraries that were sequenced using an Illumina Hiseq 4000 . The reads were mapped against the human genome ( UCSC , HG19 ) . For the analysis of differential splicing , multi-exonic genes were partitioned into features defined as “bins” corresponding to exonic regions , intronic regions and regions annotated as alternatively spliced . About 175000 bins were analyzed for each sample , 26% corresponding to intron bins and 74% to exon bins ( Fig 8B ) . After the analysis , the data was filtered using a criteria to identify reliable changes in splicing during DENV infection . Using a threshold of difference in PSI or PIR between experimental conditions >5% and false discovery rate ( FDR ) of ≤10% , ASpli identified changes in 451 and 858 events at 24 and 36 hours post-infection , respectively ( Fig 8B , S2 and S3 Tables ) . Interestingly , we observed an enrichment of intron sequences in mature RNAs at both 24 and 36 post DENV infection . Intron retention ( IR ) is the process by which an intron remains unspliced in a polyadenylated transcript , and although it was considered a rare event , increase IR rates have been recently reported in a large number of human cells [57] . In infected cells , IR increased in about 70% of the event that changed at 24 and 36 hours , indicating a reduction of intron excision in hundreds of transcripts , which could be associated to a perturbation of the splicing catalysis ( Fig 8B ) [57] . We conclude that DENV infection has an impact on splicing efficiency of a large subset of transcripts . Our proteomic data supported a novel interaction of DENV NS5 with U5 snRNP components , while in vitro and cell based assays indicated a role of the viral protein on modulating cellular splicing . In addition , transcriptome analysis of infected cells revealed splicing changes with intron removal perturbation in a large number of genes . To evaluate whether U5 snRNP composition is relevant for DENV infection , we conducted siRNA-mediated knockdowns of U5 proteins identified as NS5 binders and assessed the virus ability to infect and replicate ( Fig 9 ) . Specific siRNAs were transfected into A549 cells , 48 hours later the cells were infected with a reporter DENV virus encoding luciferase [58] , and viral replication was evaluated as a function of time . Negative and positive controls , non-related siRNAs or siRNAs directed to the luciferase gene , respectively , were included . Silencing of STAT2 was used as a control for a negative regulator of viral replication that binds NS5 . Interestingly , depletion of CD2BP2 or DDX23 showed a significant increase on viral replication . It is important to mention that the rise of viral replication detected upon silencing any of these U5 components was similar or higher than that obtained by silencing STAT2 . In contrast to the effect observed with components of U5 , silencing a core component of U2 snRNP ( SF3A2 ) greatly reduced viral replication ( Fig 9A ) , supporting the idea that viral modulation of cellular splicing must be specific to provide a beneficial environment for viral infection . To confirm this observation the impact of silencing core components of U5 on replication of the WT virus was evaluated . In this case , an increase in viral RNA replication was observed upon silencing CD2BP2 , DDX23 or EFTUD2 ( Fig 9B ) . The favorable environment for viral replication observed upon silencing U5 components could be the consequence of changes in cellular splicing , including that of antiviral and/or proviral transcripts . To evaluate whether these splicing components modulate cellular innate immunity , we induced an antiviral state by infection with an unrelated virus that lacks known counteracting mechanisms ( Newcastle disease virus . NDV ) and evaluated gene expression of antiviral factors in cells that were or not silenced for U5 components ( Fig 9C ) . Silencing EFTUD2 resulted in a significant reduction of RIGI , ISG15 , and IL-8 induction , while knocking down CD2BP2 reduced the expression of ISG15 and IL-8 . This observation provides a link between the splicing machinery , DENV infection and the host antiviral response , emphasizing the relevance of certain U5 components in restricting viral infection . Here , we discovered a novel property of the DENV NS5 protein in subverting cellular splicing . Proteomic analysis and functional studies revealed that NS5 binds to active spliceosomes in infected cells and modulates splicing . Mechanistic studies indicate that NS5 interacts with components of the U5 snRNP and reduces the efficiency of pre-mRNA splicing . Our findings support a model in which NS5-mediated regulation of specific spliceosomal components renders an advantageous cellular environment for DENV replication , providing a new function for the viral NS5 polymerase in infected cells . In this work , we present a comprehensive DENV NS5 protein-protein interaction network using a tagged infectious virus . The study revealed a significant enrichment of splicing related proteins , with particular abundance of U5 snRNP components . About 20% of the proteins identified were previously associated with NS5 or DENV infection . These proteins include STAT2 , THRAP3 , PRPF8 , SNRNP200 and UBR4 that have been recently reported to interact with DENV NS5 [21 , 33] , ERC1 that was found to be downregulated during DENV infection [59] , and ILF3 that also interacts with the viral NS3 protein [60] . Because NS5 was proposed to function as an adaptor for the ubiquitination system in association with UBR4 , and the screen also identified UBR5 as NS5 binder , it will be interesting to further evaluate whether this ubiquitin ligase is involved in NS5-mediated proteasome degradation of specific host factors . In addition , our study identified two members of the Y box binding protein family , YBX2 and YBX3 , as NS5 binders . In this regard , YBX1 is an alternative splicing regulator that has been previously associated to DENV replication [61] . Based on the enrichment of splicing factors as NS5 binders , we evaluated the impact of DENV infection on cellular splicing by analyzing first a panel of endogenous well-characterized alternative splicing events . A substantial change in inclusion/exclusion ratios was observed when comparing infected versus mock infected cells . Differential splicing during DENV infection was also examined at the wide genome level by high throughput RNA sequencing . This analysis revealed changes in splicing and defects in intron removal during viral infection ( Fig 8 ) . It is important to mention that increased intron retention was recently associated to pathological conditions [62 , 63] , and in some of these cases mutations or alterations of core splicing factors were reported [64] . We hypothesize that perturbation of splicing efficiency during DENV infection is , at least partially , explained by NS5 interaction with core splicing factors . Our findings are in agreement with a recent study using RNAseq in DENV infected cells [38] . In this particular report , a transcriptome comparison of cells infected with WT or an attenuated strain of DENV1 revealed important differences in splicing of transcripts involved in innate immune responses and cell cycle control [38] . The alteration in cellular splicing observed during DENV replication could be the result of a combination of processes that take place within infected cells . On one hand , it can be a consequence of the cellular response to infection , associated with cellular damage or to a cellular antiviral state . On the other hand , splicing alteration could be a viral mechanism to counteract the antiviral response or to favor gene expression of pro-viral factors . In this regard , there is an emerging theme that links viral infection with specific alteration of U5 snRNP components . Several lines of evidence obtained with different RNA viruses support the significance of this emerging link . In the case of picornaviruses , it has been shown recently that the viral polymerase 3D binds to PRPF8 , a core components of U5 , and blocks mRNA maturation , with the consequent accumulation of the lariat intermediate [65] . In addition , RIPseq analysis of 3D-PRPF8 complexes identified transcripts associated to cellular growth , proliferation and differentiation [65] . In the case of hepatitis C virus ( HCV ) , a siRNA screen aimed to detect genes associated to the antiviral response revealed the involvement of three specific splicing factors: EFTUD2 , PRPF8 , and SNRNP200 , which are all components of the U5 snRNP complex [37] . Further research with HCV showed that EFTUD2 contributes to the antiviral response through RIG-I/MDA5 pre-mRNA maturation and it was found that viral infection down-regulates EFTUD2 , resulting in splicing alteration of antiviral factors [30] . Our results using DENV are consistent with these previous observations and contribute to the model that viral infection perturbs cellular splicing . Despite the similarity with that described for HCV , the underlying mechanism exploited by DENV seems to be different since viral infection does not involve a significant reduction of the abundance of U5 components or a reduction in the assembly of snRNP complexes ( Figs 4 and 5 ) . In addition , the impact on splicing efficiency during DENV infections , observed by increase intron retention in polyadenylated transcripts , also provides an evidence for viral manipulation of the splicing machinery ( Fig 8 ) . Mechanistic studies suggest that DENV NS5 binding to active spliceosomes modulates splicing efficiency , without involving any known NS5 enzymatic activity neither altering localization/abundance of splicing components during DENV infection ( Figs 4 and 5 ) . In addition , our results support that NS5 interaction with the spliceosome is not mediated by RNA binding because treatment with RNases did not impair the NS5-U5 complex interaction . In vitro studies using reconstituted systems with recombinant NS5 showed a clear reduction in the amount of RNA intermediates and pre-mRNA processing in the presence of the native NS5 protein . We propose a model in which binding of NS5 to protein components of the spliceosome reduces the efficiency of the splicing reaction . In the same line , evaluation of splicing efficiency of endogenous pre-mRNAs , such as RIG-I , as well as alternative splicing events derived from a variety of endogenous genes or reporter minigenes , showed that NS5 expression was sufficient to alter both splicing efficiency and alternative splicing . In this context , it is important to highlight that a growing body of experimental evidence points out that alterations in core spliceosomal components by depletion , mutation or pharmacological treatment , alter different cellular pathways . This seems to be dependent on the differential sensitivity or differential requirement for splicing factors for the processing of distinct pre-mRNAs [66 , 67] . Although NS5 alone is able to lower the efficiency of the splicing reaction in vitro and in cells , in the context of infection , splicing modulation is extremely more complex . In this regard , our findings with transcriptome analysis provide a new angle to study the link between splicing modulation and viral infection . The question is whether there are key changes in the splicing machinery and cellular gene expression that yields a more permissive scenario for viral replication . Interestingly , manipulating the levels of core splicing factors resulted in splicing changes with different outcomes for viral infection . For instance , upon silencing certain U5 components we observed: a ) increased DENV replication and b ) reduced ability to induce the antiviral response by an unrelated virus . In contrast , silencing U2 components caused viral inhibition ( Fig 9 ) . This observation supports a model in which modulating specific splicing components leads to a more or less restrictive environment for viral replication . The finding that several RNA viruses evolved mechanisms to interfere with the U5 snRNP and that different viral proteins bind components of U5 , points to a function of this specific splicing particle in facilitating an antiviral state and is consistent with the known essential roles of U5 components during spliceosome catalytic activation . In this context , our study provides novel information strengthening this emerging viral mechanism that manipulates cellular splicing . DENV is currently the most significant insect-borne viral pathogen around the world without effective vaccines or antivirals . Ultimately , increased understanding of the host-virus interaction network at the molecular level will bring novel ideas for viral inhibition and control . Mammalian cells: A549 cells ( human lung adenocarcinoma epithelial cell line , ATCC , CCL-185 ) and Huh7 cells ( human hepatoma cell line , JCRB Cell Bank #0403 ) were maintained in Dulbecco’s Modified Eagle’s Medium/ F-12 ( Ham’s ) , HEK 293 cells ( human embryonic kidney cell lines , ATCC CRL-1573 ) were maintained in Dulbecco’s Modified Eagle’s Medium , HeLa cells ( human epithelioid cervix carcinoma cell line , ATCC , CCL2 ) ‎ were maintained in RPMI-1640 Medium , BHK-21 cells ( baby hamster kidney cell line , ATCC , CCL-10 ) were maintained in Minimum Essential Medium Alpha . All the media were supplemented with 10% fetal bovine serum , 100 U/ml penicillin , and 100 μg/ml streptomycin . Mosquito cells: C6/36HT ( Aedes albopictus cells ATCC , CRL-1660 were adapted to grow at 33°C ) were maintained in Leibovitz's L-15 Medium , supplemented with 10% fetal bovine serum , 100 U/ml penicillin , 100 μg/ml streptomycin , 0 . 3% tryptose phosphate , 0 . 02% glutamine , and 1% MEM non-essential amino acids solution . Wild-type virus was obtained from an infectious clone of the DENV2 strain 16681 ( GenBank accession number U87411 ) , plasmid pD2/ICAflII [68] . Recombinant viruses with a purification tag in NS5 ( DENV NS5-tag ) were generated by replacing sites NheI-NruI ( r1-DV ) , or StuI-AvrII ( r2-DV ) , or AvrII-AflII ( r3-DV and r4-DV ) in pD2/ICAflII with fragments derived from overlapping PCRs containing the desired tag-coding sequence ( a double Strep-tag sequence , amino acids WSHPQFEK-GGGS-WSHPQFEK ) . Viral RNA transcription and transfection was performed as previously described [69] . Briefly , DENV plasmids were linearized using XbaI and used as template for transcription with T7 RNA polymerase in the presence of cap analog ( m7GpppA ) . RNA transcripts were transfected with Lipofectamine 2000 ( Invitrogen ) into BHK-21 or C6/36HT cells . Viral replication was assessed through indirect immunofluorescence . Supernatants were harvested at different times post-transfection and used to quantify infectious DENV particles by plaque assays . Rabbit polyclonal antibodies against CD2BP2 , DDX23 , EFTUD2 , and SF3A2 were gently provided by Dr . Reinhard Lührmann ( Max Planck Institute for Biophysical Chemistry , Gottingen , Germany ) . Antibodies against SNRNP40 ( goat polyclonal ) and against STAT2 ( rabbit polyclonal ) were purchased from Santa Cruz Biotechnology ( sc-162407 and sc-476 , respectively ) . Anti-Strep-tag antibody ( mouse monoclonal ) was purchased from EMD Millipore ( 71590–3 ) . Anti-GAPDH ( mouse monoclonal ) was purchased from Abcam ( ab8245 ) . For the detection of the DENV envelope protein ( E ) , mouse monoclonal anti-E antibody E18 was used [70] . Anti-NS5 antibodies ( rabbit polyclonal and mouse polyclonal ) were obtained in our laboratory . Immunofluorescence assays were performed as previously described [71] . Briefly , A549 cells were seeded into 24-well plates containing glass coverslip . Twenty four hours later , the cells were mock infected or infected with DENV WT using a multiplicity of infection ( MOI ) of 10 . After 24 hours , coverslips were collected and the cells were fixed with paraformaldehyde 4% , sucrose 4% , in PBS pH 7 . 4 at room temperature for 20 minutes . PFA fixed cells were then permeated with 0 . 1% Triton X-100 for 4 minutes at room temperature . Images were obtained with a Carl Zeiss LSM 5 Pascal confocal microscope . Huh7 or HEK 293T ( ~2x107 ) cells were either infected ( MOI of ~1 ) with DENV WT , DENV NS5-tag or mock infected . At 48 hours post-infection , cells were harvested by scraping , washed with PBS , centrifuged at 100 rcf for 5 minutes at 4°C , and the pellet lysed with 1 ml of lysis buffer ( 50 mM Tris-HCl pH 7 . 4 , 150 mM NaCl , 1 mM EDTA , 0 . 5% NP40 , protease and phosphatase inhibitors , +/- DNase I and RNase A ) chilled on ice . Cell lysates were incubated in rotating mixer for 30 minutes at 4°C and then clarified by centrifugation at 3000 rcf for 20 minutes at 4°C . The clarified lysates were over-night incubated with 30 μl Strep-tactin sepharose beads—50% suspension ( IBA ) in rotating mixer at 4°C . Beads were recovered by centrifugation at 400 rcf for 2 minutes at 4°C and washed four times with 1 ml of wash buffer ( 50 mM Tris-HCl pH 7 . 4 , 150 mM NaCl , 1 mM EDTA , 0 . 05% NP40 ) chilled on ice . A fifth wash was conducted in absence of NP40 and beads were transferred to protein LoBind tubes ( Eppendorf ) . Beads were incubated with 40 μl of elution buffer ( 50 mM Tris-HCl pH 7 . 4 , 150 mM NaCl , 1 mM EDTA , 2 . 5 mM desthiobiotin ) for 30 minutes at room temperature , continuous vortexing ( beads in suspension ) . The eluates containing the purified protein complexes were obtained by a final centrifugation at 400 rcf for 2 minutes at 4°C and supernatant aspiration . Aliquots from the different samples were denatured in SDS sample buffer at 80°C for 15 minutes and loaded on precast 4–12% acrylamide gels ( BioRad ) . Gels were stained with Pierce Silver Stain Kit ( Thermo Scientific ) or NOVEX Colloidal Blue Staining Kit ( Invitrogen ) . Protein identification was conducted using a LTQ Orbitrap Velos ( Thermo Scientific ) mass spectrometer at the Department of Cellular and Molecular Pharmacology , University of California , San Francisco . STRING v10 ( [72]; http://string-db . org ) was used to retrieve the interactions among human proteins that co-purified with Strep-tagged NS5 and to determine enrichments for Gene Ontology ( GO ) categories . HEK 293T cells ( ~9x106 ) were transfected using Lipofectamine 2000 ( Invitrogen ) with 12 μg of pcDNA NS5-Strep-tag or 1 μg pcDNA GFP-Strep-tag plus 11 μg empty pcDNA ( as control ) . Cells were harvested 48 hours post-transfection by using a cell scraper , washed with PBS , and pelleted at 100 rcf for 5 minutes . Cell pellets were stored at -80°C before processing . The Strep-tag pull-down was conducted as described above , but introducing few modifications . Cells were lysed in 500 μl of lysis buffer , 10 μl Strep-tactin sepharose beads—50% suspension ( IBA ) were added to the clarified lysates , and after over-night incubation the beads were washed three times with 1 ml of wash buffer containing 0 . 5% NP40 plus once in absence of NP40 . To analyze by SDS PAGE/Western blot the co-precipitation of different proteins , beads were heated for 10 min at 70°C in 50 μl of SDS sample buffer . Conversely , to analyze the presence of specific RNAs , beads were treated with 500 μl of Trizol ( Invitrogen ) . A small-scale fractionation was conducted as described by [73] . Briefly , ~3 x106 HEK 293T cells were transfected with 4 μg of pcDNA-NS5-Strep-tag or 0 . 4 μg pcDNA-GFP-Strep-tag plus 3 . 6 μg empty pcDNA ( as control ) . After 48 hours , cells were harvested by using a cell scraper , washed with PBS , and pelleted at 100 rcf for 2 minutes . Cells were resuspended in 70 μl of Buffer A ( 10 mM HEPES pH 7 . 9 , 10 mM KCl , 1 . 5 mM MgCl2 , 0 . 34 M sucrose , 10% glycerol , 1 mM DTT , and protease inhibitors ) . Triton X-100 was added to a final concentration of 0 . 1% . Cells were incubated on ice for 8 minutes and then centrifuged at 1300 rcf for 5 minutes at 4°C . Supernatant ( fraction S1 ) was separated from pellet ( nuclei , fraction P1 ) . Fraction S1 was clarified by centrifugation at 20000 rcf for 5 minutes at 4°C and the supernatant ( cytoplasm , fraction S2 ) was collected . Fraction P1 was washed once with Buffer A , lysed for 30 minutes in 30 μl of Buffer B ( 3 mM EDTA , 0 . 2 mM EGTA , 1 mM DTT , and protease inhibitors ) , and centrifuged at 1700 rcf for 5 minutes at 4°C . Supernatant ( nucleoplasm , fraction S3 ) was separated from pellet ( chromatin , fraction P3 ) . Fraction P3 was washed once with Buffer B . Fractions S2 , S3 , and P3 were diluted or resuspended in SDS sample buffer , heated for 10 min at 70°C , and analyzed by SDS PAGE/Western blot . Total cellular RNA was isolated by using Trizol ( Invitrogen ) and reverse transcribed to cDNA with random deca-oligonucleotide primer mix . Quantitative PCRs ( qPCRs ) were performed by using SYBR Green dye and specific primers for different snRNAs , pre-mRNAs , and mRNAs . GAPDH mRNA was used as a control housekeeping gene . The specific primers are listed in S4 Table . Cells were mock infected or infected ( MOI ~1 ) with DENV WT and harvested 48 hours post-infection . Total RNA was extracted by using Trizol ( Invitrogen ) . The isolated RNA was subjected to RT-PCR in presence of α-32P dCTP . The specific primers are listed in S1 Table . Radiolabelled PCR products were electrophoresed in 6% polyacrylamide native gels , which were subsequently dried and exposed to X-ray films ( Agfa ) . The bands detected by autoradigraphy were excised from gel and the radioactivity was measured in a scintillation counter ( Cerenkov method ) to then calculate the ratio between the amplification products corresponding to different mRNA isoforms . The splicing reporter minigenes are described elsewhere [43 , 44 , 74] . HeLa cells ( ~0 . 5x106 ) were transfected with 100 ng of each plasmid containing the indicated minigenes with different amounts of the empty pcDNA , pcDNA NS5-Strep-tag and/or pcDNA GFP-Strep-tag ( total of 900 ng ) . Forty-eight hours later , RNA was extracted and subjected to RT-PCR in presence of α-32P dCTP . The ratio between the different splicing isoforms derived from each minigene was calculated as described above . NS5 polymerase and methyltransferease activity mutants were obtained by site-directed mutagenesis . For methyltransferase activity mutant the residue D146 was replaced by A [46 , 47] while for polymerase activity mutant the catalytic triplet GDD was replaced by AAA . MINX [32P]-labelled pre-mRNA was synthesized from a linearized MINX plasmid [54] using T7 RNA polymerase ( Ambion ) . HeLa nuclear extract was prepared according to [55] and kindly provided by the Luhrmann laboratory . Splicing reactions were performed in the presence of 40% nuclear extract , 24 mM Hepes-KOH ( pH 7 . 9 ) , 2 . 4 mM MgCl2 , 2mM ATP , 25 mM KCl and 20 mM creatine phosphate . Reactions were incubated at 30°C for the times indicated . RNA was extracted using Trizol and analyzed by 14% PAGE followed by autoradiography . Biological triplicates of DENV and mock infected cells were used for RNA-seq analysis . To construct the libraries , RNA was first extracted from cells using Trizol , followed by chloroform extraction and DNAse treatment and isopropanol precipitation . RNA quality was determined by 260/280 ratio using a Nanodrop 2000 and by agarose gel analysis of 28S and 18S rRNA bands . The resulting RNA was then PolyA purified using the BioO Scientific PolyA purification beads ( catalog number 512979 ) . PolyA purified RNA was then used to create sequencing libraries using the KAPA stranded RNA-seq kit ( catalog number KK8400 ) . Libraries were sequenced using paired-end reads on an Illumina Hiseq 4000 . Sequence reads were aligned against the Homo sapiens genome ( Hg19 ) with TopHat v2 . 1 . 1 [75] with default parameters . Count tables for the different feature levels were obtained from bam files using custom R scripts and considering the Hg19 transcriptome . Raw sequences ( fastq files ) and count tables at gene , exon , intron , AS bin and junction levels used in this paper have been deposited in the Gene Expression Omnibus ( GEO ) database ( accession no . GSE84285 ) . We followed the strategy presented in ASpli [56] for the analysis of differential splicing . Briefly , multiexonic genes were partitioned into features defined as “bins” and then classified into exon , intron and annotated alternative splicing bins . Using this strategy , the transcriptome was partitioned in 468132 bins ( S2 and S3 Tables ) . These datasets were then filtered according to several criteria applied at the gene and bin level . First , defined subgenic regions ( i . e . bins ) were considered for differential splicing analysis only if the genes with which they are associated were expressed above a minimum threshold level ( more than 10 reads per gene and RD > 0 . 05 ) in all experimental conditions . Next , bins were considered for differential splicing analysis only if they had more than 5 reads and a RD bin/RD gene ratio > 0 . 05 , in at least one experimental condition . After applying these filters , reads summarized at the bin level were normalized to the read counts of their corresponding gene . Differential bins usage was estimated using the edgeR package version 3 . 14 . 0 [76] , and resulting P-values were adjusted using a false discovery rate ( FDR ) criterion . We then computed the metrics PSI ( percent spliced-in ) and PIR ( percent intron retention ) , which were used as a filtering criteria for the splicing analysis combined with FDR ( Fig 8 ) . Bins with FDR lower than 0 . 1 and absolute delta PSI/PIR between 5 and 95% were considered differential used bins . The selected bins are included in supplementary information ( S2 and S3 Tables ) . RNA interference experiments were carried out using siGENOME ON-TARGET plus SMART pool siRNA oligonucleotides ( Dharmacon RNA Technologies , Lafayette , CO , USA ) . The control non-related siRNA used was directed against Firefly luciferase . Twenty-four hours after seeding in 24-well plates , A549 cells were transfected with the corresponding siRNA using Oligofectamine ( Invitrogen ) . Briefly , 25 pmol of siRNA in 50 μl of Opti-MEM ( Invitrogen ) were mixed with 2 μl of Oligofectamine in 50 μl of Opti-MEM and incubated for 20 minutes . The mix was added to a 50% confluent A549 cells monolayer and incubated overnight . Then , the supernatant was replaced with complete medium . After 48 hours of transfection , cells were infected with different viruses . For the reporter DENV containing the Renilla luciferase gene [71] cells were incubated for 36 hours prior to measuring luciferase activity . For infection with WT DENV ( 16681 ) , cells were infected with a MOI of 10 and RNA was purified using Trizol at 10 and 24 hpi . For Newcastle diseade virus infection ( NDV ) , transfected cells were infected with MOI of 0 . 1 and RNA was purified at 24 hpi . Quantitative real time PCR was performed as previously described using primers listed on S1 Table .	Mapping host-pathogen interactions has proven fundamental for understanding how viruses manipulate host machinery and how cellular processes are regulated during infection . Dengue virus poses a major threat to public health: two-thirds of the world’s population is now at risk from infection by this mosquito-borne virus . In this work , using a global proteomic approach in the context of viral infections with tagged dengue viruses , we constructed a comprehensive protein-protein interaction map of the multifunctional NS5 viral protein . NS5 is central for viral RNA replication and for immune evasion . Our studies revealed the interaction of NS5 with core components of the splicing machinery , specifically with proteins of the U5 small nuclear ribonucleoprotein particle , and that viral infection reduces splicing efficiency . Mechanistic studies analyzing endogenous splicing events and in vitro splicing assays indicated that NS5 binds active spliceosomes and reduces the efficiency of pre-mRNA processing . Our results provide a new function of the dengue virus NS5 protein and support a model in which manipulation of specific splicing components favors viral infection .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "viral", "transmission", "and", "infection", "gene", "regulation", "rna", "extraction", "microbiology", "alternative", "splicing", "extraction", "techniques", "research", "and", "analysis", "methods", "small", "interfering", "rnas", "genome", "complexity", "gene", "expr...	2016	The Dengue Virus NS5 Protein Intrudes in the Cellular Spliceosome and Modulates Splicing
Cilia have a unique diffusion barrier ( “gate” ) within their proximal region , termed transition zone ( TZ ) , that compartmentalises signalling proteins within the organelle . The TZ is known to harbour two functional modules/complexes ( Meckel syndrome [MKS] and Nephronophthisis [NPHP] ) defined by genetic interaction , interdependent protein localisation ( hierarchy ) , and proteomic studies . However , the composition and molecular organisation of these modules and their links to human ciliary disease are not completely understood . Here , we reveal Caenorhabditis elegans CEP-290 ( mammalian Cep290/Mks4/Nphp6 orthologue ) as a central assembly factor that is specific for established MKS module components and depends on the coiled coil region of MKS-5 ( Rpgrip1L/Rpgrip1 ) for TZ localisation . Consistent with a critical role in ciliary gate function , CEP-290 prevents inappropriate entry of membrane-associated proteins into cilia and keeps ARL-13 ( Arl13b ) from leaking out of cilia via the TZ . We identify a novel MKS module component , TMEM-218 ( Tmem218 ) , that requires CEP-290 and other MKS module components for TZ localisation and functions together with the NPHP module to facilitate ciliogenesis . We show that TZ localisation of TMEM-138 ( Tmem138 ) and CDKL-1 ( Cdkl1/Cdkl2/Cdkl3/Cdlk4 related ) , not previously linked to a specific TZ module , similarly depends on CEP-290; surprisingly , neither TMEM-138 or CDKL-1 exhibit interdependent localisation or genetic interactions with core MKS or NPHP module components , suggesting they are part of a distinct , CEP-290-associated module . Lastly , we show that families presenting with Oral-Facial-Digital syndrome type 6 ( OFD6 ) have likely pathogenic mutations in CEP-290-dependent TZ proteins , namely Tmem17 , Tmem138 , and Tmem231 . Notably , patient fibroblasts harbouring mutated Tmem17 , a protein not yet ciliopathy-associated , display ciliogenesis defects . Together , our findings expand the repertoire of MKS module-associated proteins—including the previously uncharacterised mammalian Tmem80—and suggest an MKS-5 and CEP-290-dependent assembly pathway for building a functional TZ . The eukaryotic cilium represents a functionally diverse organelle whose microtubule-based axoneme is templated from a modified centriole-termed basal body [1] . Motile cilia ( also called flagella ) propel cells or generate flow across cell surfaces , and their dysfunction results in primary ciliary dyskinesia [2] . Nonmotile or primary cilia are present in most metazoan cell types and enable sensory processes , including chemosensation/olfaction , mechanosensation , and photosensation [3 , 4] . In vertebrates , primary cilia are associated with several signalling pathways ( Hedgehog , Wnt , PDGF , cyclic nucleotide ) [5–8] . Since primary cilia have pervasive roles in signalling , disruption of their function is linked to numerous human disorders ( ciliopathies ) that affect sensory physiology ( vision , smell , hearing ) , as well as the development and function of most organs , including eyes , kidney , skeleton , and brain [4 , 9–12] . Several ciliopathies , such as Jeune asphyxiating thoracic dystrophy ( JATD ) and Bardet-Biedl syndrome ( BBS ) , result from defects in the evolutionarily conserved intraflagellar transport ( IFT ) machinery [13–16] . IFT is responsible for the formation and maintenance of cilia . It uses kinesin anterograde motor ( s ) to mobilise cargo ( e . g . , structural components , receptors ) from the basal body transition fibres to the ciliary tip , and a dynein retrograde motor to recycle components back to the base [17–19] . The IFT machinery comprises over 20 core components and a cargo-adaptor protein complex ( BBSome ) consisting of several BBS proteins [13 , 18] . Defects in core IFT and BBS proteins cause several multisystemic ailments , including cystic kidney disease , retinal degeneration , obesity , and skeletal malformations [20] . Another large macromolecular complex linked to a growing number of ciliopathies is the transition zone ( TZ ) . The TZ represents the proximal-most domain of the ciliary axoneme , found immediately distal to the basal body [21–24] . Its ultrastructure , as observed by transmission electron microscopy ( TEM ) , reveals doublet microtubules that are connected to the ciliary membrane , typically via Y-shaped structures [25] . These Y-links likely organise the so-called ciliary necklace , a repeating unit of membrane-associated proteinaceous “beads” that form rings or a spiral at the TZ membrane surface [21–23] . As detailed below , the Y-links and , presumably , the associated necklace appear to perform two principal functions: one in cilium formation , and the other as a ciliary “gate” that maintains the correct composition of the ciliary organelle . TZ-associated ciliopathies include Meckel syndrome ( MKS ) , Nephronophthisis ( NPHP ) , Joubert syndrome ( JBTS ) , Leber Congenital Amaurosis ( LCA ) , and Senior Løken syndrome ( SLNS ) [10 , 26] . Collectively , the clinical ailments of these ciliopathies partially overlap with those arising from IFT/BBS dysfunction and include brain malformations , cystic kidneys , retinal dystrophy , liver fibrosis , and polydactyly . Many genes encoding TZ-localised proteins are mutated in one or more ciliopathies; however , for most ciliopathies , not all causative genes have been identified , suggesting that additional TZ proteins remain to be discovered [26 , 27] . Genetic interaction studies of TZ genes in Caenorhabditis elegans have defined two major functional modules , termed “MKS module” and “NPHP module . ” The MKS module includes MKS-1 ( mammalian Mks1 orthologue ) , MKSR-1 ( B9d1/Mksr1 ) , MKSR-2 ( B9d2/Mksr2 ) , MKS-2 ( Mks2/Tmem216 ) , MKS-3 ( Mks3/Tmem67/Meckelin ) , MKS-6 ( Mks6/Cc2d2a ) , TMEM-231 ( Tmem231 ) , JBTS-14 ( Jbts14/Tmem237 ) , TMEM-107 , and TCTN-1 [25 , 28–32] . The C . elegans NPHP module thus far includes NPHP-1 and NPHP-4 [33] . Genetic ( synthetic ) interactions between the two modules are apparent . If one or more gene ( s ) within one module ( either MKS or NPHP ) are disrupted , ciliogenesis is essentially normal , although a subset of cilia are modestly truncated in nphp-4 mutants [29 , 34]; in contrast , deleting individual genes from both modules severely disrupts TZ/ciliary structures [28–30 , 35 , 36] . For example , the mks-6;nphp-4 double mutant has prominent ciliary phenotypes not observed in the individual single mutants [25 , 35] . These phenotypes include significantly fewer or no observable TZ Y-links , loss of basal body-transition fibre membrane attachments , and axonemal defects ( missing or shorter microtubules ) [25 , 28 , 30] . Hence , MKS and NPHP module proteins play genetically redundant roles in anchoring the basal body-ciliary axoneme and forming an ultrastructurally normal , functional TZ . Another established TZ protein , MKS-5 ( RPGRIP1L/MKS5 ) , is thought to act as a “scaffold” or “assembly factor” for most , if not all , MKS and NPHP module proteins [25 , 28 , 30 , 31] . Importantly , the genetically defined MKS and NPHP modules are consistent with physical interaction networks obtained through proteomic ( pull-down ) studies in mammalian cells , in which complexes containing the same ( orthologous ) MKS module proteins , or NPHP proteins , are observed [22 , 37–41] . In both C . elegans and mammalian cells , the modules can be further confirmed and defined hierarchically by testing for interdependent protein localisation at the TZ . Interestingly , mammalian Mks5 is found in a complex with Nphp1/Nphp4 proteins [39] , a functional association not yet observed in C . elegans . Also , in C . elegans and mammalian cells , the TZ is thought to form early during ciliogenesis , following the docking of the basal body to either a ciliary vesicle or the plasma membrane , where the incipient ciliary axoneme is subsequently extended in an IFT-dependent manner [13 , 21 , 25 , 42] . Early evidence that the TZ forms a membrane diffusion barrier which concentrates signalling machinery within cilia came from Musgrave and colleagues ( 1986 ) [43] , who presented cytological evidence suggesting that the unique composition of Chlamydomonas eugametos motile cilia is maintained by a membrane diffusion barrier . Spencer and colleagues ( 1988 ) [44] obtained similar evidence , showing that sequestration of rhodopsin to the outer segment of ciliary photoreceptors requires the TZ ( connecting cilium ) . Together , these studies hinted at a “gate” or “barrier , ” but until recently , the components involved and the mechanistic basis of this TZ functionality remained unknown . At least 14 evolutionarily conserved TZ-localised proteins are now implicated in the formation of a selective membrane diffusion barrier that precludes nonciliary proteins from entering and helps to compartmentalise ciliary proteins [21 , 22 , 45] . In C . elegans , two membrane-associated proteins , RPI-2 ( orthologue of Retinitis Pigmentosa 2 ) and TRAM-1 ( Translocating Chain-Associating Membrane Protein ) are present at the base of cilia and are excluded from the ciliary compartment unless TZ protein ( s ) are disrupted [25] . In mammals , several signalling proteins ( e . g . , Adcy3 , Arl13b , Inpp5e , Sstr3 , Pkd2 , and Smo ) either fail to enter cilia or are no longer concentrated within the organelle in various individual TZ mutants [31 , 38 , 40] . Similarly , the ciliary composition of cilia is altered in the TZ mutants of Chlamydomonas ( cep290 and nphp4 ) , Drosophila ( Cep290 ) , and ciliary photoreceptor ( Rpgr ) [46–49] . Although many TZ components have been identified and their role in establishing a ciliary gate is generally accepted , many questions regarding the TZ remain unanswered . Are there additional components of the TZ awaiting discovery , and do they fit within the known MKS or NPHP functional modules ? Aside from Mks5/Rpgrip1L , are there other essential “core” scaffolding and/or assembly factors of the MKS and/or NPHP modules at the TZ ? Finally , how are the different proteins and modules spatiotemporally assembled at the TZ ? In this study , we provide new insights into the composition , organisation , assembly , and function of the C . elegans TZ , as well as provide evidence for an expanded role for TZ proteins in ciliopathies . We identify C . elegans TMEM-218 ( mammalian Tmem218 ) as a novel TZ protein . Our genetic interaction , hierarchical , and in vivo functional analyses reveal that TMEM-218 is a new MKS module component . We show that , in addition to C . elegans MKS-5 ( Rpgrip1L/Rpgrip1 ) , another TZ protein , CEP-290 ( Cep290/Mks4/Nphp6 ) , functions as an assembly factor for not only TMEM-218 but all MKS module proteins tested . Consistent with a key role for CEP-290 in establishing TZ ultrastructure , the cep-290 mutant lacks all characteristic features of a TZ , including Y-links . Notably , removal of CEP-290 does not significantly perturb the localisation of NPHP module proteins or MKS-5 , whereas removal of MKS-5 results in CEP-290 delocalisation . Our findings suggest an assembly pathway that is initiated by MKS-5 and involves two separate branches . One branch requires CEP-290 for the assembly of MKS module proteins as well as two other TZ proteins ( TMEM-138 and Cyclin-Dependent Kinase-Like CDKL-1 ) that may form a separate module . The other branch of the pathway , which can assemble separately , involves the NPHP module . Finally , we present evidence that three human genes encoding CEP-290-dependent TZ components , TMEM17 , TMEM138 , and TMEM231 , are mutated in Oral-Facial-Digital type 6 ( OFD6 ) syndrome families; furthermore , a novel mammalian TZ protein we uncovered , Tmem80 , represents an excellent ciliopathy candidate . Notably , TMEM17 has not yet been associated with a human disorder , and patient-derived cells with the TMEM17 mutation display impaired ciliogenesis . Collectively , our work provides essential insights into the formation , organisation , and function of the TZ and expands the number of likely or potential ciliopathy-associated TZ proteins . Recently , a large-scale mutagenesis screen uncovered a knockout mouse strain with NPHP and retinal degeneration phenotypes reminiscent of SLSN ( OMIM 266900 ) , a known ciliopathy [50] . Although this strain was found to harbour a mutation in Tmem218 , the protein was not characterised at the molecular or cellular level . Tmem218 encodes a small protein with three transmembrane domains that is conserved across metazoans , including C . elegans and humans ( Fig 1A ) . Tmem218 is detectable in at least some , if not all , ciliated protists , including the green algae Chlamydomonas reinhardtii , the Chromalveolate Guillardia theta , and Choanoflagellates , the most closely related unicellular metazoan ancestors . To determine if the unstudied C . elegans orthologue of Tmem218 ( TMEM-218 ) plays a cilium-associated role , we generated an expression construct consisting of the endogenous promoter and entire coding region fused in-frame to green fluorescent protein ( GFP ) . Transgenic strains expressing this construct were observed by confocal microscopy to ascertain which cell types the gene is expressed in . We found that tmem-218 is expressed exclusively in cells that possess cilia , including amphid ( head ) and phasmid ( tail ) sensory neurons ( Fig 1B ) . This expression pattern is like that of most C . elegans ciliary genes , including those encoding IFT , BBS , and TZ proteins [3 , 15 , 25 , 36 , 51] . The GFP-tagged TMEM-218 protein is specifically concentrated at the base of cilia , immediately distal to the basal body-associated transition fibres ( Fig 1B; an IFT-dynein comarker , XBX-1::tdTomato , marks transition fibres and axoneme ) . This subcellular localisation for TMEM-218 is identical to known TZ proteins [25 , 28 , 31 , 35] . Next , we investigated whether TMEM-218 could be assigned to one of the two established genetic modules , MKS or NPHP [25] . Since a null mutant allele of tmem-218 was not available from knockout consortia or the Million Mutation Project ( MMP ) , we used transposon-mediated mutagenesis ( imprecise excision ) to generate one . Using this approach , we uncovered a deletion allele , tmem-218 ( nx114 ) , that removes the entire coding region of tmem-218 without affecting neighbouring genes ( Fig 1C ) . Mutant animals outcrossed to wild-type are viable and appear grossly normal in terms of morphology , movement , and development . To determine if tmem-218 displays genetic interactions with other TZ genes , we subjected single and double mutant strains to a dye-filling assay that tests for cilia structure defects [51–53] . Wild-type animals incubated with fluorescent DiI solution display dye-filling in several head ( amphid ) and tail ( phasmid ) sensory neurons , indicating normal exposure of cilia to the external environment ( see schematic , Fig 1D ) . As with all previously tested TZ mutants , the tmem-218 single mutant shows normal dye-filling ( Fig 1D ) . However , a double mutant with tmem-218 and an NPHP module mutant , nphp-4 , shows a prominent dye-filling phenotype ( Fig 1D ) . Such a genetic interaction is consistent with TMEM-218 acting together with NPHP-4 to facilitate TZ and cilium ultrastructure formation , a possibility that will need to be confirmed by TEM analysis . We also show that , as expected for an MKS module mutant [25] , combining the tmem-218 mutation with another MKS module mutant , mks-2 [28] , does not cause a dye-filling phenotype ( Fig 1D ) . Altogether , our findings that tmem-218 encodes a TZ protein and genetically interacts with nphp-4 but not mks-2 are consistent with TMEM-218 representing a novel MKS module protein . Our previous studies on C . elegans TZ proteins revealed that MKS-5 ( mammalian Rpgrip1L/Rpgrip1 ) plays a central role in assembling MKS module components at the TZ [25 , 28 , 30 , 31] . We therefore queried if MKS-5 is also required for the TZ localisation of TMEM-218 . To test this , we introduced the TMEM-218::GFP fusion protein into the mks-5 mutant . While TMEM-218::GFP is concentrated at the TZ in wild-type animals , it is consistently mislocalised ( absent from the TZ ) in the mks-5 mutant background ( Fig 2A ) . This finding supports the notion that TMEM-218 is functionally associated with the established network of TZ proteins , and confirms MKS-5 as a critical assembly factor for all known MKS module proteins tested thus far . To provide additional evidence that TMEM-218 is specifically associated with the MKS module , we assessed whether its TZ localisation is perturbed upon disruption of a“core” MKS module protein . Indeed , TMEM-218 is no longer present at the TZ in the mks-2 MKS module mutant ( Fig 2B ) . In contrast , and as expected for an MKS module protein , TMEM-218 remains correctly localised in the nphp-4 core NPHP module mutant ( Fig 2B ) . Reciprocal experiments were performed to query if TMEM-218 is itself required for the localisation of other TZ proteins . Our data show that TMEM-218 is not required for the localisation of MKS-5 , MKS-2 ( “core” MKS module protein ) , or NPHP-4 ( “core” NPHP module protein ) ( Fig 2C ) . Overall , our findings point to a specific association between TMEM-218 and MKS module components , potentially as a more “peripheral” TZ component compared to other “core” MKS module proteins , which include MKS-2 , MKSR-1 , MKSR-2 , and TMEM-231 ( Fig 2D ) [25 , 28 , 30 , 31] . Consistent with this possibility , these “core” TZ proteins are all necessary for TZ gate function—namely , restricting the inappropriate entry of membrane-associated TRAM-1a into cilia—but TMEM-218 , similar to the “peripheral” TZ protein MKS-3 [25] , does not appear to influence this aspect of TZ function ( Fig 2C ) . Given our discovery of a novel MKS module protein dependent on MKS-5 for TZ localisation , we wondered if there are additional proteins that participate in anchoring/assembling MKS module proteins at the TZ . We hypothesised that Cep290 , which is implicated in several ciliopathies ( including MKS and JBTS ) [54] and is suggested to be a structural component of the TZ in Chlamydomonas [48] , might represent such a protein . We sought to analyse the C . elegans homologue of Cep290 , encoded by the gene Y47G6A . 17 . To test our hypothesis , we created a construct encompassing the full-length cDNA of Y47G6A . 17 fused in-frame to GFP . As anticipated , the Y47G6A . 17::GFP translational fusion protein is specifically enriched at the TZ of all cilia ( Fig 3A ) ; importantly , we confirmed that the GFP-tagged protein is functional and , hence , its localisation is physiologically relevant ( see below ) . Based on Y47G6A . 17 consisting largely of coiled coils and having sequence homology to mammalian Cep290 , its TZ localisation , and our finding that it plays an important role in the assembly and function of the TZ ( see below ) , our data support the notion that it is indeed the functional homologue of Cep290; we therefore named the nematode protein CEP-290 . Notably , Schouteden and colleagues also recently confirmed that C . elegans CEP-290 localises to the TZ [55] . To provide evidence that C . elegans CEP-290 is implicated in ciliary gate function , we first acquired a strain ( VC30108 ) from the MMP [56] that contains a nonsense mutation ( Q638* ) in the coding region of cep-290 ( Y47G6A . 17 ) ( Fig 3B ) . Quantitative RT-PCR analyses show that the cep-290 transcript in this mutant is eliminated by nonsense-mediated mRNA decay ( NMD ) ( S1A Fig ) . The strain , cep-290 ( gk415029 ) , therefore very likely contains a null allele of cep-290 , consistent with its ciliary phenotypes being indistinguishable from those described for a different strain harboring a complete loss-of-function allele of cep-290 [55]; our strain was outcrossed to wild-type to remove unlinked background mutations . We then tested for the abnormal entry of two membrane-associated proteins into the cilia of cep-290 mutant animals . Both RPI-2 ( mammalian RP2 orthologue ) and TRAM-1a ( TRAM in mammals ) are normally excluded from cilia in wild-type animals ( Fig 3C ) [25] . In contrast , visualisation of fluorescently-tagged RPI-2 and TRAM-1a in the cep-290 mutant consistently reveals their abnormal “leaking” or accumulation in cilia ( Fig 3C ) . These findings parallel those found with several previous TZ mutants [25] and provide evidence that CEP-290 is required for the ciliary gating function of the TZ . Further support for this possibility stems from our observation of ARL-13 ( Arl13b in mammals ) , which freely diffuses within the ciliary membrane [45 , 57] . In wild-type animals , palmitoylated ARL-13 ( and other membrane-associated proteins ) are completely excluded from the TZ , a reflection of the TZ acting as a membrane diffusion barrier [30 , 45] . In the cep-290 mutant , the ARL-13 ciliary zone of exclusion ( CIZE ) at the TZ is substantially disrupted , with ARL-13 visible within the proximal region of the cilium and at the periciliary membrane ( Fig 3D ) ; this is similar to that observed for ARL-13 and other membrane-associated proteins in the mks-5 mutant or TZ double mutants [30 , 45] . Together , our two complementary TZ functional assays suggest that CEP-290 plays an essential role in maintaining a functional TZ “gate” or membrane diffusion barrier . To assess the potential role of CEP-290 in assembling TZ ultrastructure , which would help explain its role in TZ gate function , we subjected the cep-290 mutant to TEM analyses . In wild-type animals , the following structures are observed in TEM cross-sections of the amphid channel ciliary bundle: transition fibres that connect the distal end of the basal body to the base of the ciliary membrane , a TZ with prominent Y-links , a middle segment axoneme with doublet microtubules , and , finally , a distal segment with singlet microtubules ( Fig 3E ) . In the cep-290 mutant , most cilia are present throughout the pore , although one to two axonemes are missing from the distal regions , indicating that they are short ( Fig 3E ) . Like wild-type worms , cep-290 mutants also possess clearly distinguishable middle and distal segments . Strikingly , however , cep-290 mutant cilia reveal no structures characteristic of the TZ , displaying a lack of Y-link axoneme-to-membrane attachments; furthermore , the apical ring structure ( of unknown composition ) that is present inside the doublet microtubules is also absent ( Fig 3E; see also [55] ) . In addition , microtubule doublets are frequently disorganised near the ciliary base and in the periciliary membrane compartment of some neurons ( Fig 3E ) . Notably , the cep-290 null mutant TZ phenotype is essentially indistinguishable from that of the mks-5 null mutant [30] . Together , our findings reveal a critical role for C . elegans CEP-290 in the formation of TZ ultrastructure as well as in ciliary gate function . Having confirmed that CEP-290 is important for the formation of the TZ and function of the ciliary gate , we tested for its potential role in localising MKS module components ( including TMEM-218 ) as well as NPHP module proteins and MKS-5 at the TZ . We first observed that TMEM-218 is consistently absent from the TZ in the cep-290 mutant ( Fig 4A ) . Moreover , four additional MKSome components , namely , the “core” TZ proteins MKSR-1 ( mammalian B9d1 ) , MKS-2 ( Tmem216 ) , and TMEM-231 ( Tmem231 ) , as well as TMEM-17 ( Tmem17 ) , are all mislocalised in the cep-290 mutant ( Figs 3C and 4A ) . Notably , how the proteins are mislocalised may be different ( e . g . , MKS-2 “leaks'”into the axoneme , whereas TMEM-17 remains in the dendrite; Figs 3C and 4A ) ; the reason for this is unclear . In contrast , we find that MKS-5 remains largely within the proximal region of cep-290 mutant cilia in the area where the TZ would normally form ( Fig 4A ) . This suggests that while both CEP-290 and MKS-5 play critical roles in the assembly of MKS module proteins at the TZ , MKS-5 can localise independently of CEP-290 and , therefore , may function “upstream” of CEP-290 during TZ assembly . We note that the localisation of MKS-5 in the cep-290 mutant is not entirely wild-type , as it is in some cases more spread out and present within a more distal region of the axoneme ( Fig 4A ) . This suggests that CEP-290 may provide additional stability and/or specificity to MKS-5 localisation at the TZ . In addition , both NPHP module proteins ( NPHP-1 and NPHP-4 ) remain associated with the proximal region of the axoneme in the cep-290 mutant , in a manner that is very similar to MKS-5 ( Fig 4A; note that NPHP-1/NPHP-4 also localises to the basal body-transition fibres , as shown in Jensen et al . [30] ) . Hence , CEP-290 is required for the localisation of MKS module components but is not essential for the largely normal TZ localisation of the NPHP module or MKS-5 . As we elaborate in the discussion , this result is both striking and surprising: in the absence of CEP-290 , both MKS-5 and NPHP module proteins are present at the base of the cilium , in a region where the TZ is expected , even though no TZ ultrastructure ( Y-links ) is present ( Fig 3E ) . We performed reciprocal localisation experiments to ascertain if CEP-290 itself depends on other TZ proteins for its localisation . We show that CEP-290 is correctly localised in the tmem-218 mutant and the core MKS module TZ mutant , mks-2 ( Fig 4B ) . Similarly , CEP-290 localisation is not affected by disruption of the core NPHP module protein , NPHP-4 ( Fig 4B ) . In contrast , CEP-290 localisation to the TZ is lost when the core assembly factor , MKS-5 , is ablated ( Fig 4B ) . Specifically , the CEP-290::GFP signal is significantly reduced in the absence of MKS-5 and is often undetectable at the distal end of the dendrite despite the lower but not highly different cep-290 transcript levels between wild-type and mks-5 mutant ( approximately 50% lower ) ; notably , expression of cep-290 in the mks-2 mutant is slightly lower than in the mks-5 mutant , yet localisation of CEP-290 is unperturbed ( Figs 4B , 4D and S1B ) . This suggests that , in the absence of MKS-5 , the CEP-290 protein is not incorporated into to the TZ region , may not be transported to the dendritic tip , and/or may be degraded . Furthermore , of the remaining CEP-290 observable near the ciliary region , the signal is closer to the basal body-transition fibers compared to its TZ localisation in wild-type animals ( Fig 4C ) ; it is also significantly lower in intensity and occupies a smaller volume ( S1C Fig ) . Given the limiting resolving power of confocal microscopy , our results suggest that , in the absence of MKS-5 , residual CEP-290 can localise to the very distal end of the dendrite but is not incorporated within the proximal region of the cilium . To ascertain how MKS-5 might enable CEP-290 localisation to the TZ , we queried whether the coiled coil N-terminal domain of MKS-5 alone could confer this activity . Notably , the coiled coil region is both necessary and sufficient for MKS-5 localisation to the TZ , independent of its two C2 and RPGR-interacting domain regions [30] . We expressed the MKS-5 coiled coil region in the mks-5 mutant , and found that it fully rescues CEP-290 localisation at the TZ , similar to full-length MKS-5 ( Fig 4D ) . Furthermore , the MKS-5 coiled coil region restores the space ( volume ) normally occupied by CEP-290 at the TZ ( S1C Fig ) . We then sought to uncover if cep-290 is genetically associated with the MKS module using genetic interaction and dye-filling studies . The cep-290 mutant is able to dye-fill normally ( Fig 4E ) , as expected from having largely intact , environmentally-exposed ciliary axonemes ( Fig 3E ) . A synthetic dye-filling phenotype is observed when cep-290 is combined with the nphp-4 mutant , but not with the mks-2 mutant ( Fig 4E ) ; this behaviour mimics that of other MKS module genes ( e . g . , mksr-1 , mks-2 , and mks-6 ) . Rescue of the cep-290 mutant with our CEP-290::GFP translational reporter confirms that cep-290 is responsible for this phenotype and that the GFP-tagged protein is functional—making its TZ localisation biologically relevant ( S2 Fig ) . Genetically , cep-290 is , therefore , aligned with the MKS module of TZ genes ( Fig 4F ) . Together , our findings are consistent with a hierarchical organisation at the TZ , wherein MKS-5 is positioned at the very “base , ” influencing both NPHP and MKS module assembly , and CEP-290 is positioned between MKS-5 and MKS module components ( Fig 4F ) . In C . elegans , all TZ proteins tested fit genetically within the MKS or NPHP module , the exception being the central organising or assembly factor , MKS-5 , which shows interactions with both modules [25 , 28 , 30] . Here , we show that two proteins ( CDKL-1 and TMEM-138 ) that require MKS-5 for TZ localisation and do not behave as either MKS or NPHP module proteins depend on CEP-290 for their correct localisation . The C . elegans CDKL-1 kinase , homologous to mammalian Cdkl proteins ( Cdkl1/Cdkl2/Cdkl3/Cdkl4 ) , localises to the TZ and influences cilium length , similar to the recently described Chlamydomonas CDKL5 protein ( manuscript in preparation ) [58] . We find that CDKL-1 ( isoform A or C ) is still TZ-localised when core MKS or NPHP module proteins are removed ( Fig 5A ) . Furthermore , combining the cdkl-1 mutant with either the mks-2 mutant or the nphp-1 mutant does not cause a synthetic dye-filling defect ( Fig 5B ) . Hence , CDKL-1 cannot be assigned to either the MKS or NPHP module . Remarkably , however , CDKL-1 not only requires MKS-5 for its TZ localisation but also depends on CEP-290 ( Fig 5C ) . To determine if additional TZ proteins exhibit this behaviour , we examined TMEM-138 , the C . elegans orthologue of mammalian Tmem138 [59] . Similar to CDKL-1 , TMEM-138 requires MKS-5 for its localisation but does not depend on either core MKS or NPHP module proteins; furthermore , the tmem-138 mutant shows no synthetic dye-filling defect with mksr-1 or nphp-4 , making an assignment to either module not feasible [30] . Remarkably , TMEM-138 also depends on CEP-290 for its TZ localisation ( Fig 5C ) . Together , our findings suggest hierarchical/functional characteristics for CDKL-1 and TMEM-138 that differ from all other MKS module proteins but that share a dependence on both MKS-5 and CEP-290 for their TZ localisation . For the purpose of discussion below , we refer to CDKL-1 and TMEM-138 as CEP-290-associated proteins . We wondered if the two genes interact genetically , but the tmem-138;cdkl-1 double mutant does not show a synthetic ciliogenesis ( dye-filling ) phenotype ( Fig 5B ) . We further queried if there are genetic interaction ( s ) between cep-290 and tmem-138 and/or cdkl-1 . Only a small difference in dye-filling ( weaker uptake ) was found between the double mutants ( tmem-138;cep-290 and cdkl-1;cep-290 ) compared to the single mutants , consistent with the positioning of tmem-138 , cdkl-1 , and cep-290 in the same genetic module ( Fig 5D ) . The majority of known TZ proteins are associated with one or more ciliopathies , including MKS , JBTS , SLNS , and NPHP [21 , 22 , 26] . Our recent discovery that C . elegans TMEM-17 is an MKS module protein [30] that depends on CEP-290 for TZ localisation ( Fig 4A ) suggests that it may also be linked to one or more ciliopathies; however , such a possibility has not been reported . We therefore included TMEM17 in a panel of 101 ciliary genes ( S1 Table ) that underwent next-generation sequencing ( NGS ) in a cohort of 330 patients with a neuroradiologically confirmed diagnosis of JBTS and variable organ involvement ( S2 Table ) . Interestingly , we uncovered a homozygous missense mutation in TMEM17 ( p . N102K ) in two siblings whose clinical profile is consistent with OFD6 ( Figs 6A and S3A ) , a subtype of JBTS [60] . The TMEM17 variation fully segregates with the disease in the family ( Figs 6A and S3A ) and is not detected in 150 in-house controls or in over 73 , 000 samples from the public databases EXAC and EVS . Furthermore , the p . N102K variant is predicted to be pathogenic by four distinct bioinformatic predictors ( PolyPhen2 , SIFT , Mutation Assessor , Mutation Taster ) and affects a highly conserved amino acid residue ( S3B Fig ) . Importantly , we show that fibroblast cells isolated from a TMEM17-mutated sibling display a much reduced ability to form cilia compared to cells obtained from the healthy heterozygous mother ( Fig 6B and 6C ) . Together , the discovery of a homozygous mutation in a highly conserved residue of Tmem17 that is not present in control genomes and is correlated with a defect in ciliogenesis provides strong evidence that disruption of Tmem17 results in a ciliopathy , OFD6 . Notably , the C5orf42 gene was previously thought to be the main cause of OFD6 [61] , although this is now debated [62]; and a second gene related to TMEM17—TMEM216 ( MKS2 ) —has also been associated with OFD6 [63] . We therefore sought to identify additional genes that encode CEP-290-dependent TZ proteins that may be mutated in OFD6 patients . We identified individual families with missense mutations in two genes , TMEM138 ( homozygous p . M118L ) and TMEM231 ( compound heterozygous p . P179A/p . P219L ) ( Figs 6D , 6E and S4; S2 Table ) . All mutant alleles are predicted to be pathogenic in that they affect conserved residues , are not found in control cohorts or public databases , and fully segregate with the disease in the respective families ( Fig 6D and 6E ) . Collectively , our findings provide strong evidence that Tmem17 is a novel OFD6-associated protein that is necessary for ciliogenesis and suggest a potentially expanded spectrum of TZ-associated proteins ( Tmem138 , Tmem231 ) linked to this JBTS phenotype . Since not all genes linked to TZ-associated ciliopathies have been uncovered , we speculated that other TZ-associated ciliopathy proteins likely exist . Indeed , the Tmem17/Tmem216-related protein that emerged in tetrapods ( S5A Fig ) , Tmem80 , also represented an excellent , uncharacterised candidate . Using immunofluorescence analysis of wild-type and TZ mutant mouse embryonic fibroblasts , we show that mammalian Tmem80 localises to the TZ in a manner dependent on MKS module-associated proteins ( Cc2d2a/Mks6 , Tctn1 , and Tctn2 ) ( S5B Fig ) . Although we sequenced TMEM80 ( as well as TMEM218 ) in our patient cohort and did not find any pathogenic variants ( S1 Table ) , we propose that both genes represent excellent candidates for being associated with OFD6 or other related ciliopathies , including MKS and JBTS . One branch of the MKS-5-dependent pathway involves the assembly of the NPHP module , consisting of NPHP-1 and NPHP-4 ( Fig 7B ) . Our work reveals that its assembly at the TZ can occur independently of CEP-290 . Notably , when CEP-290 is disrupted , both NPHP-1 and NPHP-4 , as well as MKS-5 , colocalise within the proximal region of cilia , where the TZ is normally found ( Fig 4A ) . This represents evidence for a functional association between the NPHP-1/NPHP-4 and MKS-5 proteins , consistent with the discovery by Sang and colleagues that mammalian Nphp1/Nphp4 proteins are observed in a complex with Mks5 ( Nphp8/Rpgrip1L ) [39] . Our results also suggest the possibility that MKS-5 together with NPHP-1/NPHP-4 represent the first components to establish a region at the base of cilia that give rise to the canonical TZ . Indeed , this MKS-5-NPHP-1-NPHP-4 domain is established in the cep-290 mutant , which has no discernible TZ ultrastructure , including Y-links ( Fig 3E ) . Given these findings , we speculate that MKS-5 and the NPHP module might help nucleate the assembly of Y-links—whose structural makeup may include CEP-290 and , perhaps , other associated proteins—as well as establish the position and appropriate length of the TZ . Hence , in the cep-290 mutant , Y-links are absent , but the ciliary axoneme “foundation” for a TZ remains . Such an intriguing possibility will require further confirmation . The second branch of the MKS-5-dependent pathway first involves the assembly of CEP-290 ( Fig 7C , first step ) and then MKS module components , including “core” and “peripheral” TZ proteins , which are all predicted to be transmembrane proteins or have membrane associations via C2 or B9 domains ( Fig 7C ) . Notably , C . elegans CEP-290 is specifically required for MKS module protein assembly , as it is not needed for NPHP module protein assembly at the TZ ( Figs 3C and 4A ) . This is consistent with the finding that in Drosophila , NPHP module proteins are not present and Cep290 is required for the correct localisation of B9-domain MKS module proteins ( Mks1 , B9d1 , and B9d2 ) [47] . Pull-down studies in mammalian cells are similarly consistent with an association between Cep290 and various MKS module proteins [40 , 64] . Also consistent with the above , Nphp4 is TZ-localised in the Chlamydomonas Cep290 mutant [46] . Finally , we observe a genetic interaction between cep-290 and an NPHP module gene ( nphp-4 ) that is required for ciliogenesis , but no interaction between cep-290 and an MKS module gene ( mks-2 ) ( Fig 4E and 4F ) , precisely as with all known MKS module genes [25 , 28 , 31] . Given these findings , we propose that CEP-290 is specifically required for the assembly of the entire MKS module . Since MKS-1/MKSR-1/MKSR-2/MKS-2/TMEM-231 all depend on each other for their TZ localisation [25 , 28 , 36] , we propose that these are “core” TZ proteins that assemble “downstream” of CEP-290 ( Fig 7C , second step ) . This hierarchical definition implies that TMEM-17 , TMEM-67 ( MKS-3 ) , TMEM-218 , and TMEM-237 are more “peripheral , ” as they fail to assemble at the TZ when any “core” TZ protein is disrupted and do not themselves influence the localisation of the “core” TZ proteins ( this study and refs . [25 , 28 , 30] ) . Assembly of “peripheral” TZ proteins is likely to be “downstream” of “core” TZ proteins ( Fig 7C , third step ) . Together with these localisation studies , a genetic interaction between cep-290 and a core NPHP module mutant ( nphp-4 ) , but not a core MKS module mutant ( mks-2 ) , argues for a clear positioning of CEP-290 as the core component of the MKS module ( Fig 7C ) . Notably , our hierarchy for C . elegans CEP-290 differs from that of Schouteden and colleagues [65] , who propose that the TZ protein occupies an entirely different ( third ) module . Aside from MKS-5 , which is the core assembly factor , all C . elegans TZ-localised proteins analysed heretofore can be assigned to an MKS or NPHP module . Here , we show that CDKL-1 ( a novel TZ protein homologous to uncharacterised human CDKL family members ) and TMEM-138 ( whose human orthologue Tmem138 is mutated in JBTS [59] ) both require CEP-290 for their TZ localisation ( Fig 5C ) . Given our above-mentioned findings , this might suggest an association between CDKL-1/TMEM-138 and the MKS module . Intriguingly , however , neither protein is delocalised when core MKS module proteins are disrupted ( see Fig 5A and Jensen et al . [30] ) . Moreover , the tmem-138 and cdkl-1 genes do not behave like MKS module genes , in that their respective mutants do not show a synthetic ciliogenesis defect with NPHP genes ( Fig 5B and [30] ) . Altogether , these findings suggest that CEP-290 may itself organise two “branches” of an extended MKS module: one that harbours “canonical” MKS module proteins and another that consists of CDKL-1 , TMEM-138 , and , perhaps , additional TZ proteins ( Fig 7D ) . Interestingly , CDKL-1 does not appear to influence cilium gate function but does influence cilium length ( manuscript in preparation ) , the latter being similar to its CDKL5 homologues in Chlamydomonas [58 , 66] . Thus , CEP-290 may help organise a novel module that may be hierarchically and functionally different from the well-established NPHP and MKS modules . Interestingly , Lee and colleagues [59] provided some insights into possible behavioural differences between Tmem138 and Mks2 , including with respect to Cep290 . The Cdkl1/Cdkl2/Cdkl3/Cdkl4 family of kinases is largely unstudied , but a zebrafish knockdown study of Cdkl1 suggests ciliary-like phenotypes ( including abrogated Hedgehog signalling ) [67] . Having ascertained that CEP-290 plays a central role in the assembly of the MKS module , we confirmed that it is also essential for the formation of a functional ciliary gate . Specifically , loss of CEP-290 causes two proteins ( TRAM-1a and RPI-2 ) normally found at the periciliary membrane ( just outside of the cilium ) to accumulate abnormally within the organelle ( Fig 3C ) . This “leakiness” has been observed with most MKS and NPHP module mutants and is also consistent with changes to ciliary composition in the Chlamydomonas Cep290 mutant [48] . Furthermore , we show that ARL-13 enters the TZ and accumulates at the periciliary membrane in the cep-290 mutant ( Fig 3D ) , which represents additional evidence that CEP-290 helps form a functional TZ membrane diffusion barrier . Previous TEM analyses of C . elegans TZ gene mutants revealed that disruption of one or more MKS module component ( s ) , or NPHP module component ( s ) , does not abrogate the formation of Y-links ( with the exception of nphp-4 , which has modest defects [29] ) . However , various MKS-NPHP double mutants , such as mks-2;nphp-4 or mksr-1;nphp-4 , result in the apparently complete loss of Y-shaped structures [25 , 28 , 36] . Here we show that disruption of CEP-290 alone is sufficient to eliminate Y-links ( Fig 3E ) . This cep-290 phenotype is striking in that it influences to a great extent only the TZ region , wherein Y-links are absent . Yet , as discussed above , in the cep-290 mutant , the most “central” assembly factor , MKS-5 , together with the NPHP module , are present at the base of the cilium , in the same position expected for the TZ . Hence , removal of MKS-5 may cause a loss of Y-link structures [30] , potentially because CEP-290 itself cannot assemble at the TZ . Based on this reasoning , we suggest in our model for the assembly pathway of the TZ that Y-link structures are established at the step at which CEP-290 assembles at the TZ ( Fig 7C , first step ) . This inference is based on the fact that Y-link formation is not impaired when one or more “core” or “peripheral” MKS module proteins are disrupted , which likely results in all MKS module proteins being mislocalised [25 , 28 , 30 , 36] . Although the MKS module proteins are likely a part of the Y-links , they are all membrane-associated ( using transmembrane of C2/B9 domains ) and , therefore , may be more closely linked to the formation of the ciliary necklace ( intramembraneous “beads” [68] ) rather than the central/prominent region of the Y-shaped structure . We speculate that although Y-links are missing in the C . elegans TZ of the cep-290 mutant , similar to that found by Schouteden et al . [65] and in the Chlamydomonas Cep290 mutant [48] , structural proteins other than CEP-290 might make up these prominent Y-shaped structures . Virtually all genes encoding TZ proteins are linked to one or more ciliopathies , namely MKS , JBTS , SLNS , Leber Congenital Amaurosis , and NPHP [21 , 22 , 26 , 27] . Recently , disruption of Tmem218 in a mouse mutagenesis screen was found to cause retinal degeneration and NPHP phenotypes that resemble SLNS [50] . While the murine protein was not characterised , we show that the nematode orthologue ( TMEM-218 ) localises to the TZ , as with other ciliopathy-associated TZ proteins ( Fig 1B ) . Furthermore , TMEM-218 can be positioned genetically and hierarchically within the MKS module , likely as a peripheral component , similar to MKS-3 ( Tmem67 ) , JBTS-14 ( Tmem237 ) , and MKS-6 ( Cc2d2a ) , which are all ciliopathy-associated ( Figs 1D , 2D and 7 ) . In addition , we show for the first time that the unchracterised Tmem17/Tmem216-related mammalian protein , Tmem80 , localises to the TZ in ciliated mouse embryonic fibroblasts ( S5 Fig ) . Notably , Tmem80 depends on at least three known TZ-localised ciliopathy proteins ( Tctn1 , Tctn2 , and Mks6/Cc2d2a ) for its localisation , positioning it within a growing functional network of TZ proteins . Hence , both Tmem218 and Tmem80 represent prime candidates for being implicated in one or more of the above ciliopathies or OFD6 , as discussed below . Although sequencing of TMEM80 and TMEM218 in a large cohort of patients with JBTS-related phenotypes did not identify pathogenic mutations , the possibility that such mutations represent a rare cause of OFD6 cannot be excluded at present , especially since the number of OFD6 patients in our cohort is quite low due to the rarity of this condition ( 17 probands ) . Aside from Tmem218 and Tmem80 , one of the few TZ proteins that remains to be linked to one or more ciliopathies is Tmem17 . Here , we present evidence that a variation in an evolutionarily conserved residue of Tmem17 causes OFD6 syndrome . The mutation ( N102K ) , not present in control cohorts , is predicted to be deleterious and segregates specifically with affected siblings in the family ( Fig 6A ) . Importantly , we show that patient fibroblast cells with the pN102K mutation display a prominent ciliogenesis defect compared to control cells ( Fig 6B and 6C ) . Notably , many TZ proteins , including B9d1 , Tmem67/Mks3 , Tmem231 , Tmem237 , Cc2d2a/Mks6 , Cep290 , and Tectonics ( Tctn1 and Tctn2 ) are known to influence ciliogenesis in mammalian cells [28 , 38 , 40 , 69] , consistent with our findings . Although only present in single families , we also provide evidence that the human orthologues of two additional proteins that require CEP-290 for their TZ localisation in C . elegans , Tmem138 and Tmem231 , are also mutated in OFD6 ( Figs 6D , 6E and S4 ) . Uncovering additional families with mutations in TMEM17 , TMEM138 , and TMEM231 will be important to conclusively assign these genes to OFD6 . Finally , what TZ/ciliary defect may result in OFD6 rather than other related ciliopathies , including OFD3 , remains unclear . For example , Tmem231 is linked to OFD3 as well as MKS [31] . Understanding what components make up the TZ and how they assemble into different functional modules represent important goals for shedding light on how the TZ forms during ciliogenesis and acts as a membrane diffusion barrier . In addition to uncovering C . elegans TMEM-218 and mammalian Tmem80 as novel TZ proteins , our work illuminates the central roles of MKS-5 and CEP-290 in the assembly of NPHP and MKS modules , and of TZ ultrastructure ( Fig 7 ) . Aside from CEP-290 , which proteins constitute the Y-links remains an open question . How at least two TZ proteins ( CDKL-1 and TMEM-138 ) assemble in a CEP-290-dependent but MKS module-independent manner is unclear , but hints at the existence of a TZ module that is distinct from the MKS and NPHP modules . It will be interesting to confirm and further investigate in mammalian cells our overall model for TZ assembly ( Fig 7 ) as well as obtain evidence for the involvement of the novel TZ proteins we uncovered ( Tmem218 and Tmem80 ) in ciliopathies . The translational construct for the cep-290 ( Y47G6A . 17 ) gene was generated by fusing the bbs-8 promoter ( 941 bp ) to the cDNA of cep-290 and EGFP together with the unc-54 3’ UTR . The translational construct for tmem-218 ( T23E7 . 5 ) was generated by fusing its native promoter ( 2 , 092 bp upstream of the start codon ) and all exons and introns to EGFP together with the unc-54 3’ UTR . For cdkl-1 ( Y42A5A . 4 ) , we generated two constructs . One was made by fusing the bbs-8 promoter ( 941 bp ) to the genomic region of cdkl-1 ( isoform C ) and EGFP with the unc-54 3’ UTR . The other was generated by fusing its native promoter ( 1 , 869 bp ) and all exons and introns ( cdkl-1 isoform A ) to EGFP with the unc-54 3’ UTR . Transgenic lines were generated as reported previously [51] . All C . elegans strains used in this study ( S3 Table ) were maintained and cultured at 20°C . Those carrying mutations in C . elegans mks-5 ( tm3100 ) , tmem-231 ( tm5963 ) , cdkl-1 ( tm4182 ) , cep-290 ( gk415029 ) , tmem-138 ( tm5624 ) , nphp-1 ( ok500 ) , and nphp-4 ( tm925 ) were obtained from the C . elegans Gene Knockout Consortium or National Bioresource Project and outcrossed to wild-type ( N2 ) a minimum of five times . Standard mating procedures were used to introduce GFP-tagged protein constructs into different genetic backgrounds . Genotyping the various mutants was done by single-worm PCR . Confocal microscopy was used to assess the subcellular localisation of the various fluorescent reporters for both the wild-type and each TZ mutant strain , as indicated . A minimum of 50 animals for each strain were analysed for all reported mislocalisation phenotype . We obtained a cep-290 ( gk415029 ) allele , which has a G→A nonsense mutation at nucleotide 3989 in the cep-290 gene . The gk415029 allele is likely to be null , as it contains a stop codon 36 . 7% into the coding region . The mutagenesis protocol for generating a tmem-218 ( T23E7 . 5 ) allele was modified from Huang et al . ( 2011 ) using the ttTi20388 allele [70] , which contains a Mos1 insertion in the intron of tmem-218 . Sequencing of the tmem-218 ( nx114 ) allele revealed a 2721 bp deletion plus 4 bp insertion that removes the entire tmem-218 coding region . Before analysis , tmem-218 ( nx114 ) worms were outcrossed six times to wild type ( N2 ) . C . elegans were exposed to the lipophilic dye , DiI , to assay for uptake of dye through intact , environmentally exposed ciliary structures , as previously described [25] . Stained worms were imaged with fluorescence microscopy , and intensities were analysed with ImageJ . Young adult worms were fixed , sectioned , and imaged as described previously [71] , with the exception that worms were fixed overnight at 4°C in 2 . 5% gluteraldehyde , 1% paraformaldehyde in Sørensen's buffer . CEP-290::GFP localisation in wild-type or mks-5 mutant animals was assessed relative to the position of the basal body-transition fibres marked by the XBX-1::tdTomato IFT protein marker . Distance ( in μm ) from the peak fluorescence value of XBX-1 ( middle of the basal body-transition fibres ) to the first CEP-290 pixel above 0 . 5 ( relative intensity ) is defined here as the start of the TZ . The size of the TZ is the length ( in μm ) from the first pixel of CEP-290 fluorescence over 0 . 5 to the last pixel over 0 . 5 . All images were taken on a spinning-disc confocal of phasmid ( PHA and PHB ) cilia , as described above , with the same settings for the same exposure time ( 600 ms ) . CEP-290::GFP localisation volumes were detected by the Volocity software “find object” program with an automatic threshold offset of 200 . p-values were calculated by Dunn Kruskal-Wallis multiple comparison ( Holm-Sidak method ) . Anaesthetized L4 stage transgenic worms in 10 mM levamisole were prepared for spinning-disc confocal microscopy . Fluorescence in at least 20 pairs of cell bodies of PHA and PHB neurons from strains expressing TMEM-218::GFP , TMEM-231::GFP , or CDKL-1::GFP ( isoforms A and C ) was visualized and analyzed using Volocity software ( Perkin Elmer ) . To obtain relative fluorescence intensity for each strain , the background fluorescence signal was subtracted and the absolute fluorescence intensity measurements were divided by the median fluorescence intensity in wild type . The relative fluorescence intensity data were plotted using Box and Dot plots in R . The assessment of normal distribution of each dataset was done using the Shapiro–Wilk test . The p-value of the dataset that shows normal distribution was obtained by Tukey multiple comparisons of means of 95% family-wise confidence level . The p-value of the dataset that shows non-normal distribution was calculated using the Kruskal-Wallis test; Dunn Kruskal-Wallis multiple comparison ( Holm-Sidak method ) was used for comparing more than two groups . Results are presented in S1D–S1G Fig . Quantitative RT-PCR to measure CEP-290::GFP and cep-290 transcript levels was performed as previously described [68] using iTaq Universal SYBR green supermix ( Biorad ) , an Applied Biosystems StepOne real-time PCR system , SuperScript III reverse transcriptase ( Invitrogen ) , and the following primers: control , tba-1 forward: tcaacactgccatcgccgcc and reverse: tccaagcgagaccaggcttcag , GFP forward: atggtgttcaatgcttctcg and reverse: tgtagttcccgtcatctttga , and cep-290 forward: tgctcagcgagttgaatagg and reverse: aagcttcccaatttgctcat . The patient cohort included about 330 probands with JBTS , selected by the unique neuroimaging criterion of the molar tooth sign ( MTS ) , including 17 probands with the OFD6 phenotype . For each patient , a standardised clinical questionnaire completed by the referring clinician provided detailed information on the phenotypic spectrum and organ involvement . Written informed consent was obtained from all families , and the study was approved by the local ethics committee . All patients underwent simultaneous target sequencing of 100 genes ( including genes known to cause human ciliopathies and candidate genes derived from functional studies ) on a Solid 5500xL platform ( Life Technologies ) . Probes were designed to cover all exons , with splice-site junctions and at least 25 bp of flanking introns . Bioinformatic analyses were conducted to filter data and remove low-quality calls as well as variants with a frequency of >1% in public databases dbSNP ver . 141 ( http://www . ncbi . nlm . nih . gov/SNP/ ) and Exome Variant Server ( http://evs . gs . washington . edu/EVS/ ) or found in internal controls . The identified mutation in the TMEM17 gene ( NM_198276 ) was validated using bidirectional Sanger sequencing using the Big Dye Terminator chemistry ( Life Technologies ) and was searched against public databases dbSNP and Exome Variant Server . Potential pathogenicity was predicted using prediction software PolyPhen-2 ver . 2 . 2 . 2 ( http://genetics . bwh . harvard . edu/pph2/ ) and SIFT ( http://sift . jcvi . org/ ) . Nomenclature was assigned according to the Human Genome Variant Society ( http://www . hgvs . org/mutnomen/ ) . Sanger sequencing was also used to birectionally sequence all coding exons and exon–intron boundaries of TMEM218 in a subset of 160 Joubert patients who had tested negative at the previous NGS experiment . Probands from two large families diagnosed with OFD6 directly underwent whole exome sequencing . Three micrograms of genomic DNA per individual were subjected to whole-exome capture and sequencing using the SureSelect Human All Exon V5 kit ( Agilent ) . The resulting libraries were sequenced on a HiSeq 2000 ( Illumina ) as paired-end 76 bp reads . BAM files were aligned to a human genome reference sequence ( GRCh37/hg19 ) using BWA ( Burrows-Wheeler Aligner; v0 . 6 . 2 ) . All aligned read data were subject to the following steps: ( 1 ) duplicate paired-end reads were removed by Picard 1 . 77 , and ( 2 ) indel realignment and ( 3 ) base quality score recalibration were done on Genome Analysis Toolkit ( GATK; v2 . 1–10 ) . Variants with a quality score >30 and alignment quality score >20 were annotated with SeattleSeq SNP Annotation ( http://snp . gs . washington . edu/SeattleSeqAnnotation138/ ) . Rare variants present at a frequency above 1% in dbSNP 138 ( http://www . ncbi . nlm . nih . gov/SNP/ ) and the NHLBI GO Exome Sequencing Project or present from 312 exomes of unaffected individuals were excluded ( http://evs . gs . washington . edu/EVS/ ) . To improve exome analysis , data are crossreferenced with a list of known cilia-related genes from Ciliome Database , Cildb v2 . 1 ( http://cildb-archive . cgm . cnrs-gif . fr/ ) and transcriptomic study of cilia [72] . The analyses were focused on genes with homozygous variants in consanguineous families and with two heterozygous variants in other cases , prioritising ( 1 ) genes associated with human pathology in ClinVar ( http://www . ncbi . nlm . nih . gov/clinvar/ ) or HGMD databases ( http://www . hgmd . cf . ac . uk/ac/index . php ) , ( 2 ) cilia-related genes , and ( 3 ) other non cilia-related genes . Candidate variants and parental segregation were confirmed by DNA PCR/Sanger sequencing using targeted primers ( S5 and S6 Tables ) and following usual protocols . Fibroblasts from one patient homozygous for TMEM17 mutations and from his healthy mother ( heterozygous carrier ) were plated on coverslips and cultured in DMEM with 20% FBS until they were 80% confluent after being starved for 24 h in DMEM 0 . 1% FBS to allow cilia formation . Cells were fixed in cold methanol , and coverslips were rinsed and blocked in PBS with 10% BSA prior to incubation with antibodies . Fixed cells were incubated with the following antibodies: acetylated-α-tubulin ( 1:1000 , SIGMA ) , γ-tubulin ( 1:500 , SIGMA ) overnight followed by incubation with goat anti-mouse Alexa fluor 555 ( 1:3000 ) , and goat anti-rabbit Alexa fluor 488 ( 1:3000 ) . Nuclei were stained with Hoechst ( Invitrogen ) . Images were captured with a confocal microscopy ( C2 Confocal Microscopy System ) , by using the laser lines 488 nm ( green ) or the 561 nm ( red ) and a 60X 1 . 4 NA Plan Apo objective ( Nikon Corporation ) controlled by NIS Element AR 4 . 13 . 04 software . To count cilia , acetylated-α-tubulin and γ-tubulin positive cilia were manually counted within 15 images for each phenotype in at least three independent experiments . Variables were analysed by Student’s t test , and a value of p < 0 . 05 was deemed statistically significant . Values are expressed as standard error ( S . E . ) . Tctn1 , Tctn2 , Cc2d2a , Tmem67 , B9d1 , and Tmem231 mouse embryonic fibroblasts ( MEFs ) have already been described [31 , 40 , 73] . For Tmem80 immunofluorescence , cells were plated on glass coverslips , grown to confluency in DMEM+10%FBS , and starved for 48 h in OptiMEM . Cells were then fixed for 90 s at -20°C in freezer-cold methanol , incubated for 30 min at RT in block ( PBS+0 . 1%Triton-X100+2%BSA+1% donkey serum ) , and incubated overnight at 4°C in block containing these primary antibodies: mouse anti-Tmem80 ( Origene , TA501452 , dilution 1:200 ) , rabbit anti-detyrosinated tubulin ( Millipore , AB3201 , dilution 1:500 ) , and goat anti-γ-tubulin ( Santa Cruz Biotechnology , sc-7396 , dilution 1:200 ) . Coverslips were then rinsed thrice in PBS , incubated 1 h at RT with block containing DAPI and AlexaFluor-conjugated donkey secondary antibodies , rinsed twice more in PBS , mounted on slides using gelvatol , and visualised in a Leica TCS SPE confocal microscope .	The primary cilium is a structure found in most animal cell types . Much like an antenna , it is responsible for sensing extracellular signals , including light and small molecules , and conveying this information to the receiving cell and respective tissue or organ . At the base of the cilium is the transition zone ( TZ ) , which acts as a “gate” to regulate the entry and exit of ciliary proteins required for signal transduction . Here , we use the nematode Caenorhabditis elegans as a model system to dissect how different proteins within the TZ assemble to form a functional barrier . We find that the TZ protein MKS-5 ( Rpgrip1/Rpgrip1L orthologue ) forms the foundation for two different assembly pathways involving two distinct modules: Nephronophthisis ( NPHP ) and Meckel syndrome ( MKS ) . We show that at the base of the MKS module is CEP-290 , another TZ protein that assembles MKS module proteins , including a novel TZ protein we identify as TMEM-218 . CEP-290 also helps assemble a potentially separate submodule containing TMEM-138 and CDKL-1 . Notably , we provide evidence that the MKS module protein TMEM-17 facilitates cilium formation and is disrupted in the human disorder ( ciliopathy ) Oral-Facial-Digital Syndrome type 6 ( OFD6 ) . Together , our findings provide essential insights into the assembly pathway of the ciliary TZ and suggest further connections between the transition zone and human health .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "invertebrates", "medicine", "and", "health", "sciences", "caenorhabditis", "animals", "fibroblasts", "animal", "models", "membrane", "proteins", "caenorhabditis", "elegans", "model", "organisms", "connective", "tissue", "cells", "genetic", "interactions", "cellular", "st...	2016	MKS5 and CEP290 Dependent Assembly Pathway of the Ciliary Transition Zone
Most part of Southeast Asia is considered endemic for human-infecting Taenia tapeworms; Taenia solium , T . saginata , and T . asiatica . However , until now there was no report of the occurrence of human cases of T . asiatica in Lao PDR . This study , conducted in Savannakhet Province , Lao PDR , microscopically examined a total of 470 fecal samples by Kato Katz method and found 86% of people harboring at least one helminth . Hookworms were detected in 56% of the samples besides Opisthorchis like eggs ( 42% ) , Trichuris trichiura ( 27% ) , Ascaris spp . ( 14% ) , and Taenia spp . ( 4% ) eggs . Serology for cysticercosis showed 6 . 8% positives with results varying from 3% to 14 . 3% in Ethnic School students and Kalouk Kao village respectively . Species-specific PCR targeting mitochondrial DNA ( mtDNA ) of 28 tapeworms , recovered from 16 patients , revealed T . solium ( n = 2 ) , T . saginata ( n = 21 ) , and T . asiatica ( n = 5 ) . Two patients were confirmed to be coinfected with T . saginata and T . asiatica , indicating the endemicity of the 3 human Taenia in Lao PDR . However , nucleotide sequencing of a nuclear DNA gene , DNA polymerase delta ( pold ) revealed that all the tapeworms identified as T . asiatica using mtDNA had T . saginata type allele at pold locus , demonstrating that they are not “pure T . asiatica” but the hybrid descendants between the two species , confirming the wide distribution of hybrids of T . saginata/ T . asiatica in Southeast Asia . The high prevalence of several helminthic NTDs in east Savannakhet area even with conventional control measures indicates the importance to establish wide and multifaceted health programs to sustainably improve the quality of life of the populations living in these communities . Only human are the definitive hosts for Taenia solium , Taenia saginata , and Taenia asiatica , which are referred as the human-Taenia . The distribution of each of the 3 species of human Taenia depends on peoples’ cultural characteristics which involve the consumption of undercooked meat or organs of intermediate hosts infected with viable metacestodes [1–3] . Swine are the intermediate hosts for T . solium and T . asiatica . However , the metacestodes of these species present different tropism: usually muscle and brain for T . solium , and viscera , mainly liver , for T . asiatica [4–6] . Domestic bovine are the main intermediate hosts for T . saginata with cysticerci establishing predominantly in the muscles [7] . Southeast Asia is considered an endemic area for the 3 species of human Taenia with several reports of occurrence in human and animals [8–11] . However , there is no report of the occurrence nor evidence of T . asiatica in human in Lao PDR despite its localization , surrounded by endemic countries [12–14] . Antibody serosurveillance of four provinces in the northern area of Lao PDR in 2011 indicated high frequency of contacts with adult ( 46 . 7% ) and larval parasites ( 66 . 7% ) [14] . The existence of T . solium was confirmed by DNA sequencing of copro-PCR positive fecal samples , but no T . saginata or T . asiatica were detected [14] . Furthermore , in a recent study , 15 haplotypes of T . saginata were obtained from 30 isolates from Khammouane Province , central Lao PDR [15] . An extensive study on the prevalence of taeniasis in Lao PDR with whole country coverage reported the presence of mainly T . saginata found in all Lao PDR’s provinces and T . solium in Luang Prabang , northern area [10] . In this study , we report a high prevalence area for foodborne parasites and STHs . Furthermore , we could detect worm carriers of T . solium , T . saginata and T . asiatica by mitochondrial DNA in east Savannakhet Province , suggesting that Lao PDR as an endemic country for the 3 human-Taenia species . Moreover , we verified hybridization of T . saginata and T . asiatica is likely to be occurring in the region . The study was conducted in Sepon District , Lao PDR in March and December 2013 . The area is located in the eastern part of Savannakhet province and is bordering with Quang Tri province of Central Vietnam ( Fig 1 ) and it is covered by subtropical forests in its majority . Ancient human occupation is reported in the actual area of Lao PDR , and according to the last classification , there are 49 different ethnical groups in the country , with more than 4 groups living in Savannakhet area [16] . The participants joined the study on a voluntary basis , from an estimated population reported by the Basic Health Center of Sepon district of 743 people in the study area , 396 ( 53% ) males and 347 ( 47% ) females aged from 3 to 74 years old , and living in the 3 studied villages ( Kalouk Kao , Poung , and Ayay Yay ) besides the residents of an ethnic college located in Sepon district area , at the eastern Savannakhet province ( Fig 1 ) . A detailed explanation of the study was done in the local language for proper understanding . Adult subjects provided written informed consent , and a parent or guardian of any participant child provided informed consent on the child’s behalf and , after approving the informed consent , the participants received instructions for collecting and transporting the fecal samples . Fecal examinations were conducted at the Sepon District Hospital by Kato-Katz modified cellophane thick smear method ( KK ) [17] . Each slide was examined under a microscope , helminth eggs were counted , and the number of eggs per gram of feces ( EPG ) was calculated as previously described [18] . Blood sampling was conducted from 235 persons for serological diagnosis of cysticercosis . Feces and serum samples were then brought back to the laboratory in Thailand and Japan for further analysis . Fecal samples were submitted to KK and molecular procedures . Samples for copro PCR were added sufficient volume of RNAlater stabilization solution ( Life Technologies , USA ) and brought to the laboratory for Copro PCR analysis . Copro DNA technique was done as previously established [19] with some modification , briefly; fecal samples were homogenized , the volume estimated and 300mg of each sample was used for DNA extraction . The fecal material was disrupted with a μT-12 beads crusher ( TAITEC Co . , Koshigaya , Japan ) using 3 stainless steel beads of 4mm plus 200mg of 0 . 2mm glass beads in each tube . DNA was extracted from the homogenized solution using the QIAamp DNA Stool Mini Kit ( QIAGEN , Hilden , Germany ) following the manufacturer’s instructions . Final DNA elution was done in 30 μl of elution buffer . PCR was conducted using a T100 Thermal Cycler DNA thermocycler ( BIO-RAD , Hercules , CA , USA ) . The reaction was carried out in a final volume of 25 μl containing PCR reagent ( TOYOBO , Osaka , Japan ) and 1 `l of DNA preparation as template . The DNA samples were initially denatured at 94°C for 4 min , followed by 30 amplification cycles of denaturation at 94°C for 1 min , annealing at 60°C for 30 s , and elongation at 72°C for 2 min . PCR products were electrophoresed on a 2 . 0% agarose gel with positive samples showing amplicons of the proper size for each Taenia species [19] . The participants presenting Taenia eggs on KK or voluntarily looked-for treatment were given a single oral dose of niclosamide ( Yomesan , Bayer AG , Germany ) according to the fabricant recommendations , followed by purgation with 60 ml of saturated magnesium sulfate solution . After treatment , attention for worm expulsion was dispensed for all participants , including the people who showed the Taenia eggs in fecal examination or the people who requested treatment even with no positive results in KK exams [20] . Discharged parasites recognizable by naked eye were separated , identified , washed in saline solution ( NaCl 0 . 85% w/v ) and preserved individually in 70% ethanol or RNAlater stabilization solution ( Life Technologies , USA ) . The samples were transferred to a laboratory in Japan for further DNA analysis . The genomic DNA of each ethanol preserved tapeworm was extracted using QIAamp DNA Mini Kit ( Qiagen ) and subsequently used as a template for polymerase chain reaction ( PCR ) . For the differentiation of three human Taenia species , multiplex PCR was performed as previously described [19] , with a slight modification . Briefly , one reverse and four forward primers were used to amplify different sizes of amplicons , specific for the cytochrome c oxidase subunit I ( cox1 ) gene sequences of T . solium Asian genotype , T . solium Afro-American genotype , T . saginata and T . asiatica , respectively . The forward primer specific to T . asiatica was newly designed as TasiMpF ( 5’- TTA TTT ATT TAC GTC AAT CTT ATT G -3’ ) , instead of the originally used primer . For some of the tapeworms identified as T . saginata or T . asiatica , nucleotide sequencing of nuclear DNA gene markers was performed to examine whether they are the hybrid descendants of the two Taenia species . Partial sequences of two nuclear genes , ezrin-radixin-moesin-like protein ( elp ) and DNA polymerase delta ( pold ) , were amplified by PCR and directly sequenced [21 , 22] . In the case of double peaks in the sequencing electropherogram , PCR products were cloned using pGEM-T vector ( Promega ) transformed into Escherichia coli DH5α and plated on LB agar containing X-Gal ( 20mg/ml ) and ampicillin ( 100ug/ml ) . At least 10 positive clones from each PCR product were used for nucleotide sequence confirmation . ELISA and immunoblotting using LMWAgs were performed as previously described [23] . Briefly , for ELISA , 100 μl of 1 μg/ml of T . solium LMWAgs in PBS were loaded in 96-well microplates ( Maxisorp , Nunc , Copenhagen ) overnight at 4°C , blocked with 300 μl of blocking buffer ( 20 mM Tris–HCl , pH 7 . 6 , 150 mM NaCl , 1 . 0% casein , 0 . 1% Tween 20 ) at 37°C for 1 hour and washed twice with PBS containing 0 . 1% Tween 20 ( PBST ) . Serum samples diluted 1:100 in blocking buffer , were applied ( 100 μl/well ) in duplicates and incubated at 37°C for 1 h . After washing twice in PBST , the plates were incubated with 100 μl/well of recombinant protein G conjugated with peroxidase ( Invitrogen , USA ) diluted 1:2000 in blocking buffer at 37°C for 1 h . For color development , the plates were incubated with 100 μl of peroxidase substrate ABTS ( 2 , 2’-azino-di ( 3-ethyl-benzothiazoline-6-sulfonate ) ) ( Sigma-Aldrich ) in 0 . 2 M citric acid buffer ( pH 4 . 7 ) for 30 min at room temperature . Optical density ( OD ) was determined at 405 nm for each well using microplate reader ( Immuno Mini NJ-2300 , Biotec , Japan ) . The cut-off value was determined as the mean of OD plus 4 standard deviations of sera from 37 healthy donors . Immunoblot was used for confirmation of the ELISA positive serum samples and serum samples with OD value close to the cut-off value . Briefly , LMWAgs ( 60 μg/mini gel ) in SDS sample buffer ( 62 . 5 mM Tris-HCl , pH 6 . 8 , 2 . 0% SDS , 50 mM dithiothreitol and 10 . 0% glycerol ) were loaded in a 15 . 0% polyacrylamide gel . The separated proteins were transferred onto a polyvinylidene difluoride ( PVDF ) membrane sheet ( Millipore ) and then blocked with blocking buffer . Serum samples diluted 1:20 in blocking buffer were incubated at room temperature for 1 h . After washing 3 times for 5 minutes in blocking buffer , the membranes were incubated with the peroxidase-conjugated recombinant protein G ( Invitrogen , USA ) diluted 1:2000 in blocking buffer . The substrate ( 4-Chloro-1-Naphthol/Phosphate ) was used for color development at room temperature for 30 min , and the reaction was stopped by washing with water . All the data collected in paper forms were tabulated and the analyzed in Excel 2016 ( Microsoft ) and EpiInfo version 7 . 1 . 5 . 0 ( Centers for Diseases Control and Prevention , Atlanta , GA , USA ) . DNA sequencing data were edited and analyzed using MEGA 6 software [24] . In this cross-sectional study , the sample size was determined at a confidence level of 95% with confidence limits of 5% considering the entire population , gender and age calculated with the expected frequencies of cysticercosis and taeniasis . The data were analyzed using descriptive statistics and chi-square test to determine association between the results of the tests used in each category of data . P values <0 . 05 were considered statistically significant . This study was approved by the National Ethical Committee for Lao Health Research of the Ministry of Health , Lao PDR ( 172/NECHR ) . Stool samples analysis by KK were performed on 470 fecal samples . Analyses of the results revealed 86% of people harboring at least one species of parasite: 42% ( 196/470 ) , 31% ( 146/470 ) , 11% ( 50/470 ) and 2% ( 11/470 ) harbored 1 , 2 , 3 and 4 helminth species , respectively . The prevalences of the detected helminth eggs were 56% ( 265/470 ) for hookworms , 42% ( 199/470 ) for Opisthorchis like eggs , 27% ( 129/470 ) for Trichuris trichiura , 14% ( 66/470 ) for Ascaris spp . , and 4% ( 19/470 ) for Taenia spp . Copro-PCR to detect human Taenia DNA were performed in 163 fecal samples revealing 9 . 8% ( 16/163 ) of T . saginata , 3 . 1% ( 5/163 ) of T . solium and 1 . 8% ( 3/163 ) T . asiatica , indicating the endemicity of the 3 human Taenia in the Sepon district . Eighteen people received treatment with niclosamide , a total of 28 tapeworms were recovered from 16 taeniasis patients in 4 villages and the ethnic school . Two worms were identified as T . solium and 26 worms appeared to be T . saginata or T . asiatica morphologically . All the specimens were submitted to DNA examination . People receiving treatment for taeniasis and seropositives for cysticercosis by village , tapeworm expulsion and the results of the different diagnostic tests done are shown in Table 1 . Firstly , using multiplex PCR targeting mtDNA cox1 gene , based on mtDNA sequences , 75% ( 21/28 ) of the tapeworms were identified as T . saginata , 18% ( 5/28 ) were T . asiatica and 7% ( 2/28 ) T . solium . Two patients from Kalouk Kao village harbored multiple tapeworms . One patient had a mixed infection with three tapeworms ( two T . saginata and one T . asiatica ) . The other had two worms ( one T . saginata and one T . asiatica ) . T . asiatica was confirmed only in Kalouk Kao village , and T . solium was found in Kalouk Kao village ( n = 1 ) and in the Ethnic School ( n = 1 ) . Nuclear DNA analysis was carried out to clarify whether the tapeworms identified as T . asiatica by mtDNA were pure T . asiatica or hybrid descendants between T . saginata and T . asiatica . The tapeworms identified as T . saginata or T . asiatica in Kalouk Kao ( n = 9 ) were used for nuclear DNA study . Partial sequences of elp and pold genes ( 1164 bp and 1097 bp ) were obtained from all except three samples by direct sequencing . After cloning and sequencing , two different nucleotide sequences were obtained from each of those three samples . In the pold locus , all the alleles obtained from nine tapeworms were T . saginata types ( poldA or poldB ) as previously found [22] , and one worm was heterozygous of pold A/B alleles ( Table 2 ) . On the other hand , both T . asiatica type ( elpA ) and T . saginata type ( elpC ) alleles [21 , 25] were confirmed in elp locus . The genotype of elp locus mostly corresponded to the species identification by mtDNA , however , two worms with T . saginata type mtDNA were heterozygous of elp A/C alleles ( Table 2 ) . Serum samples collected totalized 235 samples , corresponding to 50% of the fecal samples , once not all the individuals that provided fecal samples accepted to provide blood samples . Results of ELISA confirmed by immunoblot showed the presence of positive cases of cysticercosis in all the villages with a prevalence of 7 . 2% ( 17/235 ) as shown in Fig 2 . Considering the results by village , 2 locations , Kalouk Kao and Poung presented higher prevalence with 14 . 3% ( 7/49 ) and 10 . 7% ( 6/56 ) respectively . The serum samples from Ayay Yay village and the Ethnic School students presented lower seropositives to cysticercosis when compared to the other villages ( p<0 . 05 ) with 2 cysticercosis positives samples each , corresponding to 3 . 1% ( 2/64 ) and 3 . 0% ( 2/66 ) of the studied people respectively . This is the first report of T . asiatica in Lao PDR , a country with considerable differences in latitude from the south at 13oN in Champassack to Phongsali ( Lat . 22oN ) the northern province , presenting great climatic diversity . The geographic particularities of the country , as the Boulavean plateau and the Mekong basin , created a basis for the development of different ethnicities , with many cultures and eating habits . A total of 49 ethnic groups have been recognized in Lao PDR [16] , consequently , as one could expect , parasitic infection prevalence patterns also might differ according to different areas of the country and may explain some parasitic infection particularities as we observed in Sepon , now considered an endemic area for the three human Taenia . The absence of restrictions on food consumption can contribute to the parasite infection in those communities . Additionally , we detected rates as high as 86% of the studied population harboring at least one species of helminth . In general , there is a lack of sanitary infrastructure , toilet access and other issues like no schooling and basic hygiene , factors that would be considered causes for the high infection rates of parasitic infection , as pointed out in Lao rural communities [11] . The villages of Poung and Kalouk Kao , where higher prevalence for all helminths was recorded , presented no structures for sewage treatment and the lowest education level according to the data provided by the Sepon Health Center . The lower prevalence of cysticercosis was found within the students , and it can be due to the access to education and basic infrastructure , as toilet , positive points on the protection for infection as previously reported [11] , indicating that basic and general hygiene practices could diminish the infection levels in endemic areas . Another issue that could interfere with the knowledge or understanding of the importance of hygiene habits is the adhesion of the target population to health promotion initiatives , perceived when only 49 of 129 people listed collaborated with this study in Kalouk Kao village . That is , the village with less sanitary conditions presented the lowest adhesion to the health project , as we verified with only 38% of people joining the study . Bringing the interest of more people for health actions is an important issue to be improved in health programs for better results . The occurrence of cysticercosis in the studied area was detected by serology without additional supportive data . Despite no clinical symptoms of neurocysticercosis were reported by the health center staffs or the villagers , suggesting there might be non-clinical cysticercosis cases , follow-up studies for the confirmation of such cases is an urgent task for early evaluation and treatment . In the present study , all the three human Taenia were confirmed from taeniasis patients by the conventional PCR method targeting mtDNA . However , all the tapeworms identified as T . asiatica had T . saginata type pold allele , indicating that they are not “pure T . asiatica” . It has been shown that most of the tapeworms identified as T . asiatica using mtDNA have genetic traces of T . saginata in some nuclear DNA loci , and possible “pure T . asiatica” has been confirmed only in Taiwan and Philippines until now [21 , 22 , 25] . Those tapeworms showing nuclear-mitochondrial discordance are considered to be derived from the hybrid descendants between “pure T . asiatica” and “pure T . saginata” . Although the infection of T . asiatica in humans has been associated with eating raw or undercooked pork viscera , it is uncertain which animals , cattle and/or pigs , are the intermediate hosts of the hybrids . To clarify the host affinity and tissue tropism of T . saginata , T . asiatica , and their hybrids , it is necessary to examine the cysticercus from domestic animals by using both mtDNA and nuclear gene markers . The reasons for the endemicity of the hybrids between T . asiatica and T . saginata in this region could be the proximity and the commuting of people and goods from Vietnam , a reported endemic area for T . asiatica with several human cases [26 , 27] . Ethnic overlapping occurs on all borders of Lao PDR with neighboring countries especially the Austro-Asiatic groups on both sides of the Laos-Vietnam border as well as the cultural behavior of ethnical communities which are overlapping in this Laos-Vietnam border region [16] with habits of eating pork liver in several dishes . Moreover , in the west area of Savannakhet province which is bordering with Thailand , also a known endemic country for T . asiatica [9] , there are reports for T . saginata and T . solium only [10 , 28] reinforcing the idea of the Laotian T . asiatica or its hybrids origin from Vietnam , though more studies in this issue are necessary . The causes of hybridization may include people harboring different tapeworm’s species as observed in this ( Table 1 ) and other studies [21 , 22 , 25] . Multi-species parasitism may facilitate the exchange of genetic material allowing the occurrence of hybrids . However , the consequences of this hybridization as the seriousness of disease in people and even if the hybrid descendants can produce viable generations are unknown . For further studies , to detect hybrid cysticerci in intermediate hosts , confirm hybridization , and its infectivity to intermediate hosts is necessary . MDA is the recommended strategy of the World Health Organisation to control or eliminate NTDs in endemic areas . It has been implemented widely in Southeast Asian countries and its success is related to the improvement of sanitation and education programs [29] . The Sepon region is included in the MDA programs for elimination of parasitic diseases with treatment using praziquantel . The detection of seropositive individuals for cysticercosis leads to a point of concern in this issue: the occurrence of seizures and other collateral effects after treatment due to the inflammation caused by the sudden damage or death of cysts in the brain [30–33] . Side symptoms after treatment , including seizures , were reported by the population submitted to praziquantel MDA in Lao PDR , leading the people to reject subsequent treatments and the stop of the program . Nowadays only children under 5 years of age will take mebendazole when going for vaccination at primary school under a WHO project ( Dr . Pongvongsa personal communication ) . The strict calculation of doses to be administrated should be considered in prevalent areas for cysticercosis , furthermore , a program for early diagnosis of cysticercosis in asymptomatic patients could be designed in these areas for improvement of MDA as accidental death after treatment may occur due to the extensive use of praziquantel in MDA of Asian countries [34] . Another issue in this endemic area for 3 species of human Taenia is to detect worm carriers . As we could observe in the results of worm recovering after treatment , the number of worm carriers would be higher if we combine the results of KK , Copro-PCR , and self-detection ( Table 1 ) . Unfortunately , Copro-PCR was not done in the field , so its results were not used for treatment protocols . Therefore , copro-PCR is a feasible option to be done in province’s central laboratories/health centers , where a safe treatment can be prescribed and conducted . In this study , we could note self-diagnosis as an educational alternative for detection of worm carriers; after the explanation of the study and the description of the fecal aspects and symptoms , 4 individuals who could not supply fecal samples , voluntarily came to expel worms , with 100% of worm expulsion ( Table 1 ) . This could be an excellent method for detection of worm carriers with a high rate of success as reported in a Mexican endemic area for T . solium [35] . Differently from the other studies in the central , north and northeastern areas of Lao PDR , where cases of T . asiatica were not found [12 , 14 , 15] , in the east part of Savannakhet province ( Fig 1 ) we could detect the presence of 3 species of human Taenia species in Sepon district that border Vietnam . The prevalence of Taenia eggs in KK was not high as presented by Okello et al . [14] which detected percentages as high as 46% . This result raises suspicion that other "hotspots" of T . solium hyper endemicity may exist in some regions , particularly in communities where the consumption of raw or undercooked pork is common , associated with the lack of health education . Regarding other helminths observed in this study , the high prevalence of Opisthorchis like eggs ( 43% ) is also a point of concern once Opisthorchis viverrini infection is a major cause of cholangiocarcinoma in endemic areas [36] . O . viverrini is a fishborne fluke and , like Taenia and its infection to humans is related to the consumption of raw fish meat ( cyprinid fish ) [37] and to domestic dogs , natural reservoirs of O . viverrini , which can be an important source of aquatic environment contamination because of its routine behavior to assess water sources [38] . Moreover , a high prevalence of hookworm ( 57% ) was found , a parasitic disease where dogs play an important role [38 , 39] . Dog meat is eaten in Asian countries including Lao PDR . However , considering the dog as an intermediate host of T . solium [40] , dogs survey for cysticercosis in addition to pigs’ survey may also be important to screen the risk factor for human infections . More studies on the ecological aspects of NTDs , as carried out in other localities including checking reservoir animals and using environmental DNA detection [38 , 41] , would be an interesting way to determine the level of exposure of the people living in endemic areas to agents causing diseases like O . viverrini , hookworms and other STHs found in this study as Ascaris and Trichuris . This study revealed a highly endemic area for helminthic diseases in east Savannakhet , Lao PDR including the high occurrence of STHs and foodborne trematodes . The existence of T . solium in Savannakhet province was confirmed in this study , moreover hybrids descendants between T . saginata and T . asiatica were detected , indicating the presence of 3 human-Taenia species in Lao PDR . The situation points out the importance of establishing new , wide and multifaceted health program to sustainably improve the quality of life of the populations living in these communities .	Southeast Asian Countries are endemic for several foodborne and soil-transmitted helminths occurring in different levels and areas , depending on environmental and cultural conditions . This study aimed to study the soil-transmitted helminths ( STHs ) and foodborne parasites in Savannakhet Province of Lao PDR , bordering with Vietnam . We found people infected with hookworms , roundworms , whipworms , intestinal/liver flukes , and tapeworms . We also detected antibodies against cysticercosis , an infection caused by eating the eggs of Taenia solium , the pork tapeworm . Focusing on human tapeworm infection , using molecular techniques based on mitochondrial DNA , we detected the three species of human tapeworms T . solium , T . saginata and T . asiatica . Interestingly , when we checked the same material using nuclear gene markers we noted that T . asiatica found in the region were the hybrid descendants of T . saginata and T . asiatica . The causes of hybridization may include people harboring different tapeworm’s species at the same time , allowing the exchange of genetic material but , the consequences of this hybridization are unknown including the seriousness of disease in people . Despite the deworming programs , there is a high prevalence of STHs and foodborne parasites in east Savannakhet area , therefore wide and sustainable health programs are an urgent task to improve the quality of life of the people living in the area .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "medicine", "and", "health", "sciences", "lao", "people", "education", "mitochondrial", "dna", "sociology", "tropical", "diseases", "social", "sciences", "parasitic", "diseases", "ethnicities", "forms", "of", "dna", "neglected", "tropical", "diseases", "molecular", "b...	2018	Taenia solium, Taenia saginata, Taenia asiatica, their hybrids and other helminthic infections occurring in a neglected tropical diseases' highly endemic area in Lao PDR
Post Kala-azar Dermal Leishmaniasis ( PKDL ) is a sequel of Visceral Leishmaniasis ( VL ) . The patients act as a reservoir for the causative parasite ( i . e . Leishmania donovani ) and thus should be diagnosed and treated with the utmost urgency to prevent the transmission of the disease . In this study , we tried to report the first instances of corneal complications supposedly associated with Miltefosine ( MF ) , in PKDL patients and the probable pathophysiology of such events . The recently rejuvenated National Kala-azar Elimination Program in Bangladesh has put great emphasis on monitoring all the leishmaniasis patients to investigate possible adverse drug reactions ( ADR ) . A total of 194 patients have received Miltefosine for the treatment of Post Kala-azar Dermal Leishmaniasis . So far five patients were found to have developed unilateral ophthalmic complications during the periods from May 2016 to October 2017 , after being treated with MF for PKDL . Unfortunately , one of whom had to go through complete evisceration of the affected eyeball . Despite the fact that MF is the only oral formulation of choice to treat PKDL , occurrences of such unexpected ADRs after MF administration urges the exploration of the pathogenesis of such incidents and determine measures to avert such occurrences from happening in future . Post Kala-azar Dermal Leishmaniasis ( PKDL ) is a skin manifestation which appears as a macular , papular , or nodular rash or a combination of the three , usually starting from the face but may start from any other part of the body , typically after being treated for VL [1 , 2] . Bangladesh is now in the consolidation phase for VL elimination after reaching the target of less than 1 cases /10000 population in all the endemic sub districts . As a result policymakers in Bangladesh are now keen to sustain this achievement . As PKDL cases can act as a source of infection , its early diagnosis and treatment is now a major concern [3] . Currently , Miltefosine ( MF ) , amphotericin B deoxycholate , and liposomal amphotericin B ( LAmB ) are recommended by the World Health Organization as treatment options for PKDL [4] . MF or hexadecylphosphocholine ( HePC , C21H46NO4P ) is an amphiphilic cationic phospholipid analogue comprising of two parts , a hydrophobic chain of 16 carbons and a polar head group of phosphocholine [5 , 6] . The drug was initially developed as an anti-tumor agent , particularly for the treatment of cutaneous metastases of breast carcinoma , however , it is more commonly used as an anti-leishmanial drug , and preferably to treat PKDL cases [7] . As an immunomodulator , MF stimulates T-cells , macrophages and the expression of interleukin 3 ( IL-3 ) , granulocyte-macrophage colony stimulating factor ( GM-CSF ) , and interferon gamma ( INF-gamma ) [8] . This report presents five cases of PKDL with lesions of various degrees who were prescribed with Capsule Miltefosine ( 50mg ) twice daily for 84 days . All the patients developed unusual ophthalmic complications , after a varying duration of oral MF intake ( Fig 1 ) . The drug is categorized as H319 , by GHS Hazard Statement and there are reports of retinal toxicity in some animal studies . [9 , 10] . However , to the best of our knowledge , such severe adverse event ( SAE ) manifesting corneal toxicity with suspected association with MF , either in animal or human study , are the first instances in Bangladesh so far . As if now , Miltefosine ( MF ) is the only available oral preparation for treating leishmaniasis patients . However , such SAEs certainly requires further exploration regarding its safety , more specifically when used for a longer duration . An effective pharmacovigilance activity can play a pivotal role in this regard . Despite , the scarcity of the evidence , physicians must meticulously monitor patients while MF is administered and ensure periodic follow-ups . Preemptive counseling of the patients concerning its side effects will enable them to identify complications in due time and seek appropriate help . The drug brochure should specify and elaborate these effects , so that the physicians can be aware of the possible adverse effects and coordinate with the primary care physician for consideration of dose reduction or alternative choice of therapy in the events of such complications .	PKDL is a sequel of VL , which acts as a source of leishmaniasis and should be diagnosed and treated at the earliest possible time to prevent disease transmission . In Bangladesh during the period from May 2016 to October 2017 , a total of 5 patients were diagnosed to have Miltefosine induced unilateral ophthalmic complications . Miltefosine was introduced as an anti-parasitic drug through drug repurposing , and it has a complex mechanism of acting towards the parasite killing . The mechanism itself in certain circumstances may produce ophthalmic complications by its metabolites or by causing dry eye disease . Miltefosine is the only oral drug for leishmaniasis . Hence the etiopathogenesis and prevention strategy of such severe adverse events must be explored .	[ "Abstract", "Introduction", "Results/Discussion" ]	[ "kala-azar", "medicine", "and", "health", "sciences", "physicians", "medical", "doctors", "clinical", "research", "design", "medical", "personnel", "ocular", "anatomy", "tropical", "diseases", "parasitic", "diseases", "health", "care", "research", "design", "health", ...	2018	Corneal complications following Post Kala-azar Dermal Leishmaniasis treatment
Astrocytes integrate and process synaptic information and exhibit calcium ( Ca2+ ) signals in response to incoming information from neighboring synapses . The generation of Ca2+ signals is mostly attributed to Ca2+ release from internal Ca2+ stores evoked by an elevated metabotropic glutamate receptor ( mGluR ) activity . Different experimental results associated the generation of Ca2+ signals to the activity of the glutamate transporter ( GluT ) . The GluT itself does not influence the intracellular Ca2+ concentration , but it indirectly activates Ca2+ entry over the membrane . A closer look into Ca2+ signaling in different astrocytic compartments revealed a spatial separation of those two pathways . Ca2+ signals in the soma are mainly generated by Ca2+ release from internal Ca2+ stores ( mGluR-dependent pathway ) . In astrocytic compartments close to the synapse most Ca2+ signals are evoked by Ca2+ entry over the plasma membrane ( GluT-dependent pathway ) . This assumption is supported by the finding , that the volume ratio between the internal Ca2+ store and the intracellular space decreases from the soma towards the synapse . We extended a model for mGluR-dependent Ca2+ signals in astrocytes with the GluT-dependent pathway . Additionally , we included the volume ratio between the internal Ca2+ store and the intracellular compartment into the model in order to analyze Ca2+ signals either in the soma or close to the synapse . Our model results confirm the spatial separation of the mGluR- and GluT-dependent pathways along the astrocytic process . The model allows to study the binary Ca2+ response during a block of either of both pathways . Moreover , the model contributes to a better understanding of the impact of channel densities on the interaction of both pathways and on the Ca2+ signal . Astrocytes integrate and process synaptic information and by doing so generate calcium ( Ca2+ ) signals in response to neurotransmitter release from neighboring synapses [1] . Ca2+ signals in astrocytes are largely attributed to an elevated metabotropic glutamate receptor ( mGluR ) activity , which stimulates the phospholipase C and the production of the second messenger inositol trisphosphate ( IP3 ) . The binding of IP3 to receptors at internal Ca2+ stores ( endoplasmatic reticulum ) induces IP3 and Ca2+-dependent Ca2+ release into the intracellular space [2–7] ( see mGluR-dependent pathway in Fig 1 ) . Experimental results , however , showed not only a clear attenuation of the Ca2+ signal during an inhibition of the mGluR , but also during a block of the glutamate transporter ( GluT ) [7 , 8] . The glutamate transporter itself does not influence the intracellular Ca2+ concentration , but it indirectly activates Ca2+ entry over the membrane mediated by the Na+/Ca2+ exchanger [9] ( see GluT-dependent pathway in Fig 1 ) . The uptake of one glutamate molecule mediated by the glutamate transporter is accompanied by the transport of three sodium ( Na+ ) ions into the astrocyte and one potassium ( K+ ) ion out of the astrocyte . An inwardly directed Na+ gradient and an outwardly directed K+ gradient promote the glutamate uptake by the glutamate transporter and glutamate accumulation in the astrocyte . The Na+-K+-ATPase maintains the Na+-K+ concentration gradient and favors the glutamate transport [10] . In close proximity to glutamate transporters high concentrations of Na+/Ca2+ exchangers have been observed [9] . During a rapid rise of the Na+ concentration the Na+/Ca2+ exchanger works in the reverse mode and transports Na+ out of the astrocyte while transporting Ca2+ into the astrocyte . Thereby the Na+/Ca2+ exchanger serves as an additional transient source of Ca2+ and the intracellular Ca2+ concentration increases [9] . Therefore , at least two different mechanisms contribute to the generation of Ca2+ signals in astrocytes . A closer look into Ca2+ signaling in different astrocytic compartments revealed a spatial separation of those two pathways . In the soma Ca2+ signals are mainly evoked on the mGluR-dependent pathway , whereas in perisynaptic astrocytic processes ( PAPs ) most Ca2+ signals are evoked by Ca2+ entry over the plasma membrane [11] . These results are supported by the finding , that astrocytic compartments close to the synapse are devoid of internal Ca2+ stores and the volume ratio of internal Ca2+ stores compared to the intracellular space increases towards the soma . Moreover , the surface volume ratio decreases along the astrocytic process from the PAPs towards the soma , because processes become increasingly thinner ( see Fig 2 ) [12] . Based on the findings cited above we hypothesized that the underlying mechanisms for Ca2+ signals differ between astrocytic compartments . The mGluR-dependent pathway is mainly present close to the astrocytic soma , while the GluT-dependent pathway dominates Ca2+ signals in PAPs . So far most mathematical models attribute astrocytic Ca2+ dynamics solely to mGluRs and neglect Ca2+ entry through the membrane . In order to test whether the Na+/Ca2+ exchanger serves as a source for Ca2+ signals in PAPs , we propose a mathematical model , which incorporates glutamate driven Ca2+ responses evoked by simultaneous binding of glutamate to mGluR’s and transport of glutamate by GluT while taking the volume ratio of internal Ca2+ stores into account . With the help of the model we investigated how the volume ratio between the internal Ca2+ store and the intracellular space affects Ca2+ signaling evoked on the mGluR- and GluT-dependent pathway in different astrocytic compartments along astrocytic processes from the synapse towards the soma . We consider small astrocytic compartments , which have a cylindrical shape . Each astrocytic compartment consists of three parts: the internal Ca2+ store ( endoplasmatic reticulum ) , the intracellular space , and the extracellular space ( see Fig 1 ) . The internal Ca2+ store and the intracellular space are considered as two cylinders with different diameter , which lie within each other . The volume of the intracellular space includes the volume of the internal Ca2+ store . The intracellular space is surrounded by the extracellular space . The volume of the extracellular space is set equal to the volume of the intracellular space . Flow of ions to neighboring compartments is not considered . Thus , only the curved surface area of the cylinder is considered . For the change of the ion concentration within the intracellular space or the internal Ca2+ store ( see Eq 2 ) , we consider the sum of all ionic currents carrying the respective ion ( ∑Iion ) multiplied with the area A , the ionic current is flowing through , and divided by the volume Vol of the space the ions are located in . Both A and Vol are scaled by the length l of the compartment . Therefore , the fraction A V o l does not depend on l and lateral diffusion of ions was neglected . For each astrocytic compartment the surface area and the volume of both the internal Ca2+ store and the intracellular space change along the astrocytic process . The diameter of the intracellular space increases from astrocytic compartments close to the synapse towards astrocytic compartments at the soma ( see Fig 2 ) . Thus , the surface area and the volume of the intracellular space increase from the synapse to the soma , but the surface volume ratio ( SVR ) decreases . The volume ratio between the internal Ca2+ store and the intracellular space increases from astrocytic compartments close to the synapse towards astrocytic compartments at the soma . Astrocytic compartments close to the synapse do not contain internal Ca2+ stores ( ratioER = 0 ) ( see Fig 2 ) . Within a single astrocytic compartment the diameter of the internal Ca2+ store is smaller than the diameter of the intracellular space . The volume of the internal Ca2+ store is equal to the volume of the intracellular space reduced by the factor ratioER . Consequently , the surface area of the internal Ca2+ store is reduced by the factor r a t i o E R compared to the surface area of the intracellular space . Thus , the volume ratio between the internal Ca2+ store and the intracellular space determines the change of the surface volume ratio ( S V R = A V o l ) of the internal Ca2+ store along the astrocytic process . Along the astrocyte process , the surface volume ratio ( SVR ) and the volume ratio between the internal Ca2+ store and the intracellular space depend on each other , and the relationship ( see [12] and Fig 3 ) is quantified by: r a t i o E R = 0 . 15 · e - ( 0 . 002 μ m · S V R ) 2 . 32 ( 1 ) The release of glutamate from an activated nearby synapse is calculated using the Tsodykis and Makram model [21 , 22] in its adapted form published by Wallach and colleagues [7] . r ( t ) =x ( t ) ·y ( t ) dxdt= ( 1−x ( t ) ) τrec−x ( t ) ·y ( t ) ·s ( t ) dydt=−y ( t ) τfacil+U0 ( 1−y ( t ) ) ·s ( t ) dgdt=−gτclear+ρCGT·r ( t ) , where x and y represent the fraction of resources in the recovered and active states , respectively . During each spike a fraction of active synaptic resources is released into the synaptic cleft , and the time constant τrec determines the recovery of these resources . The fraction of active synaptic resources y increases with each spike and the step increase of y is determined by U0 . In the absence of a spike y decays back to a baseline level with time constant τfacil . The product r ( t ) corresponds to the ratio of glutamate ( g ) which is released during a spike of the sequence s . The change of the glutamate concentration in the synaptic cleft is determined by the total glutamate content of readily releasable vesicles ( GT ) and the volume ratio between the synaptic vesicles and the synaptic cleft ( ρC ) . Glutamate is removed from the synaptic cleft with the time constant τclear . Values of model parameters can be found in Table 6 . The initial values of [IP3]i , the fraction h of active IP3 receptor channels , and [Ca2+]ER and the model parameters gNaleak and gKleak were determined as follows . Since the model parameters for the production and degradation of IP3 and the intracellular resting concentration of Ca2+ were known from literature , the zero of d[IP3]i/dt revealed the initial concentration of IP3 . In the same way the initial ratio of activated IP3 receptor channels , h , and the initial concentration of the Ca2+ concentration in the endoplasmatic reticulum was calculated . In this way a stable resting state was ensured . The model parameter gNaleak was calculated by setting d[Na+]i/dt equal to zero and solving the equation for gNaleak . The model parameter gKleak was calculated the same way by setting d[K+]i/dt equal to zero . All simulations were performed with Python 2 . 7 using the packages Brian [28] , NumPy and Matplotlib . The Brian Simulator used the Euler integration as numerical integration method for the non-linear differential equations with time step dt = 1ms . First , we analyzed the generation of mGluR-dependent Ca2+ signals along the astrocytic process . For this reason we varied the volume fraction of the internal Ca2+ store ( ratioER ) , which changes along the astrocytic process ( Fig 3 ) , and studied the amplitude and the frequency of the Ca2+ signals ( Fig 4 ) . All currents related to the GluT-dependent pathway ( IGluT , INKA , INCX ) were set to zero . Astrocytic compartments with a high volume fraction of the internal Ca2+ store ( ratioER>0 . 06 ) showed Ca2+ oscillations ( Fig 4b ) . These compartments corresponded to astrocytic regions close to the soma . A reduction of ratioER decreased the amplitude of the Ca2+ oscillations . This was caused by the weaker Ca2+ influx into the cytoplasm through the smaller surface area of the internal Ca2+ store . Astrocytic compartments closer to the synapse ( 0<ratioER<0 . 06 ) did not show Ca2+ oscillations , but an increase of the intracellular Ca2+ concentration . However , when the astrocytic compartment was devoid of the internal Ca2+ store ( ratioER = 0 ) , we observed an unchanged intracellular Ca2+ concentration . Different stimulation frequencies led to qualitatively similar behavior ( data not shown ) . In particular , the critical value of ratioER = 0 . 06 for the onset of oscillations remained the same . The Ca2+ entry through the plasma membrane mediated by the Na+/Ca2+ exchanger is driven by a Na+ accumulation in the intracellular space . The glutamate transporter ( GluT ) , the Na+/Ca2+ exchanger ( NCX ) and the Na+-K+-ATPase ( NKA ) determine the intracellular Na+ concentration . For this reason we analyzed the increase of the intracellular Na+ concentration as a function of the maximal pump currents of the glutamate transporter ( IGluTmax ) , the Na+-K+-ATPase ( INKAmax ) , and the Na+/Ca2+ exchanger ( INCXmax ) . The maximal pump current of the GluT ( IGluTmax ) and the NKA ( INKAmax ) had a strong effect on the accumulation of Na+ in the astrocyte , while changes of the maximal pump current of the NCX ( INCXmax ) showed no effect ( see Fig 5 ) . The accumulation of Na+ in the intracellular space was highest for a high maximal pump current of GluT and a low maximal pump current of NKA ( see Fig 5a ) . While the GluT transported Na+ into the astrocyte , the NKA counteracted this effect by pumping Na+ out of the astrocyte and led to a saturation of [Na+]i at lower concentration levels . The time until saturation was lowest for a low maximal pump current of the GluT and a high maximal pump current of the NKA ( see Fig 5b ) . A low maximal pump current of the GluT resulted in a small Na+ accumulation in the intracellular space , which saturated faster for a high Na+ transport out of the astrocyte mediated by the NKA . In experiments the increase of the intracellular Na+ concentration in response to external stimulation with glutamate ranges from 10 mM to 20 mM saturating with increasing glutamate concentrations [29] and is performed in under 60 seconds [30] . For the following simulations we chose a parameter combination of the maximal pump currents of the GluT and the NKA which revealed the desired results for the increase of the intracellular Na+ concentration and the time to saturation ( IGluTmax = 0 . 68 p A μ m 2 and INKAmax = 1 . 52 p A μ m 2 . ) . As a next step , we analyzed how the Ca2+ transport through the membrane mediated on the GluT-dependent pathway affects mGluR-dependent Ca2+ signals along the astrocytic process . Different regions of the astrocytic process were simulated by changing the volume fraction of the internal Ca2+ store ( ratioER ) . We analyzed the influence of the GluT-dependent pathway on the Ca2+ signal by changing the maximal pump currents of the Na+/Ca2+ exchanger ( INCXmax ) and the glutamate transporter ( IGluTmax ) . First , we analyzed the impact of Ca2+ transport through the membrane mediated by the Na+/Ca2+ exchanger on the intracellular Ca2+ signal along the astrocytic process ( see Fig 6 ) . During a block of the Ca2+ transport through the membrane ( INCXmax = 0 p A μ m 2 ) Ca2+ oscillations were only observed for a high volume fraction of the internal Ca2+ store ( ratioER>0 . 06 ) ( see Fig 6a and 6g ) . An increase of the maximal pump current of the Na+/Ca2+ exchanger ( INCXmax > 0 p A μ m 2 ) shifted the critical value of ratioER for the onset of Ca2+ oscillations to higher values ( see Fig 6a ) , culminating in a total suppression of the Ca2+ oscillations ( see Fig 6e ) . In astrocytic compartments , which were devoid of the internal Ca2+ store ( ratioER = 0 ) , Ca2+ was transported into the astrocyte and the intracellular Ca2+ concentration increased ( see Fig 6b and 6c ) . Second , we analyzed the influence of the maximal pump current of the glutamate transporter ( IGluTmax ) on the Ca2+ signal ( see Fig 7 ) . The impact of IGluTmax on the Ca2+ signal mainly depended on the maximal pump current of the Na+/Ca2+ exchanger ( INCXmax ) and the volume fraction of the internal Ca2+ store ( ratioER ) . In astrocytic compartments close to the soma ( ratioER≥0 . 1 ) an increase of IGluTmax increased the Ca2+ oscillation frequency until it reached a maximal value and decreased again ( see Fig 7a and 7b ) . An increase of INCXmax shifted the maximal value of the oscillation frequency to lower values of IGluTmax ( see Fig 7a ) . The increase of IGluTmax caused a higher increase of the intracellular Na+ concentration . The higher Na+ accumulation activated the Na+/Ca2+ exchanger in the reverse mode and prevented an outflux of Ca2+ into the extracellular space . The elevated Ca2+ transport into the cell preserved the Ca2+ oscillations for high values of INCXmax and resulted in an increase of the oscillation frequency . The amplitude of the Ca2+ oscillations was mainly affected by the volume fraction of internal Ca2+ stores and increased with an increase of ratioER ( see Fig 7c ) . The increase of the volume of both the internal Ca2+ store and the intracellular space with ratioER caused an enhanced Ca2+ release from the internal Ca2+ store . The interplay of the mGluR- and GluT-dependent pathways showed the experimentally observed Ca2+ fluctuations in astrocytic compartments with a low volume fraction of an internal Ca2+ store ( ratioER ) for a high pumping activity of the NCX ( INCXmax > 0 p A μ m 2 ) . However , a high maximal pump current of the NCX ( INCXmax > 0 . 01 p A μ m 2 ) evoked a suppression of the Ca2+ oscillations in regions with a high ratioER . Thus , in comparison with experimental data the simulation data suggested a low maximal pump current of the NCX for regions with a high ratioER and a high maximal pump current of the NCX in regions with a small ratioER . Moreover , an increase of IGluTmax allowed Ca2+ oscillations for high values of INCXmax ( INCXmax ≥ 1 p A μ m 2 ) . Thus , the distribution of GluTs and NCXs determines Ca2+ signal along the astrocytic process . The reason for the suppression of the Ca2+ oscillations for high ratioER was investigated in a later results section . Experiments have shown that a block of the glutamate transporter ( GluT ) leads to a clear attenuation of the Ca2+ signal [8] . For that reason we examined the impact of the GluT-driven Ca2+ signal on the overall Ca2+ response to synaptic stimulation . Fig 8 shows the dynamics of the Ca2+ signal as a function of the volume ratio between the internal Ca2+ store and the intracellular space ( ratioER ) with ( ’control condition’ ) and without ( ’block’ ) a contribution of the GluT . We observed a high impact of the GluT-driven Ca2+ signal for a high pumping activity of the Na+/Ca2+ exchanger ( INCXmax > 0 . 1 p A μ m 2 ) and a small volume ratio between the internal Ca2+ store and the intracellular space ( ratioER<0 . 1 ) ( see Fig 8b and 8e ) . With a decrease of INCXmax and an increase of ratioER the impact of the GluT-driven Ca2+ signal decreased ( see Fig 8b , 8c and 8d ) . In astrocytic compartments with a low volume fraction of the internal Ca2+ store the Ca2+ signal mainly arose by the Ca2+ transported through the membrane ( see Fig 4 ) . A block of the glutamate transporter prevented a Na+ accumulation in the intracellular space ( see S1 Fig ) . The Na+/Ca2+ exchanger remained in the forward mode and transported Ca2+ out of the astrocyte . Thus , during a block of the glutamate transporter no Ca2+ was transported into the astrocyte via the Na+/Ca2+ exchanger and a clear attenuation of the Ca2+ signal was observed in regions with a small ratioER . With an increase of the volume fraction of the internal Ca2+ store more Ca2+ was released from the internal Ca2+ store and led to a lower impact of the glutamate transporter on the overall Ca2+ signal . The extracellular glutamate concentration and the Ca2+ entry through the membrane affected the IP3 production as well as the IP3- and Ca2+-dependent Ca2+ release from internal Ca2+ stores . An increase of ratioER was accompanied with an increase of the IP3- and Ca2+-dependent Ca2+ release from internal Ca2+ stores and thus with an increase of the impact of the mGluR-dependent mechanism . For that reason , Ca2+ signals mainly evoked by the GluT-dependent mechanism were observed in regions with a small ratioER . In order to study the mechanisms underlying the interaction of the mGluR- and GluT-dependent pathways we analyzed the Ca2+ concentration in the three spaces as well as the concentration of IP3 in the intracellular space and the fraction h of open IP3 channels for different values of the maximal pump current of the Na+/Ca2+ exchanger ( INCXmax ) and the volume ratio of internal Ca2+ stores ( ratioER ) . Fig 9 summarizes the results . Oscillations of the Ca2+ concentration in the intracellular compartment ( see Fig 9b ) were reflected in all of the other dynamical variables ( see Fig 9c–9f ) . When the GluT-dependent pathway was studied in isolation and Ca2+ release from internal Ca2+ stores was neglected ( see Fig 9a ) a finite current through the Na+/Ca2+ exchanger led to an increase of [Ca2+]i when compared with the concentration without external stimulation . The stationary value of [Ca2+]i was independent of the maximal pump currents . When both the GluT- and mGluR-dependent pathway were considered ( see Fig 9b–9f ) a high INCXmax ( INCXmax > 0 . 001p A μ m 2 ) caused an increase of the concentration of IP3 and the fraction h of open IP3 receptor channels . This caused Ca2+ flux out of the internal Ca2+ store leading to a decrease of [Ca2+]ER compared to the resting concentration . The concentration of Ca2+ in the intracellular space , however , increased by 0 . 1 μM while [Ca2+]o increased by 3 μM compared to its resting concentration . For high values of the maximal pump current of the Na+/Ca2+ exchanger the Ca2+ transport into the endoplasmatic reticulum mediated by the SERCA pump was overcompensated by the highly strong outflux of Ca2+ via the Na+/Ca2+ exchanger ( see S ? ? ) . Thus , Ca2+ accumulated in the extracellular space , which prevented the generation of Ca2+ oscillations . Our computational study addresses the generation of Ca2+ signals in different astrocytic compartments along the astrocytic process . We considered two different pathways for the generation of Ca2+ signals: the metabotropic glutamate receptor ( mGluR ) - and glutamate transporter ( GluT ) -dependent pathway . We analyzed both pathways in consideration of the volume ratio between the internal Ca2+ store and the intracellular space . The volume ratio between the internal Ca2+ store and the intracellular space changes from the soma towards the synapse . Whereas astrocytic compartments at the soma have a high volume ratio between the internal Ca2+ store and the intracellular space , in astrocytic compartments close to the synapse there is a low volume ratio . There are five main findings of the study . First , while considering the mGluR-dependent pathway in isolation Ca2+ oscillations have only been observed in astrocytic compartments with a high volume ratio between the internal Ca2+ store and the intracellular space . Second , a high maximal pump current of the Na+/Ca2+ exchanger suppressed Ca2+ oscillations in regions with a high volume ratio between the internal Ca2+ store and the intracellular space . Third , the suppression of Ca2+ oscillations for a high maximal pump current of the Na+/Ca2+ exchanger in astrocytic compartments with a high volume ratio between the internal Ca2+ store and the intracellular space was due to an overcompensation of the Ca2+ influx from the internal Ca2+ store by the outflux of Ca2+ into the extracellular space via the Na+/Ca2+ exchanger . Fourth , a high impact of the GluT-dependent mechanism on the generation of Ca2+ signals was observed for a high maximal pump current of the Na+/Ca2+ exchanger in regions with a low volume ratio between the internal Ca2+ store and the intracellular space . Fifth , the GluT-dependent mechanism accounted for Ca2+ fluctuations in astrocytic compartments which were devoid of internal Ca2+ stores . In their study Srinivasan and colleagues also addressed the question which mechanism could account for Ca2+ fluctuations in astrocytic compartments close to the synapse . They discovered that a significant proportion of Ca2+ signals in astrocytic compartments close to the synapse is because of transmembrane Ca2+ fluxes . In our model we also considered Ca2+ transport from the extracellular space into the intracellular space of the astrocyte through the GluT-dependent pathway . We found that the GluT-dependent Ca2+ transport into the astrocyte could account for mGluR-independent Ca2+ fluctuations in astrocytic compartments with a low volume ratio between the internal Ca2+ store and the intracellular space . However , while analyzing both the mGluR- and GluT-dependent pathway a high maximal pump current of the Na+/Ca2+ exchanger suppressed Ca2+ oscillations in astrocytic compartments with a high volume ratio between the internal Ca2+ store and the intracellular space . Moreover , the contribution of the GluT on the generation of Ca2+ signals was highest for a large maximal pump current of the Na+/Ca2+ exchanger in astrocytic compartments with a low volume ratio between the internal Ca2+ store and the intracellular space . These simulation results suggested a change of the pumping activity of the Na+/Ca2+ exchanger along the astrocytic process . A low maximal pump current in astrocytic compartments at the soma prevented the suppression of Ca2+ oscillations . A high maximal pump current in astrocytic compartments close to the synapse allowed a high contribution of the GluT-dependent pathway on the generation of Ca2+ signals . Based on the strength of the maximal pump current the channel density of the Na+/Ca2+ exchanger can be concluded . The higher the maximal pump current is , the more ions are transported through the membrane . The same holds true for the channel density . The higher the channel density is , the more ions are transported through that channel . Experimental results confirm a concentration and colocalization of Na+/Ca2+ exchangers , Na+/K+-ATPases and GluTs in perisynaptic astrocytic processes [31 , 32] . Ca2+ transport through the plasma membrane ( e . g . via the Na+/Ca2+ exchanger ) [23 , 33 , 34] as well as by the Ca2+ diffusion within a single astrocyte [35 , 36] or between astrocytes [37] changes the intracellular Ca2+ concentration . Fluctuations of the intracellular Ca2+ concentration affect both the Ca2+ entry mediated by the Na+/Ca2+ exchanger when operating in the reverse mode [23] and the Ca2+ release probability of the endoplasmatic reticulum [38] . The current model neglects Ca2+ diffusion within the astrocyte and describes the Ca2+ dynamics in a single compartment . Thus , an extension of the current point-model to a multi-compartment model will most probably reveal deviating results for parameters such as the maximal pump current of the Na+/Ca2+ exchanger . Moreover , the volume determines the number of Ca2+ ions within an astrocytic compartment and consequently the concentration change . Thus , diffusion of Ca2+ in astrocytic compartments with a low volume , such as in the perisynaptic astrocytic processes , leads to a bigger concentration change as in compartments with a larger volume . The above named findings allow to make a prediction about the functional role of astrocytes in neural networks . Astrocytic compartments , which have a high volume ratio of internal Ca2+ stores and are capable of IP3-dependent Ca2+ release , are not located directly at the synapse . Moreover , the high surface volume ratio of the perisynaptic astrocytic processes and a slow diffusion exchange in such thin processes favors a localized Na+ accumulation and promotes Ca2+ intrusion mediated by the NCX [39] . This may indicate that store-dependent Ca2+ signals in astrocytes act as integrators of local network activity , but not as detectors of individual synaptic events [12] . GluT-dependent Ca2+ signals in perisynaptic astrocytic processes are evoked in response to individual synaptic events . Depending on the synaptic activity Ca2+ is transported into the astrocyte by the Na+/Ca2+ exchanger and diffuses within the astrocyte network . Once this Ca2+ wave reaches astrocytic compartments which are capable of store dependent Ca2+ signals an integration of the local network activity , the intracellular Ca2+ signal and the glutamate-dependent IP3 production , takes place . Our model describes the generation of Ca2+ signals in a single astrocyte compartment with respect to its morphology . However , it is of special interest how activity of single synapses and neural networks is integrated by astrocytes . It was proposed that perisynaptic astrocytic processes serve as detectors for single synaptic events , whereas astrocytic processes which contain Ca2+ stores act as integrators of neural network activity [12] . A multi-compartment model would contribute to the analysis of the integration of neural activity performed by astrocytes . This would allow the study of Ca2+ waves within a single astrocyte and in astrocyte networks as well as their impact on the surrounding extracellular space .	Astrocytes are considered as active partners in neural information processing , because they integrate and process synaptic information and control synaptic transmission . Neuronal transmitter release induces the generation of Ca2+ signals in astrocytes . The functional role of astrocytic Ca2+ signals is still under debate . However , experimental results were able to show that astrocytic Ca2+ signaling acts to control local network activity , which plays an important role in diseases like epilepsy . Thus , it is of special interest to investigate the underlying mechanisms for Ca2+ signals in astrocytes in order to understand the role of astrocytes in neural network activity . Two different mechanisms are known to be responsible for the generation of Ca2+ signals in astrocytes . These mechanisms are the release of Ca2+ from internal Ca2+ stores and the entry of Ca2+ through the plasma membrane . We studied the interaction of those two different mechanisms for the generation of Ca2+ signals and found that these mechanisms are spatially separated along the astrocytic processes .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "medicine", "and", "health", "sciences", "neurochemistry", "action", "potentials", "nervous", "system", "astrocytes", "endoplasmic", "reticulum", "cell", "processes", "membrane", "potential", "electrophysiology", "neuroscience", "macroglial", "cells", "neurotransmitters", "...	2017	Spatial separation of two different pathways accounting for the generation of calcium signals in astrocytes
Fast-spiking ( FS ) cells in the neocortex are interconnected both by inhibitory chemical synapses and by electrical synapses , or gap-junctions . Synchronized firing of FS neurons is important in the generation of gamma oscillations , at frequencies between 30 and 80 Hz . To understand how these synaptic interactions control synchronization , artificial synaptic conductances were injected in FS cells , and the synaptic phase-resetting function ( SPRF ) , describing how the compound synaptic input perturbs the phase of gamma-frequency spiking as a function of the phase at which it is applied , was measured . GABAergic and gap junctional conductances made distinct contributions to the SPRF , which had a surprisingly simple piecewise linear form , with a sharp midcycle break between phase delay and advance . Analysis of the SPRF showed how the intrinsic biophysical properties of FS neurons and their interconnections allow entrainment of firing over a wide gamma frequency band , whose upper and lower frequency limits are controlled by electrical synapses and GABAergic inhibition respectively . Rhythmic oscillations of concerted electrical activity can occur in the neocortex and hippocampus at gamma frequencies ( 30–80 Hz ) , and are thought to be associated with a variety of cognitive tasks including sensory processing , motor control , and feature binding [1] , [2] . A striking feature of gamma oscillations is their ability to be generated locally in the neocortex . Local gamma oscillations can be produced by pharmacological [3] , [4] , electrical [5] or optogenetic [6] stimulation . In vivo , synchronous gamma oscillations may be highly localized or widely distributed , even between hemispheres , with or without phase lags between different areas and layers [1] . It appears , therefore , that local neocortical circuits have an intrinsic capability for generating gamma oscillations , while sensory inputs and connections from other brain regions may shape the complex spatial patterns of oscillatory interaction . Synchronized firing of cortical inhibitory interneurons has been implicated in the production of these rhythms in many experimental and modeling studies . During spontaneous network activity of the neocortex in vivo , the power of intracellular voltage fluctuations at frequencies higher than 10 Hz is dominated by inhibitory postsynaptic potentials , which are correlated with the extracellular gamma rhythm , and which synchronously inhibit nearby pyramidal cells [7] . A recent study using conductance injection in neocortical pyramidal cells indicated that gamma-frequency-modulation of firing is almost completely determined by their inhibitory input [8] . In the hippocampus and cortex , models of interneuron activity suggest that network oscillations depend on mutually inhibitory synaptic conductances [9] , [10] , [11] . Fast-spiking ( FS ) inhibitory interneurons are coupled by electrical synapses in addition to mutual and autaptic inhibitory synapses [12] , [13] , [14] , [15] . Electrical synapses alone [12] , [13] or in combination with GABAergic synapses [14] can produce synchronous firing in pairs of these interneurons in vitro . In addition , the biophysical properties of FS neurons appear to be ideally suited to generating gamma rhythms: they have a hard ( “type 2” ) onset of regular firing at about 30 Hz [16] , which means that they can be easily entrained at this frequency . They also show a strong intrinsic drive for spike generation at gamma frequencies when stimulated with broadbrand conductance noise [17] . Recently , selective optical stimulation of FS interneurons , but not of pyramidal neurons , was shown to cause gamma oscillations [6] . Electrical synapses amongst mutually inhibitory interneurons have been found to increase the precision of synchrony in simulation studies [18] , [19] , [20] . However , the relative roles of chemical inhibition and gap-junctional coupling in shaping synchronous oscillations in the cortex are still unclear . The theory of synchronization of coupled oscillators uses the concept of phase dynamics to evaluate the stability of the relative phase of coupled oscillators in time [21] , [22] . The key to this approach is to determine the effect of a very small perturbing input on the phase of oscillation ( “phase resetting” ) , as a function of the point in the oscillation cycle at which it occurs . This is most often used , under the assumptions of weak coupling and linear summation of phase shifts , to account for how the relative phase of presynaptic and postsynaptic cells evolves from cycle to cycle . However , as described above , FS cells in the cortex are actually coupled quite strongly to other FS neighbours , with large postsynaptic conductance changes caused by each presynaptic action potential . Here , we have used synthetic conductance injection , or dynamic clamp , to directly measure the phase-resetting response to conductance inputs mimicking the effects of presynaptic action potentials , while systematically varying the relative strengths of electrical and GABAergic inhibitory conductances . The compound synaptic connections between FS neurons , together with the intrinsic spike-generating properties of FS neurons , give rise to a distinctively-shaped phase-resetting relationship , or “synaptic phase-resetting function” , which ensures rapid and precise synchronization over a large gamma-frequency range . FS cells in rat somatosensory cortical slices were identified by their morphology , action potential shape and characteristic firing pattern in response to depolarizing current injection [12] , [13] , [23] , [24] . FS cells fired high frequency , nonadapting trains of action potentials during depolarizing current steps , occasionally interrupted by pauses with subthreshold oscillations , particularly around threshold [16] ( see Methods ) . We used conductance injection/dynamic clamp [25] , [26] to reproduce the effects of electrical and chemical synapses ( Fig . 1 , see Methods ) . In FS cells , both gap junctions and GABAergic synapses from neighboring cells are located perisomatically [14] , so that point conductance injection at the soma should reasonably reproduce the electrical effects of synaptic inputs . Gap junctions were implemented as a static conductance between the recorded cell and a “voltage-clamped” trajectory of “presynaptic” membrane potential . This “voltage-source” approximation , importantly , allowed us to characterize a functional mapping between the presynaptic spike time and the influence on postsynaptic membrane potential , without considering any reverse effect of gap-junctional current on the presynaptic cell . This is valid as long as the presynaptic cell is considered to be much more strongly controlled by its other inputs , as when it is already part of a synchronous assembly ( see Discussion ) . It is estimated that each FS cell is gap-junction coupled , directly or indirectly , with a measurable coupling , to between 20 and 50 other FS neurons [27] , so that if the presynaptic cell is quite strongly-driven by a major proportion of these inputs , then the effect of any one can be neglected . At rest , this gap-junctional input produced a small postsynaptic spikelet ( Fig . 1a , left ) , very similar in size and shape to those observed with natural electrotonic coupling [12] , [13] . We also measured coupling coefficients ( the ratio of postsynaptic to presynaptic potential change ) for gap-junctional type conductance . These were similar to physiological values , and larger for step inputs ( 0 . 05–0 . 22 ) than for spike inputs ( 0 . 01–0 . 05 ) , owing to low-pass filtering by the combined effects of gap junctional conductance and membrane resistance and capacitance [28] . Many pairs of FS cells are connected by both GABAergic ( GABAA , chloride conductance ) and electrical synapses [12] , [13] , [14] . We simulated GABAergic synaptic input using conductance injection ( Fig . 1a , middle ) . The GABA reversal potential ( EGABA ) was set to −55 mV , based on gramicidin-perforated patch measurements in this cell type [10] , [29] , considerably more depolarized than in pyramidal neurons [30] . Thus , inhibition is shunting in the range of membrane potentials between spikes during repetitive firing ( Fig . 1b ) . Starting from the resting potential , the “IPSP” is a small depolarisation lasting about 40 ms , again very similar to natural IPSPs in these cells . At the resting potential , a stimulus with both electrical and GABAergic components produces a biphasic depolarizing response ( Fig . 1a , right ) with the gap-junctional potential visible just before the larger GABAergic potential . Unlike the gap-junctional spikelet , though , the amplitude of the GABAergic potential can change sign in the subthreshold , interspike range of membrane potentials , reversing around EGABA [12] . To determine how this compound synaptic input shifts the timing of periodic firing in an FS cell , we applied conductance inputs during periodic firing elicited by a maintained excitatory stimulus , a step of excitatory conductance reversing at 0 mV . An example response to a compound “synaptic” perturbation is shown in Fig . 1b . In phase-resetting analysis of synchronization , the state of the neuron is characterized by a single quantity , the phase angle , , which – in the absence of any perturbations - increases linearly with time , and which is reset to zero whenever it reaches , corresponding to the occurrence of a spike [21] . The variability of interspike intervals can be represented by adding additional noise , due to stochastic gating of ion channels and other intracellular sources of variability , to the rate of change of . To measure the phase resetting , or shift in the phase , produced by synaptic-like conductance inputs , we applied isolated single inputs during long trains of periodic firing . Fig . 1c shows the relationship between the time tp at which an input ( in this case a compound gap/GABA input ) is applied , relative to the time of the preceding spike , and the time until the next spike occurs ( tn ) . This clearly deviates from the line of slope −1 ( dotted line ) expected in the absence of any input , and has two approximately linear regions separated by a sharp transition . Note the characteristic progressive decrease in the variability of this relationship , as tp increases – this is because the earlier the input arrives , the more time is left for integrating the effects of noise before the next spike . From this relationship , we can estimate the phase at the moment that each input is applied , and the amount of phase resetting produced by the input ( see Methods ) , as shown in Fig . 1d , in which is plotted as a function of . This relationship - the total phase-resetting effect of a synaptic input as a function of the phase at which it arrives – we will refer to as a synaptic phase-resetting function ( SPRF ) , to distinguish it from a classical phase response or phase-resetting curve , which normally describes responses to very small , brief inputs , whose effects can be considered to sum linearly . We examined how the parameters of the synaptic input determine the shape of the SPRF , by varying the magnitude of gap-junctional and GABAergic conductance , applied individually or together ( Fig . 2a–f ) . These components vary physiologically , since FS cells' interconnections can be purely GABAergic ( one-way or reciprocal ) , purely gap-junctional or both [12] , [13] , [14] . In addition , there is a wide range of electrical synaptic strengths [28] . Purely GABA input produced a phase delay early in the cycle , which increased during the cycle until an abrupt critical point , beyond which it had no effect ( Fig . 2a ) . Introducing a small ( 250 pS ) gap junction , caused a linear region of phase advance ( Fig . 2b ) , as in Fig . 1d , which had an abrupt onset at a phase of about . A sharp transition marks the boundary between this region and the first , phase delay part of the phase cycle . The slope of the phase advance region became more negative , and the boundary between the regions , designated the critical phase , shifted earlier in the cycle , as gap junctional conductance increased ( Fig . 2c , d , e ) . With no GABAergic input , a phase advance region produced by gap junctional input is seen in isolation ( Fig . 2f ) . Thus , GABAergic input retards , and gap-junctional input advances the phase of firing . For the compound gap/GABA input , the early region of phase delay has a slope determined by the amplitude of inhibition , gi ( see Methods ) , and switches abruptly , midcycle , to a region of decreasing phase advance , whose slope is determined by ge , with no detectable sign of cancellation of the two regions in midcycle . The only clear interaction between the electrical and GABAergic components was that a larger gap junctional conductance shifted to earlier in the cycle . To quantify the goodness of fit of the piecewise linear SPRF , we performed a chi-square test of 130 phase response curves ( in total 6111 data points , 10 cells ) . For each SPRF , variance of phase was estimated from an unperturbed spike train within the same experiment ( median = 0 . 021 ( rad/2π ) 2 . 111 of 130 SPRFs contained no significant difference between the model fit and experimental result ( p<0 . 05 ) . The average reduced chi-square value was 0 . 80 , meaning that the overall fit of the model is extremely good , given the measured degree of variance in the phase . On the whole , the relatively simple piecewise linear model performs remarkably well . The dependencies of the slopes and breakpoint on the strengths of gi and ge were also fitted by linear relationships ( Figure 3 ) . The negative slope of the region of phase delay was proportional to inhibition ( , Fig . 3b ) , the negative slope of the phase advance region was proportional to excitation ( , Fig . 3a ) , while was weakly sensitive to ge ( , Fig . 3c ) . Average values of a and b of this piecewise linear model for the SPRF were a = 0 . 16/nS ( n = 7 cells , 3 cells providing insufficient data for analyzing this dependency ) , b = 0 . 69/nS ( n = 10 cells ) . c and d were more variable from cell to cell , and the pooled data in fact showed little overall dependence on ge ( not shown ) . Nevertheless ( e . g . Fig . 3c ) , the weak relationship is clear within individual cells . Having established that conductances resembling the synaptic input of neighboring FS cells can consistently modify spike timing , we next tested the ability of FS cells to synchronize to , or to be entrained by this input . To visualize the time course of entrainment , we examined responses stroboscopically [22] , sampling the phase of the FS cell at the times of periodic stimuli . Figure 4 shows such an experiment . Before the conductance pulses are switched on ( open circles ) , the phase changes in a “sawtooth” pattern , reflecting detuning - the continuously growing phase difference between two oscillators of different frequencies . After the conductance transients begin ( Fig . 4 , filled circles ) , the phase quickly converges on a fixed value relative to the stimulus , at about ( dashed line ) , which matched the expected equilibrium phase difference from solving Equation 2 with parameters for this cell . Thus the FS cell becomes phase-locked and frequency-locked to the stimulus train , with spikes occurring around 0 . 6π before , or equivalently 1 . 4π after each stimulus . After the end of the stimulation train , the phase reverts to the drifting detuned state . The piecewise linear SPRF could also account for the frequency band over which synchronization was possible . Fig . 5 shows an experiment in which an FS neuron firing at a steady frequency F was stimulated repeatedly with a periodic synaptic conductance input at frequency f , and an index of the synchrony of the cell with the input ( S , varying between 0 and 1 , see Methods ) was measured over a range of frequencies . As seen in Fig . 5a , this changes from a low level when f is very different from F , to a high value approaching 1 , when . Because of the effects of noise in the neuron , there is no absolute phase locking ( S<1 ) , and the change in synchrony with input frequency does not have abrupt boundaries , but falls away continuously as the difference between f and F grows . It is clear that the central region of high synchrony lies below the unperturbed or natural firing frequency F when only inhibition is applied ( Fig . 5b ) , above F when only gap-junctional conductance is applied ( Fig . 5c ) , or both above and below F when a compound input is applied ( Fig 5a ) . This observation was duplicated by the piecewise linear model of the SPRF , analysis of which ( see Methods ) predicted the 1∶1 synchronized frequency bands shown in gray , for the deterministic ( noise-free ) case – in this neuron , these boundaries corresponds to a synchrony of about 0 . 7 . The synchronized frequency band is much narrower for either gap-junctional stimulation alone ( Fig . 5b ) or GABAergic inhibition alone ( Fig . 5b ) . Iterations of the noisy stroboscopic map derived from the fitted SPRF ( Eq . 2 ) showed that it could also reproduce the distribution of S adequately ( black curves in Fig . 5a–c ) . Thus the piecewise linear model of the SPRF appears to account very well , both for the frequency range and degree of synchronization in noise . We next used the SPRF to predict the frequency ranges of entrainment for different strengths of inhibition and electrical coupling ( Fig . 6 ) , by analyzing the bifurcations at the onset of synchrony in the stroboscopic map of the phase , i . e . the map of the phase of the postsynaptic cell at successive presynaptic spike times in a regular train ( see Methods , equation 2 ) . For the deterministic ( zero noise ) case , 1∶1 entrainment corresponded to a stable fixed point of the map , labelled in the example shown in Fig . 6a . As the amount of detuning ( difference between f and F ) varies , the map shifts vertically , so that at certain stimulus frequencies , the fixed point disappears ( at a “corner-collision” bifurcation [31] ) . Thus , it is possible to plot the regions in which there is synchronization in the plane ( Fig 6b ) or the plane ( Fig . 6c , d ) . These form Arnol'd tongues [22] in which the frequency range of entrainment shrinks as the synaptic strength is reduced . This analysis shows a number of effects which are relevant to the physiological function of FS neurons . Increasing strongly increased the upper frequency limit of entrainment and weakly increased the lower limit ( Figs . 6b ) . When it is impossible to entrain firing with f<F . Conversely , with , it is impossible to entrain for f>F , and increasing strongly reduces the lower frequency limit of entrainment ( Fig . 6c , d ) . Since physiologically , entrainment must occur in the face of considerable noise , we also investigated the effect of adding noise to the phase map . It is possible to define stochastic bifurcation points of the map F , at which there is a qualitative change in the nature of the stochastic dynamics . These points coincide with the deterministic bifurcation frequencies [32] for ( see Methods for details ) . We examined the frequency extents of this kind of stochastic entrainment at different noise levels ( Fig . 6b–d ) . In all cases , increasing the noise in the phase shrinks the region of entrainment . For rad/2π , which was a typical noise level in these cells in vitro , the area of stochastic entrainment shrank to a third or less of the noise-free case . This noise-induced distortion is not symmetrical in the frequency axis . For example , Fig . 6d shows that in the absence of electrical coupling , the lower frequency limit of entrainment was highly susceptible to noise while the upper limit was not . The greater the level of electrical coupling ( ) , the more the upper limit was reduced by noise . The SPRF makes several predictions . First , FS cells receiving purely electrical synaptic input will synchronize effectively when driven at frequencies higher than F . Higher frequencies can be followed with stronger electrical input . Second , cells will synchronize to purely inhibitory input at frequencies lower than F , and stronger inhibition allows lower frequencies to be followed . Third , combined electrical and inhibitory input allows cells to synchronize to frequencies both above and below their unperturbed frequency . Although noise diminishes the frequency band of synchronization , sometimes asymmetrically , these conclusions remain valid in the presence of noise . For typical strengths of combined electrical-inhibitory synaptic connections , 20 Hz or greater bandwidths of stochastic synchronization persist even in quite high levels of noise ( σ = 0 . 1 ) . A number of previous theoretical and experimental studies have examined the phase-resetting properties of cortical neurons . Ermentrout and Kopell developed a theoretical approach to calculate what they termed the “synaptic interaction function” based on phase response curves and the assumption of weak coupling [33] . Reyes and Fetz ( 1993 ) stimulated synaptic inputs to regularly-firing pyramidal neurons to measure the phase resetting produced by EPSPs [34] , while Stoop et al . ( 2000 ) used similar measurements to predict input frequency regions for entrainment and chaos [35] . Netoff et al . used dynamic-clamp to measure phase-resetting ( or spike-time response curves ) by artificial excitatory or inhibitory conductances in excitatory stellate cells of medial entorhinal cortex , and oriens-lacunosum-molecular interneurons in the CA1 region of hippocampus [36] , and were able to demonstrate synchronization in pairs of neurons connected by artificial conductances mimicking synaptic connections , or between biological neurons and simulated neurons . In fast-spiking inhibitory cells , Mancilla et al . ( 2007 ) measured phase-resetting relationships for small current pulses ( weak coupling ) and showed that they could account quite well for synchronization of pairs of gap-junction coupled FS cells , both experimentally and in a biophysical model of FS neurons [37] . In this paper , we go further , by using conductance injection ( dynamic clamp ) to reproduce the combined effect of gap-junctional and strong synaptic connections , and using this to predict the resulting synchronized frequency bands , and their dependence on synaptic strength , including the effect of noise in the synaptic phase-resetting function on synchronization . The conductance pulses which we have used are based on the physiological properties of the synaptic connections between FS neurons . In FS neurons of a basket morphology , APs initiate in the axon [38] arising usually from a proximal dendrite , [39] and receive many of their inhibitory connections and gap junctions from other fast-spiking interneurons perisomatically [14] . Thus , dynamic clamp recordings at the soma should provide a reasonably realistic simulation of the natural gap-junctional and fast inhibitory input . In order to carry out this analysis , we have made the approximation that , between spikes , the presynaptic voltage of the gap-junctional input was held at a resting potential of −70 mV , . In other words , we have focused on the effect of gap-junctional current flow associated with the discrete event of the presynaptic spike . This approach does not take account of the way in which presynaptic membrane potential would gradually depolarize between spikes , if firing periodically . We have also ignored the two-way nature of coupling between cell pairs . In other words we model entrainment of one cell by another , rather than synchronization of a symmetrical coupled pair . Although both electrical and inhibitory coupling can often be asymmetrical [13] , [40] , they may also be quite symmetrical . However , the entrainment studied here models the situation where the presynaptic cell is already imperturbably-driven as part of a strong synchronously-firing assembly of FS neurons , so that the phase and frequency of its firing will be clamped to that of its predominant input . Thus , the SPRF that we measure should be an effective model for describing recruitment of new cells to such a synchronous assembly . It is expected that the preferred firing frequency F of the postsynaptic cell may also affect the form of the SPRF , since the timing of intrinsic ion channel kinetics will shift relative to phase as the cycle length changes . In a few experiments where we were able to address this issue , we indeed found evidence of a change in the parameters of the SPRF model . a , the dependence of phase delay on gi , increased quite strongly as firing frequency increased , and shifted earlier in the cycle as firing frequency increased . The dependence of b and d on firing frequency was not marked . The relatively strong effect on a may partly reflect the long duration of the IPSP conductance relative to the period of the cycle . The synaptic phase-resetting function , or SPRF , for compound input was distinguished by the following features: an extremely abrupt midcycle switch from phase delay to phase advance , which shifted weakly towards the early part of the cycle as the strength of electrical coupling was increased; amplification of the phase delay region by increasing inhibition; and amplification of the phase advance region by increasing gap-junctional coupling . We found that these qualitative features were also present in a biophysical model of firing in fast-spiking cells [41] ( see Methods ) , incorporating voltage-gated sodium , Kv1 . 3 and Kv3 . 1/3 . 2 potassium channels , and stimulated with exactly the same inputs as used experimentally ( Fig . 7 ) . In this fully-deterministic model , we also observed a very fine local structure of fluctuations around the main relationship , particularly in the phase delay . Despite these qualitative similarities between the model and experimental results , there were also major differences . In experiments , phase advance was produced exclusively by gap-junctional conductance and phase delay exclusively by inhibition , while in the model , gap-junctional input did affect phase delay strongly early in the cycle – this was never observed experimentally . This deficiency of the biophysical model suggests that additional conductances expressed in FS neurons somehow help to confer a complete immunity to gap-junctional stimulation in the early , phase-delay part of the cycle . We surmise that the voltage-gated potassium conductance in this part of the cycle may actually be much higher than in the model , and that this may allow phase delay and advance to be regulated completely independently . Also , because of their relative timing , the effect of inhibition will outlast that of the gap-junctional current transient – thus phase delays caused by inhibition starting early in the cycle may in fact be caused more by their persistence until later in the cycle . In addition , the model shows a pronounced curvature in the phase delay region of the SPRF which was not noticeable in any experimental recordings . This might reflect the presence of other voltage-dependent conductances in real FS cells which effectively linearize this part of the relationship . The sharp discontinuity between phase delay and advance which emerges at high synaptic strengths is a result of the particular intrinsic biophysical properties and the nature of the synaptic perturbation . It appears to be related to the “class 2” nature of the FS neuron threshold [16] , and may be sensitively determined by the potassium conductance densities and kinetics [42] , [43] . It was not observed for example in a class 1 excitable Morris-Lecar model . The discontinuity is a critical decision point , or threshold , in the progression of the membrane potential towards spike initiation , at which hyperpolarization and depolarization both exert their maximal influence . The effect of this shape of SPRF is to ensure very rapid synchronization of the cell . Maximal phase shift occurs in the middle of the cycle when the phase difference is high - the postsynaptic cell either advances or delays its phase to achieve nearly immediate in-phase firing when detuning between pre- and postsynaptic cell is small . This extremely sharp midcycle transition is not observed in conventional phase-resetting relationships to weak brief inputs in these cells [37] , [44] , and is a consequence of the integration of the strong compound input . The piecewise nature of the SPRF , with the phase advance contributed exclusively by gap-junctional input , and the delay component contributed exclusively by chemical inhibition , mean that these two types of connection have complementary roles in synchronization: gap junctions are necessary to entrain the firing of the postsynaptic cell to a frequency higher than its preferred frequency , while inhibitory synapses are necessary to entrain firing to a frequency lower than the preferred frequency ( as seen in Figures 5 and 6 ) . This can be seen as follows . Let H be the phase difference between postsynaptic and presynaptic cells ( ) . The change in H over one period of the input , i . e . from input i to input i+1 , is: . Therefore , when entrainment is achieved , , and so if F>f , then , and if F<f , then . Using the SPRF to model entrainment assumes that the effect of each stimulus in the train is the same as if it was applied in isolation . The success of the SPRF in predicting entrainment shown here demonstrates that it is at least a good approximation for this purpose , and that the arithmetic of adding effects of multiple sequential synaptic inputs behaves reasonably linearly . The SPRF assumes that the entire dynamical state of the neuron may be represented by just a single number at any time , the phase , which would imply that its dynamical state always lies on a limit cycle , along which it is kicked instantaneously forwards and backwards by the synaptic inputs . The complex dynamics of a real neuron containing a large number of different voltage-dependent conductances distributed in a complex morphology , and the strong and non-instantaneous nature of the perturbation mean that this is a considerable simplification of the reality . An indication of whether the phase approximation is reasonably valid , is to test whether there is any higher-order phase resetting , i . e . changes in the interspike interval following that during which the input is applied , or in subsequent intervals . When we analysed second order shifts , we found that they were sometimes detectable , but very small in relation to the first-order SPRF ( See Figure S1 ) , in line with the short memory of FS cells for input conductance fluctuations [17] . FS cell firing is suspected to be directly and primarily responsible for producing gamma oscillations in the neocortex [6] , [7] , [8] . Different fine-scale subnetworks of mutually-exciting pyramidal cells in layers 2 or 3 , which are driven by specific subsets of local layer 4 inputs , appear to interact with other such subnetworks via the inhibitory interneuron network [45] . Synchronization of FS cells , therefore , may be essential for linking responses of pyramidal cells very rapidly to specific features of the synaptic input , as hypothesized to occur in sensory “binding” [2] . We have shown that the effect of conductance inputs which realistically mimic single synaptic connections on the phase of FS firing is very powerful , and is capable of entraining the postsynaptic cell even against strong noise . The strikingly sharp discontinuity between phase delay and advance in the SPRF causes a very rapid jump to nearly in-phase firing . The relative strengths of electrical and inhibitory components can vary greatly from connection to connection [12] , [13] , and some pairs of FS cells connected by gap junctions can synchronize their firing , while others cannot [14] . The strengths of these components will also vary dynamically . Electrical synapses can exhibit plasticity through G protein-coupled receptor activation , intracellular calcium and phosphorylation [46] , and the GABAergic connections show strong short-term depression [12] , [13] , [14] . These effects presumably help to shape the spatiotemporal dynamics of synchronous firing . The model that we introduce here could easily accommodate independent plasticity rules for inhibition and gap junctions , by additional rules for modifying the slopes of the corresponding regions of the SPRF . In addition to such modulation , the GABAA receptor is also the target of many important neuroactive drugs , such as benzodiazepines , barbiturates and ethanol . These will be expected to influence the shape of the SPRF , and the synchronization behavior of FS cells in the gamma frequency range . The SPRF , therefore , may be a useful tool for characterizing the action of such compounds on pathological network states treated by such drugs . Firing is considerably more variable in vivo than in vitro [47] , and it is important to consider the consequences of the SPRF in strong noise . The stochastic bifurcation analysis that we carried out ( Fig . 6 ) delineated a well-defined boundary between entraining and non-entraining frequencies , based on a qualitative change in the nature of the motion of the phase [32] ( see Methods ) . The stronger the noise , the smaller the frequency region of stochastic entrainment – in line with intuition , noise acts to break down synchronization . The strength of the noise effect in controlling the boundary of the synchronized region is not symmetrical around F – thus noise can effectively shift , as well as shrink the synchronized frequency band . In conclusion , the synaptic phase-resetting function of FS cells firing at gamma frequencies , as characterized here , is very well-suited to achieving rapid synchronization , and demonstrates complementary roles of the two types of synaptic connection in determining the frequency range of synchronization . It provides a simple yet surprisingly accurate model for predicting synchronization of these cells , and should be a useful component in network models aimed at understanding the complex spatiotemporal properties of locally-synchronized gamma-frequency firing in the cortex . 300 µm sagittal slices of somatosensory cortex were prepared from postnatal day 13–19 Wistar rats , using a vibratome ( DSK Microslicer Zero 1 , Dosaka EM , Kyoto ) , in chilled solution composed of ( in mM ) : 125 NaCl , 25 NaHCO3 , 2 . 5 KCl , 1 . 25 NaH2PO4 , 2 CaCl2 , 1 MgCl2 , and 25 glucose , oxygenated with 95% O2 , 5% CO2 gas . Slices were then held at room temperature for at least 30 minutes before recording . The tissue was visualized with an Olympus BX50WI upright microscope ( Olympus UK , London ) using infrared differential interference contrast videomicroscopy . During recording , slices were perfused with oxygenated solution identical to the slicing solution , at 31–35°C ( 8 cells analysed in detail ) or 23°C ( 4 cells ) . 10 µM 2- ( 3-carboxypropyl ) -3-amino-6- ( 4-methoxyphenyl ) -pyridazinium bromide ( SR95531; gabazine ) , 10 µM D-2-amino-5-phosphonopentanoic acid ( AP5 ) , and 10 µM 6-cyano-7-nitroquinoxaline-2 , 3-dione ( CNQX ) were usually added , to block chemical synaptic transmission mediated by GABAA , N-methyl-D-aspartic acid ( NMDA ) , and α-amino-3-hydroxy-5-methyl-4-isoxazole proprionic acid ( AMPA ) receptors , respectively . Whole-cell recordings were made from the somas of nonpyramidal neurons in cortical layers 2/3 , 4 , and 5 . Cells identified as FS neurons had a mean input resistance of 202±87 MΩ ( n = 12 ) . Data from 10 fast-spiking neurons ( taken from 8 animals ) were used for analysis , with a further 12 cells showing consistent results , but which were not complete enough for analysis . The number of synaptic phase-resetting functions with different parameters of the conductance perturbations ( see below ) which could be constructed for each cell was limited by the lifetime of the recording , typically 20 to 40 minutes . Patch pipettes of 3–5 MΩ resistance were pulled from borosilicate capillary glass and filled with an intracellular solution containing ( in mM ) : 105 K-gluconate , 30 KCl , 10 HEPES , 10 phosphocreatine , 4 ATP , 4 MgCl2 , and 0 . 3 GTP , adjusted to pH 7 . 3 with KOH . Current-clamp recordings were performed using an Axon Multiclamp 700A or in a few cases , an Axopatch 200A amplifier ( Axon Instruments , Foster City , CA ) . Membrane potentials were corrected for nulling of the liquid junction potential before seal formation . Signals were filtered with a four-pole low-pass Bessel filter at −3dB cutoff frequency of 5 kHz , sampled at 20 kHz , and recorded with custom software written in MATLAB ( The Mathworks , Natick , MA ) . Recorded neurons were stimulated using artificial conductance injection [25] , [26] , [48] . An effective conductance is inserted in the recorded cell by injecting a current I according to Ohm's law , I = g ( V−Erev ) , where g is the conductance , V is the membrane potential of the cell , and Erev is the reversal potential of the conductance . A conductance injection amplifier [49] or digital signal processing system ( SM-1 or SM-2 , Cambridge Conductance , Cambridge , UK ) [50] with response times of less than 200 ns or 10 µs respectively , were used to calculate and produce the current command signal in real time for the current-clamp amplifier . Steady trains of action potentials at gamma frequencies were elicited by steps of AMPA-receptor like ohmic conductance , reversing at 0 mV , to which perturbing conductances were added as follows . Stimuli that mimicked action potentials filtered through electrical synapses were generated . An action potential ( AP ) waveform was produced using a conductance-based model of an FS cell , identical to that of [41] , except that the leak conductance was reduced to better fit the stimulus-response curves of actual FS cells ( see Fast-spiking cell conductance-based model ( section below ) This AP waveform was then used as the time-varying Erev signal for a constant conductance ge , representing the electrical synapse . The conductance of a unitary synaptic GABA event was modelled as a difference of exponentials , where is the scaling amplitude of the inhibitory conductance , and was 7 ms , and was 0 . 5 ms . In compound stimuli , the start of the GABA event was delayed by 3 ms from the start of the simulated action potential to represent synaptic latency . The reversal potential EGABA was usually set to −55 mV [29] . Spike times were determined as the times of positive-going threshold crossings of the membrane potential at a threshold set at 10 mV below the peak of action potentials . The phase at which a stimulus was applied was calculated from the time elapsed from the preceding spike , relative to the unperturbed firing period . Variability of phase was characterized by the phase order parameter , or synchrony , which varied between 0 ( phases distributed uniformly between 0 and ) and 1 ( phases all identical ) . The change in phase ( ) caused by a stimulus was calculated as follows . Let be the phase reached at the moment of perturbation , the phase immediately after , the time after the previous spike at which the perturbation is applied , the time elapsed after the perturbation before the next spike , and the average interspike interval . Then , and . The synaptic phase-resetting function ( SPRF , see Fig . 2 ) was approximated by the piecewise linear relationship: ( 1 ) where conductance values are in nS , -α is the slope in the phase advance section , -β is the slope of the phase delay section , and is the breakpoint . SPRFs were fitted to experiments by least-squares , and using Grubbs' test for outliers , to delete occasional outlying points ( in most cases none , but no more than three per SPRF ) . Entrainment of periodic spiking to periodic stimulation was simulated by the noisy map describing the evolution of the phase from stimulus n to stimulus n+1: ( 2 ) where f is the stimulus angular frequency , F is the unperturbed ( natural ) angular frequency of the cell , and is a Gaussian-distributed noise term , with variance . The biophysical simulations of Fig . 7 were carried out using the model specified by [41] , modified slightly as described above ( see Conductance injection ) . Bifurcation points , where 1∶1 entrained fixed points of the map given by Eq . 2 appear , were solved for directly . To determine the points of stochastic bifurcation , we used the definition of [32] . The stochastic map of the phase between successive stimuli on a unit circle S is represented by a Markov operator p on the phase distribution , where is the conditional probability density function of the phase at stimulus i+1 , given a phase of at stimulus i . and the distribution of phase advances from stimulus n to stimulus n+1 according to:p is approximated by a stochastic transition matrix , and the onset of stochastic entrainment is defined by the point where the second eigenvalue of this stochastic transition matrix changes from real to complex . This definition of a stochastic bifurcation coincides with the deterministic case as the noise level approaches zero , is clearly defined even when the steady-state phase distribution hardly changes , and incorporates the dynamics of the phase: the first eigenfunction gives the stationary or invariant distribution of the phase , while the second eigenfunction can be thought of as forming the principal component of the average time course of relaxations from an initial random phase distribution . A model of fast-spiking cell membrane potential ( V ) dynamics was used ( as above for generating action potentials for gap-junctional stimulation ) which was slightly modified , with a different leak conductance , from that specified in Erisir et al . , 1999 [41] ( also correcting typographical errors in the published description of the model ) . Sodium ( Na ) , Kv1 ( K1 ) and Kv3 type potassium and static leak ( L ) conductances were used in a single electrical compartment of capacitance C , as follows ( units of mV for voltage , ms−1 for rates ) :Exactly the same conductance stimuli were applied to the model as to cells experimentally ( see Conductance injection section above ) .	Oscillations of the electrical field in the brain at 30–80 Hz ( gamma oscillations ) reflect coordinated firing of neurons during cognitive , sensory , and motor activity , and are thought to be a key phenomenon in the organization of neural processing in the cortex . Synchronous firing of a particular type of neuron , the inhibitory fast-spiking ( FS ) cell , imposes the gamma rhythm on other cells in the network . FS cells are highly interconnected by both gap junctions and chemical inhibition . In this study , we probed FS cells with a synthetic conductance stimulus which mimics the electrical effect of these complex connections in a controlled way , and directly measured how the timing of their firing should be affected by nearby FS neighbours . We were able to fit a mathematically simple but accurate model to these measurements , the “synaptic phase-resetting function” , which predicts how FS neurons synchronize at different frequencies , noise levels , and synaptic connection strengths . This model gives us deeper insight into how the FS cells synchronize so effectively at gamma oscillations , and will be a building-block in large-scale simulations of the FS cell network aimed at understanding the onset and stability of patterns of gamma oscillation in the cortex .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "neuroscience/neuronal", "signaling", "mechanisms", "neuroscience/theoretical", "neuroscience", "computational", "biology/computational", "neuroscience" ]	2010	Synchronization of Firing in Cortical Fast-Spiking Interneurons at Gamma Frequencies: A Phase-Resetting Analysis
JC virus is a member of the Polyomavirus family of DNA tumor viruses and the causative agent of progressive multifocal leukoencephalopathy ( PML ) . PML is a disease that occurs primarily in people who are immunocompromised and is usually fatal . As with other Polyomavirus family members , the replication of JC virus ( JCV ) DNA is dependent upon the virally encoded protein T-antigen . To further our understanding of JCV replication , we have determined the crystal structure of the origin-binding domain ( OBD ) of JCV T-antigen . This structure provides the first molecular understanding of JCV T-ag replication functions; for example , it suggests how the JCV T-ag OBD site-specifically binds to the major groove of GAGGC sequences in the origin . Furthermore , these studies suggest how the JCV OBDs interact during subsequent oligomerization events . We also report that the OBD contains a novel “pocket”; which sequesters the A1 & B2 loops of neighboring molecules . Mutagenesis of a residue in the pocket associated with the JCV T-ag OBD interfered with viral replication . Finally , we report that relative to the SV40 OBD , the surface of the JCV OBD contains one hemisphere that is highly conserved and one that is highly variable . There are now twelve known human polyomavirus members ( e . g . , [1] , [2] ) and particularly for immuno-compromised individuals , there is an increasing association between these viruses and human diseases ( reviewed in [3] , [4] , [5] ) . For example , JC virus ( JCV ) is the causative agent of Progressive Multifocal Leukoencephalopathy ( ( PML ) ; reviewed in [6] , [7] , [8] ) ; a demyelinating disease of the central nervous system [9] , [10] . JCV is also a major opportunistic infection associated with acquired immunodeficiency syndrome [11] , occurring in up to 5% of AIDS patients [12] . Further interest in JCV , which is present in approximately 50% of the general population [13] , stems from the fact that a promising new treatment of multiple sclerosis ( the monoclonal antibody Tysabri ) is known to be associated with the induction of PML ( reviewed in [8] , [14] , [15] ) . Studies have also suggested a possible association between infection with JCV and human brain and non-central nervous system tumors [16] , [17] . Unfortunately , there is no specific treatment for JCV . Central to the JCV life cycle is the replication of its genome . The JCV origin of replication has been the topic of numerous studies ( e . g . , [18] , [19] , [20] , [21] , [22] ) . The interactions between the origin with the viral initiator , large T-antigen ( T-ag ) , has also been explored ( e . g . , [23] , [24] ) . The T-antigens encoded by polyomaviruses are multi-domain , multifunctional proteins ( reviewed in [25] , [26] ) that form hexamers and double hexamers at origins of replication ( reviewed in [27] ) . Assays designed to monitor T-ag dependent JCV replication have been reported ( e . g . , [18] , [28] ) , including a cell free replication system [29] . However , theories regarding how JCV replication takes place are largely based on studies of the replication of Simian Virus 40 ( SV40 ) ( reviewed in [26] , [30] , [31] , [32] ) . For example , an in depth understanding of the enzymology of SV40 DNA replication was obtained following many elegant studies ( reviewed in [32] , [33] , [34] ) . Related studies have focused on the roles played by the T-ag during SV40 replication ( reviewed in [25] , [27] , [35] , [36] ) . Our laboratories have focused on the multiple roles played by the central origin-binding domain ( OBD ) of the SV40 T-ag during viral replication ( reviewed in [37] , [38] ) . Functions of the OBD include site-specific binding to GAGGC sequences in the origin ( [39] , [40] ) , promoting oligomerization of T-ag ( e . g . , via the B3 motif [41] , [42] ) , melting of the central region of the core origin [43] , binding to ssDNA at replication forks [37] , [44] , [45] and recruiting cellular initiation factors ( e . g . , [46] ) . Structural studies of T-ag have provided critical insights into how this single domain can engage in so many activities ( reviewed in [37] ) . For example , structures of the SV40 T-ag OBD established how the A1 & B2 loops in the OBD bind site-specifically to the GAGGC repeats in the central region of the viral origin ( i . e . Site II ) [42] , [47] , [48] , [49] . They also established how the same A1 & B2 loops engage other DNA structures ( e . g . , duplex DNA in a non-sequence specific manner [49] and ssDNA [44] , [45] ) . Crystallography studies also established that the SV40 T-ag OBDs can bind to a fork like DNA structure [45] . The latter observation was one reason for suggesting that the SV40 T-ag OBD is eventually positioned at the replication forks ( reviewed in [37] ) . The structures of additional domains of T-ag have provided many additional insights into the interactions needed to initiate viral DNA replication ( e . g . , [47] , [50] , [51] , [52] ) . For example , structures of the C-terminal helicase domain have greatly increased our understanding of how hexameric helicases catalyze DNA replication ( reviewed in [53] ) and how the helicase and OBDs work together to interact with ds DNA [47] . The initiation of JCV DNA replication is a central event during the viral life cycle [24] . The shared nucleotide sequence identity between the T-ag genes of SV40 and JCV is 71% [54] . Therefore , it is perhaps not surprising that SV40 T-ag recognizes and binds the JCV origin both in vivo and in vitro [24] , [55] , [56] , [57] . Studies show that the converse is not true; that is JCV T-ag is inefficient at promoting replication of an SV40 origin-containing plasmid [24] . Thus , while T-ag and the other proteins encoded by these viruses are highly homologous , they likely contain subtle but important structural differences . To examine these issues , we pursued structural and biophysical studies of the JCV T-ag OBD . The results from these studies suggest how the JCV T-ag OBD binds to the viral origin and its subsequent roles in oligomerization events . They also demonstrate that the JCV OBD contains a pocket that has not been described in previous structures of the polyomavirus OBDs . Collectively , these findings provide a preliminary molecular understanding of the initiation of JC virus replication . A luciferase based assay for studies of polyomavirus DNA replication was previously reported [75] . We developed a similar assay for measuring levels of JCV replication ( unpublished ) using the pCMV JC T-ag plasmid and a second plasmid containing the JCV origin of replication that was termed pJCV ori . Additional replication reactions were conducted with JCV T-ags containing point mutations introduced at selected residues using the QuikChange Kit ( ( Agilent ) ; with oligonucleotides containing the desired mutation . Western blots , conducted with the Pab 416 antibody against T-ag ( Santa Cruz Biotechnology ) , were used to determine whether a given point mutation disrupted T-ag's stability . The JCV OBD ( residues 132–261 ( Fig . 1A ) ) crystallized in three different forms that were termed form 1 , form 2 and form 3 ( Table 1 ) . Form 1 has two molecules in the asymmetric unit cell , and together the three crystals provide four independent structures of the JCV OBD . The four structures are very similar; a superposition of the four JCV OBD structures revealed root mean-squared deviations ( RMSDs ) of less than 0 . 5 Å . Form 3 , the highest resolution structure ( 1 . 32 Angstroms ) , is shown in Fig . 1B . The topology of the JCV OBD is a five-stranded antiparallel β-sheet sandwiched between two helices on either side ( Figs . 1 A and B ) . A superposition of the four JCV OBDs structures onto the DNA-free SV40 OBD structure ( the only other polyomavirus OBD to be solved in the absence of DNA [63] ) revealed an additional low RMSD ( between 0 . 85–0 . 88 Å over 121 Cα atoms ) . JCV OBD region B3 ( residues 216–220 [40] ) is poorly ordered in crystal forms 1 and 2 , but well ordered in form 3 ( Fig . 1C ) . This loop is also poorly ordered in several of the SV40 OBD structures [45] , [76] , [77] . B3 is well ordered in form 3 because tartrate ( a component in the crystallization mixture ) modified lysine 168 in a manner analogous to lysine acetylation . The carboxyl groups of the tartrate stabilized the B3 residues via a series of backbone hydrogen bonds . There are no previous reports indicating that JCV T-ag Lys168 is acetylated and further studies are necessary to determine if the observed modification of Lys168 is functionally important . Phylogenetic studies have established that the amino acid sequence for JCV T-ag is very similar to that of SV40 T-ag ( e . g . , [37] ) . Indeed , the amino acid sequence identity between the JCV and SV40 OBDs is 81 . 5% ( 106 amino acids identical/130 amino acids ( Fig . 2A ) ) . Given that the structures of the JCV and SV40 T-ag OBDs have both been determined , it was of interest to analyze these molecules in terms of the distribution of the identical , conserved and non-conserved residues ( Fig . 2B; identical ( blue ) , conserved ( pale pink ) , non-conserved ( magenta ) ) . As might be predicted , the interior of the molecule is highly conserved as are the A1 and B2 motifs involved in both DNA binding and interface formation ( discussed below ) ( Fig . 2B; right side ) . The non-conserved residues map primarily to the hemisphere that is opposite to the one containing the A1 and B2 loops ( Fig . 2B; left side . Certain of the conserved and non-conserved residues are indicated ) . The JCV origin of replication contains multiple high affinity GAGGC sequences that serve as binding sites for the JCV T-ag OBD [22] . The GAGGC binding sites are arranged as palindromic repeats in Site II and as direct repeats in Site I ( Figure 3A ) . It has been proposed that in the context of a full-length T-ag hexamer , the high local concentration of OBDs promotes their association [42] , [84] . To better understand how the JCV T-ag OBDs may assemble during oligomerization , we examined the interactions among the OBDs within the three crystal forms . As described in this section , the largest interface between adjacent molecules is the same in all three crystals . This was unexpected because the three forms belong to different space groups and have different cell dimensions ( Table 1 ) . In light of the findings derived from our structural studies , it was of interest to determine if particular residues in JCV T-ag are needed for replication . A luciferase-based assay for measuring levels of SV40 and HPV31 DNA replication was previously described [75] . This assay has been adapted for studies of JCV replication using plasmids containing JCV T-ag and the JCV origin of replication ( materials and methods ) . Initially , we used this assay to determine whether residues in the JCV OBD pocket are critical for DNA replication . Inspection of Fig . 9A establishes that a T-ag molecule containing a pocket mutation ( i . e . , F258L: its location in the pocket is shown in Fig . 7A ) does not support DNA replication . Moreover , it is clear from Fig . 9B that the F258L mutation does not cause destabilization of JCV T-ag . ( In contrast , two additional mutations in the JCV associated pocket ( i . e . , L199N and L199R ) did cause destabilization ( data not shown ) ) . In addition , we initiated studies designed to address whether certain “non-conserved” surface residues ( Fig . 2 ) play a role in JCV replication . Therefore , additional replication assays were conducted with T-ag molecules having the Q240A mutation . Inspection of Fig . 9A establishes that relative to wt JCV T-ag , T-ag molecules containing the Q240A mutation are greatly compromised in terms of their ability to support DNA replication . It is also apparent from Fig . 9B that the decreased ability of the Q240A mutant to support replication is not due to T-ag destabilization . We also analyzed the ability of the F190A mutant to support replication . Surprisingly , this mutant consistently supported higher levels of replication than wild type T-ag ( Fig . 9A ) ; a result that is not explained by increased expression of JCV T-ag ( Fig . 9B ) . Finally , no replication of the JCV origin containing plasmid was detected in the control reaction conducted in the absence of T-ag . The full-length T-ags encoded by both SV40 [86] , [87] , [88] and JCV [89] form hexamers and double hexamers on their respective origins of replication . Based on previous biochemical and structural studies , we proposed a model for SV40 T-ag's dynamic interactions with the viral origin and its subsequent oligomerization to form double hexamers [37] . One feature of this model is the proposal that following site-specific binding to the GAGGC sequences in the core origin , the OBD domains within SV40 T-antigen rearrange to form hexameric spirals ( e . g . , [63] , [76] , [77] ) . Spiral formation is also a feature of many of the other initiators that have been used as models for studies of the initiation of DNA replication ( e . g . , [90] , [91] , [92] , [93] ) . Therefore , spiral formation by replication initiators may be a general phenomenon ( reviewed in [94] ) . In view of the structures presented herein , we propose that the JCV T-ag OBD undergoes interactions with the JCV origin that are similar to those of the SV40 T-ag OBD ( reviewed in [37] ) . Regarding the initial binding of the OBD to the GAGGC sequences , our analysis of the JCV T-ag OBD structure indicates that the A1 & B2 loops mediate site-specific binding via a mechanism that is similar to that used by the SV40 OBD ( [42] , [47] , [49]; reviewed in [37] ) . Nevertheless , the ITC studies indicate that there are differences in the interactions between the JCV and SV40 T-ag OBDs and origin sequences . For example , the binding of the JCV T-ag OBD to an oligonucleotide containing the JCV Site II is weaker than the SV40 OBD/Site II interaction [79] ( 298 . 6 nM verses 93 . 5 nM ) . Related ITC studies demonstrate that the JCV T-ag OBD binds to the GAGGC containing Site I regulatory region with a much higher affinity than Site II ( Kds of 18 . 3 nM and 298 . 6 nM; respectively . The SV40 T-ag OBD also preferentially bound to Site I [76] ) . Why the JCV and SV40 T-ag OBDs have different affinities for Site II , and such a wide range of affinities for different GAGGC containing substrates , is not known . Of interest , the B2 regions in the OBDs encoded by JCV and SV40 are identical [40] and there is only one amino acid difference in the A1 regions ( H148 in the SV40 OBD is Q149 in the JCV OBD ) . Therefore , pronounced sequence differences between the A1 & B2 motifs do not explain the observed differences in affinity; however , subtle structural differences in DNA , the OBDs , or both may play a role . Previous SV40 based studies have also established that sequences flanking the individual GAGGC sites play a significant role in modulating OBD binding affinities [95] . Thus , additional studies , including the co-structures of the JCV OBD with oligonucleotides derived from Site II and Site I , are needed to explain the observed differences in OBD affinities for origin sub-fragments . Finally , the full-length T-ag's from JCV & SV40 also have different affinities for Site II [23] , [24] . The ITC studies suggest that the differences in the affinities are , at least in part , a function of the OBDs . The ITC experiments also indicate that four JCV OBDs bind simultaneously to the four GAGGC sequences in Site II . However , in the context of full-length T-ag it is unlikely that all four pentanucleotides are initially bound by the OBDs . This conclusion is based on previous biochemical experiments with SV40 T-ag [96] , [97] and structural studies that indicate that once the helicase domain has oligomerized , the shortness of the spacer that links the helicase domain to the OBD restricts OBD binding to only the most proximate pentanucleotide [47] . The subsequent stage ( s ) during the initiation process at which the initially unbound pentanucleotides are bound by the SV40 , and presumably JCV , OBDs remain to be determined . Moreover , studies of both murine [98] and Merkel [79] polyomaviruses have established that in those systems only three pentanucleotide repeats are necessary for DNA replication; further evidence that the interactions of polyomavirus T-ags with the pentanucleotides in Site II are complex . How polyomavirus T-ags transition from their sequence specific binding mode to fully assembled hexamers and double hexamers is not understood . While the OBDs are monomeric in solution ( e . g . [48] ) , it has been proposed that in the context of T-ag hexamers and double hexamers , the high local concentration of OBDs will promote their association ( [42] , [84] ) ; reviewed in [37] ) . Consistent with this possibility , our previous structures of the SV40 T-ag OBD established that it forms a hexameric spiral within the crystal [63] , [77] . Therefore , it is of interest that our current studies have established that the JCV T-ag OBD also forms a spiral in the crystal . As in the SV40 T-ag OBD spiral [63] , the monomers in the JCV T-ag OBD spiral are arranged in a head-to-tail manner , and the A1 loops are in the DNA-free or “retracted conformation” ( reviewed in [37] ) . An additional common feature of the JCV and SV40 spirals is that they contain a very positively charged central channel that could interact with DNA in a non-sequence specific manner ( data not shown ) . Nevertheless , the spirals formed by the JCV and SV40 T-ag OBDs are not identical . For example , the JCV “spiral” contains 4 OBDs per turn while the SV40 OBD spiral has 6 OBDs/turn ( diagrammed in Fig . 8C ) . In addition , the JCV T-ag OBD forms a right-handed spiral , whereas the SV40 forms a left-handed one . These observations raise the question , “how can different spirals form from T-ag OBDs utilizing very similar interfaces ? ” Comparison of the existing spiral structures for the SV40 and JCV T-ag OBDs suggest a common “interface based” model for formation of the observed higher order structures . According to this model , the interface acts like a joint or pivot point and differences in the rotational and translational components of the interface promote the formation of the structures observed to date . For example , in the crystallographic spirals , the angles between the interfaces in the JCV and SV40 T-ag OBDs are very different ( i . e . , ∼90° and 60°; respectively ) . In addition , for a spiral to occur , instead of a flat ring structure , there is a requisite translational component ( “rise” ) to the interface ( the SV40 spiral has a rise of ∼6 Å [63] , while the rise in the JCV OBD spiral is ∼9 Å ) . The direction of the translation component relative to the principal rotational axis ( i . e . , up or down ) results in either a left or right-handed spiral ( Fig . 8C; legend ) . Furthermore , in the context of T-ag hexamers and double hexamers , the interactions between the OBDs are likely to be highly dynamic . Support for this postulate includes the relatively small size of the interfaces observed in the crystal structures and previous EM based studies showing multiple orientations of the SV40 T-ag OBDs [85] . In summary , plasticity in the OBD/OBD interface may contribute to the multiple higher-order conformations adopted by the OBD . Nevertheless , it is not known whether the tetrameric JCV T-ag OBD spiral forms in vivo or whether it can rearrange into a hexameric OBD spiral that is analogous to the one formed by the SV40 T-ag OBD . However , given the dynamic nature of the domains within T-ag , it is possible that under certain conditions ( e . g . , following assembly of the hexameric helicase domain ) , the tetrameric JCV T-ag OBD spiral rearranges to accommodate two additional OBDs . The C-terminus of the JCV T-ag OBD contains a pocket into which the A1 and B2 residues are inserted . Furthermore , our studies have established that pocket residue F258 is necessary for JCV replication . However , whether this pocket is a general feature of polyomavirus OBDs is not known . The T-ag OBD-DNA co-structures derived from Merkel ( PDB entry 3QFQ [79] ) and murine polyomavirus ( PDB entry 4FB3 [98] ) did not contain suitable electron density for tracing of the residues in the C-termini of the OBDs . Therefore no clear pocket was observed in these structures and it is concluded that there is some flexibility in the C-terminal OBD residues . Analyses of SV40 OBD structures revealed that they contain a groove in the same location , but it is not as pronounced as the one in the JCV OBD structure . Regarding evidence for the OBD pocket in larger T-ag structures; a co-structure of a SV40 T-ag dimer , containing both the OBD and the helicase domain ( PDB entry 4GDF ) interacting with DNA , was recently reported [47] . This structure revealed two completely different orientations of the linker region connecting the two domains . In the structure in which the OBD is bound to pentanucleotide 1 , the linker points away from the OBD and the relatively shallower groove is observed . In the second or “hidden site” , the linker bisects the putative pocket . Together , these observations indicate that the “pocket” in SV40 T-ag may be part of a dynamic structure . However , additional structural studies are needed to further characterize the pocket in the SV40 and JCV T-ag OBDs . Previous studies have also established that the SV40 T-ag OBD serves as a module for binding cellular proteins ( reviewed in [38] ) . For example , the RPA 70AB domain was reported to bind to the T-ag OBD via interactions that include those with R154 [46] . Furthermore , the Nbs1 subunit of the MRN complex binds to the OBD [99] . Given the central roles played by the OBDs during viral DNA replication ( reviewed in [37] ) , the surfaces on the OBDs that interact with these and related cellular replication factors have likely been conserved . Therefore , it is of interest that the JCV and SV40 T-ag OBDs contain one surface that is highly conserved . This surface contains the DNA binding A1 and B2 loops , but also many additional conserved residues that may be involved in binding to cellular proteins ( e . g . , R154 associated with RPA recruitment ) . However , it is also apparent that the opposite hemisphere contains the majority of the non-identical residues and certain of these residues ( i . e . , Q240 ) are required for JCV replication . These variable regions may simply reflect genetic drift . Alternatively , they may be binding surfaces for cellular proteins encountered in the very different cell types in which these viruses replicate ( i . e . , monkey kidney cells needed for SV40 replication versus human glial cells needed for JCV replication ) . Finally , the F190A mutation leads to higher levels of JCV DNA replication . The biochemical basis for this increase is unknown and subsequent studies are needed to address this issue . Nevertheless , a sequence comparison of JCV , SV40 and BK reveals that while JCV T-ag has a bulky aromatic amino acid at position F190 , the T-ags from SV40 and BK contain less bulky residues at comparable positions ( SV40: S189; BKV: C191 ) . The alanine substitution at JCV T-ag residue F190 introduces an amino acid that requires less space than a phenylalanine . Therefore , the F190A T-ag mutant is more analogous to the SV40 and BKV T-ags and this may be related to the observed increase in DNA replication . The initiation of JCV DNA replication , and the regulation of this process , is a complicated process . It is apparent that many additional structures will have to be determined before a molecular understanding of the initiation of JCV replication is obtained . Nevertheless , the individual structures of the proteins involved will provide considerable useful information , including potential targets for drug design , such as the pocket within the JCV T-ag OBD described herein .	Polyomaviruses have been invaluable tools for biomedical research into basic cellular processes . It is becoming increasingly clear , however , that members of this family are also involved in human diseases , particularly among the immunocompromised and the elderly . The subject of this study , the JC virus ( JCV ) , is a member of this family and the causative agent of a brain disease termed Progressive Multifocal Leukoencephalopathy ( PML ) , a disease that is often fatal and for which there is no cure . Herein we present the high-resolution crystal structure of the origin binding domain ( OBD ) from the JCV initiator protein large T-antigen . Furthermore , we propose a molecular model for the oligomerization of the JCV T-antigen OBD that is based upon the crystal structure . We also report a novel pocket that modeling studies suggest is available when the OBD is site-specifically bound to DNA and therefore may represent a possible starting point for structure-based drug design .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "biochemistry", "dna", "replication", "nucleic", "acids", "proteins", "virology", "protein", "structure", "dna", "biology", "dna-binding", "proteins", "microbiology", "viral", "replication", "biophysics" ]	2014	Insights into the Initiation of JC Virus DNA Replication Derived from the Crystal Structure of the T-Antigen Origin Binding Domain
Bothrops atrox snakes are the leading cause of snake bites in Northern Brazil . The venom of this snake is not included in the antigen pool used to obtain the Bothrops antivenom . There are discrepancies in reports on the effectiveness of this antivenom to treat victims bitten by B . atrox snakes . However , these studies were performed using a pre-incubation of the venom with the antivenom and , thus , did not simulate a true case of envenomation treatment . In addition , the local lesions induced by Bothrops venoms are not well resolved by antivenom therapy . Here , we investigated the efficacy of the Bothrops antivenom in treating the signs and symptoms caused by B . atrox venom in mice and evaluated whether the combination of dexamethasone and antivenom therapy enhanced the healing of local lesions induced by this envenomation . In animals that were administered the antivenom 10 minutes after the envenomation , we observed an important reduction of edema , dermonecrosis , and myonecrosis . When the antivenom was given 45 minutes after the envenomation , the edema and myonecrosis were reduced , and the fibrinogen levels and platelet counts were restored . The groups treated with the combination of antivenom and dexamethasone had an enhanced decrease in edema and a faster recovery of the damaged skeletal muscle . Our results show that Bothrops antivenom effectively treats the envenomation caused by Bothrops atrox and that the use of dexamethasone as an adjunct to the antivenom therapy could be useful to improve the treatment of local symptoms observed in envenomation caused by Bothrops snakes . Snake bites are a public health problem . They are associated with poverty and occupation and affect the population of rural areas [1 , 2] . Despite no longer being listed on the World Health Organization list of neglected tropical diseases , snake bites continue to affect thousands of victims every year [3 , 4] . In Brazil , approximately 26 , 000 cases occur per year , of which an average of 0 . 39% cause death and 1 . 72% heal with sequelae [4 , 5] . Northern Brazil is the region with the highest incidence of cases/100 , 000 inhabitants , and the state of Pará has the second-highest rate in the country , after Roraima state [5] . Bothrops atrox snakes are distributed through the South American Amazon region [6 , 7] . In Northern Brazil , this species causes the majority of bites , but its venom is not used as an antigen in the production of the Bothrops antivenom . The antivenom produced in Brazil is obtained using a pool of antigens containing the following venoms: 50% Bothrops jararaca , 12 . 5% B . jararacussu , 12 . 5% B . alternatus , 12 . 5% B . moojeni , and 12 . 5% B . neuwiedi . This antigen formulation was achieved studying the cross-reactivity of monovalent antivenoms to venoms of ten Bothrops species [8] . Some studies have suggested that the Bothrops antivenom produced in Brazil does not effectively neutralize the B . atrox venom [9 , 10] , but other studies have shown that the Bothrops antivenom can neutralize the toxins of this venom [11–14] . All these studies were performed using a prior incubation of the venom with the antivenom , which does not represent the actual situation where the treatment with antivenom follows the envenomation after the onset of the primary venom-induced effects . A clinical study also showed the efficacy of the Bothrops antivenom in the treatment of Bothrops atrox snakebites [15] . Similar to other Bothrops venoms , B . atrox bites cause hemostatic disturbances and local reactions including edema , hemorrhage , and necrosis , which can progress to tissue losses that lead to sequelae [15 , 16] . The specific antivenom therapy is the only recommended procedure for treating envenomation by snake bites [17 , 18] . Neutralization of the toxins by the antivenom results in an efficient restoration of the coagulation factors but does not reverse the locally established lesions or the inflammatory mediators released by the damaged tissue [19–21] . Studies have shown that eicosanoids are the major mediators of the inflammatory edema induced by Bothrops venoms [22–27] . One study showed that the use of a combination of dexamethasone with antivenom reverses the Bothrops jararaca venom-induced edema when administered after venom injection in mice . However , the individual treatments with antivenom or with dexamethasone did not decrease the edema [27] . This combined therapy was also effective in reducing the myotoxic effect induced by the venoms of B . jararaca and B . jararacussu [28] . In this study , we tested the efficacy of Bothrops antivenom to treat the signs and symptoms of experimental B . atrox venom-induced toxic effects in mice . We also assessed whether combining the antivenom with dexamethasone would be beneficial to treat the pathophysiological effects of this envenomation . Our results show that the Brazilian Bothrops antivenom effectively reverses the coagulopathy induced in mice by Bothrops atrox venom . Furthermore , the combination of dexamethasone with the antivenom improved the reversal of the inflammatory edema and the regeneration of the skeletal muscle that was damaged by the venom . All the experimental protocols used in this work were reviewed and approved by Institute Butantan Animal Care and Use Committee ( protocol n° 1249/14 ) . These procedures are in accordance with standards outlined by Brazilian laws for use of experimental animals , and with ethical principles adopted by the Brazilian College of Animal Experimentation . Venom of wild Bothrops atrox ( BaV ) was obtained from 22 snakes of both sexes , collected in Santarém , Pará , Brazil ( SISBio license 32098–1 ) . Each venom sample was extracted individually and freeze-dried , after which the samples were pooled in equal proportions , and stored at –20°C until use . Bothrops antivenom produced by Instituto Butantan ( batch 0609/68/C ) and injectable dexamethasone ( Aché , Brazil ) were used . One mL of Bothrops antivenom neutralizes 5 mg of Bothrops jararaca venom . Male Swiss mice ( 18–22 g body weight ) were obtained from the animal housing facility of the Instituto Butantan . The animals were maintained in a 12:12 hour light:dark cycle and received water and food ad libitum . Before evaluating the effectiveness of treatments in local manifestations of Bothrops atrox envenomation , the minimum doses of BaV required to induce the pathophysiological activities studied ( edema , hemorrhage , dermonecrosis and muscle damage ) were determined . The details of the methodologies for assessing these toxic activities of the BaV were previously described [12 , 29] . After determining the minimum doses required to induce the pathophysiological manifestations to be studied , the challenge doses were established for use in the protocols for treatments to be administered 10 or 45 minutes after the venom injection . The treatment groups consisted of ( 1 ) antivenom ( AV ) administered intravenously ( 0 . 2 mL/animal ) ; ( 2 ) dexamethasone ( Dexa ) administered intraperitoneally ( 1 mg/kg ) ; ( 3 ) the combined administration of AV and Dexa protocols; and ( 4 ) control group that was untreated ( 5 mice/group ) . The doses of antivenom and Dexa were those used in a previous study [27] . Edema was induced by injecting 0 . 75 μg of BaV in 30 μL of sterile saline into one hind foot pad of each mouse , and the same volume of saline was injected into the contralateral paw . The volume of both paws was verified by plethysmometry 1 , 2 , 4 , 6 and 24 hours after the venom injection . The edematogenic activity results were expressed as the percent difference between the venom- and saline-injected paws . Hemorrhage was induced by injecting 3 . 3 μg of BAV in 100 μL of sterile saline intradermally in the ventral region of the abdomen . Two hours after the injection , mice were euthanized , their skins were removed and the diameter of the hemorrhagic halo on the inner side of the skin was determined for each group . The dermonecrosis diameter was obtained in a similar protocol . The mice were injected intradermally in the ventral region with 30 μg of BAV in 100 μL of sterile saline . At 48 hours after injection , the mice were euthanized , their skins were removed and the diameter of the dermonecrotic halo on the inner side of the skin was determined for each group [29] . Myonecrosis was evaluated by quantifying the plasma creatine kinase ( CK ) activity according to the instructions for the CK-NAC Liquiform ( LabTest , Brazil ) . Each mouse received an i . m . injection of 40 μg of BaV in 40 μL of saline into the gastrocnemius muscle . At 3 and 24 h after the venom injection , blood samples were collected by an orbital puncture for determination of the serum CK levels , which were expressed as units/L . Muscle regeneration was evaluated by the quantification of the muscle tissue creatine kinase ( residual CK ) activity and by histological analysis . Each mouse received an i . m . injection of 40 μg of BaV in 40 μL of saline into the right gastrocnemius muscle and the same volume of saline into the left gastrocnemius muscle followed by the treatment protocols . Groups of mice were euthanized 1 , 4 , 7 , 14 , 21 and 28 days after the venom injection , and both gastrocnemius muscles were collected , weighed , and homogenized in 4 mL phosphate buffered saline ( pH 7 . 4 ) containing 0 . 1% Triton X-100 . After the centrifugation ( 2100 g/5 minutes ) , the supernatant was collected for determination of the residual CK ( CK-NAC Liquiform , LabTest ) . The results are expressed as the percentage of CK obtained in the venom-muscle injected ( residual CK ) compared with the CK content of the control contralateral muscle . In another set of experiments , the injected muscles were collected for histological analysis . They were fixed in Bouin’s solution , embedded in paraffin , sectioned and stained with hematoxylin and eosin ( HE ) for light microscopic analysis . Four groups of mice were envenomated as described for the analysis of the myotoxic activity ( 40 μg/40 μL , i . m . ) and treated 45 minutes later . Six hours after the envenomation , blood was collected ( 9:1 , v:v ) in 13 mmol sodium citrate with 2% of Bothrops antivenom to neutralize venom in the sample [19] . The fibrinogen concentration was evaluated in citrated plasma [30] and the platelet count was determined in EDTA-anticoagulated blood using an automated cell counter ( Mindray BC-2800 Vet , Nanshan , Shenzhen , China ) . One-way ANOVA followed by the Tukey test was used to evaluate significant differences between the treated and non-treated groups . The data are expressed as the means ± SEM , and differences were considered statistically significant when the p values were < 0 . 05 . To evaluate the effects of the treatments on the systemic disturbances induced by B . atrox venom , we assayed the plasma fibrinogen levels and the numbers of platelets . Our results showed that there was significant fibrinogen consumption in the untreated animals and those treated with dexamethasone . Six hours after the treatments with antivenom or antivenom + dexamethasone , the fibrinogen levels increased to reach hemostatic levels , and in both cases , the values were significantly higher than those of the untreated animals and those treated only with dexamethasone ( Fig 7A ) . Severe thrombocytopenia was observed in the untreated or dexamethasone-treated groups . Similar to the results for fibrinogen consumption , the numbers of platelets in the groups of mice treated with antivenom or antivenom + dexamethasone were restored ( Fig 7B ) . Our results showed that the Bothrops antivenom can counteract the systemic and local signs and symptoms of Bothrops atrox envenomation in mice . Bothrops atrox is the species of snake that causes the most bites in humans in Northern Brazil and Amazon region . However , B . atrox venom is not included in the pool of antigens that is used to obtain the antivenom . Some in vitro studies have suggested that the Bothrops antivenom does not effectively neutralize the Bothrops atrox venom [9 , 10] . In contrast , pre-clinical , antivenomic , and clinical studies have shown that the Bothrops antivenom effectively neutralizes B . atrox venom [12–15] . However , with the exceptions of the clinical [15] and the antivenomic [13] studies , all other studies of Bothrops venom have evaluated its biological and enzymatic activities , including the lethal , hemorrhagic , coagulant , defibrinating , and myotoxic effects , after preincubating the venom with the antivenom . In the present study , we used a model that mimics treatment of a snake envenomation , and our data corroborate those studies that showed the efficacy of the antivenom . Thus , the groups treated with the antivenom demonstrated a significant reduction of all the effects studied except for the local hemorrhage . The efficacy of the antivenom against the local symptoms was more obvious when it was applied soon after the experimental envenomation , as has also been observed in the clinical studies [19] . In the case of the edema , the antivenom was effective even when applied 45 minutes after the injection of the B . atrox venom into the footpads of mice . It is interesting to note that following envenomation with B . jararaca venom , the main antigen used to obtain the Bothrops antivenom , the edema was not affected when the antivenom was applied 45 minutes after the experimental envenomation [27] . All the animals injected with B . atrox venom demonstrated signs and symptoms of the envenomation , such as edema , local hemorrhage , fibrinogen consumption , and thrombocytopenia , at the times that the treatments were applied . Once the antivenom had neutralized the venom , six hours after the envenomation , the fibrinogen levels and the platelet numbers were within hemostatic levels , although lower than the levels in the non-envenomed group . Clinicians use the restoration of the hemostatic parameters as indicators of the efficacy of the antivenom therapy [31] . However , the inflammatory process and the healing of established lesions have longer time-courses and involve the participation of some endogenous mediators that are not influenced by the antivenom [21] . Eicosanoids are main mediators of inflammatory response induced by Bothrops venoms [22–27] . From this point of view , the use of anti-inflammatory drugs in combination with the antivenom therapy could reduce the time for the tissue regeneration and potentially improve the functions of the healed tissue . With this rationale , the use of dexamethasone combined to Bothrops antivenom presented a better result than specific inhibitors of cyclooxygenase or lipoxygenase pathways , such as indomethacin or NDGA , in reversing the inflammatory edema induced by the Bothrops jararaca venom into the paw of mice [27] . Edema is the most common inflammatory sign of envenomation by Bothrops snake bites in humans . It can affect the entire bitten limb and can evolve to a compartmental syndrome that increases the risk of permanent sequelae [20] . For the edema induced by B . atrox , the antivenom + dexamethasone combination was the most efficient treatment when applied 10 or 45 minutes after the envenomation . Nevertheless , in contrast to the results obtained using B . jararaca venom , the present results showed that this therapeutic combination caused a significant reduction of the edema in the second hour and subsequently after envenomation [27] . These data suggest that in addition to eicosanoids , other endogenous mediators such as histamine may participate in the snake venom-induced inflammatory edema [32] . Having confirmed the efficacy of this treatment for edema , we evaluated whether it could affect other toxic effects induced by this venom such as local hemorrhage , dermonecrosis , and myonecrosis . We also determined whether this therapeutic strategy could have an impact on the regeneration of damaged muscle and some hemostatic parameters . The local hemorrhage induced by viperid venoms is one of the most rapid pathological symptoms that occurs after contact of the venom with the connective tissue [24 , 33 , 34] . This effect is induced by Snake Venom Metalloproteinases ( SVMPs ) that are present in these venoms [35] . In the venom of the B . atrox , SVMPs are the most abundant toxins , representing more than 50% of the crude venom [13 , 14 , 36] . These early effect of the venom could explain the lack of efficacy of an antivenom , even when applied 10 minutes after venom injection , despite the fact that the antivenom has antibodies that can neutralize these SVMPs when incubated with the venom [12] . Treatment with dexamethasone , alone or combined with the antivenom , also did not affect the hemorrhagic lesion induced by this venom . This result is consistent with the fact that eicosanoids do not participate in the pharmacological mechanisms of the local hemorrhage induced by Bothrops jararaca venom [24] . The SVMPs also participate strongly in the dermonecrosis by acting directly on the components of the connective tissue [37] . The SVMPs induce their effects by generating cytokines , such as TNF-α , through the cleavage of pro-TNF-α [38 , 39] . Nevertheless , our results have shown that all the treatments were effective in reducing the necrotic area of the injected skin when applied 10 minutes after the venom injection , and none of them were effective when applied 45 after the experimental envenomation . These data indicated that the antivenom could neutralize the effects of the venom when administered early after the envenomation , which suggested that the onset of the necrotic lesions occurs later than that of the hemorrhagic lesions and that neutralization of SVMPs by antivenoms could prevent the development of necrosis . Furthermore , the group treated with dexamethasone demonstrated a significant reduction of the necrosis . This effect could be due to the corticoid inhibitory activity on the production of TNF-α [40 , 41] , corroborating previous observations that dexamethasone can prevent necrosis when applied up to 15 minutes after the injection of the B . jararaca venom in mice [42] . When injected 10 or 45 minutes after the venom , the antivenom alone or in combination with dexamethasone was efficient in reducing the muscle lesions induced by the intramuscular injection of the B . atrox venom . The creatine kinase level in the group treated with antivenom + dexamethasone was not different from that observed in the animals that were treated only with the antivenom . In this case , the neutralization of toxins by the antivenom seemed to prevent the progression of the muscle lesion and suggested that myotoxins in the B . atrox venom directly induce this muscle damage independently of the inflammatory response as was described for the venom of B . asper [43] . However , other studies have shown that the combination of antivenom and dexamethasone could reduce the muscle lesions induced by B . jararaca or B . jararacussu venoms , which suggests that an inflammatory response participates in this process [28] . Nonetheless , the inflammatory process is fundamental for appropriate muscle regeneration . The absence of the inflammatory response or the presence of an exacerbated inflammatory process may lead to poor muscle regeneration or a functional decrease in the regenerated tissue due to the replacement of muscle cells by adipose or fibrotic tissue [44 , 45] . In this context , the use of the combination of antivenom+dexamethasone proved to be the best therapeutic approach among the experimental treatments used in the present study . After the establishment of the muscle injury , from the 7th through the 21st day , only the group treated with the combination of antivenom and dexamethasone showed a significant difference from the untreated control group , which indicated a more effective protection , resulting in a faster recovery of damaged muscle . Morphologically , all groups demonstrated a time course of regeneration similar to that described for other types of muscle lesions [46] , but in the groups treated with dexamethasone , a less intense inflammatory infiltrate was observed . Despite some controversies regarding the use of corticoids in muscle regeneration [47–49] , our results suggest that the use of the combination of dexamethasone and the antivenom was beneficial to the regeneration of the muscle tissue . The use of a single dose of dexamethasone in this study could explain the difference from other studies that used anti-inflammatory drugs over a prolonged period , which affected the muscle regeneration [47–49] . Concerning the hemostatic parameters , the use of the combination of dexamethasone and antivenom did not hinder the restoration of the fibrinogen or platelet levels . It has been suggested that the combination of a thrombin inhibitor and dexamethasone could prevent the fibrinogen and platelet depletion in an experimental model of disseminated intravascular coagulation [50] . In the envenomation by Bothrops snake bites , the toxins act directly on fibrinogen or through other factors of the blood coagulation cascade including the endogenous formation of thrombin to induce the coagulopathy [51] . In conclusion , our results show that the Bothrops antivenom produced in Brazil is useful for the treatment of the experimental envenomation caused by B . atrox venom , and also suggest that the use of dexamethasone associated to antivenom reduce the time for the recovery of the edema and muscle damage . Further clinical trials are needed to confirm both , the efficacy of the Bothrops antivenom to treat Bothrops atrox snake bites from different Amazon regions , and the benefit of the use of dexamethasone associated with the antivenom therapy to treat Bothrops snake bites .	Bothrops atrox is the dominant species responsible for accidental human snake bites in Northern Brazil . The efficacy of antivenom therapy to correct the systemic disturbances , including hemostatic disorders , caused by Brazilian Bothrops is well known . However , two fundamental issues need to be addressed in this region . ( 1 ) There are concerns regarding the effectiveness of the antivenom to treat Bothrops snake bites in this region since Bothrops atrox venom is not used as an antigen to obtain the Bothrops antivenom in Brazil , and ( 2 ) the efficacy of the antivenom therapy in reversing local injuries induced by Bothrops venoms is low . Thus , our study aimed to assess the effectiveness of antivenom therapy alone or in combination with dexamethasone to treat experimental envenomation induced by Bothrops atrox venom in mice . Our results showed that the Brazilian Bothrops antivenom effectively reversed the systemic disturbances caused by this envenomation and combining the antivenom therapy with dexamethasone accelerated the regression of inflammatory edema and the regeneration of skeletal muscle that was damaged by the venom .	[ "Abstract", "Introduction", "Materials", "and", "methods", "Results", "Discussion" ]	[ "medicine", "and", "health", "sciences", "toxins", "pathology", "and", "laboratory", "medicine", "fibrinogen", "immunology", "tropical", "diseases", "vertebrates", "animals", "toxicology", "toxic", "agents", "muscle", "regeneration", "developmental", "biology", "organism"...	2017	Experimental Bothrops atrox envenomation: Efficacy of antivenom therapy and the combination of Bothrops antivenom with dexamethasone
Intracellular Local Ca releases ( LCRs ) from sarcoplasmic reticulum ( SR ) regulate cardiac pacemaker cell function by activation of electrogenic Na/Ca exchanger ( NCX ) during diastole . Prior studies demonstrated the existence of powerful compensatory mechanisms of LCR regulation via a complex local cross-talk of Ca pump , release and NCX . One major obstacle to study these mechanisms is that LCR exhibit complex Ca release propagation patterns ( including merges and separations ) that have not been characterized . Here we developed new terminology , classification , and computer algorithms for automatic detection of numerically simulated LCRs and examined LCR regulation by SR Ca pumping rate ( Pup ) that provides a major contribution to fight-or-flight response . In our simulations the faster SR Ca pumping accelerates action potential-induced Ca transient decay and quickly clears Ca under the cell membrane in diastole , preventing premature releases . Then the SR generates an earlier , more synchronized , and stronger diastolic LCR signal activating an earlier and larger inward NCX current . LCRs at higher Pup exhibit larger amplitudes and faster propagation with more collisions to each other . The LCRs overlap with Ca transient decay , causing an elevation of the average diastolic [Ca] nadir to ~200 nM ( at Pup = 24 mM/s ) . Background Ca ( in locations lacking LCRs ) quickly decays to resting Ca levels ( <100 nM ) at high Pup , but remained elevated during slower decay at low Pup . Release propagation is facilitated at higher Pup by a larger LCR amplitude , whereas at low Pup by higher background Ca . While at low Pup LCRs show smaller amplitudes , their larger durations and sizes combined with longer transient decay stabilize integrals of diastolic Ca and NCX current signals . Thus , the local interplay of SR Ca pump and release channels regulates LCRs and Ca transient decay to insure fail-safe pacemaker cell operation within a wide range of rates . The Coupled Clock Theory [1] predicts an important role for the sarcoplasmic reticulum ( SR ) in the pacemaker function of the sinoatrial node cells . Changes in the SR Ca pumping rate ( Pup ) alone can regulate the pacemaker rate within the entire physiological range , including extremely low and high rates . The important functional role of Ca pump has been also suggested in experimental studies [2 , 3] and in more recent numerical model simulations [4 , 5] . This regulation is linked to the Local Ca releases ( LCRs ) , which are SR discharges below the cell membrane [6 , 7] , and also to Ca transient decay [8] . The LCRs occur when SR gets refilled with Ca to a threshold level allowing spontaneous Ca release . The refiling time and the timing of the LCR occurrence are controlled by SERCA ( SR Ca pump molecules ) together with cell Ca available for pumping and intra-SR Ca diffusion . LCRs , in turn , accelerate the diastolic depolarization via activation of NCX current ( review [9] ) . The LCRs are generally larger in geometrical extent than isolated Ca sparks seen in ventricular myocytes , appear to be propagated locally , but do not form global waves . In simulations we generally use the term “spark” to refer to regenerative releases that appear to originate in a single couplon . While complex Ca releases also happen in ventricular myocytes ( such as abortive waves or macrosparks , review [10] ) , they are not observed under normal physiological conditions . The release distinction is probably due to the underlying RyR distribution , which in sinoatrial node cells consists of irregularly placed sub-sarcolemmal clusters of various sizes ( dubbed “hierarchical clustering” [5] ) as opposed , for example , to sarcomeric intracellular arrays in ventricular myocytes . The local geometry of Ca diffusion probably also plays a role . Finally , in sinoatrial node cells , in contrast to ventricular myocytes , Ca cycling proteins , including RyRs and phospholamban ( regulating SERCA ) , are phosphorylated in the basal state by PKA [11] . The functional consequence of the phosphorylation is higher rates of Ca pump and release resulting in occurrence of the diastolic LCRs [11] and their attendant diastolic NCX current [7 , 12 , 13] , lacking in ventricular myocytes under normal conditions . While Ca sparks and waves have been thoroughly characterized [10 , 14] , the LCRs in sinoatrial node cells have been neither unbiasedly detected , nor systematically classified and analyzed . The existing spark detection algorithms in confocal line-scanning images [14 , 15] and in 2D recordings [15–18] are mainly tuned for stereotyped ( simple ) release events in ventricular myocytes ( or in neurons [19] ) ; and as such they cannot be applied to LCRs in sinoatrial node cells featuring complex release propagation patterns ( including merges and separations ) and multiple release peaks . Thus , the aim of the present study was twofold: We discovered counterintuitive behaviors of LCRs and Ca transient that stabilize the integrals of diastolic Ca and NCX signals upon variations in SR Ca pumping rate that insures regulation and fail-safe pacemaker cell operation within a wide range of action potential firing rates . In order to detect and characterize LCR in a given single time frame ( reflecting an instant distribution of Ca ) we introduced the definitions listed below . The definitions basically outline our LCR detection algorithm ( see Fig 2 for more details ) . Frame: an image ( i . e . 2d-matrix ) generated by numerical simulation program representing an instant [Ca] in each voxel directly under cell membrane . Intensity increment ( Δ ) : an increment of [Ca] that is used to slice the frame into a set of intensity bins . Intensity bin: a mask ( Boolean array ) that reflects if [Ca] is higher of a given level defined by a multiple of the intensity increment over mean value . In our analysis the number of bins was 50 and the bins were constructed as follows: ( Bini , 1<=i<=50 ) includedvoxelswith[Ca]>mean+Δ* ( i-1 ) The lowest intensity bin ( Bin 1 , Fig 2A ) served the purpose of separating the LCRs from the background [Ca] . In other words , we considered all voxels with [Ca] above the average as a part of an LCR . Intensity cluster: a sub-set of the Intensity bin whose TRUE Boolean elements have a common border ( Fig 2B ) . LCR image: a part of a frame within the area of the lowest bin of a set of nested clusters , i . e . all clusters sharing a common lower bin . Intensity peak: [Ca] in a voxel that has the highest amplitude among its neighbors in a given intensity cluster . Simple LCR image: an LCR image that has only one intensity peak Complex LCR image: an LCR image that has multiple intensity peaks , i . e . all voxels with intensity peaks sharing a common low intensity bin within the LCR image . Signal mass ( or total Ca ) of LCR image ( in nmol of Ca ) : a total [Ca] amount of an LCR image , calculated as a sum of Ca concentrations in nmol/liter in all voxels delineating the LCR image multiplied by the voxel volume given in liter ( Vvoxel = 100 nm100 nm100 nm = 10−18 liter ) . Thus , in a single frame , our computer program finds all LCR images . Each LCR image is described as a C++ object that represents a subset of submembrane cytosolic voxels , each of which features its own [Ca] and coordinates . Each LCR image object also carries the LCR signal mass ( in nmol of Ca ) and its complexity type ( simple or complex ) . To handle all LCR images in a given frame , we create an array of objects of LCR images in that frame . Thus , each LCR image is fully referenced by its index within [1 , the total number of LCR images] in that array and time ( frame index ) used further to describe LCR dynamics In order to identify , track , and classify LCRs in a series of time frames , we find all LCR images in each frame and then perform frame-to-frame comparison . Our complete analysis is schematically illustrated in Fig 3 from bottom to top , as hierarchy and complexity of objects increase . The LCR represents the highest hierarchy object in our analysis and the algorithm uses the following definitions: Dynamic LCR ( or simply LCR ) : a set of related LCR images describing in time and space either an individual release or a collection of interacting local releases . In our computer program we describe each LCR by a C++ object holding indexes of all its LCR images in each frame and its termination type . Complex LCR: an LCR that has at least one complex LCR image . Simple LCR: an LCR that has only simple LCR images LCR image pair: A pair of LCR images that share at least one common voxel of the lowest intensity bin in subsequent frames ( i . e . in the previous and in the current frame ) . We identify all pairs to detect LCR propagation , collisions and separations . LCR collision: LCRs collide when the current frame has an LCR image that forms a pair with at least two of the LCR images in the previous frame LCR separation: an LCR undergoes separation when at least two of its LCR images in the current frame form a pair with one common LCR image in the previous frame . Important: when an LCR undergoes separation , it is still considered as the same LCR , i . e . all subsequent coupled LCR images remain within the same C++ object tracking the LCR . LCR thread: a subset of LCR images within a given LCR that represents a separate part of the LCR after its separation . Multi-thread LCR: an LCR that has a history of separations . Simple dynamics LCR: a single-thread LCR that has no history of separations . LCR max amplitude: Maximum [Ca] that is reached by the LCR during its lifetime in μM . LCR path area ( in nm2 ) : the total area that include all voxels participating in the LCR during its lifetime . Signal mass of an LCR ( in nmol*ms ) : the sum of signal masses of all LCR images constituting the given LCR multiplied by the time increment ( 5 ms ) . We define an LCR as a time-sequenced collection of LCR images . Each LCR has its birth with an initial image and then it adds more images to its collection as it progresses through frames . In the course of an LCR’s life , its ordinary behaviors include propagation , expansion , and/or contraction . A frame-to-frame comparison of LCR images is conducted to understand which LCR images in the current frame relate to the ones in the previous frame ( Fig 4 ) . Our algorithm finds all overlapping ( i . e . by their spatial extent ) LCR images in the previous frame and the current frame . Specifically , a Boolean ( AND ) matchup is made within the lowest intensity bins ( outskirt ) for all LCRs images found in the two frames . Thus our computer program creates a set of all LCR image pairs from the previous to current frame . In simple LCR dynamics , an LCR image from the previous frame overlaps with a unique LCR image from the current frame ( Fig 4B ) . The matched-up LCR image from the current frame is added to the LCR that its partner belongs to . Thus , the LCR is “built” and progressing as the algorithm goes through the frames ( technically , the LCR object simply collects indexes of its linked LCR images ) . However this is not always the case . Sometimes a pair does not have a unique LCR image from the previous frame or from the current frame , reflecting collisions or separations ( described later in the text ) . In the case when an LCR image has no overlap with any LCR image from previous frame ( Fig 4A ) , the algorithm initiates a new LCR ( i . e . an LCR birth ) . In the case when none of the LCR images of an LCR in the previous frame finds a partner in the current frame ( Fig 4C ) , the LCR is pronounced terminated by stochastic attrition ( dubbed as LCR “death” ) . Next step is to identify whether a terminated LCR is complex or simple . If the LCR has at least one complex LCR image , then it is defined as complex . If an LCR consists of only simple LCR images , then the LCR is defined as simple . Because complex LCRs embrace several release sites , they usually exhibit complex dynamic behavior and last a long time , while simple LCRs ( usually from one release site ) last a short time . If the LCR persists until the very last frame , it is not a given that the LCR terminates , thus the LCR is labeled “still alive . ” At the end of frame-to-frame comparisons , the total signal mass of each LCR is calculated by summing the signal masses of all its constituent LCR images and then multiplying by the time interval between frames ( that is 5 ms ) . With propagation in play , an LCR image from the current frame can form a pair with multiple LCR images in the previous one and vice versa , i . e . an LCR image from the previous frame can form a pair with multiple LCR images in the current one ( Fig 3 ) . These cases translate to either a collision or separation between LCRs , respectively . There are even cases where separation and collision occur simultaneously . In the event of a collision ( Fig 4D ) , we compare the signal masses of the previous LCR images of each pair , and the one with the largest signal mass has its pair’s LCR live on , while the rest have their LCR terminated , pronounced terminated by collision . The terminated LCR image pairs are also deleted from the main set of LCR image pairs , so that if LCRs collide and separate at the same time , the collision algorithm forcefully has priority over the separation algorithm . In an event of a separation ( Fig 4E ) , all the current LCR images from all the overlapping LCR image pairs are added to the LCR that the common previous LCR image belongs to . With separations , an LCR could be described in a space-time graph as a tree with multiple branches . Because branching happens in time we call the branches LCR “threads” . Thus , in this terminology , an LCR with a history of splits is a multi-thread LCR , whereas an LCR with simple dynamics ( described above ) can be described as a single-thread LCR . As a clarification to a statement made in the previous section on LCR stochastic attrition , termination of multi-thread LCRs differs from that of single-thread LCRs . A multi-thread LCR is pronounced dead only when all its threads terminate ( Fig 3 , top ) . For example , if only one of the threads terminates , the LCR continue to live via remaining threads . In fact , even if all threads terminate , but one , the LCR still lives via the remaining thread . We applied our algorithm of LCR detection and classification to get new insights into calcium pumping role in sinoatrial node cell function , specifically to identify the difference in statistics of LCRs generated by the SR at high ( 24 mM/s ) , moderate ( 12 mM/s ) , and low Pup ( 4 mM/s ) . At all Pup values , JSR ( gray traces in Fig 5 ) becomes depleted of Ca by action potential-induced Ca transient ( blue traces ) , but at higher Pup , JSR refilled quicker with Ca during diastole to reach a threshold of spontaneous Ca release . The earlier and stronger Ca releases activate stronger and earlier NCX currents ( green traces ) , which shorten diastolic depolarization and the cycle length , so that the action potential firing rate increases ( red traces ) as predicted by the coupled-clock theory [1] ( see also S1 and S2 Movies for more details , including local Ca dynamics ) . We examined Ca signals within a time window of the Maximal Diastolic Potential ( MDP ) to -50 mV , i . e . before the activation of L-type calcium current ( the windows at various Pup are shown by a blue box in Fig 5 ) . The MDP values were as follows: -67 . 2 , -69 . 1 , -68 . 5 mV for Pup 4 , 12 , and 24 , respectively . The LCR activity generally increases within this time window . However , LCRs appear rare and small at Pup 4 mM/s , but large , frequent , and propagating at Pup 24 mM/s ( Fig 6A ) . However , at the higher amplitude resolution of 0 . 25 μM local Ca distribution at low Pup shows substantially higher background Ca and multiple small LCRs at the MDP ( Fig 6Ba vs . 6Bb ) . We further examined the interplay of LCRs and Ca transient decay ( seen as decreasing Ca background ) with respect to generation of the diastolic NCX currents important for diastolic depolarization and ultimately action potential firing rate ( Fig 7 ) . Overlapped Ca signals ( synchronized at the Ca transient peak ) show clear diastolic Ca elevations ( Fig 7A ) linked to LCR activity that increases at higher Pup ( arrow “net LCR signal” ) . More important details can be seen at a smaller scale ( Fig 7B ) , i . e . the LCR signal overlaps with the background ( Ca transient decay ) , resulting in a higher nadir ( Ca minimum at about 200 nM ) at Pup = 24 mM/s vs . below 100 nM at Pup = 4 mM/s . The corresponding simulated NCX current traces show an earlier and stronger increase at 24 mM/s , but a decrease at 4 mM/s ( Fig 7C ) . To provide further insights into local Ca signaling , we illustrate typical Ca dynamics in a single voxel at various Pup ( Fig 7D–7F ) . The dynamics is rather noisy at higher Pup , reflecting LCR occurrences nearby the voxel . To evaluate the background changes , we found parts of Ca local transient decays with no LCR occurrences nearby ( shown by rectangles ) . Then we overlapped the traces to illustrate that indeed the decay is accelerated at higher Pup ( Fig 7G ) , indicating that the higher nadir of total Ca signal ( Fig 7B ) is indeed due to LCR overlap , rather than a Ca overload . We next examined the net ( integrated ) Ca and NCX signals generated by the transient and LCRs in the time window of our interest from MDP to -50mV ( Fig 8 ) . Surprisingly , both Ca signal and NCX signal integrated over the critical time window during diastolic depolarization remained basically invariant , despite huge variations in Pup and action potential firing rates . Next , we performed statistical analysis and comparisons of automatically detected LCRs in the same time window of interest ( from MDP to -50 mV ) at two extreme Pup of 4 and 24 mM/s . The analysis confirmed generation of numerous LCRs observed at the low Pup in S1 Movie and Fig 6B: in fact , nearly four times as many LCRs that appeared at higher Pup ( Table 1 ) . When we show LCRs spatial dynamics ( without Ca amplitude grading ) in S3 Movie , one can indeed observe frequent and large spreading LCRs at Pup 4 mM/s . As we show below , the majority of these LCRs is of relatively low amplitude and have a small impact on the net diastolic Ca signal . Next we applied our new LCR classification and found that the amount of complex LCRs stayed nearly the same , whereas the amount of simple LCRs was about 7 times larger at low Pup ( see Table 1 ) . With regard to the diastolic depolarization time course , most of the simple LCRs at low Pup formed closer to the start of the diastolic depolarization i . e . the MDP ( histograms in Fig 9A and 9B ) , whereas complex LCR appeared throughout the entire diastole ( see S3 and S4 Movies ) . Complex LCRs had an average birth time counted from the MDP of 221 ms , whereas simple LCRs had an average birth time of only 95 ms . The very premature small simple LCRs are actually linked to the decaying transient , provoking releases from the junctional SR ( JSR ) not completely filled with Ca . Poor SR Ca pumping at Pup 4 mM/s clears cytosolic Ca less efficiently , leaving much higher background Ca levels as clearly seen in Fig 6B ( red background vs . black background in sub-panels a vs . b ) and local decays lacking LCRs in Fig 7G . As followed from histograms in Fig 9A and 9B , such small isolated releases happen less frequently further in the diastole because the transient has been decayed and the cytosolic calcium is cleared ( Fig 7G ) , preventing premature releases and thereby allowing JSR to refill with substantial amount of Ca before the release . It is important to note , however , the early simple LCRs were much smaller and contributed much less toward the total diastolic signal vs . complex LCRs . Specifically , signal mass of all complex LCRs was 5 times larger than that of all simple LCRs at Pup 4 mM/s and 68 times larger at Pup of 24 mM/s ( see Table 1 ) . Taking into account this result we have assumed simple LCRs as negligible and focused our further analysis on complex LCRs . While the number of complex LCRs are almost the same at both Pup , at high Pup complex LCRs had larger signal mass than those at low Pup . ( Table 1 , parameter “Signal Mass per LCR” ) . Our histogram of the LCR signal mass distribution revealed an emergence of high signal mass LCRs ( Fig 9D , shown by red circle ) that are not observed at low Pup ( Fig 9C ) . We dubbed these release events “powerful LCRs” , because LCRs with higher signal mass ( i . e . larger size and amplitude ) is a known powerful mechanism to generate larger NCX currents and notably accelerate the pacemaker rate ( e . g . in numerical simulations [12] and in experiments with phosphodiesterase inhibition [13] ) . To get further insights into emergence of the powerful LCRs , we analyzed the components of the signal mass parameter: path area , duration , birth time , and amplitude . At Pup 4 mM/s the average LCR path area was larger and their duration was longer than that of Pup 24 mM/s ( Table 1 , “Average Path Area” and “Average Duration , ” Fig 10 ) . Thus , the major factor of a larger signal mass at Pup 24 mM/s was actually larger release amplitude ( Table 1 , “Average LCR Maximum Amplitude” ) . Our analysis provided more interesting results . Even though LCRs lasted shorter and their path areas were smaller at Pup 24 mM/s , the amount of collisions ( Table 1 , “Total Complex Collisions” ) was larger at the high Pup , indicating a more active recruiting process i . e . Ca-induced-Ca-release ( CICR ) , resulting in a faster propagation . Finally , to further support our findings , we performed additional LCR analysis at a moderate SR Ca pumping with Pup of 12 mM/s ( i . e . between the two extreme values of 24 and 4 mM/s discussed above ) that was assigned to a basal state action potential firing in prior numerical studies [1 , 20] . We measured parameters of LCRs ( see Table 1 ) detected by our algorithm under these conditions and illustrated LCR detection in S5 Movie . The appearance of the detected LCRs and their parameters values were found somewhere reasonably between those described above at the two extreme Pup of 4 and 24 mM/s . An important result of this additional analysis was that at all three Pup tested the number of complex LCRs remained almost unchanged , and the total signal mass of all LCRs was composed of mainly complex LCRs . Based on data in Table 1 , complex LCRs contributed 84% , 96% and 99% to the total signal mass of all LCRs at Pup 4 , 12 and 24 mM/s , respectively . The present study accomplished two goals: 1 ) we developed a novel LCR classification and computer algorithm that automatically detects and analyzes LCRs generated in silico [5]; 2 ) using the new algorithm we discovered a new mechanism of stabilization of integrals of diastolic Ca and NCX signals via LCR adoptive changes and an interplay of the LCRs and Ca transient over a wide range of SR Ca pumping rates . Our new classification of LCRs includes LCR merges and separations as well as three types of termination ( by stochastic attrition , merging into a larger release , and fusion into action-potential-induced transient ) . Using our new algorithms and our new terminology , we were able for the first time to classify ( and thus simplify ) the extremely complex appearance of intracellular Ca dynamics ( S1 and S2 Movies ) by a few typical cases of LCR behavior with their statistics and contributions into the total Ca signal . We found that signal mass of all LCRs is mainly generated by complex LCRs over a wide range of pumping rates , indicating the functional importance of CICR ( and release interactions ) in forming the LCR signal and pacemaker cell function . Simple ( non-propagating ) , small LCRs mainly occur close to the MDP at the end of the decaying transient . The incomplete transient decay , especially at low pumping rates at the beginning of diastolic depolarization ( Fig 7G ) , results in a higher cytosolic Ca that provokes the premature small releases which have a small impact to the total LCR signal . Small LCRs at the beginning of the diastolic depolarization were also observed in 2D high-speed camera recordings in rabbit sinoatrial node cells [21] . Our examination of complex LCRs revealed a larger path area and longer duration for the LCRs at low Pup ( Table 1 ) . This counterintuitive result is explained by the fact that poor SR Ca pumping imparts a higher background diastolic Ca in the cytosol ( Fig 7G ) that provokes release propagation and therefore longer duration of the complex LCRs , despite their smaller amplitude . Furthermore , while the total LCR signal mass decreases at low Pup ( Table 1 ) the duration of the diastolic depolarization becomes longer and slower Ca transient decay contribution to net NCX signal increases . This interplay of LCRs and the transient stabilizes the integrals of diastolic Ca and NCX signals ( Fig 8 ) . This finding offers a new explanation to a prior experimental result that a substantial pharmacological inhibition of SR Ca pumping ( e . g . by cyclopiazonic acid [2] ) decreases the rate , but does not cease spontaneous action potential firing of sinoatrial node cells . This experimental result is often used as a key argument against importance of local Ca releases in cardiac pacemaker function ( e . g . section “Are Ca2+ sparks/LCRs Necessary for Pacemaking ? ” in [22] ) . The present study shows that Ca releases are important to stabilize the Ca signal mass over a wide range of Pup , but especially at higher Pup . A similar counterintuitive result has been reported in NCX-deficient model , in which diastolic NCX current ( and basal pacemaker rate ) remained almost unchanged ( i . e . stabilized ) , with only 20% remaining NCX molecules [23] . Next we found that the increase in the action potential firing rate at higher Pup is associated with a more synchronized and higher amplitude LCR ensemble signal that ultimately generated a higher amplitude NCX current , accelerating the diastolic depolarization ( CasubCyt , INCX and Vm traces in S1 and S2 Movies , Fig 7A–7C ) . That is in line with the coupled-clock theory ideas [1 , 9] and prior experimental observations that at higher pumping rate ( e . g . in the presence of β-adrenergic receptor stimulation ) LCRs increase in amplitude , signal mass and emerge earlier within the cycle [11 , 24] . On the contrary , in the presence of pharmacological inhibition of SR Ca pumping with cyclopiazonic acid , LCRs decrease in amplitude and signal mass and they emerge later within the cycle [3] that is also in line with our finds here in simulated LCRs . Increase in the release amplitude ( dubbed spark current or “Ispark” in prior studies [25 , 26] ) is an important known mechanism of release recruitment via CICR [26] . Our statistical LCR analysis show increased number of LCR collisions and faster propagation at higher Pup that is an important mechanism of LCR synchronization . At the same time higher Pup depletes cytosolic Ca more efficiently , preventing propagation and premature releases . The preclusion of the premature small releases represents , in fact , yet another release synchronization mechanism . Finally , the faster SR Ca pumping is associated with faster SR refilling and more synchronous attainment of the spontaneous release threshold in different SR locations , representing the third release synchronization mechanism . There is a discrepancy , however , between simulations and experiments with respect to both the LCR number and size when SR pumping varies . Partial inhibition of SR Ca pump with cyclopiazonic acid decreased the LCR number and size , but β-adrenergic receptor stimulation increased these parameters [3 , 11 , 24] . When we varied Pup in silico , the number of complex ( meaningful ) LCRs stayed almost unchanged ( Table 1 ) , but the LCR size changed to the opposite vs . the experimental results . This apparent contradiction could be explained by the fact that our algorithm detects all releases in the model system , in which there is no noise . Experimental records of fluorescent Ca indicator signals , however , have instrumental noise that precludes detection of the low amplitude releases ( predicted by the theory ) . As LCR signal amplitude increases , the LCRs become experimentally detectable , observed as an apparent increase in LCR numbers ( as seen in S1 and S2 Movies , in a high Ca scale of 5 μM ) . There are seveal technical issues that complicate application of our LCR detection method to real pacemaker cells . Recording noise of high speed cameras and confocal microscopes is one of major issues . An additional issue is that experimental sub-sarcolemmal calcium is very difficult to visualize in real time due to dye bleaching , photo-damage , Ca buffering by indicator molecules , and scanning rate limitations , especially with 3D microscopy methods . Finally there is a motion artifact in these spontaneously contracting cells . that requires mapping the intracellular space via an additional algorithm . Some of these issues have been explored in our recent pilot study , in which we applied the LCR detection principles and terminology reported here to experimental Ca recordings performed by high speed cameras in sinoatrial node cells isolated from rabbit and guinea pig [27] . The cell motion artefact was addressed by using non-Euclidean dynamic coordinate transformations to compensate for cell contraction . Similar to simulated LCRs analyzed here , LCRs in real cells also expand , propagate , merge , separate and terminate . However , the number of LCRs per cycle varied substantially among individual cycles , cells , species , and camera types , but rarely exceeded 100 , indicating that many LCRs present in numerical simulations ( Table 1 ) are missed in the real recordings . Some smaller release signals can be hidden within the experimental noise and remain below the detection threshold . In the absence of recording noise , our modeling system detects every LCR in curved submembrane space that is perfectly scaled in terms of space coordinates and absolute Ca concentrations , whereas the experimental data are obtained from noisy 2D non-confocal images of non-linear Ca sensing dyes , which are projections through the cell of various shapes ( spindle shape , spider shape , etc . ) , convolved with the point-spread function of the microscope . Despite these numerous technical issues , our new algorithm and terminology seem to be helpful in statistical analysis and interpretation of the experimental data . Further technical advances in Ca signal recording will narrow the gap between the experimental and simulated data . In summary , we introduced new terminology , classification , and automatic algorithms to characterize complex spatiotemporal structure of LCRs in cardiac pacemaker cells . Using this new approach , we found that Ca pumping regulates LCRs and pacemaker rate via timely synchronized occurrence of LCRs creating a powerful ensemble signal activating NCX current . Specific mechanisms of LCR synchronization include CICR , suppression of small premature releases , and more synchronous SR refilling with Ca at higher Pup . At lower Pup the LCRs have smaller amplitudes , but the system manifests a counterintuitive behavior to stabilize the diastolic Ca signal mass ( and its NCX signal ) by increasing LCR size and duration and by increasing the signal from slowly and longer decaying Ca transient . More generally , these results demonstrate functional importance of complex local crosstalk among NCX , LCR , Ca pump , and L-type Ca channels . The emergent behaviors of these interactions deserve accurate and more detailed future studies . Finally , our development of new LCR detection and classification algorithms will be also helpful in creating similar approaches to unbiased detecting and analyzing experimental data on LCRs in sinoatrial node cells [27] . Specifically it would be important to compare LCR synchronization mechanisms in silico and experimental data via analysis of statistics of major LCR parameters introduced here , including path area , amplitude , duration , birth time , separations , collisions , and stochastic attritions . Highly integrated release events have also been reported in other cell types: Ca wavelets , abortive Ca waves , macrosparks , compound sparks or puffs [10] . Our new approach to detect and analyze complex local release events may be also helpful in studies of these Ca release events . The numerical simulations were performed using our recent 3D-model of rabbit sinoatrial node cells [5] . The model structure is illustrated in Fig 1 and a full model description can be found in the original paper . In short , membrane ion currents and voltage are represented by ordinary differential equations , except for the L-type calcium current , which is handled separately and stochastically . Diffusion and buffering of calcium in the cytosol , and in the free SR , are represented by partial differential reaction-diffusion equations . The free SR is treated as a fine random network , considered as a separate , continuous space co-extensive with the cytosol . The model simulates the function of individual Ca release channels , ryanodine receptors ( RyRs ) , located in a close proximity to the cell membrane in line with the prior experimental result that the primary pacemaker cells of smaller size in the center of rabbit sinoatrial sinoatrial node express RyRs mainly under cell membrane ( along cell perimeter ) with almost no presence in the bulky cytosol [5 , 28] . It is important to note that here we studied LCRs generated only by this type of “hollow” , primary pacemaker cells , but we did not model LCRs in another type of ( “peripheral” ) sinoatrial node cells with RyR clusters also located inside the cells . The dense clusters of RyRs are co-localized in the model with L-type calcium channels at 15 nm dyad junctions between JSR and the sarcolemma forming couplons ( or Ca release units ) . Each couplon has one JSR compartment containing calsequestrin . The JSR receives Ca from the adjacent free SR through a diffusion resistance representing the several fine , tubular nexi . Ca from the JSR is released into the dyadic space through open RyRs at a rate proportional to the free Ca gradient between the JSR and the dyadic space at the location of each RyR . The dyadic space of each couplon is discretized into a two-dimensional grid of 10 nm squares on which local calcium and diffusible buffer evolve according to reaction-diffusion equations , which are integrated along with the calcium concentration in the JSR compartment . RyRs are located at 30 nm spacing , and L-type channels randomly placed . Ca diffuses from the edges of the cleft into the adjacent cytosol . Joint gating of RyRs and L-type channels is simulated by a custom-modified Gillespie Monte Carlo algorithm that generates an exact realization of the high dimensional , variable-rate Markov process controlled by voltage and instantaneous local calcium [29] . In cardiac pacemaker cells the critical region for the clocks coupling is under cell membrane . It is the local Ca under the cell membrane that activates Na/Ca exchanger accelerating diastolic depolarization . Thus , while the model predicts Ca dynamics within the entire cell , here we examined Ca only under the cell membrane . Since the cell shape is approximated by a toroid , all examined signals under the membrane are represented by a 2D array of Ca intensities in respective sub-membrane voxels with a size of 100 x 100 x 100 nm ( Fig 1 ) . Examples of our model simulations of Ca dynamics ( near the cell membrane and in cross sections ) along with action potentials and NCX current at different Pup are given in S1 and S2 Movies . Here we developed new , unique computer algorithms for automatic detection and classification of LCRs . The LCRs were detected in series of consecutive images of simulated Ca dynamics within 100 nm under the cell membrane . The computer algorithms were implemented in C++ within Microsoft Visual Studio 2010 . Our original C++ source code of the algorithms is available at https://www . nia . nih . gov/research/labs/lcs/complex-lcr-detector . For visualizations we coded with OpenGL also in Microsoft Visual Studio 2010 . Rather than having a single cut off amplitude to find Ca sparks [14] , our new algorithm is capable of also differentiating between high amplitude and low amplitude neighboring release events . In other words , the new algorithm detects peaks of a wide range of amplitudes . The algorithm also tracks the LCR complex spatiotemporal evolution , including births , deaths , separations , and collisions . It is important to note that while the time tick of the model simulations was 0 . 05 ms , the simulation output ( and our analysis ) was performed with a time interval of 5 ms that satisfactory describes the LCR dynamics . This simplification substantially accelerated data output , storing , and computation . We analyzed the simulations within the time window starting at MDP and ending at -50 mV ( at the threshold of L-type channel activation and before the beginning of action potential-induced Ca transient ) .	Life’s vital rhythm , the heartbeat is guaranteed by specialized , cardiac pacemaker cells . Based upon the Hodgkin-Huxley membrane excitation theory and its application to cardiac cells , the cardiac pacemaker clock has been , for a long time , to be considered essentially a surface membrane oscillator , i . e . a membrane clock , driven by an ensemble of time- and voltage-dependent ion channels . More recent studies , however , discovered a tight integration of the membrane clock with an intracellular Ca oscillator ( dubbed Ca clock ) . The Ca clock generates rhythmic , diastolic locally propagating Ca releases from sarcoplasmic reticulum , a cell major Ca store , that is refilled with Ca via a Ca pump . The released Ca activates Na/Ca exchanger that generates an inward current and accelerates the diastolic depolarization . Despite their importance , the local releases have not been systematically studied . Here we developed a new computer algorithm for automatic detection , classification , and analysis of local Ca releases generated by a recent computational model of a sinoatrial node cell . Then we investigated how the releases are regulated by the Ca pump , a major functional component of Ca clock . We discovered counterintuitive behaviors stabilizing diastolic Ca signal mass and ensuring fail-safe pacemaker operation at various pumping rates .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "biotechnology", "medicine", "and", "health", "sciences", "action", "potentials", "depolarization", "applied", "mathematics", "membrane", "potential", "electrophysiology", "neuroscience", "simulation", "and", "modeling", "algorithms", "ion", "channels", "mathematics", "pace...	2017	Stabilization of diastolic calcium signal via calcium pump regulation of complex local calcium releases and transient decay in a computational model of cardiac pacemaker cell with individual release channels
Human neutrophil antigen 2 ( HNA-2 ) deficiency is a common phenotype as 3–5% humans do not express HNA-2 . HNA-2 is coded by CD177 gene that associates with human myeloproliferative disorders . HNA-2 deficient individuals are prone to produce HNA-2 alloantibodies that cause a number of disorders including transfusion-related acute lung injury and immune neutropenia . In addition , the percentages of HNA-2 positive neutrophils vary significantly among individuals and HNA-2 expression variations play a role in human diseases such as myelodysplastic syndrome , chronic myelogenous leukemia , and gastric cancer . The underlying genetic mechanism of HNA-2 deficiency and expression variations has remained a mystery . In this study , we identified a novel CD177 nonsense single nucleotide polymorphism ( SNP 829A>T ) that creates a stop codon within the CD177 coding region . We found that all 829TT homozygous individuals were HNA-2 deficient . In addition , the SNP 829A>T genotypes were significantly associated with the percentage of HNA-2 positive neutrophils . Transfection experiments confirmed that HNA-2 expression was absent on cells expressing the CD177 SNP 829T allele . Our data clearly demonstrate that the CD177 SNP 829A>T is the primary genetic determinant for HNA-2 deficiency and expression variations . The mechanistic delineation of HNA-2 genetics will enable the development of genetic tests for diagnosis and prognosis of HNA-2-related human diseases . Transfusion-related acute lung injury ( TRALI ) is associated with the transfusion of leukocyte alloantibodies from donors or associated with the presence of alloantibodies in recipients of blood [1 , 2] . Alloantibodies against human neutrophil alloantigenes ( HNAs ) are a very strong trigger for the development of TRALI [1 , 2] . Human neutrophil antigen 2 ( HNA-2 ) alloantibodies have been linked to the induction of TRALI and various pulmonary reactions [3–6] while anti-HNA-3 alloantibodies are frequently implicated in severe and fatal TRALI [7] . Animal models have firmly established a pathological role for HNA-2 alloantibodies in TRALI [8 , 9] . Furthermore , HNA-2 alloantibodies have been implicated in multiple human disorders such as neonatal alloimmune neutropenia , autoimmune neutropenia , drug-induced immune neutropenia , and graft failure following marrow transplantation [10–13] . Accordingly , HNA-2 is among the most important clinical antigens . HNA-2 is heterogeneously expressed on subpopulations of neutrophils and approximately 3–5% Americans do not express HNA-2 [14] . HNA-2 deficient subjects are predisposed to the production of HNA-2 alloantibodies when exposed to the HNA-2 antigen during blood transfusion , pregnancy , and bone marrow transplantation . HNA-2 is encoded by the CD177 gene that contains nine exons at Chromosome 19q13 . 31 region , where a CD177 pseudogene highly homologous to CD177 between exon 4 and 9 is also located ( Fig 1A ) [15–17] . The genetic studies of CD177 were significantly hampered by the presence of CD177 pseudogene [18 , 19] . HNA-2 is also known as PRV-1 as CD177 mRNA is over-expressed in polycythemia rubra vera patients [20] . CD177 has an open reading frame of 1311 nucleotides that encode 437 amino acids with a signal peptide of 21 residues . HNA-2 ( or CD177 ) is expressed as a GPI-linked receptor with a mature peptide consisting of residue 22 to 408 [15 , 21] . HNA-2 plays important roles in neutrophil functions and myeloid cell proliferation . The interaction between HNA-2 and PECAM-1 facilitates neutrophil transendothelial migration [22 , 23] . In addition , HNA-2 is required for the attachment of proteinase 3 ( PR3 ) to neutrophils [24–27] , which plays a pivotal role in PR3-ANCA-mediated neutrophil activation [28] . CD177 mRNA levels are elevated in several conditions associated with increased neutrophil counts [14 , 29] . Furthermore , elevated levels of neutrophil CD177 mRNA are associated with increased neutrophil production and quantitation of neutrophil CD177 mRNA is a diagnostic tool for polycythemia vera [14] . Moreover , the level of HNA-2 expression has been identified as a prognostic biomarker for gastric cancer [30] . The CD177 non-synonymous coding SNPs ( cSNPs ) were reported to associate with HNA-2 expression variations , however , the effect of those non-synonymous CD177 coding SNPs on HNA-2 expression was unknown [18 , 31 , 32] . CD177 mRNA splicing variants were found in two HNA-2 deficient donors but it remains inconclusive whether CD177 splicing abnormality was actually responsible for HNA-2 deficiency [33] . Therefore , the underlying genetic mechanism of HNA-2 deficiency has remained elusive since the observation of HNA-2 deficiency four decades ago [10] . Elucidation of the molecular genetics and basis of the HNA-2 deficiency is a prerequisite for the use of effective genetic tests in prognosis and diagnosis of HNA-2-related human diseases . In the current study , we demonstrated that a novel nonsense CD177 coding SNP 829A>T is the primary genetic determinant for HNA-2 deficiency and expression variations in humans . The percentages of neutrophils expressing HNA-2 were heterogeneous among normal healthy blood donors in flow cytometry analysis ( Fig 1B ) . In 294 normal healthy blood donors , the percentage of HNA-2-positive neutrophils ranged from 0 . 0% to 97 . 8% . Among 294 blood donors , we have identified 11 donors ( or 3 . 7% ) deficient for HNA-2 and the percentage of HNA-2 deficient blood donors is consistent with those previously reported [6 , 10 , 34] . Copy number variations ( CNVs ) are the primary cause of human neutrophil antigen 1 ( HNA-1 or FcγRIIIB ) deficiency and expression variations [35–38] . To investigate whether CD177 CNVs are involved in HNA-2 deficiency , we determined CD177 CNVs using TaqMan CNV assay kit Hs01327659_cn with the probe targeting the unique CD177 exon 1 region ( Fig 1A ) . Among 294 human subjects , 95 . 2% ( 280/294 ) of subjects were two-copy CD177 carriers and 4 . 8% ( 14/294 ) were three-copy CD177 carriers . No human subjects had CD177 gene deletions among 294 subjects . Notably , all 11 HNA-2 deficient donors identified in the flow cytometry analysis carried two copies of CD177 gene . In addition , those 11 HNA-2 deficient donors produced full-length CD177 mRNAs as demonstrated by RT-PCR ( Fig 1C ) . Our data clearly demonstrated that CD177 gene deletion ( or CNVs ) and the lack of mRNA expression are not the cause of HNA-2 deficiency . We subsequently determined CD177 cDNA sequences of all 11 HNA-2 deficient donors along with 119 HNA-2 positive donors . In addition to CD177 coding SNPs ( cSNPs ) identified previously , we discovered five novel cSNPs ( SNP 824G>C or rs17856827G>C , 828A>C or rs70950396A>C , 829A>T or rs70950396A>T , 832G>A , and 841A>G or rs201266439 ) ( S1 Table ) , which form two haplotypes ( Fig 2 ) . Most importantly , the CD177 SNP 829A>T is a nonsense polymorphism that creates a translation stop codon at amino acid position 263 ( Lysine → Stop codon change ) in CD177 open reading frame . Consequently , those two haplotypes were designated as the open reading frame haplotype ( or ORF allele: 824G/828A/829A/832G/841A ) and the stop codon haplotype ( or STP allele: 824C/828C/829T/832A/841G ) ( Fig 2 ) . To determine the origin of the novel CD177 cSNP haplotype , we have also sequenced CD177 genomic DNA PCR products . Based on genomic DNA sequencing analysis , 72 . 1% ( 212/294 ) of donors were homozygous 829AA donors and the homozygous 829TT donors accounted for 3 . 1% ( 9/294 ) in our study population . The minor allele ( 829T ) frequency is 15 . 5% ( S2 Table ) . The distribution of SNP 829A>T genotypes was consistent with the Hardy-Weinberg equilibrium in 294 blood donors ( χ2 = 0 . 76 , P = 0 . 38 ) ( S2 Table ) . To examine whether the CD177 SNP 829A>T affects HNA-2 expression , the donor genotypes and HNA-2 expressions were statistically analyzed . As shown in Fig 3A , all nine 829TT homozygous donors were negative for HNA-2 expression in flow cytometry analysis . In addition , the percentages of HNA-2 positive neutrophils from 73 heterozygous donors ( 829AT ) were significantly lower than those from 212 homozygous 829AA donors ( P < 0 . 0001 ) . Western blot analyses also confirmed the absence of HNA-2 protein in 829TT homozygous donors and significantly less HNA-2 protein being expressed in the 829AT donors when compared to the 829AA homozygous donors ( Fig 3B ) . Our data strongly support the notion that the SNP 829A>T allele is a crucial determinant for HNA-2 deficiency and expression variations . To verify our findings , we recruited an independent cohort containing 102 blood donors , among whom nine HNA-2 deficient donors were identified ( S1 Fig ) . Similar to those of the first cohort , all nine HNA-2 deficient donors in the replication cohort were SNP 829TT homozygotes as demonstrated by sequencing analysis of genomic DNA and cDNA ( S2 Fig ) . Again , the SNP 829A>T genotypes were significantly associated with the percentages of HNA-2 positive neutrophils ( S1 Fig ) and the HNA-2 protein expression ( S3 Fig ) . Our data confirmed that the SNP 829A>T is a crucial genetic determinant for HNA-2 deficiency and expression variations . Similar to all nine homozygous 829TT donors , two 829AT heterozygous donors were also negative for HNA-2 expression ( Fig 3A , empty diamonds in the middle column ) . Analysis of their CD177 cDNA sequences revealed that both HNA-2 deficient donors who were heterozygous for SNP 829A>T also had a heterozygous deletion of the guanidine nucleotide at nucleotide 997 ( 997G deletion ) . To determine haplotypes of the SNP 829A>T and the 997G deletion , we cloned and sequenced cDNA from those two HNA-2 deficient donors . As shown in Fig 4A , two species of CD177 mRNAs were found in those two donors . The SNP 829T ( STP ) allele is in the linkage disequilibrium with the wild-type CD177 997G allele while the 829A ( ORF ) allele carries the 997G deletion . Genomic DNA sequence analysis confirmed that the guanidine nucleotide deletion occurs at genomic level ( Fig 4B ) . Our data indicate that the presence of the 829T allele in combination with the deletion mutation at nucleotide 997 on another chromosome could also lead to the HNA-2 expression deficiency in an individual . However , we found that only two out of 294 blood donors carried the 997G deletion mutation at one chromosome with genomic sequencing analysis . Therefore , the allele frequency of the 997G deletion mutation is estimated to be 0 . 0034 in the study population . In those two 829AT heterozygous donors , the 997G deletion allele was coincidentally paired with the 829T allele , which facilitated the discovery of the rare 997G deletion mutation in the study . We failed to identify any donors with the CD177 997G deletion among 102 additional blood donors of the replication cohort , confirming that the 997G deletion is a rare mutation in the population . Although the genotypes of CD177 non-synonymous SNPs were reportedly associated with HNA-2 expression variations in several genetic analyses [18 , 31 , 32] , it is unknown whether those CD177 cSPNs directly affect HNA-2 expression . To examine the effect of non-synonymous CD177 cSNPs on HNA-2 expression and on the binding to HNA-2 alloantibodies , we cloned the full-length CD177 cDNA variants containing common non-conservative cSNPs ( SNP 134A>T , 652A>G , 656G>T , and 1084G>A ) within the coding region for HNA-2 mature peptide ( aa22-408 ) . As shown in Fig 5 , there were no significant differences in the expression of HNA-2 ( Fig 5A ) or in the binding to HNA-2 alloantibodies ( Fig 5B ) among four CD177 variants consisting of four non-conservative amino acid substitutions ( His31Leu , Asn204Asp , Arg205Met , and Ala348Thr ) . Our data support the notion that non-synonymous CD177 cSNPs do not have a direct role in the HNA-2 alloantibody production and expression variations . However , cells transfected with CD177 variants of either STP haplotype ( CD177-STP ) or 997G deletion ( CD177-997ΔG ) failed to express HNA-2 on cell surface ( Fig 5C ) and had no reactivity with HNA-2 alloantibodies ( Fig 5D ) . Our data confirmed that either STP allele or 997G deletion mutation will lead to the HNA-2 expression deficiency . To further confirm that the nonsense SNP 829A>T in the STP haplotype is the key factor for HNA-2 expression , we generated a CD177 expression construct carrying the sole change at SNP 829A>T position . The T substitution at nucleotide position 829 alone led to the absence of HNA-2 expression in transfection experiments ( S4 Fig ) , confirming that the SNP 829A>T is the sole determinant for HNA-2/CD177 expression in the STP haplotype . The phenomenon of HNA-2 deficiency was observed more than four decades ago [10] , however , the underlying genetic mechanism of HNA-2 deficiency has remained unknown . In the current study , we identified five common CD177 cSNPs ( SNP 824G>C , 828A>C , 829A>T , 832G>A , and 841A>G , minor allele frequency = 0 . 155 ) in complete linkage disequilibrium . Among five SNPs , the nonsense SNP 829A>T changes the amino acid codon #263 from lysine to a stop codon , which leads to the HNA-2 expression deficiency . Neutrophils from all 829T allele homozygous donors failed to express HNA-2 . In addition , the percentages of HNA-2 positive neutrophils from the SNP 829A>T heterozygous donors ( ORF/STP ) were significantly lower than those from ORF homozygous donors . In vitro , the T substitution at the nucleotide position 829 alone led to HNA-2 expression deficiency in transfection experiments , confirming that the SNP 829A>T is the sole determinant for HNA-2 expression in the STP haplotype . Our study was the first to unravel the genetic mechanism for HNA-2 deficiency , which plays critical roles in human immunological diseases including TRALI , immune neutropenia , and bone marrow graft failure [3–6 , 10–13] . The delineation of the HNA-2 genetics undoubtedly will enable the development of effective genetic and clinical diagnosis tools in human medicine . Intriguingly , similar to neutrophils from all homozygous donors of 829T allele , neutrophils from two 829AT heterozygous donors were also negative for HNA-2 expression ( Fig 3A ) . Analysis of cDNA sequences of those two 829AT heterozygous donors deficient for HNA-2 revealed that the 829A allele ( or ORF allele ) in those two donors had a guanidine deletion at the nucleotide position 997 , which leads to the CD177 reading-frame shift starting from the amino acid codon #319 ( Fig 4 ) and the creation of a stop codon at the amino acid codon #342 . The CD177 997G deletion also leads to the early termination of HNA-2 peptide translation , similar to the consequence of the 829T allele . Furthermore , the CD177 variant carrying the nucleotide 997G deletion failed to express HNA-2 on cell surface in the transfection experiments ( Fig 5C ) , confirming the contribution of the 997G deletion mutation to HNA-2 deficiency in those two specific individuals . The CD177 nucleotide 997G deletion mutation was extremely rare ( mutant allele frequency = 0 . 0034 ) and was absent in the replication cohort of 102 donors . The coincidental appearance of the 997G deletion allele and the 829T allele in the HNA-2 deficient donors facilitated the discovery of the rare 997G deletion mutation in the study . Therefore , at the presence of 829T allele , the rare CD177 997G deletion may also contribute to HNA-2 deficiency . However , the 997G deletion mutation with the allele frequency of 0 . 0034 will have much less impact on overall HNA-2 deficiency as compared to the SNP 829A>T ( the 829T allele frequency was 0 . 155 , S2 Table ) . Previous genetic studies suggested that the CD177 non-synonymous SNPs might affect HNA-2 expression [18 , 31 , 32] , however , the effect of those CD177 cSNPs on HNA-2 expression is unclear . In the current study , we carried out transfection experiments to examine whether common non-conservative cSNPs ( SNP 134A>T , 652A>G , 656G>T , and 1084G>A ) within the HNA-2 mature peptide ( aa22-408 ) affect the HNA-2 expression and the binding of HNA-2 alloantibodies . We found that the expression of HNA-2 and the binding to HNA-2 alloantibodies were not significantly different among those natural CD177 variants containing non-conservative amino acid substitutions ( His31Leu , Asn204Asp , Arg205Met , and Ala348Thr ) ( Fig 5A and 5B ) . The expression of HNA-2 in normal neutrophils is also affected by methylations of CD177 promoter and the CD177 SNP 42G>C ( rs45441892 ) at the third codon ( Pro3Ala ) of the HNA-2 signal peptide was associated with methylation levels of CD177 promoter [39] . However , we found no association between the SNP 42G>C genotypes and the percentages of HNA-2 positive neutrophils in our study ( ANOVA , P = 0 . 1209 , S5 Fig ) . Taken together , those non-synonymous CD177 cSNPs do not seem to have a significant effect on HNA-2 deficiency and expression . The CD177 mRNA splicing abnormality was previously suggested to be the cause of HNA-2 deficiency as alternatively spliced CD177 mRNA species were detected in two HNA-2 deficient donors [33] . However , no further evidence was provided to support the alternative splicing hypothesis of HNA-2 deficiency in the report [33] . Alternative mRNA splicing is a physiological process and is an essential mechanism to produce different products from a single human gene [40–42] . It seems unlikely that HNA-2 deficient subjects have an abnormal mRNA splicing machinery as HNA-2 deficient donors appear healthy [6] . We have detected full-length CD177 mRNAs in all 11 HNA-2 deficient donors in the main study ( Fig 1C ) and in all nine HNA-2 deficient donors from the replication study . The combination of the alternative spliced CD177 mRNA isoforms and the regular CD177 mRNA isoform occurred only in two out of nine SNP 829TT homozygous donors in our replication cohort ( S2 Fig ) . Our data refute the notion that the alternative splicing is a major cause of HNA-2 deficiency . Although gene deletions or copy number variations ( CNVs ) are the primary cause for HNA-1 ( or FcγRIIIB ) deficiency [35–38] , we did not find any CD177 gene deletion in our blood donors . We found that all HNA-2 deficient donors expressed full-length CD177 mRNAs . We also found that the SNP 829T allele was in complete linkage disequilibrium with SNP 134A , 156G , 593G , 652A , 656G , 671C , 782C , 793C , 824C , 828C , 832A , 841G , 1084G , and 1333G . Our data clearly demonstrated that the gene deletion or the lack of mRNA expression is not responsible for HNA-2 deficiency , in contrast to the HNA-1 deficiency . Interestingly , we found that all heterozygous donors of the SNP 829A>T determined by genomic DNA analysis primarily produced the SNP 829A allele ( or ORF allele ) mRNA based on their cDNA sequences . The nonsense SNP 829T allele tracer peak barely above the background was typically considered as sequence noise in the cDNA sequence analysis for heterozygous donors . This observation suggests that the CD177 mRNAs containing the nonsense 829T allele are much less stable than the CD177 mRNAs containing the common 829A allele within the same donor . This may explain the observation of associations between expression variations and certain CD177 cSNPs and the inability to discover the SNP 829A>T using the cDNA sequencing strategy in previous studies [31–33] . After transcription , the CD177 mRNA of the nonsense 829T allele may be quickly degraded by the mechanism of nonsense-mediated mRNA decay [43] , which will lead to the low abundance of CD177 829T allele mRNA and the dominance of CD177 829A allele mRNA in the heterozygous individual . The nonsense-mediated mRNA decay mechanism per se may contribute to the CD177 mRNA expression deficiency in humans with different diseases , which may explain that the partial HNA-2 peptide was undetectable from those HNA-2 deficient donors in a previous study [44] and in the current study using multiple anti N-terminus of HNA-2 mAbs and HNA-alloantibodies ( S6 Fig ) . Therefore , HNA-2 alloantibodies likely target the whole mature peptide of CD177 in HNA-2 deficient subjects . In heterozygous donors for the SNP 829A>T , only the 829A allele is able to express HNA-2 . CD177 promoter DNA methylation regulates HNA-2 expression under physiologic conditions [39] . Non-selective methylation on the 829A allele alone is sufficient to effectively abrogate the HNA-2 expression in a specific cell during granulopoiesis , which may explain that the percentages of HNA-2 positive granulocytes were significantly lower in the 829A>T heterozygous donors than those in the 829A ( or ORF ) allele homozygous donors ( Figs 3 and S1 ) . Therefore , our data strongly support the concept that the SNP 829A>T is also a primary genetic factor for HNA-2 expression variations in humans . As an important biomarker , HNA-2 ( CD177 ) is over-expressed in neutrophils from patients with myeloproliferative disorders including polycythemia vera , essential thromobocythemia , idiopathic myelofibrocythemia , and hypereosinophilic syndrome [6 , 14] . HNA-2 was an indicator of increased erythropoietic activity in thalassemia syndromes as HNA-2 expression was significantly elevated in β-thalassemia patients compared to healthy controls [45] . HNA-2 overexpression may also have a direct role in the pathogenesis of myeloproliferative disorders as HNA-2 enhances cell proliferation in vitro [46 , 47] . Not surprisingly , the low percentage of HNA-2 positive neutrophil is significantly associated with myelodysplastic syndrome and chronic myelogenous leukemia [48 , 49] , suggesting that the reduced levels of membrane-bound HNA-2 may decrease the proliferation and differentiation potentials of myeloid cells . It is possible that the selection pressure to limit the spread of myeloproliferative disorders during evolution may be an important factor in maintaining the CD177 nonsense polymorphism in humans . Therefore , the CD177 SNP 829A>T may be an important genetic risk factor for various myeloproliferative disorders . Approximately 3% of Caucasians , 5% of African Americans , and 1–11% of Japanese are HNA-2 deficient [14] . In the current study , we found that between 3 . 7% ( main cohort ) and 8 . 8% ( replication cohort ) blood donors ( >98% of them were Caucasians from the State of Minnesota ) were HNA-2 deficient . Our data indicate that percentages of HNA-2 deficient humans may vary in different regions and be affected by sample sizes . In summary , the elucidation of the molecular mechanism of HNA-2 deficiency and expression variations fills the critical knowledge gap in the genetics of HNA-2 antigen system . Our findings will enable the development of reliable genetic assays for HNA-2 system and will facilitate the diagnosis and prognosis of HNA-2-associated human disorders . The human study was approved by the Institutional Review Board for Human Use at the University of Minnesota with Study #1301M26461 . Memorial Blood Centers ( 737 Pelham Boulevard , St . Paul , Minnesota 55114 ) provided healthy donor blood samples without identifications for research purpose as a service and no consent form was provided per the Memorial Blood Centers policy . Healthy American blood donors were recruited at the Memorial Blood Center in St . Paul , Minnesota . The age of healthy blood donors ranged from 19 to 84 years-old and >98% of donors in the study were self-declared Caucasians living in the State of Minnesota . The expression of HNA-2 and the percentage of HNA-2 positive neutrophils were determined with flow cytometry analysis . Leukocytes stained with either FITC-conjugated anti-CD177 mAb ( MEM-166 , mIgG1 , Thermo Scientific ) or mIgG1-FITC isotype control were analyzed on a FACS Canto flow cytometer ( BD Biosciences ) . The FlowJo software ( Tree Star Inc . ) was used to evaluate flow cytometry data . Characteristic light-scatter properties were used to identify neutrophils in flow cytometry . Using the same criteria as in the literature [31] , donors had less than 5% of granulocytes positive for MEM-166 staining in flow cytometry analysis were called as HNA-2 deficient . Peripheral blood leukocytes ( 2 × 107 cells ) were lysed in PBS containing 1% NP-40 and 1× protease inhibitor cocktail ( Roche , Indianapolis , ID ) for 1 hr on ice . The total proteins ( 50 μg ) from each donor were used for Western blotting analysis under non-reducing condition with mouse anti-CD177 mAbs and rabbit anti-actin mAb ( LI-COR Biosciences , Lincoln , NE ) . IRDye 800CW-labeled goat anti-mouse and IRDye 600-labeled goat anti-rabbit antibodies were used for imaging analysis with the instrument software on an Odyssey Infrared Imager according to vendor’s instructions ( LI-COR Biosciences ) . Human genomic DNA was isolated from EDTA anti-coagulated peripheral blood using the Puregene DNA isolation kit ( Gentra Systems , Minneapolis , MN ) by following the vendor’s instruction . Total RNA was purified from peripheral blood leukocytes using TRIzol total RNA isolation reagent ( Invitrogen , Carlsbad , CA ) . The CNVs of CD177 gene were determined using TaqMan Copy Number Assay kit ( the probe location of the assay ID Hs01327659_cn is shown in Fig 1A ) ( Applied Biosystems , Foster City , CA ) and RNase P reference assay ( Applied Biosystems , Part# 4403326 ) . Duplex quantitative real-time PCR reactions were carried out on an Applied Biosystems 7500 Real-Time PCR System according to the manufacturer’s instructions . All samples were tested in duplicates , and fluorescence signals were normalized to ROX . TaqMan assay quantitative PCR amplification curves were analyzed using 7500 Software on a plate by plate basis , and the CN was assigned from the raw Cq values using CopyCaller software ( version 2 . 0; Applied Biosystems ) . Five μg of total RNA was used for cDNA synthesis with the SuperScript Preamplification System ( Invitrogen ) . The 1411-bp cDNA fragment covering the entire CD177 coding region was amplified with RT-PCR using the sense primer ( 5’-CTGAAAAAGCAGAAAGAGATTACCAGCCACAG-3’ ) and anti-sense primer ( 5’-GTCCAAGGCCATTAGGTTATGAGGTCAGA-3’ ) . The PCR reaction was performed with 2 μl of cDNA , 200 nM of each primer , 200 μM of dNTPs , 2 . 0 mM of MgSO4 , and 1 U of Platinum Taq DNA polymerase High Fidelity ( Invitrogen ) in a 25 μl reaction volume . Platinum Taq High Fidelity DNA polymerase was used as it allows the amplification of complex cDNA or DNA templates with high accuracy and yield . The ABI Veriti 96-well Thermal Cycler was used for the PCR reaction starting with 94°C for 3 min , 35 cycles of denaturing at 94°C for 30 s , annealing at 56°C for 45 s , extension at 68°C for 1 min and 30 s with a final extension at 72°C for 7 min . All the PCR products , treated with ExoSAP-IT ( Affymetrix , Santa Clara , CA ) , were assessed by direct Sanger sequencing on an ABI 3730xl DNA Analyzer with BigDye v3 . 1 Sequencing kit ( Applied Biosystems ) . CD177 cDNA was also directly cloned into pCR2 . 1-TOPO vector ( Invitrogen , Carlsbad , CA ) . Multiple clones containing CD177 cDNA were selected and subsequently sequenced to confirm CD177 SNPs . Two sense primers and two antisense primers were used to sequence the full-length CD177 cDNA coding region ( sequencing primers are listed in S3 Table ) . The electropherogram data , aligned by the DNASTAR software ( DNAStar , Madison , WI ) were used for the identification of gene polymorphisms . Since CD177 and its pseudogene contain a highly homologous region between exon 4 and 9 ( Fig 1A ) [16 , 17] , we used the long-template PCR strategy to obtain the CD177-specific products for sequence analyses . Long-template PCR was carried out to amplify the CD177 genomic DNA containing all 9 exons using the sense primer ( 5’-CTGAAAAAGCAGAAAGAGATTACCAGCCACAG-3’ ) and antisense primer ( 5’-GTCCAAGGCCATTAGGTTATGAGGTCAGA-3’ ) . The PCR reaction was performed with 200 ng DNA , 200 nM of each primer , 200 μM of dNTPs , 2 . 0 mM of MgSO4 , and 2 U of Platinum Taq DNA polymerase High Fidelity ( Invitrogen ) in a 25 μl reaction volume . The ABI Veriti 96-well Thermal Cycler was used for the PCR reaction starting with 95°C for 3 min; 10 cycles of denaturing at 94°C for 30 s , annealing at 64°C for 30 s , extension at 68°C for 8 min and 30 s; 30 cycles of denaturing at 94°C for 30 s , annealing at 54°C for 30 s , extension at 68°C for 8 min and 30 s; with a final extension at 68°C for 5 min . The CD177 DNA fragment ( 8728 base pairs ) was sequenced with a primer ( 5’-TCTTTGCCCCACACTAAACA-3’ ) on an ABI 3730xl DNA Analyzer with BigDye v3 . 1 Sequencing kit . The human HNA-2 expression constructs were generated by cloning Hind III/Xba I-flanked RT-PCR products containing full-length CD177 coding region ( nucleotide position 25 to 1419 , GenBank accession number: NM_020406 . 2 ) into the eukaryotic expression vector pcDNA3 ( Gibco BRL ) . The Hind III/Xba I-flanked CD177 cDNA was amplified from the synthesized cDNA of a blood donor using the upper primer 5’-CCCAAGCTTACCAGCCACAGACGGGTCATGAG-3’ and the lower primer 5’-TGCTCTAGAGAGGTCAGAGGGAGGTTGAGTGTG-3’ . The changes at nucleotide position 134 , 652 , 656 , 824 , 828 , 829 , 832 , 841 , 997 , and 1084 were generated using QuikChange Site-Directed mutagenesis kit ( Stratagene , La Jolla , CA ) and primer sets listed in the S3 Table . The 293 cells ( human embryonic kidney cell line ) from ATCC ( ATCC#CRL-1573 , Manassas , VA ) were maintained in the DMEM medium supplemented with 10% fetal calf serum and L-glutamine ( 2 mM ) in 5% CO2 . Transfection reactions were carried out in the 100 mm cell culture dishes with the plasmid DNA ( 20 μg ) purified with OMEGA Plasmid Maxi Kit ( Omega Bio-Tek , Norcross , GA ) and 40 μl of Lipofectamine 2000 reagent ( Invitrogen ) . Transfected cells were cultured in DMEM medium supplemented with 10% fetal calf serum for two days before harvesting the cells for HNA-2 expression or the selection of stable cell lines with the supplement of G418 ( final concentration: 1 mg/ml ) . The polyclonal cells surviving the G418 selection were sorted with Stemcelll EasySep Cell Sorter for equivalent HNA-2 expression . The expression of HNA-2 on the transfected 293 cell lines was determined with FITC-conjugated anti-CD177 mAb as described previously . In addition , five defined HNA-2 alloantibodies from the American Red Cross Neutrophil Serology Laboratory were used to evaluate the binding of HNA-2 to the cell lines expressing CD177 variants . The ANOVA and the nonparametric t-test ( Mann-Whitney test ) were used to determine whether HNA-2 positive cell population sizes and the HNA-2 deficiency are statistically associated with the nonsense CD177 cSNPs . The χ2 test was used to determine whether the observed genotype frequencies are consistent with Hardy-Weinberg equilibrium .	Human neutrophil antigen 2 ( HNA-2 ) is coded by CD177 gene that involves in human myeloproliferative disorders . HNA-2 expression varies among humans and about 3–5% people lack HNA-2 expression . HNA-2 deficient people are susceptible to produce HNA-2 alloantibodies , which play a pathological role in various human diseases including transfusion-related acute lung injury , neonatal alloimmune neutropenia , autoimmune neutropenia , drug-induced immune neutropenia , and graft failure following marrow transplantation . The level of HNA-2 expression has also been identified as a prognostic biomarker for the gastric cancer . Although HNA-2 is among the most important clinical antigens , the underlying genetic mechanism of HNA-2 deficiency and expression variations has remained unknown . Here , we demonstrate that HNA-2 deficiency and expression variations are primarily caused by a novel CD177 genetic polymorphism that disrupts HNA-2 expression . The illumination of genetic mechanism for HNA-2 deficiency and expression variations will enable the development of effective HNA-2 genetic tests . Our findings will facilitate prognosis and diagnosis of HNA-2-related human disorders .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[]	2015	Genetic Mechanism of Human Neutrophil Antigen 2 Deficiency and Expression Variations
Natronomonas pharaonis is an archaeon adapted to two extreme conditions: high salt concentration and alkaline pH . It has become one of the model organisms for the study of extremophilic life . Here , we present a genome-scale , manually curated metabolic reconstruction for the microorganism . The reconstruction itself represents a knowledge base of the haloalkaliphile's metabolism and , as such , would greatly assist further investigations on archaeal pathways . In addition , we experimentally determined several parameters relevant to growth , including a characterization of the biomass composition and a quantification of carbon and oxygen consumption . Using the metabolic reconstruction and the experimental data , we formulated a constraints-based model which we used to analyze the behavior of the archaeon when grown on a single carbon source . Results of the analysis include the finding that Natronomonas pharaonis , when grown aerobically on acetate , uses a carbon to oxygen consumption ratio that is theoretically near-optimal with respect to growth and energy production . This supports the hypothesis that , under simple conditions , the microorganism optimizes its metabolism with respect to the two objectives . We also found that the archaeon has a very low carbon efficiency of only about 35% . This inefficiency is probably due to a very low P/O ratio as well as to the other difficulties posed by its extreme environment . Natronomonas pharaonis is a polyextremophilic archaeon that can be isolated from soda lakes , where it has to cope with two extreme conditions: high salt concentration and an alkaline pH . The microorganism thrives at an optimal pH of 8 . 5 , and remains viable up to a pH of about 11 . Two strains of Natronomonas pharaonis have been described so far: strain Gabara from lake Gabara in Egypt ( DSM 2160 ) [1] , which was used in this study , and strain SP1 from lake Magadi in Kenya ( DSM 3395 ) [2] . Among other results , we show that the microorganism is able to grow on a single carbon source , such as acetate , glutamate and pyruvate , unlike the more well-studied halophilic archaeon Halobacterium salinarum . This greatly simplifies several experimental protocols , in particular those based on stable isotope labeling . The combination of this capability , a completely sequenced genome [3] and the recent successful development of a transformation protocol ( separate publication ) will increase the prominence of Natronomonas pharaonis in the study of extremophilic life . The metabolic network of an organism can be reconstructed at the genome scale through the combination of genomic , biochemical and physiological data , using bioinformatics methods and literature review [4]–[8] . The resulting network , comprised of the known and hypothesized reactions that take place within the organism , is valuable in that it forms a knowledge base of cellular metabolic capabilities . For example , inspection of the network would allow one to draw hypotheses regarding nutritional requirements and biosynthetic capabilities . Moreover , metabolic reconstructions can also serve as convenient starting points for crafting genome-scale , constraints-based models of metabolism , which allow more detailed computational/formal analysis . Indeed , constraints-based models have emerged as important alternatives to kinetic models because they do not require the detailed kinetic information needed by the latter . Rather , constraints-based models require only generally available physicochemical information such as stoichiometry , reversibility and energy balance [9]–[11] , data which are already typically included in , or at least could easily be derived from , metabolic reconstructions . More sophisticated data , such as flux ( reaction velocity ) limits , could also be easily integrated . The repertoire of computational methods available under , or related to , the constraints-based framework include extreme pathways [12] , elementary modes [13] , and flux-balance analysis with its derivatives [14]–[16] . Genome-scale models of metabolism have been constructed and analyzed for various organisms [6] , yielding interesting results , such as the prediction of E . coli metabolic mutant phenotypes up to an accuracy of 86% [17] , a simulation and characterization of E . coli's secreted metabolites under various nutrient and oxygenation conditions [14] , a study of the methanogenic growth of M . barkeri [18] , and an analysis of growth on a relatively complex medium for L . plantarum [19] and Halobacterium salinarum [20] , [21] . In this work , we present a manually curated , genome-scale metabolic reconstruction for Natronomonas pharaonis , and report on the development and use of a constraints-based model from it . This is the third curated reconstruction for a microorganism belonging to the archaeal domain of life; the first being for M . barkeri [18] and the second for Halobacterium salinarum [20] , [21] . We also report on the experimental determination of several parameters relevant to growth , including material uptake , respiratory rates , and the distribution of carbon in the biomass ( e . g . , amino acids , nucleotides , etc . ) . The reconstructed Natronomonas pharaonis metabolic network is composed of 683 reactions and 597 distinct metabolites . It covers 654 genes , not including those with known transport function but with unclear substrate specificity . Reactions were added to the network based on either genetic ( e . g . , homologs of known enzyme-coding genes ) or literature ( e . g . , enzyme assays , labeling studies ) evidence . Figure 1 shows the distribution of the reactions based on the former type of supporting data . Specifically , it shows the number of reactions , grouped according to general functional categories , for which: ( 1 ) enzyme-coding genes could be reliably assigned; ( 2 ) only genes with general functional annotation ( i . e . , with unclear substrate specificity ) could be associated; and ( 3 ) no genetic evidence could be found . Given the relatively recent isolation of Natronomonas pharaonis , its literature base , particularly for subjects relevant to metabolism , is still quite small . This is the reason why most reactions in the network only have bioinformatic support . Nevertheless , at least 168 ( 24% ) of the reactions are associated with literature ( experimental ) evidence from related haloarchaeal species ( e . g . , Halobacterium salinarum ) . Note that reactions with neither genetic nor literature support were nevertheless added to the network in order to fill pathway “gaps” . We obtained an initial draft of the metabolic network by merging the reaction database ( LIGAND ) from KEGG [22] with the Natronomonas pharaonis genome annotation found in Halolex [23] , [24] . The latter resource is a genome information system that specializes in halophilic microorganisms . Because of the procedure used , most of the reactions in our reconstructed network are defined according to the definitions found in the LIGAND database . Nevertheless , the network also contains reactions that had to be manually defined , such as newly characterized pathways that are not yet contained in KEGG . This is particularly relevant since the archaea regularly use pathways that are different from , or modifications of , those studied in most model organisms , which are often from the bacterial domain of life , and it typically takes a while before these are reflected in databases . For example , it was recently shown that aromatic amino acid biosynthesis in Methanocaldococcus jannaschii does not use the classical precursors erythrose 4-phosphate and phosphoenolpyruvate , but rather proceeds via an alternative pathway that begins with the condensation of 6-deoxy-5-ketofructose 1-phosphate with aspartate 4-semialdehyde [25] . Genetic evidence suggests that the same ( or a similar ) pathway is in operation in Natronomonas pharaonis . This alternative pathway was still not in KEGG when we did our reconstruction . Another example is the modified mevalonate biosynthesis pathway that was also demonstrated in Methanocaldococcus jannaschii [26] . The modification to the classical pathway , essentially a change in ordering between a decarboxylation and a phosphorylation step , is still not reflected in KEGG . Again , genetic evidence suggests that this pathway is also in operation in Natronomonas pharaonis . The reconstructed network is provided as supplementary information , both in tabular ( Table S1 ) and SBML format ( Text S2 ) . Although in silico gene deletion simulations were not part of our analysis , information on putative logical relationships between genes and reactions was nevertheless included in the reconstruction . For example , the oxidation of pyruvate to acetyl-CoA and CO is catalyzed by the polymeric enzyme , pyruvate-ferredoxin oxidoreductase ( EC 1 . 2 . 7 . 1 ) . This protein consists of two distinct subunits , which in Natronomonas pharaonis are encoded by NP4044A ( beta subunit ) and NP4046A ( alpha subunit ) . Accordingly , the reaction was assigned the gene-logic formula NP4044A NP4046A , because , presumably , both subunits are required in order for the enzyme to function . We should note , however , that given the current lack of appropriate genomic data , such as genome-scale single deletion lethality assays , the logical relationships were mostly inferred only from the genome annotation ( e . g . , keywords such as “subunit” or “component” in the case of complex enzymes ) . Accordingly , this information should be taken with a grain of salt . For example , four genes are currently annotated in the genome as “probable aspartate aminotransferase ( EC 2 . 6 . 1 . 1 ) ” , and accordingly appear as disjuncted terms in the gene-logic of the reaction catalyzed by the enzyme EC 2 . 6 . 1 . 1: However , even if all four genes are indeed aspartate transaminases , there is a good probability that not all of them actually catalyze the reaction defined above . Some could work with an acceptor different from 2-Oxoglutarate , and as such are not true isoenzymes . All of the reactions in the reconstructed network are mass and charge balanced , except for 10 . The latter could not be balanced because some reactants are still unknown . Seven of the reactions are involved in cofactor biosynthesis , and the rest belong to amino acid degradation pathways . The ionization state of the metabolites reflected in the network is that of the most abundant microspecies under a pH of 9 . 0 . In flux balance analysis , growth is typically simulated through the use of a growth reaction , which is a pseudo reaction where the reactants are the components of the biomass in the proper ratios and the product is a unit of population ( e . g . , mg DW , ODml ) . Clearly , a prerequisite for the definition of this reaction for Natronomonas pharaonis is an approximation of the microorganism's average biomass composition , at least under the conditions used . For this we determined the total organic carbon content of samples taken from cultures at different population levels ( optical densities ) . The data is summarized in Figure 2 . Overall , we calculated a linear correlation of 18 . 22 . 6 mmol organic carbon per ODL , compared to 23 . 1 mmol per ODL in the closely related halophile Halobacterium salinarum [21] . Determination of the total organic carbon content of the biomass is only the first step . The next is to decompose it into the different building blocks . Given that proteins typically account for a large fraction , if not a majority , of the organic mass of most microorganisms , we began our characterization of the Natronomonas pharaonis biomass by analyzing its amino acid content . Specifically , we took detailed quantitative measurements of the amino acid composition of cell cultures at different population levels ( Figure 3; see Growth on Acetate ) . From these measurements , we were able to observe good linear correlations between the amino acids and the optical density , which implies that the average cellular amino acid composition remains reasonably constant throughout growth . This observation reinforces the validity of using a static definition of growth in flux balance models , such as the one used in this study . Note that due to experimental limitations , measurements for aspartate and glutamate are already inclusive of asparagine and glutamine , respectively , and cysteine and tryptophan could not be reliably determined . For these amino acids , statistical analysis of the predicted Natronomonas pharaonis proteome was used to specify coefficients in the growth reaction ( see Growth on Acetate ) . The amount of amino acids in the biomass added up to about 335 mg/ODL , which represents approximately 75% of the total organic mass or 80% of the total organic carbon content . The reason for this very high amino acid ( protein ) composition is unclear . A similar relationship between the amino acids and the organic mass was also observed in Halobacterium salinarum ( data not shown ) . 2-Sulfotrehalose has been demonstrated to be an osmolyte used by several haloalkaliphilic archaea , including Natronomonas pharaonis [27] . This molecule was reported to accumulate in amounts over 2 . 5 mol per mg protein [27] , which translates to about 855g of the compound per mg protein . Clearly , such a high intracellular concentration of 2-sulfotrehalose can be excluded under the current conditions , given that proteins already account for 75% of the total organic mass . Thus , the likely scenario is that under the conditions used , cells rely predominantly , if not completely , on inorganic ions to counter osmotic stress; a situation that has already been observed for Natronomonas pharaonis under other conditions [27] . Indeed , using NMR to analyze intracellular solutes ( see Detection of 2-Sulfotrehalose ) we were not able to detect any dissacharide . In addition to amino acids , the cellular biomass is also made up of other molecules such as lipids ( e . g . , archaeaol ) , sugars ( e . g . , S-layer glyco moieties ) , cofactors ( e . g . , NAD/P , coenzyme-A , retinal ) , and other small molecules . Due to lack of experimental data , the stoichiometric coefficients that we used for these molecules in the growth reaction were simply taken from the approximations used in the Halobacterium salinarum model [21] , calculated in proportion to the ratio of the measured amino acid contents of the two halophiles . Finally , the remaining amount of organic mass ( from the total organic carbon analysis ) was assumed to comprise of nucleotides ( approximately 20% ) . Separate coefficients for the different nucleotide molecules were calculated in proportion to the 63 . 4% GC content of the genome . The final growth reaction we used in our model is shown in Table 1 . It was previously believed that any growth medium for Natronomonas pharaonis would require the presence of leucine because of a disruption in the 2-isopropylmalate synthase gene ( NP2206A ) [3] , which is involved in the biosynthesis of the amino acid . However , in the course of optimizing the carbon sources supplied in the medium , which we undertook in the interest of designing stable isotope experiments , we observed that the haloalkaliphile was able to grow on a medium that contained just acetate . Subsequent resequencing of the 2-isopropylmalate synthase gene showed that the gene is still interrupted , so the reason for the phenotype is uncertain . Nevertheless , it is now clear that Natronomonas pharaonis is capable of growing ( aerobically ) on a single carbon source . Indeed , although the rest of this study deals with growth on acetate , we were able to find other possible substrates , including glutamate and pyruvate . In order to simulate aerobic growth on acetate using the constraints-based model , three parameters needed to be further defined: ( 1 ) the consumption rate of acetate , ( 2 ) the consumption rate of oxygen , and ( 3 ) the maintenance energy . Under certain assumptions however ( see Computational Analysis for details on the calculations ) , it turns out that what is critical to the first two parameters with respect to the analysis is just the ratio between the two , i . e . , the ratio between acetate and oxgyen usage . Scaling the actual values simply results in a similar scaling of the output of the model ( in this case , the growth flux ) . Accordingly , for most of the analysis we could simplify the calculations by reducing the parameter space to just two instead of three; namely , the acetate to oxygen consumption ratio and the maintenance energy . The model , through the definition of the growth reaction , already accounts for the energy involved in synthesizing the basic building blocks of the biomass ( see Table 1 for a complete list ) from the supplied materials . This is accomplished in an implicit manner based on the pathways defined in the metabolic network . However , there are numerous other energy-consuming processes performed by cells , such as assembling the building blocks into larger structures , motility , repair , cellular division , etc . These other processes are represented by the maintenance energy parameter . The theoretical maximum level of growth as a function of the acetate∶oxygen parameter and the maintenance energy is plotted in Figure 4 . The latter parameter is defined in terms of the equivalent mols of ATP hydrolized in producing an ODml worth of biomass ( i . e . , mols of ATP per ODml ) . It is immediately recognizable from the figure that growth is not possible for a very high ( 7∶3 ) or a very low ( 3∶7 ) acetate to oxygen consumption ratio . Indeed , the model clearly indicated that having such ratios is physiologically impossible . Formally speaking , the linear programs do not have feasible solutions under such settings of the parameter . In the case where the ratio is too high , i . e . , too much acetate , energy cannot be produced in amounts sufficient to process all of the consumed material . Note that all of the carbon ( acetate ) taken up by cells will have to be either: ( 1 ) incorporated into the biomass; ( 2 ) secreted as the respiratory byproduct CO; or ( 3 ) secreted as some other metabolite after conversion . Accordingly , if the acetate to oxygen consumption ratio is too high , enough energy that would allow any distribution of the consumed carbon into the possible fates simply cannot be generated . In an analogous way , a very low acetate to oxygen consumption ratio is also impossible because under such conditions , there is not enough organic material that can be oxidized in order to allow the conversion of all the consumed oxygen to HO . Optimal acetate to oxygen consumption ratios as a function of the maintenance energy are shown in Figure 4 using the red broken curve ( see also figure inset ) . They are optimal in the sense that they are the consumption ratios that would theoretically allow the highest level of growth for the different values of the maintenance energy ( i . e . , the amount of energy required to produce a unit of biomass ) parameter . Notice that as the maintenance energy parameter is increased , the optimal acetate to oxygen ratio decreases . This is because under the conditions used , energy can only be derived through the oxidation of acetate . Thus , increasing the amount of energy needed to produce a unit of biomass means that a greater fraction of the consumed carbon ( acetate ) will have to be converted into energy . In addition , oxidization of a greater amount of acetate would require a greater amount of oxygen , and this also decreases the optimal ratio . It is clearly of interest to compare how the theoretical results summarized in Figure 4 compare to the real-world behavior of Natronomonas pharaonis . Accordingly , we prepared actual cultures where we grew the archaeon aerobically on acetate . The experimental data is summarized in Figure 2 of Text S1 ( supplementary information ) . During the early stages of growth , we observed an acetate∶oxygen comsumption ratio of about . This is represented in Figure 4 by the green shaded region . In addition , we also approximated the maintenance energy by calculating the amount of oxygen consumed per unit increase in population size ( optical density ) , and then multiplying the result with a P∶O ratio of 1∶1 . This calculation resulted in a maintenance energy of about mols ATP per ODml , which is represented in the figure by the orange shaded region . Not much is currently known regarding the biochemistry of the respiratory chain of Natronomonas pharaonis . Nevertheless , the pathway has been demonstrated to be functional , by showing increases in both ATP level and membrane potential in response to aeration [3] . With respect to the genome , complete sets of ORFs encoding analogs of Complex II and Complex IV genes are present , but none could be assigned for Complex III . Homologs of most Complex I subunits are present , but subunits comprising the NADH acceptor module ( nuoEFG ) could not be assigned . Indeed , NADH dehydrogenation likely occurs via a non-proton pumping type II NADH dehydrogenase NP3508A , which is a homolog of the Acidianus ambivalens ndh gene [28] . However , note that the presence of the type II NADH dehydrogenase does not preclude the possibility of another donor molecule for which proton translocation is possible . Because of the relatively scarce information available for Natronomonas pharaonis , the P∶O ratio that we used to calculate the maintenance energy above is actually the experimental value for the closely related Halobacterium salinarum , which was determined using oxygen pulse experiments [29] , [30] . Thus , the maintenance energy value reported above is still subject to scaling . Nevertheless , interpreting it in terms of the O consumed per unit increase in biomass ( i . e . , mols O per ODml ) , instead of ATP , should represent a fairly accurate approximation of the parameter . The intersection of the regions corresponding to the experimentally determined values of the acetate to oxygen consumption ratio ( green region in Figure 4 ) and the maintenance energy ( orange region ) could be thought of as the area representing experimentally observed growth behavior ( shaded yellow in the figure ) . It is immediately recognizable that this region is within the space that the model considers to be physiologically permissible with respect to the two parameters . However , what is more interesting is the fact that all of the points in this region are at least reasonably close to the optimality curve ( red curve in the figure ) . This means that the cells must be using an acetate to oxygen consumption ratio that is at least near-optimal with respect to growth and energy production ( 90% based on theoretical calculations ) , whatever the actual maintenance energy may be . Indeed , the region is located on a relatively flat area of the surface , where “near optimality” is robust with respect to the two parameters . While the observations above , by themselves , say nothing directly about the optimality of the actual fluxes used by the cells , it is at least an argument in favor of the hypothesis that Natronomonas pharaonis , under the conditions , optimizes its metabolism with respect to growth and energy production . Indeed , it has been demonstrated in Escherichia coli that under similar , simple growth conditions , a theoretical optimization of the fluxes with respect to energy provides a reasonable approximation of the actual fluxome [31] , [32] . We mentioned earlier that all of the carbon consumed by cells will have one of the following three fates: ( 1 ) incorporated into the biomass; ( 2 ) secreted as the respiratory byproduct CO; or ( 3 ) secreted as some other metabolite after conversion . Given that only acetate was supplied in the medium , then the only other possible way through which the cells could incorporate carbon is through CO fixation . This capability is suggested by the presence of various enzymes in the genome of Natronomonas pharaonis , so we checked if it is indeed a contributing factor under the conditions used . Specifically , we supplied 13C sodium carbonate in the medium , and then tried to see if label would turn up in amino acids ( which represent 75% of the biomass ) . Results of the stable isotope experiments showed that no net fixation seems to occur ( data to appear in a separate publication ) . Accordingly , the amount of acetate that disappears from the medium is a direct measure of the total carbon consumption of the cells ( i . e . , two carbon atoms per acetate molecule ) . By correlating this consumption , which is represented as “total uptake” ( red borken curve ) in Figure 5 , with the total amount of carbon in the biomass ( see Analysis of biomass composition ) , we find that only about 35% of the total carbon consumed was incorporated into the cells ( i . e . , fate 1; blue curve in Figure 5 ) . This very low carbon efficiency is likely due to a very small P/O ratio as well as to the other difficulties presented by the extreme environments of the microorganism . The respiratory exchange ratio ( RER ) is the ratio between the production of CO and the consumption of O . This ratio can be theoretically calculated for the complete oxidation of any metabolite . In the case of acetate ( and some sugars in general ) , the RER is typically close to one . We used this fact to approximate the CO production of our cultures , specifically by multiplying the experimentally determined oxygen consumption with it . We found respiratory-related CO production to account for about 63% of the total carbon consumption . The sum of fates 1 ( incorporation ) and 2 ( CO production ) is plotted ( green curve ) in Figure 5 to provide a comparison with the total carbon consumption ( red curve ) . Clearly , the difference between this sum and the total carbon consumption corresponds to material that was consumed but neither incorported into the biomass nor oxidized to CO . While this relatively small difference , which is represented in the figure by the red shaded region , could potentially be due to carbon being secreted in some other form ( i . e . , fate 3 ) , it is more likely to simply be due to small methodological inaccuracies , such as deviations in the actual RER ratio or errors in the measurements . That is , under the conditions used , biomass incorporation and CO production together likely fully account for carbon consumption . We presented a genome-scale , manually curated metabolic reconstruction for the polyextremophile Natronomonas pharaonis . The reconstruction itself represents a summary of the knowledge regarding the haloalkaliphile's metabolism . As such , it would greatly assist future investigations on archaeal pathways . This is particularly relevant since the archaea are known to , quite frequently , use novel or modified pathways from those in existing model organisms [24] . An existing knowledge base that gives an overview of metabolism , such as this reconstruction , could , for example , assist in identifying knowledge “holes” or “gaps” , which are promising directions for further study . Indeed , this is very timely as research on haloarchaeal metabolism , up until recently , has been limited by the experimental protocols available for its model organisms . For example , while genetic transformation was possible for Halobacterium salinarum , the microorganism required multiple carbon sources in the medium , and this complicated other protocols such as stable isotope labeling . In the case of Natronomonas pharaonis , work was limited by the absence of a procedure for genetic modification . Accordingly , the recent development of a transformation protocol for Natronomonas pharaonis ( separate publication ) means that we now have a haloarchaeon for which genetic manipulation is possible and labeling experiments are relatively simple . In order to complete a constraints-based model for Natronomonas pharaonis , we experimentally determined several physiological parameters . These include a characterization of the biomass composition , and a quantification of carbon and oxygen consumption under typical conditions . The data was integrated with the metabolic reconstruction to create a computational model that we used to analyze the behavior of Natronomonas pharaonis when grown on a single carbon source . Among other results , we found that the archeaon , when grown aerobically on acetate , uses an acetate to oxygen consumption ratio that is theoretically near-optimal with respect to growth and energy production . This supports the hypothesis that , under simple conditions , Natronomonas pharaonis optimizes its metabolism with respect to these two objectives . We also found that the archaeon has a very low carbon efficiency of only about 35% , likely due to a very low P/O ratio and the other difficulties brought about by the haloalkaliphilic character of its environment . Stable isotope labeling of small-molecules ( e . g . , C-acetate in this case ) is a powerful tool for characterizing metabolic pathways . However , this approach requires detailed information on the fates of individual atoms in each reaction , and these can be very difficult to collect in a useful form . For this reason , one of the things that we are currently working on is to add carbon fate data to the metabolic network reconstruction . Our eventual goal is to elucidate some of the currently unclear archaeal pathways , for example to identify the exact precursors of archaeal aromatic amino acid biosynthesis [25] , and to directly measure the internal fluxes of Natronomonas pharaonis [31] , [32] , such as in response to different stimuli . A prior partial reconstruction of the Natronomonas pharaonis metabolic network was downloaded from the PATHNET database of the Halolex system [23] , [24] . We used this as the starting point of our reconstruction . Evidence used to support the inclusion of a reaction/enzyme into the network is of two types: bioinformatic evidence , e . g . , high similarity to a known enzyme; and literature evidence , such as experimental reports of enzyme assays , labeling studies , and nutrient uptake ( for transporters ) . Pathway gaps , i . e . , reactions devoid of literature or genetic indicators but belonging to pathways that seem to be present in the microorganism , were added to the network regardless ( and marked accordingly ) , particularly if they are involved in the production of a compound essential to growth . Since the current body of metabolic literature for Natronomonas pharaonis is still quite small , data from closely related microorganisms was used , particularly from the intensively studied halophile , Halobacterium salinarum . Reaction reversibilities were taken from literature sources whenever available . Particularly helpful in this regard is the BRENDA database [33] , [34] , which , among other things , stores reversibility information from various organisms . In the absence of such data , we used the assignments made in existing reconstructions as well as those made by Ma and Zeng [35] , who annotated the reversibility of reactions defined in KEGG [22] based on biochemical principles . As a consequence of the reconstruction procedure that we used , most of the reactions in the network are based on the definitions in KEGG . However , a number had to be manually added because some reactions , particularly those from newly characterized archaeal pathways , still do not exist in the database . Moreover , KEGG reactions had to be modified in order to correct errors or to achieve elemental and charge balance . ORFs with a known general function but unclear substrate specificity were linked to all the possible reactions that they could affect ( marked accordingly ) . Microspecies distributions with respect to ionization states at pH 9 . 0 were approximated using the pKa Plugin for Marvin ( Marvin 5 . 3 . 0 , 2009 , ChemAxon; http://www . chemaxon . com ) . The reader interested in more information on metabolic reconstructions is directed to the reviews by [5] and [4] . A metabolic network can be conveniently represented by a stoichiometric matrix , where each column corresponds to a reaction and each row to a metabolite . The entries of are the stoichiometric coefficients that define the relationships between the reactions and compounds . A positive value for indicates that compound is produced in the left to right direction of reaction , while a negative value indicates that it is consumed . A set of fluxes that is both consistent with the known constraints and optimal with respect to some objective function can be obtained by solving the linear program ( 1 ) where is a vector of fluxes defining the flux through each reaction , is the objective function , is the set of reversible internal reactions , is the set of irreversible internal reactions , is the set of exchange fluxes associated with ubiquitous metabolites , and is the set of exchange fluxes that correspond to metabolites that were treated as parameters during analysis ( e . g . , acetate and oxygen uptake ) . The set of ubiquitous compounds that we used include CO , HO , Na , Cl , K and H . For the purpose of the analysis , sulfate , orthophosphate , as well as a number of other ions were also assumed to be available in abundance . To account for the possibility that cells produce and accumulate certain metabolites in the medium , for example in the case of overflow metabolism , the set of one-way exchange reactions , , was included . In the analysis of growth with respect to the acetate∶oxygen consumption ratio and the maintenance energy , Equation 1 was solved for different combinations of the two parameters using the growth reaction as objective function ( see Analysis of biomass composition ) . Specifically , the acetate∶oxygen consumption ratio was allowed to vary from 1∶9 to 9∶1 , and the maintenance energy was allowed to vary from 0 to 100 mol ATP per ODml . Most of our calculations were carried out using Matlab ( TheMathworks , Massachusetts , USA ) . However , the default linear programming package was unable to handle our network due to its size , so we had to integrate the GNU Linear Programming Kit ( GLPK ) to handle flux analysis . For the analysis of growth , Natronomonas pharaonis strain Gabara ( DSM 2160 ) was grown under aerobic conditions using chemically-defined media prepared according to Table 1 of Text S1 ( supplementary information ) . Preparatory cultures were grown in 100 ml flasks containing 35 ml of the medium , from which inoculants were taken to start successive cultures . This was done repeatedly to adapt cells to the growth conditions . All cultures were prepared in flasks which had side arms to measure turbidity ( cell density ) via a Klett photometer , and were carried out at least in duplicates . Cell suspensions were shaken at 105 rpm at 40C in the dark . At specific points during the growth period , samples were collected , and these were stored at −20C . To separate the cells from the medium , samples were centrifuged for five minutes at 15 , 000 rpm , using a SS34 rotor . Amino acid analysis was performed on the pellets to determine the amino acid composition of the biomass , using a Biotronik LC 3000 analyser ( Biotronik , Maintal , Germany ) . Oxygen saturation in the medium was monitored using the “Fibox 3-trace v3 , fiber-optic oxygen meter” from Precision Sensing GmbH ( Regensburg , Germany ) . Details on the calculation of actual oxygen consumption are provided as supplementary information ( Text S1 ) . Due to experimental limitations , measurements of the aspartate and glutamate content of the biomass were already inclusive of asparagine and glutamine , respectively . For each of these pairs , specific coefficients in the growth reaction were assigned by dividing the combined total according to the relative abundance of the correspoding amino acids in all protein-coding genes ( calculated assuming one copy per gene ) . Similarly , the individual contributions of cysteine and tryptophan to the biomass were approximated using their relative abundance compared to the rest of the amino acids ( for which measurement was possible ) . The formulation of the growth medium ( Table S4 ) used in the analysis of aerobic growth is the result of optimizing previous synthetic media used for Natronomonas pharaonis . In particular , we tested different carbon sources , including all of the amino acids , acetate , glycerol , citrate cycle intermediates , and combinations thereof , in different concentrations using 96-well microtiterplates ( Greiner bio-one ) . The total organic carbon ( TOC ) content of the Natronomonas pharaonis biomass was determined by adapting a procedure typically used to determe the TOC of soils . The following solutions were used: Solution 1 was prepared by dissolving 12 . 5 grams of silver sulfate in 40 ml of HSO; Solution 2 was prepared by dissolving 98 . 08 grams of KCrO in deionized water to a total volume of 1 L; and Solution 3 was prepared by dissolving 11 . 2 grams of FeSO*7HO in deionized water containing 6 ml of concentrated HSO to a final volume of 400 ml . The procedure is as follows: the precipitate of each sample was collected after centrifugation , and this was resuspended in 500 L of deionized water . Next , 1 . 5 ml of Solution 1 was added in order to precipitate Cl with silver . This step is necessary because Cl , which is abundant in the biomass , interferes with a reaction downstream of the process . The resulting solution was centrifuged for 30 minutes , and 1 ml of the supernatant was transferred to a reaction vial for the succeeding steps . After adding 500 L of concentrated HSO and 500 L of Solution 2 to the vial , the solution was heated to 135C , and then set aside to cool slowly . Once cooled to room temperature , the remaining amount of KCrO ( added with Solution 2 ) , which oxidizes the organic carbon in the biomass , was then determined via titration with Solution 3 . This value is inversely related to the amount of carbon in the biomass . Titration was carried out using the TitroLine easy automatic titrator from Schott Instruments ( Mainz , Germany ) . Natronomonas pharaonis was grown in larger volumes ( 200 ml ) to mid-log phase ( OD 0 . 3–0 . 4 ) , after which cells were harvested by centrifugation ( 6000 g , 30 min ) . Cell pellets were washed three times with 3 . 4 M NaCl , and then dried in a speed vac centrifuge . The pellets were suspended in 1 . 5 ml 70% ( v/v ) EtOH and vortexed for 5 min . Sonication for 5 min in a water bath followed , before centrifugation at 11 000 g for 10 min at 4C in an Eppendorf microcentrifuge . The extraction step was repeated twice , and the ethanolic supernatants were pooled in a small pointy glas flask . Most of the solvent was removed by rotary evaporation , and final sample drying was achieved by centrifugation in a speed vac . For NMR analysis , samples were redissolved in 160 l DO ( 99 . 9% ) and centrifuged for 10 min at 10 000 g in an Eppendorf centrifuge . A thin top layer was discarded , and the rest of the aqueous supernatant taken for measurements . A H WALTZ-decoupled C-NMR spectrum was recorded using a Bruker DRX 500 spectrometer at 500 MHz with a 5 mm TXI probe head . Acquisition parameters included 33 . 3 kHz sweep width , 32 k datapoints , 90 pulse angle , 100 k transients and a temperature of 298 K . Chemical shifts were measured relative to dioxane ( 67 ppm ) .	Extremophiles are organisms that thrive in physically or geochemically extreme conditions that are detrimental , even lethal , to the majority of life on Earth . Natronomonas pharaonis is one that has been able to adapt to both high salt concentration and an alkaline pH . In this study , we investigate the chemical reactions that occur within the microorganism , collectively referred to as its metabolic network , that allow it to convert the nutrients in its environment to biomass and energy . Specifically , we reconstructed the network by collecting evidence for the existence of reactions from the literature , and then supplemented them with computational approaches , for example by searching the genome of Natronomonas pharaonis for genes that could potentially encode analogs of known enzymes from other organisms . Finally , with the network in hand , we developed a computational model which we used to simulate growth . Among other results , we found indications that Natronomonas pharaonis regulates its metabolism such that energy production and growth are maximized . Despite this however , we also found that Natronomonas pharaonis is only able to incorporate a very small fraction of the total carbon that it consumes ( approximately 35% ) , likely in no small part due to the difficulties posed by its environment .	[ "Abstract", "Introduction", "Results/Discussion", "Materials", "and", "Methods" ]	[ "microbiology/microbial", "physiology", "and", "metabolism", "computational", "biology/metabolic", "networks" ]	2010	Characterization of Growth and Metabolism of the Haloalkaliphile Natronomonas pharaonis
The zebrafish adult pigment pattern has emerged as a useful model for understanding the development and evolution of adult form as well as pattern-forming mechanisms more generally . In this species , a series of horizontal melanophore stripes arises during the larval-to-adult transformation , but the genetic and cellular bases for stripe formation remain largely unknown . Here , we show that the seurat mutant phenotype , consisting of an irregular spotted pattern , arises from lesions in the gene encoding Immunoglobulin superfamily member 11 ( Igsf11 ) . We find that Igsf11 is expressed by melanophores and their precursors , and we demonstrate by cell transplantation and genetic rescue that igsf11 functions autonomously to this lineage in promoting adult stripe development . Further analyses of cell behaviors in vitro , in vivo , and in explant cultures ex vivo demonstrate that Igsf11 mediates adhesive interactions and that mutants for igsf11 exhibit defects in both the migration and survival of melanophores and their precursors . These findings identify the first in vivo requirements for igsf11 as well as the first instance of an immunoglobulin superfamily member functioning in pigment cell development and patterning . Our results provide new insights into adult pigment pattern morphogenesis and how cellular interactions mediate pattern formation . Pigment patterns are among the most striking of vertebrate traits and nowhere are these patterns more diverse than in teleost fishes [1]–[4] . In this group , a stunning array of pigment patterns function in predation avoidance , shoaling , and mate choice and are thought to have played important roles in speciation [5]–[8] . Among teleosts , the zebrafish Danio rerio has emerged as a useful model organism for uncovering the genetic and cellular bases of pigment pattern development . The zebrafish adult pigment pattern comprises a series of dark horizontal stripes that include black melanophores , alternating with lighter “interstripes” that include yellow–orange xanthophores; a third class of pigment cells , the iridescent iridophore occurs in both stripes and interstripes . Development of this pattern occurs during the larval-to-adult transformation between ∼2–4 weeks post-fertilization [9]–[12] . At this time , latent precursor cells of presumptive neural crest origin migrate from peripheral nerves and possibly other locations to the hypodermis , between the epidermis and the myotome , where differentiation occurs and the initially intermingled cells organize into stripes [10] , [13] . Mutational analyses have identified several loci that are required for the development of adult melanophores [9] , [14]–[17] , xanthophores [18] , and iridophores [19] , [20] , and these and other approaches have revealed important roles for cellular interactions , particularly between melanophores and xanthophores , in organizing the adult stripe pattern [21]–[24] . Remarkably , these interactions meet the predictions of Turing models of pattern formation that rely on dynamics driven by processes of reaction diffusion with lateral inhibition [25] , [26] . Nevertheless , the molecular mechanisms that drive cellular behaviors during stripe formation have remained obscure . Of particular interest for understanding the genetic mechanisms and cellular behaviors underlying stripe formation are mutants that retain all three classes of pigment cells while nevertheless developing abnormal adult pigment patterns . To date , two such mutants have been analyzed most extensively . The jaguar mutant exhibits fewer stripes than wild-type fish and , within these stripes , melanophores and xanthophores are intermingled [27]–[29] . The jaguar phenotype arises from mutations in kir7 . 1 , encoding an inwardly rectifying potassium channel , expressed and required by cells of the melanophore lineage [28] , [30] . By contrast , the leopard mutant [9] , [12] , [22] , [29] exhibits spots rather than stripes of melanophores , a defect arising from mutations in the gap junction gene , connexin41 . 8 , which is expressed by melanophores and xanthophores [31] . The presumed functions of both gene products raised the possibility that physiological ion fluxes contribute to pattern formation; indeed wild-type , but not jaguar ( kir7 . 1 ) mutant melanophores depolarize as a result of contacts with xanthophores in vitro [28] . Nevertheless , it has remained unclear to what extent genes classically known to regulate other morphogenetic processes are required specifically during pigment stripe formation . In this study , we analyze the seurat mutant phenotype , consisting of an irregular spotting pattern similar to that of the leopard mutant . We chose the seurat mutant because , unlike some adult pigment mutants [19] , [32] , [33] , defects are found in both body and fin pigment patterns , suggesting the affected locus may function normally in a core aspect of pattern formation . We show that seurat corresponds to immunoglobulin superfamily member 11 ( igsf11 ) , encoding a cell surface receptor containing two immunoglobulin-like domains . We find that igsf11 is expressed by the melanophore lineage , promotes the migration and survival of these cells during adult stripe development , and mediates adhesive interactions in vitro . Our results are the first demonstration of igsf11 functions in vivo , and , more generally , are the first to implicate a major family of “classical” cell adhesion molecule in adult pigment stripe formation . In turn , these findings set the stage for future investigations into how physiological and morphogenetic mechanisms affecting cell migration and survival interact to generate the adult pigment phenotype of zebrafish and other teleosts . We isolated the recessive , homozygous viable allele seuratutr15e1 from the inbred ABwp genetic background during a forward genetic screen for ENU-induced mutations affecting adult pigment pattern development . In comparison to the wild-type , seurat homozygotes develop fewer adult melanophores , which form irregular spots rather than stripes ( Figure 1A , 1B ) ; embryonic and early larval pigment patterns are indistinguishable between wild-type and seurat mutants ( not shown ) . We isolated two additional ENU-induced alleles , seuratwp15e2 and seuratwp15e3 , in the wik genetic background by non-complementation screening against seuratutr15e1 . These additional alleles were phenotypically indistinguishable from one another and exhibited less severe phenotypes than seuratutr15e1 ( Figure 1C; Figure S1 ) . Gross deficiencies in xanthophore or iridophore numbers were not apparent . For all phenotypic analyses below , we used the stronger allele , seuratutr15e1 ( hereafter seurat ) . To test if seurat acts autonomously to the melanophore lineage in promoting adult pigment stripe formation , we transplanted cells at the blastula stage from phenotypically wild-type Tg ( bactin:GFP ) embryos to homozygous seurat mutant embryos and reared the resulting chimeras until adult pigment patterns had formed . If seurat acts within the melanophore lineage , we anticipated that wild-type ( GFP+ ) melanophores would form patches more organized than the irregular spots formed by seurat mutant melanophores; regions of rescued pattern should include wild-type ( GFP+ ) melanophores but also might include seurat mutant ( GFP− ) melanophores , some of which develop where stripes would normally form ( Figure 1B and see below ) . Consistent with these predictions , we found that wild-type→seurat mutant chimeras in which wild-type melanophores developed exhibited large spots or rescued stripes , comprising both wild-type ( GFP+ ) melanophores as well as some seurat mutant ( GFP− ) melanophores ( Figure 2 ) . We did not observe these organized patches of melanophores in chimeras that failed to develop wild-type melanophores despite the presence of wild-type epidermis , iridophores , or nerves; we did not observe chimeras that developed donor xanthophores . To further assess the cell autonomy of seurat activities , we transplanted wild-type or seurat mutant cells to nacrew2 mutant embryos . nacre mutants fail to develop melanophores owing to a mutation in the mitfa transcription factor , which is required autonomously for specifying melanophore fate [15] . Any melanophores developing in these chimeras are thus donor-derived [16] . nacre mutants do , however , develop xanthophores and iridophores [27] . If seurat acts autonomously to the melanophore lineage , wild-type melanophores should form stripes in the nacre mutant background , whereas seurat mutant melanophores should fail to do so . Alternatively , if seurat effects on melanophore organization are non-autonomous , perhaps acting via xanthophores or another cell type , then both wild-type and seurat mutant melanophores should organize into stripes in the nacre mutant background [16] , [21] . Phenotypes of seurat mutant→nacre mutant chimeras support an autonomous role for seurat within the melanophore lineage , as donor , seurat mutant melanophores failed to organize into stripes and instead developed in dispersed patterns , as in the seurat mutant ( Figure 2C , 2D ) . Together these data support a model in which seurat acts within melanophores or their precursors to promote the organization of these cells into stripes . To identify the gene affected in seurat mutants , we mapped the mutant phenotype to a telomeric region of chromosome 15 between microsatellite markers Z10193 ( 45 . 6 Mb ) and Z8551 ( 46 . 5 Mb ) ( Figure 3A ) . Fine-mapping using single nucleotide polymorphisms ( S1 , S2 , S3 , S4 ) within this region revealed a critical genetic interval containing six complete or partial open readings frames . By sequencing exons and cDNAs of each locus and comparing resulting sequences to pre-mutagenized ABwp and wik genetic backgrounds , as well as single nucleotide polymorphisms in the Ensembl database , we identified novel , ENU-induced lesions in immunoglobulin superfamily member 11 ( igsf11; sc:d812 ) in each of the three seurat alleles ( GenBank accession number JQ796184 ) , and found no such lesions in the other candidate genes within this interval . Analyses of the inferred , 442 amino acid Igsf11 peptide sequence revealed a signal sequence , two immunoglobulin-like domains , a transmembrane domain , and a cytoplasmic domain ( Figure 3B ) . The zebrafish peptide sequence exhibited 64% identity and 77% similarity to human IGSF11 . In seuratutr15e1 a T→C transition leads to a substitution , S151P , located within the second immunoglobulin domain ( Figure 3B; Figure S2 ) . Mutations in the weaker alleles , seuratwp15e2 ( T29P ) and seuratwp15e3 ( V28E ) were found at the boundary between the predicted signal sequence and the beginning of the first immunoglobulin domain . These findings suggested that mutations in igsf11 cause the seurat mutant phenotypes . To further test the allelism of igsf11 and seurat , we asked if the seurat mutant phenotype could be rescued by expressing wild-type igsf11 cDNA within pigment cells or their precursors . To this end , we constructed transgenes to drive igsf11 with the mitfa promoter [34] , [35] , which is expressed by precursors to adult melanophores ( and possibly iridophores ) and newly differentiated melanophores during the larval-to-adult transformation , as well as xanthophores and undifferentiated cells that may be precursors to multiple pigment cell classes in the late larva and adult ( Figure S3 ) . We then injected seuratutr15e1 embryos with mitfa:igsf11 or mitfa:nlsVenus-V2a-igsf11 transgenes and reared fish through completion of the adult pigment pattern . These genetically mosaic fish expressed nuclear-localizing Venus within the melanophore lineage and exhibited partially rescued stripes ( Figure 4A–4C ) . After screening for germ line carriers , we additionally found that stable transgenic lines expressing mitfa:igsf11 exhibited stripes nearly indistinguishable from those of the wild-type ( Figure 4D–4G ) . These results and those of positional cloning analyses confirm that seurat corresponds to igsf11 . In conjunction with the results of cell transplantation analyses , these phenotypes also suggest that igsf11 promotes normal melanophore stripe development in part by acting through melanophores , or their undifferentiated and possibly multipotent precursors , though we do not exclude the possibility of contributory igsf11 functions within other lineages as well . The above analyses suggest that igsf11 should be expressed by adult melanophores and perhaps their precursors , though widespread expression in the early embryo [36] suggests the potential for expression more broadly as well . During the larval-to-adult transformation , in situ hybridization revealed igsf11 transcripts in relatively rare , scattered cells in the hypodermis , where stripe formation takes places between the skin and muscle ( Figure 5A , 5B ) , in extra-hypodermal locations where pigment cell precursors are found [34] , and in cells within the spinal cord ( Figure S4 ) . A polyclonal antiserum raised against a zebrafish Igsf11 peptide ( Figure S4 ) revealed an identical distribution of Igsf11-immunoreactive cells . To determine if scattered Igsf11+ cells might be pigment cell precursors , we examined Tg ( mitfa:GFP ) w47 fish [34] , [35] . These analyses revealed that many mitfa:GFP+ cells coexpressed Igsf11 ( Figure 5C , 5E ) , consistent with the autonomous activity of igsf11 within the pigment cell or melanophore lineages demonstrated by genetic mosaic analyses . Analyses at adult stages further revealed Igsf11 immunoreactivity of isolated melanophores ( Figure 5D ) and igsf11 transcripts expressed in isolated cells highly enriched for melanophores and xanthophores ( Figure 5F ) . We did not detect gross differences in levels of Igsf11 immunoreactivity between wild-type and seurat mutants , either in sections of larvae or in isolated melanophores , consistent with similar translational efficiency of the wild-type protein and S151P mutant protein ( data not shown ) . Finally , we also detected igsf11 transcript in several other adult tissues , including the eye , brain , heart , skin , fin , testis and ovary ( Figure 5G ) , presumably reflecting expression by other cells types , or pigment cells or their precursors resident in some of these tissues . Together , RT-PCR , in situ hybridization , and immunohistochemistry support the conclusion that Igsf11 is expressed in adult pigment cells and their precursors in post-embryonic zebrafish . Immunoglobulin superfamily members mediate a wide range of adhesive interactions . To test if zebrafish Igsf11 also might contribute to adhesive interactions , and the potential of seurat mutations to disrupt such interactions , we transfected K562 human myeloid leukemia cells with wild-type or seurat mutant igsf11 cDNAs . In rotary cultures , cells expressing wild-type Igsf11 adhered to one another to form large aggregates within two hours ( Figure 6; Figure S5 ) . By contrast , mock transfected cells or cells transfected with S151P ( seuratutr15e1 ) or T29P ( seuratwp15e2 ) mutant igsf11 cDNAs failed to form large aggregates . These findings support a model in which Igsf11 can mediate adhesive interactions in vivo and further demonstrate that both mutant forms of Igsf11 are compromised for such activity . Our finding that Igsf11 can mediate adhesive interactions in vitro , and the well-known roles of adhesive interactions in promoting cell migration and survival , led us to ask if either of these morphogenetic behaviors were compromised in seurat mutants . We repeatedly imaged homozygous seurat mutants and heterozygous wild-type siblings through the larval-to-adult transformation . These image series indicated that melanophores in seurat mutants tend to be more punctate than in the wild-type and exhibit reduced rates of migration and an increased likelihood of death as compared to wild-type melanophores ( Figure 7A; Videos S1 , S2 ) . seurat mutants also exhibited a progressively more severe deficiency in melanophore numbers as the larval–to–adult transformation progressed ( Figure S6 ) . To further assess a role for igsf11 in promoting melanophore migration we compared the motility of wild-type and igsf11 mutant melanophores in vitro . Similar to phenotypes in vivo , seurat mutant melanophores attached poorly to their substrate resulting in a more rounded appearance , and seurat mutant melanophores that did attach migrated significantly shorter distances than wild-type melanophores ( Figure 7B , 7C; Videos S3 , S4 ) . Finally , to determine if igsf11 is required for the migration and survival of melanophore precursors , in addition to differentiated melanophores , we crossed Tg ( mitfa:GFP ) w47 into the seurat mutant background and examined cell behaviors by ex vivo imaging [34] . As for differentiated melanophores , these analyses revealed significantly reduced migration and survival of mitfa:EGFP+ cells in seurat mutants as compared to the wild-type ( Figure 8 , Videos S5 , S6 , S7 , S8 ) . Together , these analyses demonstrate a requirement for igsf11 in promoting the migration and survival of melanophores and their precursors , supporting a model in which melanophore organization into stripes is mediated in part through Igsf11-dependent adhesive interactions . The results of this study identify critical roles for the immunoglobulin superfamily member Igsf11 in the development of zebrafish adult pigment stripes . We found that lesions in igsf11 are responsible for the seurat mutant phenotype of irregular melanophore spots . We demonstrated that Igsf11 promotes adhesive interactions of heterologous cells in vitro , and that seurat mutant forms of Igsf11 harboring missense mutations are defective for this activity . By cell transplantation and cell-type specific rescue experiments , we additionally found that igsf11 acts autonomously to pigment cell lineages in promoting melanophore stripe formation . Finally , our analyses of cellular behaviors in vivo , in vitro , and in ex vivo explants indicate that igsf11 promotes both the migration and survival of melanophores and their precursors . Whereas roles for cell adhesion molecules in the development of specific pigment patterns have long been suspected [37]–[41] , our study is the first to implicate a particular locus expressed by pigment cells in these processes . Our study expands the known developmental roles of immunoglobulin superfamily ( IgSF ) proteins , which include such well-studied members as N-CAM , DSCAM and ICAM-1 , and provides the first in vivo model system for dissecting the functions of Igsf11 specifically . The immunoglobulin superfamily is an especially diverse set of transmembrane proteins [42] , [43] , the functions of which have been analyzed most extensively in the nervous system , where they mediate axon guidance and fasciculation , target recognition , and dendrite patterning [44]–[47] , as well as in the immune system , where they are required for mediating interactions between immune cells and their environments , and for mounting immune responses [48]–[50] . IgSF members also play important roles in regeneration [51] and in cancer , acting as tumor suppressors or enhancers of invasion [52] , [53] . Although IgSF members are not known to be expressed abundantly by normal human melanocytes , several of these genes are dysregulated in melanoma and associated with melanoma progression and metastasis [54] , [55] and immunoreactivity using an anti-N-CAM antibody has been detected in xanthophores of some species [56] . IGSF11 was first identified in mouse and human and shown to be expressed highly in brain and testes ( for this reason being named originally Brain- and Testes-specific-IgSF , BT-IgSF ) [57] . IGSF11 was also identified independently as a gene up-regulated frequently in intestinal-type gastric cancers [58] . Our finding that igsf11 promotes melanophore morphogenesis represents the first identified function for an igsf11 orthologue in vivo as well as the first evidence of an IgSF member contributing to normal pigment cell development and patterning . Our finding that igsf11 is expressed and required by cells of the melanophore lineage and promotes adhesive interactions in vitro suggests two complementary models for the cellular bases of Igsf11-dependent interactions during adult pigment pattern formation . First , Igsf11 could mediate adhesive interactions specifically amongst differentiated melanophores as these cells organize into stripes . Second , Igsf11 could promote stripe development by mediating interactions between melanophores or their possibly multipotent precursors and their environments , either through homophilic or heterophilic adhesive interactions . Our analyses cannot yet speak to the first model , but results of the present study do support the second model of Igsf11-dependent interactions between melanophores and neighboring cell types . For example , we found that mitfa:GFP+ cells exhibited defects in migration and survival in seurat mutants , prior to melanization and stripe formation . Likewise , seurat mutant melanophores attached poorly to a collagen type IV substrate in serum-containing medium , and exhibited reduced motility independent of interactions with other melanophores . The biochemical mechanisms for Igsf11-dependent interactions remain unknown . Mammalian IGSF11 mediates homophilic adhesive interactions in vitro [59] and such interactions could occur in vivo during adult pigment pattern formation . Yet , our findings that Igsf11 acts autonomously to melanophores or their precursors in cell transplantation and genetic rescue experiments , and that igsf11 transcripts and protein are not detected in the environment through which these cells migrate , suggest that Igsf11 interacts with one or more heterophilic binding partners to promote melanophore lineage morphogenesis . Indeed , the coxsackie and adenovirus receptor , encoded by CXADR , which is the closest homologue of IGSF11 , mediates both homophilic and heterophilic adhesive interactions [60] . Consistent with the existence of additional Igsf11 ligands is our observation that different seurat mutant forms of Igsf11 equally abrogate cellular aggregation in vitro , despite having either severe or more mild pigment pattern defects in vivo; this outcome suggests that different mutant forms of Igsf11 may be differentially affected in their adhesive interactions with heterologous factors present in vivo but not in vitro . The existence of other Igsf11 interaction partners also seems likely from our demonstration that isolated wild-type and seurat mutant melanophores differed in their motility on Type IV collagen ( though we cannot exclude the possibility that IgSF11 could have been present in serum ) . Future studies aim to elucidate the mechanisms responsible for Igsf11-dependent adhesion , including the identification of cis- and trans-interacting proteins . In conclusion , our results identify a new gene required for adult pigment pattern formation and suggest an essential role for Igsf11-dependent adhesive interactions in promoting the morphogenesis of melanophores and their precursors during development of the adult form . It will be especially interesting to learn how pathways dependent on “classical” cell adhesion molecules of the sort identified here interact with physiological mechanisms mediated by kir7 . 1 and other factors [28] to orchestrate pigment pattern formation in zebrafish and other teleosts . All work in this study was conducted in accordance with guidelines and approved protocols for animal care and use at the University of Washington and Osaka University . seuratwp15e1 was isolated in a forward genetic , early pressure screen for N-ethyl-N-nitrosourea ( ENU ) induced mutations in the ABwp genetic background and was subsequently maintained in the same , unmutagenized background . Additional alleles , seuratwp15e2 and seuratwp15e2 , were isolated as ENU-induced mutations in the wik genetic background by screening against seuratwp15e1 , with subsequent backcrosses of non-complementing individuals to confirm allelism of new mutations . Chimeric embryos were generated by transplanting cells at blastula stages ( 3 . 3–3 . 8 hours post-fertilization ) and then rearing through late juvenile stages by which time an adult pattern has formed [11] . The Tg ( bactin:GFP ) transgenic line was provided by Ken Poss . seurat was mapped to chromosome 15 by bulked segregant analyses of fish derived from mapping crosses constructed using seurat ( ABwp genetic background ) and wik , then subsequently mapped between microsatellite markers Z10193 and Z8551 . Additional single nucleotide polymorphisms ( S1 , S2 , S3 , S4 ) were identified within this region of chromosome 15 ( 45 . 8∼46 . 1 Mb ) and were used to narrow the critical genetic interval in additional mapping crosses generated in Tubingen and AB genetic backgrounds . Differences in total numbers of individuals tested reflect background-specific polymorphisms and numbers of informative individuals analyzed . Gene predictions were derived from Ensembl ( Sanger Institute ) . cDNA sequences for all genes in the critical interval were compared to those of the un-mutagenized ABwp as well as other backgrounds . To test for lesions that might affect mRNA splicing , exons and flanking intronic sequences of these loci were examined from genomic DNA as well , though the presence of numerous repetitive elements in this telomeric region precluded complete sequencing of some splice junctions . Protein domains were predicted using Pfam , CLC Main Workbench 6 . 6 . 1 ( CLC bio , Muehital Germany ) and SignalP 4 . 0 [61] and by alignment and structural comparison with the closely related coxsackie and adenovirus receptor [60] , [62] . Structures were illustrated using Cn3D ( http://www . ncbi . nlm . nih . gov/Structure/CN3D/cn3d . shtml ) . Two different plasmid DNAs were generated for rescue experiments . For one construct , mitfa:igsf11 , the pT2AL200R150G vector [63] was modified by replacing the ef1a promoter with a 1 . 3 kb fragment of the mitfa promoter followed by the igsf11 coding sequence . The second construct , mitfa:nlsVenus-V2a-igsf11 , was generated using the Gateway Tol2kit pDestTol2pA2 vector [64] , and included a 2 . 2 kb fragment of the mitfa promoter followed by a composite open reading frame generated by overlap extension PCR [65] that consisted of a nuclear-localizing Venus fluorophore and the igsf11 coding sequence , linked by a V2a peptide breaking sequence to allow the production of separate peptide products [66] . Rescue constructs and Tol2 mRNA synthesized in vitro were injected into homozygous seuratutr15e1 embryos at the one-cell stage . Effects of transgenes were evaluated in these injected , mosaic fish , and in the non-mosaic F1 progeny of germ-line carriers for the mitfa:igsf11 transgene , in which genomic incorporation of the transgene was verified by PCR . Zebrafish adult tissues were harvested following euthanasia by methyl methane sulfonate ( MMS , Sigma ) overdose . Total RNAs were obtained using the RNeasy Protect Mini Kit ( Qiagen ) , and cDNA generated with SuperScript III CellsDirect cDNA Synthesis System ( Invitrogen ) . 4 . 4 ng of the cDNAs ( RNA equivalent ) obtained from each organ were used in PCRs to detect expression of igsf11 expression or β-actin as a positive control . PCR amplifications were performed for 32 cycles for igsf11 and 27 cycles for β-actin at 95°C for 30 s , at 60°C for 30 s , and 72°C for 30 s . To test for igsf11 expression in adult melanophores and xanthophores , fin pigment cells were isolated and cDNAs synthesized . Zebrafish were anesthetized with MMS and then fin regions containing melanophores and xanthophores , respectively , were dissected under a stereomicroscope . Fin clips were treated with solution containing 2 . 5 mg/ml trypsin liquid ( Worthington ) , 1 . 2 mg/ml BSA ( Sigma ) and 1 mM EDTA ( Wako ) in PBS for 10 min at 28°C . Trypsin solution was then removed and the tissue rinsed several times with PBS , after which samples were incubated for 60 min at 28°C with solution containing 1 mg/ml collagenase I ( Worthington ) , 0 . 1 mg/ml DNase I ( Worthington ) , 0 . 1 mg/ml STI ( Worthington ) , 1 . 2 mg/ml BSA , and 100 nM epinephrine ( Sigma ) in PBS . Suspension solutions were filtered with 25 µm mesh , followed by density-gradient centrifugation at 30× g for 15 min at room temperature in 50% Percoll ( Sigma ) , precipitating separate populations of melanophores and xanthophores to ∼95% purity from other contaminating cell types such as epidermis . To evaluate cross-contamination of melanophores and xanthophores , we tested for expression of dct and aox3 , which are specifically expressed by melanophores and xanthophores , respectively [18] , [67] . PCR amplifications were performed for 40 cycles for igsf11 , 34 cycles for dct and aox3 , and 38 cycles for β-actin at 95°C for 30 s , at 60°C for 30 s , and 72°C for 30 s . Primer sets were designed to span introns . For igsf11: 5′-TCTGATCGCGGGCACCATCG-3′ , 5′-TAGGTGTTGGTGGACGTCAGAGTG-3′; β-actin: 5′-CGGTTTTGCTGGAGATGATG-3′ , 5′-CGTGCTCAATGGGGTATTTG-3′; dct: 5′-ATCAGCCCGCGTTCACGGTT-3′ , 5′-ACACCGAGGTGTCCAGCTCTCC-3′; aox3: 5′-AGGGCATTGGAGAACCCCCAGT-3′ and 5′-ACACGTTGATGGCCCACGGT-3′ . Polyclonal antisera for zebrafish Igsf11 were generated in mouse . The peptide immunogen selected , PTYAWEKQESVPKLPHN , occurs within the second predicted immunoglobulin domain of Igsf11 . This antiserum did not recognize a specific fragment of the predicted size in Western blots , therefore its specificity was assessed by injecting embryos at the one-cell stage with a morpholino oligonucleotide ( Gene Tools , LLC ) targeting the igsf11 translational start site ( igsf11-MO: CATGTTTCCCAGCGAAAGTCGTCGT ) to test for reduced immunoreactivity at 24 hours post-fertilization . Morpholino was injected at 2 ng or 4 ng per embryo , as determined by an absence of toxicity at these doses using a 5 base pair mismatch control morpholino ( igsf11-MM: CATcTTTgCCAcCGAAAcTCcTCGT ) . Antiserum was used at 1∶500–1∶1000 following fixation in 4% paraformaldehyde and detected using goat anti-mouse Alexa 555 or Alexa 568 secondary antibodies . For immunohistochemistry of larvae , individuals of 7 . 0–9 . 0 standardized standard length ( SSL; [11] ) were fixed in 4% paraformaldehyde containing 1% DMSO in PBS , embedded in OCT , and sectioned by cryostat at 18–20 µm . In situ hybridization on vibratome sections followed [17] using a full length ( 1329 bp ) igsf11 cDNA for synthesis of antisense and sense riboprobes . The human myeloid leukemia cell line ( K562 ) was maintained with RPMI-1640 medium ( Sigma ) containing 10% FBS ( Invitrogen ) . pIRES2-igsf11 plasmid for nucleofection , a modified electroporation technique , was generated by cloning of full length Igsf11 fragment into pIRES2-AcGFP ( Clontech ) . Cells transiently expressing Igsf11 proteins ( wild-type or mutant ) were obtained using a 4D-Nucleofector ( Lonza ) with pIRES2-igsf11 plasmid . Twenty four hours after the nucleofection , cultures were suspended into single-cells by repeated pipetting , centrifuged and then resuspended in HBSS ( Gibco ) containing 1 mM CaCl2 at a density of 6×104 cells/ml . 500 µl of cell suspensions were transferred into a 24-well culture plate . The plate was then rotated on a gyratory shaker ( 80 rpm ) at 37°C for 2 h . Three independent experiments were performed and 5 random fields of view for cells in each treatment were imaged every 30 min . The degree of adhesion was evaluated as the ratio of the number of cell clusters over the total number of cells . Data were analyzed for effects of treatment and replicate by analyses of variance ( ANOVA ) in JMP 8 . 0 . 2 ( SAS Institute , Cary NC ) after arc sin transformation to control for heteroscedasticity of residuals that is common for ratio data [68] . Differences between specific treatments were assessed by Tukey Kramer post hoc comparisons . In additional experiments , transfected cells were split after centrifugation , with half of each sample used for aggregation assays and the other half used either for verifying the equivalence of transfection efficiencies of wild-type and mutant igsf11 constructs by fluorescence activated cell sorting ( BD FACS Calibur , BD Biosciences ) for GFP expression ( 5000 cells per sample ) . Similar levels of wild-type and mutant Igsf11 protein expression were also examined by immunocytochemistry . Larvae were viewed and imaged with an Olympus SZX-12 stereomicroscope and Axiocam HR camera . For time-course analyses , individual fish from a seurat/+ backcross were imaged daily and genotypes determined retrospectively . Fish were reared individually and imaged after brief anesthetization with MMS . Complete image series were obtained for 5 wild-type and 8 seurat mutant individuals . For determination of melanophore numbers , all melanophores were counted between the dorsal and ventral margins of the flank in a region bounded by the anterior margin of the dorsal fin and the posterior margin of the anal fin . Counts were obtained from individual larvae at selected standardized standard lengths during the larval-to-adult transformation with genotypes and binned sizes analyzed as fixed effects in analyses of variance . For depicting pattern development in animations shown in Videos S1 and S2 , all images were aligned and rescaled to control for growth using Adobe Photoshop CS5 . Melanophores were isolated from adult fish as described above and then re-suspended in L15 ( Sigma ) without FBS . Cells were cultured in 96-well culture dishes that had been coated with type IV collagen ( BD Biosciences ) . After one overnight incubation at 28°C , culture medium was changed with fresh L15 containing 5% FBS and the cells were imaged using an Olympus IX71 microscope equipped with an Olympus DP72 digital camera and Lumina Vision software ( Mitani Corporation ) . Melanophores that attached , survived at least 48 h , and did not interact with other melanophores were chosen for analysis . Melanophore centroid positions were obtained in ImageJ ( http://rsb . info . nih . gov/ij/ ) and plotted every hour between 12–36 h after medium change . Rectilinear migration distance was defined as the length between the beginning and ending positions for each cell . To image the morphogenetic behaviors of presumptive melanophore precursors ex vivo , 7 . 0 SSL larvae were rinsed with 10% Hanks medium and then anesthetized and then decapitated with a razor blade . Larval trunks were then placed on 0 . 4 µM transwell membranes ( Milipore ) in glass bottom dishes containing L15 medium , 3% fetal bovine serum , and penicillin/streptomycin . The trunks were equilibrated for 3 h at 28 . 5°C then imaged at 30 min intervals for 18–26 h on a Zeiss Observer inverted epifluorescence microscope with an Axiocam MRm camera . Z-stacks of 10–15 planes collected at 4 µm intervals were merged for final analyses .	Vertebrate pigment patterns are stunningly diverse and have been an important model of pattern formation for more than a century . Nevertheless , we still know remarkably little about the genes and cell behaviors that underlie the generation of specific patterns . To elucidate such mechanisms , a large number of pigment pattern mutants have been isolated in the genetically tractable zebrafish . Instead of the normal horizontal stripe pattern , many of these mutants exhibit spots of varying sizes and degrees of organization . Here , we show that one such mutant , seurat , named for the 19th century pointillist , George Seurat , exhibits lesions in the gene encoding a classical cell adhesion molecule ( CAM ) of the immunoglobulin superfamily , Igsf11 . We find that Igsf11 mediates cell adhesion and promotes the migration and survival of melanophores and their precursors during adult stripe formation . These results are exciting because they are the first time that a CAM has been implicated in pigment pattern formation , despite the long-standing expectation that such molecules might be required to regulate adhesive interactions during these events . These cellular phenotypes further represent the first known in vivo functions for Igsf11 and point to the potential for similar activities amongst the rich diversity of immunoglobulin superfamily members .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "developmental", "biology", "model", "organisms", "genetics", "biology", "genetics", "and", "genomics" ]	2012	Melanophore Migration and Survival during Zebrafish Adult Pigment Stripe Development Require the Immunoglobulin Superfamily Adhesion Molecule Igsf11
A model of the gene-regulatory-network ( GRN ) , governing growth , survival , and differentiation of melanocytes , has emerged from studies of mouse coat color mutants and melanoma cell lines . In this model , Transcription Factor Activator Protein 2 alpha ( TFAP2A ) contributes to melanocyte development by activating expression of the gene encoding the receptor tyrosine kinase Kit . Next , ligand-bound Kit stimulates a pathway activating transcription factor Microphthalmia ( Mitf ) , which promotes differentiation and survival of melanocytes by activating expression of Tyrosinase family members , Bcl2 , and other genes . The model predicts that in both Tfap2a and Kit null mutants there will be a phenotype of reduced melanocytes and that , because Tfap2a acts upstream of Kit , this phenotype will be more severe , or at least as severe as , in Tfap2a null mutants in comparison to Kit null mutants . Unexpectedly , this is not the case in zebrafish or mouse . Because many Tfap2 family members have identical DNA–binding specificity , we reasoned that another Tfap2 family member may work redundantly with Tfap2a in promoting Kit expression . We report that tfap2e is expressed in melanoblasts and melanophores in zebrafish embryos and that its orthologue , TFAP2E , is expressed in human melanocytes . We provide evidence that Tfap2e functions redundantly with Tfap2a to maintain kita expression in zebrafish embryonic melanophores . Further , we show that , in contrast to in kita mutants where embryonic melanophores appear to differentiate normally , in tfap2a/e doubly-deficient embryonic melanophores are small and under-melanized , although they retain expression of mitfa . Interestingly , forcing expression of mitfa in tfap2a/e doubly-deficient embryos partially restores melanophore differentiation . These findings reveal that Tfap2 activity , mediated redundantly by Tfap2a and Tfap2e , promotes melanophore differentiation in parallel with Mitf by an effector other than Kit . This work illustrates how analysis of single-gene mutants may fail to identify steps in a GRN that are affected by the redundant activity of related proteins . An important participant in the gene-regulatory-network ( GRN ) that governs the differentiation of melanocytes from neural crest precursors ( i . e . , the melanocyte GRN ) is the class III receptor tyrosine kinase Kit . In mouse embryos , binding of this growth-factor receptor by its ligand , stem cell factor ( SCF ) , promotes the growth , survival , migration , and possibly terminal differentiation of melanocytes [1] . Mouse embryos homozygous for hypomorphic alleles of Kit completely lack melanocytes ( embryos homozygous for Kit null alleles die prior to pigmentation ) [2]–[6] . While ligand-bound Kit stimulates many signal transduction pathways , its effects on melanocyte growth and differentiation appear to occur via the Ras/Raf/Map Kinase pathway . Activity of this pathway results in phosphorylation of Microphthalmia transcription factor ( Mitf ) ; phosphorylation of Mitf regulates its activity and stability [7] , [8] . Within melanoblasts , Mitf promotes a ) cell-cycle exit , by activating expression of the p21WAF1 , a cyclin-dependent kinase inhibitor [9] , b ) cell survival , by upregulating the expression of BCL2 [10] , and c ) melanin synthesis , by activating expression of Tyrosinase ( Tyr ) , Tyrosinase-related protein 1 ( Tyrp1 ) , and Tyrosinase-related protein 2 ( Tyrp2 , also known as Dopachrome tautomerase , Dct ) [11]–[14] . Thus , Kit signaling is essential for normal melanocyte development , at least in part via its ability to stimulate Mitf activity . Of note , KIT levels are reported to be lower in metastatic melanoma cell lines than in benign nevi , and forced expression of KIT in these cells has been shown to induce apoptosis [15] . These findings highlight the importance of understanding the regulation of Kit expression within the melanocyte lineage . While there is evidence that the KIT gene is dependent on direct stimulation by the Transcription Factor Activator Protein 2 alpha ( TFAP2A ) in melanoma , analyses of mutant model organisms indicate a more complex regulatory scenario within embryonic melanocytes . TFAP2A and other members of the TFAP2 family control cell fate specification , cell differentiation , cell survival and cell proliferation within neural crest , skin , breast epithelium , and other embryonic cell types and stem cells [16] , [17] . Gel shift experiments showed that TFAP2A can bind an element 1 . 2 kb upstream of the KIT transcription start site , and expression driven by this enhancer in melanoma cells is lost when the TFAP2 binding sites are deleted [18] . Moreover , forced expression of the TFAP2A DNA binding domain , which presumably unseats endogenous TFAP2A and thus acts as a dominant negative AP2 , prevents expression of KIT in these cells [18] . Mice lacking the Tfap2a gene do not live long enough to develop melanocytes , due to failure of body wall closure [19] , [20] . However , in embryos with Wnt1-CRE-mediated deletion of Tfap2a specifically within the neural crest , melanocytes are absent from the belly [21] . Interestingly , this phenotype resembles that of heterozygous , not homozygous , Kit loss-of-function mutants , suggesting that loss of Tfap2a leads to a reduction rather than complete loss of Kit expression . Zebrafish have two orthologues of mammalian Kit , known as kita and kitb; only kita is expressed in the melanophore lineage [22] . In kita homozygous null mutants ( i . e . , kita mutants ) relative to their wild-type counterparts , embryonic melanophores are reduced in number by about 40% , migrate less , and eventually undergo apoptosis [23] . In zebrafish tfap2a homozygous null mutants ( i . e . , tfap2a mutants ) , kita expression is reduced and embryonic melanophores exhibit reduced migration [24] , [25] . However , in contrast to the melanophores in kita mutants , those in tfap2a mutants do not appear to die , at least as long these animals survive [23] , [26] . The simplest explanation for this difference is that kita expression in melanophores is initially dependent on tfap2a but later becomes independent of it . How can the dominant negative AP2 block Kit expression while loss of Tfap2a only diminishes or delays it ? Because many Tfap2 family members have the same DNA binding affinity , it is possible that another such family member cooperates with Tfap2a to activate Kit expression . Here we show that Tfap2e , a homolog of Tfap2a with the equivalent DNA binding specificity , is expressed in zebrafish melanoblasts and in cultures of primary human melanocytes . With single and double knockdown studies , we show that while Tfap2e is not required for the development of embryonic melanophores , it functions redundantly with Tfap2a in maintaining kita expression in embryonic melanophores . Interestingly , in contrast to the situation in kita mutants , the melanophores in embryos doubly deficient for tfap2a/e fail to differentiate . These results imply that Tfap2 activity has targets other than kita that are important for melanophore development . We find that forced expression of mitfa partially restores melanophores in embryos lacking tfap2a and tfap2e , implying that the targets of Tfap2a/e function to stimulate Mitfa activity or act in parallel with it . These findings reveal unexpected roles for Tfap2 activity in the melanocyte GRN . To determine if a second Tfap2 family member is expressed in the melanoblast lineage , we identified orthologues of Tfap2b , Tfap2c , Tfap2d , and Tfap2e in a database of expressed sequence tags ( www . ensembl . org ) , amplified partial clones of at least 1 kb from each to make a probe for in situ hybridization , and examined the expression of each in embryos that ranged in stage from 0 . 5 hours post fertilization ( hpf ) , revealing maternal expression , to 48 hpf . Expression patterns of tfap2b and tfap2c have previously been reported [27] , [28] . We did not detect expression of tfap2b , tfap2c , or tfap2d in melanoblasts or melanophores ( Figure S1 ) , so we did not pursue these orthologues in the context of melanophore development . In 8-cell zebrafish embryos , maternal tfap2e transcripts were detected by both in situ hybridization and semi-quantitative RT-PCR ( not shown ) . At 24 hpf , tfap2e expression was detected in several regions of the brain , including presumed olfactory bulb , as in mouse embryos [29] , [30] ( Figure 1A ) , and also within dispersed cells in the trunk that we assumed to be a subset of migrating neural crest cells ( Figure 1B and 1D ) . At this stage , tfap2e expression was detectable in early-differentiating melanophores close to the ear ( Figure 1C ) , suggesting that the dispersed , non-melanized cells expressing tfap2e were melanoblasts . To test this possibility , we probed homozygous mitfa null mutant embryos ( i . e . , mitfab692 ) , which are devoid of melanoblasts [31] , and found that tfap2e expression was absent from the dispersed cells in the trunk ( Figure 1E-1G ) . This result was consistent with expression of tfap2e in melanoblasts . However , because mitfa is co-expressed with xdh and fms , two markers of xanthophore precursors [32] , it was conceivable that tfap2e was expressed in the xanthophore lineage , in an Mitfa-dependent fashion . To test whether tfap2e is expressed in xanthophores , we processed embryos to simultaneously reveal expression of tfap2e mRNA and Pax7 protein , a marker of the xanthophore lineage [33] . We did not detect overlap of the two signals , which implies that tfap2e is not expressed in xanthophores ( Figure 1H ) . In wild-type embryos at 36 hpf , tfap2e expression was present in the forebrain and presumed optic tectum , and expanded in the hindbrain relative to earlier stages ( Figure 1I and 1J ) . However , at this stage expression was not detected in melanophores ( Figure 1K ) . At 48 hpf , high-level tfap2e expression was also observed in the retina ( Figure 1L ) . To assess if melanocyte-specific expression of TFAP2E is conserved in humans , we performed quantitative RT-PCR on cDNA generated from various human cell lines . We detected higher levels of TFAP2E message in three independent isolates of primary melanocytes , consistent with microarray data indicating expression of TFAP2E in melanocytes and melanoma cell lines [34] . Expression in melanocytes was 2–10 fold higher than in a keratinocyte cell line , and approximately 50–100 fold higher than in a lymphocyte cell line ( Figure 1M ) . In summary , tfap2e is expressed in zebrafish melanoblasts and in human melanocytes . As discussed in the Introduction , KITA has been reported to be a direct target of TFAP2A , and a dominant negative AP2 variant was found to block KIT expression in cultured cells [18]; however the status of kita expression in tfap2a mutants has not been fully investigated . In zebrafish tfap2a mutants or tfap2a MO-injected embryos at 28 hpf , kita expression in the melanophore lineage is reduced to undetectable levels as assessed by in situ hybridization [24] , [25] . However because melanophores undergo cell death in kita mutants but do not do so in tfap2a mutants , it has been proposed that kita is expressed in the melanophore lineage of tfap2a mutants at a later stage [24] . To test this prediction , we crossed heterozygote tfap2a null mutants ( i . e . , lockjaw , tfap2ats213 ) and identified homozygous mutant offspring ( hereafter , tfap2a mutants ) at 28 hpf by virtue of their pigmentation phenotype . We fixed a fraction of these embryos at 28 hpf , and incubated the remainder in water containing phenylthiourea ( PTU ) to prevent melanin synthesis , until 36 hpf . We then processed all embryos to reveal kita expression . In tfap2a mutants at 28 hpf , kita expression in melanophores was undetectable by in situ hybridization ( Figure 2C ) , as previously reported . However , at 36 hpf , kita expression was clearly visible in cells present in the dorsum of these embryos ( Figure 2E ) . Thus normal kita expression in melanoblasts at 28 hpf is dependent on tfap2a , but later becomes independent of it . To explain these observations we hypothesized that Tfap2e compensates for the loss of Tfap2a and activates kita expression by 36 hpf . To test whether Tfap2e maintains kita expression in tfap2a mutants , we first assessed tfap2e expression in tfap2a mutants , and found that it was expressed on schedule in migrating neural crest , as in wild-type embryos ( Figure S2 ) . Next we injected embryos with a morpholino ( MO ) targeting the tfap2e exon 3 splice donor site ( i . e . , tfap2e e3i3 MO ) ( Figure 2A ) . To confirm the efficacy of this MO towards its intended target , we harvested RNA from embryos injected with the tfap2e e3i3 MO , generated first-strand cDNA , and performed PCR using primers in exon 1 and exon 4 . Sequencing of the major aberrant splice product revealed that the e3i3 MO causes deletion of exon 3 in its entirety , resulting in a frame shift and a severe truncation of the predicted protein that eliminates the DNA binding domain ( Figure 2A ) . By semi-quantitative PCR , this MO appears to inhibit normal splicing of the majority of tfap2e transcripts at 36 hpf , but to act with greatly reduced efficiency at 3 days post fertilization ( dpf ) ( Figure 2A ) . By 24 hpf , wild-type zebrafish embryos injected with tfap2e e3i3 MO showed evidence of cell death in the central nervous system ( CNS ) , i . e . , patches of opacity in the brain and spinal cord , but no other gross morphological defects; possibly this was due to non-specific toxicity of the MO to the embryo . Despite this cell death , the melanophores that developed in such embryos looked normal and were normally distributed ( see below and Figure S3 ) . tfap2e e3i3 MO-induced CNS cell death was reduced by co-injection of p53 MO , implying that Tfap2e has a role in cell survival in the CNS , or that the tfap2e e3i3 MO has non-specific toxicity towards the nervous system , which is true of many MOs ( Figure S3 ) [35] . To preserve the morphology of embryos , in all experiments discussed hereafter we have included p53 MO with tfap2e e3i3 MO . Interestingly , in tfap2a mutants injected with the tfap2e e3i3 MO ( hereafter , tfap2a/e doubly-deficient embryos ) , kita was absent from the dorsum at 36 hpf , although kita expression was readily detected in the cloaca and pharyngeal pouches ( Figure 2G and not shown ) . These findings imply that in absence of Tfap2a , Tfap2e promotes kita expression in the melanophore lineage . Because of the sustained loss of kita expression in tfap2a/e doubly-deficient embryos , we expected that the phenotype in these embryos would be similar to that of kita homozygous null mutants , although perhaps not as severe because MO-mediated inhibition of gene expression is transient and partial; instead , however , we detected a much more severe phenotype . At 36 hpf , compared to the embryonic melanophores in their non-mutant siblings ( Figure 3A ) , those in kita null mutants ( i . e . , kitab5 ) ( Figure 3B ) appeared normally melanized , but were reduced to about 60% of their normal numbers ( because of a presumed defect in cell division ) and did not migrate as extensively as their wild-type counterparts [23] , [36] . In control MO-injected tfap2a mutants ( Figure 3C ) , embryonic melanophores exhibited these same phenotypes . In tfap2e MO-injected sibling embryos ( Figure 3D ) there was no apparent melanophore phenotype . However , in tfap2a/e doubly-deficient embryos there were far fewer melanophores than present in control MO-injected tfap2a mutant embryos . Compared with control MO-injected tfap2a mutants , tfap2a/e doubly-deficient embryos had fewer pigmented melanophores in the dorsum and almost no visible melanophores on the lateral sides of the trunk or on the yolk sac ( Figure 3E ) ; this difference was still apparent at 84 hpf ( not shown ) . In summary , whereas wild-type embryos injected with the tfap2e MO developed normally until at least 4 dpf , tfap2a/e doubly-deficient embryos displayed melanophore defects more severe than those of tfap2a or kita mutants . These findings suggest that Tfap2a and Tfap2e have partially redundant function in zebrafish melanophore development , and that this function exceeds the simple maintenance of kita expression . To confirm the specificity of the tfap2e e3i3 MO-induced melanophore phenotypes , we co-injected mRNA encoding a glucocorticoid-fused version of Tfpa2a ( tfap2aGR ) , whose nuclear transport is dexamethasone-inducible , or lacZ as a control , into embryos injected with MOs targeting tfap2a , tfap2e , and p53 ( hereafter also termed tfap2a/e doubly-deficient embryos ) . Dexamethasone was added to both groups at 70% epiboly to avoid gastrulation defects caused by tfap2a over-expression [28] . Embryos were then scored for the rescue of under-melanized melanophores , seen in tfap2a/e doubly-deficient embryos , at 36 hpf . We found that tfap2aGR mRNA effectively rescued melanophores in tfap2a/e doubly-deficient embryos , whereas lacZ did not ( Figure S4G and S4H ) . As an alternative approach for testing specificity , we purchased two additional independent tfap2e MOs—one targeting the exon 2 splice donor site ( i . e . , e2i2 MO ) and the other the translation start site of the tfap2e gene ( i . e . , AUG MO ) ( Figure 2A ) . Injection of either the tfap2e e2i2 MO or the tfap2e AUG MO into wild-type embryos had no effect on melanophore development , although both induced some degree of nervous-system cell death . Upon injection of either the tfap2e e2i2 MO or tfap2e AUG MO into embryos derived from tfap2a mutant heterozygous parents , about one fourth of embryos exhibited the melanophore phenotype seen with the tfap2e e3i3 MO ( Figure S4A-S4F ) ; co-injection of p53 MO did not alter the melanophore phenotypes although it reduced nervous system cell death ( not shown ) . These multiple tests of specificity strongly argue that the melanophore phenotypes we observe in tfap2e MO-injected embryos result from inhibition of tfap2e expression and not from off target effects . The reduced number of melanophores in tfap2a/e doubly-deficient embryos relative to tfap2a mutants could reflect a role for Tfap2a/e activity in the specification of melanoblasts or , alternatively , in either survival or differentiation of melanophores . To distinguish among these possibilities , we examined the expression of mitfa , an early marker of the melanoblast and xanthoblast lineages [31] , [32] . At 29 hpf , mitfa-expressing cells are visible in the head and trunk of wild-type embryos injected with a control MO ( Figure 4A ) . The number of mitfa-expressing cells is reduced by about half in tfap2a mutant embryos injected with a control MO ( Figure 4B ) ; this reduction results at least in part from the absence of kita in such mutants at this stage , because melanophores are reduced by this amount in kita mutants [23] , as are mitfa-expressing cells ( our unpublished observations ) . In tfap2e MO-injected , wild-type embryos , the number of mitfa cells is not grossly different from that in control MO-injected , wild-type embryos ( Figure 4C ) . Interestingly , in tfap2a/e doubly-deficient embryos , the number of mitfa-expressing cells did not appear to be further decreased relative to that in control MO-injected tfap2a mutants ( Figure 4D ) . To confirm these impressions , we counted mitfa-expressing cells over the hind yolk ( see Materials and Methods ) at 24 hpf , and compared the results for tfap2a mutants injected with control MO versus those injected with tfap2e MO; we found no significant difference ( See Figure 4 legend for numbers ) . In addition , we used fluorescence-activated cell sorting ( FACS ) to count GFP-positive cells in dissociated mitfa:egfp transgenic embryos injected with MOs , and this analysis supported our findings from histology [37] . Thus , GFP-expressing cells were similarly reduced in tfap2a MO-injected and tfap2a/e doubly-deficient mitfa:egfp embryos ( i . e . , to about 40% of the number in controls ) , although the number of differentiated melanophores in tfap2a/e doubly-deficient embryos was clearly reduced relative to that in tfap2a MO injected embryos ( Figure 4E , histogram ) . These findings imply that Tfap2 activity , provided by the redundant actions of Tfap2a and Tfap2e , is involved in a step of melanophore development that occurs subsequent to specification of the mitfa-positive lineage . To determine which step in melanophore development depends on Tfap2 activity , we analyzed the expression of genes involved in melanophore differentiation: tyr , tyrp1b and dct [12] . In tfap2a mutant embryos at 29 hpf , the number of cells expressing each of these melanophore markers was reduced by about half relative to that in siblings , consistent with the previously described decrease in melanophores in tfap2a mutants ( Figure 5A , 5E , 5I and 5C , 5G , 5K ) [24] , [25] . In tfap2e MO-injected embryos , the number of cells expressing each of these genes appeared to be normal ( Figure 5B , 5F , and 5J ) , while in tfap2a/e doubly-deficient embryos their numbers were further reduced relative to that in tfap2a mutant embryos ( Figure 5D , 5H , and 5L ) . To quantify this effect , we counted cells in embryos processed for in situ hybridization . We discovered that the reduction in gene expression was not equal in all cases . The number of cells expressing dct was most clearly and most consistently reduced in tfap2a/e doubly-deficient embryos , i . e . , by approximately 47% relative to the number in tfap2a mutant embryos ( Figure 5A-5D , and 5M ) . The reduction in tyrp1b and tyr expressing cells was more variable , with an average reduction of approximately 30% and 23% , respectively ( Figure 5E-5L , and 5M ) . The results described above indicate that when the expression of tfap2a and tfap2e is reduced , melanoblasts express mitfa but fail to progress to a stage at which they express normal levels of melanophore differentiation genes , such as dct , tyrp1b , and tyr . To test this model more quantitatively , we injected mitfa:egfp transgenic embryos [37] with either tfap2a MO or both tfap2a MO and tfap2e MO , dissociated them at 29 hpf , sorted and collected GFP-expressing cells , and measured the levels of various transcripts by quantitative RT-PCR ( Figure 5N ) . Using this method , we saw a trend similar to that observed in the histology analysis: in GFP-positive cells sorted from tfap2a/e MO-injected embryos relative to those sorted from tfap2a MO-injected embryos , dct expression was reduced by approximately 45% , tyrp1b expression was reduced by 17% , and unexpectedly , tyr expression was not reduced . Taken together with the cell counts , these results reveal that Tfap2 activity , redundantly provided by Tfap2a and Tfap2e , promotes the differentiation of embryonic melanophores . We tested the possibility that the loss of differentiated melanophores in tfap2a/e doubly-deficient embryos results from a fate switch of melanophores to xanthophores , because mitfa is co-expressed with c-fms , a marker of xanthophore precursors [32] . We injected embryos with a control MO , tfap2a MO , tfap2e MO , or tfap2a/e MOs , and at 36 hpf processed them to reveal expression of anti Pax7 IR , a marker of xanthophores [33] ( Figure 6A-6C and not shown ) . While the numbers of xanthophores in these groups did not differ significantly ( Figure 6D ) , melanophore differentiation was clearly affected in tfap2a/e doubly-deficient embryos . These findings suggest that loss of Tfap2 activity in the melanophore lineage does not result in a cell fate switch . We also assessed whether melanophores in tfap2a/e doubly-deficient embryos undergo cell death , i . e . despite the presence of p53 MO . First , we co-injected embryos with MOs targeting tfap2a and tfap2e and with an mRNA encoding Bcl2 , an inhibitor of apoptosis [38] . Injection of bcl2 mRNA reduced the number of cells expressing a marker of programmed cell death in control embryos at 25 hpf ( Figure 6E and 6F ) , but had no effect on the melanophore phenotype in tfap2a/e doubly-deficient embryos ( Figure 6G and 6H ) . Secondly , embryos were incubated in acridine orange ( AO ) , which is taken up by dying cells , from 16 hpf to 30 hpf and assessed for the presence of AO-containing cells in the dorsal neural tube and migratory neural crest . Relative to control MO-injected wild-type embryos , control MO-injected tfap2a mutants had an elevated number of such cells , but these numbers were not detectably increased in tfap2e MO-injected tfap2a mutants ( data not shown ) . These findings suggest that loss of Tfap2 activity in melanophores does not result in either a switch in cell fate specification or promotion of cell death , but more likely in inhibition of normal melanophore differentiation . In tfap2a mutants and MO-injected embryos , embryonic melanophores initially appear somewhat under-melanized [24] , [25] . The tfap2a gene is expressed both in skin and neural crest , and we have reported evidence based on transplant studies that Tfap2a has both cell-autonomous and cell non-autonomous effects on melanophore differentiation [25] . Because tfap2e is expressed in melanoblasts but not skin , we assumed that the even poorer differentiation of melanophores in tfap2a/e doubly-deficient embryos is primarily a consequence of a cell autonomous role for Tfap2 activity . To confirm this prediction , we created genetic chimeras by carrying out transplantations at the blastula stage . Specifically , we transplanted cells from 4 hpf wild-type donors , which had been injected with a biotin-dextran as a lineage tracer , into 4 hpf hosts injected with tfap2a/e MO . We then reared the transplanted hosts to 48 hpf , and processed them for biotin staining to reveal the donor-derived cells . Melanophores lacking lineage tracer were indistinguishable from those seen in the untransplanted tfap2a/e MO-injected controls ( Figure 7C-7F , arrows ) , whereas those positive for the lineage tracer were clearly darker , similar to wild-type controls ( Figure 7A and 7B ) , indicating an increase in the level of melanin . In addition , they displayed a more normal morphology ( Figure 7E and 7F , arrowheads ) . These findings indicate that normal melanophores can develop from wild-type cells that are flanked by tfap2a/e-deficient epidermis . This supports a cell-autonomous requirement for Tfap2a/e activity in melanophore differentiation . Several signals are known to modulate Mitf transactivation activity [39] , [40] . If Tfap2a/e is required for the expression of a component of such a signaling pathway , Mitfa activity might be reduced in tfap2a/e doubly-deficient embryos despite levels of mitfa mRNA being similar to those in tfap2a mutants . Alternatively , the Tfap2a/e effector required for melanocyte differentiation might be co-activated by Mitf . In either of these scenarios , forced mitfa expression might rescue melanophore differentiation in tfap2a/e doubly-deficient embryos . We injected tfap2a/e doubly-deficient embryos with a plasmid in which the sox10 promoter drives mitfa expression ( sox10:mitfa ) [41] , and found sox10:mitfa-injection increased the number of tfap2a/e doubly-deficient embryos with differentiated melanophores ( compare Figure 8B to Figure 8C , 8D ) . We observed an increase in the number of darkly-pigmented melanophores in tfap2a/e doubly-deficient embryos injected with sox10:mitfa compared to in tfap2a/e doubly-deficient embryos alone ( Figure 8E ) . We also quantified the mean gray value of single melanophores in these embryos ( as a measure of pigment density ) , within a defined region , using ImageJ software . We found that there was a significant reduction in the pigment density of tfap2a/e doubly-deficient embryo melanophores , compared to control MO-injected embryo melanophores , and that this density was restored in doubly-deficient embryos co-injected with sox10:mitfa ( Figure 8F ) . Since sox10 is expressed throughout the neural crest , we considered the possibility that sox10:mitfa might induce a conversion of neural crest to the melanoblast lineage , and that if this were to occur in neural crest that expressed another Tfap2 family member , normally differentiated melanophores might emerge in tfap2a/e doubly deficient embryos . However , arguing against this alternative model , we did not detect an increase in the number of melanophores in control-MO injected embryos co-injected with the sox10:mitfa plasmid ( Figure 8E ) . Moreover , in this alternative model , tfap2b is the best candidate Tfap2 family member , as it is expressed in Rohon Beard sensory neurons [27] , which are closely related to trunk neural crest [42] , [43] . However , we found that even in embryos triply depleted of tfap2a/b/e using MOs , co-injection of sox10:mitfa plasmid elevated the number of normal-looking melanophores ( our unpublished observation ) . Together these observations support the model that over-expression of mitfa can compensate for the role in melanophore differentiation normally played by Tfap2a/e , implying that the effector of Tfap2a/e-type activity necessary for melanophore differentiation acts upstream or in parallel with Mitfa . Here we have presented two new findings relevant to the gene-regulatory-network ( GRN ) that governs the differentiation of zebrafish embryonic melanophores . First , kita expression in embryonic melanophores is positively regulated by Tfap2e , at least when Tfap2a levels have been reduced . Expression of tfap2a is present throughout the neural crest starting at the neurula stage , while the expression of tfap2e starts at approximately the time of neural crest delamination and appears to be restricted to melanoblasts [24] , [25] . The relative timing of tfap2a and tfap2e expression explains why kita expression ( in melanophores ) in tfap2a mutants is reduced at 28 hpf , but present at later stages; Tfap2e compensates for the absence of Tfap2a but only after 28 hpf . The presence of TFAP2E expression in human melanocytes suggests that TFAP2A and TFAP2E have redundant or partially redundant function in mammalian melanocytes , as in fish melanophores . If so it would explain the observation , mentioned in the Introduction , that the coat color phenotype in mice with neural crest-specific deletion of Tfap2a is less severe than that of Kit homozygous null mutants [21] . The second unexpected finding is that Tfap2 activity ( provided by Tfap2a and Tfap2e ) promotes the differentiation of embryonic melanophores . This was revealed by reduced expression of the dct and tyrp1b mRNAs , as well as of melanin—changes that are evident in tfap2a mutants and more pronounced in tfap2a/e doubly-deficient embryos . Does Tfap2 activity also direct neural crest cells to join the melanophore sublineage ? There is precedent for such a possibility , because Tfap2 activity provided by Tfap2a and Tfap2c appears to direct ectodermal precursors to join the neural crest lineage [28] , [44] . In tfap2a single mutants , neural crest induction appears to occur normally , but mitfa-expressing cells , which are primarily melanoblasts , are reduced in number . This reduction may reflect a role for Tfap2 in melanophore specification or alternatively a reduction of Kita-mediated proliferation of melanoblasts . Whatever the explanation for reduced melanoblasts in tfap2a mutants , simultaneous reduction of tfap2a and tfap2e leads to a further reduction of melanophore numbers without a further reduction of mitfa-expressing cells , arguing Tfap2 promotes differentiation of melanoblasts to melanophores . While a reduction of melanophores without a reduction in mitfa-expressing cells might have been consistent with a cell fate change of melanophores to xanthophores ( because markers of melanoblasts and xanthoblasts are briefly co-expressed [32] ) , xanthophore numbers are equivalent in tfap2a deficient and tfap2a/e doubly-deficient embryos , arguing against such a fate transformation . Does Tfap2 also promote survival of melanophores ? We did not detect evidence of cell death of melanophores shortly after their differentiation in tfap2a/e doubly-deficient embryos . We predict that in embryos permanently deprived of both Tfap2a and Tfap2e melanophores would die as a consequence of the absence of Kita . However , because melanophores persist for several days in kita mutants , and this is longer than MOs are effective ( see Figure 2A ) , it will be necessary to isolate a tfap2e mutant to test this prediction . Together these observations reveal that Tfap2 activity has multiple roles in melanophore development , including promoting melanophore differentiation . Another result that will be important to revisit when a tfap2e mutant is available is the apparent heightened Tfap2-dependence of dct expression relative to tyr expression . Consistent with differential regulation of these related genes , in mice , Dct expression appears prior to Tyr expression , and this has also been suggested to be the case in zebrafish [45] , [46] . However , because we knock-down tfap2e expression with an MO , the stronger effect on dct expression relative to on tyr expression may simply reflect loss of MO effectiveness over time . There may be a similar explanation for the inconsistent findings regarding tyr expression between the RNA in situ hybridization and the quantitative RT-PCR analyses . The cell dissociation protocol required for quantitative RT-PCR introduces a delay in the analysis of gene expression relative to that obtained using the RNA in situ hybridization protocol , giving further time for the MO to lose efficacy . Nevertheless , these results reveal that Tfap2 activity , redundantly provided by Tfap2a and Tfap2e , promotes the differentiation of embryonic melanophores . How does Tfap2 activity , mediated by Tfap2a and Tfap2e , effect melanophore differentiation ? In tfap2a/e doubly-deficient embryos , melanophore differentiation fails but can be rescued by forced expression of mitfa . One model to explain these findings is that Mitfa and Tfap2 normally co-activate genes important for melanophore differentiation , but in the absence of Tfap2 , elevated levels of Mitfa can suffice to do so ( Figure 9A ) . Thus , Tfap2 family members may directly activate genes involved in melanin synthesis , such as dct , tyrp1b , and possibly tyr , all of which are known to be Mitfa targets [47]–[49] . Consistent with this possibility , recent studies have identified conserved DNA elements adjacent to the dct and tyrp1b genes that have melanocyte enhancer activity [13] , and some of these contain putative Tfap2 binding sites . Simultaneous inhibition of tyrp1a and tyrp1b blocks melanization of zebrafish melanophores , suggesting that tyrp1a/b may partially mediate Tfap2a/e activity within these cells [50] . A variation of this model is that , rather than Tfap2 itself functioning as a co-activator with Mitfa , the protein product of a gene stimulated by Tfap2 does so . For instance , Tfap2 activates expression of estrogen receptor alpha ( ERα ) [51] , [52] . ERα , together with p300 , interacts with Mitf to strongly activate the Dct promoter [53] . It is also possible that the effector of Tfap2 activity is an enzyme that alters the activity , translation , or longevity of the Mitfa protein ( Figure 9B ) . Thus , perhaps mitfa RNA levels are the same in tfap2a deficient vs . tfap2a/e deficient embryos , but Mitfa activity is reduced in the latter . For instance , the Tfap2-effector may be a receptor tyrosine kinase ( RTK ) whose activity results in posttranslational activation of Mitfa , i . e . similar to a proposed role of Kit [7] , [8] . Supporting such a possibility , Kita itself is necessary for differentiation of embryonic melanophores in zebrafish in certain experimental conditions [54] [23] . A variety of RTKs are candidates for the Tfap2 effector in melanophore differentiation , including Erbb3 [55] , [56] , IGF1R [57] , FGF receptor [58] , c-Ret [59] , and c-MET [60] . Two G-protein coupled receptors , which like RTKs can stimulate the MAP Kinase pathway , are also candidates . First , Endothelin receptor b ( Ednrb signaling ) promotes melanocyte differentiation in mammals , in part by activating MAP Kinase signaling and Mitfa [61]–[64] . While embryonic melanophores differentiate normally in zebrafish ednrb1 mutants [65] , uncharacterized ednrb homologues are present in the zebrafish genome ( e . g . , on chromosome 9 ) and may function in embryonic melanophores . Second , Melanocortin 1 receptor ( Mc1r ) is necessary for normal levels of pigmentation in zebrafish [66] and in mammals [67] , and MC1R expression may be directly regulated by TFAP2A , because it has been shown that TFAP2A binds DNA adjacent to the MC1R gene in HeLa cells ( chromatin immunoprecipitation results ) [68] . Finally , Tfap2 could normally repress expression of an Mitfa phosphatase , alter processing of the mitfa transcript , change Mitfa translation or change Mitfa protein stability . All these scenarios would result in similar mitfa mRNA levels in situ but weaker Mitfa activity when Tfap2 levels are reduced , and would potentially be by-passed by over-expression of mitfa mRNA . The direct targets of Tfap2 in melanocytes are currently under investigation . Zebrafish embryos and adults were reared as described previously [69] , in the University of Iowa Zebrafish Facility . Embryos were staged by hours or days post fertilization at 28 . 5°C ( hpf or dpf ) [70] . Homozygous mutant embryos were generated from heterozygous adults harboring a presumed null allele of tfap2a ( lockjaw , tfap2ats213 ) [26] , mitfa ( mitfab692 ) [31] , or kita ( kitab5 ) [23] , as indicated . First-strand cDNA was synthesized from total RNA harvested from embryos at 4 hpf and 24 hpf as described [25] . A 1 . 4 kb full-length zebrafish tfap2e cDNA was amplified from the wild-type cDNA using the following primers: forward , 5′-GGA TTC ATG TTA GTC CAC TCC TAC TC-3′ , reverse , 5′-TTA TTT GCG GTG CTT GAG CT-3′ . This cDNA includes the entire open reading frame and was inserted into the pCR4-TOPO vector ( Invitrogen , Carlsbad , CA ) . A 1 . 3 kb fragment of zebrafish tyrp1b cDNA was amplified from the wild-type 24 hpf cDNA using the following primers: forward , 5′-GAG AGC GGA TGA TAT AAG GAT GTG G-3′ , reverse , 5′-GCC CAA TAG GAG CGT TTT CC-3′ . This cDNA was inserted into pSC-A vector ( Stratagene , La Jolla , CA ) . In designing a tfap2e construct in which expression is disrupted , the exon 2 splice donor site and the exon 3 splice donor sites had to be inferred from comparison of the cDNA to the corresponding genomic sequence ( http://uswest . ensembl . org/Danio_rerio/Info/Index ) . MOs complementary to these sites were ordered: tfap2e e2i2 MO , 5′-ATA CAA GAG TGA TTG AAC TCA CCT G-3′; tfap2e e3i3 MO , 5′-CAC ATG CAG ACT CTC ACC TTT CTT G-3′ ( Gene Tools , Philomath , OR ) . In addition , a MO targeting the tfap2e translation start site ( AUG MO ) was designed , 5′-GCT GGA GTA GGA GTG GAC TAA CAT C-3′ . MOs were reconstituted to 5 mg/ml in water and stored at room temperature ( 25°C ) . Immediately before use , they were diluted to 0 . 5 mg/ml in 0 . 2 M KCl . MOs ( 4–8 nl of diluted stock ) were injected into the yolk underlying the blastomeres of embryos at the 1–4 cell stage . Upon injection of 3 ng or more of either MO , we saw evidence of non-specific toxicity , i . e . , patches of opacity in the brain and spinal cord that did not develop when 5 ng of a p53 MO ( 5′-GCG CCA TTG CTT TGC AAG AAT TG-3′ ) was injected [71] . To assure strong penetrance while preventing non-specific toxicity , we used 3 ng/embryo of tfap2e e3i3 MO plus 5 ng/embryo of p53 MO to generate tfap2a−/eMO embryos . For double MO experiments ( tfap2aMO/tfap2eMO ) , 3 ng of tfap2e e3i3 MO , 5 ng tfap2a e2i2 MO ( 5′-GAA ATT GCT TAC CTT TTT TGA TTA C-3′ ) and 5 ng of p53 MO were injected together . To test the efficacy of the tfap2e MOs , we used a pair of primers flanking a 305 bp fragment between exon 2 and exon 4 of tfap2e for RT-PCR ( forward , 5′-CAC CAC GGC CTG GAT GAT ATT-3′; reverse , 5′-AGG ACT CCT CCA AGC AGC GA-3′ ) . Additionally , where noted , a control MO ( controlMO ) was used for comparison ( 5′-CCT CTT ACC TCA GTT ACA ATT TAT A-3′ ) . To create genetic chimeras , we injected donor embryos with 5 nl of 1% lysine-fixable biotinylated-dextran , 10 , 000 MW ( Sigma , St . Louis , MO ) . At the sphere stage ( 4 hpf ) , about 100 cells were withdrawn from each donor embryo using a manual-drive syringe fitted with an oil-filled needle ( Fine Science Tools , Vancouver , BC ) , and about 20 cells were inserted into each of several host embryos at the same stage . In placing these cells , we aimed for a position near the animal pole , to target clones to the ectoderm [72] . Host embryos were allowed to develop to 48 hpf , fixed , images taken , and then processed using an ABC kit ( Vector Labs , Burlingame , CA ) and DAB to reveal biotin as previously described [73] , and subsequently photographed . The following restriction fragments were used to generate DIG-labeled antisense RNA probes ( Roche Diagnostics , Mannheim , Germany ) for whole mount in situ hybridization: tfap2e , NotI/T3; tyrp1b , BamHI/T3; dct , EcoRI/T7 [74]; mitfa , EcoRI/T7 [31] . Standard procedures were followed as previously described [75] . For total cell counts , 10–20 embryos were analyzed per group ( see figure legend ) . For immunohistochemistry , a monoclonal anti-Pax7 antibody [33] was used at a 1∶25 dilution ( supernatant obtained from the Developmental Studies Hybridoma Bank at the University of Iowa , USA ) . The primary antibody and an anti-DIG antibody were added during routine whole mount in situ hybridization . Following development of whole mount in situ hybridization with NBT/BCIP , the embryos were blocked and then incubated with an Alexa-488 conjugated goat-anti-rabbit secondary antibody , as previously described [76] . After several washes , the embryos were mounted in 50% glycerol/PBST , and photographed . Cell counts were performed on ten embryos per group , along the entire length of the hind yolk . Live embryos were reared to an appropriate stage , homogenized with a pestle , and dissociated with PBS containing trypsin and EDTA for 30 minutes at 33°C . After dissociation , cells were resuspended in PBS plus 3% fetal bovine serum ( FBS ) . EGFP-positive cells were counted using a Becton Dickinson FACScan . For cell sorting , cells were dissociated as previously described , and subsequently sorted , on a Becton Dickinson FACS DiVa , directly into buffer RLT and β-mercaptoethanol for subsequent RNA isolation ( RNeasy Plus Mini Kit , QIAGEN , Valencia , CA ) . FACScan cell counting , FACS DiVa cell sorting , and data analyses were conducted at the University of Iowa Flow Cytometry Facility . The isolation and culture of normal melanocytes and keratinocytes was performed as described previously , Mel 1 and Ker [77] , [78] , Mel 2 , 3 [79] ( see Figure 1M ) . Total messenger RNA was isolated using an RNeasy Plus Mini Kit ( QIAGEN , Valencia , CA ) , along with on-column DNase digestion according to the manufacturer's instructions . Lymphocytes ( Jurkat cells , clone E6-1 ) were obtained ( ATCC , Manassas , VA ) , and total RNA was isolated using the PerfectPure RNA Kit ( following manufacturer's instructions , 5 PRIME Inc . , Gaithersburg , MD ) . RNA concentrations were determined using a NanoDrop spectrophotometer ( Thermo Scientific ) and diluted to equal concentrations . For complementary DNA ( cDNA ) reactions , approximately 200 ng of total RNA was added to 0 . 5 µg random hexamers , plus 2 . 5 µl of 10 mM dNTPs ( Invitrogen; Carlsbad , CA ) , and brought to 30 µl with nuclease-free water . Reactions were heated to 65°C for 5 minutes , and cooled to 4 . 0°C for 5 minutes in a PTC-200 Peltier Thermo Cycler ( MJ Research; Ramsey , MN ) . We then added 19 µl of a master mix containing 10 µl of 5x First-Strand buffer ( Invitrogen ) , 5 µl of 0 . 1 M dithiothreitol , 20 units of RNasin ( Promega , Madison , WI ) , and nuclease-free water to a volume of 19 µl . Reactions were incubated at 25°C for 10 minutes , and then at 37°C for 2 minutes . Then 1 µl of Moloney-murine leukemia virus Reverse Transcriptase ( New England Biolabs , Ipswich , MA ) or 1 µl nuclease-free water was added to each reaction . Reactions were carried out at 37°C for 2 hours , followed by incubation at 75°C for 15 minutes . PCR reactions ( 25 µl ) were prepared with approximately 10 ng of cDNA , using the SYBR Green kit ( Applied Biosystems , Foster City , CA ) following the manufacturer's instructions . The following primers were used at a final concentration of 200 nM in separate PCR reactions: human TFAP2E ( forward: 5′-AAT GTG ACG CTG CTG ACT TC-3′; reverse: 5′-GGT CCT GAG CCA TCA AGT CT-3′ ) ; or human GAPDH ( forward: 5′-AGG TCG GAG TCA ACG GAT TTG-3′; reverse: 5′-GTG ATG GCA TGG ACT GTG GT-3′ ) . Quantitative real-time PCR in Low 96-well plates ( Bio-Rad , Hercules , CA ) was conducted using a Bio-Rad thermal cycler ( CFX96 Real-Time PCR Detection System ) and following the default protocol . Primers were designed to flank large exon-intron boundaries to avoid the potential amplification of contaminating genomic DNA . Also , RNA samples not reverse-transcribed ( -RT ) were used as a negative control . The 2ΔΔCT method was used to determine relative levels of gene expression between samples ( normalized to GAPDH ) [80] . Experiments were performed in triplicate and mean and standard error were calculated . Following real-time PCR , melt-curve analysis was performed to determine reaction specificity . Similar methods were used for qRT-PCR of sorted cells , with the exception that approximately 20 ng of RNA was used for cDNA synthesis . The following primers were used at a final concentration of 200 nM in separate PCR reactions: tyr ( forward: 5′-GGA TAC TTC ATG GTG CCC TT-3′; reverse: 5′-TCA GGA ACT CCT GCA CAA AC-3′ ) ; tyrp1b ( forward: 5′-TAT GAG ACA CTG GGC ACC AT-3′; reverse: 5′-CAC CTG TGC CAT TGA GAA AC-3′ ) ; dct ( forward: 5′-CCT CGA AGA ACT GGA CAA CA-3′; reverse: 5′- CAA CAC CAA CAC GAT CAA CA-3′ ) ; and β-actin ( forward: 5′-CGC GCA GGA GAT GGG AAC C-3′; reverse: 5′-CAA CGG AAA CGC TCA TTG C-3′ ) . Again , the 2ΔΔCT method was used to determine relative levels of gene expression between samples , first normalizing both samples to β-actin , and then comparing relative gene expression levels in tfap2a/e doubly-deficient cells to those in tfap2a deficient cells . Apoptotic cell death was revealed in whole embryos by terminal transferase dUTP nick-end labeling ( TUNEL ) as described [81] . The terminal transferase reaction was terminated by incubation at 70°C for 30 min , and embryos were processed with anti-FITC-alkaline phosphatase antibody and developed with NBT/BCIP , as for an RNA in situ hybridization . For tfap2aGR mRNA rescue experiments , approximately 5 nL of 0 . 075 mg/mL tfap2aGR or lacZ encoding mRNA , transcribed in vitro ( mMessage mMachine kit , Ambion , Austin , TX ) was injected into one of four cells of embryos previously injected with tfap2a/e/p53 MOs ( similar concentration as indicated before ) . Embryos were raised until they reached approximately 75% epiboly , at which point dexamethasone ( dissolved in EtOH ) was added to the fish water at a final concentration of 40 µM . For DNA rescue experiments , 5 nL of a 0 . 025 mg/ml plasmid encoding 4 . 9 Kb of the sox10 promoter driving full length mitfa [41] was injected at the one cell stage , followed by co-injection of various MO combinations ( control MO and p53 MO or tfap2aMO , tfapeMO ) . Embryos were then raised until approximately 36 hpf and fixed in 4% paraformaldehyde overnight . Finally , embryos were rinsed in PBST , mounted in 3% methylcellulose , and photographed . To analyze the mean gray value of melanophores , embryos were first fixed at the appropriate stage in 4% paraformaldehyde overnight . Embryos were then rinsed in PBST and mounted in 3% methylcellulose , and images of single melanophores were taken near the otic vesicle at 40x . All lighting conditions remained constant throughout image capturing . 6–10 melanophores were imaged per embryo , and 10 embryos were analyzed per group ( roughly 70–80 melanophores per group ) . Images were converted to a 32 bit gray image and then processed using the auto threshold function in ImageJ software ( Version 1 . 40 g , National Institutes of Health , Bethesda , MD ) , creating an outline of the melanophore being analyzed . After application of the auto threshold function , a selection was created of the pixels highlighted , and a measurement reporting mean gray value for the given area was taken . An inverse of the selection was then created , highlighting the background ( area not occupied by the melanophore ) , and a similar measurement was taken , reporting the mean gray value of the surrounding background . The difference was then calculated between the mean gray value of the melanophore and the surrounding background , resulting in the normalized mean gray value of the melanophore . Averages were then calculated for all melanophores measured per group , and standard deviation was calculated . For mitfa-positive and TUNEL-positive cell counts , the entire region overlying the hind yolk was counted . For melanophore cell counts in sox10:mitfa rescue experiments , the total number of melanophores in the embryo body ( excluding yolk and hind yolk ) were counted . Embryos were fixed in 4% paraformaldehyde overnight , washed in PBST , and mounted in 3% methylcellulose for counts . Embryos were mounted and then counted blindly by an independent observer .	Neural crest-derived pigment cells , known as melanocytes , are important to an organism's survival because they protect skin cells from ultraviolet radiation , camouflage the organism from predators , and contribute to sexual selection . Networks of regulatory proteins control the steps of melanocyte development , including lineage specification , migration , survival , and differentiation . Gaps in our understanding of these networks hamper progress in effective prevention and treatment of diseases of melanocytes , including metastatic melanoma and vitiligo . Studies conducted in tissue-culture cells and mouse embryos implicate regulatory proteins including the transcription factor TFAP2A , the growth factor receptor KIT , and the transcription factor MITF as being important for multiple steps in melanocyte development . Abnormalities in TFAP2A , KIT , and MITF expression in melanoma highlight the importance of this pathway in human disease . Here we show that a gene closely related to TFAP2A , tfap2e , is expressed in zebrafish embryonic melanocytes and human melanocytes . We provide evidence that Tfap2e cooperates with Tfap2a to promote expression of zebrafish kita in embryonic melanocytes . Further we show that an effector of Tfap2a/e activity other than Kita is required for melanocyte differentiation and that this effector acts upstream or in parallel with Mitfa activity . These findings reveal unexpected complexity to the gene-regulatory network governing melanocyte differentiation .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "genetics", "and", "genomics", "developmental", "biology" ]	2010	Differentiation of Zebrafish Melanophores Depends on Transcription Factors AP2 Alpha and AP2 Epsilon
Alternative splicing is commonly used by the Metazoa to generate more than one protein from a gene . However , such diversification of the proteome by alternative splicing is much rarer in fungi . We describe here an ancient fungal alternative splicing event in which these two proteins are generated from a single alternatively spliced ancestral SKI7/HBS1 gene retained in many species in both the Ascomycota and Basidiomycota . While the ability to express two proteins from a single SKI7/HBS1 gene is conserved in many fungi , the exact mechanism by which they achieve this varies . The alternative splicing was lost in Saccharomyces cerevisiae following the whole-genome duplication event as these two genes subfunctionalized into the present functionally distinct HBS1 and SKI7 genes . When expressed in yeast , the single gene from Lachancea kluyveri generates two functionally distinct proteins . Expression of one of these proteins complements hbs1 , but not ski7 mutations , while the other protein complements ski7 , but not hbs1 . This is the first known case of subfunctionalization by loss of alternative splicing in yeast . By coincidence , the ancestral alternatively spliced gene was also duplicated in Schizosaccharomyces pombe with subsequent subfunctionalization and loss of splicing . Similar subfunctionalization by loss of alternative splicing in fungi also explains the presence of two PTC7 genes in the budding yeast Tetrapisispora blattae , suggesting that this is a common mechanism to preserve duplicate alternatively spliced genes . Gene duplication is thought to be a major source of evolutionary innovation . Although the fate of duplicated genes is incompletely understood , it is thought to fit one of three patterns: nonfunctionalization , neofunctionalization or subfunctionalization . Of these nonfunctionalization is thought to be the most common . Immediately after duplication , duplicated genes are typically redundant . Thus , purifying selection cannot provide selective pressure to maintain both . The absence of selective pressure generally leads to loss of function mutations ( nonfunctionalization ) in one of the copies , followed by loss of that copy of the gene . In neofunctionalization , one of the duplicated copies acquires a new advantageous function that is different from the ancestral function maintained by the other copy [1] . Subfunctionalization can occur when an ancestral gene carries out more than one function . If one duplicated copy mutates so that it loses one of the functions , and the other copy mutates so that it loses a separate function , selective pressure can subsequently maintain both copies by selecting for both functions [2] , [3] . Multiple functions in this context can mean being expressed in multiple cell types , encoding proteins localized to different compartments , encoding proteins with distinct biochemical activities , etc . Saccharomyces cerevisiae is an excellent model organism to study the fate of duplicated gene pairs because an ancestor underwent a whole genome duplication ( WGD ) approximately 100 million years ago resulting in a transient increase in genome size from around 5000 protein coding genes to 10 , 000 [4] , [5] , [6] . Following genome duplication most duplicated genes were lost ( nonfunctionalized ) , but 544 duplicated gene pairs that arose from WGD remain [4] . The genomes of many related species have been sequenced , which revealed through synteny which genes were duplicated as part of the WGD [6] , [7] , [8] . The related genomes also provide a large amount of sequence information on the duplicated genes and their non-duplicated homologs . The pattern of gene retention in these genomes revealed that nonfunctionalization after WGD is random such that different post-WGD species retained different subsets of duplicated genes [9] . In addition , gene function can be easily assayed in S . cerevisiae . Using these advantages , we have previously shown that subfunctionalization is a major mechanism by which duplicated S . cerevisiae genes were retained , which was confirmed by others [10] , [11] , . One example of subfunctionalized genes resulting from the WGD event is provided by SKI7 and HBS1 [11] . The S . cerevisiae Ski7 and Hbs1 proteins both recognize ribosomes stalled during translation and initiate degradation of the mRNA . However , they recognize different stalled ribosomes and initiate mRNA degradation differently . When an mRNA lacks an in frame stop codon , the ribosome is thought to translate until it reaches the 3′ end of that mRNA [13] . The stalled ribosome is then recognized by Ski7 , which recruits the RNA exosome to degrade that mRNA . In contrast , Hbs1 recognizes ribosomes stalled within the coding region , for example due to a structure or damage in the mRNA [14] . Recognition by Hbs1 causes cleavage of the mRNA in an RNA exosome-independent manner [14] , [15] . Although the SKI7 and HBS1 genes of S . cerevisiae perform distinct functions , we have previously shown that the single ancestral gene performed both functions [11] . The related budding yeast Lachancea kluyveri diverged from S . cerevisiae before WGD and thus contains a single ortholog to SKI7 and HBS1 , which we will call SKI7/HBS1 . Our key finding was that when the L . kluyveri SKI7/HBS1 gene was introduced into S . cerevisiae it could complement the defects caused by both ski7Δ and hbs1Δ , thus indicating that this single gene carried out both functions [11] . Since the function of duplicated genes can diverge from each other through neo- or subfunctionalization , gene duplication may be one way to generate a more diverse proteome . The proteome can also be diversified through alternative splicing , where one gene generates multiple distinct mRNAs that each encode a distinct protein . Although alternative splicing is important to diversify the proteome in metazoans , it is much rarer in the fungal kingdom . Most fungal alternative splicing events that have been described are of the intron retention type , where the spliced mRNA encodes a functional protein , and the unspliced mRNA is nonfunctional . For example , transcriptome sequencing of Aspergillus oryzae identified only 8 . 6% of the genes as alternatively spliced , which is 10-fold lower than in humans and 92% of the Aspergillus alternative splicing was intron retention [16] . A well-studied and typical example of fungal intron retention is the S . cerevisiae CYH2 gene . The CYH2 mRNA encodes a 17 KDa ribosomal protein . The intron in the CYH2 pre-mRNA is retained approximately 50% of the time , which results in an mRNA that codes for a 2 KDa peptide with no known function . Furthermore , intron-retained mRNAs are typically very rapid degraded by the nonsense-mediated mRNA decay pathways [17] . In these cases instead of diversifying the proteome , intron retention may function to regulate gene expression . Similarly , the S . cerevisiae SRC1 gene is alternatively spliced using alternative 5′ splice sites , but only the longer splice isoform has been shown to be functional [18] , [19] . To the best of our knowledge , the only case in which intron retention or alternative splicing leads to two functional mRNAs in S . cerevisiae is in PTC7 , which contains one intron and encodes a protein phosphatase subunit . If this intron is spliced out , the mRNA is translated into a protein that is imported into the mitochondria , while after intron retention the mRNA is translated into a protein that is inserted into the nuclear envelope [20] . A few other cases have been described were fungi use alternative splicing to target a protein to multiple locations [21] , [22] , [23] . As mentioned above , there are multiple ways a gene can be multifunctional in the context of subfunctionalization . A corollary of that is that subfunctionalization can occur through distinct molecular changes . In yeast , subfunctionalization through changes in the coding region seem to be common [10] , [11] , [12] . In these cases a single amino acid change can be responsible for subfunctionalization [10] , [12] . In multicellular organisms , genes that are expressed in multiple cell types , in response to multiple stimuli , or by multiple transcription factors can be subfunctionalized through changes in expression pattern [e . g . ref 3] . Subfunctionalization through changes in splicing patterns have been described in a few cases [e . g . refs 24] , [25] . In these cases , an alternatively spliced gene upon duplication results in two genes where one gene follows one ancestral splicing pattern and the other follows another ancestral splicing pattern . However , the function of these alternative splicing isoforms is often not clear . Thus while loss of alternative splicing happens at the same time as some gene duplications , whether they cause subfunctionalization has not been experimentally demonstrated . Here we show that the pre-WGD ancestor of SKI7 and HBS1 was alternatively spliced . We also show that in most extant fungi , including ascomycetes and basidiomycetes , the SKI7/HBS1 gene is still alternatively spliced , thereby describing the by far most conserved fungal alternative splicing event . The L . kluyveri alternative splicing isoforms are functionally distinct , such that one spliced mRNA encodes a functional Hbs1 , while an alternatively spliced mRNA encodes a functional Ski7 . Sequence analysis indicates that a very similar subfunctionalization occurred in an ancestor of the Schizosaccharomyces genus . Finally , while the S . cerevisiae PTC7 gene encodes two differently localized proteins through intron retention , in a related species this gene is replaced by a pair of duplicated genes that arose form WGD . Thus , evolution of a fungal ancestral alternatively spliced gene into two subfunctionalized genes occurred at least three times: twice for the SKI7/HBS1 gene , and once for PTC7 . This further suggests that alternative splicing and gene duplication are not independent mechanisms to diversify the proteome , but instead are interrelated . Several genomes of Saccharomycetaceae have been sequenced , but incompletely annotated . Upon careful analysis of these sequences we noticed that the SKI7/HBS1 genes in five pre-WGD Saccharomycetaceae each have a potential intron ( Figure 1A ) . In contrast , S . cerevisiae and five other post-WGD species lack introns in both SKI7 and HBS1 . Furthermore , each pre-WGD gene has two potential 3′ splice sites , resulting in the potential to encode two different conserved proteins . In the case of L . kluyveri these proteins are predicted to be 70 and 96 KDa ( Figure 1A ) . Although alternative splicing is rare in fungi , we speculated that the SKI7/HBS1 gene may be alternatively spliced . We used rt-PCR to show that both predicted splice sites are indeed used . Use of the proximal 3′ splice site was confirmed using rt-PCR with a primer upstream of the 5′ splice site and a primer downstream of the proximal 3′ splice site and sequencing the resulting PCR product ( Figure 1B left panel ) . Use of the distal 3′ splice site was similarly confirmed using a primer downstream of the distal 3′ splice site ( Figure 1B right panel ) . Thus , the L . kluyveri gene is indeed alternatively spliced through the use of alternative 3′ splice sites . To determine whether both spliced mRNAs were used to generate a protein , we generated a plasmid that introduced the HA epitope at C-terminus of the L . kluyveri ORF . A western blot shows that two proteins of the expected size are indeed made in L . kluyveri ( Figure 1C; second lane ) . We further modified the HA-tagged plasmid by deleting the intron . In one construct we deleted sequences between the 5′ and distal 3′ splice sites , such that only the short 70 KDa splice isoform could be expressed . Figure 1C ( third lane ) shows that the encoded protein comigrates precisely with the smaller of the two species seen when the intron is included . Conversely , in another plasmid ( fourth lane ) we deleted sequences between the 5′ and proximal 3′ splice sites and as expected only the large 96 KDa isoform was made . Therefore , the rt-PCR and Western blot data show that the single SKI7/HBS1 gene of L . kluyveri is used to generate two distinct proteins through use of alternative 3′ splice sites . We have previously shown that the L . kluyveri SKI7/HBS1 can carry out both the Ski7 and Hbs1 functions by showing that the L . kluyveri gene can complement both a ski7Δ and an hbs1Δ in S . cerevisiae [11] . To test whether both L . kluyveri proteins were generated in this context , we introduced the same HA-tagged constructs describe above into a wild-type S . cerevisiae strain . Western blot analysis indicates that when the L . kluyveri SKI7/HBS1 gene is introduced into S . cerevisiae , both L . kluyveri proteins are made ( data not shown ) . To determine whether one splice isoform carries out the Ski7 function and the other splice isoform carries out the Hbs1 function , plasmids expressing one or both proteins were introduced into both a dcp1-2 ski7Δ strain and an rps30aΔ hbs1Δ strain . A ski7Δ by itself does not result in a growth phenotype , but in combination with dcp1-2 , results in a failure to grow at 30°C [26] . This ski7Δ phenotype can be complemented by the unmodified L . kluyveri gene ( Figure 2A third row top panel ) and by the long splice isoform ( fourth row ) but not by the short splice isoform ( fifth row ) . Thus , only the long splice isoform can perform the Ski7 function . hbs1Δ by itself does not result in a growth phenotype but in combination with rps30aΔ results in slow growth at room temperature [27] . This hbs1Δ phenotype can be complemented by the unmodified L . kluyveri gene ( Figure 2B third row bottom panel ) and by the short splice isoform ( 5th row ) , but not by the long splice isoform ( 4th row ) . We conclude that alternative splicing generates two functionally distinct polypeptides , with the long splice isoform functioning similar to Ski7 and the short splice isoform similar to Hbs1 . Multiple sequence alignment ( Figure S1 ) identified several sequence elements that correlated with Hbs1 function in short splice isoforms of pre-WGD Saccharomycetaceae SKI7/HBS1 genes and post-WGD HBS1 genes . The structure of S . cerevisiae Hbs1 has been solved , which shows a structured N-terminal domain and a C-terminal GTPase domain connected by a flexible linker [28] , [29] , [30] . The structured N-terminal domain is conserved in other post-WGD Hbs1 proteins but not in post-WGD Ski7 . In the alternatively spliced pre-WGD homologs , this domain is encoded by exon 1 , and thus present in both splice isoforms . The unstructured linker of Hbs1 is poorly conserved , with the exception of one sequence motif ( Motif H1 in Figure 2C ) . A similar sequence motif is encoded in pre-WGD SKI7/HBS1 genes , but has diverged very much in post-WGD SKI7 genes . The GTPase domain is also highly conserved in post-WGD Hbs1 proteins and pre-WGD Ski7/Hbs1 proteins ( Motifs G1 to G5 in Figure S1 ) , consistent with previous findings that Hbs1 GTPase activity is important for its function [31] , [32] , [33] . In contrast , Ski7 has not been shown to be an active GTPase and the domain has diverged rapidly post-WGD . Specifically , a catalytically important His residue in motif G3 is changed to Ser , Asn or Asp in post-WGD Ski7s . The structure of Ski7 has not been experimentally determined , but the N-terminus is known to be important for interaction with the RNA exosome and three other Ski proteins [34] . Multiple sequence alignment indicated that although the Ski7 N-terminus is generally poorly conserved , it contains three conserved sequence motifs ( Figure 2C and Figure S1; motif S1 , S2 , and S3 ) . Alignment of pre-WGD Ski7/Hbs1 sequences shows that these motifs are also conserved in the pre-WGD species . Motif S1 and S2 are encoded between the two alternative 3′ splice sites and thus are only present in the longer splice form . These observations suggest that the short isoform of pre-WGD SKI7/HBS1 genes may fail to carry out Ski7 function because they lack motifs S1 and S2 . Most ascomycetes contain a single SKI7/HBS1 gene . To determine whether alternative splicing of SKI7/HBS1 was restricted to pre-WGD Saccharomycetaceae such as L . kluyveri or is a more ancient feature , we next looked at the more distantly related ascomycetes . The phylum Ascomycota can be divided in three subphyla , the Saccharomycotina ( which includes Saccharomyces and Lachancea ) , the Pezizomycotina and the Taphrinomycotina . We therefore used the same rt-PCR and sequencing approach described above to analyze SKI7/HBS1 splicing in Aspergillus nidulans and Saitoella complicata , which are members of the Pezizomycotina and Taphrinomycotina , respectively . Figure 3A shows that Aspergillus nidulans also uses alternative 3′ splice sites in SKI7/HBS1 to generate two distinct mRNAs . The single A . nidulans SKI7/HBS1 gene contains four introns . The second intron contains two predicted alternative 3′ splice sites , and rt-PCR and sequencing indicated that both are used . Similarly , the single SKI7/HBS1 gene in Saitoella complicata contains seven introns , and the second intron contains two predicted 3′ splice sites . Figure 3B shows that both 3′ splice sites are used . Notably , the alternative spliced introns in Lachancea , Aspergillus and Saitoella are in the same position , just upstream of motif S3 characteristic of Ski7 . Therefore , the capacity to use alternative 3′ splice sites in SKI7/HBS1 is conserved throughout the phylum Ascomycota . Although we were able to predict alternative 3′ splice sites in the single SKI7/HBS1 gene of many other fungi , we noted four notable differences in the Schizosaccharomyces genus , a subset of the CTG clade ( Saccharomycetales that use CUG as a serine codon instead of the canonical leucine ) , and the basidiomycetes Cryptococcus neoformans and Ustilago maydis . Species within the Schizosaccharomyces genus have duplicated SKI7 and HBS1 genes ( see next section ) , while species in the CTG clade and the two basidiomycetes each contain a single SKI7/HBS1 gene , but these genes lack obvious alternative 3′ splice sites . The CTG clade can be divided into two smaller clades . One clade contains Candida guilliermondii , Debaryomyces hansenii , and Candida lusitaniae . In all three species there are two potential 3′ splice sites in locations similar to L . kluyveri ( Figure S2 ) . Thus , these three species appear to use the same mechanism as other ascomycetes to express two splice isoforms from a single SKI7/HBS1 gene . The other clade includes C . albicans , C . dubliniensis , C . tropicalis , and C . parapsilosis . These four Candida species also contain a predicted intron within their single HBS1/SKI7 gene , however we only detected one potential 3′ splice site , which corresponds to the distal 3′ splice site of other ascomycetes . rt-PCR and sequencing confirmed that this 3′ splice site is used to generate an mRNA that is equivalent to the short splice isoform of L . kluyveri ( Figure 4A ) . Although a proximal 3′ splice site is absent in these species , the capacity to encode Ski7-like sequence upstream of the distal 3′ splice site is conserved in these four species . In all four species motif S1 starts with a methionine . Since motif S1 is at the extreme N-terminus of the protein in post-WGD SKI7 genes , we tested the hypothesis that the four Candida species generated a distinct mRNA that uses the AUG codon at the beginning of motif S1 as start codon . Figure 4A shows that 5′RACE indeed identified an mRNA with a 5′ end five nucleotides upstream of the conserved Ski7 motif S1 . Thus , C . albicans uses alternative transcription start sites/first exons instead of alternative 3′ splice sites to generate two distinct mRNAs from the single SKI7/HBS1 gene . The single SKI7/HBS1 gene in basidiomycetes also appears to be alternatively spliced , although the details differ from the ascomycete situation . The Cryptococcus neoformans and Ustilago maydis genes have 9 and 4 annotated exons , respectively . Several EST sequences indicate that exons 4 and 5 in C . neoformans and exon 2 in U . maydis are skipped to generate an Hbs1-like protein ( Figure 4B and 4C ) . Existing annotations suggest that a Ski7-like protein can be encoded by inclusion of these exons . However , EST , RNA sequencing and protein sequence similarity suggests an alternative where the annotated intron 4 of C . neoformans ( and intron 2 of U . maydis ) is not a true intron ( See Text S1 ) . This alternative mechanism encodes a truncated protein that resembles the N-terminus of Ski7 , but is missing the GTPase domain ( Figure 4B and 4C ) . This mechanism is strikingly similar to potential alternative splicing of the metazoan homolog ( See Text S1 and Figure S5 ) . Overall , while it is clear that the basidiomycetes use alternative splicing of their single SKI7/HBS1 gene , it is not entirely clear which mechanism they use to generate a Ski7-like protein . The fourth exception to conserved alternative 3′ splice sites in SKI7/HBS1 occurs in the Schizosaccharomyces genus . The Hbs1 protein of Schizosaccharomyces pombe has been previously studied and appears to function similarly to S . cerevisiae Hbs1 [29] . In addition we found an uncharacterized paralog in S . pombe ( Systematic name SPAP8A3 . 05 ) that encodes amino acid sequence motifs characteristic of Ski7p ( labeled S1 , S1′ and S3 in Figure S3 ) . We therefore refer to this S . pombe gene as SKI7 . Thus , an ancestor to S . pombe must have independently duplicated its SKI7/HBS1 gene . The other three Schizosaccharomyces species with sequenced genomes each contain one clear ortholog of HBS1 and one clear ortholog of SKI7 ( Figure S3 ) , which suggests this duplication occurred before the Schizosaccharomyces species diverged from each other . Of the sequenced fungal genomes , the most closely related species with a single SKI7/HBS1 gene is S . complicata . As discussed above , S . complicata has an alternatively spliced SKI7/HBS1 gene , and thus the duplication in the Schizosaccharomyces genus appears to have occurred after it diverged from S . complicata , but before the Schizosaccharomyces species diverged from each other . Our above observations indicate that S . cerevisiae SKI7 and HBS1 evolved from a single alternatively spliced ancestral gene and that they correspond to the different splice isoforms of the pre-WGD ancestor . This conclusion suggests that a similar mechanism may apply to other alternatively spliced genes . The only S . cerevisiae gene known to use splicing to generate two different functional proteins is PTC7 [20] . The PTC7 gene contains an intron that can either be spliced out or retained . Both the spliced and unspliced mRNAs encode type 2C protein phosphatases ( PP2C ) [20] . It has previously been noted that an intron of 3n nucleotides without any in frame stop codons is conserved in the PTC7 gene of twelve species within the Saccharomycetaceae , both pre- and post-WGD ( [20]; Figure 5 and Figure S4 ) . We searched for PTC7 genes in additional yeast genomes and noticed that the only species within the Saccharomycetaceae that did not follow this pattern is Tetrapisispora blattae . The genome of this species contains two PTC7 genes ( which we will call PTC7a and PTC7b ) . The synteny pattern ( http://wolfe . gen . tcd . ie/ygob/ ) indicates that after WGD the T . blattae lineage maintained both copies of PTC7 , while one copy was lost in the S . cerevisiae lineage . We searched for potential introns in PTC7a and PTC7b , but failed to find one in the PTC7a gene , while PTC7b contains a 103 nucleotide intron . The spliced PTC7b mRNA is predicted to encode a functional protein . In contrast to other post-WGD species , translation of the PTC7b unspliced mRNA does not encode a functional PP2C: translation starting from the normal start codon would end after 20 amino acids at a stop codon within the intron , while the only other in frame AUG codon is only 8 amino acid upstream of the normal stop codon . Thus , unlike other Saccharomycetaceae that encode two proteins from one alternatively spliced PTC7 gene , the T . blattae PTC7a and PTC7b genes each can only encode a single protein . The two splice isoforms of S . cerevisiae PTC7 are targeted to different compartments . The spliced S . cerevisiae PTC7 mRNA encodes a protein that is localized to the mitochondria , while the intron-retained mRNA encodes a protein localized to the nuclear envelope . Targeting to the nuclear envelope has been attributed to a predicted trans-membrane helix ( TM ) that is encoded by the retained intron [20] . We used the TMHMM 2 . 0 server ( http://www . cbs . dtu . dk/services/TMHMM/ ) to predict TMs in Ptc7 proteins of various Saccharomycetaceae . Each of the intron-retained mRNAs from post-WGD species and the T . blattae PTC7a gene encodes a single predicted TM near the N-terminus , suggesting that all of these proteins are targeted to the nuclear envelope . In contrast , none of the spliced mRNAs or PTC7b encode a predicted TM . We also used the PSORT II server ( http://psort . hgc . jp/form2 . html ) to predict TMs and protein localization . The TM results agreed with the TMHMM server . In addition , all of the spliced isoforms and PTC7b were predicted to contain a mitochondrial targeting sequence that was absent from the unspliced isoforms . Thus , T . blattae PTC7a encodes a single PP2C that is predicted to be targeted to the nuclear envelope , like the protein encoded by intron-retained PTC7 mRNA in S . cerevisiae , while T . blattae PTC7b gene appears to encode a single PP2C that is predicted to be targeted to the mitochondria , like the protein encoded by spliced PTC7 mRNA in S . cerevisiae . While separate functions of the S . cerevisiae PTC7 splice isoforms have not been defined , these results strongly suggests that the PTC7a and PTC7b genes are subfunctionalized , and thus that subfunctionalization by loss of splicing isoforms is not restricted to SKI7/HBS1 . We describe alternative splicing in fungal SKI7/HBS1 genes that is unusual in two respects . First , unlike in Metazoa , most fungal alternative splicing events do not produce two different proteins , but instead either have no known function or function to quantitatively regulate gene expression . Both of the mRNAs that are produced through SKI7/HBS1 alternative splicing are predicted to encode functional proteins , western blot analysis indicates that both predicted proteins are produced , and complementation of S . cerevisiae mutants shows that the two proteins are functionally distinct . Second , most alternative splicing events that have been described in fungi are not widely conserved but instead have only been described in one species [MDH1 , Ref . 22] , genus [GND1 , Ref . 21] or family [PTC7 Ref . 20] . Besides SKI7/HBS1 , the most conserved fungal alternative splicing events were recently reported for PGK1 in the Ascomycota and GAPDH in the Basidiomycota [23] . In contrast , alternative splicing of SKI7/HBS1 most likely arose before the divergence of the Ascomycota from the Basidiomycota , and it might even have arisen before fungi and animals diverged ( Text S1 and Figure S5 ) . Thus , this alternative splicing event has been maintained for at least 500 million years . It has been suggested that the Ski7 function is peculiar to S . cerevisiae and close relatives [6] , [35] . The finding of ancient alternative splicing indicates that Ski7 function is much older than appreciated and suggests that the ability to produce both Hbs1 and Ski7 is very important to fungi . Interestingly , one of the reasons why conserved alternative splicing in fungi has not been previously reported is that S . cerevisiae and S . pombe have been chosen somewhat arbitrarily as model fungi , and in both of these species the alternatively spliced SKI7/HBS1 gene has been replaced with duplicate genes . Although alternative splicing of SKI7/HBS1 is conserved in diverse fungi , we have characterized changes in expression strategies for Ski7 and Hbs1 , which are summarized in Figure 6A . The common ancestor of the ascomycetes and basidiomycetes appears to have had an alternatively spliced SKI7/HBS1 gene . Although the exact nature of SKI7/HBS1 alternative splicing event in basidiomycetes remains to be determined , it is clear that the mechanism by which a Ski7-like protein is expressed is different . Fully characterizing this event will require additional data from the basidiomycetes and/or additional early branching fungi . Independent duplications in the Saccharomyces and Schizosaccharomyces lineages allowed loss of alternative splicing ( Figure 6 events 2 ) . In the Saccharomyces lineage this duplication was part of a WGD , but in the Schizosaccharomyces lineage this duplication appears to be restricted to a single gene . The fourth evolutionary change occurred in the Candida clade in which a single SKI7/HBS1 gene gained an alternative initiation codon for Ski7 ( Figure 6 event 3 ) . Interestingly , after duplication , the S . cerevisiae , and S . pombe SKI7 genes also appear to have gained a new initiation codon ( see below ) . A major mechanism for intron loss in S . cerevisiae involves a transposon-encoded reverse transcriptase that converts spliced mRNA into cDNA . This cDNA then recombines with the gene , resulting in precise deletion of the intron [36] . Multiple sequence alignment shows that the intron in post-WGD Saccharomycetaceae is precisely deleted ( Figure S1 ) , consistent with it being deleted by this mechanism . Similarly , the short isoform from Saitoella complicata aligns very well with Hbs1 sequences from four Schizosaccharomyces species , indicating that the alternatively spliced intron was precisely deleted ( Figure S3 ) . The Saitoella SKI7/HBS1 gene contains seven introns . Recombination with a cDNA preferentially deletes introns near the 3′ end of the gene while introns near the 5′ end are more likely to be retained [36] . Consistent with intron loss by recombination with cDNA , HBS1 genes from all four sequenced Schizosaccharomyces species contain two introns that correspond to the first two S . complicata introns . Thus , the alternatively spliced intron was precisely deleted in both the Saccharomyces and Schizosaccharomyces lineage , possibly by reverse transcription of the mRNA into cDNA and recombination . In contrast to HBS1 , the intron in post-duplication SKI7 genes was not precisely deleted . Multiple sequence alignment ( Figure S1 ) of post-duplication Saccharomycetaceae showed that although the Ski7 N-terminus is generally poorly conserved , it contains three conserved sequence motifs ( motif S1 , S2 , and S3 ) . In post-WGD SKI7 genes in both the Saccharomycetaceae and in Schizosaccharomyces , motif S1 is located at the extreme N-terminus of Ski7 , starting with the Met translated from the start codon . This is most consistent with the model that after duplication S . cerevisiae SKI7 lost exon 1 and gained a new initiation codon . Similarly , Schizosaccharomyces Ski7 appears to have lost exons 1 to 4 , and gained a new initiation codon . This deletion of both the SKI7 intron and the first exon ( s ) in Saccharomyces and Schizosaccharomyces is inconsistent with simple recombination with a cDNA . It has previously been noted that a significant number of the genes that were duplicated and retained in S . cerevisiae after WGD are also duplicated in S . pombe [37] , suggesting parallel subfunctionalization events in the two species . Our observations provide striking similarities of the evolution of the Hbs1 protein in Saccharomycetaceae and Schizosaccharomyces and of the Ski7 protein in these same clades and in Candida albicans . Thus , after independent duplication in these lineages , a similar sequence of changes occurred , intron deletion through recombination ( HBS1 ) or generation of an alternative start site ( SKI7 ) . The only known example of a S . cerevisiae gene encoding functional proteins from both unspliced and spliced mRNA is PTC7 . This capacity to encode two proteins is conserved in most Saccharomycetaceae , but not in other Saccharomycetales ( including Candida , Pichia and Yarrowia species ) [20] . The time of divergence of the Saccharomycetaceae has not been carefully defined , but estimates indicate that it preceded divergence of mice from humans 75 million year ago . Strikingly , only about 30% of alternative splicing events are conserved from mice to human . Therefore , although alternative splicing of PTC7 is not nearly as well conserved as that of SKI7/HBS1 , it still has been conserved for a longer period than many human alternative splicing events . Since PTC7 homologs outside the Saccharomycetaceae lack an intron in the same position , the alternatively spliced PTC7 intron appears to have been gained by an ancestor of the Saccharomycetaceae ( Figure 6B event 4 ) . The PSORT server ( http://psort . hgc . jp/form2 . html ) predicts that Ptc7 proteins outside the Saccharomycetaceae ( i . e . from Candida , Pichia , and Yarrowia species ) localize to the mitochondria . Thus , the gain of an intron and the alternative splicing of this intron provides the Saccharomycetaceae with a PP2C localized to the nuclear envelope . After WGD ( Figure 6 event 5 ) , one copy of PTC7 was lost in the Saccharomyces lineage and the remaining copy maintained the capacity to encode two proteins ( Figure 6 event 6 ) . In contrast , in T . blattae both duplicated PTC7 genes were maintained , but each lost the ability to encode two PP2C splice isoforms ( Figure 6 event 7 ) and subfunctionalized into one gene for a mitochondrial PP2C and one gene for a PP2C in the nuclear envelope . Our combined bioinformatic and experimental analysis shows that alternative splicing and gene duplication may be interrelated events in a cycle that diversifies the proteome . In this cycle , gain of alternative splicing , duplication , and loss of alternative splicing and subfunctionalization result in functionally distinct paralogs . Although there have been some previous descriptions of subfunctionalization by loss of alternative splicing [e . g . ref 24] , our findings extend these descriptions in three important ways . First , previous descriptions are generally limited to two closely related species and thus cover only part of the evolutionary history of the gene . The ever-increasing number of sequenced fungal genomes allowed us to analyze SKI7/HBS1 and PTC7 gene structure and expression in diverse fungi thereby identifying when alternative splicing arose and was lost . The whole cycle of gain of an alternative splicing event , duplication , and loss of alternative splicing can be observed in the PTC7 gene of Saccharomycetaceae . In contrast , although we have not been able to identify when alternative splicing of SKI7/HBS1 arose , we have described independent subfunctionalization events by loss of alternative splicing in the Schizosaccharomyces and Saccharomyces lineages . Second , our observations suggest that subfunctionalization by loss of alternative splicing occurred very similarly in the Saccharomyces and Schizosaccharomyces lineages . Thus , unlike previously described isolated examples this phenomenon appears to have occurred multiple times . Third , in most previously described cases of loss of alternative splicing in duplicated genes it was not clear whether the splicing isoforms have distinct functions , and thus it is not clear in those cases that subfunctionalization and loss of alternative splicing are causally linked . Similarly , lack of one PTC7 splice isoform does not cause an easily identifiable phenotype under lab conditions , making it impossible to test whether the duplicate T . blattae genes can substitute for one but not the other splice isoform . In contrast , SKI7 and HBS1 have well-described functions , allowing us to demonstrate that the splice isoforms of L . kluyveri SKI7/HBS1 are functionally distinct . The S . cerevisiae , C . albicans , and wild-type L . kluyveri strains have been described [11] , [38] . The L . kluyveri ura3 mutant strain FM628 was a kind gift of Mark Johnston . The S . complicata type strain Y-17804 was obtained from the USDA ARS culture collection . A draft sequence of the S . complicata genome is available [39] . BLAST analysis using the S . pombe Hbs1 identified two non-overlapping contigs that encoded N- and C-terminal parts of a S . complicata homolog . Extensive BLAST analysis with other queries did not reveal additional homologs . We hypothesized that these contigs represented different parts of the same gene . We used PCR to close the gap between the contigs and sequenced the PCR product directly . The assembled sequence of the S . complicata SKI7/HBS1 gene has been submitted to Genbank ( Accession number JQ928880 ) . Since exons proved difficult to predict due to their small size , we sequenced rt-PCR products to determine the gene structure depicted in Figure 3B , which was then used to align the encoded protein with Schizosaccharomyces homologs . L . kluyveri and S . complicata were grown in YPD and RNA was extracted using our standard method for S . cerevisiae . C . albicans growth and RNA extraction was performed as described [40] . Aspergillus RNA isolated from strain R21 was a kind gift from Taylor Schoberle and Greg May ( UT MD Anderson Cancer Center ) . When contamination with genomic DNA proved to be a problem , we treated the RNA with DNase ( Promega ) . rt-PCR was done using a commercial kit per manufacturer's instructions ( Sigma-Aldrich ) . To close the gap between the two S . complicata contigs , we used the same kit but omitted the reverse transcriptase to amplify genomic DNA . 5′ RLM-RACE was done using a commercial kit per manufacturer's instructions ( Invitrogen ) . All PCR , rt-PCR and RACE products were sequenced directly ( Genewiz ) and exactly confirmed the predicted splice sites . pAv231 has been described [11] . It contains the L . kluyveri SKI7/HBS1 gene , including the promoter , intron and 3′UTR sequences . pAv844 and pAv847 have the long form and short form of the intron removed , respectively . They were generated by overlap PCR using the forward overlap oligonucleotides oAv963 ( tgctcaaccaaagcaagaag aagagaagaaattatctaaactgg ) for pAv844 and oAv965 ( tgctcaaccaaagcaagaag ccaaaaaacaagctatctctaatttc ) for pAv847 and their reverse complements oAv964 and oAv966 . To generate the HA-tagged SKI7/HBS1 gene , a plasmid encoding the C-terminus and triple HA tag was chemically synthesized ( by Genewiz ) . This fragment was used to replace the Bcl I to Bgl II restriction enzyme fragment of pAv231 , generating pAv888 , which contains the entire L . kluyveri SKI7/HBS1 gene with a C-terminal triple HA tag . A Bam HI Xba I fragment of pAv888 was then used to replace the Bam HI Xba I fragment of pAv844 and pAv846 to generate pAv903 and pAv905 . Thus , pAv903 encodes a C-terminally HA-tagged version of the long Ski7-like isoform of L . kluyveri SKI7/HBS1 , while pAv905 encodes the tagged short Hbs1-like isoform . Plasmids carrying the HA-tagged L . kluyveri SKI7/HBS1 gene , or empty vector controls , were transformed into S . cerevisiae or L . kluyveri strains using a standard method [41] . Transformants were selected on SC-URA , and then grown overnight in SC-URA . Total protein was isolated using the glass bead method and analyzed by western blotting using anti-HA antibodies ( Roche ) . As a control for the western blotting we used a S . cerevisiae strain with the HA epitope integrated at the C-terminus of the endogenous SKI7 locus . Fungal SKI7 and HBS1 homologs were initially identified by BLAST in the sequenced genomes and the predicted proteomes from 14 Saccharomycetaceae , 10 species from the CTG clade , 4 Pezizomycotina , 5 Taphrinomycotina and 2 basidiomycetes . In none of the cases was the gene annotated as alternatively spliced , and in a number of cases introns were not annotated or incorrectly annotated and were corrected based on our rt-PCR analysis . Multiple sequence alignments of various subsets of protein sequences were generated with the help of the ClustalW ( http://www . ch . embnet . org/software/ClustalW . html ) BOXSHADE ( http://www . ch . embnet . org/software/BOX_form . html ) , and WebLogo ( weblogo . berkeley . edu/ ) servers . The species trees in Figure 6 and Figures S2 and S3 are adapted from [42] , [43] , and [44] .	The role of duplicated genes in originating new functions is an important question in evolution . Almost all species have duplicated genes that carry out similar but not identical functions . Similar proteins that perform different functions can also be generated when one gene generates multiple mRNAs by alternative splicing that are translated into multiple similar proteins . This alternative splicing is prevalent in animal cells , but much rarer in fungi . Here we show that most fungi use alternative splicing to make a Ski7 protein and a Hbs1 protein from the same gene . Two fungi , budding yeast and fission yeast , have been much better characterized than other fungi , and co-incidentally they both have duplicated this alternatively spliced gene , resulting in two similar genes that are no longer alternatively spliced . Finally , we describe another example where two duplicate genes replace one alternatively spliced gene , suggesting that this is a common mechanism to divide functions among duplicate genes .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "model", "organisms", "molecular", "cell", "biology", "genetics", "biology", "microbiology", "evolutionary", "biology", "genetics", "and", "genomics" ]	2013	Alternative Splicing and Subfunctionalization Generates Functional Diversity in Fungal Proteomes
The covalent attachment of adenosine monophosphate ( AMP ) to proteins , a process called AMPylation ( adenylylation ) , has recently emerged as a novel theme in microbial pathogenesis . Although several AMPylating enzymes have been characterized , the only known virulence protein with de-AMPylation activity is SidD from the human pathogen Legionella pneumophila . SidD de-AMPylates mammalian Rab1 , a small GTPase involved in secretory vesicle transport , thereby targeting the host protein for inactivation . The molecular mechanisms underlying Rab1 recognition and de-AMPylation by SidD are unclear . Here , we report the crystal structure of the catalytic region of SidD at 1 . 6 Å resolution . The structure reveals a phosphatase-like fold with additional structural elements not present in generic PP2C-type phosphatases . The catalytic pocket contains a binuclear metal-binding site characteristic of hydrolytic metalloenzymes , with strong dependency on magnesium ions . Subsequent docking and molecular dynamics simulations between SidD and Rab1 revealed the interface contacts and the energetic contribution of key residues to the interaction . In conjunction with an extensive structure-based mutational analysis , we provide in vivo and in vitro evidence for a remarkable adaptation of SidD to its host cell target Rab1 which explains how this effector confers specificity to the reaction it catalyses . Microbial pathogens have developed a diverse spectrum of mechanisms to manipulate the human host and cause disease . Many bacterial proteins post-translationally modify host factors in order to alter their function . The covalent attachment of adenosine monophsophate ( AMP ) to threonine or tyrosine side chains within proteins , a process known as AMPylation ( adenylylation ) , was discovered more than 40 years ago in the Escherichia coli protein glutamine synthetase adenylyl transferase ( GS-ATase ) which regulates the enzyme glutamine synthetase through reversible AMPylation [1] . This post-translational modification recently re-emerged with the discovery of several virulence proteins from Gram-negative bacteria such as Vibrio parahaemolyticus , Histophilus somni , and Legionella pneumophila that AMPylate host proteins [2] , [3] , [4] . Surprisingly , each of these AMPylators was shown to target host cell GTPases of the Rho or Rab family . VopS from V . parahaemolyticus and IbpA from H . somni covalently modify Rho GTPases such as Cdc42 and Rac1 with AMP , thereby causing a collapse of the host cell actin cytoskeleton resulting in cell rounding [2] , [3] . In contrast , SidM ( DrrA ) from L . penumophila AMPylates host cell Rab GTPases [4] thereby exploiting intracellular vesicle trafficking routes . The finding that host cell GTPases are a preferred target of bacterial AMPylators can be attributed to the fundamental role these proteins play in all eukaryotic cells . Rab proteins regulate virtually all aspects of vesicle transport [5] , [6] . They function as molecular switches that cycle between an inactive GDP-bound state with predominantly cytosolic distribution and an active GTP-bound form that is associated with organelle membranes [7] , [8] , [9] . Rab activation requires a guanine nucleotide exchange factor ( GEF ) which promotes replacement of GDP with GTP to enhance the recruitment of downstream ligands , whereas Rab inactivation requires GTPase-activating proteins ( GAPs ) that stimulate the hydrolysis of GTP to GDP . Inactive GDP-bound Rabs are subsequently extracted from the membrane by a GDP dissociation inhibitor ( GDI ) and maintained in the cytosol for the next recruitment cycle . The opportunistic pathogen L . pneumophila , the causative agent of a severe pneumonia known as Legionnaires' disease , subverts membrane dynamics of the host cell by intercepting and modulating Rab1 [10] , [11] , [12] , [13] , the regulator of endoplasmic reticulum ( ER ) to Golgi vesicle transport . The organism infects human alveolar macrophages and multiplies within a specialized compartment called the Legionella-containing vacuole ( LCV ) . To ensure intracellular survival , L . pneumophila uses a specialized translocation machine known as the Dot/Icm type IV secretion system ( T4SS ) which mediates the delivery of over 200 proteins , termed effectors , from its own cytosol into the host cytoplasm [14] . The effector SidM ( DrrA ) binds phosphatidylinositol 4-phosphate present in the LCV membrane [15] and exhibits GEF as well as GDF activity towards host cell Rab1 [16] , [17] , thereby accumulating active GTP-Rab1 on the LCV surface . SidM then AMPylates tyrosine-77 located in the switch II region of Rab1 ( Y77Rab1 ) [4] . The bulky AMP moiety is believed to sterically interfere with the ability of Rab1 to interact with downstream ligands , most importantly GAPs such as the L . pneumophila Rab1GAP LepB , thereby making Rab1 insensitive to inactivation and maximizing its accumulation on the LCV . Notably , activated AMP-Rab1 is gradually removed from the compartment in a process that depends on the L . pneumophila effector protein SidD [18] , [19] . SidD is delivered into the host cell later than the AMPylase SidM and catalyzes AMP removal from Rab1 , a reaction referred to as de-AMPylation ( or de-adenylylation ) . Once Rab1 has been de-AMPylated , it becomes accessible to binding and inactivation by LepB and subsequent GDI-mediated extraction from LCV membranes [18] , [19] . The ability of L . pneumophila to regulate Rab1 membrane cycling through AMPylation and de-AMPylation provides a precedent for how reversible post-translational modification may be used by pathogens to precisely control the function of small Rab GTPases within host cells . To our knowledge , L . pneumophila SidD and the N-terminal domain ( AT-N ) of the E . coli GS-ATase are the only known enzymes with de-AMPylation activity , yet the reactions they catalyze differ significantly: While AMP removal performed by AT-N is strictly dependent on the presence of orthophosphate and produces ADP [20] , Rab1 de-AMPylation by SidD is phosphate-independent and generates AMP [18] , indicative of two fundamentally different mechanisms of de-AMPylation . SidD lacks any obvious sequence homology with the AT-N or other known proteins , although fold recognition analysis of the N-terminal portion of SidD predicted limited resemblance with members of the metal-dependent protein phosphatase ( PPM ) family . The conserved aspartate residues at position 92 and 110 , which are crucial for the activity of other phosphatases , also contribute structurally or chemically to SidD's catalysis [19] . Nonetheless , the molecular mechanism of AMP removal and the structural determinants for Rab1 recognition by SidD have remained largely unexplored . In this study , we use a multidisciplinary approach to characterize the structural and molecular details that determine substrate recognition and catalysis by SidD . We discover a unique mechanism by which SidD identifies AMPylated Rab1 but not Rho GTPases and that it performs de-AMPylation but not the chemically related de-phosphocholination reaction . The primary sequence of SidD consisting of 507 amino acids shows no homology to other proteins . Thus , it was unclear which part of SidD possessed de-AMPylation activity and if the protein potentially exhibited additional functions . Guided by secondary structure predictions we created N- or C-terminally truncated variants of SidD , purified them from E . coli , and tested their ability to catalyze removal of radiolabeled [α32P]AMP from Rab1 in vitro ( Figure 1A , B ) . We found that none of the C-terminal fragments and only the longest N-terminal variant spanning amino acid 1 to 379 ( SidD1–379 ) displayed catalytic activity comparable to the full length protein . We also noticed that several of the shorter variants ( SidD1–321 , SidD1–260 , or SidD1–164 ) were produced either as insoluble or unstable proteins in E . coli ( data not shown ) , suggesting that proper folding of these fragments was compromised by the truncations . To reduce folding or stability problems that might occur during protein production in E . coli , we employed a mammalian cell-based assay to analyze the de-AMPylation activity of the SidD variants within their host environment . We previously described that production of SidM in transiently transfected COS1 cells causes Golgi fragmentation and subsequent cell rounding and that this phenomenon can be partially repressed by simultaneously producing SidD in the same cell [18] , consistent with the fact that SidD's de-AMPylation activity antagonizes SidM's AMPylation activity . When analyzing GFP-tagged SidD variants in this rescue assay we found that none of the truncated proteins was capable of efficiently preventing SidM-induced COS1 cell rounding ( Figure 1A ) , not even SidD1–379 , the longest N-terminal fragment that exhibited full Rab1 de-AMPylation activity in vitro ( Figure 1B ) . The failure to rescue cell rounding was not due to the absence or instability of the truncated SidD variants ( Figure S1 ) . Rather , we noticed a difference in the intracellular localization pattern of some SidD fragments compared to that of full length GFP-SidD which , as we reported earlier , colocalizes with marker proteins of the Golgi and trans-Golgi network [18] . Upon closer examination , we found that only SidD variants containing the C-terminal 185 residues ( amino acid 322 to 507 ) displayed colocalization with the Golgi marker giantin similar to that of full length SidD ( Figure 1C ) . None of the N-terminal fragments were enriched at the Golgi but instead showed a predominantly cytosolic distribution pattern . Thus , the C-terminal region spanning amino acid 322–507 possessed the ability to target SidD to the Golgi by interacting with a yet unknown factor on this compartment , and failure of GFP-SidD1–379 to properly localize to the correct target organelle may explain the inability of this catalytically active SidD fragment to rescue SidM-mediated cell rounding ( Figure 1A ) . Based on these results we divided SidD into two functional regions: an N-terminal domain with de-AMPylation activity ( aa 1–379 ) and a C-terminal targeting region ( aa 322–507 ) . To further investigate the molecular basis for Rab1 recognition and de-AMPylation by SidD we initiated its structural characterization by X-ray crystallography . We identified a proteolytically resistant N-terminal domain ( residues 37–350; SidD-NT ) that fell within the domain borders of the largest catalytically active domain discovered above ( Figure 1 ) and that crystallized readily . The structure of SidD-NT assumed an α/β fold formed by two stacked six-stranded antiparallel β-sheets flanked by α-helices ( Figure 2A ) . Pairwise alignment using the DALI server [21] revealed a notable resemblance to metal-dependent protein phosphatases ( PPMs ) , including human PP2Cα [22] and the bacterial PstP [23] which are considered the defining members of this family ( Figure S2A ) . Despite the overall similarity to PPMs , SidD-NT exhibits several major structural differences ( Figure 2A , B ) . First , SidD-NT contains two extra β strands at the N-terminus ( β1 and β2 ) that contribute to extend the central β-sandwich as compared to the classical β-sandwich of the PstP bacterial phosphatase . A second difference resides within the region equivalent to the flap subdomain of PPMs . In most prokaryotic enzymes , the flap subdomain consists of a loop and two helical stretches connecting the last two strands of the central β-sandwich . The length and orientation of the flap region relative to the catalytic pocket is variable between different phosphatases and appears to regulate substrate binding and catalysis [24] . In the case of SidD-NT , the corresponding flap segment ( residues 209–236 ) is completely repositioned by a large hinge bent tangential to the catalytic groove . Furthermore , the N-terminal section of the equivalent flap segment in SidD-NT contains a β strand ( β11 ) that is part of a novel three-stranded antiparallel β-sheet adjacent to the active site . The other two strands ( β14 and β15 ) of this extra β-sheet correspond to an insertion between α6 and β13 . Finally , the third main structural difference corresponds to two additional insertions ( residues 73–78 and 311–325 ) that contribute to a noticeable extension of helix α6 and the formation of a two-stranded β-sheet ( β4 and β16 ) . The extended α6 and the extra β-sheet form a stalk-like protrusion positioned on one side of the catalytic pocket . In summary , the crystallographic structure of SidD-NT assumes a PPM fold with some conformational rearrangements and the presence of additional subdomains , most of them being grouped around the negatively charged active site ( Figure 2C , D ) . The active site of SidD-NT is located in a negatively charged cleft between the central β-sheets and comprises a relatively well-preserved binuclear metal center . The first metal ( M1 ) is coordinated by four water molecules and residue D110 , whereas the second metal ( M2 ) is hexa-coordinated with the classical octahedral geometry formed by four water molecules , D110 , and the main chain carbonyl of G111 ( Figure 3A , B ) . The M1 position is slightly shifted as compared to other PPMs which can be attributed to the incomplete coordination derived from the absence of a highly conserved aspartic acid residue ( Figure 3C ) . In this regard , D192 could accomplish the M1 hexa-coordination but the extended distance would require a conformational closure of the catalytic site . Interestingly , the conserved aspartic acid residue that is missing in the catalytic site of SidD coordinates a third ion ( M3 ) in most bacterial PPMs ( Figure 3C ) . The absence of this aspartic acid residue in SidD precludes a similar M3 coordination and no additional metal binding site is observed . Thus , in contrast to other bacterial PPM phosphatases , SidD appears to lack the capacity of binding a third ion at the equivalent M3 position . Most PP2C phosphatases require either magnesium ( Mg ) or manganese ( Mn ) ions for their activity , with distinct preferences [25] . Quantitative Mg2+ analysis by inductively coupled plasma-optical emission spectrometry ( ICP-OES ) revealed a stoichiometry of Mg2+ relative to SidD of 1 . 7 to 1 ( data not shown ) suggesting that the active site of SidD contains two Mg2+ ions . Furthermore , mutation of D110A in SidD , which directly coordinates both M1 and M2 in the crystallographic structure , resulted in a dramatic reduction in the amount of Mg2+ to nearly negligible values . In this regard , the result from the quantitative ICP-OES analysis for Mg2+ correlates well with the two ions observed in the catalytic pocket of the SidD-NT structure . The presence of Mg2+ ions within the catalytic pocket of SidD implied an important role of metal ions for the enzyme's activity . Consistent with this , we found that pre-incubation of SidD with the metal chelator ethylenediaminetetraacetic acid ( EDTA ) efficiently interfered with Rab1 de-AMPylation catalyzed by SidD ( Figure 4A ) . Furthermore , the activity of SidD was fully restored by complementing the reaction with MgCl2 but not by adding other divalent ions such as calcium ( Ca2+ ) or copper ( Cu2+ ) ( Figure 4B ) . A partial recovery of the de-AMPylation activity of EDTA-treated SidD was achieved when the reaction was supplemented with Mn ions ( MnCl2 ) . Together , these results indicate a strong preference of SidD for Mg2+ over other divalent ions which is further supported by the observation that SidD regained its maximum de-AMPylation activity at concentrations of 0 . 8–1 . 0 mM Mg2+ ( Figure 4C ) which correspond well with the physiological level of free Mg2+ [26] . Another notable difference between the catalytic site of SidD and other PPMs is the absence of a highly conserved arginine residue equivalent to R33 in PP2C , R17 in MspP , R20 in PstP and R13 in tPphA ( Figure S2B ) , thought to play an important role for binding and neutralizing the negative charge of the phosphate monoester group during the catalysis [27] . The absence of this arginine side chain in SidD might reflect the difference in electrostatics between monoesterase and diesterase reactions , whereby the greater negative charge on the monoester ( such as phospho-Tyr ) relative to the diester phosphate ( such as AMP-Tyr ) might explain the necessity of an arginine side chain for stabilization . The pH dependency of PP2Cα in the presence of Mg2+ revealed the existence of two ionizable groups with pKa values of 7 . 2 and 8 . 9 [28] . The lower pKa has been interpreted as the binuclear bridging water which , in the form of a hydroxide ion , could attack the phosphorus substrate in a SN2-like mechanism . Using endpoint assays , we examined the pH-dependency of SidD's activity and observed two optimal pH values at ∼7 . 25 and ∼9 . 0 which suggest a ionization dependent catalytic mechanism ( Figure 4D ) . Although the identity and protonation state of the amino acid side chains directly involved in the catalytic activity remains to be determined , the lower apparent pKa value of SidD is comparable to that of PP2C [28] , consistent with a similar binuclear bridging water acting as the reaction-initiating nucleophile . Indeed , D326 in the crystal structure of SidD , like D282 in PP2C , is appropriately positioned to accept the proton from the bridging water when the hydroxide ion is generated ( Figure 3A ) . According to this interpretation , the de-AMPylation reaction performed by SidD involves a hydrolytic cleavage of the adenylyl-O-tyrosyl linkage , whereas the catalysis of the E . coli GS-ATase , the only other known de-AMPylase , utilizes a phosphorolysis mechanism in which a phosphate ion , not a hydroxyl ion , carries out the nucleophilic attack ( Figure S3A , B ) . The structure of the active site revealed the presence of two Mg2+ ions coordinated by D92 , D110 , D326 , the main chain carboxyl of G111 , and several water molecules ( Figure 3B ) . In addition , the nearby residue D192 could potentially fulfill the coordination of M1 . In order to confirm the role of these residues in metal ion coordination , we created SidD mutant proteins in which each of the four aspartate residues was replaced with either alanine or with a similarly charged glutamate ( Figure 5A; Figure S4 ) . When assayed for [α32P]AMP removal in vitro we found that even a conservative substitution of aspartate for glutamate attenuated de-AMPylation activity of the SidD mutants considerably ( D92E , D192E ) or severely ( D110E , D326E ) ( Figure S4 ) . Upon a more drastic substitution of aspartate for alanine no residual activity was detected in three out of four SidD mutants ( D92A , D110A , D326A ) ( Figure 5A ) . The recombinant mutant proteins displayed no detectable change in stability or solubility ( Figure S4 ) , suggesting the absence of major structural disturbances . In fact , we determined the crystallographic structure of SidD ( D110A ) at 1 . 9 Å resolution and confirmed the absence of coordinated Mg2+ ions within the catalytic pocket of the mutant protein without noticing any significant effect on its overall fold ( Figure S3 ) . Next , we validated our in vitro de-AMPylation results in two independent mammalian cell-based assays . First , we analyzed the SidD point mutants for their ability to prevent SidM-induced COS1 cell rounding and cytotoxicity ( Figure 5B ) . As expected , wild type GFP-SidD , which showed full de-AMPylation activity in vitro , prevented SidM-induced cytotoxicity in COS1 cells . In contrast , SidD ( D92A ) and SidD ( D110A ) were not capable of reducing the percentage of rounded cells that simultaneously produced SidM ( Figure 5B ) , consistent with their lack of de-AMPylation activity in vitro ( Figure 5A ) . SidD ( D92E ) which possessed residual de-AMPylation activity in vitro prevented morphological changes in twice as many COS1 cells as GFP alone ( 20% vs . 10% , respectively ) . Notably , the failure of SidD mutants to efficiently rescue cell rounding was not due to their inability to target to the Golgi compartment ( Figure S4 ) . In a second in vivo approach , we determined the effect of aspartate substitutions on the ability of SidD to catalyze de-AMPylation and , thus , removal of Rab1 from LCVs during the infection process ( Figure 5C ) . Consistent with earlier reports [18] , [19] , L . pneumophila mutants lacking sidD showed a significantly prolonged colocalization with host cell Rab1 four hours post infection compared to LCVs containing wild type bacteria ( 36% vs 11% Rab1-positive vacuoles ) , in agreement with the failure of a sidD deletion strain to de-AMPylate Rab1 and to initiate Rab1 inactivation and removal from the LCV membrane by Rab1GAPs and GDI , respectively . The Rab1 removal defect of an L . pneumophila ΔsidD mutant was fully complemented by plasmid-encoded SidD but not by the catalytically inactive protein SidD ( D92A ) . Remarkably , complementation with plasmid-encoded SidD ( D92E ) fully rescued the phenotype of a ΔsidD mutant , a phenomenon most likely attributable to the residual activity of this enzyme ( Figure S4 ) which may have been further amplified by its overproduction from the high-copy plasmid within L . pneumophila . Taken together , our mutational analysis confirmed that the four aspartate residues at position 92 , 110 , 192 , and 326 are crucial for SidD function both in vivo and in vitro ( Figure 5 ) most likely by properly positioning the two catalytically essential Mg2+ ions inside the active site . Despite significant efforts we were unsuccessful in obtaining crystals of the complex between SidD and either AMPylated and non-AMPylated Rab1 . Furthermore , any attempts to crystallize SidD or the catalytically inactive mutant SidD ( D92A ) in the presence of AMP analogues such adenosine , adenosine 5′- monophosphate , 5′- ( 4-Fluorosulfonylbenzoyl ) adenosine hydrochloride , adenosine 5′- ( α , β-methylene ) diphosphate , S- ( 5′-Adenosyl ) -L-homocysteine , or 5′-Tosyladenosine were unsuccessful . Thus , to explore the interaction between both proteins we performed an energy-based rigid-body docking experiment with unmodified Rab1 . By using the crystal structure of cacodylate bound to the MspP phosphatase [29] as initial constraint , we found that the docking solution with the Tyr77 hydroxyl O atom closest to the Mg ions ( 5 . 4 Å ) was able to accommodate the AMP moiety in the same crystallized conformation without steric clashes . We then applied molecular dynamics ( MD ) in order to refine the SidD-Rab1 ( AMP ) docking model as well as to evaluate its stability . In this regard , the initial docking showed only small fluctuations along the MD simulation indicating a stable SidD-Rab1 interaction ( Figure S5A ) . Similarly , the Mg2+-phosphate interaction at the catalytic site remained constant during the MD simulation ( Figure S5B ) . These results further attested a good structural complementarity between SidD and AMPylated Rab1 with a buried surface of approximately 1 , 300 Å2 and unrestricted access to the catalytic pocket without the need of large conformational rearrangements ( Figure 6A , B ) . Next , we used in silico alanine scanning on the interfacial residues of SidD to predict relevant hotspots for Rab1 recognition . Interestingly , the residues with higher contribution to the binding free energy are grouped asymmetrically around the catalytic pocket ( Figure 6C , and Table S2 in Text S1 ) . Indeed , the average structure from the last nanosecond of the MD shows that F112SidD and Y113SidD form extensive hydrophobic interactions with Y77Rab1 ( Figure S5C ) . Another participating residue is K217SidD which is facing the phosphate group of AMP and may function as proton donor for the leaving phosphate . More peripherally , Y223SidD contributes to the hydrophobic burial of Y109Rab1 . Other residues such as E168SidD and D221SidD form hydrogen bonds with R79Rab1 ( Figure S5C ) . Finally , the docking model shows that the AMP moiety is accommodated in a groove adjacent to the catalytic pocket of SidD without being detached from Rab1 ( Figure 6A ) . The adenine base of AMP rests against F74SidD and K88SidD and lacks additional specific interactions whereas the ribose hydroxyl groups interact with R323SidD . To validate the binding hotspot found in SidD , we mutated several residues that contribute to the interaction with Rab1 and examined the effect on the ability of SidD to remove [α32P]AMP from Rab1 in vitro . In agreement with the interactions described above , the single residue substitutions F112A , Y113A , K217A , Y223A , and Y253E as well as the double exchange F74A/K88A strongly affected the ability of SidD to de-AMPylate Rab1 ( Figure 6D ) without compromising the overall protein stability or solubility ( Figure S5 ) . Only the R323A mutant , designed to disrupt the interaction with the ribose of AMP , had no apparent effect on SidD activity , which may reflect a redundant interaction as a consequence of nearby contacts . Notably , while replacement of Y113 with glutamine strongly reduced Rab1 de-AMPylation by SidD , substitution with the structurally similar phenylalanine had no apparent effect on activity ( Figure S4 ) , consistent with phenylalanine but not glutamine being capable of mediating π stacking interactions with Y77Rab1 . We also examined additional mutations outside the binding hotspot such as I321S , D271A , or H87A and even the triple mutant T261A/E264A/R281A and , as expected , observed no obvious reduction in SidD activity in vitro ( Figure 6C , D ) or in vivo ( H87A; Figure 5B ) which further validated the SidD-Rab1 docking model . Overall , the experimental results are in remarkable agreement with the model complex , with the majority of the mutations at the binding hotspot severely attenuating or preventing SidD-mediated de-AMPylation of Rab1 . The structure of SidD displays a stalk like protrusion on one side of the active site cleft and a binding hotspot on the other side , features that may contribute to recognizing and properly orienting Rab1 in a way that the AMPylated Y77Rab1 is correctly positioned inside the catalytic pocket . We speculated that this topological design might allow SidD to distinguish AMPylated Rab1 from similarly modified substrates such as Rho GTPases . To validate our hypothesis , we generated [32P]AMP-labeled Cdc42 by incubating it with either V . parahaemolyticus VopS , which AMPylates Cdc42 at Y32 , or with H . somni IbpA , which AMPylates the neighboring T35 ( Figure 7A ) , and tested the ability of SidD to convert Cdc42 back into the unmodified form . In contrast to [32P]AMP-Y77Rab1 which was efficiently de-AMPylated by SidD , neither [32P]AMP-T35Cdc42 nor [32P]AMP-Y32Cdc42 showed any detectable decrease in the AMPylation level in the presence or absence of SidD ( Figure 7B ) . Thus , SidD did not accept Cdc42 as substrate for de-AMPylation in vitro even if the AMP modification in Cdc42 was located on a tyrosine residue ( Y32 ) , as it is the case in AMPylated Rab1 ( Y77 ) . Similar results were obtained in an in vivo assay where SidD failed to prevent rounding of COS1 cells transiently producing GFP-tagged VopS ( Figure S6 ) confirming that Rho GTPases AMPylated by VopS cannot be de-AMPylated by SidD . Finally , we determined if SidD could remove posttranslational modifications other than AMP from Rab1 . The L . pneumophila effector AnkX/LegA8 covalently attaches phosphocholine to serine-76 in Rab1 ( S76Rab1 ) , the residue located immediately adjacent to Y77Rab1 , the target of AMPylation by SidM [30] . Like AMP , phosphocholine is connected to Rab1 via a phosphodiester bond . Its removal requires the L . pneumophila effector Lem3 [31] , [32] , [33] which we predict assumes a PP2C-like fold similar to SidD ( data not shown ) . Given the similarity of the two removing enzymes and of the chemical bond they hydrolyze we explored whether SidD or Lem3 are capable of catalyzing the other enzyme's reaction . While AMP was efficiently removed by SidD and phosphocholine by Lem3 , neither modification was affected by the presence of the opposite enzyme ( Figure 7C ) . Together , these data favor the idea that SidD from L . pneumophila ( and probably Lem3 as well ) has evolved to exclusively recognize its host cell substrate and to remove only a particular post-translational modification from a specific side chain location . To our knowledge , SidD from L . pneumophila is the first known microbial effector protein with de-AMPylation activity . Together with the AMPylase SidM it forms an enzymatic cascade that enables the pathogen to post-translationally modify host cell Rab1 in a transient rather than permanent manner . Despite limited sequence homology , the crystal structure of the de-AMPylation domain of SidD revealed a notable similarity to Serine/Threonine phosphatases of the PPM family . However , in addition to the conserved PPM core , SidD-NT exhibits additional structural elements like a repositioned flap domain as part of a new three-stranded antiparallel β-sheet and a stalk-like protrusion , both derived from sequence insertions located around the catalytic site , thus with potential regulatory functions ( Figure 2A , B ) . The finding that SidD is a PPM phosphatase with de-AMPylation activity constitutes a clear example of how L . pneumophila has adapted a common enzymatic fold and mechanism to effectively hydrolyze an unusual substrate . In contrast , AT-N from the E . coli GS-ATase assumes a nucleotidyl transferase fold , indicating that de-AMPylases have developed more than once during microbial evolution . From a chemical perspective , a common feature of enzymes that hydrolyze phosphate monoesters and diesters is the presence of a binuclear metal center . PPM phosphatases share an invariant M1 and M2 whereas the presence of an additional M3 in bacterial homologs is associated with a small flap subdomain adjacent to the catalytic site . The role of this M3 is still unclear , although it has been proposed to modulate the flap orientation and , thus , substrate binding [34] . More recently , the M3 has been associated with the activation of a water molecule that might function as a proton donor for the leaving phosphate [35] . The crystal structure of SidD-NT shows the absence of an absolutely conserved aspartate in the catalytic site that in other PPMs coordinates M1 and M3 . This absence not only produces a slight shift in the M1 position but also compromises the coordination of a third ion . Indeed , the quantitative ICP-OES analysis of SidD together with the ion-dependent enzymatic activity assay ( Figure 4 ) support the presence of two Mg2+ ions that are essential for Rab1 de-AMPylation . Collectively , the absence of an M3 , the M1 shifted position , the strict requirement of Mg2+ ions for catalysis , and the absence of an arginine side chain to interact with the phosphate group appear to be variations through which L . pneumophila SidD has been converted into an enzyme that de-AMPylates Rab1 , capitalizing on the existing PPM active site . The crystal structure of cacodylate bound to the MspP phosphatase shows a direct interaction with M1 and M2 by bidentate coordination which has been interpreted as a mimicking phospho-substrate intermediate during the catalysis [29] . By using this metal-phosphate coordination as initial constraint in docking AMPylated Rab1 into the catalytic pocket of SidD , we found a remarkable surface complementarity ( Figure 6 ) . Subsequent analysis of the docking model by molecular dynamic simulations showed that both the root mean square deviation ( RMSD ) of the complex with respect to the initial model as well as the distance between the two Mg2+ ions and the phosphate group of AMP experienced only small fluctuations during the simulation process ( Figure S5 ) . These observations not only confirm the stability of the docking prediction but , more importantly , evidence a good structural complementarity between SidD and AMPylated Rab1 without the need for large conformational rearrangements to access the catalytic pocket . It should be noted that although our docking model is energetically favorable , the AMP-Tyr side-chain could adopt alternative conformations relative to the crystallized AMPylated Rab1 and that the actual protein complex may experience additional structural rearrangements beyond what has been sampled in our simulations . We also analyzed the interface features of SidD that enable the initial recognition of AMPylated Rab1 . Using computational alanine scanning , we identified a hotspot in which the binding energy is largely concentrated on a few amino acids near the catalytic pocket . Indeed , the majority of individual mutations introduced at the binding hotspot severely attenuated or prevented the catalytic activity of SidD in vitro ( Figure 6D ) , which is in remarkable agreement with the qualitative description of the SidD-Rab1 interaction derived from our docking model . Our structural , computational and mutational analysis revealed the existence of distinctive features in SidD such as the binding hotspot flanking the catalytic site or the stalk-like protrusion that appear to be absent from generic phosphatases . We hypothesized that through this topological design SidD can distinguish AMPylated Rab1 from similarly modified substrates . Accordingly , we demonstrated that AMPylated Rho GTPases were not recognized by SidD under any of the conditions where Rab1 was efficiently de-AMPylated ( Figure 7B , Figure S6 ) . Likewise , we found that phosphocholinated Rab1 did not serve as substrate for SidD ( Figure 7C ) even though its post-translational modification was comparable to AMPylated Tyr77Rab1 with respect to its location ( Ser76Rab1 ) and chemical linkage ( phosphodiester bond ) . The fact that the activity of the de-phosphocholinase Lem3 was similarly restricted from targeting AMPylated Rab1 ( Figure 7 ) suggests that these L . pneumophila effectors have adapted their catalytic activity towards their correct host target thought the acquisition of specific topological determinants . Based on our domain mapping and cellular localization studies ( Figure 1 ) we predict the existence of a second functional region in SidD that assists in localizing the protein to membranes , more precisely the LCV membrane within L . pneumophila-infected cells or the Golgi compartment within transiently transfected cells . The exact mechanism of membrane targeting of SidD , however , remains unclear . Several L . pneumophila effectors have been shown to specifically interact with phospholipids such as PI ( 3 ) P or PI ( 4 ) P in order to associate either with the LCV membrane or with other host cell compartments [36] . Indeed , targeting of effectors to a specific cellular compartment constitutes an additional mechanism to enhance substrate specificity . Using protein-lipid overlay assays we were unable to detect binding of SidD to any of the most common phosphoinositides ( data not shown ) , suggesting that membrane targeting of the C-terminal region is mediated by binding to another lipid or proteinaceous host factor . Several attempts to demonstrate a stable association between purified recombinant SidD and Rab1 by pulldown studies failed , indicating that the interaction between both proteins is of transient nature . Nonetheless , it is likely that even a weak interaction with Rab1 is sufficient to retain the majority of SidD molecules in close proximity to the LCV membrane after their translocation by the L . pneumophila T4SS . It is worth mentioning that the prenylation anchor of Rab1 and the C-terminal targeting region of SidD are located on the same side of the complex thus with the potential to simultaneously contact the LCV membrane during catalysis , further strengthening the likelihood of our modeled complex ( Figure S7 ) . Future studies should help to reveal the mechanistic details of SidD membrane targeting and substrate detection during host cell infection . In summary , the study presented here provides an important first look at the structure and catalytic mechanism of SidD and reveals that this L . pneumophila effector differs in many aspects from the E . coli GS-ATase , the only other known de-AMPylase . The finding that SidD is a converted phosphatase equipped with structural elements designed to distinguish AMPylated Rab1 from similar host cell substrates demonstrates the versatility of the phosphatase fold and suggests that it may have served as blueprint for a variety of thus far uncharacterized de-modifying enzymes capable of targeting an array of different post-translational modifications . Native SidD37–350 was concentrated to 8 mg/ml and used for initial crystal screening . All crystallization conditions were carried out in a sitting drop setup of 0 . 1 µL protein solution mixed with 0 . 1 µL of mother liquor . Visible crystals appeared in several comparable conditions after 5 days at 18°C . Further optimization yielded good quality crystals in 1 . 6–1 . 8 M NaCl , 0 . 1 M NaOAc pH 4 . 8 , 20% glycerol using 2 µL sitting drops with equal protein/mother-liquor ratio . The D110A mutant crystallized under the same conditions as the native SidD37–350 . The structure of SidD37–350 was solved by single anomalous dispersion with isomorphous replacement ( SIRAS ) using a single gadolinium derivative [37] . Gd positions were determined using the SHELX software [38] . The initial electron density map was then calculated with experimental phases derived from the Gd positions with phenix . phaser [39] . A preliminary model was automatically traced by phenix . autobuild and completed by hand in Coot [40] . The model was improved through alternating cycles of manual rebuilding using Coot and refinement using phenix . refine . A final refinement cycle was performed with REFMAC5 [41] , [42] . This model was subsequently used for molecular replacement with the high-resolution diffraction data using phenix . phaser . Additional model building and refinement was performed using Coot , phenix . refine and REFMAC5 . The final models have good structural geometries with no residues in disallowed regions of the Ramachandran plot . Statistics on data collection and refinement are provided in Table S1 in Text S1 . All the molecular representations were prepared with PyMOL ( The PyMOL Molecular Graphics System , Schrödinger , LLC ) and ChemDraw ( PerkinElmer ) . We built a structural model of Rab1/SidD complex by rigid-body docking , based on the FFT-based docking program Zdock2 . 1 [43] and the energy-based pyDock scoring scheme [44] . Details of this procedure are described in Supplemental Materials and Methods . In order to refine the docking model of the Rab1 ( AMP ) /SidD complex we performed molecular dynamics ( MD ) simulation in explicit solvent using the force field AMBER parm99 of the AMBER10 package [45] , [46] . Details of this procedure are described in Supplemental Materials and Methods . We performed in silico alanine scanning on the MD refinement of the selected docking model to identify the key residues responsible for the binding process . The MMPBSA . py script in AMBER12 [47] was used to carry out all binding energy calculation using the MM-GBSA method on 200 snapshots extracted from the last 2 ns of the MD trajectory of the selected docking model . Each SidD interface residue was mutated to alanine and then we estimated the binding free energy change ( ΔΔG ) as the difference between the binding ΔG of the wild type and the mutated complex ( van der Waals and electrostatic energy by the MM force field , electrostatic contribution to the solvation free energy by GB method , and nonpolar contribution to the solvation free energy by an empirical model ) . The conformational entropy contribution to binding was not included here , given the difficulty of computing it for a large protein-protein complex , and the small effect when calculating relative system free energies . Rab1a ( 25 µM ) was phosphocholinated at room temperature for 4 h in the presence of His-AnkX ( 0 . 25 µM ) in a buffer containing 20 mM HEPES pH 7 . 5 , 100 mM NaCl , 1 mM CDP-choline , 1 mM MgCl2 , and 1 mM ATP . The reaction mixture was then incubated with 60 µl of HisLink beads ( Promega ) to remove His-AnkX before purification by gel filtration on a HiLoad 16/60 Superdex 75 pg column ( GE Healthcare ) at 4°C . Fractions containing phosphocholinated Rab1a ( Rab1-PC ) in either PBS-MM were pooled , concentrated , and stored at −80°C . For de-phosphocholination , Rab1-PC ( 10 µM ) was incubated for 2 h at room temperature with increasing amounts of the purified Lem3 or SidD in PBS-MM . Immunoblot analysis was used as described above for the de-AMPylation assays using the anti-phosphocholine-specific antibody TEPC-15 ( Sigma ) to detect phosphocholinated Rab1a . Immunofluorescence microscopy was performed as previously described [18] . The structural coordinates of SidD37–350 and SidD37–350 ( D110A ) have been deposited in the Protein Data Bank ( http://www . rcsb . org . pdb ) with the accession codes 4IIK and 4IIP respectively .	The covalent attachment of adenosine monophosphate ( AMP ) to proteins , a process called AMPylation ( adenylylation ) , has recently emerged as a novel theme in microbial pathogenesis . While AMPylases from various pathogenic microorganisms have recently been characterized , the only virulence protein with de-AMPylation activity known to date is the Legionella pneumophila effector SidD which catalyzes AMP removal from the host GTPase Rab1 . Thus , both AMPylation and de-AMPylation constitute a novel catalytic mechanism to precisely control the function and membrane dynamics of a host Rab GTPase . In spite of this pivotal role , the molecular mechanism of AMP removal and the structural determinants for Rab1 recognition by SidD have remained largely unexplored . Here , we present the crystal structure of the de-adenylylation domain of SidD and reveal the catalytic mechanism of Rab1 de-adenylylation . Surprisingly , the structure of SidD is not related to the other known enzyme with de-AMPylation activity , the Escherichia coli GS-ATase . Instead , the catalitic domain of SidD is remarkably similar to that of the metal-dependent protein phosphatases ( PPMs ) , however with distinctive structural features to distinguish AMPylated Rab1 from similarly modified substrates . Importantly , we provide a model for the SidD-Rab1 complex which sheds light into the specific details of substrate recognition and catalysis by this virulence factor .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "medicine", "cellular", "structures", "subcellular", "organelles", "enzymes", "microbiology", "bacterial", "diseases", "infectious", "diseases", "signaling", "in", "cellular", "processes", "legionella", "microbial", "pathogens", "membranes", "and", "sorting", "ras", "sign...	2013	Structural Basis for Rab1 De-AMPylation by the Legionella pneumophila Effector SidD
Singapore experiences endemic dengue , with 2013 being the largest outbreak year known to date , culminating in 22 , 170 cases . Given the limited resources available , and that vector control is the key approach for prevention in Singapore , it is important that public health professionals know where resources should be invested in . This study aims to stratify the spatial risk of dengue transmission in Singapore for effective deployment of resources . Random Forest was used to predict the risk rank of dengue transmission in 1km2 grids , with dengue , population , entomological and environmental data . The predicted risk ranks are categorized and mapped to four color-coded risk groups for easy operation application . The risk maps were evaluated with dengue case and cluster data . Risk maps produced by Random Forest have high accuracy . More than 80% of the observed risk ranks fell within the 80% prediction interval . The observed and predicted risk ranks were highly correlated ( ρ ≥0 . 86 , P <0 . 01 ) . Furthermore , the predicted risk levels were in excellent agreement with case density , a weighted Kappa coefficient of more than 0 . 80 ( P <0 . 01 ) . Close to 90% of the dengue clusters occur in high risk areas , and the odds of cluster forming in high risk areas were higher than in low risk areas . This study demonstrates the potential of Random Forest and its strong predictive capability in stratifying the spatial risk of dengue transmission in Singapore . Dengue risk map produced using Random Forest has high accuracy , and is a good surveillance tool to guide vector control operations . Dengue is a viral infection caused by one of the four closely related yet antigenically distinct virus serotypes ( DENV-1 , DENV-2 , DENV-3 and DENV-4 ) , and transmitted by Aedes mosquitoes , primarily the Ae . aegypti and Ae . albopictus [1 , 2] . Infection confers lifelong immunity to the infecting serotype [3] . However , it increases risk for dengue haemorrhagic fever ( DHF ) and dengue shock syndrome ( DSS ) , a deadly form that present with severe complications , in subsequent infections [4] . Since the publication of the GBD 2010 , it was estimated that 390 million dengue infections occur each year globally , of which 500 , 000 develop into DHF [5 , 6] . Dengue poses a substantial public health threat globally , especially throughout the tropical and subtropical regions [7 , 8] . Located one and a half degrees north of the equator and lying in the dengue belt , Singapore is prone to dengue transmission , with all four dengue serotypes co-circulating and frequent introduction of new genotype virus [9] . Though intensive vector control efforts have successfully suppressed the Aedes population , from an Aedes house index of over 50% in the 1960’s to the present 1–2% , Singapore remains susceptible to dengue outbreaks [10–12] . The increased in human population density and the low herd immunity resulting from sustained period of low dengue transmission are factors that may have contributed to the resurgence of dengue in Singapore [13 , 14] . A significant amount of funding and resources has been allocated for dengue every year [15] . The estimated economic and disease burden of dengue were 9–14 disability-adjusted life years ( DALYs ) per 100 , 000 population and US$41 . 5 million per annum [16] . A dengue temporal model was developed in 2013 by the Environmental Health Institute , a research institute of the Singapore’s National Environment Agency ( NEA ) in collaboration with the National University of Singapore ( NUS ) to aid vector control measures . The model predicts trends and incidence up to 12 weeks ahead , providing early warnings of outbreak and facilitating public health response to moderate impending outbreak [17] . This model was able to accurately project an upward trend of dengue cases in 2013 and 2014 , predicting the two major outbreaks [18] . NEA has been using the model in planning vector control and public communication [19] . However , a limitation of the model is the missing spatial resolution as it does not highlight areas with high risk of dengue transmission . Given that NEA’s key strategy in dengue control is preventive surveillance and larval source reduction , a labour-intensive activity that requires effective deployment of a limited pool of skilled vector control officers , spatial risk profiling of dengue transmission is thus necessary for effective deployment of resources , and achieving maximum impact . In this paper , we describe a new approach for spatial risk stratification of dengue transmission in Singapore . Using Random Forest , we quantify the risk of dengue transmission in different areas and categorize them into different risk groups to guide the pre-emptive source reduction exercise conducted by NEA vector control officers . Predictive performance of the model is evaluated with both dengue cases and clusters . Proposed by Leo Breiman , Random Forest is an ensemble machine learning method that uses an ensemble of decision trees [20] . In Random Forest , several ( N = 1000 ) bootstrap samples are drawn from the training set data , and an unpruned decision tree fn ( x ) , is fitted to each bootstrap sample . At each node of the decision tree , variable selection is carried out on a small random subset of the predictor variables , so as to avoid the “small n large p” problem . The best split on these predictors is used to split the node . The predicted response is obtained by averaging the predictions of all trees , i . e . 1N∑n = 1Nfn ( x ) ( Fig 1 ) . Random Forest was used to predict the percentile rank of dengue case count in 1km2 grids , with past dengue exposure ( total number of cases in previous year , total number of cases in neighbouring grids in previous year and number of non-resident cases in previous year ) , human population ( estimated population density ) , vector population ( estimated ratio of Aedes aegypti mosquitoes out of all Aedes moquitoes—breeding percentage ) and environmental data ( vegetation index , connectivity index and ratio of residential area ) . The predicted percentile ranks are then categorized and mapped to four color-coded risk groups ( RG1-4 , lowest risk as RG1 and highest risk of dengue transmission as RG4 ) for easy operation application . Although administrative boundaries are more compatible with ground operation , 1km2 grids were used as study units as they are more consistent in area size and do not change over time . We use residential grids exclusively for the analysis and risk mapping . Random Forest analyses were performed using the randomForest package implemented in the R statistical language [21] . Data from 2006 to 2013 were used to parameterize the model , and performance of the model is evaluated with new dengue case data from 2014 to 2016 . Apart from visually comparing the risk map and distribution of dengue cases , we applied the following quantitative metrics to evaluate the model: 1 . correlation between predicted and observed percentile ranks , 2 . coverage of prediction intervals , 3 . summary statistics of the number of cases within each risk group , and 4 . weighted ( square ) Kappa agreement coefficients of risk grouping . In addition to using dengue case data , data on dengue cluster , which indicates possible transmission within the locality , were considered for model evaluation as well . We investigated the odds of clusters forming in high ( RG 3 and 4 ) and low ( RG 1 and 2 ) risk areas , and examined if transmission intensity , comprising of cluster’s growth rate , transmission duration and cluster size differ between high and low risk areas . Differences were analysed using Kruskal-Wallis tests . Table 1 shows the various risk factors considered for the risk mapping . The risk factors were identified from literature review and examined with historical data [11 , 22 , 23] . All data ( Dengue , Population and Entomological ) were aggregated to the 1km2 grids . The time period used for all variables was January 2006 to December 2016 , and their sources are: Associations between covariates and dengue burden were examined through partial dependence plot . Consistent with our prior knowledge , all covariates are associated with dengue burden , as contrasted by the flat line partial effect of random noise ( Fig 2 ) . Among the covariates , the number of residential units , dengue burden in previous year and the breeding percentage in previous year are top-ranked in terms of variable importance ( Fig 3 ) , and impose a larger influence on model accuracy , relative to the other covariates . This , therefore , suggests that population density , dengue burden and abundance of Ae . aegypti are significant risk factors for dengue transmission . The predicted percentile ranks were categorized and mapped to four color-coded risk groups based on the three quartiles so that the number of grids in each risk group is approximately the same . The distribution of risk groups is comparable in all three years , with high risk groups ( RG 3 and 4 ) congregating in the eastern part of Singapore . When dengue cases were overlaid onto the risk maps , we observed good agreement between the cases and risk groups ( Fig 4 ) . Majority of the cases fell in risk group 3 and 4 . There was strong positive correlation between the observed and predicted risk ranks , a correlation of 0 . 86 ( P <0 . 01 ) , 0 . 87 ( P <0 . 01 ) and 0 . 88 ( P <0 . 01 ) in 2014 , 2015 and 2016 respectively . In addition , the risk level commensurate with case density . The predicted risk levels were in excellent agreement with the case density , a weighted Kappa coefficient of 0 . 814 ( P <0 . 01 ) in 2014 , 0 . 839 ( P <0 . 01 ) in 2015 and 0 . 821 ( P <0 . 01 ) in 2016 . This is further supported by the increasing trend of dengue case count from risk group 1 to 4 ( Table 2 ) . Fig 5 shows the predicted percentile ranks and its 80% prediction interval . 82% and 83% of the observed percentile ranks fell within the 80% prediction interval in 2014 and 2015 respectively . In 2016 , 81% of the observed percentile ranks fell within the 80% prediction interval . Overall , cases in 2015 have slightly better agreement with the risk map than in 2014 and 2016 . Evaluation of risk maps with 2014 to 2016 clusters data shows that the number of dengue clusters in high risk areas was almost 8 times the low risk areas ( Fig 6 ) . Each year , close to 90% of the dengue clusters were found in high risk areas , which represent 22% of Singapore land area and 50% of residential areas . The odds of cluster forming in high risk areas was higher than in low risk areas for all three years . The odds ratios were 11 . 1 ( P <0 . 01 ) , 14 . 6 ( P <0 . 01 ) and 12 . 1 ( P <0 . 01 ) for 2014 , 2015 and 2016 respectively . Clusters were further stratified by the number of serotypes into single serotype and multiple serotypes clusters . High risk areas have a larger proportion of multiple serotypes clusters than low risk areas , and interestingly , 3-serotypes clusters were only present in high risk areas , especially in RG4 ( Fig 6 ) . Transmission intensity , comprising of cluster’s growth rate , transmission duration and cluster size were significantly different between single serotype and multiple serotypes clusters ( P <0 . 01 ) . Clusters with more serotypes present have a faster growth rate , longer transmission duration and larger cluster size ( Table 3 ) . The same characteristics were seen when we grouped the clusters by high and low risk areas . Though there were less clusters in low risk areas , the transmission intensity of clusters in these low risk areas was of no significant difference ( P >0 . 1 ) when compared with those in high risk areas ( Table 3 ) . This study demonstrates the potential of Random Forest and its strong predictive capability in stratifying the spatial risk of dengue transmission in Singapore . Dengue risk map produced using Random Forest has high accuracy , and is a good tool to guide vector control operations , allowing targeted preventive measures before and in times of dengue outbreak . Valuable resources can then be deployed in a strategic manner , mitigating the spread of dengue transmission .	Dengue fever , the most prevalent mosquito-borne viral disease today , is caused by Dengue virus ( DENV ) and transmitted to human by Aedes mosquitoes , primarily the Ae . aegypti and Ae . albopictus . The key approach to mitigating dengue transmission is to control the Aedes population , and this often involve vector control strategies such as larval source reduction and preventive surveillance that are labour-intensive and require effective deployment of valuable resources . Spatial risk profiling of dengue transmission is therefore necessary to ensure the optimal utilization of limited resources , and achieving maximum impact of dengue vector control . Here , we developed a dengue risk map by stratifying the spatial risk of dengue transmission in Singapore . Random Forest was used to predict the risk rank of dengue transmission in 1km2 grids , and the predicted risk ranks are then categorized and mapped to color-coded risk groups . The dengue risk map is a good surveillance tool to guide vector control operations . Valuable resources can be deployed in a strategic manner , mitigating the spread of dengue transmission .	[ "Abstract", "Introduction", "Materials", "and", "methods", "Results", "Discussion" ]	[ "invertebrates", "medicine", "and", "health", "sciences", "geographical", "locations", "tropical", "diseases", "animals", "geoinformatics", "singapore", "artificial", "intelligence", "neglected", "tropical", "diseases", "infectious", "disease", "control", "insect", "vectors...	2018	Mapping dengue risk in Singapore using Random Forest
Chagas disease , caused by the unicellular parasite Trypanosoma cruzi , claims 50 , 000 lives annually and is the leading cause of infectious myocarditis in the world . As current antichagastic therapies like nifurtimox and benznidazole are highly toxic , ineffective at parasite eradication , and subject to increasing resistance , novel therapeutics are urgently needed . Cruzain , the major cysteine protease of Trypanosoma cruzi , is one attractive drug target . In the current work , molecular dynamics simulations and a sequence alignment of a non-redundant , unbiased set of peptidase C1 family members are used to identify uncharacterized cruzain binding sites . The two sites identified may serve as targets for future pharmacological intervention . American trypanosomiasis , also known as Chagas disease , is endemic to Central and South America , where 90 to 100 million people are at risk of infection [1] , 10 to 20 million people are infected [1] , [2] , and 50 , 000 die annually [3] . The disease is caused by the unicellular parasite Trypanosoma cruzi ( T . cruzi ) , an organism transmitted by insects of the Reduviidae family . After drawing a blood meal from its human host , the insect reflexively releases feces containing the parasite into the resulting wound [4] . Once blood borne , the parasites infiltrate host cells and replicate . Following replication and maturation , host cells burst open , releasing new T . cruzi parasites into the bloodstream [5] . The acute phase of the disease , which typically persists for two months and has a fatality rate of 2 to 8% , is characterized by the mononuclear inflammation and necrosis of parasitized cells , especially in the heart [6] . The chronic stage of the disease is characterized by dilated cardiomyopathy; indeed , American trypanosomiasis is the leading cause of infectious myocarditis in the world [7] . New therapies for Chagas disease are urgently needed . Current treatments , nitrofurans like nifurtimox and benznidazole , are highly toxic [6] , [8] , [9] , and drug resistance has been reported [10] . Furthermore , one recent study demonstrated that these compounds neither eradicate the parasite nor prevent cardiomyopathy over the long term [11] . The major cysteine protease of T . cruzi , called cruzain or , alternatively , cruzipain , is one attractive drug target [12] . A member of the peptidase C1 protein family , cruzain is present and essential in all stages of T . cruzi development [2] , [13] . Over-expression of cruzain enhances the transformation of the parasite into the infective form [14] , and reduced protease activity prevents infection in wild-type mice [9] . Additionally , cysteine protease inhibitors block both the replication and the differentiation of the parasite in vitro and in vivo [12] , [15]–[22] . Cruzain inhibitors can cure infection in cell , mouse , and dog models [18] , [23] . The future rational design of improved cruzain inhibitors necessitates a better understanding of the flexibility and conformational changes characteristic of the cruzain active site . Molecular dynamics ( MD ) simulations , in which the forces that act on the atoms of a molecular system are approximated using Newton's laws of motion , can be powerful tools for better understanding protein flexibility and conformational sampling relevant to drug design . For example , one recent MD study of HIV integrase revealed a previously uncharacterized binding trench that was subsequently exploited in the design of Isentress ( raltegravir ) , an HIV drug approved by the FDA in 2007 [24] . Importantly , this trench was not evident in the then available crystal structures; it was only by studying active-site flexibility via MD that the trench was initially identified . Additional novel sites of enzymatic , allosteric , or structural importance can be identified computationally by comparing the sequence of the target protein with evolutionarily related enzymes . Critical protein residues are often conserved across multiple members of the same protein family; once multiple sequences are aligned , conserved patches of protein residues can be easily identified . Additional experimental studies can then characterize the pharmacological significance of these patches . Given the urgent need for novel antichagastic therapeutics , we here use computational methods , including molecular dynamics ( MD ) simulations and a sequence alignment of a non-redundant , unbiased set of peptidase C1 family members , to identify previously uncharacterized binding regions that may serve as sites for future pharmacological intervention . To prepare cruzain for MD simulations , hydrogen atoms were added to a high-resolution cruzain crystal structure ( PDB: 1ME4 ) [25] using PDB2PQR to approximate protein protonation at pH 5 . 5 , the pH of the reservosome where cruzain is located in the epimastigote stage of the parasite [26]–[28] . Protonation states were subsequently verified manually . Hydrogen atoms were added to the bound hydroxymethyl-ketone inhibitor using Discovery Studio ( Accelrys ) . The LEaP module of the AMBER9 suite [29] was used to solvate the system by submerging the protein in a TIP3P water box [30] that extended 10 Å beyond the protein in all directions . All crystallographic water molecules were maintained . Ten sodium cations were added to make the system electrically neutral; additional ions were then added to simulate a more physiological 20 mM NaCl solution . The system was parameterized using the generalized and FF99SB AMBER force fields [31] , [32] . NAMD2 . 7b1 [33] was used for all MD simulations . Periodic boundary conditions were employed with the particle mesh Ewald method to account for electrostatic effects ( smoothing cutoff: 14 Å ) . Langevin dynamics were applied to maintain the temperature , and a modified Langevin piston Nosé-Hoover thermostat was used to maintain 1 atm pressure . The initial structure was minimized in four distinct steps; hydrogen atoms were first relaxed for 5 , 000 steps; hydrogen atoms , water molecules , and ions were next relaxed for 5 , 000 steps; hydrogen atoms , water molecules , ions , and protein side chains were then relaxed for 10 , 000 steps; and , finally , all atoms were relaxed for 25 , 000 steps . Following minimization , the system was equilibrated with an NPT-ensemble at 310 K using stepwise harmonic-constraint force constants of 4 , 3 , 2 , and 1 kcal/mol/Å2 on the protein backbone . 250 , 000 steps of MD simulation were executed for each force constant ( 1 fs time step ) . Following minimization and equilibration , five distinct 20-ns productive runs were performed ( 107 steps of 2 fs ) with distinct random seeds in order to sample many protein configurations . The RMSD-based gromos clustering algorithm , as implemented in the GROMACS++ computer package ( g_cluster ) , was used to cluster the conformations sampled during the five 20-ns MD simulations [34] . Structures were first extracted from the trajectories every 50 fs , generating 4 , 002 snapshots total . These snapshots were aligned by their Cα atoms and clustered on the 73 residues of the cruzain active site , defined as all residues within 10 Å of the ligand: 18–31 , 50 , 53–54 , 57–72 , 74 , 91 , 93–98 , 115 , 117 , 120 , 136–142 , 144–145 , 158–165 , 181–184 , 203–210 . The gromos clustering algorithm was first described by Daura et . al . [35] . In brief , for each protein conformation in a pool of conformations , the RMSD distance between the atoms of the aforementioned residues and the corresponding atoms of every other protein conformation in the pool ( potential “neighbors” ) is calculated . The conformation with the most neighbors within a user-specified distance cutoff ( “close neighbors” ) is then selected . This conformation , together with its close neighbors , constitutes the first cluster . The protein conformations of the first cluster are then removed from the pool , and the process is repeated with the remaining conformations until none are left . When a cutoff of 0 . 95 Å was used , this procedure produced 24 clusters . The central member of each cluster was considered most representative; the set of all central members is said to constitute an ensemble . To derive beta factors from the motions sampled during the MD simulations , all trajectories were concatenated , and the RMSF of each protein residue was calculated using the AMBER 9 ptraj module [29] . These RMSF values were converted into beta factors by multiplication , where β = RMSF * 8π2/3 . A small-molecule library was prepared from the ligands of the NCI Diversity Set II , a set of freely available , diverse , drug-like molecules . The Schrödinger LigPrep program ( Schrödinger ) was used to assign protonation states at pH 5 . 5 and to identify and generate tautomers and stereoisomers . One ligand could not be processed with LigPrep; instead , hydrogen atoms were added to this ligand and its geometry was optimized using Discovery Studio ( Accelrys ) . The ligands of this small-molecule library were docked into a 1 . 20 Å cruzain crystal structure ( PDB ID: 1ME4; [25] ) . Hydrogen atoms were added using PDB2PQR [27] , [28] at pH 5 . 5 . At this pH , C25 and H159 formed the thiolate/imidazolium pair required for the catalytic mechanism [36] . An initial virtual screen was performed using the CDOCKER docking software ( Accelrys ) with a docking sphere 15 Å in diameter centered on the coordinates of the crystallographic ligand , as that program was able to recapture the crystallographic poses of two known hydroxymethyl-ketone cruzain inhibitors [25] . The CDOCKER-predicted pose of each of the ligands was rescored using the PLP2 scoring function [37] . The best ligands as evaluated by PLP2 were compiled into a new small-molecule ligand library enriched for potential cruzain inhibitors . To account for receptor flexibility , we subsequently used the relaxed-complex scheme [38] , a protocol that has been used previously to identify inhibitors of FKBP [39] , HIV integrase [24] , and T . brucei RNA editing ligase 1 [40] . The compounds of the enriched small-molecule library were docked into the 24 members of ensemble , again using CDOCKER ( Accelrys ) . Each of these compounds was rescored with the PLP2 [37] scoring function . For each ligand , a PLP2-based ensemble-average score was calculated according to the following equation: ( 1 ) where is the weighted ensemble-average score , wi is the size of cluster i , and Ei is the best score of the ligand , independent of tautomeric or stereoisomeric form , docked into the centroid of cluster i . Cruzain was compared to other members of the peptidase C1 family . First , the UniProt database [41] was used to identify reviewed members of the peptidase C1 family , as defined by the MEROPS classification [42] , that had structures deposited in the Protein Data Bank [43] . All amenable sequences except those of cruzain were then aligned using ClustalW in the MultiSeq extension of VMD [44]–[46] . A non-redundant set was selected from these aligned peptidase C1 sequences ( sequence QR: 75; GF: 1 . 0 ) . Gaps in the sequences were then removed , and ClustalW was used to align the corresponding sequences to a cruzain crystal structure ( PDB: 1AIM ) that was chosen as a non-redundant structure from the set of all cruzain structures aligned using STAMP [47] . The following sequences were aligned: 1A6R , 1AEC , 1CJL , 1DEU , 1FWO , 1JQP , 1K3B , 1M6D , 1PCI , 1XKG , 2C0Y , 2CB5 , 2FO5 , 2O6X , 2WBF , 3PBH , 7PCK , and 8PCH . Residues were colored by similarity according to the BLOSSUM30 matrix . While four of the five 20-ns MD simulations equilibrated , as judged by convergent RMSD values , the RMSD plot of the first simulation suggested that several conformational states had been sampled ( Figure 1 ) . A careful examination of the trajectory revealed that a mobile N-terminal tail was entirely responsible for the non-convergent RMSD values of the first simulation . In the crystal structure ( PDB: 1ME4 ) [25] , as in four of the five MD simulations , the N-terminal tail is held against the protein via hydrogen bonds between the A3 backbone carbonyl and the D167 backbone amine , and between the P2 backbone carbonyl and the Y166 side-chain hydroxyl group . In the first MD simulation , however , the hydrogen bond between P2 and Y166 broke after 6 . 7 ns . After 14 . 3 ns , the bond between A3 and D167 broke , allowing the N-terminal tail to rotate such that new hydrogen bonds were formed between the D167 side-chain carboxylate group and the backbone amines of both A3 and A4 . After 18 . 7 ns , the N-terminal tail returned to its original position . While these conformational changes are interesting , they occur far from the peptide binding site and so are probably not relevant to drug design . Importantly , when the first three residues of the protein are omitted from the RMSD calculation , the RMSD plot of the first MD simulation is convergent , similar to the RMSD plots of the other four simulations . The MD simulations were subsequently used to study the flexibility of the cruzain active site . Cruzain , like other cysteine proteases , contains seven subsites that bind peptide amino acids . Four subsites on the acyl side of the cleaved peptide bond , named S4 , S3 , S2 , and S1 , bind the peptide amino acids P4 , P3 , P2 , and P1 . Three subsites on the amino side of the bond , named S1′ , S2′ , S3′ , bind the peptide amino acids P1′ , P2′ , and P3′ ( Figure 2 ) [48] . The only well defined subsites of these seven are S2 , S1 , and S1′ , and only S2 and S1′ demonstrate significant specificity [49] . To judge the flexibility of the cruzain active site , the beta factor of each protein residue was calculated from the molecular motions sampled during the MD simulation . In general , the active site was remarkable for its great stability , likely in part due to the bound hydroxymethyl-ketone inhibitor [25] . To better distinguish between the many conformational states sampled by the MD simulations , 4 , 002 protein configurations were extracted from the simulations at regularly spaced intervals and grouped into 24 clusters by RMSD using the gromos clustering algorithm [35] . The centroid member was selected from each cluster , and the set of all centroid members , representative of the many conformations sampled by the MD simulations , is said to constitute an ensemble . To test the potential physiological relevance of the ensemble-member active-site conformations , CDOCKER ( Accelrys ) was used to dock the compounds of the NCI Diversity set II , a set of freely available , diverse , drug-like molecules , into both the cruzain crystal structure and the 24 protein conformations of the ensemble . A full account of the results of this virtual screen is forthcoming; however , one of the predicted inhibitors warrants further discussion here . Compound 1 ( clorobiocin , Figure 3 ) was the best predicted novel cruzain inhibitor as evaluated by the PLP2 scoring function [37] in both the screen against the static crystal structure and the relaxed-complex screen against the ensemble of 24 conformations . As positive controls , two hydroxymethyl-ketone inhibitors ( PDB: 1ME3 ) [25] were included in the relaxed-complex screen . After rescoring with an ensemble-average PLP2 score , these compounds ranked even better than compound 1 , confirming that the PLP2 scoring function is well suited to this particular protein receptor . We note with interest that previous studies have demonstrated that compound 1 antagonizes T . cruzi amastigote growth [50] , [51] . The primary protein target of clorobiocin is thought to be T . brucei topoisomerase II , but the idea of a polypharmacophoric mechanism that includes cruzain inhibition is interesting . The PLP2 scores of compound 1 docked into the central members of the first , second , and third most populated clusters were 95 . 07 , 117 . 7 , and 115 . 68 , respectively . To understand why compound 1 binding to the second ensemble conformation was favored , the pose of the ligand docked into that conformation was analyzed . While docking poses should never be blindly accepted , this particular pose seemed promising . Aside from having the best PLP2 score , this binding mode placed a conjugated ring in the S2 pocket , similar to the binding modes of some known ligands ( e . g . some vinyl sulfone inhibitors [52] ) and of some native substrates [53] . Importantly , the docked pose also suggested that one of the ligand rings binds in a previously uncharacterized , druggable pocket immediately beyond the S2 subsite ( Figure 3 ) . The beta factors of the protein residues that from this previously uncharacterized pocket revealed significant protein flexibility . Two of the residues that form the distal wall of the S2 subsite , L67 and E205 , were somewhat flexible ( Figure 4D ) , and two other protein residues beyond the S2 subsite , N69 and E112 , were also mobile ( Figure 4D ) . Together , these four flexible residues comprise two “gates” ( L67-E205 and N69-E112 ) that , when open , form the walls of a previously uncharacterized druggable pocket that medicinal chemists have yet to exploit . Published cruzain crystal structures hint at the existence of this additional pocket . A crystal structure of cruzain bound to a vinyl sulfone derived inhibitor ( PDB: 2EFM ) demonstrates a closed configuration ( Figure 4B ) , while a crystal structure of cruzain bound to a hydroxymethyl-ketone inhibitor ( PDB: 1ME3 ) [25] demonstrates a semi-open configuration ( Figure 4F ) . The crystal structures , however , do not fully capture the extent of opening demonstrated by the MD simulations . The central member of the top cluster , which accounted for 82 . 5% of the trajectory , had a closed conformation ( Figure 4A ) . The central member of the second cluster , accounting for 6 . 9% of the trajectory , had a semi-open conformation ( Figure 4C ) , and the central member of the third cluster , accounting for 3 . 6% of the trajectory , was fully open ( Figure 4E ) . As shown in Figures 4C and 4E , molecular docking demonstrates that both the semi-open and the fully open conformations can easily accommodate small molecular fragments . To further characterize the opening and closing of the first gate , the distance between the L67 γ carbon atom and the E205 δ carbon atom ( d1 ) was monitored over all 100 ns of trajectory . A histogram of these distances was bimodal ( Figure 5A ) and suggested that the gate was open ( d1>6 . 25 Å ) 70% of the time ( d1 = 7 . 6 Å±0 . 6 ) and closed ( d1<6 . 25 Å ) 30% of the time ( d1 = 5 . 4 Å±0 . 4 ) . As a reference , this same distance is 7 . 8 Å and 8 . 4 Å in the semi-open and fully open conformations , respectively , both of which can accommodate a ligand ( Figures 4C and 4E ) . We note that the presence of the hydroxymethyl-ketone inhibitor may have affected the dynamics of the first gate by largely immobilizing the E205 residue in an open conformation . E205 plays a unique role in substrate binding . The cruzain S2 subsite , like that of cathepsin B , differs from other cysteine proteases in that it can bind both hydrophobic and basic amino acids [2] , [53] . The E205 residue acts as a highly mobile switch . When a basic amino acid like arginine occupies the S2 subsite , the acidic E205 carboxylate group swings into S2 to interact with the guanidino group , a conformation that can be seen in the crystal structure of cruzain bound to benzoyl-arginine-alanine-methyl ketone ( PDB: 2AIM ) [2] , [53] . When a hydrophobic amino acid occupies the S2 subsite , E205 rotates away from S2 and interacts with the solvent [2] , [53] , a conformation evident in the crystal structure of cruzain bound to WRR-99 ( PDB: 1EWL ) . The hydrophobic phenyl group of the hydroxymethyl-ketone inhibitor present in the S2 subsite of the MD simulation locked E205 in the open , solvent exposed conformation ( Figure 4D ) . Consequently , the dynamics of the first gate were mostly determined by L67 . A histogram of the dihedral angle defined by the backbone amino nitrogen atom and the α , β , and γ carbon atoms of L67 , measured over the course of the trajectory , demonstrated that the side chain of this important residue rotated freely ( Figure 5B ) . Visual inspection confirmed that gate opening occurred when the dihedral angle ( θ1 ) was roughly −60° ( d1 = 7 . 6 Å±0 . 7 when −140°<θ1<60° ) , and that gate closing occurred when the dihedral angle was roughly 180° ( d1 = 5 . 7 Å±0 . 7 when θ1<−140° or θ1>60° ) . By this metric , the first gate was open 67% of the time , a value that matches that found by measuring the distance between the L67 γ carbon atom and the E205 δ carbon atom directly . To assess the opening and closing of the second gate , the distance between the N69 side-chain amino nitrogen atom and the E112 δ carbon atom ( d2 ) was monitored over all 100 ns of trajectory . A histogram of these distances was again bimodal ( Figure 5A ) and suggested that the second gate is open ( d2>4 . 5 Å ) 43% of the time ( d2 = 6 . 0 Å±1 . 1 ) , and closed ( d2<4 . 5 Å ) 57% of the time ( d2 = 3 . 7 Å±0 . 4 ) . As a reference , this same distance is 5 . 3 Å in the fully open conformation , which can accommodate a ligand ( Figure 4E ) . Both N69 and E112 , which form the second gate , are mobile . Of these two residues , E112 is particularly flexible ( Figure 4D ) . Visual inspection of the trajectory confirmed that N69 and E112 interact with each other via a transient hydrogen bond between the N69 side-chain amino nitrogen atom and the E112 carboxylate oxygen atoms ( Figure 5C ) . A hydrogen bond between these two residues ( distance cutoff of 3 . 5 Å ) was present in roughly 30% of the frames extracted from the trajectory . As this hydrogen bond was transient , E112 often flipped out into the solvent , where the carboxylate group interacted with water molecules . A histogram of the dihedral angle defined by the backbone amino nitrogen atom and the α , β , and γ carbon atoms of E112 ( θ2 ) , measured over the course of the trajectory , confirmed that the side chain of this important residue can freely rotate ( Figure 5B ) . Visual inspection demonstrated that gate opening occurred at two rotameric states , when the dihedral angle was 60° or 180° ( d2 = 5 . 9 Å±1 . 4 when θ2>0° or θ2<−140° ) . Gate closing occurred when the dihedral angle was −60° ( d2 = 4 . 1 Å±0 . 9 when −140°<θ2<0° ) . By this metric , the second gate was open 34% of the time . To determine what kinds of molecular fragments would best fit into the previously uncharacterized pocket immediately beyond the S2 subsite , we examined the predicted binding poses of NCI compounds docked into the third ( fully open ) ensemble conformation ( Figure 4E ) . Roughly two-dozen ligands were predicted to occupy the previously uncharacterized pocket and to bind cruzain with high affinity . With some exceptions , the molecular fragments occupying the previously uncharacterized pocket were generally aromatic rings or aliphatic chains , often with hydroxyl groups that formed hydrogen bonds with the E205 carboxylate oxygen atoms . Numerous FDA-approved drugs contain hydroxylated rings ( e . g . masoprocol , carbidopa , acetaminophen , etc . ) and/or aliphatic chains ( e . g . penciclovir , ethambutol , and miglitol ) , and so these fragments can be considered drug like . Having used MD simulations to identify a previously uncharacterized binding pocket immediately beyond the S2 subsite , we next used a bioinformatics approach to identify other possible sites of importance . Residues critical to protein function , like those of an enzymatic or allosteric active site , like those that participate in essential protein-protein interactions , or like those that play important structural roles , are often conserved across multiple homologous members of the same protein family . To identify these critical residues , cruzain was compared to other members of the peptidase C1 family . As expected , the residues of the seven subsites of the proteolytic binding pocket are generally conserved ( Figure 6A , S1 ) . The S2 subsite , critical for specificity , is an important exception; this site , like the S2 subsite of cathepsin B , differs from other cysteine proteases in that it can bind both hydrophobic and basic amino acids [2] , [53] . Additionally , the six cysteine residues involved in disulfide bonds are conserved , suggesting that these bonds are critical for protein tertiary structure . A natural mutation in human cathepsin C , a related cysteine protease with the same papain fold , confirms this importance . Patients with a cathepsin C C291Y mutation , equivalent to a cruzain C56Y mutation , develop Papillon-Lefèvre syndrome due to cathepsin C dysfunction [54] . Surprisingly , there are two patches of highly conserved residues on the side of the protein opposite the proteolytic active site . The first , patch one , is comprised of Y88 , P87 , E83 , Y86 , Q51 , and S49 . The second , patch two , is comprised of Y186 , R8 , V16 , G11 , D6 , and V13 ( Figure 6B , S1 ) . Both of these patches lie in a long shallow groove , formed largely by several disordered loops , which traverses the protein surface ( Figure 6B ) . These loops include the loops spanning G11 to G23 ( loop11–23 ) , G42 to S48 ( loop42–48 ) , Y86 to T101 ( loop86–101 ) , and N175 to G185 ( loop175–185 ) . Though disordered , these loops are held rigid by the conserved residues of the two patches , which bind the loops to stable tertiary structures and/or to each other . This rigidity may serve to maintain the shape of the traversing groove . To the best of our knowledge , this groove and its associated conserved patches , which are common to members of the peptidase C1 family , have not been previously characterized . These highly conserved patches may play roles in allosteric regulation or structural stability . Additionally , the shallow traversing groove formed by these two patches may also constitute a surface amenable to protein binding . We first turned our attention to the first patch of conserved residues . S49 and Q51 are highly conserved buried residues that belong to a stable helix spanning S49 to L56 ( helix49–56 ) . Interactions between these residues and residues of loop86–101 help to pin the loop against the stable helix , thereby imparting stability to part of the traversing groove . The S49 side-chain hydroxyl group hydrogen bonds with the backbone carbonyl oxygen of Y86 , another conserved residue . Additionally , the side-chain carbonyl oxygen atom of Q51 forms two hydrogen bonds , one with the backbone amine of A89 and one with the side-chain hydroxyl group of S89 . The side-chain amine of Q51 hydrogen also hydrogen bonds with the side-chain hydroxyl group of S89 ( Figure 6C ) . Though cruzain mutagenesis data is absent from the literature , studies of the closely related human cathepsin C protein likewise suggest that S49 and Q51 have important roles . Patients with a cathepsin C S284N mutation , analogous to a cruzain S49N mutation , develop Papillon-Lefèvre syndrome [55] , and patients with a cathepsin C Q286R mutation , analogous to a cruzain Q51R mutation , develop Haim-Munk syndrome . Both these syndromes are caused by cathepsin C dysfunction [56] . Y86 and P87 , also conserved residues of the first patch , likewise seem to play an important role in imparting rigidity to the disordered loop86–101 . The Y86 side-chain hydroxyl group forms two hydrogen bonds with T96 , helping to maintain the hairpin shape of loop86–101 . P87 does not participate in any hydrogen bonds , but the conformational rigidity of the proline backbone may contribute to the overall rigidity of loop86–101 as well ( Figure 6C ) . The rigidity of loop86–101 is in part transferred to loop11–23 and loop42–48 via the conserved residues Y88 and E83 , respectively . The Y88 side-chain hydroxyl group hydrogen bonds with the Q19 backbone carbonyl oxygen atom , and the E83 backbone amine hydrogen bonds with the N47 backbone carbonyl oxygen atom , thereby holding all these loops rigid relative to one another . It is also interesting to note that the side-chain carboxylate group of N47 is solvent exposed and potentially capable of interacting with other proteins or small-molecule compounds that may bind in the traversing groove ( Figure 6C ) . We next turned our attention to the second patch of conserved residues . The conserved residues of this patch likewise serve to hold disordered loops rigid against underlying secondary structures . For example , a hydrogen bond exists between the backbones of two highly conserved residues , G11 and R8 , that anchors part of loop11–23 to a small helix spanning W7 to R10 ( helix7–10 ) . Helix7–10 is in turn positioned relative to an underlying beta sheet by multiple hydrogen-bond interactions between the conserved residues R8 and D6 . These interactions are likely critical for protein function; D6 is analogous to the cathepsin C residue D236 , and patients with D236Y mutations develop Papillon-Lefèvre syndrome , again suggesting cathepsin C dysfunction ( Figure 6D ) [54] . The conserved residues of the second patch also impart some structure and rigidity to loop175–185 . The backbone of the conserved residue Y186 , part of a stable underlying beta sheet , forms two hydrogen bonds with the backbone of E183 . These interactions not only hold part of loop175–185 fixed relative to the beta sheet , but also help stabilize a sharp turn at the sheet-loop junction . Additionally , the phenol group of Y186 forms an interesting π-cation interaction with R8 , also conserved . This interaction may help impart curvature to the underlying beta sheet , contributing to the overall curvature of the traversing groove ( Figure 6D ) . Two additional conserved residues of the second patch , V13 and V16 , do not participate in any hydrogen-bond interactions and have no obvious structural importance . Nevertheless , V16 likely has a critical , albeit unknown , role in protein function . V16 is analogous to the cathepsin C residue V249 . Patients with V249F mutations develop Papillon-Lefèvre syndrome , again suggesting cathepsin C dysfunction [57] . Clearly , additional research is needed to further characterize these conserved patches and the traversing groove in which they are located ( Figure 6D ) . If we accept the hypothesis that in vivo the traversing groove constitutes a surface positioned at an important protein-protein interface , small molecules that target specific residues critical for protein binding may be able to disrupt the protein-protein interaction and potentially inhibit cruzain function . We note , however , that many of the residues that form the traversing groove are homologous with residues of human cathepsin C , and so relevant cruzain inhibitors are likely to inhibit cathepsin C as well . However , several residues , located between the two conserved patches ( Figure 6B ) , are not themselves conserved . For example , the cathepsin C residues homologous to cruzain A15 and N47 are I258 and P224 , respectively . It may therefore be possible to design cruzain-specific inhibitors that bind to non-conserved residues like A15 and N47 . Chagas disease , caused by the unicellular parasite T . cruzi , claims 50 , 000 lives annually [3] and is the leading cause of infectious myocarditis in the world [7] . As current antichagastic therapies like nifurtimox and benznidazole are highly toxic [6] , [8] , [9] , ineffective at parasite eradication [11] , and subject to increasing resistance [10] , novel therapeutics are urgently needed . Cruzain , the major cysteine protease of T . cruzi , is one attractive drug target [12] . In order to further the development of cruzain inhibitors , we here used MD simulations to identify a previously uncharacterized druggable pocket adjacent to the S2 subsite and a sequence alignment of a non-redundant , unbiased set of peptidase C1 family members to identify two conserved patches that may play roles in allosteric regulation , structural stability , or protein-protein interactions . Future directions include using computer-aided drug design to identify and characterize cruzain inhibitors that exploit the previously uncharacterized pocket immediately beyond the S2 subsite . Considerably more effort is required to characterize and exploit the two conserved patches opposite the peptide-binding site . While several of the residues of these patches are known to be critical for the function of cathepsin C , a cruzain homologue , mutagenesis studies are needed to directly confirm that they play an essential role in cruzain function as well . Once established as important , experiments are needed to further characterize the role these patches play . Two-hybrid screening or co-immunoprecipitation may identify other T . cruzi proteins that interact with cruzain . X-ray crystallography could then be used to determine whether or not these protein partners bind to the traversing groove formed by the two conserved patches identified in the current study . Additionally , small-molecule compounds that bind these patches may be useful tools for probing possible allosteric effects and/or disrupting critical protein-protein interactions .	Chagas disease , an infection that afflicts millions of people in Central and South America , is caused by the unicellular parasite Trypanosoma cruzi . In the chronic stage of the disease , patients' hearts are adversely affected . Chagas is the leading cause of infectious heart disease in the world . The current drugs used to treat Chagas disease are highly toxic , unable to eradiate the parasite , and subject to increasing drug resistance . Consequently , researchers are actively looking for new treatments . One attractive drug target is a Chagas protein called cruzain , which is required for the parasite's survival . Drugs that can inhibit the correct functioning of cruzain within the parasite may one day serve as powerful treatments in the fight against this devastating tropical disease . To design drugs that will be effective against cruzain , we need to know what portions of the protein are crucial for its functionality . For example , portions of the protein that bind to other proteins or to small molecules are likely to be critical . These regions are called “binding sites . ” In the current work , we identify two uncharacterized cruzain binding sites . With this knowledge in hand , future researchers may be able to design drugs that target these sites .	[ "Abstract", "Introduction", "Methods", "Results", "and", "Discussion" ]	[ "infectious", "diseases/neglected", "tropical", "diseases", "computational", "biology/comparative", "sequence", "analysis", "biochemistry/theory", "and", "simulation", "infectious", "diseases/protozoal", "infections", "computational", "biology/molecular", "dynamics", "biochemistry/...	2010	Computational Identification of Uncharacterized Cruzain Binding Sites
Individual differences in brain functional networks may be related to complex personal identifiers , including health , age , and ability . Dynamic network theory has been used to identify properties of dynamic brain function from fMRI data , but the majority of analyses and findings remain at the level of the group . Here , we apply hypergraph analysis , a method from dynamic network theory , to quantify individual differences in brain functional dynamics . Using a summary metric derived from the hypergraph formalism—hypergraph cardinality—we investigate individual variations in two separate , complementary data sets . The first data set ( “multi-task” ) consists of 77 individuals engaging in four consecutive cognitive tasks . We observe that hypergraph cardinality exhibits variation across individuals while remaining consistent within individuals between tasks; moreover , the analysis of one of the memory tasks revealed a marginally significant correspondence between hypergraph cardinality and age . This finding motivated a similar analysis of the second data set ( “age-memory” ) , in which 95 individuals , aged 18–75 , performed a memory task with a similar structure to the multi-task memory task . With the increased age range in the age-memory data set , the correlation between hypergraph cardinality and age correspondence becomes significant . We discuss these results in the context of the well-known finding linking age with network structure , and suggest that hypergraph analysis should serve as a useful tool in furthering our understanding of the dynamic network structure of the brain . Functional connectivity ( FC ) analyses based on fMRI data are effective tools for quantifying and characterizing interactions between brain regions . Many approaches borrow methods from the field of graph theory , in which FC is used to build graphs that model the brain as a complex network , treating brain regions as nodes and using functional connections ( pairs of nodes with significantly related BOLD signal dynamics ) to determine the edge structure of the network [1 , 2] . Individual differences in both underlying FC and the complex network structure resulting from graph theory approaches have been investigated for a variety of task states , developmental stages , and clinical diagnoses [3–5] . Certain characteristics of FC have been found to vary consistently over the course of normal human aging . The loss of clear segmentation between neural systems is widely reported: many intrinsic functional connectivity networks in the brain tend to become less internally coherent with age , and the functional differences between these intrinsic networks generally become less pronounced [6–8] . These changes are most commonly reported in the default mode network ( DMN ) [9–15] , although they have also been observed in other networks , including those associated with higher cognitive functions [9 , 11 , 14–16] . In addition , inter-network connectivity between the DMN and other regions of the brain has been found to increase , diminishing the ability to discriminate between networks based on FC [13 , 15] . There are some intrinsic functional networks , however , that show no changes or even increased intra-network connectivity with age , such as sensory networks [10 , 12 , 14] . The bulk of studies on age-related changes and other individual differences in FC , including those that use methods from complex networks and graph theory to represent FC patterns , are performed using static FC analysis , which represents the similarities of brain region activity ( or some other measure of concordance ) aggregated across an entire data set . In the present investigation , we build upon recent advances in network science to study individual differences in human brain activity and behavior from a dynamic network science perspective [17] . Dynamic functional connectivity ( DFC ) extends FC to examine how functional organization evolves over time [18 , 19] , allowing investigation of the changes in FC during the course of a cognitive task or scanning session . Efforts to probe the dynamics of functional brain networks have revealed that functional structure reconfigures over time in response to task demands [20–24] and spontaneously at rest [18 , 25] . DFC methods have also been used to inform understanding of individual differences related to aging . In particular , dynamic community structure was found to vary significantly with age [26] and amplitude of low-frequency fluctuations of FC ( ALFF-FC ) was used to show age-dependent changes in the dynamics of interactions between networks [27] . Both studies imply that functional dynamics should be considered when investigating how aging affects brain network organization . To address this , we use hypergraph analysis , a method from dynamic graph theory , to examine individual differences in DFC network structure in fMRI data acquired as subjects perform cognitively demanding tasks . The method is based on a generalization of standard graph theoretical techniques . In particular , by defining the standard node-node FC graph in successive temporal epochs , we construct a set of edge timeseries—that is , a vector of how the edge changes over time . The edge-edge DFC graph is constructed by treating these edge timeseries analogously to the node timeseries in the first step , and computing the relationship between every edge pair . Finally , we focus on “hyperedges , ” which are connected components of the absolute valued edge-edge DFC graph ( described in more detail in Methods ) [28] . To contextualize hypergraph analysis , we define the graph theoretic elements used to construct hypergraphs as follows: Node: As in the FC literature , nodes denote brain regions , or groups of voxels . Edge: Also corresponding to the FC literature , edges denote correlations in activity between pairs of nodes over time . A node-node graph G = {V , E} on N nodes will have ( N 2 ) edges , because each pair is considered . Unlike the majority of FC analyses , the edges are not thresholded for significance in the hypergraph analysis . Links: Links denote significant correlations in activity between pairs of edges over time . An edge-edge graph G′ = {V′ , E′} on ( N 2 ) edges will have ( ( N 2 ) 2 ) possible links , but tends to be sparse in practice . Hyperedge: A hyperedge denotes a group of links connecting two or more edges with significantly correlated temporal profiles . Hyperedges are the simplest form of link community , since they are simply the connected components of the edge-edge graph G′ = {V′ , E′} , where V is the set of edges and E is the set of links . Hypergraph: A hypergraph is a set of hyperedges . The hypergraph analysis is a simple first step toward understanding the structure of functional dynamics . Hyperedges are the connected components of the edge-edge graph , and so avoid the introduction of additional unconstrained parameters , unlike many common FC and DFC methods such as community detection . The groups of brain regions that comprise hyperedges are not necessarily strongly active or strongly interconnected brain regions . Rather , correlations in the dynamic connectivity of these regions are the defining characteristics that determine hyperedge structure . As a result , hyperedge analysis is able to identify groups of dynamic connections that change from strong to weak ( or vice versa ) cohesively together over time , providing complementary information to other DFC methods that focus on only the strongest node-node correlations , such as dynamic community detection [26 , 29 , 30] . Note that our choice of hyperedge metrics , as opposed to any other graph theoretic measure , is due to the simplicity of the hyperedge . Although it is beyond the scope of the present investigation , other graph properties of the edge-edge graph are likely to provide insight into dynamic brain network structure along other relevant dimensions . Nonetheless , hyperedges have some appealing intuitive validity in terms of the neural properties they might uncover—that is , in defining collections of nodes ( or more technically , edges ) on the basis of their similar dynamics . In previous work , we demonstrated that hyperedges discriminate between diverse task states in a group-level analysis of an fMRI data set spanning four tasks , which we refer to as the “multi-task” data set [24] . We also observed notable variation in descriptive hypergraph measures across individuals . However , given the level of abstraction involved in the construction of the hypergraph , an important first question is whether the method is able to capture well-known phenomena . In this paper , we investigate the relationship between the variability in hypergraph cardinality and other individual difference measures . We develop and employ hypergraph measures that capture individual differences in functional brain dynamics to determine correspondences between dynamics and specific demographic and behavioral measures . In the multi-task data set , we find that hypergraph cardinality—the number of distinct hyperedges within a subject’s hypergraph—exhibits marked variation across individuals . At the same time , we find this measure is consistent within individuals , across overall hypergraphs and those associated with specific tasks . To elucidate the drivers of this striking variation in hypergraph metrics observed across subjects , we explore systematic relationships between hypergraph cardinality and individual difference measures spanning distinct domains such as demographics , cognitive strategy , and personality . In the multi-task data set , we find a suggestive relationship between hypergraph cardinality and participant age . This relationship is confirmed with an independent analysis of a data set with participants who range in age from 18 to 75 , which we refer to as the “age-memory” data set . We report a strong positive relationship between age and hypergraph cardinality: older participants are significantly more likely to have a larger number of distinct hyperedges in their hypergraph . This agrees with the widely reported phenomenon of the loss of cohesion within intrinsic functional brain systems , because an increase in the number of distinct hyperedges linking various brain regions points to interconnections between functional groups evolving in time [13 , 15] . Thus , the hypergraph method agrees with previous descriptions of age-related brain changes , while capturing information about dynamics that adds a novel dimension to previous studies . This work further recommends the hypergraph as a useful tool in studying structure in dynamic functional connectivity . Informed written consent was obtained from each participant prior to experimental sessions for the multi-task and age-memory experiments . All procedures were approved by the University of California , Santa Barbara Human Participants Committee . The majority of the methods are identical to those discussed for the multi-task data set . Below , we point out aspects that differ between the two analyses . A previous study of the multi-task data identified measures that capture significant differences in population-level hypergraph structure across tasks [24] . Furthermore , extensive variation was observed in several hypergraph measures , including hypergraph cardinality , across individuals . These results emphasize that hypergraph structure can be used to differentiate between task states and motivates our investigation of the correspondence between hypergraph structure and individual difference measures . Fig 3 depicts the empirical cumulative hyperedge size distributions for all hyperedges found across all subjects in the multi-task data set . As a null test , we shuffle the data over time and find no hyperedges of size greater than one . There is a rough power law for the smaller sizes ( s < 100 ) , followed by a gap in the distribution from about 100 to 1000 and a sharp drop at the system size ( s = ( 194 2 ) = 18721 ) . The shape of the distribution is due to the consistent hypergraph structure across individuals; the majority of subjects in this study have a hypergraph composed of one large hyperedge and many small hyperedges . While this characteristic structure is common to most subjects in the study , the size of the largest hyperedge varies across individuals . This size is closely related to the hypergraph cardinality , defined as the number of hyperedges in a hypergraph , a measure which also exhibits large variation . Fig 3 also depicts task-dependent differences in the cumulative size distributions of task-specific hyperedges . Memory-specific hyperedges tend to be more numerous than those specific to the rest and attention tasks . However , the total number of task-specific hyperedges for any task is at least ten times fewer than the total number of hyperedges . Our strict definition of task specificity includes only hyperedges specific to a single task and discards those associated with more than one task . This approach is conservative , and likely leaves some meaningfully task-related hyperedges unclassified . However , it reduces the dimension of the task-specific results , and provides greater confidence that any hyperedges classified as task-specific are indeed providing truly task-driven information due to coherence within that task alone , rather than coherence due to an unrelated driver that is common to several tasks . There are significant differences in the spatial organization of task-specific hyperedges over all individuals that are visualized in Fig 4 . The plots depict task-specific hyperedge degree across the brain for each of the four tasks . In addition to the differences in magnitude between word memory and the other tasks , the locations of high hyperedge concentration vary with task . These significant differences in hypergraph structure between the tasks confirm that hypergraph structure varies between task states . However , persistent variability in hypergraph measures across individuals indicates that the hypergraph method reflects innate differences beyond the current task state . The work presented here follows this line of inquiry , beginning with an analysis of individual differences in the multi-task data set . Here , we illustrate and quantify the wide variation in hypergraph measures across individuals in the multi-task data . In brief , we identify a particular measure , hypergraph cardinality , that demonstrates large variance across all individuals but is consistent within individuals . Following this , we investigate relationships between the variation in individual difference measures and the variation in hypergraph cardinality . The results from this study are not statistically significant due to the limited variation in individual difference measures and strict corrections for multiple comparisons . However , we report a marginally significant result relating demographics and word-memory hyperedge cardinality that motivates further analyses on the age-memory data set . To supplement the findings from the multi-task data set , we perform a parallel set of analyses on the age-memory data set . The data set includes participants with ages ranging from 18 to 75 , a range three times larger than the range of ages in the multi-task study . Furthermore , the age-memory study uses an almost identical task to the multi-task word-memory task . In this section , we combine hypergraph results for all participants in the age-memory data set and obtain a distribution of hyperedge size over all participants with similar features to the hyperedge size distribution from the word-memory task of the multi-task data . We then identify and test specific drivers of individual variation in hypergraph cardinality for the age-memory study participants . We find a strong correspondence between age and hypergraph cardinality that confirms the preliminary result from the multi-task study . As we showed in the Multi-Task Analysis , the hypergraph cardinality varies widely across individuals , but is consistent between task states . Previous work on the multi-task data set found that the probability for hypergraphs to appear in a particular network configuration over individuals was significantly different depending on task state [24] . Consistent spatial organization rules for each task existed at the level of the group . There were similarities in the spatial arrangement of hyperedges in the brain for differing tasks , but certain properties were found to vary significantly between tasks . Brain areas in the occipital lobe in particular were highly likely to participate in the hypergraph across individuals and across tasks , likely due to the visual nature of most of the cognitive tasks studied . Here , we study hypergraph cardinality , which displays high variability across individuals and consistency across tasks within individuals ( Fig 5 ) . This indicates that hypergraph cardinality serves as an individual signature of a subject’s brain dynamics . The similarities across subjects in the spatial distributions of hypergraphs described in [24] capture information orthogonal to the information summarized by hypergraph cardinality . For example , there are some individuals for whom the visual brain regions are linked by many hyperedges , and some for whom those same regions are linked by relatively few hyperedges , but these regions are more likely than others to be included in hypergraphs in the majority of subjects . This suggests that , for some subjects , brain regions tend to be more dynamically integrated in general , with co-varying functional relationships across many brain circuits; in other subjects , connectivity dynamics are more fragmented across the brain . The high degree of variability in hypergraph cardinality across subjects and consistency within subjects , combined with the significant differences in spatial hyperedge arrangement across tasks , indicate that hypergraphs are a useful analysis tool for investigating both individual and task-based differences in brain function in a variety of settings . At the same time , hypergraphs can provide a view of dynamic patterns that complements other commonly used DFC methods . For example , many FC methods exclusively investigate the structure of strong correlations in functional data [29 , 66–68]; hypergraph analysis captures information about both strongly and weakly correlated dynamics and how sets of brain regions transition between them [28] . Although they are highly informative , many of the hypergraph metrics we study here are representative measures that greatly reduce the dimension of the hypergraph and only reveal a small part of the information contained in its structure . Further development of methods to utilize more of the information that hypergraphs provide will allow characterization of the consistency of particular hyperedges and dynamic modes , an understanding of which are important for behavior , or influenced by demographics or disease . Future work is also needed to further quantify the spatial differences in hypergraph arrangement across both individuals and tasks , to clarify the extent of overlap between the two types of information , and to determine whether the individual variability in cardinality can be mapped to individual spatial differences in hypergraph structure . FC studies have established clear trends associated with aging , including a decrease in connectivity within functional networks and an increase in connectivity across different functional networks in resting and task states [15 , 69–72] . Many of these studies have considered resting-state FC , because the absence of task stimulus provides a simple and reliable setting for comparison between subjects [73] , although recent studies have successfully used FC networks to study various cognitive proceses [74] . The default mode network ( DMN ) and similar resting-state analyses may miss functional changes evoked by task states; while the DMN FC decreases with age , task-related sensorimotor network FC has been shown to increase with age [12 , 14] . Similarly , FC in memory tasks shows increased segmentation with age [75] . Extending these analyses to incorporate the dynamics of functional interactions is a necessary step towards quantifying individual changes in functional brain dynamics associated with age . Several efforts have been made to capture individual age-related differences with methods from dynamic FC . Dynamic community structure and amplitude of low-frequency fluctuation of FC were both found to be strongly correlated with age , illustrating that functional dynamics are closely linked with aging [26 , 27] . In the dynamic community detection analysis , functional communities were found to be more fragmented with age , which agrees with the hypergraph cardinality result presented here [26] . A multi-scale community detection analysis uncovered similar fragmentation with age for small scales [76] . Our finding that hypergraph cardinality also increases with age aligns with this result and provides further information based upon its ability to capture higher-order dynamic patterns across larger ensembles of brain regions . Not only do the functional similarities of communities of brain regions themselves become less distinct as humans age , but the temporal profiles of these functional similarities also become less integrated across brain regions . The agreement of this result with known age-related changes in FC [6–8 , 13 , 15] demonstrates the ability of hypergraph methods to capture and quantify major brain changes . Moreover , since the hypergraph analysis is not limited to strong correlations , our analysis further suggests that age is related not only to the organization of functional activity in groups of brain regions with strongly coherent activity , but also to the coordination between groups of regions that transition from being strongly to weakly correlated over time ( or vice versa ) . The reported correspondence between age and hypergraph cardinality is significant in the age-memory data set , but our analysis did not include data that could verify this relationship for cognitive tasks other than the word memory task . Although memory is a cognitive ability known to decline with age in many individuals , it is unlikely that the specific task studied in the age-memory data set drives this result . Rather , the consistency of hypergraph cardinality across tasks seen in the multi-task data set in Fig 5 ( B ) suggests that similar hypergraph cardinalities may be found during other tasks in data sets with higher age variability , and that the relationship between age and cardinality is unlikely to depend primarily on the behavioral task . Further investigation is needed to determine whether individual differences in hyperedge structure have any significant relationship to behavioral or cognitive performance on any particular task . Here , we have shown that the considerable differences in functional connectivity dynamics across individuals are closely linked with age . The hypergraph method is presented as an analysis tool that captures information about group-level similarities that differ between task states as well as individual differences that are consistent within individuals , across tasks . Further investigation into a single hypergraph metric ( hypergraph cardinality ) that varies across individuals uncovers a significant relationship between hypergraph cardinality and age . Specifically , there are a greater number of hyperedges in older individuals’ hypergraphs , suggesting that there are more small groups of regions with cohesively evolving dynamics and indicating a loss of coherence across larger , spatially distributed intrinsic functional connectivity networks . This complements widely reported relationships between FC and human aging by providing new insight into how FC activity and the co-evolution of FC activity are altered with increasing age , including the loss of large groups of co-evolving brain regions in older individuals . The correspondence with and extension of classic FC results to new dynamic regimes , along with the unique capacity of hypergraphs to probe multiple dimensions of both strong and weak dynamic variability , show that hypergraph analysis is a valuable tool for understanding age-related changes and other individual differences in dynamic brain function .	Complex patterns of activity in each individual human brain generate the unique range of thoughts and behaviors that person experiences . Individual differences in ability , age , state of mind , and other characteristics are tied to differences in brain activity , but determination of the exact nature of these relationships has been limited by the intrinsic complexity of the brain . Here , we apply dynamic network theory to quantify fundamental features of individual neural activity . We represent functional connections between brain regions as a time varying network , and then identify groups of these interactions that exhibit similar behavior over time . The result of this construction is referred to as a hypergraph , and each grouping within the hypergraph is called a hyperedge . We find that the number of these hyperedges in an individual’s hypergraph is a trait-like metric , with considerable variation across the population of subjects , but remarkable consistency within each subject as they perform different tasks . We find a significant correspondence between this metric and the subject’s age , indicating that the dynamics of functional brain activity in older individuals tends to be more dynamically segregated . This new insight into age-related changes in the dynamics of cognitive processing expands our knowledge of the effects of age on brain function and confirms our methods as promising for quantifying and examining individual differences .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "medicine", "and", "health", "sciences", "diagnostic", "radiology", "functional", "magnetic", "resonance", "imaging", "neural", "networks", "demography", "social", "sciences", "neuroscience", "learning", "and", "memory", "magnetic", "resonance", "imaging", "regression", ...	2016	Individual Differences in Dynamic Functional Brain Connectivity across the Human Lifespan
Kaposi’s sarcoma-associated herpesvirus ( KSHV ) latently infects host cells and establishes lifelong persistence as an extra-chromosomal episome in the nucleus . To persist in proliferating cells , the viral genome typically replicates once per cell cycle and is distributed into daughter cells . This process involves host machinery utilized by KSHV , however the underlying mechanisms are not fully elucidated . In present study , we found that N-Myc downstream regulated gene 1 ( NDRG1 ) , a cellular gene known to be non-detectable in primary B cells and endothelial cells which are the major cell types for KSHV infection in vivo , was highly upregulated by KSHV in these cells . We further demonstrated that the high expression of NDRG1 was regulated by latency-associated nuclear antigen ( LANA ) , the major viral latent protein which tethers the viral genome to host chromosome and plays an essential role in viral genome maintenance . Surprisingly , knockdown of NDRG1 in KSHV latently infected cells resulted in a significant decrease of viral genome copy number in these cells . Interestingly , NDRG1 can directly interact with proliferating cell nuclear antigen ( PCNA ) , a cellular protein which functions as a DNA polymerase clamp during DNA replication . Intriguingly , we found that NDRG1 forms a complex with LANA and PCNA and serves as a scaffold protein bridging these two proteins . We further demonstrated that NDRG1 is critical for mediating LANA to recruit PCNA onto terminal repeat ( TR ) of KSHV genome , and facilitates viral DNA replication and episome persistence . Taken together , our findings suggest that NDRG1 plays an important role in KSHV viral genome replication , and provide new clues for understanding of KSHV persistence . Kaposi’s sarcoma-associated herpesvirus ( KSHV ) , a human oncogenic DNA gammaherpesvirus , is known for its causal association with human cancers , including endothelial-derived Kaposi’s sarcoma ( KS ) , a B cell malignancy named primary effusion lymphoma ( PEL ) , and a plasmablastic form of the B lymphoproliferative disorder named multicentric Castleman disease ( MCD ) [1–4] . KSHV infection of host cells is predominantly latent , and the virus establishes lifelong persistence of its genome in proliferating cells , which contributes to tumorigenesis . During latency , KSHV exists as a circular extrachromosomal episome tethered to the host chromosome [5–8] . To persist in cells , KSHV typically replicates once , accompanied by host replication , and is distributed into daughter cells along with the host chromosomes [5 , 6 , 8–12] . KSHV-positive PEL cells , such as BCBL1 , BC3 , and JSC1 , are cultured cell lines established from KSHV-infected PEL samples , and there are multiple copies of viral episomes in these cells , ranging from approximately 50 to 200 per cell . The copy number of the KSHV genome remains constant with cell division , suggesting that there are mechanisms by which KSHV persists in these cells . KSHV utilizes various viral and cellular factors for perpetuation in host cells . An indispensable viral factor is the latency-associated nuclear antigen ( LANA ) , one of the few viral proteins expressed during latency , which is required and sufficient for KSHV episome persistence in the absence of other viral genes . LANA , encoded by the KSHV open reading frame 73 ( ORF73 ) , is an 1162-amino-acid protein . This protein tethers viral episomes to cellular chromosomes and ensures episomal DNA replication during each cell division event and segregation of viral DNA into progeny nuclei [6 , 8 , 10 , 11 , 13] . The C terminus of LANA directly binds to the KSHV latent replication origin-terminal repeat ( TR ) DNA to mediate viral DNA replication [5 , 8 , 14–19] . The N terminus of LANA interacts with the host chromosome and is critical for the efficiency of LANA-mediated viral DNA replication and episome persistence [20–24] . However , LANA lacks the enzymatic activity required for DNA replication . To achieve this function , LANA recruits numerous cellular replication factors , such as the origin recognition complex ( ORC1-6 ) , replication factor C ( RFC ) , minichromosome maintenance complex ( MCM ) , topoisomerase II beta ( TopoII beta ) , structure-specific recognition protein 1 ( SSRP1 ) , and proliferating cell nuclear antigen ( PCNA ) [19 , 23 , 25–28] . PCNA is a DNA clamp that is highly conserved in eukaryotic species and is essential for DNA replication , acting as a processivity factor for DNA polymerase epsilon . PCNA encircles the DNA and executes its processivity as a scaffold , enrolling proteins that participate in DNA synthesis or repair [29–31] . Recently , it was found that PCNA is loaded onto the KSHV TR region by LANA , which is a rate-limiting step in viral DNA replication [23] . The recruitment by LANA of PCNA for loading onto the KSHV latent replication origin allows increased efficiency of viral replication and persistence . Although LANA does not interact with PCNA directly , this antigen recruits PCNA via several adaptors , such as the replication factor RFC and the cellular mitotic kinase Bub1 , to assist viral replication and persistence during latency . Depletion of RFC or Bub1 has a negative impact on LANA’s ability to replicate and maintain viral episomes in KSHV-infected cells or in cells containing the TR plasmid , thus leading to loss of virus [23 , 27] . To investigate how KSHV utilizes cellular factors to maintain viral persistence during latency , we compared primary rat embryonic metanephric mesenchymal precursor cells ( MM cells ) and KSHV-immortalized MM cells ( KMM cells ) [32] and identified a host protein , named N-Myc downstream regulated gene 1 ( NDRG1 ) , that was distinctly upregulated in KMM cells . NDRG1 is a multifunctional protein that is involved in cell growth , differentiation , development , stress response , etc . [33–35] . However , whether NDRG1 associates with KSHV during infection and plays a role in KSHV persistence has not yet been determined . NDRG1 is widely expressed in numerous human tissues but is nondetectable in certain cell types , such as primary B cells and endothelial cells [36] . Remarkably , the prevailing view is that KS is derived from endothelial cells and that PEL is derived from B cells that are permissive to KSHV infection [37 , 38] , suggesting that KSHV might specifically upregulate NDRG1 in these cells to facilitate infection . In this study , we demonstrated that NDRG1 is highly expressed in KSHV-positive cells , including KSHV-infected KS and PEL cells . We further demonstrated that the high expression of NDRG1 in KSHV-positive cells is induced by KSHV infection and that LANA is essential for promotion and maintenance of NDRG1 expression in KSHV-infected cells . Surprisingly , knockdown of NDRG1 dramatically reduces the viral copy numbers and leads to loss of viral genomes in cells latently infected with KSHV . Interestingly , we found that NDRG1 directly interacts with PCNA and that NDRG1 also forms a complex with LANA and PCNA . We further showed that NDRG1 acts as a scaffold protein , mediating the recruitment of PCNA by LANA onto the KSHV TR , leading to the promotion of viral DNA replication and episome persistence during latency . KSHV utilizes various cellular factors to persist in proliferating host cells [10 , 11] . To investigate the underlying mechanisms , we performed RNA-seq , microarray analysis and iTRAQ to identify factors that may affect the persistence of viral latency using primary rat embryonic MM cells and KMM cells as a pair of uninfected and KSHV-infected cells , respectively [32] . We performed an integrative analysis of transcriptomic and proteomic changes and screened out a series of differentially expressed candidate genes . Thirty-two genes ( S1 Table ) were identified as being differentially expressed at both the RNA and protein levels by comparing the microarray and iTRAQ databases . Fifty-seven genes ( S2 Table ) were generated by assessing the RNA-seq and iTRAQ data , and seventeen genes ( S3 Table ) were screened out by analyzing the microarray , RNA-seq , and iTRAQ databases ( Fig 1A ) . To verify the candidates , we tested these genes in KMM and MM cells via qPCR ( S1 Fig ) . Surprisingly , we found that a cellular gene named NDRG1 , known to be nondetectable in primary B cells and endothelial cells , which are permissive to KSHV infection [37] , was significantly upregulated in KMM cells at the RNA level ( Fig 1B ) . Consistently , NDRG1 , which was detected approximately 43kDa , was markedly upregulated in KMM cells at protein level , while it was non-detectable in MM cells ( Fig 1B ) . These results proved that NDRG1 is at higher levels in KSHV-positive KMM cells than in KSHV-negative MM cells , indicating that NDRG1 might play a role in viral persistence . We next tested the endogenous expression level of NDRG1 in KSHV-positive human PEL cell lines ( BCBL1 , BC3 , and JSC1 ) , which were established from KSHV-infected PEL tissue samples [39–41] , and KSHV-negative human B lymphoma cell lines ( DG75 , Raji , Loukes , and Ramous ) . The results were consistent with those observed in rat cells . NDRG1 was highly expressed in all KSHV-positive B cells but barely detected in KSHV-negative B cells at both the RNA and protein levels ( Fig 1C ) . Given that NDRG1 is known to be nondetectable in primary B cells [36] , the results suggested that the presence of NDRG1 in KSHV-positive cells might be important for persistent KSHV infection in not only rat cells but also human cells . We further examined NDRG1 expression in KSHV-positive KS tumor samples and KSHV-negative normal skin tissues by immunohistochemical assays . LANA is one of the major latent viral proteins expressed in KS spindle cells [37 , 38] , so we used LANA as a marker to monitor the KSHV-positive cells in KS tissues . As expected , no LANA staining was observed in normal skin tissues . NDRG1 is usually expressed in the basal cells of the epidermis but not in the cells of the dermis . In contrast to normal skin tissues , there were strong signals for NDRG1 in the dermal cells of the KS tissues , which also stained positive for LANA ( Fig 1D ) . To further confirm this result , we performed a double staining assay for both NDRG1 and LANA in the same KS tissue sample section . NDRG1 staining was observed in cells that stained positive for LANA ( Fig 1E ) . These results strongly suggest that the elevated expression of NDRG1 in KSHV-positive cells is correlated with KSHV infection and that NDRG1 may play a critical role in the persistence of viral latency . Although NDRG1 is highly expressed in KSHV-positive cells , whether the expression of this protein is induced by KSHV infection is not clear . To explore this effect , primary human umbilical vein endothelial cells ( HUVECs ) were infected de novo with KSHV virions , and the expression of NDRG1 was examined 48 hr post infection ( hpi ) . We found that the RNA level of NDRG1 in the KSHV-infected group was upregulated approximately 10-fold compared to that in the mock group ( Fig 2A ) . The protein level of NDRG1 in KSHV-infected HUVECs was markedly increased , while NDRG1 protein was slightly or barely detectable in the mock HUVECs ( Fig 2A ) , indicating that KSHV infection leads to substantial enhancement of NDRG1 expression in cells . To confirm these results , SLK cells were infected de novo with KSHV virions under similar conditions . Similarly , distinct augmentation of NDRG1 expression at both the RNA and protein levels was observed in KSHV-infected SLK cells compared to mock SLK cells ( Fig 2B ) . These results confirmed that KSHV infection significantly upregulates NDRG1 expression in cells . To further ascertain whether KSHV-induced NDRG1 was dependent upon KSHV gene expression rather than KSHV virion proteins or only the cellular stress response , HUVECs were subjected to infectious KSHV virions or UV-inactivated virions . Previous studies have demonstrated that UV-inactivated KSHV virion particles do not lessen the ability of the virions to bind or enter host cells , but the virions do so without viral gene expression [42–44] . Here , we assessed the RNA levels of NDRG1 in cells 6 , 24 , 48 hpi via qPCR . As shown in Fig 2C , a gradual increase in NDRG1 mRNA levels was observed in the live KSHV-infected group , while barely and changes in NDRG1 mRNA expression were observed in the UV-inactivated KSHV group . We also measured the protein levels of NDRG1 in cells 48 hpi . Consistent with the RNA levels observed , the NDRG1 protein could be detected in live KSHV-infected cells but was barely expressed in UV-inactivated KSHV-infected cells or mock cells ( Fig 2D ) . The efficiency of infection of cells with KSHV is shown in S2 Fig and the protein level of LANA could be detected in KSHV-infected cells ( Fig 2D ) , indicating KSHV de nove infected successfully . These data suggest that KSHV viral gene expression activates the expression of NDRG1 . To further confirm this hypothesis , SLK cells were also infected de novo with UV-inactivated KSHV virions and live KSHV virions under similar conditions . The results were similar to those obtained in HUVECs . The RNA levels of NDRG1 in SLK cells could be activated and increased by KSHV infection , while the levels remained unchanged in UV-inactivated KSHV-infected cells ( Fig 2E ) . The protein levels of NDRG1 in SLK cells infected with UV-inactivated KSHV and live KSHV at 48 hpi were determined , and the results showed again that high expression signals could be detected in only KSHV-infected cells , not UV-inactivated KSHV-infected cells ( Fig 2F ) . All these data indicated that KSHV viral gene expression , rather than virion proteins or the cellular stress response , activates the expression of NDRG1 during KSHV infection . The above results demonstrated that the products generated by KSHV genes regulate the expression of NDRG1 . Next , we sought to identify the viral protein that regulates NDRG1 expression . The latent phase of KSHV is characterized by the expression of a limited number of viral genes . LANA is the most abundant viral protein expressed during latency and is required for various vital functions for persistence of KSHV infection [45] . Hence , we speculated that LANA might be responsible for regulating the expression of NDRG1 . To test this possibility , we first investigated the role of LANA in the maintenance of NDRG1 expression . We silenced LANA in KSHV-infected SLK cells via RNA interference . Indeed , the expression of NDRG1 was greatly decreased when LANA expression was knocked down ( Fig 3A ) , which is consistent with the result in PEL cells ( S3 Fig ) In addition , the NDRG1 mRNA level was significantly decreased , which was accompanied by reduction of LANA expression ( Fig 3A ) . These data suggested that LANA plays a role in maintaining NDRG1 expression . Next , we explored the role of LANA in the induction of NDRG1 in KSHV infection . For this purpose , we first constructed two stable SLK cell lines expressing SLK-shcon and SLK-shLANA . The purified KSHV virus was used to infect these two stable cell lines , and NDRG1 was detected at both the protein and RNA levels . The efficiency of infection of SLK-shcon and SLK-shLANA cells with KSHV were shown in S4 Fig . Gradual augmentation of NDRG1 expression was observed in the SLK-shcon group , and the peak NDRG1 level was observed at 48 hpi , along with an increase in LANA expression , during KSHV infection ( Fig 3B ) . However , NDRG1 failed to be induced sufficiently when LANA was inhibited in SLK-shLANA cells ( Fig 3B ) , suggesting that LANA plays a key role in inducing NDRG1 during KSHV infection . To further validate these results , SLK cells were infected with KSHV and LANA-depleted KSHV [46] or were not infected . Similarly , NDRG1 expression was not induced at the protein level when LANA was eliminated ( Fig 3C ) , although the RNA level of NDRG1 was slightly enhanced compared with that in uninfected cells ( Fig 3C ) , suggesting that LANA is responsible for inducing NDRG1 during KSHV infection . Taken together , these results demonstrate that LANA plays an essential role in the upregulation of NDRG1 expression . To explore the role of NDRG1 in KSHV infection , we knocked down endogenous NDRG1 in KMM cells . KMM cells were transduced with a lentivirus containing two specific shRNAs ( sh-NDRG1-1# and sh-NDRG1-2# ) and a control shRNA ( sh-control ) , followed by selection . The NDRG1 expression were greatly decreased in KMM-shNDRG1-1# and 2# compared with KMM-shcon ( Fig 4A ) . Because the previous data indicated that NDRG1 may play a critical role in the persistence of viral latency , we hypothesizes that the expression of viral genes might be changed in the absence of NDRG1 expression . We checked the transcriptional levels of a representative latent gene ( LANA ) and a lytic gene ( RTA ) in KMM-shNDRG1 cells . Surprisingly , both LANA and RTA levels were significantly diminished upon the reduction of NDRG1 levels in KMM cells ( S5 Fig ) , indicating that NDRG1 might regulate the KSHV DNA levels , thereby simultaneously affecting both latent and lytic gene expression at the RNA level . To validate this hypothesis , we assessed the DNA levels of the viral genomes in KMM cells approximately 20 days after NDRG1depletion in KMM cells by qPCR . The results showed that the DNA levels of TR , LANA , and K9 in KMM-shNDRG1 cells were markedly lower than those in KMM-shcon cells ( Fig 4B ) . We further assessed the DNA levels of the viral genomes at days 1 , 5 , 10 , and 20 after NDRG1 knockdown in KMM cells . There was a distinct loss of KSHV episomal genomes in the absence of NDRG1 , whereas KSHV episome levels remained relatively stable in the control cells ( Fig 4C ) . To obtain direct evidence , we performed fluorescence in situ hybridization ( FISH ) analysis . Indeed , the FISH results showed that the levels of TR DNA , a region on the KSHV viral genome , were decreased in the NDRG1 knockdown KMM-shNDRG1 cells ( S6 Fig ) . These findings suggested that silencing NDRG1 hampers viral genome persistence in cells latently infected with KSHV . To further verify this phenomenon , we examined the effect of NDRG1 downregulation on the maintenance of viral episomes in BCBL1 , which is a PEL cell line latently infected with KSHV . We constructed stably transfected BCBL1 cell lines , including BCBL1-shcon , BCBL1-shNDRG1-1# , and BCBL1-shNDRG1-2# . NDRG1 expression in BCBL1-shNDRG1 cells was successfully reduced ( Fig 4D ) . Indeed , the DNA levels of the intracellular viral genomes were decreased in BCBL1-shNDRG1 cells , which were tested approximately 20 days after NDRG1depletion in cells ( Fig 4E ) . As LANA colocalizes with KSHV episomes on chromosomes in PEL cells [6] , we also detected LANA by immunostaining to determine the changes in viral episomes . In NDRG1 knockdown cells , the LANA dots were significantly reduced , and the percentage of cells containing LANA dots was approximately 30% in BCBL1-NDRG1-1# and approximately 12% in BCBL1-NDRG1-2# compared with BCLB1-shcon cells ( Fig 4F ) . Taken together , these results demonstrated that inhibition of NDRG1 in KSHV-infected cells results in greatly decreased persistence of the KSHV genome and suggested that NDRG1 plays a role in the maintenance of the viral genome in KSHV-infected cells . To explore the potential mechanism by which NDRG1 regulates the persistence of viral episomes during KSHV latency , we performed tandem affinity purification/mass spectrometry ( TAP-MS ) to identify proteins that interact with NDRG1 . Strep-FLAG ( SF ) -tagged NDRG1 was expressed in KSHV-positive iSLK . RGB cells . Then , the cell lysates were subjected to affinity purification with streptavidin beads , followed by immunoprecipitation ( IP ) with FLAG M2 beads ( Fig 5A ) . The purified eluates were analyzed by MS . Several nucleoproteins , shown in S4 Table , were identified by peptide correlation with the International Protein Index database , suggesting that NDRG1 may play important roles in cell nuclei . The classified and predicted functions of the identified proteins are shown in the pathway pie chart in Fig 5B , which was created using the PANTHER system ( http://www . pantherdb . org/ ) . Interestingly , there were approximately 13 . 6% potential NDRG1-binding proteins related to DNA replication , which would influence the viral copy numbers in KSHV-infected cells . PCNA , among these proteins , is a replication processivity factor and functions as a DNA polymerase clamp that is essential for DNA replication [29 , 30] . We speculated that NDRG1 might interact with PCNA and be involved in regulating viral DNA replication , thereby contributing to the maintenance of viral copy numbers in KSHV-infected cells . To verify this hypothesis , we performed a co-IP assay , and the results showed that NDRG1 interacted with PCNA ( Fig 5C ) . To further determine whether this interaction was direct , we investigated the interaction using an in vitro binding assay . As shown in Fig 5D , IVT-PCNA ( in vitro translated PCNA ) was directly bound to full-length NDRG1 ( 1–394 aa ) . To support the results , we also performed an immunofluorescence ( IF ) analysis to assess the localization of endogenous PCNA and NDRG1 in KSHV-positive BCBL1 cells . The results showed that NDRG1 colocalized with PCNA in the nuclei of BCBL1 cells ( Fig 5E ) . Taken together , these results demonstrate that NDRG1 directly interacts with PCNA in vitro and in vivo , suggesting that NDRG1 might play a role in DNA replication . PCNA can be loaded onto the KSHV TR region by LANA indirectly , and this loading is a rate-limiting step in viral DNA replication [23 , 27] . Based on the above data , we hypothesized that NDRG1 might form a complex with LANA and PCNA and regulate the replication of KSHV episomal DNA . To determine whether NDRG1 forms a complex with LANA and PCNA in cells naturally infected with KSHV , we performed an IF assay in PEL cells . We found that NDRG1 colocalized with PCNA and LANA in the nuclei of BCBL1 , BC3 , and JSC1 cells ( Fig 6A ) . We also performed a co-IP assay in BCBL1 cells . The results showed that endogenous LANA could interact with PCNA and NDRG1 in cells naturally infected with KSHV ( Fig 6B ) . Previous studies have been reported that LANA can bind TR region of episomes itself [16] and PCNA can be recruited onto TR region [23 , 27] . In order to exclude the possibility that NDRG1 and PCNA are associated with KSHV episomes but not because of a meaningful complex with LANA , we also preformed an anti-CTCF IP as a control , which is an antibody against a KSHV episome-associated antigen that does not involve targeting LANA [47–49] . The results showed that NDGR1 and PCNA are enriched in an anti-LANA pulldown in BCBL1 cells , compared to an anti-CTCF pulldown , indicating that endogenous LANA-specific associates with NDRG1 and PCNA ( S7 Fig ) . To further confirm this hypothesis , we also performed a co-IP assay in HEK293T cells . The results showed that NDRG1 coimmunoprecipitated with PCNA and LANA ( Fig 6C ) . Next , we altered the amount of NDRG1 while keeping the amount of LANA and PCNA unchanged to check whether the quantity of PCNA that coimmunoprecipitated with LANA would vary . For this purpose , we first generated a stably transfected HEK293T-Strep-Flag-LANA cell line that constantly expressed SF-LANA , endogenous NDRG1 and PCNA . Then , a NDRG1-containing plasmid was transfected into the HEK293T-SF-LANA cells to upregulate NDRG1 expression in the cells . The results showed that the quantity of PCNA that coimmunoprecipitated with LANA increased when the amount of NDRG1 was increased ( Fig 6D ) . These data indicated that NDRG1 forms a complex with LANA and PCNA and that NDRG1 mediates the amount of PCNA recruited by LANA . Because LANA cannot directly bind to PCNA [23 , 27] , we speculated that NDRG1 acts as an adaptor , connecting LANA and PCNA . First , we performed an in vitro binding assay to test whether LANA binds to PCNA . As shown in Fig 6E , LANA did not directly bind to IVT-PCNA , which is consistent with previous reports [23 , 27] . Next , we performed an in vitro binding assay to test whether LANA binds to NDRG1 . The results showed that in vitro translated NDRG1 ( IVT-NDRG1 ) directly bound to the N-terminal domain of GST-fused LANA instead of the C-terminal domain ( Fig 6F ) . To test whether LANA would bind PCNA in vitro in the presence of NDRG1 , IVT-PCNA and IVT-NDRG1 were incubated with purified GST-fused LANA beads . The results showed that LANA could bind PCNA in the presence of NDRG1 ( Fig 6G ) , suggesting that NDRG1 bridges the interaction between LANA and PCNA . Taken together , these data demonstrate that NDRG1 forms a complex with LANA and PCNA in vivo and in vitro . Based on the above results , we speculated that NDRG1 may mediate the recruitment of PCNA by LANA onto the TR DNA and facilitate viral DNA replication . To test this hypothesis , we performed an in vitro pull-down assay by using TR biotin-labeled DNA . NDRG1 and/or LANA were transfected into BJAB cells , which express endogenous PCNA but lack NDRG1 and LANA . Twenty-four hours later , the cells were lysed , incubated with the biotin-TR DNA fragment , and immobilized to streptavidin beads . The inputs and the pull-down products were analyzed by western blotting . As shown in Fig 7A , the presence of NDRG1 in BJAB did not alter the levels of PCNA and LANA . In such cases , we found that the amount of PCNA bound with TR did not change when NDRG1 was present but LANA was absent . Consistent with previous reports [23 , 27] , the loading of PCNA onto DNA was dramatically enhanced in the presence of LANA ( Fig 7A ) . In addition , NDRG1 was capable of enhancing the loading of PCNA onto TR in the presence of LANA ( Fig 7A ) . The full-length western blot images of Fig 7A cutoff bands are shown in S8 Fig . These results indicated that the enhancement of PCNA enrichment on TR by NDRG1 is dependent on LANA-mediated recruitment of PCNA onto the TR DNA . We further assessed whether NDRG1 mediates the recruitment of PCNA to TR DNA by LANA in cells by a chromatin immunoprecipitation ( ChIP ) assay . NDRG1 and/or LANA were transfected into BJAB cells along with p8TR plasmids . The cells were lysed for the PCNA-ChIP assay 24 hr post transfection . As shown in Fig 7B , PCNA was substantially enriched at the TR DNA in the presence of LANA . The enrichment of PCNA at the TR was predicted to be further enhanced in the presence of both NDRG1 and LANA . To confirm this hypothesis , we performed a PCNA-ChIP assay in HEK293T-LANA cells . Similarly , the results showed that the enrichment of PCNA at the TR was enhanced at high NDRG1 levels in the cells ( Fig 7C ) . To determine the role of NDRG1 in the recruitment of PCNA to the KSHV genome , we first performed ChIP to assay the presence of LANA , NDRG1 , and PCNA at the TR DNA in KSHV-infected BCBL-1 cells . As expected , LANA and PCNA were enriched at the TR DNA , as previously reported [23 , 27] . In addition , we also found that NDRG1 was enriched at the TR DNA , suggesting that NDRG1 may play a role in the KSHV genome ( Fig 7D ) . To determine whether NDRG1 mediates the recruitment of PCNA to the KSHV genome , we performed a PCNA-ChIP assay in BCBL1-shcon and BCBL1-shNDRG1 cells . As shown in Fig 7E , the level of PCNA enrichment to the KSHV genome was hampered when NDRG1 was knocked down , suggesting that NDRG1 mediates the recruitment of PCNA to the KSHV genome by LANA . To assess the role of NDRG1 in LANA-mediated DNA replication , we performed a LANA-mediated DNA replication assay [23 , 50] . The p8TR plasmid was transfected into BJAB cells along with LANA and NDRG1 or vector plasmids . First , to ensure the efficiency of transfection , we collected cells at 24 hr post transfection and assayed the expression of LANA , NDRG1 , and PCNA by western blotting ( Fig 8A left ) and the amount of p8TR DNA by qPCR and DNA gel analysis ( Fig 8A right ) . These results indicated similar transfection efficiencies of the LANA and TR plasmids in both groups of BJAB cells . Next , we collected cells at 72 hr post transfection and determined the effects of NDRG1 on LANA replication . As shown in Fig 8B , DNA replication activity was significantly enhanced in the cells expressing NDRG1 compared with those lacking NDRG1 , indicating that NDRG1 assists LANA-mediated DNA replication . The process of TR DNA replication affects the maintenance of viral episomes in dividing cells . To further assess the role of NDRG1 in DNA replication and maintenance of viral episomes , we measured the short-term retention rates of TR plasmid levels post transfection in the absence of selective agents . The p8TR plasmid was transfected into BJAB cells along with LANA and NDRG1 or vector plasmids , and the total TR level was measured at days 1 , 2 , 3 , and 4 post transfection by qPCR . Consistent with previous reports [51] , the total TR plasmid levels in the cells remained stable for 48 hr post transfection and gradually decreased with the reduction of transfected LANA ( Fig 8C ) . Interestingly , NDRG1 enhanced the total levels of TR plasmids maintained in the cells , especially at 48 hr , and the levels then gradually decreased . These results indicated that NDRG1 plays a role in DNA replication and maintenance of viral episomes . To further confirm these results , the p8TR plasmids along with NDRG1 or vector plasmids were transfected into HEK293T-LANA cells stably expressing LANA and low levels of endogenous NDRG1 . The results in Fig 8D show the efficiency of transfection . Cells were collected at 72 hr post transfection , and the effects of NDRG1 on LANA-mediated DNA replication were determined . Consistent with the above results , DNA replication activity was significantly enhanced in the cells expressing high levels of NDRG1 ( Fig 8E ) . Short-term retention of TR plasmids post transfected in the absence of selective agents was also measured . The results showed that NDRG1 further reduced the loss of TR plasmids ( Fig 8F ) , indicating that NDRG1 plays a role in DNA replication and may also have an effect on TR maintenance . To test whether NDRG1 assists LANA in long-term retention of TR , the p8TR plasmid was transfected into HEK293T-LANA cells along with NDRG1 or vector plasmids . Because TR plasmids can be maintained long term in subpopulations in the presence of LANA without drug selection [51] , the cells were cultured without drug selection . Considering the loss of NDRG1 plasmids in dividing cells post transfection , we replenished the NDRG1 plasmids at days 3 and 7 . Higher levels of TR DNA were detected at day 10 in the presence of both LANA and NDRG1 than in the presence of LANA but absence of NDRG1 ( Fig 8G ) , suggesting that NDRG1 plays a role in TR plasmid persistence . In cells latently infected with KSHV , the viral genome levels are constant . Similar to other gammaherpesviruses , KSHV is distinctly adept at establishing stable latent infections and maintaining a constant copy number of viral episomes in proliferating host cells throughout the life of the host . For instance , KSHV-positive PEL cells contain 50–200 copies of viral episomes per cell , and the copy number of the KSHV genome remains constant after multiple cell division events [39–41] . To maintain this stable episome in growing cells , the viral genome typically replicates once per cell cycle and is evenly distributed into daughter cells [5 , 6 , 8–13] . However , the regulation of this process is complicated . Anything that interferes with the replication of viral episomes along with cell DNA or hampers efficient distribution of replicated episomes into daughter nuclei will hinder the maintenance of viral episomes . This process involves the utilization of host machinery by KSHV; however , the underlying mechanisms have not been fully elucidated . In our study , we found that the cellular protein NDRG1 is highly upregulated by KSHV infection . Silencing of NDRG1 results in greatly decreased KSHV episome persistence . In terms of mechanism , we found that NDRG1 serves as a scaffold protein that bridges the interaction between LANA and PCNA and mediates the recruitment of PCNA to the KSHV genome by LANA . Finally , we demonstrated that NDRG1 enhances LANA-mediated replication and thus influences the persistence of the KSHV genome . The proposed working model is summarized in Fig 9 . Previously , numerous studies have shown that the host protein NDRG1 is a multifunctional protein that is involved in carcinogenesis , differentiation , stress response , immunity , etc . [33–36] , but none of these studies linked this protein to viral DNA replication . Recently , some groups have reported that NDRG1 might be associated with viral infection . It has been described that miRNAs encoded by EBV can downregulate NDRG1 , a suppressor of metastasis , to promote EBV-mediated epithelial carcinogenesis , suggesting that NDRG1 plays a negative role in EBV-induced cancer [52] . Another study also showed that NDRG1 restricts hepatitis C virus propagation by regulating lipid droplet formation and viral assembly [53] . On the other hand , a study published this year reported that NDRG1 facilitates influenza A virus replication by suppressing canonical NF-kappa B signaling [54] . All of these studies suggest that NDRG1 may participate in the viral life cycle , but the functions of this protein appear to be diverse . Regardless of the role of NDRG1 in other viruses , the role of this protein in KSHV infection has not been studied . NDRG1 is a member of the NDRG family , currently consisting of NDRG1 , NDRG2 , NDRG3 and NDRG4 . In our study , the omics data of MM and KMM cells showed NDRG1 but not NDRG2-4 is distinctly upregulated in KMM cells , hence we focused on exploring the role of NDRG1 in KSHV infection . We demonstrated that the high expression of NDRG1 in KSHV-positive cells is induced by KSHV infection . The expression of NDRG1 is regulated by a diverse range of effectors . Many of the effectors are the transcription factors , like EGR-1 , Sp-1 , Ap-1 , and ETS , which are known to regulate NDRG1 expression [33] . The latent phase of KSHV is characterized by the expression of a limited number of viral genes . LANA , a transcription factor , is the most abundant viral protein expressed during latency and is required for various vital functions of KSHV infection [45] . Our data suggested that LANA is essential for upregulation of NDRG1 during KSHV infection . We analyzed the promoter of NDRG1 and found a DNA sequence ( CGCTCAGGGCGTGGCGC , -410 of the promoter ) similar to LBS1 ( CGCCCGGGCATGGGGC ) , which is the LANA DNA binding motif in TR region of KSHV genome . So it is possible that LANA may regulate NDRG1 expression through interacting with its promoter . We have preliminary data ( S9 Fig ) indicating that LANA alone can upregulate the mRNA level of NDRG1 slightly but in the meantime the protein level of NDRG1 remains unchanged . This suggests that LANA may regulate the expression of NDRG1 at both transcriptional and post transcriptional levels . Some other viral factors may also be involved in the regulation of NDRG1 . The detailed mechanisms for the upregulation of NDRG1 during KSHV infection would be interesting to be further studied in the future . DNA replication of the KSHV episome along with each cell division event during latency is one of the key components of episome persistence . Perturbation of viral episomal DNA replication will negatively impact viral DNA maintenance [10 , 11] . LANA , one of the few viral proteins expressed during latency , is essential for genome maintenance as well as viral episomal DNA replication [5 , 6 , 8 , 9 , 12] . LANA directly binds to the TR region , which is the KSHV latent replication origin region , via its C-terminal end to mediate viral DNA replication [5 , 8 , 14–19] . Because LANA lacks the enzymatic activity required for DNA replication , this protein interacts with and recruits diverse cellular proteins for latent viral genome replication , including the origin recognition complex ( ORC1-6 ) , poly ( ADP-ribose ) polymerase 1 ( PARP1 ) , minichromosome maintenance complex ( MCM ) , Bromodomain containing 2 ( BRD2 ) , H4-specific histone acetylase ( HBO1 ) , replication factor C ( RFC ) , topoisomerase II beta ( TopoII beta ) , and structure-specific recognition protein 1 ( SSRP1 ) [5 , 6 , 19 , 23 , 25–28 , 55–63] . Despite this LANA-mediated viral DNA replication , some evidence also suggests that latent KSHV DNA replication can be initiated at origins that are not required for LANA replication [12 , 64] . However , these studies also showed that LANA-mediated DNA replication is the predominant viral DNA replication pattern during latency [64] . Similar to host DNA replication , the main steps in the process of latent KSHV DNA replication include initiation , elongation , and termination . During the elongation step , new viral DNA strands are formed . Recent studies have shown that LANA enhances the loading of PCNA onto the KSHV TR DNA region and enhances the efficiency of latent viral DNA replication . PCNA is a cellular protein that is a component of the DNA replication machinery . This protein encircles the viral DNA and cooperates with host DNA polymerase to increase the efficiency of viral DNA elongation . However , LANA cannot directly interact with PCNA , and LANA recruits PCNA via other proteins , such as RFC and the cellular mitotic checkpoint kinase Bub1 . Knockdown of the expression of RFC or Bub1 will reduce the replication and persistence of the viral DNA [23 , 27] . In our study , we proved that silencing of NDRG1 results in greatly decreased KSHV episome persistence . We further demonstrated for the first time that NDRG1 serves as a scaffold protein , bridging the interaction between LANA and PCNA . NDRG1 mediates the recruitment by LANA of PCNA for loading onto the KSHV genome and enhances LANA-mediated replication , thus affecting the persistence of the KSHV genome . This finding again indicates that the recruitment of PCNA onto the KSHV genome by LANA is important for latent viral DNA replication and persistence . LANA is 1162-amino-acid protein that can be divided into an N-terminal domain , an internal repeat domain , and a C-terminal domain . The LANA-N region ( 1–32 ) interacts with the cellular histones H2A and H2B and is responsible for tethering viral episomes to the host genome [5 , 65] . Studies have also shown that the LANA-N domain is crucial for LANA-mediated DNA replication and viral genome persistence . The LANA-N region ( 1–32 ) recruits host TopoII beta to the TR region was found to be crucial for KSHV DNA replication in KSHV infected cells as well as in transient replication system [25] The LANA-N domain ( 262–320 ) interacts with RFC to mediate efficient latent viral replication and persistence [23] . In our study , we found that the LANA-N domain ( 1–340 ) interacts with NDRG1 to mediate latent viral replication . This finding is consistent with the finding that the LANA-N domain participates in LANA-mediated DNA replication and viral genome persistence . In addition , to the best of our knowledge , the role of NDRG1 in the nuclei remains unclear . The finding that NDRG1 directly interacts with the nuclear localization protein PCNA , a cellular DNA replication-related protein , indicates a novel function of NDRG1 related to DNA replication in host cells . In conclusion , our work showed that KSHV infection upregulated the expression of the host cellular protein NDRG1 , and silencing of NDRG1 resulted in greatly decreased viral episome persistence . NDRG1 serves as an adaptor that mediates the recruitment by LANA of PCNA for loading onto the KSHV genome and hence affects KSHV genome persistence . Because NDRG1 is critical for efficient LANA-mediated DNA replication and KSHV episome persistence , this work implicates NDRG1 as an attractive target for disruption . In addition , NDRG1 is nondetectable in normal B cells and endothelial cells , which are permissive to KSHV infection . Therefore , strategies that inhibit NDRG1 in KSHV-infected cells may be effective for virus eradication . The clinical tissue specimens from patients with KS were collected from Xinjiang Province , northwestern China . The protocols were reviewed and ethically approved by the Institutional Ethics Committee of the First Teaching Hospital of Xinjiang Medical University ( Urumqi , Xinjiang , China; Study protocol no . 20082012 ) . Written informed consent was obtained from all participants , and all samples were anonymized . All participants were adults . The MM , KMM , SLK , iSLK . RGB , iSLK . LANAstop , and HEK293T cell lines were cultured in DMEM ( HyClone ) supplemented with 10% FBS ( HyClone ) , antibiotics ( penicillin and streptomycin , HyClone ) , and the appropriate selective pressures ( hygromycin , 0 . 5 mg/ml; puromycin , 1 . 5 μg/ml; G418 , 0 . 5 mg/ml ) . KSHV-positive B lymphoma cell lines ( BCBL1 , JSC1 , BC3 ) and KSHV-negative B lymphoma cell lines ( DG75 , Raji , Loukes , Ramous ) were cultured in RPMI 1640 ( HyClone ) supplemented with 10% FBS ( HyClone ) and 1% antibiotics ( penicillin and streptomycin , HyClone ) . HUVECs were maintained in EGM ( Lonza ) . All the cell lines were grown at 37°C in a humidified environment supplemented with 5% CO2 . We are grateful to Dr . Shoujiang Gao ( University of Southern California ) for the MM and KMM cell lines [32] and iSLK . LANAstop cell line [46] , which produces LANA-depleted KSHV after induction; Dr . Jae Jung ( University of Southern California ) and Dr . Fanxiu Zhu ( The Florida State University ) for the SLK and iSLK . RGB cell lines [66]; Dr . Erle S Robertson ( University of Pennsylvania , USA ) for the BCBL1 , JSC1 , BC3 , DG75 , Raji , Loukes , BJAB , and Ramous cell lines . The HEK293T cell line was from our laboratory stock . The shLANA and control plasmids have been previously reported [67] . Two target-specific shRNAs complementary to NDRG1 ( 5’-ACC TGC ACC TGT TCA TCA A-3’ and 5’-CGC TGA GGC CTT CAA GTA C-3’ ) were cloned into the pLKO . 1 vector as previously reported . A pLKO . 1 vector with a scrambled sequence against NDRG1 was used as a control ( 5’-GGA ATC TCA TTC GAT GCA TAC-3' ) [68 , 69] . Full-length fragments of NDRG1 ( NM_001135242 . 1 ) and PCNA ( NM_002592 . 2 ) were amplified from a HEK293T cDNA library . A DNA construct expressing SF-tagged NDRG1 was generated by cloning full-length NDRG1 to a modified pCDH-SF-IRES-Blast vector , which introduced a fragment encoding a tandem Strep-tag II and FLAG peptide . pcDNA3 . 1-NDRG1 and pcDNA3 . 1-PCNA were constructed by cloning full-length NDRG1 and PCNA into pcDNA3 . 1 ( + ) , respectively . pcDNA3 . 1-HA-NDRG1 was constructed by subcloning the HA-NDRG1 fragment into pcDNA3 . 1 ( + ) from pCMV-HA-NDRG1 , which was constructed by subcloning the NDRG1 fragment from pcDNA3 . 1-NDRG1 into the pCMV-HA vector . Similarly , pcDNA3 . 1-HA-PCNA was constructed by subcloning the HA-PCNA fragment into pcDNA3 . 1 ( + ) from pCMV-HA-PCNA , which was constructed by subcloning the NDRG1 fragment from pcDNA3 . 1-PCNA into the pCMV-HA vector . The plasmids pCAGGS-HA-LANA and pCDH-SF-LANA , encoding FLAG-tagged LANA , were described previously [70 , 71] . The truncated LANA constructs GST-LANA-N ( 1–340 aa ) and GST-LANA-C ( 1022–1162 aa ) were described previously [72] . GST-fused full-length NDRG1 was obtained by cloning the corresponding fragments into the pGEX-4T-1 vector . The p8TR-gB plasmid , which contains eight TR copies was described previously [50] . All the primers used for gene amplification and qPCR are listed in S5 Table . The following primary antibodies were used: anti-NDRG1 rabbit monoclonal antibody ( Abcam , ab124689 ) , anti-PCNA ( P10 ) mouse monoclonal antibody ( CST , #2586 ) , anti-PCNA rabbit polyclonal antibody ( ABclonal , A4006 ) , anti-α-tubulin antibody ( Sigma , T6199 ) , anti-FLAG antibody ( Sigma , F1804 ) , anti-CTCF mouse monoclonal antibody ( Santa Cruz , sc-271514 ) , anti-CTCF rabbit polyclonal antibody ( ABclonal , A1133 ) , anti-LANA rat monoclonal antibody ( Advanced Biotechnology Inc , 13-210-1000 ) , and anti-LANA mouse monoclonal antibody ( 1B5 ) , which was prepared in our laboratory [73] . The secondary antibodies used for western blotting and the IF assay were goat anti-mouse IRDye 800CW ( Li-Cor , 926–32210 ) ; goat anti-rabbit IRDye 680RD ( Li-Cor , 926–68071 ) ; goat anti-rabbit antibodies conjugated with Alexa Fluor 488 ( Thermo Fisher Scientific , A-11094 ) , 555 ( Thermo Fisher Scientific , A27017 ) , and 680 ( Thermo Fisher Scientific , A27020 ) ; goat anti-mouse antibodies conjugated with Alexa Fluor 488 ( Thermo Fisher Scientific , A-11001 ) and 555 ( Thermo Fisher Scientific , A-21422 ) ; and goat anti-rat antibodies conjugated with Alexa Fluor 488 ( Thermo Fisher Scientific , A-11006 ) . The other reagents used ( and their sources ) were as follows: anti-FLAG M2 affinity gel ( Sigma , A2220 ) , Strep-Tactin Sepharose ( IBA , 2-1201-010 ) , desthiobiotin ( IBA , 2-1000-001 ) , recombinant protein A agarose ( Invitrogen , 15948–014 ) , recombinant protein G agarose ( Invitrogen , 15920–010 ) , glutathione Sepharose 4B ( GE Healthcare , 17-0756-01 ) , the TNT T7 Quick Coupled transcription/translation system ( Promega L1170 ) , Dynabeads M-280 streptavidin ( Invitrogen , 11205D ) . To determine the RNA levels or genomic DNA levels , quantitative real-time PCR was used . To analyze RNA levels , the cDNA was reverse transcribed with the Genomic DNA Eraser RT Kit ( TaKaRa ) from the total RNA extracted from cells harvested with TRIzol reagent ( Life Technologies ) . To analyze DNA levels , total DNA was extracted from the cells with the Genomic DNA Extraction Kit ( Tiangen ) . Relative KSHV episomal copy numbers were calculated by qPCR amplification of the terminal repeats as previously described [61] . qPCR was performed with a SYBR Green Master Mix Kit ( Toyobo ) on a 7900HT system ( Life Technologies ) . Relative mRNA levels and relative DNA levels were normalized to actin or GAPDH and calculated by the ΔΔCT method . The samples were tested in triplicate . The primers are listed in S5 Table . The expression of LANA and NDRG1 in the tissue samples was analyzed by both immunohistochemistry ( IHC ) and IF as previously described [75] . Primary antibodies consisting of either anti-LANA antibody ( 1:500 ) or anti-NDRG1 antibody ( 1:200 ) were used . For immunofluorescence , cells were fixed with 4% paraformaldehyde ( PFA ) for 30 min and permeabilized with 0 . 2% Triton X-100 for 30 min . Next , the cells were blocked in 10% goat serum ( Life Technologies ) for 1 hr , followed by incubation with anti-NDRG1 ( 1:100 ) , anti-PCNA ( 1:100 ) or anti-LANA ( 1:250 ) antibody overnight at 4°C and further staining with the secondary antibodies at a 1:1 , 000 dilution for 1 hr . Cell nuclei were stained with DAPI ( Sigma , D5942 ) . Slides were photographed using a digital camera and software ( FV-1200; Olympus ) . Wild-type KSHV virions were acquired by inducing iSLK . BAC16 . RGB cells with doxycycline ( DOX ) and valproic acid as described previously [76] . Briefly , iSLK . BAC16 . RGB cells were induced with DOX ( 1 g/ml ) and valproic acid ( 1 mM ) in the absence of hygromycin , puromycin , and G418 . Four days later , the supernatant was collected by centrifugation ( 1 , 500×g at 4°C for 30 min ) , followed by filtration . Virus particles were precipitated with 8 . 8% polyethylene glycol 6000 by centrifugation ( 1 , 500×g at 4°C for 1 hr ) . The MOI of the concentrated viral stock was determined by infection of HEK293T cells for 24 hr . Infection of SLK cells or HUVECs was achieved by centrifugation at 1 , 250×g at 37°C for 2 hr after the addition of concentrated virus to the medium , and the medium was changed at 2 hpi . UV-inactivated KSHV ( UV-KSHV ) was used in this study . Replication-incompetent KSHV was prepared by exposing wild-type KSHV for 30 min under UV light as described earlier [77] . Wild-type and LANA-depleted KSHV were also used in this study . LANA-depleted KSHV virions were acquired from iSLK . LANAstop cells , which produce LANA-depleted KSHV after induction . Titration of KSHV stocks was performed as described previously [46] . To transduce SLK , KMM , and BCBL1 cells , pCDH plasmids were packaged in HEK293T cells by cotransfection with the Δ8 . 9 packaging plasmid and a plasmid expressing vesicular stomatitis virus G protein ( pVSV-G ) as described previously [76] . Three days later , the supernatant was collected and cleared by filtration . Lentiviral particles were precipitated with 7% polyethylene glycol 6000 and pelleted by centrifugation at 1 , 500×g for 1 hr at 4°C . The transduction of SLK cells was achieved by centrifugation at 1 , 250×g for 2 hr after addition of the concentrated virus to the medium . The transduction of KMM and BCBL1 cells was achieved by addition of the concentrated virus to the medium for 24 hr . HEK293T cells were transfected with polyethylenimine as described previously [67] . BJAB cells were transfected by nucleofection . Cells were washed with PBS and resuspended in 100 μl of transfection buffer from the Ingenio Kit ( MIR 50118 , Mirus ) and nucleofected with plasmids using the Amaxa Nucleofector II system ( Lonza ) according to the manufacturer’s instructions . DNA FISH assays were performed as previously described [8 , 63] , with minor modifications . Briefly , cells were fixed in 4% PFA at room temperature for 30 min and then fixed in 70% ethanol overnight at -20°C . Fixed cells were permeabilized with 0 . 5% Triton X-100 . Subsequently , the cells were treated with 100 μg/ml of RNase A in 2×SSC ( 1×SSC contains 0 . 15 M NaCl plus 0 . 015 M sodium citrate ) . The slides were overlaid with in situ hybridization solution containing 20 ng of KSHV TR probe labeled with DIG using the DIG High Prime DNA labeling system ( Roche ) according to the manufacturer’s instructions . After denaturation of the DNA at 93°C for 5 min , the slides were incubated for 24 hr at 42°C , washed in 2×SSC for 30 min at 45°C , and blocked in TNB ( 0 . 1 M Tris-HCl ( pH 7 . 5 ) , 0 . 15 M NaCl , 0 . 5% blocking reagent ) at room temperature for 30 min . The slides were then incubated with anti-DIG antibody ( 2 μg/ml ) for 2 hr and with goat-anti-mouse 555 ( 1:1000 ) for 1 hr at room temperature . The slides were washed and counterstained with DAPI , followed by mounting with antifade and visualizing with a digital camera and software ( FV-1200; Olympus ) . TAP-MS was used to identify NDRG1 protein complexes . The procedures were as described previously [73 , 78] . Briefly , SF-tagged NDRG1 or SF-tagged vector was overexpressed in iSLK . RGB cells . Cells were lysed , and the extract was loaded into a Strep-Tactin Sepharose column ( IBA ) . The column was washed with buffer W ( 50 mM Tris ( pH 7 . 9 ) , 100 mM KCl , 10% glycerol , 0 . 2 mM EDTA , 0 . 5 mM DTT , 0 . 1% Triton-X100 , 0 . 2 mM PMSF ) and eluted with buffer E ( buffer W containing 2 . 5 mM D-desthiobiotin ) . The elute was then subjected to a second round of affinity purification on an anti-FLAG M2 affinity gel for 4 hr at 4°C . The beads were washed with buffer W 4 times and eluted with 3×FLAG peptide in buffer W . The elute was monitored by SDS-PAGE and subjected to mass spectrometry . Cells were lysed in radio immunoprecipitation assay ( RIPA ) buffer ( 50 mM Tris-HCl ( pH 7 . 4 ) , 150 mM NaCl , 0 . 5% Triton X-100 , 1 mM PMSF ) for 1 hr on ice with brief vortexing every 15 min . The cells were then ere centrifuged at 12 , 000×g at 4°C for 30 min to remove cell debris . Five percent of the cell lysates were kept as inputs . The remainder was precleared with protein A- or protein G-coupled Sepharose ( Life Technologies ) for 2 hr at 4°C and then immunoprecipitated with the corresponding antibodies overnight at 4°C . Immunoprecipitates were washed four times with RIPA buffer and then boiled in SDS loading buffer for western blot analysis . For western blotting , protein samples were analyzed by SDS-PAGE and transferred onto nitrocellulose membranes , followed by blocking and probing with the indicated antibodies for detection . The procedures were performed as described previously [70] . Briefly , GST or GST fusion proteins were expressed in Escherichia coli strain BL21 ( DE3 ) , which was grown in LB medium at 37°C to exponential phase and cultured overnight at 16°C after induction with isopropyl thiogalactopyranoside ( IPTG ) . The cells were harvested and resuspended in ice-cold PBS , followed by sonication lysis ( Sonics; 3/5 s , pulse cycle; 35% , amplitude ) . The cell lysates were centrifuged at 12 , 000×g to obtain the supernatant , which was combined with Sepharose 4B-glutathione resin for affinity purification according to the manufacturer’s instructions . In vitro translated proteins produced by the TNT-coupled transcription/translation system were incubated with purified GST-fusion-protein-bound beads in RIPA buffer for 12 hr . After washing with RIPA buffer four times , the pull-down products were analyzed by western blotting . To determine whether NDRG1 mediates PCNA loading onto TR DNA in the presence of LANA , a TR biotin-labeled DNA pull-down assay was adapted from previously described methods [23 , 79] . Briefly , the biotin-labeled 801-bp TR DNA fragment was amplified by PCR using 5’-biotin-labeled primers and the p8TR plasmid as the template . The forward primer sequence was 5’ GCGCCTGGTCCCGCCCCCGCCCGC 3’ , and the reverse primer sequence was 5’ CGGCCGCGCCGGGCCCTGAGGCGGC 3’ . Ten million cells were lysed in RIPA buffer ( 50 mM Tris-HCl ( pH 7 . 4 ) , 150 mM NaCl , 0 . 5% Triton X-100 , 1 mM PMSF ) for 1 hr on ice , vortexing very briefly every 15 min . Cells were centrifuged at 12 , 000×g at 4°C for 30 min to remove cell debris . Five percent of the cell lysate was kept as input . The remainder of the cell lysate was incubated with the purified biotin-TR DNA fragment ( 100 ng ) supplemented with 1 μg/ml poly dI/dC for 4–5 hr at 4°C and then immobilized on 15 μg of streptavidin beads following the manufacturer’s protocol ( Dynabeads M-280 streptavidin; Invitrogen ) for 4–5 hr at 4°C . Following this immobilization , the supernatant was removed . The beads were washed four times with PBS and then boiled in SDS loading buffer for western blot analysis . ChIP assays were performed as described previously [23 , 70] . Briefly , cells were fixed in medium with 1% formaldehyde for 30 min ( LANA , NDRG1 and PCNA ) at room temperature and quenched with 0 . 125 M glycine . For revealing differences in PCNA loading on TR region with or without NDRG1 , cells were especially immersed in hypotonic buffer containing 0 . 1% Triton X-100 for 10 min before fixation [80] . After fixation , the cells were washed with PBS twice and lysed in SDS lysis buffer ( 50 mM HEPES , 1 mM EDTA , 1% SDS , 1 mM PMSF ) supplemented with protease inhibitor cocktail for 30 min on ice . The lysates were subjected to sonication to obtain 200-500-bp DNA fragments ( Sonics; 2/6 s , pulse cycle; 30–35% , amplitude ) and then centrifuged to obtain the supernatants . The supernatants were diluted with RIPA buffer . Samples were precleared with pretreated protein A or G beads ( 1 mg/ml BSA , 1 mg/ml sperm DNA , 20% beads ) for 2 hr at 4°C . A small fraction of the supernatants was kept as input , and the remainder was divided into groups according to the experiment . The aliquots were incubated with pretreated protein A or G beads and the corresponding antibody overnight at 4°C . After extensive washing with RIPA buffer , wash buffer ( 20 mM Tris-HCl ( pH 8 . 0 ) , 1 mM EDTA , 250 mM LiCl , 0 . 5% NP-40 , 1 mM PMSF ) and TE buffer ( 10 mM Tris-HCl ( pH 8 . 0 ) , 1 mM EDTA ) , three to four times each , the beads were resuspended in TE buffer . The resuspended beads were subjected to RNaseA and proteinase K digestion , and the crosslinking was reversed at 65°C overnight . The DNA was recycled using a DNA purification kit . Transient replication assays in uninfected cells were used to assess LANA-mediated DNA replication , and the procedures were performed as described previously [50] . In brief , for the DNA replication assay performed in BJAB cell lines , ten million cells were cotransfected by nucleofection with 6 μg of p8TR-gB , 6 μg of NDRG1 and 6 μg of LANA or empty vector plasmids . After twenty-four hours of transfection , five million cells were collected and used to normalize the transfection efficiency , and the other cells were further cultured . Seventy-two hours post transfection , 10 million cells were collected and subjected to DNA extraction by Hirt’s method [81] . For the DNA replication assay in HEK293T-LANA cells stably expressing LANA , cells in 10-cm dishes were cotransfected with 20 μg of p8TR and 10 μg of NDRG1 or an empty vector plasmid . Similarly , after twenty-four hours of transfection , cells were collected and used to normalize the transfection efficiency , and the other cells were further cultured . Seventy-two hours post transfection , the cells were collected and subjected to DNA extraction by Hirt’s method . To detect replicated p8TR-gB , the Hirt DNA was digested overnight at 37°C with DpnI in NEB buffer #2 , which was followed by exonuclease III ( ExoIII ) treatment for 30 min at 37°C . The digested DNA was assayed for replication using real-time PCR as previously described [50] . The results represent the average values obtained from three experiments . For the plasmid maintenance assay performed in BJAB cells , ten million cells were cotransfected by nuclofection with 6 μg of p8TR-gB , 6 μg of NDRG1 and 6 μg of LANA or empty vector plasmids . For the plasmid maintenance assay performed in HEK293T-LANA cells , 20 μg of the p8TR plasmid was transfected into HEK293T-LANA cells along with 10 μg of NDRG1 or vector plasmids . After twenty-four hours of transfection , five million cells were collected and used to normalize the transfection efficiency , and the other cells were further cultured for several days . To detect total p8TR-gB DNA levels , the cells were collected on days 1 , 2 , 3 , and 4 , followed by extraction of genomic DNA . Relative TR DNA levels of the p8TR plasmid were calculated by qPCR amplification of the TRs , normalized to GAPDH levels , and analyzed by the ΔΔCT method as previously described [61] . TR plasmids can be maintained long term in subpopulations in the absence of drug selection [51] . To test whether NDRG1 assists LANA retain TR in the long term , part of the transfected HEK293T-LANA cells described above were further cultured without drug selection for 10 days . As described above , the cells were collected and analyzed by qPCR [61] . Data are presented as the mean ± standard deviation ( x ± SD ) . The level of significance was set at P-value<0 . 05 , as determined by Student’s t-tests . All experiments were carried out independently at least three times , and representative results are presented .	KSHV latently infects cells and establishes lifelong persistence , but the underlying mechanisms of this process has not been fully elucidated . Here , we find a novel host protein NDRG1 is highly up-regulated by KSHV infection and the viral protein LANA is essential in this process . NDRG1 is a multiple functional protein , but the role in KSHV infection remains unknown . Our findings show that NDRG1 functions as a scaffold protein that forms a complex with PCNA and LANA , thereby helping LANA load PCNA onto the viral genome and facilitating the replication and persistence of KSHV . Since NDRG1 is non-detectable in primary B cells and endothelial cells , which are the major cell types susceptible to KSHV infection in vivo , NDRG1 might be a candidate of therapeutic target for inhibition of KSHV persistence in malignant cells .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "transfection", "medicine", "and", "health", "sciences", "pathology", "and", "laboratory", "medicine", "pathogens", "microbiology", "viral", "structure", "plasmid", "construction", "viruses", "dna", "replication", "dna", "viruses", "viral", "genome", "dna", "constructio...	2019	NDRG1 facilitates the replication and persistence of Kaposi’s sarcoma-associated herpesvirus by interacting with the DNA polymerase clamp PCNA
Natural killer ( NK ) cells are essential immune cells against several pathogens . Not much is known regarding the roll of NK cells in Orientia tsutsugamushi infection . Thus , this study aims to determine the level , function , and clinical relevance of NK cells in patients with scrub typhus . This study enrolled fifty-six scrub typhus patients and 56 health controls ( HCs ) . The patients were divided into subgroups according to their disease severity . A flow cytometry measured NK cell level and function in peripheral blood . Circulating NK cell levels and CD69 expressions were significantly increased in scrub typhus patients . Increased NK cell levels reflected disease severity . In scrub typhus patients , tests showed their NK cells produced higher amounts of interferon ( IFN ) -γ after stimulation with interleukin ( IL ) -12 and IL-18 relative to those of HCs . Meanwhile , between scrub typhus patients and HCs , the cytotoxicity and degranulation of NK cells against K562 were comparable . CD69 expressions were recovered to the normal levels in the remission phase . This study shows that circulating NK cells are activated and numerically increased , and they produced more IFN-γ in scrub typhus patients . Orientia tsutsugamushi is an obligate intracellular bacterium that causes scrub typhus in humans . It is a mite-borne , endothelium-targeting intracellular bacterium . Scrub typhus is prevalent in Asia , Northern Australia , and the Indian subcontinent . Most patients may recover from scrub typhus without complications if provided with an early diagnosis and management [1] . However , some patients develop fatal complications with median mortality of 6 . 0% unless they are treated sufficiently early in the course of illness [2] . O . tsutsugamushi resides in the cytoplasm of host cells , which are mainly endothelial cells , macrophages , monocytes and dendritic cells [3 , 4 , 5 , 6] . Related studies have found elevated plasma concentrations of interferon ( IFN ) -γ , IFN-γ-inducing cytokines ( e . g . , interleukin [IL]-12 , IL-15 , IL-18 , and tumor necrosis factor [TNF]-α ) , and chemokines induced by IFN-γ ( e . g . , IFN-γ-inducible protein 10 and monokine induced by IFN-γ ) . These are well known for recruiting natural killer ( NK ) cells and T cells in patients with scrub typhus [7 , 8] . Based on these findings , a combination of innate and adaptive immune responses likely contribute to host defense against O . tsutsugamushi . Natural killer ( NK ) cells are essential effectors within our innate immunity system . They mediate the elimination of target cells directly or indirectly through secretion of effector molecules such as perforin/granzyme , cytokines ( mainly IFN-γ ) , and chemokines [9 , 10] . NK cells were discovered in the mid-1970s as they showed the ability to lyse tumor cells without prior exposure [11 , 12] . It is now well established that NK cells are also effective against several viruses , fungi , parasites and some intracellular bacteria such as Salmonella , Listeria and Chlamydia [10 , 13 , 14] . NK cell-mediated cytotoxicity is a complex process that involves receptor-mediated binding and signaling , synapse formation , granule polarization , and granule release [15] . Infection by intracellular pathogens leads to a decreased expression of major histocompatibility complex ( MHC ) class I antigens in host cells . This decrease reduces the infected host cell’s ability to interact with NK cells’ inhibitory receptors . In turn , the infected cell’s becomes more susceptible to lysis by NK cells , which leads to the destruction of the intracellular pathogen [16] . Similarly , K562 cells ( which lack the MHC complex required to inhibit NK activity ) are easily killed by NK cells . For this reason , these cells are often used for detection of NK cytotoxicity [17] . In murine models of Rickettsial infection , the clearance of bacteria was found to be significantly associated with NK cell activity and mice with NK cell deficiency showed increased susceptibility to infection [18] . However , studies have yet to explore the role of NK cells in O . tsutsugamushi infection in humans . Accordingly , this study aims to examine the level and function of NK cells in patients with scrub typhus , as well as the clinical relevance of NK cell levels . The study cohort comprised 56 patients with scrub typhus ( 30 women and 26 men; mean age ± SD , 66 . 8 ± 13 . 0 years ) and 56 age- and sex-matched healthy controls ( HCs; 28 women and 28 men; mean age ± SD , 62 . 7 ± 8 . 03 years ) . Patients’ diagnoses required detecting O . tsutsugamushi antibodies via a passive hemagglutination assay ( PHA ) using Genedia Tsutsu PHA II test kits ( GreenCross SangA , Yongin , Korea ) . Patients were positive for infection if test results showed a titer of ≥ 1:80 in a single serum sample , or at least a four-fold rise in antibody titer at a follow-up examination . According to the number of dysfunctional organs involved , severity of scrub typhus was classified into the following 3 grades as previously described [19 , 20]: mild disease ( no organ dysfunction ) ; moderate disease ( one organ dysfunction ) ; and severe disease ( ≥ 2 organ dysfunctions ) . Organ dysfunction is: ( 1 ) renal dysfunction , creatinine ≥ 2 . 5 mg/dL; ( 2 ) hepatic dysfunction , total bilirubin ≥ 2 . 5 mg/dL; ( 3 ) pulmonary dysfunction , bilateral pulmonary infiltration on chest X-rays with moderate to severe hypoxia ( PaO2/FiO2 < 300 mmHg or PaO2 < 60 mmHg or SpO2 < 90% ) ; ( 4 ) cardiovascular dysfunction , systolic blood pressure < 80 mmHg despite fluid resuscitation; and ( 5 ) central nervous system dysfunction , significantly altered sensorium with Glasgow Coma Scale ( GCS ) ≤ 8/15 . All healthy controls were recruited in the Jeollanam-do area , which was the same area as the areas where the patients have developed . Controls had no history of autoimmune disease , infectious disease , malignancy , chronic liver or renal disease , diabetes mellitus , immunosuppressive therapy , or fever within 72 hours prior to enrollment . The Institutional Review Board of Chonnam National University Hospital approved this study’s protocol . All participants provided written informed consent in accordance with the Declaration of Helsinki . This study used the following mAbs and reagents: Allophycocyanin ( APC ) -Cy7-conjugated anti-CD3 , APC-conjugated anti-CD3 , APC-conjugated anti-CD69 , and fluorescein isothiocyanate ( FITC ) -conjugated anti-CD45 , FITC-conjugated anti-CD56 , FITC-conjugated anti-CD107a , FITC-conjugated anti-IFN-γ , phycoerythrin ( PE ) -conjugated anti-CD3 , PE-conjugated anti-CD56 , PE-conjugated anti-CD69 , PE-Cy7-conjugated anti-TNF-α , PerCP-conjugated anti-CD3 , PerCP-conjugated anti-CD45 , FITC-conjugated mouse IgG isotype and PE-conjugated mouse IgG isotype control ( all from Becton Dickinson , San Diego , CA ) . Cells were stained with combinations of appropriate mAb for 20 minutes at 4°C . Stained cells were analyzed on a Navios flow cytometer using Kaluza software ( version 1 . 1; Beckman Coulter , Brea , CA ) . Peripheral venous blood samples were collected in heparin-containing tubes , and PBMCs were isolated by density-gradient centrifugation using Ficoll-Paque Plus solution ( Amersham Biosciences , Uppsala , Sweden ) . NK cells were phenotypically identified as CD3-CD56+ cells by flow cytometry , as previously described [21 , 22] . NK cells were isolated using CD56 MicroBeads ( Miltenyi Biotec , Bergisch Gladbach , Germany ) . The purity of CD3-CD56+ cells was greater than 95% as analyzed by flow cytometry . For sorting of CD69+ and CD69- NK cells , PBMCs were stained with APC-conjugated anti-CD3 , FITC-conjugated anti-CD56 , PerCP-conjugated anti-CD45 , and PE-conjugated anti-CD69 mAb , and were sorted to obtain CD45+CD3-CD56+CD69+ and CD45+CD3-CD56+CD69- NK cells using a FACS Aria I sorter ( BD Biosciences , Mountain View , CA ) at purities of > 98% . IFN-γ and TNF-α expression in NK cells were detected by intracellular cytokine flow cytometry as previous described [20] . Briefly , freshly-isolated PBMCs ( 1 × 106/well ) were incubated in 1 mL complete media . The media consisted of RPMI 1640 , 2 mM L-glutamine , 100 units/mL of penicillin , and 100 μg/mL of streptomycin . It was supplemented with 10% fetal bovine serum ( FBS; Gibco BRL , Grand Island , NY ) . The incubation period was 24 hours in the presence of IL-12 ( 50 ng/mL; Miltenyi Biotec , Bergisch Gladbach , Germany ) and IL-18 ( 50 ng/mL; Medical and Biological Laboratories , Woburn , MA ) . For intracellular cytokine staining , we added 10 μL of brefeldin A ( GolgiPlug; BD Biosciences , San Diego , CA ) . The final concentration of brefeldin A was 10 μg/mL . After incubation for an additional four hours , cells were stained with APC-Cy7-conjugated anti-CD3 , PE-conjugated anti-CD56 and APC-conjugated anti-CD69 mAbs for 20 minutes at 4°C , fixed in 4% paraformaldehyde for 15 minutes at room temperature , and permeabilized with Perm/Wash solution ( BD Biosciences ) for 10 minutes . Cells were then stained with FITC-conjugated anti-IFN-γ and PE-Cy7-conjugated anti-TNF-α mAbs for 30 minutes at 4°C and analyzed by flow cytometry . Isolated PBMCs or purified NK cells and K562 cells ( CCL-243; ATCC ) were used as effector and target cells , respectively . Cytotoxicities of PBMCs and NK cells were evaluated by flow cytometry at an effector-to-target ( E:T ) cell ratio of 20:1 and 4:1 , respectively , as previously described [17 , 21] . Briefly , isolated PBMCs and purified NK cells were cocultured with K562 cells for 4 hours . Mixed effector and target cells were stained with FITC-conjugated anti-CD45 mAb for 20 minutes at 4°C , then washed once in phosphate buffered saline ( PBS ) . They were afterwards resuspended in 0 . 5mL of PBS containing 20 μL of 1 μg/mL propidium iodide ( Becton Dickinson ) , and then incubated for 15 minutes at room temperature . A flow cytometry determined the ratio of dead K562 cells . Luminex assay was performed according to the manufacturer’s instructions . Briefly , 25 μL of each supernatant from Quantiferon tubes was thawed and analyzed undiluted to determine the concentrations of cytokines , including IFN-γ , IL-17 , and TNF-α , by Luminex assay using Milliplex MAP human Cytokine/Chemokine panel ( Millipore , Billerica , CA ) on a Bio-Plex 200 system with Bio-Plex Manager Software ( version 4 . 1 . 1; Bio-Rad , Hercules , CA ) . Plasma levels of IL-12p40 and IL-18 were measured using a commercially available ELISA kit ( R&D Systems Inc , Minneapolis , MN ) according to the manufacturer’s instructions . Degranulation of NK cells in response to K562 cells was determined by flow cytometry as previously described [23 , 24 , 25] . Briefly , freshly isolated PBMCs were stained with FITC-conjugated anti-CD107a or isotype control mAb and then incubated with or without K562 cells at an E:T ratio of 20:1 . After 1 hour , monensin ( GolgiStop; BD Biosciences , San Diego , CA ) and brefeldin A were added , and the cells were incubated for an additional 4 hours at 37°C in 5% CO2 . After the incubation , the cells were stained with PerCP-conjugated anti-CD3 and PE-conjugated anti-CD56 mAb for 20 minutes at 4°C , fixed in 4% paraformaldehyde for 15 min at room temperature , and analyzed by flow cytometry . All comparisons of percentages and absolute numbers of NK cells , expression levels of CD69 and CD107a in NK cells , and cytotoxicity were performed by analysis of covariance after adjusting for age and sex using Bonferroni correction for multiple comparisons . Expression levels of IFN-γ and TNF-α in NK cells were analyzed using unpaired t-test . The Mann-Whiney U test was used to compare plasma levels of cytokines in scrub typhus patients versus age- and sex-matched HCs . Linear regression analysis tested associations between NK cell levels and clinical or laboratory parameters . A Wilcoxon matched-pairs signed rank test compared changes in NK cell levels and activation according to disease activity . P values less than 0 . 05 were considered statistically significant . SPSS version 18 . 0 software ( SPSS , Chicago , IL ) performed the statistical analysis . GraphPad Prism version 5 . 03 software ( GraphPad Software , San Diego , CA ) performed graphic works . The clinical and laboratory characteristics of scrub typhus patients are summarized in Table 1 . A total of 56 scrub typhus patients were included in this study . According to disease severity criteria that is based on number of organ dysfunctions , 33 patients ( 58 . 9% ) had mild disease; 14 patients ( 25% ) had moderate disease; and 9 patients ( 16 . 1% ) had severe disease . Flow cytometry determined the percentages and absolute numbers of NK cells in the peripheral blood samples of 56 patients with scrub typhus and 56 HCs . NK cells were defined as CD3-CD56+ cells ( Fig 1A ) . Percentages of circulating NK cells were significantly higher in scrub typhus patients than in HCs ( median 36 . 3% versus 19 . 8% [p < 0 . 0005] ) ( Fig 1B ) . Absolute numbers of NK cells were calculated by multiplying NK cell fractions by total lymphocyte numbers ( per microliter of peripheral blood ) . Patients with scrub typhus had significantly higher absolute numbers of NK cells than HCs ( median 507 . 6 cells/μL versus 380 . 8 cells/μL [p < 0 . 05] ) ( Fig 1C ) . Based on the relative expression of the surface marker CD56 , NK cells can be subdivided into CD56bright and CD56dim cells , as they exhibit different phenotypical and functional characteristics [26] . Previous studies have revealed that NK cell percentages , especially the proportion of CD56dim NK cell subset , were increased in multiple organ failure syndrome after trauma and metastatic melanoma [27 , 28] . To determine whether increased NK cell numbers in scrub typhus were the consequence of the increased proportion of CD56dim NK cell subset , the ratios of CD56bright and CD56dim NK cell subsets were measured by flow cytometry . No significant difference was observed in the ratio of CD56bright/CD56dim NK cell subsets between scrub typhus patients and HCs ( Fig 1D ) . We used a regression analysis to investigate the correlation between NK cell percentages in the peripheral blood and other laboratory or clinical parameters ( Table 2 ) . The aim was to evaluate the clinical relevance of NK cell levels in 56 patients with scrub typhus . In the univariate linear regression analysis , circulating NK cell percentages positively correlated with age , leukocyte count , neutrophil count , and disease severity ( p = 0 . 012 , p = 0 . 001 , p = 0 . 001 , and p = 0 . 008 , respectively ) . Meanwhile , circulating NK cell percentages inversely correlated with serum protein level and serum albumin level ( p = 0 . 003 and p = 0 . 004 , respectively ) . We observed no significant correlation between NK cell percentages and lymphocyte count , hemoglobin level , platelet count , total bilirubin level , aspartate aminotransferase level , lactate dehydrogenase level , C-reactive protein level , or the erythrocyte sedimentation rate ( Table 2 ) . To determine whether NK cells were activated during infection by O . tsutsugamushi , we used flow cytometry to examine the expression of CD69 in circulating NK cells from 31 scrub typhus patients and 18 HCs . Percentages of CD69+ NK cells were significantly higher in scrub typhus patients than in HCs ( median 14 . 9% versus 2 . 9% [p < 0 . 005] ) ( Fig 2A and 2B ) . Scrub typhus patients had significantly higher absolute numbers of CD69+ NK cells than HCs ( median 122 . 7 cells/μL versus 8 . 2 cells/μL [p < 0 . 0001] ) ( Fig 2C ) . A variety of cytokines can regulate NK cell functions , including activation , cytokine release , and cytotoxicity . Next , we measured plasma levels of IFN-γ , IL-17 , and IFN-γ-inducing cytokines , such as IL-12 , IL-18 and TNF-α , in 25 patients with scrub typhus and 15 age- and sex-matched HCs using Luminex assay and ELISA . Scrub typhus patients had significantly higher plasma IFN-γ levels than HCs ( median 43 . 8 pg/mL versus 6 . 3 pg/mL [p < 0 . 001] ) ( Fig 3A ) . Plasma levels of all IFN-γ-inducing cytokines , including IL-12 , IL-18 , and TNF-α , were found to be significantly higher in scrub typhus patients than in HCs ( medians: 236 . 4 pg/mL versus 33 . 1 pg/mL [p < 0 . 0001]; 2536 pg/mL versus 503 pg/mL [p < 0 . 0001]; and 63 . 8 pg/mL versus 8 . 9 pg/mL [p < 0 . 005] , respectively ) ( Fig 3C , 3D and 3E ) . However , plasma IL-17 levels were comparable between scrub typhus patients and HCs ( Fig 3B ) . Based on our observation that IFN-γ-inducing cytokines were increased in scrub typhus patients , we next examined whether NK cells might be activated after stimulation with IL-12 and IL-18 . We cultured PBMCs obtained from 3 HCs with IL-12 and IL-18 for 24 hours . We then determined the CD69+ cell levels in NK cells by flow cytometry . Percentages of CD69+ NK cells were significantly higher in IL-12- and IL-18-treated cultures than in untreated cultures ( mean ± SEM 26 . 2 ± 4 . 6% versus 2 . 2 ± 0 . 6% [p < 0 . 05] ) ( Fig 4A and S1A Fig ) . Furthermore , to determine whether the production of IFN-γ by NK cells might be linked to the activation of NK cells , we measured the expression of IFN-γ in the CD69+ and CD69- NK cell populations of HCs in the presence of IL-12 and IL-18 for 24 hours at the single-cell level by intracellular cytokine flow cytometry . Percentages of IFN-γ+ NK cells were found to be significantly higher in CD69+ NK cell population than in CD69- NK cell population ( mean ± SEM 25 . 5 ± 3 . 2% versus 12 . 6 ± 3 . 8% [p < 0 . 05] ) ( Fig 4B and S1B Fig ) . NK cells are a critical component of the innate immune response because of their capacity to produce a variety of cytokines . Among the most prominent cytokines produced by NK cells are IFN-γ and TNF-α [29] . To investigate the expression of these cytokines in NK cells , we incubated PBMCs obtained from five patients with scrub typhus and 10 HCs for 24 hours in the presence of IL-12 and IL-18 . We then examined the expressions of IFN-γ and TNF-α in the NK cell populations at the single-cell level by intracellular cytokine flow cytometry . Percentages of IFN-γ+ NK cells were higher in scrub typhus patients than in HCs ( mean ± SEM 29 . 1 ± 12 . 0% versus 6 . 9 ± 1 . 6% [p < 0 . 05] ) ( Fig 4C and 4D ) . Similar results were obtained even in an experiment calculated by the absolute numbers of IFN-γ+ NK cells ( mean ± SEM 125 . 3 ± 65 . 3 cells/μl versus 14 . 2 ± 5 . 2 cells/μl [p < 0 . 05] ) ( S2 Fig ) . However , all of the scrub typhus patients and HCs exhibited low levels of TNF-α+ NK cells , which were comparable between the two groups ( S3 Fig ) . To examine the cytotoxic effect of NK cells on K562 cells , we used PBMCs and purified NK cells obtained from 20 patients with scrub typhus and 30 HCs . Cytotoxicities of PBMCs and purified NK cells were evaluated by flow cytometry at an effector-to-target ( E:T ) cell ratio of 20:1 and 4:1 , respectively . The cytotoxicities were found to be comparable between scrub typhus patients and HCs ( Fig 5A and 5B ) . Upon stimulation with K562 cells , the CD107a expression in NK cells was also comparable between scrub typhus patients and HCs ( Fig 5C ) . Based on our observation that the expression level of IFN-γ was higher in CD69+ NK cells than in CD69- NK cells , we hypothesized that CD69+ NK cells could have an enhanced capacity to kill K562 cells . Thus , we determined cytotoxicities of purified CD69- and CD69+ NK cell subsets in scrub typhus patients by flow cytometry . However , we observed no significant difference in cytotoxicity between CD69- and CD69+ NK cell subsets in scrub typhus patients ( Fig 5D ) . We observed that circulating NK cell levels and CD69 expressions increased in scrub typhus patients; thus , we sought to determine circulating NK cell levels and CD69 expressions in the active and remission phases of the illness . Active and remission phases were defined as the period from the onset of symptoms to the start of antibiotic therapy on admission and the resolution of all presenting symptoms of scrub typhus after antibiotic treatment , respectively . Eleven scrub typhus patients were available for follow-up examination . As shown in Fig 6A , no significant changes in NK cell levels were found according to disease activity . However , CD69 expression was found to be significantly reduced when the disease was in remission than when it was active ( median 3 . 5% versus 15 . 1% [p < 0 . 005] ) ( Fig 6B ) . This is the first study to measure the level and function of NK cells in scrub typhus patients and to examine the clinical relevance of NK cell levels . The present study showed that circulating NK cell levels , together with elevated expression levels of CD69 and IFN-γ , increased in scrub typhus patients . Increased percentages of circulating NK cells reflected disease severity . Moreover , the plasma levels of IFN-γ , IL-12 , IL-18 and TNF-α were significantly higher in scrub typhus patients than in HCs . In particular , with stimulation of NK cells with IFN-γ-inducing cytokines ( i . e . , IL-12 and IL-18 ) , CD69 expression in NK cells was found to be increased , and CD69+ NK cells produced more IFN-γ than CD69- NK cells . Elevated CD69 expression in the active phase was normalized in the remission phase . However , the cytotoxicity and degranulation ability of NK cells were comparable between scrub typhus patients and HCs . Taken together , these findings suggest that cytokine-mediated activated NK cells accelerate the production of IFN-γ in scrub typhus patients . Our results showed that percentages and absolute numbers of total NK cells in peripheral blood increased in scrub typhus patients . This result is substantiated by previous studies that found increased circulating NK cell levels in various human acute viral infections [30 , 31 , 32] , whereas NK cell levels in peripheral blood have been reported to be decreased in influenza or HBV infection [33 , 34] . However , little is known about the circulating NK cell levels in human bacterial infection as compared with those in viral infection . Giamarellos-Bourboulis et al . have reported increased circulating NK cell levels in Gram-negative severe sepsis in humans [35] , whereas others have reported a decline in NK cells in severe sepsis or septic shock [36 , 37 , 38] . Interestingly , only one study reported no evidence of change in the frequency of total NK cells in scrub typhus patients [39] . These discrepant findings about circulating NK cell levels might be due to the differences in pathogen , severity , time point at which the sample was obtained , and cohort selection bias including age and sex studied . In particular , age and sex are well-known confounding factors affecting NK cell levels in humans [40 , 41] . In the present study , comparative analyses showed that NK cell levels were compensated by age and sex , and that NK cell levels were still significantly higher in scrub typhus patients . Furthermore , elevated NK cell numbers have been found to be a direct consequence of induced proliferation in humans infected with hantavirus [31] . When considered together , these findings suggest that the numerical increase in NK cells might be due to NK cells proliferating in response to O . tsutsugamushi infection . Further study would be needed to confirm whether increased NK cell numbers are due to proliferation or recruitment to the blood . Our data showed that CD69 expression and IFN-γ production in circulating NK cells were increased in O . tsutsugamushi infection , consistent with previous studies using several intracellular bacterial pathogens , including Rickettsia conorii , Haemophilus ducreyi , Salmonella typhimurium , Mycobacterium tuberculosis , M . bovis BCG , and Listeria monocytogenes [18 , 29 , 42 , 43 , 44] . It has been relatively well established that NK cell activation is closely related to cellular crosstalk with APCs such as dendritic cells ( DCs ) and macrophages as well as endothelium stimulated or infected by bacterial pathogens [29] . Based on the previous mechanistic study on NK cell activation by Haemophilus ducreyi [42] , we speculated that O . tsutsugamushi-infected APCs possibly promote activation and IFN-γ production of NK cells through secretion of IL-12 and IL-18 . The notion is supported by our observation that plasma levels of IL-12 and IL-18 were increased in scrub typhus patients and that addition of IL-12 and IL-18 upregulated CD69 expression in NK cells . Further study is required to determine whether APCs and NK cells form conjugates during cocultures with O . tsutsugamushi . In addition , Kang et al . have reported that the induced production of IFN-γ in NK cells affected the production of inducible nitric oxide synthase ( iNOS ) in APCs [45] , which might contribute to intracellular bacterial killing in Listeria infection . Collectively , these findings suggest that increased production of IFN-γ by activated NK cells via cellular crosstalk with O . tsutsugamushi-infected APCs contribute to intracellular bacterial killing in scrub typhus . NK cells kill virally-infected or malignant-host cells by degranulation of granzymes and perforin [46 , 47] . In several previous studies , cytotoxicity of NK cells against bacteria-infected APCs was found to be increased [18 , 43] . Our data showed that cytotoxicity and degranulation of NK cells against malignant cells ( e . g . , K562 ) were comparable between scrub typhus patients and HCs , suggesting that O . tsutsugamushi infection does not enhance the cytotoxic effects of NK cells . We observed a positive correlation between circulating NK cell levels and age , leukocyte count , neutrophil count , and disease severity . This positive correlation has also been described in Epstein-Barr virus infection [48] , whereas a negative correlation has been reported in influenza [49] . As shown in S1 Table , the univariate linear regression analysis showed that the absolute numbers of total NK cells were also positively correlated with leukocyte count and lymphocyte count ( p = 0 . 008 and p = 0 . 001 , respectively ) , but lost its correlation with age , neutrophil count , total protein level , albumin level , and severity . However , little is known about the dynamics of circulating NK cells related to disease severity in bacterial infection as compared with viral infection . Interestingly , CD69+ NK cell levels were also found to be positively correlated with leukocyte count , neutrophil count , and disease severity ( S2 Table and S4 Fig ) . Taken together , these findings indicate that the activated subset rather than total NK cells contribute more to the correlation between the absolute numbers of NK cells and severity or inflammatory parameters in scrub typhus . Considering the previous observation obtained using animal models that overzealous activation of NK cells was related to organ damage , regardless of the infection itself [50 , 51] , there is a possibility that an increased level of NK cells might be a cause of severe scrub typhus infection . However , this notion was contradicted by our data which demonstrated that increased CD69+ NK cell levels in the acute phase of scrub typhus infection were recovered to the normal levels in the remission phase , implicating that NK cell activation and expansion might be a consequence of severe scrub typhus infection . In this study , increase in NK cell number levels did not change over our observation time , which was a relatively short time interval . This result is consistent with data from our previous study on mucosal-associated invariant T cell levels . In that previous study , mucosal-associated invariant T cell levels also did not change over the same observed time period [20] . A long-term follow-up study could determine whether circulating NK cell levels recover to normal levels after scrub typhus infection . Such a follow-up study would be in line with other previous studies , which found that normalization of NK cell levels in acute viral infections took a relatively long time [48 , 49] . In summary , the present study demonstrates that circulating NK cells are activated and numerically increased , and they produced more IFN-γ in patients with scrub typhus . In addition , increased NK cell levels reflect disease severity .	Orientia tsutsugamushi is an obligate intracellular bacterium . It primarily invades endothelial cells , macrophages , monocytes , and dendritic cells . Plasma concentrations of interferon ( IFN ) -γ , several cytokines and chemokines , which are known to recruit natural killer ( NK ) cells and T cells , were found to be increased in scrub typhus patients . NK cells are known as essential immune cells against several pathogens . In murine models of Rickettsial infection , the clearance of bacteria was found to be significantly associated with NK cell activity . Not much is known regarding NK cells’ role in O . tsutsugamushi infection in humans . This study is very possibly the first to measure NK cells’ level and function of in scrub typhus patients , or to examine NK cell levels’ clinical relevance . This study’s results demonstrate that circulating NK cells are activated and numerically increased in scrub typhus patients . Notably , increased production IFN-γ by NK cells of scrub typhus patients suggests their contribution to enhancement of intracellular bacterial killing in infected antigen presenting cells . Moreover , disease severity corresponded to increased NK cell levels . These findings importantly suggest that NK cells play a role in protecting the host against O . tsutsugamushi infection .	[ "Abstract", "Introduction", "Methods", "Results", "Discussion" ]	[ "blood", "cells", "flow", "cytometry", "typhus", "medicine", "and", "health", "sciences", "innate", "immune", "system", "immune", "cells", "immune", "physiology", "cytokines", "pathology", "and", "laboratory", "medicine", "body", "fluids", "immunology", "toxicology", ...	2017	Increased level and interferon-γ production of circulating natural killer cells in patients with scrub typhus
The origin of Plasmodium falciparum , the etiological agent of the most dangerous forms of human malaria , remains controversial . Although investigations of homologous parasites in African Apes are crucial to resolve this issue , studies have been restricted to a chimpanzee parasite related to P . falciparum , P . reichenowi , for which a single isolate was available until very recently . Using PCR amplification , we detected Plasmodium parasites in blood samples from 18 of 91 individuals of the genus Pan , including six chimpanzees ( three Pan troglodytes troglodytes , three Pan t . schweinfurthii ) and twelve bonobos ( Pan paniscus ) . We obtained sequences of the parasites' mitochondrial genomes and/or from two nuclear genes from 14 samples . In addition to P . reichenowi , three other hitherto unknown lineages were found in the chimpanzees . One is related to P . vivax and two to P . falciparum that are likely to belong to distinct species . In the bonobos we found P . falciparum parasites whose mitochondrial genomes indicated that they were distinct from those present in humans , and another parasite lineage related to P . malariae . Phylogenetic analyses based on this diverse set of Plasmodium parasites in African Apes shed new light on the evolutionary history of P . falciparum . The data suggested that P . falciparum did not originate from P . reichenowi of chimpanzees ( Pan troglodytes ) , but rather evolved in bonobos ( Pan paniscus ) , from which it subsequently colonized humans by a host-switch . Finally , our data and that of others indicated that chimpanzees and bonobos maintain malaria parasites , to which humans are susceptible , a factor of some relevance to the renewed efforts to eradicate malaria . Malaria infections have influenced the development of human civilizations , and have shaped the genetic make-up of current human populations . There are four globally distributed Plasmodium protozoan parasites that are responsible for malaria in humans ( P . falciparum , P . vivax , P . malariae and P . ovale ) . Molecular phylogenetic analyses have demonstrated that these four parasites are not monophyletic [1] , [2] , indicating that they independently colonised hominids [3]–[6] . The timing of their appearance in Homo sapiens , however , remains unresolved . This is of some importance to current efforts to control malaria , because it will affect how observed patterns of genetic diversity in the parasite populations are interpreted . For example , several evolutionary genetic approaches rely on reliable phylogenetic information to detect putative adaptive genetic variation , thereby identifying genes that might be involved in pathogenesis or in the evasion of host immune responses . Addressing these issues is a matter of great importance for P . falciparum , the parasite responsible for a substantial proportion of the global malaria mortality and morbidity [7] . It is now generally accepted that P . falciparum underwent a population expansion in humans [4] , [6] , [8]–[11] , though how , when and from where humans first acquired P . falciparum , is less well established . Suggestions of a host-switch from a chimpanzee parasite received recent support , albeit without resolving the likelihood or timing of this event [4] , [10] , [12] . The accuracy and robustness of conclusions derived from comparative analyses ( phylogenetic or genomics ) will be significantly enhanced if data from all of the evolutionary close parasites were to be included . In the context of parasites of humans , this data would be best obtained from Plasmodium species that infect our nearest relatives , the African Apes , because two of the parasite species , P . reichenowi and P . rodhaini , that have been reported in Pan and Gorilla are morphologically very similar to P . falciparum and P . malariae respectively , while the third , P . schwetzi , corresponds to P . vivax or P . ovale [13] , [14] . Studies of the malaria parasites of African Apes have been limited to few observations made mainly in the 1920s–1950s , and very little is known of their natural history . Nonetheless , it is known that chimpanzees are susceptible to infection by the four parasite species of humans , while humans have been infected with P . rodhaini and P . schwetzi [13] , [14] . The origin and evolutionary history of the malaria parasites in chimpanzees and gorillas are speculative [13] , [14] mainly because the molecular data has been restricted to sequences derived from a single P . reichenowi isolate [3] , [4] , [8] , [15] , [16] until very recently [12] . In another recent publication , a novel parasite lineage close to , but distinct from , P . reichenowi was reported from chimpanzees sampled in Gabon [17] . This raises the important question as to whether Plasmodium species close to P . falciparum , other than the two described so far , occur in non-human higher primates . We were afforded a rare opportunity to analyze blood samples collected independently from chimpanzees and bonobos for the presence of Plasmodium parasites . Such a collection of fresh isolates would provide sequence data for improved phylogenetic analyses . Here we report on our findings of a genetically diverse set of Plasmodium parasites found in some of the samples we analyzed , and we discuss the insights they have provided into the origin of the Plasmodium falciparum . Blood samples were obtained from 49 chimpanzees , Pan troglodytes , in Uganda and the Democratic Republic of the Congo ( DRC ) , and from 42 bonobos , Pan paniscus , in the DRC . Blood smears were not made available , so the presence and level of Plasmodium parasites were assessed solely by a highly sensitive PCR assay , where a small fragment of the small subunit ribosomal RNA ( ssrRNA ) genes is amplified using oligonucleotides that target sequences conserved in all known Plasmodium species [18] . Parasites were detected in 18 animals: 3/3 Pan t . schweinfurthii living wild in Kibale National Park in Uganda , and in 3/8 Pan t . troglodytes and 12/42 Pan paniscus cared for in sanctuaries in the DRC . Parasitaemias were quite low ( <100 parasites per µl of blood ) , consistent with previous observations of naturally infected apes [13] , [14] . We opted to conduct our analyses on the DNA purified directly from the blood samples , because whole genome amplification could lead to artefactual recombination between DNA molecules from different strains or species of parasites , should any be present in a given sample . Given the low parasite densities in the samples and the limited blood volumes available , efforts were directed at characterizing a small number of genes that have been used in recent phylogenetic analyses . Specifically , we targeted the mitochondrial genome using oligonucleotide primers that correspond to sequences conserved in Plasmodium . Since we were particularly interested in lineages related to P . falciparum , we used oligonucleotides based on sequences from P . falciparum to target two nuclear genes: dihydrofolate reductase-thymidylate synthase ( dhfr-ts ) , and the gene encoding the merozoite surface protein 2 ( msp2 ) because this gene is not known to have orthologues outside P . falciparum and P . reichenowi [15] . We specifically targeted the block 3 of msp2 , because we hypothesized that the extensive polymorphisms observed for this region in P . falciparum might also occur in orthologous genes that could be present in closely related species , and this could provide an indication of genetic diversity in these parasites . In order to minimize artefacts , nearly all the sequences obtained for the dhfr-ts and the msp2 block 3 fragments were derived from duplicate amplifications . The mitochondrial genome sequences were also derived from duplicate amplification of a single 5800 bp fragment , which spans nearly the complete mitochondrial genome of ca . 6 kb . This avoided any ambiguities in a final assembly of overlapping fragments that might arise from a sample with multiple parasite lineages . Indeed , it was not possible to combine the dhfr-ts , msp2 and mitochondrial data sets in the subsequent phylogenetic analyses , because mixed infections were common in our samples . Finally , we are confident that cross-contamination during amplification was highly unlikely because similar sequences for the different chimpanzee parasite lineages were derived from samples collected independently in Uganda or the DRC , and then processed in France or in the USA , respectively . Successful amplification was not achieved for all the genes targeted from each sample , and this was particularly noted for the samples from the bonobos . Nonetheless , the sequence data obtained revealed a rich diversity of species and strains ( Table S1 ) , in particular for the individual samples collected from the two Pan troglodytes subspecies . Sixteen near-complete mitochondrial genomes that coalesce in six distinct lineages were obtained from 12 of the 18 samples positive for Plasmodium ( Fig . 1 ) . All our phylogenetic analyses lead to identical topologies ( see Methods ) , and only the Bayesian phylogenetic tree is reported ( Fig . 1 ) . Two lineages shared a recent common ancestor either with the P . malariae clade ( two bonobos ) or with the P . vivax clade ( one chimpanzee from Uganda and one from the DRC ) . Another lineage , found in the bonobo samples , clustered with P . falciparum . One lineage from a DRC chimpanzee shared a recent common ancestor with P . reichenowi , while the two remaining lineages found in chimpanzees sampled in Uganda and the DRC , were novel and formed a monophyletic group with those of P . falciparum and P . reichenowi . For the sake of clarity , we have used the name Laverania to refer to this monophyletic clade , a generic name previously proposed to distinguish P . falciparum and P . reichenowi from the other malaria parasite species ( International Commission on Zoological Nomenclature , Opinion 283 ) . We hypothesized that the two new lineages in the Laverania clade correspond to two distinct Plasmodium species . This hypothesis was further supported by three other analyses . First , the extent of divergence in the genetic distances between these two novel Laverania lineages , as calculated from the mitochondrial genomes ( Table 1 ) , is comparable to that observed between well-established species in the rodent malaria clade , or between P . falciparum and P . reichenowi . Second , the topology of the phylogenetic tree constructed using dhfr-ts sequences from the same isolates reproduces that obtained for the mitochondrial genome ( Fig . 2 ) . Indeed , it would appear that an insert coding for eight amino acids is specific to the Laverania lineages ( P . falciparum , P . reichenowi and the two new lineages ) , which further supports our conclusion that these lineages form a monophyletic group . Finally , the samples that harboured the two novel lineages and the P . reichenowi lineage , yielded msp2 block 3 sequences that could be grouped into five distinct allelic families , of which one was similar to that previously published for P . reichenowi ( Fig . 3 ) , while the other four were novel . By a way of comparison , only two allelic families have been identified for the P . falciparum msp2 block 3 despite extensive sampling . Six of the eight bonobos positive for Plasmodium , harboured parasites that yielded sequence data for dhfr-ts and/or msp2 . The msp2 and all the dhfr-ts sequences were indistinguishable from known P . falciparum sequences . This confirmed that bonobos were infected with P . falciparum , as had been indicated by the sequences of the mitochondrial genomes derived from four of these six bonobos ( Fig . 1 ) . Interestingly , we found significant differences in the genetic diversity of the P . falciparum mitochondrial lineages derived from bonobos as compared with that previously noted for large set of mitochondrial P . falciparum lineages obtained from human isolates collected worldwide [9] . Indeed , the P . falciparum lineages in bonobos ( n = 4 , π = 0 . 0048 ) were ten times more diverse that those found in humans ( n = 96 , π = 0 . 00034 ) . Furthermore , there were no fixed differences between the P . falciparum from bonobos and those from humans . In other words , the four mitochondrial P . falciparum haplotypes we obtained from the bonobos had each a distinctive set of mutations such that none of these haplotypes were represented in the extensive P . falciparum mitochondrial haplotype database . This is clearly illustrated in the mitochondrial genome haplotype network ( Fig . 4 ) . The P . falciparum populations from bonobos and from humans , though related , have undergone some level of differentiation . Moreover , the haplotype network indicates that the four haplotypes from the bonobo do not form a monophyletic group , which suggests a scenario where bonobos and humans exchanged parasites in relatively recent times . The sum of our knowledge on the Plasmodium parasites of African Apes derives from observations , nearly all made before the 1960s , on fewer than 50 naturally infected animals captured primarily in Cameroon , Sierra Leone or the Congo . Given the highly protected status of African Apes , prospects to extend this knowledge are restricted to molecular analyses of blood samples , mainly collected during medical examination of Apes cared for in sanctuaries , or upon recovery from poachers or villagers . The results from three such surveys published this year [12] , [17] , [19] have provided new glimpses into the diversity of malaria parasites in chimpanzees , and have allowed testing of hypotheses concerning the evolution of P . falciparum [12] , [17] , [19] . Here we present the outcome of two further independent surveys , one of which is distinguished by the inclusion of samples from bonobos and from wild-living chimpanzees . The molecular data we present demonstrate that the Pan genus naturally harbours a rich Plasmodium fauna , including two novel lineages close to P . falciparum , one related to P . vivax , and one related to P . malariae . Furthermore , it brings to light the presence of a population of P . falciparum in bonobos that appears to differ from those in humans . The observations add new perspectives to the evolutionary hypotheses formulated for the Plasmodium parasites of African Great Apes and humans . From a parasitological point of view , the fact that the three samples collected from Eastern Chimpanzees ( Pan t . schweinfurthii ) living wild in a community of 44 animals , were all positive and harboured complex mixed strain/species infections ( Table S1 ) , suggests that prevalence of infections under natural conditions of transmission is high . This view is supported by our observations of a similar level of parasite diversity in three of the eight Central Chimpanzees ( Pan t . troglodytes ) that were independently sampled in the DRC ( Table S1 ) . It would be interesting to establish whether the other two chimpanzee subspecies , the Western Chimpanzee ( Pan t . verus ) and the Nigeria-Cameroon Chimpanzee ( Pan t . vellerosus ) also harbour the same parasite species . The bonobos cared for in a sanctuary also had high parasite prevalence , with Plasmodium detected via ssrRNA amplification in 12 of the 42 sampled ( 28 . 5% ) . The parasites related to P . vivax-like found in chimpanzees from the DRC and Uganda might correspond to the chimpanzee parasite P . schwetzi . Preliminary evidence from partial dhfr sequences obtained for the chimpanzees we sampled in Uganda suggests that these parasites could be related to P . vivax ( data not shown ) . Unfortunately , at present a P . schwetzi isolate is not available for comparative molecular analysis . Whether this species in Pan results from a past host switch from humans into chimpanzee , or whether it corresponds to P . vivax parasites recently reported in Equatorial Africa [20] , [21] , remains a matter of speculation . It might be that the dynamics of P . vivax and related species in African hominids , including humans , are more complex than previously thought . The quartan malaria parasites , P . brasilianum in South American primates and P . rodhaini in the chimpanzee , have long been considered to be strains of P . malariae [13] , [14] . Thus , it was interesting that the mitochondrial genomes of the parasites related to P . malariae found in two bonobos conform a sister clade and carry a six nucleotide insert that has not been observed for P . malariae or the South American parasite P . brasilianum . This could indicate that the parasites in bonobos might correspond to P . rodhaini , a species that would then be distinct rather than synonymous with P . malariae . Confirmation that this might indeed be the case awaits further molecular data from a larger set of P . malariae lineages from humans and Apes . Three parasite lineages related to P . falciparum were found in both the chimpanzees collected from DRC and those collected from Uganda . One of these lineages clearly corresponds to P . reichenowi . We propose that the other lineages may represent two distinct Plasmodium species . Given the data from the near-complete mitochondrial genome sequences , and the support from dhfr-ts and msp2 sequences , we consider it reasonable to ascribe specific status to the parasites in the two novel lineages observed in chimpanzees . We propose to name the parasites of one of the novel lineages Plasmodium billcollinsi Krief et al . n . sp . , and those of the other Plasmodium billbrayi Krief et al . n . sp . , in honour of the distinguished malariologists William E . Collins and “Bill” Robert Stow Bray ( 1923–2008 ) , respectively . The type material would be the mitochondrial genome sequences ( holotype and paratype ) , with a distribution in Uganda and the DRC in Pan t . troglodytes and P . t . schweinfurthii as hosts . While we were finalizing this manuscript for submission , a publication describing a novel lineage related to P . falciparum was reported from two Pan troglodytes sampled in Gabon [17] . Based on mitochondrial DNA sequences , the authors have also proposed that this lineage be considered a new species , P . gaboni [17] . When the mitochondrial sequence submitted for P . gaboni was compared with the mitochondrial sequence presented here , it could be concluded that P . gaboni and P . billbrayi shared a recent common ancestor ( Fig . S1 ) . However , the differences were of sufficient importance ( e . g . P . gaboni has a unique insert ) to lead us to consider P . gaboni as a possible other additional member of the Laverania clade . Nonetheless , this assessment is at present mitigated by the fact that the contiguous mitochondrial sequence provided for the K isolate of P . gaboni ( GenBank Accession No FJ895307 ) was assembled from discontinuous fragments that were amplified separately , hence the unavoidable gaps . Furthermore , if the animal from which the sample was obtained harboured a mixed infection , as did many of the chimpanzees that we sampled , the different fragments used for assembly might have originated from different species or lineages . Consequently , we opted not to consider the P . gaboni mitochondrial sequence in our phylogenetic analyses until such a time that the mitochondrial sequence from this lineage is confirmed , a view also adopted by Rich et al . [12] . We are aware that the validity of a species described only by sequences of one or more genes is open to debate , as this does not conform to current acceptable criteria . It would have been desirable to obtain some morphological data to provide a classical description of a novel species . The description of a new Plasmodium species is classically made after microscopic examination of Giemsa-stained infected erythrocytes , most often showing all asexual and sexual developmental stages . In some cases , it is necessary to examine the form of the parasite in the insect vector and/or during the hepatic stages , while for others differentiation from known species requires establishing one or more biological characteristics such as host specificity , the course of infection , or the ability to breed true . In the case of Plasmodium parasites that infect highly protected hosts ( such as chimpanzees , gorillas and orang-utans ) invasive sampling is highly restricted . On rare occasions it is possible to obtain a blood sample , but experimental infections of such animals are now nearly universally legally proscribed . Thus , the likelihood to obtain the morphological and biological data required to define and name a novel Plasmodium species for such hosts is highly remote . Furthermore , the presence in a single sample of multiple species would make it difficult to derive reliable conclusions from observations of a few blood smears . This is further exacerbated when parasite levels are low because this restricts microscopic examination to a few forms in thick smears where parasite morphology is poorly preserved . In our case , the six chimpanzees we sampled had low parasite loads , and four of them had mixed species infections . Had we had the opportunity to examine blood smears , a crescent-shaped gametocyte distinctive of P . falciparum and P . reichenowi might have been observed , but it would not have been possible to ascribe it with any degree of confidence to any one of the lineages detected by PCR amplification . Therefore , in the case of blood dwelling protozoan parasites of African Apes or other protected species , molecular data become the only accessible and reliable taxonomic features . In our study , we have considered that the phylogenetic analysis and genetic diversity comparisons based on the near-complete mitochondrial genomes , combined and supported with similar data from two nuclear genes , provided sufficient grounds to propose the description of two new species . The fact that similar sequence analyses correctly predict the specific status of well-established Plasmodium species ( Fig . 1 and Table 1 ) , adds to our confidence in the validity of P . billbrayi and P . billcollinsi as bona fide species . We nonetheless consider that it would be worthwhile for the community to agree on standardized parameters derived from defined molecular data that could serve to describe Plasmodium species for which no morphological or biological data are likely to become available . The findings we present in this manuscript advocate a reappraisal of current views on the evolution and origin of P . falciparum . When it was thought that P . reichenowi and P . falciparum were unique among all primate malaria parasites , two hypotheses for the origin of P . falciparum as a parasite of humans were considered: co-speciation in their respective hosts , or a host switch followed by independent evolution . Grounds for favouring one hypothesis over the other shifted with time , as the weight of evidence that could support one hypothesis over the other was limited , principally by the availability of only a single P . reichenowi isolate . Recent analyses of data from parasites sampled from eight chimpanzees provided clear support for the host-switch scenario [12] . The data we present further support this finding and provide a more detailed account of the events leading to the origin of P . falciparum as a parasite of humans . When the tree topologies derived from the dhfr-ts and mitochondrial sequences ( Fig . 1 & Fig . 2 ) are considered , the most parsimonious interpretation is that P . falciparum belongs to a monophyletic group of malarial parasites that have evolved in African Apes . We proceeded to estimate the divergence time of the most recent common ancestor for the Laverania clade . We agree that the use of molecular clocks is not without pitfalls , even when good time points can be used for calibration [22] , [23] . In the particular case of parasitic organisms , an assumption of some level of host specificity ( though not necessarily co-speciation ) is needed in order to use host evolution for estimating the parasite mutation rates . Therefore , we estimated times of divergence of the mitochondrial sequences using models that allow the use of relaxed molecular clocks [24] . Although the Homo/Pan divergence time has been commonly used as a point of calibration for the falciparum-reichenowi divergence ( e . g . [9] , [17] ) , we excluded it in order to avoid circularity in the analyses . Thus , we estimated the mutation rates under two previously used scenarios: a ) , the Plasmodium spp . currently found in macaques radiated with their primary hosts , the genus Macaca [5] , and b ) P . gonderi , a parasite from African monkeys , and macaque parasites co-diverged when Macaca branched from other Papionina [25] . It is worth noting that neither of these two time points requires co-speciation ( i . e . where specific malarial parasites co-speciate with specific non-human primate lineages generating phylogenies with identical topologies ) , but simply that several malarial parasites started their radiation with a major groups of non-human primates allowing for extensive host-switches . Such timeframes can be estimated even in the absence of good phylogenetic trees [26] . It is interesting that our time estimates ( Table 2 ) that did not use the Homo-Pan divergence as a calibration point , were not substantially different from those estimated by others [9] , [17] who used the P . falciparum - P . reichenowi divergence assuming co-speciation with Homo-Pan . The estimates of the divergence times for the Laverania clade members ( Table 2 ) indicated that all the four lineages might have originated between 6 . 0 and 19 million years ago ( Mya ) . Regardless of the wide confidence interval , this time frame is consistent with the origin of the genus Pan , but it clearly indicated that the Laverania lineages may have started to diverge long before the divergence Pan-Homo [27] . In addition , the phylogeny clearly indicates that the human parasite , P . falciparum , is the only Homo parasite among several Pan species in the Laverania clade . Given the phylogeny , a Pan host appears as an ancestral characteristic of the lineage . Therefore , when both phylogenies and estimated times of divergence are considered , a co-evolutionary origin of P . falciparum as a parasite of humans can be confidently excluded . Consequently the hypothesis that P . falciparum originated as a result of a host-switch between humans and Apes becomes favoured . However , our data indicate more complex scenarios that can only be addressed when data from multiple isolates of the parasite lineages currently present in both the hosts involved are included in the analyses . The mitochondrial haplotype map ( Fig . 4 ) provides evidence that the sub-population of four P . falciparum parasites in bonobos were genetically more diverse that of the extensive P . falciparum population in humans available to date . The most parsimonious interpretation of this line of evidence is that P . falciparum originated as a human parasite via a host-switch from Pan paniscus . When the human P . falciparum mitochondrial sequences alone are considered , our estimate of the time to the most recent common ancestor ( TMRCA ) was 78 , 000–330 , 000 years ago . While we cannot rule out that the available sample of P . falciparum mitochondrial genomes properly represent the genetic diversity of the species , this time frame is consistent with one expected for a parasite expanding early in human history . However , when considered together , the two distinct P . falciparum populations of humans and bonobos are estimated to have diverged from other members of the Laverania clade between 1 . 0 and 3 . 1 Mya . This timeframe coincides with the divergence of bonobo from the common chimpanzee [28] , [29] . The estimated TMRCA of 0 . 4 to 1 . 6 Mya for the P . falciparum found in bonobos coincides with the origin of bonobos [28] . Taken together , our analyses indicate that P . falciparum , as a species , has long been associated with Pan paniscus and only subsequently switched into humans . The topology of the mitochondrial haplotype network ( Fig . 4 ) is consistent with this interpretation and suggests that few lineages expanded in the human population after this event . The parasites we obtained over a short period from a single bonobo community probably constitute a biased sample set . A reliable estimate of the timing for the host-switch and the number of times this event might have taken place would require the inclusion of sequences from a larger set of P . falciparum parasites from bonobos from diverse locations . Assuming that there was no sampling bias with respect to the P . falciparum populations collected by others from humans , the limited data from bonobo parasites we present here can be most conservatively interpreted to support a single switching event , though it does not allow excluding multiple events . It is also possible that host switching still occurs today in areas where humans and bonobos are in close epidemiological contact . The presence of double or triple mutations associated with resistance to pyrimethamine in the four dhfr sequences obtained for the P . falciparum of bonobos is consistent with this , because these mutations are common in P . falciparum collected in 2008 from residents around Kinshasa [30] . At present , we cannot rule out the possibility that these dhfr mutations might have been selected independently in bonobos during the three months treatments with Bactrim™ ( trimethoprim + sulfamethoxazole , two drugs that target the same enzymes of the folate pathway as the antimalarial combination of pyrimethamine and sulfadoxine ) to which apes in the sanctuary were occasionally subjected . Finally , it could be speculated that the parasites in bonobos and in humans have recombined sexually . The scenario we propose for the origin of P . falciparum in humans differs in several respects from a very recently formulated hypothesis that proposed that this species originated from a single transfer of P . reichenowi from chimpanzees to humans [12] . These conclusions were based on the analysis of the genetic diversity and tree topologies derived from fragments of the mitochondrial cytochrome b gene ( 528 bp ) , the apicoplast caseinolytic protease ( 316 bp ) , and the nuclear small subunit ribosomal RNA gene ( 371 bp ) , obtained from eight Plasmodium-infected chimpanzees ( three from Pan t . verus , and five from Pan t . troglodytes ) . One assumption was that these sequences were derived from a single parasite specie , P . reichenowi , found in Pan troglodytes sp . This was a fair supposition to make since these short sequences did not provide sufficient resolution to distinguish their lineages from that of the only known P . reichenowi isolate . However , when these partial cytochrome b sequences are compared to the homologous region in the mitochondrial genomes that we obtained , there are clear indications that some might correspond to P . reichenowi , but also that most cluster either with the P . billbrayi or the P . billcollinsi lineages reported here ( Fig . S1 ) , which differ to such an extent from P . reichenowi that they could be considered as distinct species . Indeed this is evident on examination of the topology and branch lengths in the phylogenetic tree presented for the cytochrome b fragment ( see Fig . 4 of [12] ) , where the eight isolates cluster into three groups removed from P . reichenowi . Our data provides evidence of a contrasting and more complex evolutionary scenario where P . falciparum evolved as a species in bonobos ( Pan paniscus ) where it was one of at least four parasite species that radiated in the genus Pan before it switched into humans . The infections of bonobos by P . falciparum were not associated with any overt clinical signs , nor would the levels of parasitaemia have allowed detection by microscopy , suggesting a state of chronic malaria typical of infections in natural hosts . This is consistent with previous observations , including some made on splenectomised chimpanzees with high parasite levels [13] , [14] , in which chimpanzees experimentally infected with various parasite species including P . falciparum showed few clinical signs whether at peak parasitaemias or during the subsequent lengthy chronic infections [13] , [14] . This minor impact on the health of chimpanzees was recently supported by the failure to detect a signature of positive selection in their G6PD genes , despite a long association with Plasmodium parasites [31] . The contrasting parasitological and clinical evolutions of P . falciparum in its two hosts , humans and bonobos , which have highly similar genomes , provides an excellent opportunity for comparative genomic studies to uncover the genetic or molecular basis for its higher virulence in humans . Such knowledge could be exploited to devise novel approaches to reduce the substantial global morbidity and mortality burdens . It is likely that bonobos , in which we have found significant numbers to be naturally infected with P . falciparum or P . malariae , are also susceptible to infections by P . ovale and P . vivax , as is the case for chimpanzees [13] , [14] , [19] . One can now , therefore , justifiably explore whether bonobos and chimpanzees could act as a reservoir for all Plasmodium species that afflict humans . The potential impact of zoonotic malaria transmission on human health has been recently exemplified by a stable focus of potentially lethal P . knowlesi , a benign parasite of macaques , in inhabitants of Malaysian Borneo [32] , [33] . Such a possibility has not been considered for sub-Saharan Africa . A zoonotic reintroduction of malaria into communities that live in hyperendemic areas is likely to be of little consequence . However , this would hinder efforts to eradicate malaria and might possibly lead to epidemic foci in formerly malarious regions whose inhabitants have lost immunity acquired against malaria . Furthermore , humans have been shown to be susceptible to infection by two of the parasite species of African Apes ( P . rodhaini and P . schwetzi ) [13] , [14] , and the meagre data available does not exclude the possibility that humans can be infected by P . reichenowi or the two new species we describe here . Using the sequence data we obtained from chimpanzee parasites , it will now be possible to seek these parasites in groups of humans that are in contact with African Apes . In conclusion , the data gathered from a limited molecular analysis of a modest number of chimpanzee blood samples have not only significantly added to our knowledge of Plasmodium in our closest relatives , bonobos and chimpanzees , but also provided tantalizing insights into the evolutionary history of the malaria parasites of humans . We urge the scientific and the wildlife conservation communities to devote some resources to archive the parasites of Great Apes , which are at present likely to remain only amenable to molecular investigations , and to develop in vitro and/or ex-vivo methods to preserve and maintain them . These studies might provide novel approaches that could help control and eventually eradicate pathogens that have long exacted devastating global health , economic and social burdens . For all samples , genomic DNA was extracted from aliquots of 200 µl of whole blood using the Qiagen DNeasy Blood and Tissue Kit ( Qiagen , Germany ) , and the DNA obtained resuspended in 200 µl of buffer . Blood smears were not available for microscopic examination , thus parasite levels were estimated using PCR analysis of a 10-fold serial dilution series of the DNA purified from the positive samples . The nested PCR detection assay used was based on the small subunit ribosomal RNA gene ( ssrRNA ) , using oligonucleotide primers that were specific to , and conserved in , all known Plasmodium species [18] . This established that the parasite burdens in these animals were very low ( <10–100 parasite per µl of blood , and in one case <1000 parasites per µl ) . Approximately 5 , 800 bp ( out of 6 , 000 ) of the parasites' mitochondrial genome were amplified using the oligos Forward 5′-GAGGATTCTCTCCACACTTCAATTCGTACTTC and Reverse 5′-CAGGAAAATWATAGACCGAACCTTGGACTC with Takara LA Taq™ Polymerase ( TaKaRa Takara Mirus Bio ) , ( 1 cycle 94°C for 1 min , then 30 cycles of 94°C for 30 sec and 68°C for 7 min− 1 cycle 72°C for 10 min ) . PCR products were cloned in the PGem®-T vector ( Promega ) . In the case of the mitochondrial genome , we report sequences deposited in GenBank ( Accession numbers are in parentheses following species name ) for the Asian macaque parasites P . inui strain Taiwan II ( GQ355483 ) , P . inui strain Leaf Monkey II ( GQ355482 ) , and for P . brasilianum ( GQ355484 ) from South American primates . Other sequences were reported in other studies: P . inui Mulligan ( AB354572 ) , P . fieldi ( AB354574 ) , P . simiovale ( AB434920 , AY800109 ) , P . knowlesi ( NC_007232 ) , P . cynomolgi ( AY800108 ) , P . fragile ( AY722799 ) and P . coatneyi ( AB354575 ) ; P . hylobati ( AB354573 ) from gibbons , P . simium ( AY800110 ) , P . gonderi from African monkeys ( AY800111 ) , and the parasites of humans P . ovale ( AB354571 ) and P . malariae ( AB354570 ) . Additional information about these species , including their description , basic biology , geographic distribution and host-range can be found elsewhere [13] . Additional sequences of Plasmodium mitochondrial genomes were obtained from the GenBank ( Accession numbers are in parentheses following species name ) : the avian malarial parasites P . gallinaceum ( NC_008288 ) , P . juxtanucleare ( NC_008279 ) , and P . relictum ( AY733088–AY733090 ) ; the rodent malarial parasites P . yoelii ( M29000 ) , P . berghei ( AF014115 ) , P . chabaudi ( AF014116 ) ; the non-human primate malarial parasite P . reichenowi ( NC_002235 ) ; the human malarial parasite P . falciparum ( AY282930 ) and P . vivax ( AY598140 ) . The avian parasite Leucocytozoon sabrazesi ( NC_009336 ) was used as outgroup . The gene encoding dihydrofolate reductase-thymidylate synthase ( dhfr-ts ) from P . falciparum or related species in samples collected from chimpanzees in Uganda and bonobos in the DRC was obtained as two overlapping fragments amplified by nested PCR using the following primer pair for the primary reaction: Pfdhfrts-F 5′-ATGATGGAACAAGTCTGCGACGTTTTCG and Pfdhfrts-R 5′-GCAGCCATATCCATTGAAATTTTTTCATG , ( 2 . 5 mM Mg2+ , annealing at 58°C ) The two separate secondary reactions were initiated with 1 µl of the product from the primary reaction using the following primer pairs Pfdhfrts-F and Pfdhfrts-NR 5′-GGGAAATATTGACTTAAATCAAATTTC ( 1 . 5 mM Mg2+ , annealing at 58°C ) that amplifies the fragment coding for the DHFR and linker domains , or Pfdhfrts-NF 5′-CAAAGTGATCGAACGGGAGTAGGTG and Pfdhfrts-R ( 3 . 5 mM Mg2+ , annealing at 58°C ) that amplifies the fragment encoding the TS domain . All reactions were initiated with 1 µl of template ( equivalent to ca . 1 µl of whole blood ) in a total reaction volume of 40 µl ( final concentrations of 125 µM dNTP , 250 nM of each oligo , and 2 units/100 µl AmpliTaq polymerase ) , with the following cycling conditions: 95°C for 5 min , then 30 cycles of 2 min annealing ( see above for temperatures used for each primer set ) , 2 min extension at 72°C and 1 min denaturation at 94°C , after a final annealing step followed by a 5 min extension step , the reaction temperature was brought down to 25°C before storage at −20°C . The gene encoding the dhfr-ts from parasites related to P . falciparum in samples collected from chimpanzees in the DRC , was amplified using the primers: Forward 5′-ATGATGGAACAAGTCTGCG and Reverse 5′-TTAAGCAGCCATATCCATTG . The PCR conditions were: a partial denaturation at 94°C for 3 min and 35 cycles with 1 min at 94°C , 1 min at 53°C–55°C and 2 min extension at 72°C , a final extension of 10 min was added in the last cycle . Aligning dhfr-ts sequences among distantly related species of Plasmodium was difficult due to several insertions-deletions . We performed two analyses , one including only P . falciparum-like sequences on 1789 bp and a second including P . gallinaceum ( AY033582 ) , P . chabaudi ( M30834 ) , and P . yoelii ( XM_719562 ) with only 1690 bp . The fragment encoding the block 3 polymorphic domain of merozoite surface protein 2 ( msp2 ) from P . falciparum or related species in samples collected from chimpanzees in the Uganda and bonobos in the DRC was by nested PCR amplification using the following primer pairs: primary reaction M2-P1 5′-GAAGGTAATTAAAACATTGTC and M2-P2 5′-GAGGGATGTTGCTGCTCCACAG , and a secondary reaction were initiated with 1 µl of the product from the primary reaction using M2-N1 5′-CTAGAACCATGCATATGTCC and M2-N2 5′-GAGTATAAGGAGAAGTATG . All reactions were initiated with 1 µl of template in a total reaction volume of 40 µl ( final concentrations of 1 . 0 mM Mg2+ , 25 µM dNTP , 250 nM of each oligo , and 2 units/100 µl AmpliTaq polymerase ) , with the following cycling conditions: 95°C for 5 min , then 30 cycles of 30 sec annealing at 50°C , 1 min extension at 72°C and 30 sec denaturation at 94°C , after a final annealing step followed by a 5 min extension step , the reaction temperature was brought down to 25°C before storage at −20°C . In the majority of cases these sequences were derived from two or more independent amplifications . All the sequences obtained and reported here were submitted to GenBank ( Accession numbers and the corresponding gene fragments are presented in the Table S1 ) . Initial Neighbor Joining ( NJ ) trees were inferred under Tamura-3P model of nucleotide substitution [34] in Mega4 [35] . Maximum likelihood ( ML ) search of a tree topology was implemented in PAML4 [36] under a General Time Reversible ( GTR ) + I + Γ4 substitution model , chosen based on likelihood ratio tests [37] , and employing the NJ method to generate an initial tree . Bayesian support for the nodes was inferred in MRBAYES [38] , under a General Time Reversible ( GTR ) + I + Γ4 substitution model , using 4 Markov chains and 10 , 000 , 000 Markov Chain Monte Carlo ( MCMC ) steps , discarding the first 3 , 000 , 000 steps ( 30% ) as a burn-in . Sampling was performed every 500 generations . Mixing of the chains and convergence was properly checked after runs . The recovered ML and Bayesian trees were identical . Although a total of eight distinct near-complete mitochondrial genomes were obtained from the parasites found in the bonobos , we stringently excluded any where the accuracy of the sequence obtained was not optimal , thus only 4 sequences were included in the phylogenetic and other analyses . The mutation rates that have been widely used in Plasmodium evolutionary genetic studies have used the Homo/Pan divergence time as a point of calibration for the falciparum-reichenowi divergence ( for e . g . [9] , [39] ) . However , using such rates will make whatever argument we put forward about the origin of P . falciparum and P . reichenowi circular . Thus , in order to avoid tautological arguments , we estimated mutation rates by considering time of divergence under two scenarios: i ) the Plasmodium currently found in macaques radiated with the genus Macaca [5] , which allows the estimation of a substitution rate of 2 . 83E-09 subs/site/year; ii ) assuming that P . gonderi and macaque parasites co-diverged when Macaca branched from other Papionina [25] , which allows the estimation of a mutation rate of 5 . 07E-09 . It is worth noting that these mutation rates were not particularly off other estimates obtained for Plasmodium mitochondrial genomes ( for e . g . [9] ) indicating that , at least as first approximations , these scenarios are reasonable . We employed a Bayesian approach with a relaxed clock [40] as implemented in BEAST [24] . The estimations of times of divergence for the clades of interest were performed by running 4 independent runs of 10 , 000 , 000 Markov Chain Monte Carlo ( MCMC ) steps after discarding the first 30% of the steps as burn-in , and sampling being performed every 1 , 000 steps . Previous runs showed that this burn-in was sufficient for the chains to reach stationary distribution . For the relaxed version of the clock we assumed a lognormal distributed clock for the mutation rate , with an average mutation rate according to each scenario mentioned in the previous paragraph , under a Yule prior for the simulation of the lineages during tree reconstruction . Results of the runs were analyzed with Tracer v1 . 4 [41] and estimates of average divergence times and confidence intervals were recovered . We checked the adequate mixing of the MCMC chains for each run in and the effective sample size of the estimates , making sure that all of them were above 100 . The runs were combined in Tracer to generate the final estimates of time of divergence and their 95% confidence intervals . The following sequences were submitted to the GenBank: Near-complete Plasmodium mitochondrial genomes from parasites of chimpanzees , bonobos and other primate hosts GQ355468–GQ355486; msp2 block 3 from parasites collected from Pan t . schweinfurthii ( Uganda ) GU075719–GU75726 , and from parasites collected from Pan t . troglodytes ( DRC ) GU131994–GU131995; dhfr-ts sequences from parasites collected from Pan t . troglodytes ( DRC ) GQ369532–GQ369536; P . falciparum msp2 block 3 sequences from bonobo samples GU075709–GU075718; P . falciparum dhfr-ts partial sequences from bonobo samples GQ859592–GQ859595 ) .	Chimpanzees and gorillas are known to have malaria parasites ( genus Plasmodium ) similar to those that infect humans . It is likely that detailed molecular studies of these parasites will help understand important aspects of the malaria disease and of immune defences in humans , and could then guide the development of novel control measures . However , few studies of parasites in African Apes have been conducted to date . Here we present the results of a survey of malaria parasites in chimpanzees and bonobos , our closest relatives . In chimpanzees , we identified two new parasite species closely related to P . falciparum , the most dangerous of the parasites in humans . We also found that bonobos harbour malaria parasites including P . falciparum . Phylogenetic analyses of these parasites strongly suggested that P . falciparum evolved in bonobos , and that it was introduced into humans from bonobos at a later date . Overall , our findings have substantially altered our perception of the origin of malaria parasites in humans .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "infectious", "diseases/protozoal", "infections", "evolutionary", "biology/microbial", "evolution", "and", "genomics", "microbiology/microbial", "evolution", "and", "genomics", "microbiology/parasitology" ]	2010	On the Diversity of Malaria Parasites in African Apes and the Origin of Plasmodium falciparum from Bonobos
Transcription , replication , and repair involve interactions of specific genomic loci with many different proteins . How these interactions are orchestrated at any given location and under changing cellular conditions is largely unknown because systematically measuring protein–DNA interactions at a specific locus in the genome is challenging . To address this problem , we developed Epi-Decoder , a Tag-chromatin immunoprecipitation-Barcode-Sequencing ( TAG-ChIP-Barcode-Seq ) technology in budding yeast . Epi-Decoder is orthogonal to proteomics approaches because it does not rely on mass spectrometry ( MS ) but instead takes advantage of DNA sequencing . Analysis of the proteome of a transcribed locus proximal to an origin of replication revealed more than 400 interacting proteins . Moreover , replication stress induced changes in local chromatin proteome composition prior to local origin firing , affecting replication proteins as well as transcription proteins . Finally , we show that native genomic loci can be decoded by efficient construction of barcode libraries assisted by clustered regularly interspaced short palindromic repeats and CRISPR-associated protein 9 ( CRISPR/Cas9 ) . Thus , Epi-Decoder is an effective strategy to identify and quantify in an unbiased and systematic manner the proteome of an individual genomic locus by DNA sequencing . The chromatin at any given location in the genome is a dynamic entity that likely involves the interaction of many different proteins and nonprotein factors . The essential and complex processes of transcription , replication , and repair require major chromatin rearrangements to access and use the genome [1 , 2] . Furthermore , the genome’s chromatin is under the influence of signals that can relay information of cellular events or states to the genome and vice versa [3 , 4] . Fully understanding the chromatin regulatory mechanisms in the cell will require comprehensive knowledge of the full set of proteins that bind at individual genomic loci . Although many chromatin factors have already been identified by genetics and protein–protein interaction studies , direct , unbiased , and comprehensive analyses of chromatin interactions at specific genomic loci has remained a major challenge [5 , 6] . A commonly considered strategy towards solving this problem is introducing an affinity handle at a locus of interest , purifying the locus ( capture ) , and analysing the copurified proteins by mass spectrometry ( MS ) [7 , 8] . However , a major challenge with capture MS is the need for very high levels of enrichment of the locus of interest versus the rest of the large genome while at the same time obtaining sufficient amounts of material for comprehensive and quantitative MS analysis [5 , 6] . For example , in a model organism with a small genome such as yeast , purification of a 1 kb locus with its associated proteins from an entire genome of approximately 30 Mb requires a 30 , 000-fold purification . Here , we present an independent approach , Epi-Decoder , with which the interactome of a single-copy locus is determined by DNA sequencing instead of MS . In this approach developed in yeast , a library of clones is generated in which each clone harbours at least one DNA barcode at a fixed locus and one protein tagged at its endogenous locus with a common epitope tag . Following chromatin immunoprecipitation ( ChIP ) for the common tag on a pool of cells , the barcodes of the coimmunoprecipitated DNA are counted by high-throughput sequencing , and the amount of barcode recovered serves as a read-out for the amount of the corresponding specific protein binding at the barcoded locus . In contrast with proteins , DNA barcodes can be amplified prior to counting , boosting the sensitivity of detection . Epi-Decoder of a single transcribed locus in yeast identified more than 400 chromatin-interacting proteins , enabled the identification of differences in protein binding between the 5’ and 3’ end of a gene , and demonstrated a chromatin rewiring in response to physiological changes . We also demonstrate that Epi-Decoder can be applied to a locus of interest by harnessing clustered regularly interspaced short palindromic repeats and CRISPR-associated protein 9 ( CRISPR/Cas9 ) -mediated genome engineering . Thus , Epi-Decoder provides an efficient method to provide a comprehensive map of the dynamic proteome of a single-copy locus . Moreover , it is an orthogonal approach to capture MS because acquisition of quantitative and qualitative information on protein binding does not involve MS but is obtained by DNA sequencing . To develop a strategy for systematic and comprehensive decoding of the interactome of a single genomic locus , we asked whether DNA barcode technology can be harnessed to solve this challenging proteomics problem . We and others previously showed that short DNA barcodes ( ≤20 bp ) integrated in the genome and embedded in chromatin can serve as molecular identifiers of the chromatin state they are in [9–11] . Building on that notion , we used synthetic genetic array ( SGA ) technologies [12 , 13] to create arrayed yeast libraries that allow for high-throughput and direct assessment of chromatin states of many cell clones in parallel . In these libraries , each clone contains a pair of known unique barcodes flanking a constitutively transcribed kanamycin resistance gene ( KanMX ) under control of the Ashbya gossypii ( Ag ) TEF1-promoter and -terminator at the HO locus [14] as well as a known Tandem Affinity Purification ( TAP ) -tagged protein expressed from its endogenous locus [15] ( Fig 1A ) . The Epi-Decoder libraries used here cover approximately 4 , 000 TAP-tagged proteins ( S1 Table ) of the approximately 5 , 600 proteins encoded in the yeast genome [16] . Following pooling of the barcoded TAP-tagged strains , cells are incubated with formaldehyde to crosslink proteins to DNA and subjected to ChIP ( Fig 1B ) . From the coimmunoprecipitated DNA and the input samples , the barcoded regions are amplified ( S1 Fig ) , and the barcodes are identified and quantified by parallel sequencing ( Fig 1B ) . In this set-up , which we call Epi-Decoder , the abundance of a barcode ( ChIP/input ) reports on the crosslinking of the corresponding TAP-tagged protein to its own barcoded region for every TAP-tagged clone in the pool ( Fig 1C ) . The reporter gene at the HO locus is flanked by 2 barcodes that can be analysed in parallel but that have different chromatin contexts: the upstream barcode ( BC_UP ) is located in the promoter region , whereas the downstream barcode ( BC_DN ) lies in the terminator region and in close proximity to an origin of replication ( Fig 1C ) . Thus , Epi-Decoder analysis of the barcoded KanMX gene at the HO locus results in an inferred binding score for the vast majority of the proteins present in the yeast genome at 2 proximal but distinct genomic loci . For BC_UP and BC_DN , we identified the immunoprecipitated barcodes with significantly different counts compared to background ( red dots in Fig 2A ) , based on 6 biological replicates ( S2A Fig ) . The vast majority of significantly different barcodes showed a positive binding score ( Fig 2A , red dots , ChIP/input > 0 ) . Here , we refer to the factors associated with those barcodes as binders . Together , we identified 469 binders of which 18 were specific for BC_UP and 273 for BC_DN ( Fig 2B , S1 Table ) . Significantly depleted factors ( Fig 2A , red dots , ChIP/Input < 0 ) were not expected because proteins cannot bind less to their barcodes than negative controls such as nonexpressed or nontagged proteins . The low number of significantly depleted factors indicates a low rate of false discovery , in agreement with our stringent cut-off ( false discovery rate [FDR] < 0 . 01 ) . Among the binders , the 4 canonical histone proteins ( represented by 7 histone genes in the library ) were among the most enriched proteins at both barcodes . Furthermore , 267 out of the 469 binders had gene ontology ( GO ) terms related to DNA binding ( Fig 2C , S2B Fig ) , confirming that Epi-Decoder provides an effective approach for identifying chromatin-interacting factors . In addition to known DNA-interacting proteins , we found a substantial number of factors involved in RNA processing and cellular metabolism . RNA processing events such as capping , splicing , cleavage , and polyadenylation are all processes known to occur in close conjunction with transcription and hence in proximity to DNA [17] , providing an explanation for the recovery of barcodes associated with RNA processing proteins . The presence of factors involved in cellular metabolism cannot be explained by general associations with transcription , although recent studies have suggested various roles of metabolic enzymes in the nucleus [4 , 18 , 19] ( and see Discussion ) . Given the fact that some of these factors are known to be highly expressed , we investigated the possibility that their chromatin interactions were determined by protein abundance . Overall , chromatin binders were generally more highly expressed than nonbinders ( Fig 2D ) . However , high protein expression level alone was not sufficient for binding—many chromatin binders were not highly abundant , and many highly abundant proteins were nonbinders . For example , ribosomal proteins are highly abundant , but most of them were not detected as chromatin interactors even though ribosomal proteins traffic through the nucleus to form ribonucleoprotein complexes for ribosome assembly . Furthermore , for a selected panel of factors reflecting different classes of binders and different expression levels , we could quantitatively confirm the positive and negative Epi-Decoder results by ChIP-quantitative PCR ( qPCR ) and immunoblot analysis of individual clones ( Fig 2E , S2C and S2D Fig ) . Finally , we excluded the possibility that leaking of cytosolic factors into the nucleus due to the presence of methanol in formaldehyde solutions was a cause of the crosslinking of metabolic enzymes to chromatin . Using formaldehyde with or without methanol as a stabilizer resulted in nearly identical binding profiles ( S2E Fig ) . Whether or not the binding of metabolic enzymes to chromatin is biologically significant will require further in-depth studies and functional experiments . Together , our results suggest that the barcode counts obtained by Epi-Decoder accurately and quantitatively report on the efficiency of protein crosslinking at the barcoded loci and capture histones and other core chromatin proteins , as well as additional factors . We note that in the Epi-Decoder set-up , proximity to DNA , genomic distance from the barcode , protein abundance , and cell-to-cell variation are among the factors that can influence the measured quantitative binding scores . To determine whether Epi-Decoder can be used to identify locus-specific proteins , we next compared the significantly enriched factors for BC_UP to those of BC_DN . For this purpose , we plotted the binding scores for BC_UP versus BC_DN and color-coded enriched factors according to the main chromatin functions or complexes expected at this locus ( Fig 3A ) . We observed that proteins within the same complex or process tend to cluster together . This strongly suggests that Epi-Decoder does not just detect binding events but also provides quantitative information about protein occupancy . Several factors and complexes were shared between the 2 locations . Besides histones , strongly enriched factors at both barcodes include subunits of RNA polymerase II and several other factors and complexes with a well-known role in transcription elongation such as DSIF , Elf1 , FACT , Spt6 , and the PAF-C complex ( Fig 3A ) [20 , 21] . However , BC_UP and BC_DN also showed substantial quantitative interactome differences reflecting their different functional states and suggesting that many proteins show locus-specific binding behaviour ( Fig 3B ) . At BC_UP at the 5’ end of the gene , histone variant H2A . Z ( Htz1 ) was strongly enriched , which is in agreement with the enrichment of H2A . Z in promoter regions [22] , which we confirmed for this locus using an antibody that recognizes the endogenous H2A . Z protein ( S3A Fig ) . In addition , BC_UP showed enrichment for general regulatory factors Reb1 and Rap1 . This is in agreement with the known Rap1 binding site in the TEF1 promoter of the KanMX gene [23] and the role of Reb1 as a general chromatin organizer of regulatory regions [20] . BC_UP also showed enrichment for the basal transcription factors TFIID , -E , -F , and -H , which form the pre-initiation complex together with RNA polymerase II [21 , 24] . At the 3’ end , BC_DN showed enrichment for factors involved in transcription termination and mRNA cleavage and polyadenylation , an important step in mRNA 3’ end formation [25] . The Cohesin complex was also enriched at BC_DN . Finally , we observed strong binding of the evolutionarily conserved origin recognition complex ( ORC ) and minichromosome maintenance ( MCM ) complexes at BC_DN , which is in agreement with the proximal origin of replication sequence at which ORC and then MCM are loaded to assemble the prereplication complex [26 , 27] . BC_DN-specific binding of Orc1 was confirmed by ChIP-qPCR ( S2C Fig ) . Thus , in addition to common binders , Epi-Decoder revealed locus-specific enrichment of factors ( Fig 3C ) . This confirms that , while the barcodes are in the same genomic region , they are in distinct functional contexts that can be separated by this approach even though they are only 1 . 5 kb apart . This level of resolution was robust because the BC_UP- and BC_DN-specific binding pattern was maintained when the average chromatin fragment size was increased or decreased by applying shorter and longer shearing time , respectively ( S3B and S3C Fig ) . Not all factors followed the expected distribution ( Fig 3A ) . Tfa2 , the small subunit of the heterodimeric basal transcription factor TFIIE , and Ssl2 , the double-stranded DNA ( dsDNA ) translocase subunit of the basal transcription factor TFIIH , were found at BC_UP but were also strongly enriched at BC_DN . 3’ binding of Tfa2 and Ssl2 has been observed at other genes as well [24 , 28] , but the significance remained uncertain . Because in Epi-Decoder all proteins have the same tag and are assessed simultaneously in a pool in a quantitative manner , the deviant binding pattern observed here cannot be explained by antibody issues or experimental and strain differences and therefore strongly suggests that Tfa2 and Ssl2 have special roles or positions within their complex or have noncanonical functions outside their complex . The chromatin interactome defined by Epi-Decoder confirmed that the barcodes flank an actively transcribed gene in close proximity to and transcribing towards an origin of replication that is licensed with the prereplicative complex . This conformation is of special interest because , upon firing of the origin , replication fork progression will require negotiation with the transcription machinery , potentially leading to collisions [29–33] . It has been proposed that , in order to avoid collisions , RNA polymerase II is degraded at a subset of genes that is about to be replicated [34] . Because this recently described process of chromatin adaptation is still poorly understood , we used Epi-Decoder to determine , in a comprehensive and unbiased manner , how the barcoded HO locus interactome changes when cells are arrested in early S phase in hydroxyurea ( HU ) ( S4A Fig ) , the condition in which RNA polymerase II degradation has been observed [34] . HU affects replication-fork progression by reducing the supply of deoxyribonucleotide triphosphates ( dNTPs ) and by the generation of reactive oxygen species [35] . We first confirmed previous observations [36] that , in HU , early origins have fired but that the middle to late barcoded HO origin ( ARS404 ) has not fired yet ( Fig 4A ) . Despite the absence of initiation of replication , Epi-Decoder revealed multiple changes in the chromatin proteome; 40 proteins showed a significantly different binding score in HU at BC_UP , and 79 were significantly different at BC_DN—their altered abundance could not be explained by changes in protein levels ( S4B Fig ) . Firstly , we observed lower mRNA levels of the barcoded KanMX gene compared to the HphMX gene , which expresses a different mRNA ( Hph instead of Kan ) from the same AgTEF1 promoter but at a different genomic location ( HphMX replaces the CAN1 gene at chromosome V ) and not in proximity to an origin of replication ( Fig 4B ) . This was accompanied by a reduction in occupancy of multiple RNA polymerase II subunits at the barcoded gene , extending the previous findings that RNA polymerase II is degraded at a subset of genes that has been or is about to be replicated ( Fig 4C ) . Secondly , the chromatin response was not restricted to RNA polymerase II because other general transcription proteins involved in capping , initiation , elongation , and termination were also reduced , showing that the reduction of transcription involves a broad range of changes in the ( co ) transcriptional machinery ( Fig 4C ) . Thirdly , we observed additional changes that were not directly related to transcription but indicate topological alterations ( Fig 4C and S4C Fig ) . Topoisomerase II ( Top2 ) binding was strongly increased at both ends of the KanMX gene in HU . Top2 is the main enzyme releasing topological stress during S phase [37] and is targeted to nucleosome-free DNA during replication stress [38] . Pds5 , the cohesin maintenance factor was also increased at BC_DN , and a closer inspection of the cohesin complex indicates that 3 of the 4 cohesin subunits in the library showed moderately increased occupancy as well , suggesting stabilisation of cohesin binding in this region ( S2 Table ) . Finally , all the subunits of ORC in the library showed decreased binding at BC_DN , while occupancy of MCM was unaltered . ORC proteins have been suggested to remain bound to origins throughout the yeast cell cycle [39 , 40] , possibly being negatively influenced by MCM proteins in G1 [41] . The direct and quantitative comparison of all origin-proximal factors by Epi-Decoder suggests that the interaction of ORC proteins with the origin is compromised in S phase prior to firing but that this cannot be explained by increased MCM–protein occupancy . Therefore , our results show that ORC subunits interact more dynamically with chromatin throughout the yeast cell cycle than expected but are in agreement with the behaviour of ORC in other organisms [39 , 42] . How lower ORC binding influences origin firing and subsequent fork progression of this locus remains unknown . Recent in vitro reconstitution experiments demonstrated that the MCM complex is stably bound to DNA once assembled and also competent for replication , even after removal of ORC proteins , suggesting that origin firing per se may not be affected [43] . In summary , Epi-Decoder uncovered condition-dependent composition of a local chromatin proteome . At a transcribed gene next to a licensed origin , arrest in HU led to tuning down of transcription by a general reduction of transcription protein binding , which is accompanied by increased occupancy of topology factors and altered stoichiometry of replication proteins . We showed that Epi-Decoder effectively measures DNA–protein interactions at a barcoded reporter gene . Next , we aimed to establish versatility , i . e . , the ability to apply Epi-Decoder to any native genomic locus of interest . To this end , we set up a protocol that facilitates the generation of new barcoded libraries by implementation of CRISPR/Cas9 ( Fig 5A ) . In Saccharomyces cerevisiae , a double-strand break induced by targeted Cas9 is generally lethal unless it can be repaired by homologous recombination [44–47] . Therefore , we expected that transforming a strain with Cas9 , guide RNA ( gRNA ) , and a barcode-containing repair template would result in efficient barcode incorporation ( Fig 5A ) . For a proof of concept , we chose to insert a 15-bp barcode in the transcribed region of the ADE2 gene ( BC_5’-ADE2 ) , 57 bp downstream of the start of the ADE2 coding sequence . Because disruption of ADE2 causes the accumulation of a red pigment in the cell , it provided a direct visual read-out to confirm the efficiency of the barcode insertion [44 , 48] . Random colonies were selected and transferred to an arrayed format to enable crossing to the TAP-tagged library and , in parallel , to identify the integrated barcode at each location . The vast majority of the colonies turned red and contained a barcode insertion , confirming that CRISPR/Cas9 can provide a highly efficient barcoding strategy . The new BC_5’-ADE2 collection contained 1 , 974 strains with a unique barcode at the 5`end of ADE2 . It was crossed with the TAP-tagged protein collection , which resulted in 3 , 604 barcode-TAP-tagged protein combinations . A NatMX cassette integrated downstream of the ADE2 gene was used to select for the barcoded locus during the genetic crosses without interfering with the chromatin context of the barcode . Here , the Epi-Decoder screen for BC_5’-ADE2 was performed on diploid cells , in which the TAP-tagged allele and the barcoded allele are both present in a heterozygous fashion . As expected , histones and other general chromatin factors were detected ( Fig 5B and S5 Table ) . Furthermore , we detected at BC_5’-ADE2 proteins associated with active transcription such as RNAPII subunits , transcription initiation factors , and transcription elongation factors , consistent with the promoter-proximal location of the barcode . Indeed , the binding pattern of BC_5’-ADE2 was very similar to that of BC_UP , which is also in a promoter-proximal location ( Fig 5B ) . Together , this confirmed the successful application of Epi-Decoder to a new , native barcoded locus . BC_UP and BC_5’-ADE2 also showed some differences . Rap1 , which is a strong binder at the Rap1-controlled AgTEF1 promoter [23] , was not detected at the 5’ end of the ADE2 gene , which is in agreement with what is known about Rap1 binding across the genome [49] . Furthermore , transcription-initiation and -elongation proteins were generally recovered with lower scores at BC_5’-ADE2 than at BC_UP , which is in agreement with the lower transcript levels found for ADE2 versus TEF1 and AgTEF1 promoter-controlled transcripts ( S5B Fig and [50] ) . Thus , by taking advantage of efficient CRISPR/Cas9-mediated barcode library construction , Epi-Decoder can uncover the local chromatin proteomes of native loci of interest . Strategies aimed at identifying proteins bound at specific genomic loci frequently involve affinity capture combined with MS [5 , 6] . Affinity handles generally involve engineering the locus by introducing binding sites for a tagged protein , capturing a native locus by using oligo-capture , or targeting inactive Cas9-fusion proteins . Important progress has been made for multicopy DNA loci such as repetitive DNA elements [7 , 8] . Recently , oligo-capture [51] and CRISPR/Cas9 technology [52 , 53] have been applied to identify proteins at single-copy genomic loci . However , the problem of assessing a unique locus in a comprehensive and quantitative manner by MS has not been solved , and many challenges remain , a major one being the high degree of purification of the locus of interest while obtaining sufficient material for good coverage in MS analysis [6] . Epi-Decoder provides a powerful and orthogonal strategy in which the very challenging proteomics problem of decoding the chromatin proteome of a single genomic locus is addressed by DNA sequencing . We demonstrate that Epi-Decoder enables an unbiased , comprehensive , and quantitative analysis of protein occupancy of a single locus in budding yeast . In addition to known and expected core chromatin proteins , we observed significant binding scores for many proteins that do not have canonical DNA-related functions . Among the unexpected factors are several metabolic enzymes , oxidative stress response factors , the yeast ubiquitin-activating enzyme , chaperones , proteasome subunits , and RNA-processing factors ( S1 Table ) . Some of these unexpected factors are highly expressed; their binding score may reflect nonspecific binding . On the other hand , not all unexpected interactors are highly expressed or show equal binding scores . In addition , for a substantial number of the identified metabolic enzymes ( e . g . , glyceraldehyde 3-phosphate dehydrogenase/Tdh3 , pyruvate kinase/Cdc19 , homocitrate synthase/Lys20 , Lys21 ) and protein chaperones ( e . g . , the HSP70 chaperones Ssa1 and Ssa2 ) , there is evidence that they have active roles in transcription and DNA repair [54–58] . Furthermore , the binding scores of some metabolic enzymes changed upon treatment with HU ( S2 Table ) , indicating that their interaction with chromatin is not constitutive . Indeed , there is a growing body of evidence that enzymes in the cytoplasm might have moonlighting functions in the nucleus , for example , by local supply of cofactors of histone-modifying enzymes [19 , 54 , 59–62] . Alternatively , the interaction with chromatin could somehow influence the activity of metabolic enzymes . The protein interactions that we describe here for 3 barcoded locations provide a rich resource for further unravelling the biological meaning of noncanonical chromatin interactions . Epi-Decoder is more than a discovery tool . It offers the possibility to compare different loci or to analyse 1 locus under different physiological or genetic conditions in a systematic and quantitative manner . Because all proteins are examined in a pooled fashion using the same affinity handle , differences in purification and cell backgrounds can be excluded and binding scores can be directly compared . This is exemplified by the basal transcription factors Tfa2 and Ssl2 , which are quantitative outliers compared to their complex members ( TFIIE and TFIIH , respectively ) , suggesting noncanonical functions . Another example is the chromatin rewiring we observed upon treatment with HU , which uncovers large-scale quantitative changes in interactions of transcription and replication proteins as well as other factors . We note that Epi-Decoder generates independent binding scores for proteins with ( nearly ) identical protein sequences but encoded by different genes , such as the histone proteins or other protein paralogs such as Ssa1 and Ssa2 , which are 98% identical . This can provide extra information about the differential use of related proteins that cannot , or not easily , be distinguished by MS methods . Protein isoforms that arise through post-translational modification cannot be distinguished by Epi-Decoder . The Epi-Decoder strategy will be applicable to a broad range of fundamental epigenetic questions . A few requirements need to be met for successful application . First , Epi-Decoder requires a library of tagged proteins . For budding yeast , several tagged protein libraries are already available in addition to the carboxy-terminal TAP-tagged library used here [63–65] . This will allow for extending the current Epi-Decoder analyses to a nearly complete coverage of the proteome as well as to alternative ( amino-terminal ) tags , together enabling an unprecedented deep analysis of chromatin proteomes . Tagged protein libraries are also becoming available for many other organisms [66–71] . A second requirement for Epi-Decoder applications is the integration of DNA barcodes proximal to the locus of interest . Given their small size , DNA barcodes generally only minimally disrupt a locus , but this needs to be verified for the locus of interest , as is the case for strategies involving targeting affinity handles by Cas9 or other approaches . Fortunately , genomic barcoding has come within reach for many research questions due to the availability of emerging genome engineering strategies , in yeast and other organisms [46 , 47 , 72 , 73] . Indeed , we demonstrate that CRIPSR/Cas9 can be used to generate markerless barcoded DNA libraries at high efficiency for a new locus of interest [46] . Furthermore , our findings with the barcoded native ADE2 locus demonstrate that Epi-Decoder is not restricted to haploid cells but can also be used in diploid cells in which the tagged allele of a protein and the barcoded allele are present in a heterozygous state . This greatly facilitates creating Epi-Decoder libraries in yeast as well as other cell systems . In summary , Epi-Decoder is powerful and versatile strategy for decoding the protein interactome of a single genomic locus . We expect that the Epi-Decoder strategy and derivatives thereof will enable the decoding of dynamic proteomes of different loci in different organisms and will be of high value for addressing a broad range of important chromatin biology questions in future applications . Yeast media was prepared as described previously [12 , 74] . For screening in untreated conditions , yeast strains were grown in YEPD ( 1% yeast extract , 2% bacto peptone , and 2% glucose ) in log phase . S phase arrest was achieved by adding 1 volume of YEPD plus 360 mM HU to log-phase cells ( OD660 of 0 . 4 ) in YEPD to achieve 180 mM HU final concentration , and cells were harvested after 2 hours . The arrest was verified by flow cytometry . Yeast strains used in this study are listed in S6 Table . Library manipulations on solid media were performed using SGA technology [12] combined with robotics using a RoToR machine ( Singer Instruments , Watchet , UK ) . Library manipulations in liquid media to generate pools for barcode identification were performed using a Hamilton Microlab Star ( Hamilton , Germany ) . A collection of barcoded TAP-tagged strains ( library NKI4217 ) was generated as follows . In order to cross the TAP-tagged and Barcoder libraries , the TAP-tagged collection [15] was made compatible for SGA and converted to mating type α . This was done by crossing the TAP-tagged library with NKI4212 , which was derived from strain Y8205 by replacing the STE2pr-Sp_his5 cassette at the CAN1 locus by HphMX to enable selection during SGA for the TAP-His3MX6 alleles by using histidine prototrophy . This resulted in the MATα TAP-tagged collection ( NKI4214 ) divided over 4 different 1 , 536-well agar plates . Each of the TAP-tagged plates was mated with the set of 1 , 140 barcoder MATa strains containing a pair of known unique barcodes flanking a constitutively transcribed kanR marker gene ( KanMX ) under control of the Ashbya gossypii TEF1 promoter and terminator at the HO locus [14] . Diploids were obtained by G418 and Hygromycin double selection on rich media and transferred to sporulation media . After sporulation , the combination library was obtained by selecting twice for MATα haploids and then twice for MATα haploids containing a TAP tag , the barcoded KanMX cassette , and the HphMX marker . The barcodes and TAP tags were verified for a few selected strains . The effects of sonication time and the presence of methanol in formaldehyde solutions were assessed by performing 3 biological replicate analyses of the pool of plate 2 of the KanMX Epi-Decoder library . Five extra barcoded control strains ( TAP-tagged versions of Hht2 , Htb2 , Rpl13a , Ste2 , and Bar1 ) were generated by using the respective clones from the NKI4214 library and introducing unused barcodes from MATa haploid gene knockout library ( Open Biosystems , Huntsville , AL ) at the HO locus of these strains . Epi-Decoder libraries were grown on YEPD plates overnight , and the colonies of each plate were pooled together in liquid culture . The cultures were grown until log phase ( OD660 of approximately 0 . 4 ) and crosslinked for 20 minutes with one-tenth of the volume of freshly prepared Fix Solution ( 1% formaldehyde [Sigma-Aldrich 252549-500ML] , 50 mM Hepes-KOH [pH 7 . 5] , 100 mM NaCl , 1 mM EDTA ) and subsequently quenched for 5 minutes with Glycine ( 125 mM final concentration ) . Where indicated , methanol-free formaldehyde ( Thermo Scientific; 28908 ) was used for crosslinking . Cells were washed once in cold TBS with 0 . 2 mM PMSF , and the pellet was frozen at −80 °C . Cells from frozen pellets of approximately 1 . 5 × 109 cells were lysed by bead beating in 200 μL breaking buffer ( 100 mM Tris [pH 7 . 9] , 20% glycerol , protease inhibitor cocktail EDTA-free ) with Zirconia/silica beads . The lysate was washed twice in 1 mL FA buffer ( 50 mM HEPES-KOH [pH 7 . 5] , 140 mM NaCl , 1 mM EDTA , 1% Triton X-100 , 0 . 1% Na-deoxycholate , protease inhibitor cocktail EDTA-free ) and sonicated using the Bioruptor PICO ( Diagenode ) for 10 minutes at 30-second intervals . Where indicated , chromatin was sonicated for 5 or 20 minutes . Chromatin was cleared by centrifugation for 5 minutes at 4 °C at 4 , 000 rpm . The amount of 100 μL chromatin was used as input material . For ChIP of the Barcoded KanMX gene , immunoglobulin G ( IgG ) Sepharose 6 Fast Flow beads ( GE Healthcare ) were washed 3 times with PBSB ( PBS containing 5 mg/mL BSA ) and incubated with 1 mL chromatin for 6 hours on a turning wheel at 4 °C . For ChIP of the barcoded ADE2 locus , Dynabeads M-270 Epoxy beads ( ThermoFisher number 14301 ) were used . Rabbit IgG ( 0 . 2 mg ) from serum ( Sigma , I5006-100MG ) was coupled to 10 mg epoxy-activated Dynabeads in phosphate buffer with ammonium sulphate according to the manufacturer’s protocol . After a PBS wash , IgG-coupled beads were resuspended in 1 . 2 mL PBS containing 0 . 02% sodium azide . Of a prepared beads solution , 80 μL was used for 800 μL chromatin . Samples were washed twice in FA buffer , twice in high-salt FA buffer ( 500 mM NaCl ) , twice in RIPA buffer ( 10 mM Tris [pH 8] , 250 mM LiCl , 0 . 5% NP-40 , 0 . 5% Na-deoxycholate , 1 mM EDTA ) and once with TE buffer ( 10 mM Tris [pH 8] , 1 mM EDTA ) . With each wash step , the Sepharose beads were spun for 2 minutes at 3 , 000 rpm at 4 °C . IP samples were eluted for 10 minutes at 65 °C in 100 μL elution buffer ( 50 mM Tris [pH 8] , 10 mM EDTA , 1% SDS ) . IP and input samples were digested with 0 . 5 μL RNase A ( 10 mg/mL ) and 10 μL ProtK ( 10 mg/mL ) in 70 μL TE for 1 hour at 50 °C and subsequently kept overnight at 65 °C to reverse crosslinks . DNA was purified using the QIAquick PCR purification kit ( Qiagen ) . BC_UP , BC_DN and BC_5’_ADE2 were amplified separately with specific primers ( S7 Table: 2B_UP_Fw + 2B_UP_Rv_all; 2B_DN_Fw +2B_DN_Rv_all; P5-XL-ADE2_g39_REV + P7L-IDn-ADE2_g39_REV ) . PCR products were mixed in an equimolar fashion and purified from an agarose gel with a size selection of 100 to 150 bp . The purified DNA was sequenced ( single read , >50 bp ) on a HiSeq2500/MiSeq platform ( Illumina , San Diego , CA ) , using one or a mix of custom sequencing primers ( S7 Table: 2B_UP_Seq , 2B_DN_Seq or Seq-XL-ADE2_g39_REV ) . Barcodes were extracted from the sequencing reads by using the Perl script eXtracting Counting and LInking to Barcode References ( XCALIBR ) . The code and detailed descriptions of the functions are available at https://github . com/NKI-GCF/xcalibr . Briefly , XCALIBR locates the constant regions ( U2 and D2 of the barcoded KanMX gene or the constant region next to the barcode at ADE2 ) in each amplicon and reports the 6 bp upstream of the index and 20 bp downstream of the KanMX barcode or 15 bp downstream of the ADE2 barcode . The resulting table contains counts for each barcode–index combination . Counts below 10 were removed from this table before further preprocessing and filtering steps . Input ( IN ) and immunoprecipitated DNA ( IP ) of each plate were amplified separately with a unique index . Even though we aimed to mix each sample equimolarly , plate-specific differences in counts could still occur . This was corrected by normalizing each plate by its median . The counts table was log2 transformed , and barcode-index combinations were matched with the ORF names . Factors with low input counts were removed because these were likely to be missing from the library or the barcode failed to amplify due to technical reasons . We manually validated several factors for the presence of a TAP tag by PCR with primers in the specific ORF and the TAP tag . This revealed that ORC4 , MCM3 , MCM7 , and RAD6 were not properly tagged; we therefore removed these strains for further analysis . The final set of the barcoded KanMX library contained information on 3 , 994 BC_UP and 3 , 955 BC_DN barcode clones . We noticed that the dynamic range was slightly different between biological replicates . To overcome this problem , we performed quantile normalisation for the replicates of BC_UP and BC_DN separately . This method is used to generate similar distributions by first ranking each replicate based on the counts and then replacing the counts of each rank by the mean count of that rank . Factors with IP counts that were significantly enriched over input were identified by using the Limma R/Bioconductor software package [80] . p-Values were adjusted for multiple testing by converting them to FDRs using the Benjamini-Hochberg procedure . Factors with a positive fold-change and an FDR < 0 . 01 were selected as significantly enriched . For untreated conditions , 6 independent biological samples were used . The running sum scores for the enrichment plot ( Fig 3B ) were calculated with the gseaScores function from the HTSanalyzeR package . For identifying differential binders upon HU treatment , 3 biological samples were used for treatment and no treatment . Here , we only considered factors that were significant binders in at least one of the conditions ( FC > 0 and FDR < 0 . 05 ) . Limma was used to select factors with significantly different IP counts ( FDR < 0 . 05 ) . Barcode counts were compared with protein abundance data measured by GFP intensity . This data was obtained from the CYCLoPs database [81] . GO slim process terms are condensed versions of the full GO ontology [82 , 83] . GO slim process terms were downloaded from the Saccharomyces Genome Database [84] , and enrichment analysis was performed by using the fisher . test function in R with option alternative ‘greater’ . We manually assigned the following categories based on GO slim terms listed here: DNA binding: DNA-templated transcription , initiation , DNA-templated transcription , elongation , DNA-templated transcription , termination , transcription from RNA polymerase I promoter , transcription from RNA , polymerase II promoter , chromatin organisation , histone modification , DNA replication , DNA recombination , DNA repair , cellular response to DNA damage stimulus , regulation of DNA metabolic process , nuclear transport , and chromosome segregation; RNA binding or processing: mRNA processing , rRNA processing , tRNA processing , RNA modification , RNA splicing , and RNA catabolic process; and metabolism: carbohydrate metabolic process , cofactor metabolic process , nucleobase-containing small-molecule metabolic process , monocarboxylic acid metabolic process , cellular amino acid metabolic process , generation of precursor metabolites and energy , oligosaccharide metabolic process , and lipid metabolic process . ChIP experiments were performed similar to TAG-ChIP-Barcode-Seq but with 20 μl bed volume of IgG Sepharose 6 Fast Flow beads and 200 μl chromatin . For untagged-H2A . Z ChIP , the Htz1-specific antibody ( Active Motif , 39647 ) was coupled to Protein G Dynabeads ( Thermo Fisher Scientific ) . ChIP was performed with 20 μl Dynabeads and 200 μl chromatin . Factors were selected such that they reflect different classes of binders and different expression levels , in addition to negative controls Bar1 ( not expressed in these cells ) and Rpl13A ( a ribosomal subunit ) . qPCR was performed on the purified DNA with SYBR green master mix ( Applied Biosystems or Roche ) or SensiFAST SYBR master mix ( Bioline ) according to the manufacturer’s protocol and analysed on LightCycler 480 II ( Roche ) . The binding was analysed with specific primers in close proximity to the BC_UP and BC_DN ( S7 Table ) . Each sample was measured in 2 technical duplicates in the qPCR , and the average value of these 2 was taken as 1 value when combining biological replicates . For comparison of the ChIP-qPCR and BC-seq results in a quantitative manner , the negative control ( Bar1 ) was used to normalize the raw values . Flow cytometry samples were prepared to monitor cell cycle progression and verify S phase arrest by HU . A total of 1 × 107 cells were collected and fixed with 70% ethanol and stored at −20 °C . Flow cytometry was performed as previously described [76] after staining DNA with Sytox green ( Molecular Probes ) . Flow cytometry measurements were taken on a FACSCalibur with CellQuest software ( Becton Dickinson ) and further analysed with FlowJo software ( Treestar ) . Two independent strains ( NKI8529 and NKI8587 ) were selected for genomic DNA isolation to determine the replication status of genomic loci . Genomic DNA was isolated as described previously [85] . Briefly , 2 × 107 cells were spun down and resuspended in 200 μl of 2% Triton X-100 , 1% SDS , 100 mM NaCl , 10 mM Tris [pH 8 . 0] , and 1 mM EDTA . Then , 200 μl phenol-chloroform-isoamyl alcohol ( 25:24:1 ) was added , and cells were lysed by bead beating with 300 μl Zirconia/silica beads . TE 200 μl was added , the cells were spun for 5 minutes at maximum speed , and the aqueous layer was transferred to a new tube . The pellet was washed in 1 mL 100% ethanol and resuspended in 400 μl TE . To precipitate the DNA 10 , approximately 14 M ammonium acetate and 1 mL 100% ethanol were added , and the tube was inverted to mix contents and spun 2 minutes at 13 , 000 rpm . The pellet was air dried and suspended in 50 μl TE . The gDNA was then purified by sending through an Isolate II genomic DNA kit ( Bioline ) . The amount of 20 μl of ethanol-precipitated gDNA was resuspended in 180 μl lysis buffer GL and 1 μl RNaseA ( 10 mg/mL ) and kept at room temperature for 20 minutes before proceeding with step 3 of the manufacturer’s protocol . DNA was eluted in 50 μl elution buffer G . The gDNA was sheared with a Covaris E220 ultrasonicator ( Covaris ) to obtain fragments of 150 to 200 bp and sequenced on the HiSeq2000 platform ( Illumina , San Diego , CA ) . Single-end reads of 65 bp were aligned to the S . cerevisiae reference genome version R64 ( UCSC SacCer3 ) by using BWA [86] . The BAM files with aligned reads were filtered for mapq 37 and converted to bedgraph with bins of 250 bp by using deepTools [87] . By using custom R scripts , the region of unstable transcript XUT_12F-188 was removed because it contains repetitive sequences . The average coverage in each 250 bp bin was plotted across the entire genome by using custom R scripts . The replication timing in the absence of HU was obtained from a density transfer experiment [36] and downloaded from S1 Table in that study . This table contains the percentage of the genome that had become hydrid in density ( percent HL DNA ) in cells collected at different time points [36] . The plots were generated for both samples separately and showed similar patterns . RNA was isolated using the RNeasy Mini Kit ( QIAGEN ) using the protocol for yeast cells , with a few modifications . Briefly , 2 × 107 cells were spun down ( 5 minutes at 3 , 000 rpm ) , and pellets were dissolved in 600 μL cold RLT buffer . Cells were broken by bead beating with 400 μL Zirconia/silica beads , and debris was separated by centrifuging 2 minutes at 13 , 000 rpm . The supernatant ( approximately 350 μL ) was collected and mixed with one volume ( approximately 350 μL ) 70% EtOH and transported to RNeasy columns . Following the buffer RW1 and buffer RPE wash steps , RNA was eluted in 50 uL elution buffer . Eluted RNA was treated with DNase I ( QIAGEN ) to remove genomic DNA . Next , cDNA was prepared using SuperScript II reverse transcriptase ( Invitrogen ) . Reverse transcription PCR ( RT-PCR ) was performed with the primers in S7 Table . Each sample was measured in 2 technical duplicates , and the average value of these 2 was taken as 1 value when combining biological replicates . Quantitative observations that underlie the data summarised in the graphs are shown in S1 Data .	DNA is packaged by proteins into a structure called chromatin . This packaging plays an important role in gene transcription , DNA replication , and DNA repair . The regulation of these processes is determined by the interplay of proteins that physically interact with specific loci . Although many chromatin proteins have been identified , it remains a challenge to comprehensively identify all the chromatin interactions at any specific locus by protein-based methods such as capture mass spectrometry ( MS ) . Here , we take advantage of DNA barcoding technologies and high-throughput sequencing to tackle this problem . We developed Epi-Decoder , a method that identifies chromatin–protein interaction at a single-copy locus in budding yeast . Epi-Decoder was successfully applied at 3 different genomic loci , resulting in comprehensive overviews of local chromatin interactomes . Most factors identified are known to be involved in chromatin-associated processes such as transcription and DNA replication . We also identified unexpected chromatin–protein interactions . Furthermore , capturing the dynamic chromatin interactome by changing physiological conditions provided insight into the local response to stress . In conclusion , Epi-Decoder is an effective tool for measuring the local proteome of a specific genomic locus by DNA sequencing .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "sequencing", "techniques", "taxonomy", "methods", "and", "resources", "dna-binding", "proteins", "genomic", "library", "construction", "data", "management", "fungi", "dna", "replication", "molecular", "systematics", "dna", "construction", "epigenetics", "dna", "molecular...	2018	Decoding the chromatin proteome of a single genomic locus by DNA sequencing
Rodents move their whiskers to locate and identify objects . Cortical areas involved in vibrissal somatosensation and sensorimotor integration include the vibrissal area of the primary motor cortex ( vM1 ) , primary somatosensory cortex ( vS1; barrel cortex ) , and secondary somatosensory cortex ( S2 ) . We mapped local excitatory pathways in each area across all cortical layers using glutamate uncaging and laser scanning photostimulation . We analyzed these maps to derive laminar connectivity matrices describing the average strengths of pathways between individual neurons in different layers and between entire cortical layers . In vM1 , the strongest projection was L2/3→L5 . In vS1 , strong projections were L2/3→L5 and L4→L3 . L6 input and output were weak in both areas . In S2 , L2/3→L5 exceeded the strength of the ascending L4→L3 projection , and local input to L6 was prominent . The most conserved pathways were L2/3→L5 , and the most variable were L4→L2/3 and pathways involving L6 . Local excitatory circuits in different cortical areas are organized around a prominent descending pathway from L2/3→L5 , suggesting that sensory cortices are elaborations on a basic motor cortex-like plan . Sensation in the rodent vibrissal system relies on active whisking for interactions with the environment [1] , [2] . Motor circuits control whisker movement , while sensory afferents collect information about contact with objects . Interactions between motor and sensory systems are necessary for object localization and identification [3]–[5] . Ascending sensory and descending motor pathways interact at multiple levels including the brainstem [6] , thalamus [7] , and cortex [8] . Three areas in the cerebral cortex are activated by whisker stimulation . Primary somatosensory cortex ( vS1 ) responds with short latencies [9] , whereas secondary somatosensory cortex ( S2 ) and vibrissal motor cortex ( vM1 ) respond 10–20 ms later [10] . These areas are also strongly interconnected in a bidirectional manner [8] , [11] . In rodents , some of the cytoarchitectonic features of vM1 , vS1 , and S2 are area-specific , such as the presence of “barrels” in layer ( L ) 4 of vS1 , and others are not , such as the presence of most cortical layers , including L1 , L2/3 , L5A , L5B , and L6 [12] . Here , to explore the synaptic organization of cortical circuits in these three areas , we used glutamate uncaging and laser scanning photostimulation ( LSPS ) to map the local sources of excitatory synaptic input to individual excitatory neurons in vM1 , vS1 , and S2 . We recorded from postsynaptic neurons distributed across L2–6 ( i . e . , all the cortical layers that contain excitatory neurons ) and , for each one , stimulated presynaptic neurons also distributed across L2–6 . The collection of synaptic input maps for each area was analyzed to extract a laminar connectivity matrix representing the local pathways between excitatory neurons in each area [13] , [14] . These connectivity matrices provide a quantitative survey of the interlaminar organization of local excitatory networks in each of these three cortical areas . We identified vibrissal motor cortex ( vM1 ) , primary somatosensory ( barrel ) cortex ( vS1 ) , and secondary somatosensory cortex ( S2 ) based on anatomical coordinates , cytoarchitectonic features , anatomical labeling experiments , and in the case of vM1 , optical microstimulation mapping . vM1 ( Figure 1A ) was located in the posteromedial part of frontal agranular cortex , anteromedial to the barrel cortex [10] , [15]–[17] . When anterograde tracers were injected into vM1 , fluorescently labeled axons were observed in brainstem nuclei involved in whisker motor control ( Figure S1 ) [18] . Furthermore , microstimulation mapping using channelrhodopsin-2 ( ChR2 ) revealed that vM1 had the lowest thresholds for whisker movements ( 14 ) [19] , [20] . vS1 ( Figure 1B ) was identified by the presence of cytoarchitectonic “barrels” in L4 [21] . S2 ( Figure 1C ) was located in dysgranular cortex , lateral to the barrel cortex [8] , [10] , [22] . Axons projected from vS1 to S2 , and from S2 to vM1 and vS1 ( Figure S1 ) . These experiments enabled us to target our mapping experiments to specific cortical locations corresponding to vM1 , vS1 , and S2 . We prepared coronal brain slices containing vM1 , vS1 , or S2 ( Figure 1A–C ) and used LSPS with glutamate uncaging [23]–[25] to map excitatory inputs to excitatory neurons ( Figure 1D–G ) . We excited small clusters of neurons at each site in an array of locations while recording from individual excitatory neurons ( Figure 1D , E ) , obtaining maps of local intracortical sources of excitatory input ( Figure 1F , G ) . To calibrate LSPS , we recorded in cell-attached mode from excitatory neurons , while uncaging glutamate on a grid around the cell ( Figure 2A , B ) . The spatial distribution of action potentials ( APs ) evoked by uncaging ( the “excitation profile” ) provides a measure of the effective spatial resolution of photostimulation ( Figure 2A , B ) . These data were used to estimate neuronal photoexcitability ( Figure 2C ) and the spatial resolution of LSPS ( Figure 2D ) for photostimulating neurons in different cortical layers and areas . Photostimulation-evoked APs always occurred in perisomatic regions ( Figure 2B , Figure S2 ) with short latencies , and almost always as singlets . Stimulation of strong synaptic pathways , such as L4→L3 in vS1 , did not cause APs in the target location ( Figure S2 ) , indicating that synaptic activity did not cause APs in neurons that were not directly photostimulated . Ultraviolet ( UV ) attenuation in scattering tissue causes photoexcitation to decline as a function of depth in the slice; consistent with this , excitation was not observed for neurons deeper than 100 µm ( Figure S3 ) [26] . The total number of neurons excited per stimulus , estimated from the excitation profiles and measured densities of neurons ( Figure 2 and Figure S4 ) , was in the range 50–200 , consistent with previous results [13] , [26] . Only a small fraction of these neurons were synaptically connected to the recorded postsynaptic neuron [27] . An input map represents the aggregate functional synaptic connectivity between small clusters of presynaptic excitatory neurons at the stimulus locations and individual postsynaptic neurons . Pixels in input maps do not represent the strengths of unitary connections; rather , they measure average monosynaptic excitatory responses to a single uncaging event ( see Text S1 , Equations 1–4 ) [26]: ( 1 ) where ρcell is the neuronal density at the point of uncaging ( neurons/µm3 ) , Vexc is the volume of excited neurons ( µm3 ) , and SAP is number of APs fired per presynaptic neuron ( AP/neuron ) . The average strength of a synaptic connection ( qcon ) is calculated from equation ( 1 ) . The collection of qcon for different neuronal populations defined by laminar location is the basis of connectivity matrices . We first present the mapping data for each area in the more familiar form of average input maps . In subsequent sections we summarize connectivity in laminar connectivity matrices , which take into account the parameters in equation ( 1 ) . Unlike vS1 , vM1 lacks a distinct granular L4 . The superficial layers L2/3 and L5A are compressed , and deeper layers L5B and L6 are expanded , consistent with vM1's location at the crest of a cortical convexity [28] . In addition , L1 was thicker than in the other areas ( Table 1 ) . Both superficial and deep L5 neurons had dense basal dendrites and a single apical dendrite extending to L1 , and L6 neurons had apical dendrites that did not extend to L1; in some cases , these were inverted pyramids ( Figure 3A ) . We recorded from 95 excitatory neurons located in all layers ( i . e . , from upper L2 to lower L6 ) and mapped the local sources of excitatory synaptic input with LSPS using a stimulus grid that spanned vM1 ( Figure 3B; Figure S5 ) . We pooled neurons into groups by dividing the cortex into 10 equal distance bins; the top-most bin was empty , because L1 lacks excitatory neurons . We averaged the maps in each bin ( Figure 3C ) . The strongest pathway was a descending projection , L2/3→ upper L5 . Weaker ascending projections , within L5 and L5A→L2/3 , were also found ( Figure 3C ) . On average , neurons in the lower one-third ( 0 . 7–1 . 0 ) of vM1 showed weak inputs . However , individual neurons in this deeper range received strong inputs , but these tended to be spatially dispersed and sparse ( Figure S5 ) . We recorded from 80 excitatory neurons in vS1 , using a different stimulus grid matched to the cortical thickness ( Figure 4; Figure S6 ) . In vS1 , laminar boundaries were distinct , allowing pooling of cytoarchitectonically defined groups for binning ( Table 1; Figure S8 ) . The ascending L4→L3 pathway and the descending L2/3→L5 pathway were both prominent ( Figure 4; Figure S6 ) . Similar to vM1 , L6 neurons had relatively weak inputs ( mainly from L4 ) . L4 neurons also showed little intracortical interlaminar input [29] . In addition , we further distinguished sub-layers within L2/3 and L5B based on patterns of connectivity observed in the input maps . For example , L2 constituted a narrow superficial layer of neurons lacking strong input from L4 , but with input from L5A [30] . Binning with a simple three-layer scheme ( ‘supragranular-granular-infragranular’; Figure 4 ) conveyed the main feedforward local excitatory connections in vS1 . S2 abuts the lateral edge of vS1 , where the barrel pattern terminates ( Figure S1 ) . The cytoarchitectonic layers appeared similar in S2 and vS1 , except that the cortex was thinner and L5A thicker . L4 included neurons with a sparse apical dendrite , and neurons lacking an apical dendrite ( Figure 5A ) . L5 neurons had many basal dendrites and an apical dendrite that ramified in L1; L6 neurons' apical dendrites did not extend above L4 . We recorded input maps for 100 excitatory neurons in S2 ( Figure 5B , C; Figure S7 ) . Similar to vS1 , an ascending pathway to more superficial layers ( L4→L3 ) was present but was not the strongest projection . Instead , the descending projection L2/3→L5 was predominant . L5 also received substantial ascending input from L6 . Connectivity matrices represent local circuits in a compact manner [13] , [14] , [27] , [31] , [32] . Each element ( i , j ) in the matrix ( Wi , j ) corresponds to the strength of a connection ( qcon; Equation 1 ) from the jth presynaptic location ( along the rows ) to the ith postsynaptic location ( along the column ) . Distance is measured in normalized units along the radial directions ( pia , 0; white matter , 1 ) . Because of the curvature of vM1 at the cortical flexure ( Figure 6A , B ) , we converted map data from the coordinates of the slice image ( x , y ) to coordinates corresponding to an unfolded cortex ( h , r ) , where h is the horizontal distance along the laminar contour and r is the distance along the radial axis . Figure 6 provides a graphical illustration of the process of converting the pixels in an input map from x-y coordinates ( Figure 6A ) , using a spatial transform defined on the basis of the radial structure of the cortex ( Figure 6B ) , into r-h coordinates ( Figure 6C , D ) . This approach allowed us furthermore to convert input maps into vectors , by averaging input across the horizontal dimension ( h ) at a given presynaptic radial distance ( r ) into bins ( Figure 6E–G; a similar analysis in the horizontal dimension is given in Figure S9 ) . This is identical to averaging along the rows of input maps , except that it takes into account the curvature of the cortex . One neuron's input vector ( Figure 6G ) thus represents the inputs to one neuron from different laminar locations; i . e . , the horizontal dimension has been collapsed . Each neuron was also assigned a postsynaptic radial distance . This allowed us to group all the input vectors and then sort them by the postsynaptic neuron's depth in the cortex ( Figure 6H ) . Stacking the vectors on top of each other , sorted by depth , provided a raw connectivity matrix , Wraw ( rpost , rpre ) , describing connectivity between neurons at different locations along the radial axis ( Figure 6H , Figure 7A ) . The rows in such a connectivity matrix represent synaptic input to a particular laminar location , and the columns represent synaptic output from that laminar location . Intralaminar connections lie along the main diagonal . We note that intralaminar connectivity was undersampled because of direct excitation of the postsynaptic neurons' dendrites . In addition to deriving matrices based on the collections of input vectors ( Figure 7A , D , G ) , we further analyzed the data in terms of the excitation parameters given in equation ( 1 ) . To compute the average connectivity matrix at the level of individual neurons ( Wneuron ) , we binned the data and applied correction factors to derive the strength of input per presynaptic neuron per AP . We divided the connection strength in the raw connectivity matrix by the mean number of APs per uncaging event at the presynaptic region ( Figure 7B , E , H; Figure S10; Text S1 ) and the number of presynaptic neurons stimulated . The number of stimulated neurons was obtained from measurements of ρcell ( Figure S4 ) and Vexc . To compute the connectivity matrix at the level of cortical layers ( Wlayer ) we multiplied the neuron→neuron connections by the number of presynaptic and postsynaptic neurons per layer ( Figure 7C , F , I; a detailed calculation is illustrated in Figures S11 and S12 ) . Values for all connectivity matrices are provided in Table S1 and Dataset S1 . The connectivity matrix description allows us to directly contrast local circuits in different cortical regions . The elements ( pixels ) in the neuron→neuron connectivity matrices , Wneuron ( Figure 7B , E , H ) , represent the mean strength of postsynaptic response in a single neuron extrapolated to a single presynaptic AP in a single cell of the indicated layer ( qcon ) . Pixel values were 10–100 times lower than typical unitary EPSCs , reflecting both the generally low probability of connections between excitatory neurons in cortical circuits ( typically 0 . 1–0 . 2 ) [27] , [33]–[35] , and the fact that the current amplitude in the maps represents a mean over 50 ms rather than the peak of the EPSC . In contrast , the elements in the layer→layer connectivity matrices , Wlayer ( Figure 7C , F , I ) , represent the average strength of connections extrapolated to the entire projection from one layer to another . The Wlayer matrices differ from the Wneuron matrices in that they enhance thicker and more neuron-dense layers and diminish thinner and less neuron-dense layers . For example , because in vS1 the L5A is thin ( Table 1 ) and both L5A and L5B are low in neuronal density ( Figure S4 ) , the projections to and from L5 , such as L5A→L2/3 and L2/3→L5B , are relatively strong at the level of neuron→neuron connectivity ( Figure 7E ) but relatively weak at the level of layer→layer connectivity ( Figure 7F ) . Interestingly , in rat vS1 the L4→L2/3 projection is functionally weak compared to the structural density of L4 axons and L2/3 dendrites , while the converse holds for the L5A→L2/3 projection [36] . Our results here show how weak neuron→neuron connections may be strong in aggregate at the layer→layer level . Further structure-function analyses will be required to determine whether it is generally the case that larger and more neuron-dense layers have weaker neuron→neuron but stronger layer→layer projections . The connectivity matrix representations of vM1 show strong descending projections from L2/3→upper L5 ( Figure 7A–C ) , similar to the forelimb area of mouse M1 [13] , [14] , [37] . This input straddled the L5A/B border . L5B received an additional hotspot from itself , which appeared strong when considered as an entire layer ( Figure 7C ) . The deepest one-third of vM1 ( consisting mostly of L6 ) had weak inputs and outputs . The vS1 excitatory circuits were more complex ( Figure 7D–F ) . The major ascending pathway from L4→L3 was paralleled by an ascending component from L5A . The high cell density in L4 made the L4→L3 connection prominent in the laminar analysis ( Figure 7F ) . Another prominent projection was from L2 and L3 to L5A and L5B; inputs originating in more superficial regions of L2/3 targeted relatively more superficial regions of L5A/L5B ( note the diagonal shape of the L2/3→L5 hotspot in Figure 7E ) . On a neuron→neuron basis , the L3→L5B connection was stronger than L4→L3 , although the layer→layer analysis showed a reduction in cell density relative to L4 . L2 received input from L3 , and weaker input from L5A . However , L2 was thin and thus contributed little to Wlayer . As in vM1 , deep layers had weak inputs and outputs . In S2 ( Figure 7G–I ) , an ascending L4→L2/3 pathway and descending L3→L5 pathway were present . Neurons on the L5A/L5B border also showed strong intralaminar connections . The L6 output evident in the input maps ( Figure 5; Figure S7 ) also supplied potent input to L5B . Although not as strong at the single cell level , the entire L6 excited L5B as much as L3 ( Figure 7I ) . L6 was enhanced in S2 relative to other regions as both a source of synaptic output and a recipient of synaptic input , due to the relatively high density of neurons ( Figure S4 ) and their relatively low photoexcitability ( Figure 2C–E ) . The functional connectivity in the local excitatory circuits of all three regions is simplified into quantitative laminar wiring diagrams ( Figure 8 ) . LSPS with glutamate uncaging simultaneously excites a group of presynaptic neurons , while the postsynaptic response is measured . To derive average connection strength per neuron ( qcon ) , the number of excited neurons needs to be estimated , based on the excitability ( SAP ) , neuron density ( ρcell ) , and excitation volume ( Vexc ) at the uncaging location ( Equation 1 ) . The accuracy of the estimate of qcon is limited by our measurement of ρcell and neuronal excitation ( SAP , Vexc; Text S1 Equations 3–4 ) : Measurements of neuronal density vary by a factor of two [27] , [38] , [39] . Although excitation profiles give a direct measure of evoked APs in brain slices under the relevant recording conditions ( Figure 2 ) , excitation varies across neurons and somewhat across cortical areas , and decreases with depth in the slice; these effects together introduce uncertainty roughly on the order of a factor of two ( Figure S3 ) . Despite these uncertainties , our estimates of qcon are broadly consistent with those derived from pair recordings ( Figure S13 ) . Because LSPS excites many neurons , this strong stimulus allows weak pathways to be detected . However , the average connection strength , qcon , reflects both the connection probability and unitary connection strength: ( 2 ) It is therefore not possible to separate connection probability and unitary connection strength directly . Furthermore , pcon is inversely related to the horizontal separation between cell pairs [34] . LSPS averages inputs from a range of presynaptic locations with varied horizontal offset . For each cell class , a broad distribution of pcon values contributes to LSPS maps . In addition , by computing the average connection strength , we average out the underlying distribution of unitary connection strength , which is a skewed distribution of numerous weak and a few strong connections [27] , [33] . This inherent averaging also makes LSPS insensitive to certain non-laminar aspects of cell-type specificity in cortical connectivity [14] , [33] , [35] , [40]–[43] . Comparison of our neuron→neuron connectivity matrix with a pair-recording study [27] reveals qualitative similarities ( Figure S13 ) . After both methods are corrected to similar units ( peak amplitude in pA/AP ) , the general shape of the connectivity matrix and values for neuron→neuron connectivity are similar . The major interlaminar pathways are L2/3→L5 and L4/5A→L2/3 . However , local intralaminar connections are underestimated in our data set due to direct responses to uncaging . Furthermore , descending projections from L4→L5A and from L5A→L5B may be underestimated in LSPS relative to pair recording due to exclusion of direct responses along the apical dendrite of the postsynaptic neuron ( see L5A and L5B maps in Figure S6 ) . Under-sampling of connected pairs in low-pcon pathways , such as L4→L6 , may account for differences from LSPS , where many L4 neurons are excited during each L6 recording . Lastly , L2 connectivity differs in part because of differences in the definition of this layer . We compared the matrices for the four areas so far studied , vM1 ( present study ) , the forelimb region of somatic M1 [13] , vS1 ( also the present study ) [27] , and S2 ( present study ) . Overall , the main differences are attributable to the presence of a distinct granular layer in somatosensory cortex . Specifically in vS1 , L4 outflow contributed strongly to the connectivity matrix . L4→L2/3 is also a major pathway in rodent V1 [44] . In S2 , the local excitatory circuit differs from vS1 most prominently in that the L4→L3 pathway is reduced . LSPS analyses of auditory cortex circuits have found L4→L2/3 inputs [45] , [46] , which is adjacent to S2 . However , ascending pathways were not unique to vS1 , as a similar but weaker L3/5A→L2/3 pathway was prominent in forelimb M1 , and present but weaker still in vM1 ( Figure 3C and Figure S5C , leftmost panels ) . The upward compression of layers in vM1 , typical of cortical convexities [28] , may be why L3/5A→L2 was less distinct in vM1 than in forelimb M1 ( e . g . , it was more prone to masking by dendritic responses of L2 neurons ) . However , inspection of individual maps and traces ( Figure S5C ) showed that these ascending pathways were present for some L2 neurons . A second main interlaminar hotspot in vS1 was the descending pathway ( s ) L2/3→L5 , which was the predominant hotspot in the two motor areas . We noted that this pathway was present in all three cortical regions studied here and was similarly prominent in somatic M1 [13] . Indeed , it was the predominant pathway in S2 . Thus , a strong supragranular to infragranular descending connection emerged as a common element of local cortical circuits examined here . Superficial L5B neurons and deep L5A neurons at the laminar border were most strongly activated , suggesting that the cytoarchitectonic boundaries identified do not correspond well with functional gradient within L5 . Perhaps an alternative molecular marker , such as Etv1 ( Figure S8 ) , better denotes this functional division . In three of the four areas , L6 neither received nor sent strong projections ( but vS1 neurons in L6 received a weak projection from L4 ) . L6 output is provided by an ascending connection to L4 in cat visual cortex [31] , [32] , [47] but was absent or reduced in all vibrissal areas we studied . L6→L4 projections studied in mouse somatosensory and auditory cortical areas have “modulator” rather than “driver” properties , including paired pulse facilitation [48] . Although deeper neurons tend to have relatively small dendritic arbors [49] , which may account for a reduction ( but not absence ) of inputs , this difference in arbor size is not of sufficient magnitude to account for the paucity of inputs . Similarly , the paucity of outputs was not due to lack of photoexcitability of these neurons . Channelrhodopsin-assisted circuit mapping experiments [50] have shown that the supragranular layers indeed connect preferentially to upper rather than lower infragranular neurons . Thus , the lack of inputs was not due simply to slice-related artifacts such as severing of pathways . Consistent with weak local inputs , in vivo recordings in cat motor cortex suggest that a large number of L6 neurons are virtually silent , even during motor activity [51] . Thus , the sources and modes of excitation for L6 neurons remain to be determined [49] , [52] . However , L6 was more engaged in local circuits in S2 , supplying a measurable output to L5A and L5B and to other L6 neurons . In addition to input from L5B , L6 neurons in S2 collected inputs from a wide horizontal distance , sometimes >300 µm ( Figure S7B , C at right ) . Thus , S2 may be better suited for studying L6 function . One major difficulty in making a comparison of connectivity between two cortical areas is selecting the laminar position of pre- and postsynaptic neurons for the comparison . Is it better to compare identical relative laminar depths between cortical areas , not accounting for the decreased thickness of superficial layers , and increased thickness of deep layers , in motor areas ? How shall we treat the presence or apparent absence of a distinct layer 4 ? We present a direct quantitative comparison of three major areas identified in our study , based on cytoarchitectonic laminar divisions ( Table 1 , Figure S8 ) ( Figure 9 ) . In vS1 [53] and vM1 ( Tianyi Mao , BMH , GMGS , KS , unpublished observations ) these layers correspond to distinct cell types with different projection patterns . The descending projection from L2/3→L5A/B was prominent in all areas , but the strength of the pathways at a neuron→neuron level varied by a factor of four between the areas . Ascending projections from middle layers to superficial ones ( L4→L2/3 in vS1 and S2; L5A→L2/3 in vM1 for comparison ) were also present in all regions but were the least prominent in agranular vM1 . Lastly , the L6→L5 projection identified in S2 was more than twice as strong at the neuron→neuron level than in vS1 ( and the difference was greater with vM1 ) . Our approach provides a defined framework for measuring similarities and differences between cortical microcircuits in a quantitative manner . We use the term radial to refer to the axis defined by the apical dendrites of pyramidal neurons; this axis is approximately normal to the cortical surface . Normalized radial distance is along the radial axis , bounded by the pia and the white matter , where pia = 0 and the L6/white matter border = 1 . Vertical is synonymous with radial . Horizontal , or lateral , refers to planes normal to the radial axis , approximately parallel to layers , or laminae ( Figure 6 ) . Oblique refers to off-axis interlaminar connections . Mice were decapitated at postnatal day 20–25 under isofluorane anesthesia , and the brain rapidly placed in ice cold choline solution ( in mM: 110 choline chloride , 25 NaHCO3 , 25 D-glucose , 11 . 6 sodium ascorbate , 7 MgCl2 , 3 . 1 sodium pyruvate , 2 . 5 KCl , 1 . 25 NaH2PO4 , 0 . 5 CaCl2 ) . Coronal brain slices ( 300 µm ) were cut ( Microm HM 650V ) , incubated 30 min at 37°C in oxygenated ACSF ( in mM: 127 NaCl , 25 NaHCO3 , 25 D-Glucose , 2 . 5 KCl , 2 CaCl2 , 1 . 25 NaH2PO4 , 1 MgCl2 ) , and maintained in a holding chamber at 22°C for up to 5 h during recording . For vM1 slices , the brain was pitched upward ∼10° to optimize alignment with the radial axis of vM1 , and slices ∼0 . 7–1 . 3 mm anterior to bregma were used; for vS1 and S2 slices , the brain was cut coronally , and slices ∼1–2 mm posterior to bregma were used ( Figure 1A , B ) . To determine the optimal slice angle for each area , we used the appearance of the intact apical trunk at high magnification to select slices for recording and avoided those sections where the apical dendritic trunk was at an angle with respect to the slice plane . Thus , only one or two sections per animal could be used for recording . Separate experiments in our laboratory using the photostimulation methods in vS1 [42] and vM1 ( unpublished data ) measure input to L1 dendrites of L5 pyramidal neurons , confirming the apical trunk is intact using this slice angle . We added biocytin to visualize a subset of dendritic arbors , some of which are reconstructed in Figure 3 and Figure 5 . These neurons appeared radially symmetric , with arbors ranging from 300–500 µm in diameter . Since the neurons were 50–100 µm deep in the slice , a portion of the apical and basal dendrites are truncated by slicing and the deep half of the arbor is intact . Recordings were performed at room temperature ( 22°C ) in ACSF . Neuronal excitability was reduced by increased divalent ions ( 4 mM CaCl2 and 4 mM MgCl2 ) , and NMDA receptor blockade with 5 µM 3- ( ( R ) -2-carboxypiperazin-4-yl ) -propyl-1-phosphonic acid ( CPP; Tocris ) . Patch pipettes were fabricated from borosilicate glass with filament ( 4–6 MΩ ) . Intracellular solution contained ( in mM ) : 128 K-gluconate , 10 HEPES , 10 sodium phosphocreatine , 4 MgCl2 , 4 Na2ATP , 3 sodium L-ascorbate , 1 EGTA , and 0 . 4 Na2GTP ( pH 7 . 25; 290 mOsm ) . To visualize dendritic arbors , 20 µM Alexa 594 or 488 ( Molecular Probes ) was added to the internal solution . In some cases , 2–3 mg/mL biocytin was included . Electrophysiological signals were amplified with an Axopatch 700B amplifier ( Molecular Devices ) , filtered at 4 KHz , and digitized at 10 KHz . Data I/O were controlled by Ephus , a suite of custom Matlab-based ( Mathworks ) software tools available online ( https://openwiki . janelia . org [54] ) . Neurons were selected based on pyramidal appearance , or in the case of L4 recordings in vS1 , either pyramidal or stellate appearance . In vS1 , recordings were generally made in the middle of the barrel field and not a specific whisker barrel . Following patching , a family of current steps was presented to determine firing properties . Neurons with narrow APs and high firing rates were rejected for analysis as presumed interneurons . Methods followed published procedures [13] , [26] . MNI-glutamate ( 0 . 2 mM; Tocris , MO [55] ) was added to a recirculating bath . Photolysis was performed by shuttering ( 1 . 0 ms pulse ) the beam of an ultraviolet ( 355 nm ) laser ( DPSS Lasers , San Jose , CA ) , ∼20 mW in the specimen plane , set by a combination of a gradient neutral density filter wheel and Pockels cell ( electro-optical modulator; Conoptics ) . A 16×16 standard stimulus grid for input maps had row and column spacing of 110 µm for vM1 and 90 µm for vS1 and S2 recordings . Maps were recorded in voltage clamp at −70 mV . Inhibitory input amplitude was minimized by recording near the chloride reversal potential . The 256 grid sites were visited in a sequence that optimized the spatiotemporal separation between sites [25] . The sequence was repeated 2–4 times per neuron . In vS1 and S2 , the map was aligned to the top of the pia and centered on the soma . In vM1 , the medial edge of the map was aligned to the interhemispheric fissure , and the top to the dorsal-most edge of pia . To convert each map's set of traces into an array of pixels that represent response amplitudes , we calculated the average current over a 50 ms post-stimulus window . Direct dendritic responses were excluded on the basis of temporal windowing [56] , rejecting traces with events ( detected as >3 SD above baseline ) with onset latencies of <7 ms . At locations where some maps had direct responses at a given pixel while others did not , the average of the non-direct responses was used; pixels were excluded from the average raw input map for a given neuron if all traces had direct responses . We measured excitation profiles using loose-seal recordings with the amplifier in voltage-follower mode , to gauge the efficacy of photostimulation for neurons in the different layers in the three areas . Excitation profiles were recorded and analyzed following previous methods [13] , [25] , [26] , [30] . To characterize the size of the excitation profile , we calculated the mean weighted distance from the soma of AP generating sites as: Σ ( APs × distance from soma ) /Σ ( APs ) . Procedures build on [13] . A transformation step was added , to account for cortical curvature , which was especially strong in vM1 . As described in Results , we assigned each point in the stimulus grid a normalized radial distance and horizontal offset ( Figure 6 ) . Individual recorded neurons were also assigned a postsynaptic radial distance based on the same criteria . Individual input maps for a given neuron could then be averaged together based on postsynaptic radial distance . Furthermore , when computing the input to a given neuron for the purpose of determining the connectivity matrix , a presynaptic point would be averaged into a bin appropriate to its location . Most aspects of local connectivity were robust to changes in binning . Subsequent corrections to the connectivity matrices were made to account for variations in excitability between layers , and the number of neurons in pre- and post-synaptic layers; these were then presented as neuron→neuron connectivity matrices and layer→layer connectivity matrices ( see Results and Text S1 ) . To make quantitative comparisons between the strength of pathways in different areas , we determined both the average strength of pathways and their variability using a bootstrap-based analysis ( Figure 9 ) . After selecting the pre- and postsynaptic neuron populations by relative laminar depth , the strength of corresponding pixels in the input map ( limited to maps from neurons in the postsynaptic layer , and pixels in the presynaptic layer within 300 µm horizontal distance of the soma ) were averaged for each map . We resampled the individual map averages 10 , 000 times with replacement and resampled other factors contributing to the individual neuron→neuron strength ( number of APs from cortical area's excitation profiles and neuron density ) . Pathways were presented with the average strength and SD from the bootstrap analysis . In vS1 and S2 , we performed morphometric measurements of cortical landmarks in video images of brain slices . Along a radial line , we marked the locations of the soma , pia , white matter , and major laminar boundaries and calculated the absolute and normalized radial distances of these locations . The bottom extent of cortex was marked at the border between L6 ( including the subplate zone ) and white matter [57] . The distances to lower borders of layers ( ±SD ) are given in Table 1 . The division between L2 and L3 in vS1 was drawn between groups of neurons that did not receive appreciable L4 input ( L2 [58] ) and those that did . Since this functional division was not clear in S2 , L2/3 was divided in half . These values are bracketed in the table . vM1 appearance was similar to somatic motor cortex , with a prominent clear zone in the upper middle part of the cortex , corresponding to L5A [13] . Thus , landmarks indicating the border between L1 , a compressed L2/3 , and the bottom of L5A were apparent in video images and used to measure laminar boundaries in vM1 . The division between L5B and L6 was estimated as the radial distance where cell density increased ( Figure S4; Table 1 ) , as a clear border was not apparent based on image contrast . Alternative methods of determining cortical layers in motor and sensory cortex were performed on images of gene expression patterns from the Allen Brain Atlas ( Figure S8 ) .	The neocortex of the mammalian brain is divided into different regions that serve specific functions . These include sensory areas for vision , hearing , and touch , and motor areas for directing aspects of movement . However , the similarities and differences in local circuit organization between these areas are not well understood . The cortex is a layered structure numbered in an outside-in fashion , such that layer 1 is closest to the cortical surface and layer 6 is deepest . Each layer harbors distinct cell types . The precise circuit wiring within and between these layers allows for specific functions performed by particular cortical regions . To directly compare circuits from distinct cortical areas , we combined optical and electrophysiological tools to map connections between layers in different brain regions . We examined three regions of mouse neocortex that are involved in active whisker sensation: vibrissal motor cortex ( vM1 ) , primary somatosensory cortex ( vS1 ) , and secondary somatosensory cortex ( S2 ) . Our results demonstrate that excitatory connections from layer 2/3 to layer 5 are prominent in all three regions . In contrast , strong ascending pathways from middle layers ( layer 4 ) to superficial ones ( layer 3 ) and local inputs to layer 6 were prominent only in the two sensory cortical areas . These results indicate that cortical circuits employ regional specializations when processing motor versus sensory information . Moreover , our data suggest that sensory cortices are elaborations on a basic motor cortical plan involving layer 2/3 to layer 5 pathways .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "neuroscience/motor", "systems", "neuroscience", "neuroscience/sensory", "systems" ]	2011	Laminar Analysis of Excitatory Local Circuits in Vibrissal Motor and Sensory Cortical Areas
Coxiella burnetii is the agent of the emerging zoonosis Q fever . This pathogen invades phagocytic and non-phagocytic cells and uses a Dot/Icm secretion system to co-opt the endocytic pathway for the biogenesis of an acidic parasitophorous vacuole where Coxiella replicates in large numbers . The study of the cell biology of Coxiella infections has been severely hampered by the obligate intracellular nature of this microbe , and Coxiella factors involved in host/pathogen interactions remain to date largely uncharacterized . Here we focus on the large-scale identification of Coxiella virulence determinants using transposon mutagenesis coupled to high-content multi-phenotypic screening . We have isolated over 3000 Coxiella mutants , 1082 of which have been sequenced , annotated and screened . We have identified bacterial factors that regulate key steps of Coxiella infections: 1 ) internalization within host cells , 2 ) vacuole biogenesis/intracellular replication , and 3 ) protection of infected cells from apoptosis . Among these , we have investigated the role of Dot/Icm core proteins , determined the role of candidate Coxiella Dot/Icm substrates previously identified in silico and identified additional factors that play a relevant role in Coxiella pathogenesis . Importantly , we have identified CBU_1260 ( OmpA ) as the first Coxiella invasin . Mutations in ompA strongly decreased Coxiella internalization and replication within host cells; OmpA-coated beads adhered to and were internalized by non-phagocytic cells and the ectopic expression of OmpA in E . coli triggered its internalization within cells . Importantly , Coxiella internalization was efficiently inhibited by pretreating host cells with purified OmpA or by incubating Coxiella with a specific anti-OmpA antibody prior to host cell infection , suggesting the presence of a cognate receptor at the surface of host cells . In summary , we have developed multi-phenotypic assays for the study of host/pathogen interactions . By applying our methods to Coxiella burnetii , we have identified the first Coxiella protein involved in host cell invasion . Coxiella burnetii is an obligate intracellular Gram-negative bacterium responsible of the worldwide neglected zoonosis Q fever [1] , [2] . Acute forms of the disease are characterized by a febrile illness associated with severe headache , pneumonia and hepatitis . In a small percentage ( 2–5% ) of cases , acute Q fever develops into a chronic infection that may lead to endocarditis and chronic fatigue syndrome [1] , [3] . Coxiella resists environmental stress by generating small cell variants ( SCVs ) that facilitate its airborne dissemination; during infections , this pathogen converts into a metabolically active large cell variant ( LCV ) with a unique resistance to the degradative machinery of host cells [4] , [5] . These factors contribute to the extreme infectivity of this microbe , making of Coxiella a serious health concern , especially in rural areas where outbreaks are likely to occur and are accompanied by heavy economic burdens [6] , [7] . Moreover , the development of Coxiella as a potential bioweapon during and since World War II , has ascribed this pathogen among class B biothreats [7] . Coxiella has two antigenic phases: phase I organisms , isolated from natural sources of infection , are extremely virulent . Phase II bacteria originate from spontaneous mutations after several in vitro passages of phase I organisms and present a truncated lipopolysaccharide ( LPS ) [8] . These non-reversible mutations result in a strong attenuation of virulence in vivo [9] , [10] . Phase II Coxiella organisms are internalized more efficiently than phase I organisms by both professional macrophages and non-phagocytic cells [9] , [11] , however , once internalized , both antigenic phases replicate within host cells with similar kinetics . A phase II clone ( Nine Mile phase II clone 4 or NMIIC4 ) , which has been authorized for biosafety level 2 ( BSL-2 ) manipulation , represents therefore an optimal model to study Coxiella infections [2] , [5] . In natural infections , Coxiella has a tropism for alveolar macrophages [1] , [2] , however , infection of epithelial and endothelial cells has also been reported [12] , [13] . Indeed , in vitro , Coxiella invades and replicates in a wide variety of phagocytic and non-phagocytic cells [5] . Coxiella internalization within host cells is a passive , endocytic process , which involves the remodeling of the host cell actin cytoskeleton [14] , [15] and αVβ3 integrins have been reported as Coxiella receptors in THP-1 cells [11] . However the Coxiella factors that mediate interactions with host cell surfaces , as well as the bacterial host receptor on epithelial cells remain unknown . During the first 48 hours following internalization , bacteria reside into tight-fitting vacuoles , positive for early endosomal and autophagosomal markers [16] . As Coxiella-containing vacuoles mature along the endocytic pathway , the drop in vacuolar pH triggers the translocation of bacterial proteins by a Dot/Icm type 4b secretion system ( T4SS ) [17] . Effector translocation is essential for the biogenesis of a large parasitophorous vacuole ( PV ) that occupies the majority of the host cytoplasm [18] , [19] . Such large membranous structures are highly dynamic and fusogenic and the host endocytic SNARE Vamp7 is required for optimal PV development [20] . Importantly , mature Coxiella PVs are positive for lysosomal markers and contain active degradative enzymes [5] , [16] . Coxiella infections are not lytic and bacteria-filled PVs persist within infected cells , which are protected from apoptosis by a Dot/Icm-dependent mechanism [19] , [21]–[25] . Importantly , due to the obligate intracellular nature of this pathogen , the microbial factors involved in host/pathogen interactions remain to date largely unknown . The homology between the T4SS of C . burnetii and L . pneumophila allowed the in silico identification of 354 candidate Coxiella effectors based on the presence of a conserved Dot/Icm regulatory motif ( PmrA ) [26]–[28] , C-terminal translocation signals ( E-block ) [26]–[28] , and eukaryotic-like domains [29]–[31] . Dot/Icm–dependent secretion has been validated for 108 of these using either Coxiella or Legionella as a surrogate host [18] , [26]–[31] . Recent advances in Coxiella axenic culture techniques [32] rendered this pathogen genetically tractable [33] , allowing for the first time to couple bioinformatics analysis to morpho-functional assays and investigate the role of candidate Coxiella virulence determinants in intracellular replication [18] , [19] , [27] . To date , 20 Coxiella genes encoding Dot/Icm substrates have been mutated to investigate their role in Coxiella replication within the host [27] . Here we have set up new , integrative approaches that combine transposon mutagenesis with genomics , bioinformatics and fluorescence-based functional assays aiming at the large-scale identification of intracellular bacteria virulence factors . Our approach is designed for the simultaneous investigation of multiple key steps of Coxiella infections and is based on the identification and characterization of transposon-induced phenotypes . We have generated and isolated 3000 Coxiella transposon mutants , 1082 of which have been sequenced and screened in the present study . Our analysis revealed important insights into the functionality of the Coxiella Dot/Icm apparatus and revealed a variety of bacterial factors involved in 1 ) internalization within host cells , 2 ) PV biogenesis and intracellular replication , and 3 ) protection of the infected cell from apoptosis . By focusing our analysis on the early events of Coxiella infections we identified the first Coxiella invasin that plays an essential role in bacterial internalization by non-phagocytic cells . To identify the Coxiella factors involved in host-pathogen interactions , we have undertaken the generation of a library of GFP-tagged bacterial mutants by transposon mutagenesis . We have modified the Himar1-based transposon system initially developed by Heinzen and colleagues [33] , [34] , by inserting the enhanced green fluorescent protein ( egfp ) gene under the regulation of the Coxiella promoter P311 , upstream of the chloramphenicol resistance cassette , thus generating pITR-CAT-ColE1-P311-GFP . To obtain stable mutants , Nine Mile Phase II clone 4 ( NMIIC4 ) Coxiella ( hereafter referred to as wt Coxiella ) were electroporated using a two-plasmid system , where the transposase is encoded by a suicide plasmid that is lost during bacterial replication [34] . The eGFP-tagged Coxiella mutants thus generated were isolated on ACCM-2 agar plates in the presence of chloramphenicol and further amplified for 7 days in liquid ACCM-2 supplemented with chloramphenicol . The final concentration of each bacterial culture was calculated using the Quant-iT PicoGreen dsDNA assay . Transposon insertion sites were identified by single-primer colony PCR followed by DNA sequencing . Using the primer SP3 we amplified DNA fragments including a 278 bp region upstream of the 3′ Inverted Terminal Repeat ( ITR ) of the inserted transposon ( Fig . 1A ) . The amplified fragments were then sequenced using the transposon-specific primer P3 , which recognizes a sequence in the 3′ region of the Chloramphenicol Acetyltransferase ( CAT ) gene ( Fig . 1A ) . The obtained sequences were then aligned on the Coxiella burnetii RSA493 annotated genome using automated sequence analysis software . The genome of Coxiella burnetii RSA493 contains 1849 coding sequences ( CDS ) , 1814 in the bacterial chromosome and 35 in the cryptic plasmid QpH1 [35] . To date we have isolated 3000 transposon mutants , 1082 of which have been sequenced , annotated and analyzed for this study ( Fig . 1B ) . Transposon insertions were homogeneously distributed throughout the Coxiella chromosome and plasmid , with seven “hot spots” of preferential transposon insertion ( identified and annotated from 1 to 7 , Fig . 1B ) and a large , 52 CDS region , upstream of hot spot n . 2 , which remained non-mutated . Of note , region n . 7 corresponds to the locus that hosts T4SS core genes ( dot/icm genes ) whereas the non-mutated region between CBU_0215 and CBU_0272 is enriched in genes encoding ribosomal proteins . Overall , 926 transposon insertions were found within Coxiella annotated CDS and 156 in intergenic regions of the Coxiella genome ( excluding insertions within the first 100 bp upstream of a CDS; Fig . 1C ) . Frequency distribution analysis revealed that mutations occurred in 483 CDS on the Coxiella chromosome and 8 CDS on the QpH1 plasmid ( corresponding to 26 . 6% and 22 . 8% of the total CDS present on chromosome and plasmid respectively; Fig . 1C ) . The mutated CDS were then clustered according to their predicted function based on the data available on the Pathosystems Resource Integration Center ( PATRIC , www . patricbrc . org; Fig . 1D ) . Isolated transposon mutants were used to infect Vero cells at comparable multiplicities of infection ( MOI ) . Non-infected Vero cells were used as negative control and cells infected with GFP-NMIIC4 Coxiella ( GFP-Coxiella ) [33] were used as positive controls . Variations of GFP fluorescence associated with intracellular bacterial growth over 7 days of infection were monitored using a multimode micro-plate reader to obtain intracellular growth curves for all the Coxiella mutants screened ( Fig . 2A ) . Seven days after infection , plates were imaged using an automated fluorescence microscope and images were analyzed using the automated image analysis software CellProfiler ( www . cellprofiler . org ) . 9 morphological features were extrapolated from an average of 14000 cells for each condition ( Fig . 2A ) . Nuclei features were used to assess the overall conditions of host cells and the potential cytotoxicity of Coxiella mutations . Coxiella features were used to score the intracellular replication of bacteria . Finally , the number of Coxiella colonies was divided by the number of host cell nuclei to estimate the efficiency of bacterial internalization within cells . We validated our approach by comparing the growth curves and morphology of GFP-Coxiella to those of 6 mutants ( Tn2384 , Tn1948 , Tn292 , Tn2184 , Tn514 , Tn270 ) carrying independent transposon insertions in the gene CBU_1648 , which encodes DotA , an essential component of the Coxiella Dot/Icm secretion system [36] ( Fig . 2B ) . Axenic growth of the 6 dotA mutants was comparable to that of wt Coxiella ( Fig . S1A ) . GFP signal analysis indicated that intracellular growth of GFP-Coxiella followed a typical growth curve , with bacterial replication clearly detectable from day 2 post-infection and increasing until day 7 post-infection ( Fig . 2C ) . Morphological analysis of GFP-Coxiella-infected Vero cells reported bacterial colonies with an average area of 80 . 71±4 . 65 microns2 and an average number of Coxiella colonies/cell of 1 . 13±0 . 12 ( n = 21656; Fig . 2D ) . As expected , the GFP signal originated by all the DotA transposon mutants failed to increase significantly during the 7 days of infection ( Fig . 2C ) . Accordingly , morphological analysis indicated a strong reduction in the average area of intracellular Coxiella colonies , which ranged from 5 . 86±1 . 85 to 32 . 03±4 . 83 microns2 ( n = 19066 and 14879 respectively , Fig . 2D ) . Interestingly , the majority of dotA mutations corresponded to an increased number of colonies/cell , which , in the case of mutant Tn514 , reached 3 . 48±0 . 61 ( Fig . 2D ) . This phenotype suggests that , in the absence of a functional secretion system , the Coxiella-containing vacuoles fail to coalesce to form the typical , large PV . Moreover , two independent transposon insertions in the Dot/Icm gene icmV ( Tn2445 and Tn2214 ) , which is located in the same operon as dotA , produced a comparable , strong replication phenotype ( data not shown ) . The data derived from the multi-phenotypic analysis of Coxiella infections of host cells were mined to identify mutations that perturbed: 1 ) host cell invasion , 2 ) intracellular replication and 3 ) host cell survival . To screen for invasion and replication phenotypes , we plotted the average area of Coxiella colonies against the average number of colonies/cell ( Fig . 3A ) . Statistical analysis was used to define regions in the resulting scatter plot corresponding to mild ( −4<Z-score≤−2 ) and severe ( Z-score≤−4 ) phenotypes . The 1082 analyzed Coxiella mutants were found in 3 well-defined clusters: one included mutants whose phenotype did not vary significantly from that of GFP-Coxiella ( Fig . 3A green dots ) . A second cluster was clearly shifted towards a reduction in the size of Coxiella colonies and an increase in the average number of Coxiella colonies/cell , representative of mutations that affect Coxiella intracellular replication without affecting host cell invasion ( Fig . 3A light and dark red dots ) . A third cluster was shifted towards a reduced number of colonies/cell , indicating mutations that affect host cell invasion ( Fig . 3A light and dark blue dots ) . In parallel , the average area of Coxiella colonies was plotted against the total number of host cells surviving the 7 days of infection , to identify Coxiella genes that are potentially involved in the protection of the host cell from apoptosis ( Fig . 3B ) . As above , statistical analysis was used to define regions corresponding to mild ( −4<Z-score≤−2 ) and severe ( Z-score≤−4 ) phenotypes . The vast majority of the mutants analyzed did not affect cell survival , regardless of bacterial replication within host cells ( Fig . 3B green dots ) . 37 mutations mildly affected host cell survival ( Fig . 3B light red dots ) , and 7 mutations were particularly detrimental to host cell survival ( Fig . 3B dark red dots ) . Next , the phenotypic data from mutations within CDS were integrated with the annotated transposon insertions in the Coxiella genome . We thus clustered the screened mutations according to the mutated CDS and assigned to each mutant a color-coded map based on the intensity of their replication ( R ) , internalization ( I ) or cytotoxic ( C ) phenotype ( Table S1 ) . Intergenic transposon insertions were retained in a separate table ( data not shown ) . Importantly , mutant Tn1832 carries an intergenic transposon insertion and phenocopies wt Coxiella and GFP-Coxiella ( Fig . S2 ) . This mutant has been used in our validation experiments as additional control . Intracellular replication of Coxiella relies on the functionality of a Dot/Icm T4SS , which is highly homologous to that of L . pneumophila [2] , [37]–[39] . The Coxiella genome contains 23 homologues to the 25 known dot/icm genes of the Legionella T4SS . Accordingly , Legionella has been used as a model organism to test the secretion of candidate Coxiella effector proteins . Recently , it has been reported that the Coxiella dot/icm genes icmD , dotA , dotB and icmL . 1 are essential for Coxiella replication within host cells , proving for the first time the functionality of the Coxiella T4SS [18] , [19] , [36] . The enrichment of transposon insertions in dot/icm genes ( Fig . 1B , region n . 7 ) , prompted us to analyze the phenotype of 38 Coxiella mutants carrying single transposon insertions in 16 Dot/Icm genes ( Fig . 4A; Fig . S1 ) . First , we followed the axenic growth of the 38 Dot/Icm transposon mutants and found it to be indistinguishable from that of wt Coxiella and of the control transposon mutant Tn1832 ( Fig . S1A ) . Multi-phenotypic analysis confirmed the previously reported observations that icmD , dotA and icmL . 1 are essential for Coxiella replication within host cells [18] , [19] , [36] ( Fig . 4B , C; Fig . S1B ) . Moreover , we could observe that 12 dot/icm genes ( dotA , dotB , icmV , E , D , G , J , N , C , P , K , X , L . 1 ) are essential for bacterial replication within the host , whereas mutations in icmB and icmS showed an intermediate phenotype , which corresponded to partial intracellular replication as assessed by morphological analysis ( Fig . 4B , C; Fig . S1B ) . Of note , transposon insertions in dot/icm genes present in operons produced consistent phenotypes . Each mutation was then trans complemented by challenging host cells in combination with wt Coxiella as previously described [19] . The intergenic mutant Tn1832 was used as positive control . As illustrated in Figure 4C , mutations resulting in severe replication phenotypes were efficiently complemented by the presence of wt Coxiella in the PV occupied by Dot/Icm mutants ( black bars ) . We have also isolated and screened 63 transposon mutants in 31 CDS encoding predicted Coxiella Dot/Icm substrates ( Fig . 3C ) . The products of 22 of these CDS have been previously reported to be positive for Dot/Icm secretion [18] , [26] , [27] , [29] , [30] , [31] ( Fig . 3C ) . Our analysis indicated that mutations in 17 of these genes produced replication phenotypes ( Fig . 3C , CDS in red ) , further suggesting that these genes encode Coxiella effectors . Interestingly , a transposon insertion in CBU_1639 resulted in a strong cytotoxic phenotype ( Fig . 3C CDS in blue ) , suggesting that this gene plays a role in the Coxiella-mediated inhibition of apoptosis . The Coxiella genome encodes 4 two-component systems: PhoB-PhoR ( CBU_0367-CBU_0366 ) , GacA-GacS ( CBU_0712-CBU_0760 ) , QseB-QseC ( CBU_1227-CBU_1228 ) and an RstB-like system ( CBU_2005-CBU_2006 ) [2] , [35] . In particular , the QseB-QseC system has been reported to be homologous to the L . pneumophila PmrA-PmrB system , a fundamental regulator of Dot/Icm secretion and its role has been indirectly confirmed [26] , [40] . Consistent with these observations and with our analysis on Coxiella Dot/Icm genes , 6 independent transposon insertions in CBU_1227 ( qseB ) and one insertion in CBU_1228 ( qseC ) , significantly impaired Coxiella replication within host cells ( Table S1 ) . A single transposon insertion in CBU_2006 , part of the RstB-like system also produced a significant replication and entry phenotype ( Table S1 ) , however , the role of this two-component system in Coxiella remains to be defined . The annotation of the Coxiella burnetii NMI RSA493 genome revealed the presence of 207 pseudogenes , which are not conserved among different Coxiella isolates [35] . Recent whole transcriptome analysis ( RNA-Seq ) has validated the previous annotation confirming that these sequences do not encode complete open reading frames ( Prof . Howard Shuman personal communication ) . Interestingly , we have isolated 85 transposon mutants in 56 CDS annotated as pseudogenes and mutations in 32 of these resulted in a strong replication phenotype ( Fig . S3 ) . Finally , we could observe that 71 out of the 151 transposon insertions in intergenic regions of the Coxiella genome exhibited a significant replication phenotype ( not shown ) . This may reveal small RNA-mediated regulation of Coxiella virulence . Indeed , 11 transposon insertions fall within intergenic regions where putative Coxiella sRNAs have been identified by RNA-Seq ( Prof . Howard Shuman personal communication ) . Further investigations are currently aiming at an in-depth characterization of these new candidate factors that may regulate intracellular replication of Coxiella . As mentioned above , our multi-phenotypic screen identified 7 transposon mutants that exhibited a strong cytotoxic phenotype when incubated with Vero cells ( Fig . 3B , Fig . S4 ) . To further analyze the phenotype of these mutants , we investigated their intrinsic capacity of triggering apoptosis and their potential of protecting infected cells from staurosporine-induced apoptosis . HeLa cells were preferred to Vero cells for this assay as they displayed a more consistent response to staurosporine and all Coxiella strains tested displayed similar replication phenotype in these cells ( data not shown ) . Cells were either left unchallenged or incubated with wt Coxiella , the control transposon mutant Tn1832 , the DotA mutant Tn270 and the 7 cytotoxic mutants ( Tn881 , Tn616 , Tn946 , Tn926 , Tn1226 , Tn1232 , Tn1233 ) . Three days post-inoculation , cells were either fixed in paraformaldehyde or incubated with 1 µM staurosporine for 4 h prior to fixation . The percentage of apoptotic cells was then evaluated for each condition by the TUNEL assay . Very few TUNEL-positive cells were observed among untreated cells , these were increased to 50% of the total cell population upon staurosporine treatment . As expected , incubation of cells with wt Coxiella did not increase the number of TUNEL-positive cells as compared to untreated cells and wt Coxiella-colonized cells were efficiently protected from staurosporine-induced apoptosis . Cells challenged with the control transposon mutant Tn1832 presented the same phenotype as cells incubated with wt Coxiella and conversely , the DotA mutant Tn207 failed to protect infected cells from induced apoptosis ( Fig . S4 ) . Mutant Tn881 , which carries a transposon insertion in CBU_0485 , exhibited a partial protection of infected cells from induced apoptosis whereas the remaining mutants failed to effectively protect cells from the effects of staurosporine ( Fig . S4 ) . Interestingly , incubation of HeLa cells with mutants carrying transposon insertions in CBU_1639 , CBU_1366 and CBU_0307a significantly increased the number of TUNEL-positive cells also in the absence of staurosporine , indicating that these mutants may possess intrinsic cytotoxic properties ( Fig . S4 ) . As for all obligate intracellular pathogens , Coxiella invasion of host cells is a priming step of the infection . However , since the bacterial factors that mediate Coxiella invasion of host cells remain unknown , we sought to identify bacterial factors whose mutations affect Coxiella internalization . High-content screening identified 48 mutations in 37 Coxiella CDS that resulted in a significant reduction in the number of infected cells after 7 days of infection ( Fig . S5 ) . Of these , 18 CDS involved in bacterial metabolism and transcription were excluded . Among the remaining 19 candidate CDS ( Fig . S5 , CDS boxed in red ) , we have identified 5 independent transposon insertions ( Tn175 , Tn208 , Tn27 , Tn907 and Tn749 ) in CBU_1260 , all sharing a consistent , strong internalization phenotype ( Fig . S5 ) . CBU_1260 is a 747 bp CDS on the positive strand of the Coxiella chromosome encoding a hypothetical protein of a predicted size of 23 kDa . Importantly , the gene is not part of an operon , indicating that the phenotype observed for the 5 mutants analyzed in this study was indeed due to the inactivation of CBU_1260 alone ( Fig . 5A ) . Axenic growth of the 5 transposon mutants was comparable to that of wt Coxiella and of the control transposon mutant Tn1832 ( Fig . S6A ) . As expected , intracellular growth curve analysis of the 5 transposon mutants in CBU_1260 indicated that GFP fluorescence failed to increase during the 7 days of infection ( Fig . 5B ) . Indeed , all 5 mutations in CBU_1260 reduced the number of cells presenting Coxiella colonies at 7 days of infection by 60–70% compared to cells challenged with GFP-Coxiella ( Fig . 5C ) . We next used the online analysis software i-TASSER ( http://zhanglab . ccmb . med . umich . edu/I-TASSER/ ) and Phyre2 ( http://www . sbg . bio . ic . ac . uk/phyre2/ ) to predict the structure of the hypothetical protein encoded by CBU_1260 . Bioinformatics analysis predicted 8 transmembrane beta sheets forming a beta barrel domain , an N-terminal alpha helix and 4 unstructured loops ( L1 to L4 ) , exposed at the cell surface ( Fig . 5A , D ) . This prediction was confirmed by analyzing the sequence of the protein using TMpred ( http://www . ch . embnet . org/software/TMPRED_form ) and the BetAware software [41] . Sequence analysis indicated that transposon insertions occurred in the distal part of the CDS , within the 4th and 6th beta sheets , with mutants Tn27 and Tn907 presenting insertions at the same site ( Fig . 5A ) . The predicted transmembrane domain of CBU_1260 is typical of Outer Membrane Protein A ( OmpA ) family of proteins , which are found in several bacteria and mediate adhesion and/or internalization within host cells [42]–[50] . We therefore named the product of CBU_1260 OmpA . Coxiella encodes 3 hypothetical proteins that contain predicted OmpA-like domains: CBU_0307 , CBU_0936 and CBU_1260 ( ompA ) . Sequence alignment of these three hypothetical proteins showed high degree of homology at the level of the transmembrane domains and the N-terminal alpha helix ( Fig . S7 ) . However , little homology was observed at the level of the 4 unstructured loops of OmpA ( Fig . S7 ) . Accordingly , a transposon insertion in CBU_0307 produced a cytotoxic phenotype whereas 4 transposon insertions in CBU_0936 produced a replication phenotype ( Table S1 ) . An OmpA-specific antibody was then raised against the predicted extracellular domain of the protein . To this aim , a 15 amino acid peptide in predicted loop 1 ( KKSGTSKVNFTGVTL ) was used for its immunogenic potential as compared to peptides in the other loops . When tested by Western blot on lysates from wt Coxiella and the control mutant Tn1832 , the anti-OmpA antibody revealed a major band at the expected size of 23 kDa ( Fig . 5E , arrow ) and a faint , background band of lower molecular weight ( Fig . 5E ) . When incubated on lysates from the 5 transposon mutants in CBU_1260 , the anti-OmpA antibody only recognized the background band of lower molecular size . We next performed membrane fractionation assays on wt Coxiella and the OmpA mutant Tn208 to validate the outer membrane localization of OmpA . The protein was highly enriched in the outer membrane fraction of wt Coxiella and , as expected , absent in lysates from the OmpA mutant Tn208 ( Fig . 5F ) . Taken together , our data indicate that CBU_1260 encodes an outer membrane protein with a predicted OmpA-like structure that plays a relevant role in host cell invasion . To further investigate the role of OmpA in Coxiella invasion of host cells , we validated the transposon insertion in the OmpA mutant Tn208 by PCR and used a GFP-probe to confirm by Southern blot that Tn208 contained a single transposon insertion ( Fig . S6B , C ) . Next , non-phagocytic epithelial cells ( A431 ) , THP-1 ( PMA-differentiated ) , J774 and RAW 264 . 7 macrophages were challenged either with wt Coxiella , the OmpA mutant Tn208 or the control transposon mutant Tn1832 . Differential labeling of extracellular and intracellular bacteria was used to assess the efficiency of Coxiella internalization within host cells at 15 , 30 , 45 and 60 minutes post-infection ( Fig . 6A , E , top panels ) . Longer time points ( 5 and 6 days post-infection ) were analyzed to investigate the intracellular development of internalized OmpA mutants ( Fig . 6A , E , bottom panels ) . Automated image analysis was then used to analyze approximately 8000 bacteria per condition and quantify the efficiency of bacterial internalization as well as the area occupied by intracellular Coxiella colonies . In A431 non-phagocytic cells , the internalization of Tn208 was strongly reduced as compared to that of wt Coxiella or Tn1832 , which shared similar kinetics ( Fig . 6B ) . Interestingly , when the same internalization experiment was performed using macrophages , the three bacterial strains tested ( wt Coxiella , Tn208 and Tn1832 ) were internalized with comparable efficiency ( Fig . 6F; Fig . S8A , D ) , with a concomitant local rearrangement of the actin cytoskeleton ( Fig . 6E , top panels ) . Of note , despite the strong inhibition of bacterial internalization in A431 cells , OmpA mutants retained the capacity to adhere to host cells ( Fig . 6A ) . This suggests that if OmpA plays a role in bacterial adhesion , this may be masked by the presence of alternative factors involved in Coxiella adhesion to host cells . At longer time points of infection we observed a decrease in the number of Tn208 colonies per cells in A431 cells as compared to wt Coxiella and Tn1832 colonies , which confirmed our previous observations in Vero cells ( Fig . 6C ) . On the contrary , the number of Coxiella-colonized cells was not affected when macrophages were challenged either with wt Coxiella , the control mutant Tn1832 , or the OmpA mutant Tn208 ( Fig . 6G , Fig . S8B , E ) . Remarkably however , the average area of OmpA mutant Coxiella colonies was significantly reduced as compared to wt Coxiella and Tn1832 , regardless of the cell line used for the experiment ( Fig . 6D , H; Fig . S8C , F ) . Interestingly , this is in agreement with previously reported roles of OmpA proteins in intracellular survival of bacterial pathogens [43] , [45] , [46] , [51]–[53] . To further dissect the role of OmpA in Coxiella interaction with host cell surfaces , we 1 ) purified the recombinant protein to coat inert latex beads and 2 ) ectopically expressed OmpA in E . coli , to assess the capacity of OmpA to confer adhesiveness and invasiveness to inert particles and extracellular bacteria , respectively . Histidine-tagged , recombinant OmpA was produced by E . coli BL21-DE3 star pLysS transformed with the pET28a vector containing ompA32-248 . The first 31 amino acids of OmpA corresponding to the intracellular N-terminal alpha helix were excluded to increase the solubility of the protein . Red fluorescent latex beads were then coated with 100 µg/ml His-OmpA32-248 or GST as control and used to challenge A431 cells for 1 hour at 37°C . Unbound beads were removed by rinsing cells in PBS and cells were fixed in paraformaldehyde . Cells were then probed with an anti-histidine antibody without permeabilization to differentially label adherent and internalized beads and with fluorescent phalloidin to define the cell perimeter and volume . Alternatively , after fixation samples were further processed for scanning electron microscopy . Confocal microscopy analysis of cross sections of cells incubated with His-OmpA32-248-coated beads revealed a fraction of beads adhering to the cell surface , hence positive for the anti-histidine staining , and another fraction within the cell volume ( as defined by the actin labeling ) and negative to the anti-histidine staining ( Fig . 7A ) . Three-dimensional reconstruction of confocal sections coupled to surface rendering confirmed the presence of adhering and internalized beads ( Fig . 7A ) . GST-coated beads failed to adhere to and invade A431 cells significantly ( Fig . 7B ) . Scanning electron microscopy analysis corroborated these observations: several His-OmpA32-248-coated beads were adhering to the surface of A431 cells ( Fig . 7C , green inset ) whereas others were clearly covered by the cell plasma membrane ( Fig . 7C , red inset ) . Very few GST-coated beads were observed at the surface of cells and none seemed to be internalized ( data not shown ) , confirming our observations by fluorescence microscopy . We next assessed the capacity of OmpA to trigger the internalization of non-invasive bacteria by non-phagocytic cells using the gentamicin protection assay . To this aim E . coli BL21-DE3 star pLysS were transformed with pET27b-OmpA , which allowed the IPTG-regulated expression and periplasmic targeting of full-length OmpA . The expression and outer membrane localization of OmpA were verified by Western blot using the OmpA-specific antibody on transformed E . coli cultures induced overnight with IPTG and processed to separate the bacterial outer membranes from the inner membranes and cytoplasmic components ( Fig . 7D ) . Non-induced , transformed bacteria were used as control . We could observe that IPTG-induced bacteria efficiently produced OmpA , which was enriched in the outer membrane fraction of E . coli lysates ( Fig . 7D ) . Importantly , induction of OmpA expression conferred E . coli a 20-fold increase in invasiveness as compared to the non-induced bacteria ( Fig . 7E ) . Next , we used the E . coli ectopic expression approach to dissect the role of the 4 predicted extracellular loops of OmpA . By replacing each loop with a myc tag , we generated 4 OmpA mutants ( OmpAΔL1 , OmpAΔL2 , OmpAΔL3 , OmpAΔL4 ) that were used to test their capacity to confer invasiveness to E . coli in a gentamicin protection assay . Similar to wt OmpA , all mutated proteins were detected in the outer membrane fraction of induced E . coli ( not shown ) . Interestingly , only the exchange of loop 1 with a myc tag significantly reduced E . coli internalization by non-phagocytic cells ( Fig . 7E ) . Collectively , these data suggest that OmpA is necessary and sufficient to mediate Coxiella internalization within non-phagocytic cells and that loop 1 is primarily involved in interacting with a potential cognate receptor at the surface of host cells . To determine whether OmpA function requires the interaction with host cell surface factors , we sought to block candidate ligand/receptor interactions , either by saturating potential OmpA receptors at the surface of host cells or by masking OmpA at the surface of bacteria , prior to infection . In the first case , A431 cells were incubated with 100 µg/ml His-OmpA32-248for 1 hour at 4°C prior to challenging with wt Coxiella and the efficiency of bacterial internalization was determined by differential bacterial labeling . A431 cells incubated in the same conditions with GST or with buffer alone were used as controls . Indeed , pretreating A431 cells with His-OmpA32-248 effectively inhibited wt Coxiella internalization as compared to buffer- or GST-treated cells ( Fig . 8A , B ) . Alternatively , wt Coxiella were incubated with increasing concentrations ( 0 . 1 , 1 and 5 µg/ml ) of either anti-OmpA antibody or naïve rabbit serum prior to infection and bacterial differential labeling was used to determine Coxiella invasiveness in A431 cells . Confirming our previous observations , the pre-treatment of bacteria with the anti-OmpA antibody , but not naïve rabbit serum , inhibited Coxiella internalization in a concentration dependent manner ( Fig . 8C , D ) . These observations suggest the presence of a receptor for OmpA at the surface of host cells , which remains to be identified , and that OmpA/receptor interactions are essential to mediate Coxiella internalization within host cells . Larvae of the wax moth Galleria mellonella are an emerging , efficient model for the study of host/pathogen interactions in vivo . Like other insects , Galleria larvae present essential aspects of the innate immune response to microbial infections , which are conserved in mammals . In particular , insects possess humoral and cellular defense responses , the first including antimicrobial peptides ( galiomycin , gallerimycin and lysozyme in the case of Galleria ) and the latter consisting of specialized phagocytic cells , known as hemocytes or granulocytes [54]–[57] . Importantly , in the case of several bacterial pathogens , typical phenotypes observed in mammalian infection models were efficiently reproduced using Galleria [58]–[61] . It has been recently demonstrated that the Coxiella closely-related pathogen Legionella pneumophila invades Galleria hemocytes and replicates within large membranous compartments that present the same morphology and characteristics of Legionella-containing vacuoles generated during infection of macrophages and amoeba [62] . Infections by L . pneumophila result in severe damage to insect organs , which is accompanied by an immune response , including larvae melanization and nodule formation [62] . Moreover , the role of bacterial virulence factors previously characterized in higher mammalian models is conserved during infections of Galleria mellonella [61] , [62] . Importantly , Galleria mellonella larvae are also susceptible to phase II Coxiella infections ( Norville et al . Unpublished data ) . We thus investigated the phenotype associated with the OmpA mutation carried by Tn208 in the context of Galleria infections . Larvae were exposed to Coxiella by injecting 106 bacteria ( either wt Coxiella , Tn1832 or Tn208 ) in the upper right proleg and larvae were incubated at 37°C up to 300 h post-infection to determine survival rates . Alternatively , larvae were incubated up to 24 and 96 h prior to fixation in paraformaldehyde and processing for immunofluorescence . In all cases , larvae injected with PBS were used as a negative control . Larvae injected with PBS alone did not show any survival defect throughout the incubation time , whereas larvae infected with wt Coxiella or the control mutant Tn1832 died significantly faster compared to those infected with the OmpA mutant Tn208 ( Fig . 9A ) . Immunofluorescence analysis revealed that at 96 h post-inoculation , wt Coxiella as well as the control mutant Tn1832 organisms triggered the formation of large , highly infected nodules of hemocytes ( Fig . 9B ) . These were often juxtaposed to larval organs that also appeared severely infected and damaged ( Fig . 9C , top panels ) . In contrast , when Galleria were challenged with the OmpA mutant Tn208 fewer nodules of smaller size were observed throughout the larvae and only a small fraction of these presented signs of infection ( Fig . 9B ) . When these nodules were observed at higher magnification , we could detect small Tn208 colonies ( Fig . 9C bottom panels ) that were reminiscent of what we had previously observed in cultured macrophages . Our observations indicate that the OmpA-associated phenotypes observed in cultured cells can be reproduced during in vivo infections . Infection by bacterial pathogens depends on the subversion of host functions , which is tightly orchestrated by an array of bacterial proteins referred to as virulence factors . In the last decade , cellular microbiology has stressed the importance of studying pathogens in relation to their host , however , the effective , global identification of bacterial virulence determinants and the characterization of their diverse mechanisms of action requires the development of new high-throughput and high-content screens ( HTS and HCS respectively ) [63] . Here , we have set up protocols for the multi-phenotypic screen of bacterial factors that are involved in host cell invasion and colonization . Our approach integrates transposon mutagenesis , genomics , bioinformatics and fluorescence-based functional assays that have been adapted for the large-scale identification of virulence factors from virtually any intracellular bacterium . The advantage of our screening technique lies in the possibility of analyzing every bacterial mutations for multiple phenotypes , such as 1 ) internalization within the host/cell , 2 ) intracellular replication and 3 ) cytotoxicity , simultaneously . This analysis , integrated with the map of genome mutations , allows a global overview of bacterial genes involved in host/pathogen interactions . The emerging bacterial pathogen Coxiella burnetii is an excellent model system to apply our strategy . To date , very little is known about the bacterial factors that regulate Coxiella interactions with the host; however , the recent development of axenic culture techniques now allows genetic manipulation of this microbe and in-depth analysis of its virulence factors [32] , [64] . Moreover , previous in silico identification of putative Coxiella virulence factors [18] , [26]–[31] , provides an excellent database to cross-reference bioinformatics analysis to our functional assays . Importantly , our assay allows the identification of Coxiella virulence factors on a whole-genome scale , based on the phenotypes associated with the random mutagenesis of Coxiella CDS . Our aim being the generation of the first bank of C . burnetii transposon mutants , we chose to sequence all isolated mutants independently of their phenotype during infections . This has provided a global survey of the distribution of transposon insertions and allowed us to pinpoint also those genes , suspected to encode virulence factors , which failed to produce a phenotype during Coxiella colonization of host cells . Mapping of transposon insertions revealed an overall homogeneous distribution of mutations , with regions of preferential transposon insertion as well as other regions that remained non-mutated . Of note , the lack of transposon insertions in the region between CBU_0678 and CBU_0698 is expected , having used the annotated genome of Coxiella NMI ( RSA493 ) to map transposon insertions in Coxiella NMII ( RSA439 ) in which this region is missing . Interestingly however , we failed to isolate transposon mutants from the large region between CBU_0215 and CBU_0272 , enriched in essential genes encoding ribosomal proteins . This suggests that non-mutated regions may have been targeted by the transposon but gene disruption was lethal for the bacterium . Non-mutated regions may be thus exploited to map essential genes in the Coxiella chromosome and plasmid . When analyzing transposon mutants exhibiting a strong replication phenotype , we have occasionally observed internalization phenotypes that were not consistently reproduced in all transposon mutants isolated for a given gene . We believe that this is due to a technical limitation of our screening technique , imposed by the extremely different sizes and associated fluorescence of Coxiella colonies . In some cases the signal to noise ratio was very close to the threshold imposed to the image analysis software and resulted in colonies and/or bacteria that were not detected when images were segmented . Ongoing implementation of our automated analysis pipeline will allow the identification of these outlier phenotypes . Independent studies have formally proven the essential role of the Coxiella Dot/Icm secretory apparatus by isolating mutants in 4 of the 23 dot/icm Coxiella genes ( dotA , dotB , icmD and icmL . 1 ) [18] , [19] , [36] . In all cases dot/icm mutants retained the capacity of invading host cells but failed to generate large PVs and replicate therein [18] , [19] , [36] . The enrichment of transposon insertions in the region of the Coxiella genome hosting dot/icm core genes allowed us to validate our assays exploiting existing data and , more importantly , provided a comprehensive overview of the role of the different components of the Coxiella T4SS . Interestingly , we have identified mutations resulting in intermediate phenotypes that allowed partial bacterial replication . Of particular interest is a mutation in icmS , which confers a multivacuolar phenotype to mutants . The icmS gene encodes a chaperone protein that mediates the secretion of a subclass of bacterial effectors [40] . The observation of a multivacuolar phenotype suggests that in Coxiella , IcmS may be involved in the secretion of effectors that mediate membrane fusion events required for the biogenesis of the PV . To facilitate the identification of putative IcmS substrates , a machine learning approach is currently being used to identify other transposon mutants that share the same multivacuolar phenotype . To date , candidate Coxiella effectors have been identified by bioinformatics analysis based on conserved Dot/Icm regulatory motif ( PmrA ) [26] , [27] , C-terminal translocation signals ( E-block ) [27] , [28] , and eukaryotic-like domains . So far , 354 candidate effectors have been thus identified , however Dot/Icm-dependent translocation assays using either Coxiella or Legionella as a surrogate model , indicated that the majority of these might be false positives [18] , [26]–[28] . In addition , the lack of efficient methods for the genetic manipulation of Coxiella severely hampered the functional study of these putative effectors . In a recent study , transposon mutants were obtained from 20 Coxiella candidate effectors with 10 exhibiting a significant replication phenotype [27] . Here we report the entry , replication and cytotoxic phenotype of 63 transposon mutants in 31 previously identified Coxiella candidate Dot/Icm substrates . Indeed , some of these candidates play a role during infection whereas some others fail to produce a phenotype , stressing the importance of coupling high-content screens to in silico analysis to identify bacterial effectors . Moreover , further studies of other Coxiella genes sharing none of the features of Dot/Icm substrates , can be exploited to enrich existing databases for the bioinformatics-based identification of Coxiella effectors , thus creating a feedback loop that would significantly improve and accelerate the study of Coxiella pathogenesis . A considerable number of transposon insertions were mapped outside Coxiella CDS . By excluding mutations that affected the first 100 bp upstream of annotated genes ( to exclude mutations that might affect promoter regions of genes ) , we obtained a list of 151 intergenic transposon insertions . Interestingly , 71 of these resulted in a significant replication phenotype , suggesting that these non-coding regions of the Coxiella genome may play a role in host/pathogen interactions . sRNAs are emerging as regulators that enable pathogens to adapt their metabolic needs during infection and timely express virulence factors [65] , [66] . However , recent studies in other organisms revealed the existence of a number of putative sRNAs higher than initially expected , suggesting the presence of many non-functional sRNAs and complicating the identification of relevant sRNAs . The functional data obtained by our screening approach are being cross referenced with a list of putative Coxiella sRNAs identified by RNA-seq to facilitate the identification of Coxiella sRNAs that may coordinate host/pathogen interactions . Similarly , the interesting observation that a number of mutations in Coxiella CDS annotated as pseudogenes have an effect in host cell infection suggests that these genomic regions may have an important regulatory role . Bacterial adhesion and invasion of host cells is a fundamental step of the infection by intracellular bacterial pathogens [67] . These processes can be active or passive depending on the nature of the pathogen . “Triggering” bacteria commonly use a type 3 secretion system ( T3SS ) to inject effectors across the host cell plasma membrane to trigger actin rearrangements and pathogen internalization by phagocytosis , whereas “zippering” bacteria use surface proteins that interact with cognate receptors at the surface of host cells [67] . This activates a ligand/receptor signaling cascade that leads to the internalization of large particles by an endocytosis-like mechanism [68]–[71] . The lack of a T3SS in Coxiella suggests that these organisms adhere to and invade host cells by a zippering mechanism . Indeed , it has been reported that Coxiella is passively internalized by host cells by a yet undefined mechanism , which is accompanied by the local rearrangement of the actin cytoskeleton [11] , [14] , [15] . αVβ3 integrins have been shown to mediate Coxiella adhesion to THP-1 cells [11] , however , the lack of these integrins at the surface of epithelial cells , which are effectively colonized by Coxiella , suggest the presence of additional/alternative receptors . Similarly , the Coxiella surface determinants for host cell adhesion and invasion remain to be defined . Here we have identified the product of CBU_1260 as the first Coxiella invasin . Predictive analysis on the primary sequence of CBU_1260 revealed the presence of 8 beta sheets forming an OmpA-like domain highly homologous to that identified and characterized in several other bacterial pathogens [49] , [50] . Examples are the OmpA proteins encoded by E . coli K1 [44] , [47] , Yersinia pestis [45] , Francisella tularensis [46] , Klebsiella pneumoniae [48] and Shigella flexneri [72] . These outer membrane proteins are involved in bacterial adhesion and/or internalization within host cells , as well as in the NF-κB-mediated modulation of the immune response to infection , which is required for intracellular bacterial development . Importantly , OmpA-like proteins with similar functions in bacterial adhesion and internalization have been reported in other bacterial pathogens such as Rickettsia conorii , Anaplasma phagocytophilum and Ehrlichia chaffeensis [73]–[75] , however these proteins share no structural homology with the OmpA proteins described above . In agreement with in silico predictions , membrane fractionation experiments performed in Coxiella as well as in E . coli ectopically expressing OmpA , showed that the protein is indeed enriched in the outer membrane fraction of bacterial lysates . Our multi-phenotypic analysis revealed that five independent transposon insertions that disrupted CBU_1260 sequence severely affected Coxiella internalization and replication within host cells . The internalization phenotype was specific of non-phagocytic cells , whereas OmpA mutants were still internalized by phagocytic cells . This observation indicated that , in the absence of an active phagocytic process , OmpA is able to actively trigger Coxiella internalization by means of ligand/receptor interactions . Importantly however , the intracellular replication of OmpA mutants was severely affected in both epithelial and macrophage cell lines . This phenotype is in line with a reported role of OmpA proteins in facilitating bacterial survival within host cells [46] , [48] , [51] , [53] . Importantly , OmpA-like proteins share conserved transmembrane domains but are characterized by extremely variable extracellular domains , which are unique to each pathogen , and confer specific functions [50] . OmpA was predicted to have 4 unstructured loops exposed at the bacterial surface . By replacing each loop with a myc tag , we have generated 4 OmpA mutants ( OmpA ΔL1 , ΔL2 , ΔL3 and ΔL4 ) and showed that loop 1 is essential to confer invasiveness to E . coli ectopically expressing the OmpA mutants . Accordingly , a specific antibody against loop 1 effectively blocks OmpA function . Bioinformatics analysis indicated the presence of 2 additional OmpA-like proteins in the Coxiella genome , CBU_0307 and CBU_0936 , sharing with OmpA a good degree of homology at the level of the transmembrane OmpA-like domain but no significant homology in the 4 extracellular loops . In line with these observations , CBU_0307 and CBU_0936 failed to produce internalization phenotypes when mutated by transposon insertions . Finally , experiments aiming at blocking potential OmpA interactions with a cognate receptor , effectively blocked Coxiella internalization , indicating the presence of an interacting partner at the surface of host cells , which remains to be identified . Using Galleria mellonella larvae as a surrogate in vivo model system we could reproduce the OmpA mutant phenotypes observed in cultured cells . Indeed , only wt Coxiella and the control mutant Tn1832 were able to induce the formation of nodules that were abundantly colonized by Coxiella and disrupt the organization of Galleria peripheral organs . The OmpA mutant Tn208 induced a milder formation of nodules that presented few , isolated bacteria . Accordingly , larvae infected with the OmpA mutant survived Coxiella infections longer than those infected with the control mutants . In summary , multi-phenotypic screening of host/pathogen interactions is an efficient method for the study of infectious diseases . Here we have applied this method to Coxiella infections and identified a bacterial protein that is essential for Coxiella internalization within non-phagocytic cells . Understanding how intracellular bacteria adhere to and invade their host is essential to 1 ) understand the cell biology of infection and identify the candidate targets of anti-infectious molecules and 2 ) to develop targeted vaccines . Of note , bacterial OmpA proteins are considered as new pathogen-associated molecular patterns ( PAMPs ) and are among the most immuno-dominant antigens in the outer membrane of Gram-negative bacteria [50] , [76] . Our laboratory currently investigates the possibility of using OmpA to develop a synthetic vaccine against Q fever . Strains used in this study are listed in Fig . S9 . Escherichia coli strains were grown in Luria-Bertani ( LB ) medium supplemented with ampicillin ( 100 µg/ml ) , kanamycin ( 50 µg/ml ) or chloramphenicol ( 30 µg/ml ) as appropriate . Coxiella burnetii NMII and transposon mutants were grown in ACCM-2 [77] supplemented with kanamycin ( 340 µg/ml ) or chloramphenicol ( 3 µg/ml ) as appropriate in a humidified atmosphere of 5% CO2 and 2 . 5% O2 at 37°C . Cells were routinely maintained in RPMI ( Vero , THP-1 , J774 and RAW 264 . 7 ) or DMEM ( A431 and HeLa ) , containing 10% fetal calf serum ( FCS ) in a humidified atmosphere of 5% CO2 at 37°C . For experiments , THP-1 were allowed to differentiate into macrophages for 2 days in the presence of 200 nM phorbol myristate acetate ( PMA , Sigma ) . Hoechst 33258 , rabbit anti poly-His , anti mouse and anti-rabbit HRP-conjugated antibodies and Atto-647N phalloidin were purchased from Sigma . Rabbit anti Coxiella NMII antibodies were kindly provided by Robert Heinzen . Synthesis and production of the peptide KKSGTSKVNFTGVTL , as well as the generation of the cognate antibody in rabbit ( named anti-OmpA in this study ) were performed by Eurogentec , Belgium . Mouse and rabbit IgG conjugated to Alexa Fluor 488 and 555 as well as Prolong Gold antifade mounting reagent were purchased from Invitrogen . Paraformaldehyde was provided by Electron Microscopy Sciences , PA . Plasmids and primers used in this study are listed in Fig . S9 . DNA sequences were amplified by PCR using Phusion polymerase ( New England Biolabs ) and gene-specific primers ( Sigma ) . To create the plasmid pITR-CAT-ColE1-P311-GFP , the promoter of CBU_0311 ( P311 ) was amplified from Coxiella RSA439 NMII genomic DNA using primers P311-XhoI-Fw and P311-Rv , GFP was amplified from pEGFP-N1 ( Clontech ) using primers EGFP-Fw and GFP-PITR-Rv , and P1169-CAT-ColE1 was amplified from plasmid pITR-CAT-ColE1 using primers GFP-PITR-Fw and XhoI-PITR-Rv . PCR fragments P311 , GFP and P1169-CAT-ColE1 were fused by overlapping PCR . The resulting PCR product was digested with XhoI and ligated to obtain circular pITR-CAT-ColE1-P311-GFP . OmpA32-248 was amplified from Coxiella RSA439 NMII genomic DNA using primers OmpA32-248-BamHI-Fw and OmpA-EcoRI-Rv and cloned into pET28a to obtain pET28a-OmpA32-248 . OmpA was amplified from Coxiella RSA439 NMII genomic DNA using primers OmpA-BamHI-shift-Fw and OmpA-EcoRI-Rv and cloned into pET27b to obtain pET27b-OmpA . Plasmids pET27b-OmpAΔL1 , pET27b-OmpAΔL2 , pET27b-OmpAΔL3 and pET27b-OmpAΔL4 were generated by PCR using pET27b-OmpA as template and primer pairs loop1-myc-HindIII-Fw/loop1-myc-HindIII-Rv , loop2-myc-HindIII-Fw/loop2-myc-HindIII-Rv , loop3-myc-HindIII-Fw/loop3-myc-HindIII-Rv , loop4-myc-HindIII-Fw/loop4-myc-HindIII-Rv . The PCR products were digested with HindIII and ligated to obtain the corresponding plasmids . C . burnetii RSA439 NMII organisms were electroporated with the pITR-CAT-ColE1-P311-GFP and pUC19::Himar1C9 plasmids [34] using the following setup: 18 kV , 400 Ω , 25 µF . Bacteria were then grown overnight in ACCM-2 supplemented with 1% FBS and the following day 3 µg/ml chloramphenicol were added to bacterial cultures . Bacteria were then amplified for 4 days and plated on solid ACCM-2 for clone isolation . Seven days post-inoculation , colonies were isolated and amplified for 6 days in liquid ACCM-2 supplemented with 3 µg/ml chloramphenicol . The concentration of each isolated mutant was quantified using the PicoGreen ( Invitrogen ) assay according to manufacturer's instructions . To map transposon insertions , single primer colony PCR was performed on 1 µl of C . burnetii transposon mutant in stationary phase in ACCM-2 . The PCR mix contained 1× HF buffer ( New England Biolabs ) , 200 µM dNTPs , 1 µM primer SP3 and 1 U of Phusion polymerase ( New England Biolabs ) . The PCR cycle consisted in initial denaturation ( 98°C , 1 min ) , 20 high stringency cycles ( 98°C , 10 sec; 50°C , 30 sec; 72°C , 90 sec ) , 30 low stringency cycles ( 98°C , 10 sec; 30°C , 30 sec; 72°C , 90 sec ) and 30 high stringency cycles ( 98°C , 10 sec; 50°C , 30 sec; 72°C , 90 sec ) followed by a final extension at 72°C for 7 min . PCR products were then sequenced at Beckman Coulter Genomics ( Stansted , UK ) using primer P3 . Insertion sites were mapped on the annotated C . burnetii RSA493 NMI genome using MacVector ( MacVector Inc . ) and recorded in a relational database ( FileMaker ) . 106 GE/ml of bacteria were inoculated in 4 ml ACCM-2 and allowed to grow for 8 days in a humidified atmosphere of 5% CO2 and 2 . 5% O2 at 37°C . Where needed , 3 µg/ml chloramphenicol were added to bacterial cultures . At the indicated time points bacterial concentrations were evaluated from 100 µl of cultures , using the PicoGreen ( Invitrogen ) assay according to manufacturer's instructions . Vero cells were seeded into triplicate 96-wells plates ( Greiner Bio one ) 2 days prior to infection . Cells were then challenged with C . burnetii RSA439 NMII transposon mutants at an MOI of 100 . For each plate , cells in well A1 were left uninfected and cells in wells A2 and A3 were incubated with GFP-C . burnetii RSA439 NMII at multiplicities of infection of 100 and 200 . Bacterial contact with cells was promoted by centrifugation ( 10 min , 400 g , RT ) and cells were incubated in a humidified atmosphere of 5% CO2 at 37°C . Unbound bacteria were removed after 1 h of incubation and cells were further incubated in fresh culture medium for 7 days . Plates were analyzed at a 24-hours interval using a TECAN Infinite 200 Pro operated by the Magellan software ( TECAN ) to monitor the variations of GFP fluorescence associated with the intracellular growth of Coxiella . Raw data were analyzed for background subtraction , normalization and quality control among triplicates using in-house developed methods . Seven days after infection , plates were fixed in 3% paraformaldehyde in PBS at room temperature for 30 minutes , rinsed in PBS and incubated in blocking solution ( 0 . 5% BSA , 50 mM NH4Cl in PBS , pH 7 . 4 ) . Cells were then incubated in Hoechst 33258 diluted 1∶200 in blocking solution for 30 minutes at room temperature , rinsed and incubated in PBS . Images were acquired with an Arrayscan VTI Live epifluorescence automated microscope ( Cellomics ) equipped with an ORCA ER CCD camera . 6 fields/well of triplicate 96-wells plates were imaged with a 20× objective in the GFP , DAPI and Bright-field channels . Images were then processed and analyzed using CellProfiler . Briefly , the GFP channel was subtracted from the corresponding DAPI channel to avoid false identification of large Coxiella colonies as host cell nuclei , images were thresholded using the Otsu global method and host cell nuclei as well as Coxiella colonies were identified and segmented . The number , form factor and fragmentation of host cell nuclei and the number , form factor , area , perimeter , GFP intensity and compactness of Coxiella colonies were then calculated per object and per image . An average of 14000 cells per condition ( infection with a given Coxiella mutant ) were thus analyzed . Raw data were processed for background subtraction , normalization and quality control among the 6 fields per well and plate triplicates using in-house developed methods . Data were recorded in a relational database ( FileMaker ) that allowed clustering of phenotypes according to the annotated transposon insertions . Cells were fixed in 3% paraformaldehyde in PBS at room temperature for 30 minutes , rinsed in PBS and incubated in blocking solution ( 0 . 5% BSA , 50 mM NH4Cl in PBS , pH 7 . 4 ) . When appropriate , 0 . 05% saponin was added to the blocking solution for cell permeabilization . Cells were then incubated with the primary antibodies diluted in blocking solution for 1 h at room temperature , rinsed five times in PBS and further incubated for 45 min with the secondary antibodies diluted in the blocking solution . Fluorescent phalloidin was added to the secondary antibodies to label actin , where needed . After labeling , coverslips were mounted using Prolong Gold antifade mounting medium ( Invitrogen ) supplemented with Hoechst 33258 for DNA staining . For differential labeling , extracellular bacteria or beads were stained using specific antibodies without permeabilizing the cells . Intracellular bacteria or beads were visualized by green and red fluorescence , respectively . Alternatively , a second staining was performed after cellular permeabilization . Secondary antibody labeling using two different fluorochromes ( before and after permeabilization ) allowed discrimination between adherent extracellular bacteria/beads and those that have been internalized . Samples were analyzed with a Zeiss Axioimager Z1 epifluorescence microscope ( Carl Zeiss ) connected to a Coolsnap HQ2 CCD camera . Images were acquired alternatively with 63× or 40× oil immersion objectives and processed with MetaMorph ( Universal Imaging Corp . ) . Image J and CellProfiler software were used for image analysis and quantifications . 3D reconstruction and surface rendering were performed using the Osirix software . Transposon insertions in Dot/Icm core genes were complemented in trans as previously described [19] . Briefly , Vero cells grown on 96-wells plates were either challenged with the transposon mutants alone or in combination with wt Coxiella at a 1∶1 ratio for a total MOI of 100 . Bacterial contact with cells was promoted by centrifugation ( 10 min , 400 g , RT ) and cells were incubated in a humidified atmosphere of 5% CO2 at 37°C . Unbound bacteria were removed after 1 h of incubation and cells were further incubated in fresh culture medium for 7 days . Plates were then fixed in 3% paraformaldehyde in PBS at room temperature for 30 minutes , rinsed in PBS and incubated in blocking solution ( 0 . 5% BSA , 50 mM NH4Cl in PBS , pH 7 . 4 ) . Cells were then labeled with the anti-NMII antibody to detect wt Coxiella and with Hoechst 33258 as described above for DNA labeling . 96-well plates were imaged and analyzed essentially as described above for the mutant library screening protocol with the additional acquisition of the TRITC channel to detect and segment wt Coxiella colonies . Object masking was applied using CellProfiler to specifically calculate the area of mutant Coxiella colonies growing within wt Coxiella-occupied PVs . The terminal deoxyribonucleotidyl transferase-mediated triphosphate ( dUTP ) -biotin nick end labeling ( TUNEL ) method was used for detection of DNA fragmentation of nuclei using the In Situ Cell Death Detection Kit TMR ( Roche ) according to the manufacturer's instructions . Briefly , HeLa cells grown on 96-well plates in triplicate were either left untreated or challenged with the indicated Coxiella strains at an MOI of 100 and incubated at 37°C for 3 days . Cells were then either fixed and permeabilized as described above or incubated with 1 µM staurosporine for 4 hours prior to fixation . Samples were then incubated 1 h at 37°C in the dark with TUNEL reaction mixture . Samples were then washed three times with PBS , and incubated with Hoechst 33258 for DNA staining and with an anti-NMII antibody to detect bacteria in the samples infected with wt Coxiella . 96-wells plates were analyzed with an Arrayscan VTI Live epifluorescence automated microscope ( Cellomics ) equipped with an ORCA ER CCD camera . 10 fields/well were imaged with a 20× objective in the GFP ( bacteria ) , TRITC ( TUNEL ) , DAPI ( nuclei ) and Bright-field ( cells ) channels . Images were then processed and analyzed using CellProfiler . Briefly , the GFP channel was subtracted from the corresponding DAPI channel to avoid false identification of large Coxiella colonies as host cell nuclei , images were thresholded using the Otsu global method and host cell nuclei , Coxiella colonies and fragmented nuclei were identified and segmented . The percentage of fragmented nuclei over the total number of nuclei was then calculated on an average of 6000 cells per condition . Cells were washed in PBS and fixed with 2 . 5% glutaraldehyde in Sorensen buffer , pH 7 . 2 for an hour at room temperature , followed by washing in Sorensen buffer . Fixed samples were dehydrated using a graded ethanol series ( 30–100% ) , followed by 10 minutes in graded Ethanol-Hexamethyldisilazane and finally Hexamethyldisilazane alone . Subsequently , the samples were sputter coated with an approximative 10 nm thick gold film and then examined under a scanning electron microscope ( Hitachi S4000 , at CRIC , Montpellier France ) using a lens detector with an acceleration voltage of 20 kV at calibrated magnifications . Galleria mellonella larvae were fixed overnight in paraformaldehyde 4% in PBS ( pH 7 . 4 ) . Larvae were then rinsed 3 times in PBS and cryo protected by successive incubations in PBS containing increasing concentrations of sucrose ( 10% , 20% , 30% ) . Samples were then frozen in isopentane at −80°C using a SnapFrost machine ( Excilone ) . Consecutive 20 µm sections were then obtained from each sample using a Leica CM3050S cryostat . For indirect immuno-fluorescence , sections were permeabilized and blocked in PBS , 10% goat serum , 0 . 3% Triton X-100 for 1 h at room temperature . Samples were then incubated 48 hours at 4°C with the anti GFP antibody , then rinsed in PBS . Samples were then incubated with the appropriate secondary antibodies , Atto-647N phalloidin and Hoechst 33258 for 24 hours at 4°C . Samples were then washed in PBS and mounted on glass slides for microscopy analysis . Samples were analyzed either with an EVOS microscope ( AMG ) or with an ApoTome-equipped Zeiss Axioimager Z1 epifluorescence microscope ( Carl Zeiss ) connected to a Coolsnap HQ2 CCD camera . Images were acquired alternatively with a 10× objective ( EVOS ) or with a 63× oil immersion objective ( Axioimager Z1 ) and processed with Image J and AxioVision ( Carl Zeiss ) . Genes cloned into pET27b or pET28a vectors were expressed in E . coli BL21-DE3 star pLysS ( Invitrogen ) . Bacterial cultures were grown at 25°C to mid-exponential phase ( OD600 nm = 0 . 5 ) and were induced overnight with 400 µM isopropyl-β-D- thiogalactopyranoside ( IPTG ) . For GST expression , E . coli XL1-blue were transformed with pGEX-4T1 , grown at 37°C to mid-exponential phase ( OD600 nm = 0 . 5 ) and induced for 4 h with 1 mM IPTG . Bacteria were harvested by centrifugation , resuspended in lysis buffer ( 20 mM Tris pH 8 , 300 mM NaCl , 5% glycerol , complete anti-protease ( Roche ) ) and lysed with BugBuster ( Novagen ) following the manufacturer's recommendations . Lysates were then cleared by centrifugation ( 11 000 g , 20 min , 4°C ) . Proteins were purified by gravity flow using Ni2+ agarose His-select resin column ( Sigma ) for His-tagged proteins or glutathione-sepharose ( Sigma ) for GST . His-tagged and GST proteins were eluted with lysis buffer supplemented with 250 mM imidazole or 25 mM reduced glutathione , respectively . For Coxiella membrane fractionation , 100 ml of wt Coxiella or Tn208 mutant grown in ACCM-2 for 7 days were pelleted and resuspended in 200 µl 20 mM Tris pH 8 containing 1× Complete protease inhibitor ( Roche ) . For E . coli membrane fractionation , 30 ml of IPTG-induced or non-induced E . coli BL21-DE3 star pLysS pET27b-OmpA , pET27b-OmpAΔL1 , pET27b-OmpAΔL2 , pET27b-OmpAΔL3 or pET27b-OmpAΔL4 were pelleted and resuspended in 5 ml 20 mM Tris pH 8 containing 1× Complete protease inhibitor ( Roche ) . Bacteria were sonicated using a Branson Sonifier S-450 ( 6 pulses of 20 s at 40% intensity ) and cleared by centrifugation at 10000 g for 5 min at 4°C . Inner membrane proteins were extracted by incubation with sarkosyl ( 0 . 5% final concentration ) at RT for 15 min . Outer membrane proteins were pelleted by ultracentrifugation ( TLA-100 , 32000 r . p . m . , 30 min , 4°C ) and resuspended in 2× Laemmli sample buffer . Insoluble , soluble/sarkosyl-solubilized and outer membrane fractions were resolved by SDS-PAGE and analyzed by Coomassie staining ( Sigma ) or immunoblotting with anti-OmpA and anti NMII antibodies . 0 . 5 µm fluorescent red sulfate-modified polystyrene beads ( Sigma ) were washed three times with 25 mM MES pH 6 . 1 ( MES buffer ) . The sulfate-modified beads ( 7 . 2×109 ) were then mixed with either 100 µg/ml purified GST or His-OmpA32-248 and incubated at room temperature ( RT ) for 4 h . The GST- or His-OmpA32-248-coated beads were then washed three times with MES buffer and resuspended in MES buffer containing 1% BSA . For fluorescent beads internalization assay , 7×107 GST- or His-OmpA32-248-coated beads in DMEM were applied to 1×105 A431 cells seeded onto glass coverslips in 24-well plates and contact was promoted by centrifugation ( 10 min , 400 g , RT ) . Cells were incubated in a humidified atmosphere of 5% CO2 at 37°C . Cells were washed three times with PBS and fixed in 4% paraformaldehyde before being processed for immunofluorescence staining . Bacteria internalization assays were performed as follow: 6×106 C . burnetii NMII GE were applied to cells ( MOI 100 ) and contact was promoted by centrifugation ( 10 min , 400 g , RT ) . Cells were then fixed in 4% paraformaldehyde before being processed for immunofluorescence staining . For protein blocking experiments , A431 cells were pre-incubated for 1 h at 4°C with either 100 µg/ml of GST or 100 µg/ml of His-OmpA32-248 prior to the internalization assay described above . For antibody inhibition experiments , 6×106 C . burnetii RSA439 NMII GE were incubated at 4°C with increasing concentrations of either naïve rabbit serum or anti-OmpA antibodies ( 0 . 1 to 5 µg/ml ) prior to the internalization assay described above . Gentamicin protection assays were performed as follow: IPTG-induced and non-induced BL21-DE3 star pLysS pET27b-OmpA , pET27b-OmpAΔL1 , pET27b-OmpAΔL2 , pET27b-OmpAΔL3 or pET27b-OmpAΔL4 were diluted to OD600 nm = 0 . 5 in DMEM and applied to cells at an MOI of 10 and contact was promoted by centrifugation ( 10 min , 400 g , RT ) . Following 1 h incubation at 37°C/5% CO2 , cells were washed 5 times with PBS and incubated for 2 h with DMEM supplemented with 100 µg/ml gentamicin . Cells were then washed 5 times with PBS , lysed with 1 ml 0 . 1% Triton X-100 in ddH2O and serial dilutions were plated onto LB agar plates supplemented with the appropriate antibiotic for colony-forming units ( CFU ) assessment . Internalization frequency was determined as the number of CFU surviving the gentamicin challenge out of the total bacterial input . The results are representative of at least three independent experiments .	Infectious diseases are among the major causes of mortality worldwide . Pathogens‚ invasion , colonization and persistence within their hosts depend on a tightly orchestrated cascade of events that are commonly referred to as host/pathogen interactions . These interactions are extremely diversified and every pathogen is characterized by its unique way of co-opting and manipulating host functions to its advantage . Understanding host/pathogen interactions is the key to face the threats imposed by infectious diseases and find alternative strategies to fight the emergence of multi-drug resistant pathogens . In this study , we have setup and validated a protocol for the rapid and unbiased identification of bacterial factors that regulate host/pathogen interactions . We have applied this method to the study of Coxiella burnetii , the etiological agent of the emerging zoonosis Q fever . We have isolated , sequenced and screened over 1000 bacterial mutations and identified genes important for Coxiella invasion and replication within host cells . Ultimately , we have characterized the first Coxiella invasin , which mediates bacterial internalization within non-phagocytic cells . Most importantly , our finding may lead to the development of a synthetic vaccine against Q fever .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "gram", "negative", "microbial", "mutation", "emerging", "infectious", "diseases", "gene", "identification", "and", "analysis", "microbial", "pathogens", "molecular", "genetics", "host-pathogen", "interaction", "microbiology", "biology", "pathogenesis", "computational", "bi...	2014	Identification of OmpA, a Coxiella burnetii Protein Involved in Host Cell Invasion, by Multi-Phenotypic High-Content Screening
Copy number variants ( CNVs ) have recently been recognized as a common form of genomic variation in humans . Hundreds of CNVs can be detected in any individual genome using genomic microarrays or whole genome sequencing technology , but their phenotypic consequences are still poorly understood . Rare CNVs have been reported as a frequent cause of neurological disorders such as mental retardation ( MR ) , schizophrenia and autism , prompting widespread implementation of CNV screening in diagnostics . In previous studies we have shown that , in contrast to benign CNVs , MR-associated CNVs are significantly enriched in genes whose mouse orthologues , when disrupted , result in a nervous system phenotype . In this study we developed and validated a novel computational method for differentiating between benign and MR-associated CNVs using structural and functional genomic features to annotate each CNV . In total 13 genomic features were included in the final version of a Naïve Bayesian Tree classifier , with LINE density and mouse knock-out phenotypes contributing most to the classifier's accuracy . After demonstrating that our method ( called GECCO ) perfectly classifies CNVs causing known MR-associated syndromes , we show that it achieves high accuracy ( 94% ) and negative predictive value ( 99% ) on a blinded test set of more than 1 , 200 CNVs from a large cohort of individuals with MR . These results indicate that this classification method will be of value for objectively prioritizing CNVs in clinical research and diagnostics . Improvements in microarray resolution and hybridization robustness have resulted in the widespread implementation of genomic microarray technologies in medical research and diagnostics . This technology is most effective in detecting genomic deletions and duplications larger than 1kb , known as copy number variants ( CNVs ) . Genomic microarrays are commonly used to identify rare , but highly penetrant , and commonly single CNVs in patients suffering from neurological disorders such as autism [1]–[3] , schizophrenia [4]–[6] and mental retardation ( MR; also known as learning disability ) [7]–[9] . However CNVs have also been recently recognized as a common form of genomic structural variation: high resolution microarrays and sequencing approaches are able to identify 600–900 CNVs in a single individual [10]–[14] . Current clinical interpretation therefore needs to contrast the frequencies of a CNV in affected versus unaffected individuals , as well as determining the inheritance of CNVs via parental analysis [15] , [16] . The identification of a CNV that is ( 1 ) relatively large , ( 2 ) overlaps genes , ( 3 ) is rare , and ( 4 ) de novo in a patient provides a strong indicator of clinical significance , because this combination is extremely rare in the normal population owing to a low structural mutation rate outside of hypervariable ‘hot spot’ regions [10] , [17] , [18] . Increases in microarray resolution are revealing both a much higher rate of rare CNVs than previously thought [19] and an increasing number of genomic loci being reported that show variable inheritance and penetrance . Such examples have been reported for CNVs at 1q21 . 1 [20] , [21] , 15q13 . 3 [22] , [23] , and 16p13 . 11 [24] , [25] . These loci demonstrate that there are limitations in considering CNVs as either benign when common and inherited , or causal when rare and de novo . At present up to 5% of the human genome has been shown to vary in large scale copy number in numerous healthy controls [13] , [26] and novel CNVs continue to be identified [27] . In Nguyen et al . ( 2008 ) [28] we reported a number of genomic features whose frequencies are significantly different in apparently benign CNV regions compared with the genome as a whole . In particular , CNV regions are enriched in repetitive sequences of near identical DNA known as segmental duplications [29] and are less prone to recombination . Furthermore , these CNV regions are characterized by tendencies to coincide with between-species break-points in synteny and to be prone to elevated nucleotide substitution rates , whilst their encoded proteins tend to exhibit elevated evolutionary rates . In a separate study we compared a large set of rare de novo CNVs associated with MR with CNVs identified in healthy control individuals . This study demonstrated that MR-associated CNVs are significantly enriched in genes whose mouse orthologues , when disrupted , result in abnormal axon or dopaminergic neuron morphologies , and in genes from neurodegenerative disease pathways [30] . Importantly , we showed that benign CNVs do not display such properties . Such observations can thus now be used to prioritize dosage-sensitive candidate genes for MR . Of relevance to this study is that these distinctions of MR-associated CNVs may be exploited to aid the development of an objective method for distinguishing disease-associated CNVs from benign CNVs that does not rely solely on allele inheritance and frequency in the normal population . Although a large number of methods are available for the computational prioritization and classification of genomic data [31]–[34] , none thus far has been developed specifically for CNV data . For this study we implement a Naïve-Bayesian Tree classifier ( NBTree ) . This hybrid approach combines a decision tree with Naïve-Bayesian classifiers , and exploits the segmentation of decision trees and the accumulation of Naïve-Bayes evidence . There are four major advantages of decision-tree classifiers for assigning pathogenicity to CNVs . These classifiers are ( i ) fast and ( ii ) their results are easily comprehensible . They are ( iii ) very robust to irrelevant features and ( iv ) classification takes into account evidence from many attributes in arriving at a final prediction [35] . In this study our aim was to validate the use of an NBTree , based upon genomic features , to accurately separate disease-associated CNVs from benign CNVs . We identified a total of 16 genomic features as suitable attributes for the classifier which could be divided into either: ( 1 ) structural features such as segmental duplication density , and ( 2 ) functional features , such as gene density ( Table 1 ) . These genomic attributes were also considered to be either continuous or categorical features . To compensate for the dependencies of CNV length on the frequencies of features ( e . g . LINE , SINE , segmental duplication and gene numbers ) we also calculated the densities of LINEs , SINEs , segmental duplications and ENSEMBL gene models . A categorical feature was created to be set as ‘true’ when a CNV contains at least one gene whose mouse orthologue , when disrupted exhibits a mouse nervous system phenotype ( and otherwise ‘false’ ) . Previously we have shown that MR CNV genes are enriched in the KEGG neurodegenerative pathway ( namely , hsa01510 ) . This feature was also represented in the classifier , specifically as a categorical feature when at least one CNV gene is a member of this KEGG pathway . Finally , we incorporated in the classifier information regarding the gene expression variance from microarray expression experiments performed in 176 HapMap EBV cell lines , reasoning that dosage-sensitive genes tend to show less variable expression levels [36] , [37] . In this study we present a novel computational method to objectively identify clinically relevant CNVs using an NBTree classifier and 13 diverse genomic features . This is the first description of such a method applied to CNVs that can significantly improve interpretation of this important class of genomic variation . Our classification method has been validated on a set of 1 , 203 CNVs detected in 584 patients with MR , achieving a high accuracy ( 94% ) , with a sensitivity of 88% and a specificity of 94% ( Figure 3a ) . Several other computational methods have been developed previously to predict if disruption or disturbance of genomic elements have pathogenic consequences . Often these methods are focused on identifying disease genes or on predicting if mutation or splicing events are pathogenic [31]–[34] . Such methods make use of protein structure and stability measures , and phylogenetic or sequence conservation data [38] , [39] , and often cross-validate their predictions using OMIM ( Online Mendelian Inheritance in Man ) data [40] . These approaches may be less applicable for larger structural variants such as CNVs because they predict the effect of a single change on a single disease gene , rather than a large change involving many genes . Our approach differs in that we directly predict the causal CNV from genome-wide copy number scans on the basis of the distinguishing features of benign and disease-causing CNVs . In addition , OMIM does not provide a suitable source for validating the performance of a classification method for CNVs as dosage-sensitive genes are largely underrepresented in this database ( <5% of the entries describe haploinsufficient genes [41] ) , and because a precise mapping of CNVs in OMIM is lacking . In contrast to OMIM , the Decipher database list of known syndromes ( https://decipher . sanger . ac . uk ) provides a suitable list of CNVs for external validation of the classifier with high-resolution mapping of their genomic locations . Our classification method correctly identified all the CNVs listed in this database as causing MR-associated syndromes . The classifier incorporated specific knowledge about CNVs via 13 diverse structural and functional genomic features ( including a number of different transposable element types ) . The proximity of these elements to CNVs has been reported previously and it has been hypothesized that they mediate the formation of recurrent CNVs [18] , [26] , [42] . We confirm previous results that benign CNVs are enriched in both LINE and segmental duplication elements [13] , [28] and show that both the LINE density and the segmental duplication density substantially contribute to the classifier's accuracy ( Table S2 ) . Previous studies have also reported that CNV gains are enriched in many of the same features as CNV losses [30] . Our feature contribution results support this finding: when the CNV type was removed from the classifier only a 3 . 7% decrease in accuracy was observed , and 7 additional features had a greater contribution to the classifier's accuracy . In addition to these transposable elements , we included functional genomic elements which have recently been shown to assist in distinguishing benign from MR-associated CNVs [30] , [43] . The significant enrichment of MGI mouse nervous system phenotypes in MR loss CNVs has previously been reported [30] . We show that the MGI mouse knock-out phenotype feature is effective in distinguishing benign from MR-associated CNVs: 80% of all MR-associated CNVs contain one or more genes whose unique orthologue's disruption in mouse reveals a nervous system phenotype , whereas benign CNVs only rarely contain such genes ( Table S2 ) . Despite the MGI mouse phenotype dataset being incomplete , this feature contributes greatly to the classifier's accuracy ( 5% ) . To date , gene knockout experiments with recorded ontology based phenotype information have been performed for approximately 5 , 000 of the possible 15 , 287 genes with mouse 1∶1 orthologues [44] , [45] . Furthermore the MGI phenotype data are included in the classifier as a binary feature ( which is labelled as ‘true’; when a CNV contains 1 or more genes exhibiting a nervous system phenotype; MP:0003631 ) . However , as the MGI phenotype dataset is incomplete , our approach is conservative with respect to missing values . This is because CNVs overlapping genes whose disruption does not result in a nervous system phenotype are weighted equally to those CNVs overlapping genes whose disruption phenotypes are currently unknown . Thus , we expect that increased coverage by the MGI mouse knock-out dataset will significantly improve the accuracy of the classifier . In addition , further genomic features such as CpG islands or conserved non-coding regions [46] can now be tested for their potential to improve the accuracy of this approach . Nevertheless , as the densities of many genomic features are strongly correlated [28] , it is likely that the addition of further features to the classifier will not result in a substantial improvement in predictive power . Most of the CNVs we used to train the classifier were identified on low-resolution ( BAC–based ) microarray platforms . In contrast , the replication set contained CNVs collected solely from Affymetrix 250k SNP microarrays . Despite the different microarray technologies used , only a negligible decrease in classification accuracy ( −1 . 7% ) was observed between the training and the replication set . This indicates that the classifier is platform-independent and will not require retraining when used on data generated from comparable microarray platforms . MR-associated CNVs discovered thus far are , in general , larger than benign CNVs [30] . Previously developed CNV risk assessments for identifying disease-associated CNVs use a length greater than 3Mb as a distinguishing criterion [16] . Closer inspection of the MR-associated CNVs from our validation study indeed revealed a larger mean length ( 6 . 8Mb ) compared to the benign CNVs ( 474kb ) . Despite this large size , 25% of the MR-associated CNVs in the validation set were smaller than 1 . 1Mb . We separately tested the accuracy of the classifier on CNVs smaller than 1 . 1Mb which revealed it to exhibit a decrease in sensitivity ( −18% ) but still a high accuracy ( 93% ) . As might be expected , small MR-associated CNVs showed a decrease in the number of MGI knock-out genes displaying a nervous system phenotype , but their SINE and gene densities are comparable to those of larger MR-associated CNVs ( Table S2 ) . Importantly , the classifier was still able to correctly classify 9 of the 13 small MR-associated CNVs , demonstrating the advantage of the classifier in comparison to conventional interpretation methods which often are unable to clearly identify clinically relevant CNVs unless specific information about their genomic content is known [47] . Although current clinical interpretation of CNVs focuses on large , rare and de novo CNVs , an increasing number of genomic loci being reported show variable inheritance and penetrance [20]–[24] . Our replication study contained a number of such CNVs , including CNVs at 1q21 . 1 and 15q13 . 3 which , in addition , show variation in genomic size and content [20]–[23] . Three rare inherited CNVs encompassing the 1q21 . 1 critical region were all classified as associated with MR , even though their genomic breakpoints differed . Two rare de novo CNVs in the 15q13 . 3 region were classified differently , one as benign and one as pathogenic . In addition , three inherited CNVs at this locus were all classified as benign . Interestingly , the distal breakpoint for all five CNVs was identical whereas the proximal breakpoint of the four CNVs classified as benign was extended by an additional 150kb . This difference in classification is explained by the fact that the 150kb region showed a higher repeat element count and density due to repetitive elements surrounding the 15q13 . 3 critical region ( Table S2 ) [23] . This particular example highlights the current challenge in clinical interpretation of CNVs which relies on the availability of large control datasets . We do not claim that our classification method replaces the need for such datasets . Our method does show that 27 out of 41 ( 66% ) rare inherited CNVs identified in patients contain genomic features similar to previously recognized MR-associated CNVs , a significant proportion when compared to the remainder of the genome ( Figure 3b ) . This provides independent support for the clinical relevance of this group of CNVs and shows that the interpretation of CNVs should not be limited to rare de novo CNVs with a fully penetrant dominant effect [48] . Furthermore , in the set of 53 rare CNVs with unknown inheritance , 46 CNVs were classified as being MR-associated , the vast majority with high confidence . These rare CNVs with unknown inheritance demonstrate strong similarities to rare de novo CNVs in that they have a low segmental duplication density , a high SINE density , often contain genes whose mouse knockouts result in nervous system phenotypes , have similar gene expression values and similar synonymous substitution rates . This suggests that these rare CNVs with unknown inheritance are indeed similar in pathoetiology to rare de novo CNVs and thus can be considered strong candidates for being causal CNVs . The ability of the classifier to identify such CNVs of unknown inheritance should be of great benefit to the diagnostic communities . This CNV classifier may also be informative of disorders other than mental retardation . This is of particular relevance because CNVs have recently been associated with other neurodevelopmental disorders such as autism and schizophrenia [1] , [4] , [5] but screening for causal CNVs in these diseases has yet to be implemented in most clinics . Interestingly , many of the CNVs associated with autism and schizophrenia , as well as mental retardation , contain genes whose proteins are involved in neurotransmission or in synapse formation and maintenance . This supports the existence of shared biological pathways that are disrupted in each of these neurodevelopmental disorders [49] . Our CNV classifier trained on MR CNVs may therefore already have predictive power for CNVs in other neurological disorders . It is likely , however , that this predictive power can be further optimized by retraining the classifier using disease-specific CNVs . In addition , the KEGG and MGI features selected for the MR patient cohort are also easily configurable for pathways and phenotypes which are more relevant to these other disease cohorts . For this reason we have made the Java source code of the CNV classifier , called GECCO , freely available ( see Materials and Methods ) . In conclusion , we have developed a novel objective method to identify disease-associated CNVs which has overcome several limitations with current CNV interpretation methodology . Our NBTree classifier is able to distinguish between MR-associated CNVs and benign CNVs with high accuracy without the use of data from large control cohorts or parental samples . Results indicate that computational classification methods can be used for objectively prioritizing CNVs in clinical research and diagnostics . The tool for classifying CNVs , called GECCO ( Genomic Classification of CNVs Objectively ) , as well as the Java source code , are readily available online . The benefits of such methods will increase with advancements in microarray technology , which already identifies many thousands of such structural variants per individual [50]–[53] , and in whole genome resequencing technology , . Establishing objective criteria and methods for interpretation of these genomic variants will be crucial for implementation of these technologies in a clinical setting . In this study we investigate if rare de novo CNVs and commonly inherited CNVs could be successfully classified without the use of inheritance information . In order to achieve this we collected from the literature a large number of rare CNVs known to be de novo ( n = 164 ) and a number of common CNVs known to be benign ( n = 1 , 413 ) . These CNVs were used for training and testing the classifier . A total of 20 genomic features were initially investigated . Initially 16 features were selected as attributes during the development of the classifier , which was then further optimized to a set of 13 features ( Table 1 ) . To test the accuracy of the classifier we first tested the classifier on a set of CNVs previously identified as being associated with MR ( Decipher known syndromes ) , and then created an independent validation set containing rare de novo and common inherited CNVs , collected from routine diagnostics , to be used in a validation study ( MR diagnostic CNVs ) . Finally two application sets were created containing CNVs without a clinical interpretation that were either a ) candidate CNVs , due to unavailability of parental samples , or b ) rare privately inherited CNVs . The CNVs used during the training and test phase ( 164 rare de novo CNVs termed “MR-associated CNVs” and 1 , 413 common inherited CNVs termed “benign CNVs” ) were identified on a number of different microarray platforms in previously published studies [15] , [19] , [28] , [30] . All aberrations were mapped using HG17 coordinates and converted when necessary using UCSC liftOver [54] . The Decipher known syndromes' ( https://decipher . sanger . ac . uk/ ) dataset contained 32 pathogenic CNVs based on microarray studies and associated with MR . The remaining 26 syndromes were excluded as they do not have either mental retardation as a prominent phenotype or a fully penetrant phenotype . MR Diagnostics and application datasets were collected through in-house routine diagnostics using Affymetrix 250k SNP microarrays ( Affymetrix , Santa Clara , USA ) , and consisted of 584 samples containing 1 , 297 CNVs . Regions were excluded that contained fewer than 5 microarray targets , that were smaller than 10kb in size or that were the result of a mosaic or complex chromosomal aberration . In total , the validation/application set contained 49 rare de novo CNVs , 41 rare inherited CNVs , 53 candidate CNVs and 1 , 154 common inherited CNVs . Initially a training set was created by randomly selecting 82 of the 164 rare de novo CNVs with an equal number of commonly inherited CNVs . The remaining CNVs were placed in the test set . The NBTree classification algorithm as implemented in Weka 3 . 6 . 0 [55] was selected and incorporated into our Java based tool called GECCO ( Genomic Classification of CNVs Objectively ) . An executable version and all source code for GECCO are readily available via http://genomegecco . sourceforge . net . NBTree is a hybrid method combining a decision tree with Naïve-Bayesian classifiers . The Naïve-Bayesian classifiers calculate the posterior probability ( a distance function ) that the CNV belongs to either class ( MR-associated CNV or benign CNV ) . The definition of the training set was then investigated . Given the imbalance that exists in the data ( see Results ) we sought to incorporate this prior into the training set . We tested increasingly imbalanced versions of the training set , starting with the most unbalanced training set , by placing half of all available CNVs in the training set ( 164 de novo and 2826 common inherited ) , and gradually decreasing the imbalance until the training set contained only 5% ( n = 143 ) of all available common inherited CNVs . The training set imbalance was then further tested in 1% decrements until the minimum was reached of 82 rare de novo CNVs and 28 common inherited CNVs . Once an optimal balance of CNV classes in the training set was identified the optimal subset of the CNVs in the training set was determined . This was achieved by randomly selecting CNVs as training and test instances over 10 , 000 iterations and then identifying the set that produced the maximum accuracy . In addition , enrichment analysis of the rare inherited CNVs was performed by generating 1 , 000 sets of random genomic regions matched for size against the rare inherited CNVs and the proportion of sets with greater than or equal to 27 CNVs classified as being MR was calculated . In total 20 different genomic features were investigated as potential classifier attributes . The variance inflation factor ( VIF ) was used to measure the co-linearity within the model across the repeat , gene and evolution measures ( simple repeats , repeat masker , LINE , SINE , long terminal repeats , RNA gene elements , segmental duplications , ENSEMBL genes , mean non-synonymous substitution rate ( dN ) , synonymous substitution rate ( dS ) and the dN/dS ratio of genes ) . Based on the VIF , features were removed until the model contained only independent features resulting in 16 different structural and functional genomic features that were used subsequently for training the classifier ( Table 1 ) . The included structural features were CNV type ( loss∶gain ) , CNV length , the numbers of LINE , SINE and segmental duplication elements lying within the CNV , as well as the densities of the LINE , SINE and segmental duplication elements . The density values were determined as the number of elements per base pair . Segmental duplications were downloaded from the UCSC table genomicSuperDups . The numbers of LINE and SINE elements were extracted from the UCSC table from rmsk and the RNA gene elements from sno/miRNA . The functional genomic features consisted of the gene count , gene density and the variance in gene expression levels , the mean non-synonymous substitution rate ( dN ) , synonymous substitution rate ( dS ) and the dN/dS ratio . In addition KEGG pathway and MGI knockout phenotypes were added as features . Genes involved in the KEGG ( Kyoto Encyclopedia of Genes and Genomes ) neurodegenerative pathway ( hsa01510 ) [56] were added as a categorical feature . This pathway includes KEGG genes belonging to KEGG Pathways section 5 . 2 , namely Alzheimer's disease ( KEGG pathway 05010 ) , Parkinson's disease ( KEGG pathway 05020 ) , Amyotrophic Lateral Sclerosis ( KEGG pathway 05030 ) , Huntington's disease ( KEGG pathway 05040 ) , Dentatorubropallidoluysian atrophy ( KEGG pathway 05050 ) and Prion Diseases ( KEGG pathway 05060 ) . KEGG genes were mapped to NCBI Entrez genes using associations provided by KEGG . Genes which were annotated as having the MGI mouse knockout phenotype , MP:0003631: nervous system phenotype were also added as a categorical feature . These genes were identified via human NCBI genes whose mouse orthologue's disruption had been assayed and were obtained from the Mouse Genome Informatics ( MGI ) resource ( http://www . informatics . jax . org , version 3 . 54 ) [45] . Substitution rates were obtained from EPGD [57] . The stable expression was calculated via the standard deviation of log2 intensities across 176 Hapmap cell lines ( CEU and YRI ) hybridized onto an Affymetrix GeneChip Human Exon 1 . 0 ST array ( GSE7761 ) .	Rare copy number variants ( CNVs ) are a frequent cause of neurological disorders such as mental retardation ( MR ) . However CNVs are also commonly identified in healthy individuals . It is therefore crucial for both diagnostic and research applications to be able to distinguish between disease-causing CNVs and “benign” CNVs occurring as normal genomic variation . Separating these two types can take advantage of significant differences in their genomic contents . For example , benign CNVs are enriched in repetitive sequences . By contrast , CNVs associated with MR tend to have high densities of functional elements , including genes whose mouse orthologues , when knocked-out , lead to specific nervous system abnormalities . We have developed a novel objective approach that is effective in distinguishing MR-associated CNVs from benign CNVs based on the presence of 13 genomic attributes . This method is able to achieve high accuracies in a cohort of CNVs known to cause MR and in a cohort of individuals with unexplained MR . The development of this technique promises to substantially improve the methodology for determining the pathogenicity of CNVs .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "genetics", "and", "genomics/genetics", "of", "disease", "genetics", "and", "genomics/bioinformatics", "genetics", "and", "genomics/medical", "genetics" ]	2010	Accurate Distinction of Pathogenic from Benign CNVs in Mental Retardation
Spinal cord injury often results in permanent functional impairment . Neural stem cells present in the adult spinal cord can be expanded in vitro and improve recovery when transplanted to the injured spinal cord , demonstrating the presence of cells that can promote regeneration but that normally fail to do so efficiently . Using genetic fate mapping , we show that close to all in vitro neural stem cell potential in the adult spinal cord resides within the population of ependymal cells lining the central canal . These cells are recruited by spinal cord injury and produce not only scar-forming glial cells , but also , to a lesser degree , oligodendrocytes . Modulating the fate of ependymal progeny after spinal cord injury may offer an alternative to cell transplantation for cell replacement therapies in spinal cord injury . Transplantation of different types of stem cells improves functional recovery after spinal cord injury in rodents and primates . The beneficial effects appear to be mediated by several mechanisms , including replacement of lost cells , secretion of neurotrophic factors , and probably most importantly , the generation of oligodendrocytes that remyelinate spared axons in the vicinity of a lesion [1 , 2] . Neural stem cells present in the adult spinal cord can be propagated in vitro [3 , 4] , and promote functional recovery when transplanted to the injured spinal cord [5] . Endogenous neural stem cells could therefore be attractive candidates to manipulate for the production of desired progeny after spinal cord injury as an alternative to stem cell transplantation . This approach would offer a noninvasive strategy that avoids the need for immune suppression , but has been held back by difficulties in identifying adult spinal cord neural stem cells and developing rational ways to modulate their response to injury . Studies using indirect techniques have suggested that the neural stem cell potential in the adult rodent spinal cord resides in the white matter parenchyma [6 , 7] or close to the central canal , either in the ependymal layer [8] or subependymally [9] . We have employed genetic fate mapping to characterize a candidate neural stem cell population in the adult spinal cord and show that close to all in vitro neural stem cell potential resides within the population of ependymal cells . Ependymal cells give rise to a substantial proportion of scar-forming astrocytes as well as to some myelinating oligodendrocytes after spinal cord injury . Modulating the fate of ependymal cell progeny after injury could potentially promote the generation of cell types that may facilitate recovery after spinal cord injury . In order to fate map candidate neural stem cells close to the central canal , we generated two transgenic mouse lines expressing tamoxifen-dependent Cre recombinase ( CreER ) under the control of FoxJ1 ( HFH4 ) or Nestin regulatory sequences . FoxJ1 expression is specific to cells possessing motile cilia or flagella [10–13] . In the adult forebrain , a subset of astrocytes in the subventricular zone contact the ventricle and have an immotile primary cilium [14] , but FoxJ1 expression is restricted to cells with motile cilia [10–13] . Nestin is expressed in central nervous system stem and progenitor cells during development and in adulthood [15–19] . In the adult spinal cord , nestin is expressed by cells lining the central canal , endothelial cells , and sparse white matter glial cells [20] . The second intron enhancer in the Nestin gene allows for selective expression of CreER in the neural lineage [21] , eliminating expression in for example endothelial cells . CreER expression in the adult spinal cord is limited to cells lining the central canal in both the FoxJ1-CreER and Nestin-CreER mouse lines ( Figure 1 ) . Administration of tamoxifen to mice on an R26R [22] or Z/EG [23] Cre reporter background allows inducible , permanent . and heritable genetic labeling by the expression of β-galactosidase ( β-gal; R26R ) or GFP ( Z/EG ) in cells expressing CreER ( the strategy is schematically depicted in Figure S1 ) . Recombination in the absence of tamoxifen was exceptionally rare ( <1 cell/30 coronal 20-μm-thick sections in both transgenic lines ) and limited to CreER-expressing cells in the ependymal layer . Administration of tamoxifen ( five daily injections ) resulted in recombination of the reporter allele ( Figure 1A–1D ) in 82 ± 4% of transgene-expressing cells in Nestin-CreER mice and 88 ± 4% in FoxJ1-CreER mice ( mean ± standard deviation [SD] , n = 6 mice for each mouse line ) . The cells at the central canal expressing CreER protein from the Nestin-CreER or FoxJ1-CreER transgene are immunoreactive to Crocc , a marker for ciliated cells ( Figure S2 ) . They contain the intermediate filaments nestin and vimentin , associated with immature neural cells [15] , but notably not glial fibrillary acidic protein ( GFAP ) ( Figures S2 and S3 ) , which is present in some neural stem cells in the adult forebrain [24] . The transgene expressing cells display other markers associated with neural stem/progenitor cells such as CD133/prominin-1 , Musashi1 , PDGFR-α , Sox2 , Sox3 , and Sox9 but are negative for the oligodendroglial progenitor marker Olig2 ( Figures S2 and S3 ) . All above-mentioned proteins appear uniformly expressed by the cells lining the central canal , and we have not found any molecular marker delineating any subpopulations . Immunoelectron microscopy established that the Nestin-CreER and FoxJ1-CreER transgenes are expressed in identical cell populations by the central canal; their expression is restricted to lumen-contacting cells with motile cilia ( 9 + 2 axonemes ) , and all such cells express both transgenes ( Figures 2A–2C , S4 , and S5 ) . Ultrastructural analysis in serial sections revealed morphological heterogeneity among the lumen-contacting ciliated cells , with some cells displaying typical cuboidal ependymal cell morphology and others a tanycyte morphology [25] ( Figures 2C , 2D , S4 , and S5 ) . In addition , there is a less numerous third cell type , which we refer to as a radial ependymal cell . Radial ependymal cells share the morphology of the cytoplasm , and often nucleus , with ependymal cells , but have a long basal process ( Figures 2B , 2D , and S5 ) . The radial ependymal cells almost invariably reside in the dorsal or ventral pole of the ependymal layer , with a basal process oriented along the dorsoventral axis ( Figures 2A and S5 ) . Although the lumen-contacting ciliated cells can be subdivided into these three groups , cells with intermediary phenotypes are frequent ( Figure 2E and Table S1 ) , which together with their homogeneous molecular profile suggests that they are closely related . The naming of ependymal cell types is based solely on morphological criteria and does not imply any function . The central canal-contacting ciliated cells have in common that they reside in the ependymal layer , thus we collectively refer to them as ependymal cells . Adult spinal cord neural stem cells can be propagated in vitro [3] , but their precise identity has been difficult to establish unequivocally [6–9] . We utilized our genetic labeling paradigms to ask whether adult spinal cord ependymal cells have neural stem cell properties in vitro . Adult FoxJ1-CreER and Nestin-CreER mice on Cre recombination reporter background ( R26R or Z/EG ) received five daily injections of tamoxifen to induce recombination , and primary cultures were initiated after an additional 6 d without tamoxifen ( Figure 3A ) . Tamoxifen and its active metabolite 4-hydroxytamoxifen have a half-life of 6–12 h in the mouse [26] , and accordingly CreER protein was no longer detectable in the nucleus after 6 d without tamoxifen ( Figure 1B and 1D ) . Spinal cords were dissociated and plated at clonal density in standard conditions that allow for neurosphere formation ( Figure 3B ) . We found that 76 ± 5 . 7% of neurospheres from Nestin-CreER and 85 ± 2 . 2% from FoxJ1-CreER mice were recombined and thus derived from recombined ependymal cells ( mean ± SD , n = 6 mice analyzed separately per line , Figure 3C and 3E ) . Neurospheres were either homogeneously recombined or not recombined , verifying their clonal origin . Since recombination never was fully penetrant in the ependymal cells , the analysis of the proportion of neurospheres that were recombined may underestimate the true contribution of this cell population to neurosphere formation . If there is a stochastic distribution of recombination within the CreER expressing cell population , rather than recombination demarcating a subpopulation that differs with regard to neurosphere-forming potential , one can estimate the contribution of the cell population to neurosphere formation by normalizing it to the observed recombination rate . To estimate the theoretically maximal proportion of neurosphere-initiating cells that are ependymal cells , we analyzed the recombination frequency in CreER-expressing cells in sections from each spinal cord sample that was used for neurosphere cultures ( Figure 3B ) . Normalizing the recombination frequency in neurospheres to the recombination frequency in the CreER-expressing cells in vivo , suggested that close to all neurosphere-initiating potential could reside within the ependymal cell population under these conditions ( Figure 3D ) . Progenitor cells with limited self-renewal capacity can give rise to neurospheres , but are incapable of generating new neurospheres when passaged more than twice [27 , 28] . We found that 100% of the recombined neurospheres from both Nestin-CreER and FoxJ1-CreER could be serially passaged at least eight times to give rise to new neurospheres ( n = 6 neurospheres per 4 transgenic mice ) . The number of cells increased exponentially during passaging ( Figure S6 ) . Analysis of the differentiation potential of ependymal cell-derived neurospheres after three passages revealed that 100% of the neurosphere clones were multipotent and differentiated into neurons , astrocytes , and oligodendrocytes ( Figure 3F ) . We also isolated prospectively identified ependymal cells by flow cytometry independently of Cre-mediated recombination by utilizing the green fluorescent protein ( GFP ) expression under the bicistronic control of the FoxJ1 promoter ( Figures 3G and S1 ) . Flow cytometric isolation of adult spinal cord cells substantially reduced neurosphere formation , and 0 . 18 ± 0 . 06% ( mean ± SD from in average 1 , 600 GFP-positive ( GFP+ ) cells/mouse , n = 6 mice analyzed separately ) of GFP+ ependymal cells formed neurospheres ( Figure 3H–3K ) . In contrast , not a single neurosphere developed from the same number of cells in the GFP− non-ependymal fraction from any animal in the same experiments . Thus , the neural stem cell potential in the adult spinal cord , at least under the conditions employed here , largely resides within the ependymal cell population . Cells in the adult spinal cord ependymal layer proliferate , albeit slowly or rarely [8] . Continuous administration for one month in the drinking water of 5-bromo-2-deoxyuridine ( BrdU ) , which is incorporated into DNA in cells in S-phase , resulted in labeling of 19 . 9 ± 4 . 2% of ependymal cells ( mean ± SD from three mice , Figure 4A and 4B ) . The BrdU-labeled ependymal cells constituted 4 . 8 ± 0 . 9% of all BrdU-labeled cells in a spinal cord segment ( mean ± SD from three mice ) . The majority of BrdU-labeled ependymal cells were found in pairs , indicating that most mitoses resulted in self-duplication rather than the generation of another cell that had left the ependymal layer ( Figure 4A and 4B ) . In line with this , analysis of the distribution of recombined cells up to 8 mo after tamoxifen administration in the FoxJ1-CreER and Nestin-CreER mice did not provide evidence for the generation of cells that leave the ependymal layer under normal conditions ( unpublished data ) . Whether a specific cell population is derived from another cell type or it is maintained through self-duplication can be established by analyzing the genetic labeling frequency at different time points after induction of recombination [29 , 30] . There was no reduction in the proportion of recombined ependymal cells for up to 10 mo after tamoxifen administration ( Figure 4C–4E ) , indicating that ependymal cells are maintained by self-renewal and are not replenished by another cell population . We next assessed the response of ependymal cells to spinal cord injury . We used the same labeling paradigm as before ( Figure 3A ) , with a 6-d period between the last tamoxifen dose and the injury . This ensures that all recombination occurs prior to the insult and that even if other cells than ependymal cells would start to express the FoxJ1-CreER or Nestin-CreER transgene in response to the injury ( nestin is indeed expressed by reactive astrocytes [20] ) , it would not result in recombination . An incision in the dorsal funiculus , which does not compromise the integrity of the ependymal layer , dramatically increased the proliferation of ependymal cells ( Figures 5A , 5B , and S7 ) . In contrast to the uninjured spinal cord , where proliferation of ependymal cells appears largely limited to self-renewing divisions , recombined cells started to migrate and were located outside the ependymal layer 4 d after the injury ( Figure 5C–5I ) . Migrating recombined cells lost their ependymal phenotype as judged by the loss of immunoreactivity to Sox2 and Sox3 and lack of CreER expression from the FoxJ1 promoter ( Figure 3D and unpublished data ) . Most emigrating cells expressed Sox9 and some the astrocyte marker GFAP ( Figure 5F , 5H , and 5I ) . Ultrastructural analysis revealed that ependymal cell morphology was largely unaltered by the injury , with the exception of a darker cytoplasm due to a higher content of filaments ( Figure 5J and 5K ) . Ependymal progeny migrated towards the injury site in the dorsal funiculus and an increasing number of recombined cells accumulated in the forming glial scar over several weeks and remained there for at least 10 mo after the insult ( Figure 6A–6C ) . The recombined ependyma-derived cells occupied 18 . 3 ± 6 . 9% ( mean ± SD from three FoxJ1-CreER mice ) of the area in the scar tissue 2 wk after the injury , which is likely to be a slight underestimate of the true contribution since recombination never was fully penetrant . The ependyma-derived cells were not evenly distributed throughout the injury site , but the scar consisted of patches of recombined and unrecombined cells ( Figure 6H and 6I ) . The reaction of the ependymal cells was restricted to the injured segment and was absent in adjacent segments ( Figures 5A–5H and 6A–6C ) , which are indirectly affected by the severance of axons and Wallerian degeneration . There were no recombined cells outside the ependymal layer in animals in which only the spinal cord was exposed but no lesion was made ( sham lesion , Figure S8 ) , and a lesion did not induce recombination in animals that had not received tamoxifen ( unpublished data ) . Since some ependymal cells extend processes along the dorsolateral midline , it was possible that the activation of ependymal cells by a dorsal funiculus incision was triggered by the severance of such processes . To investigate this , we performed incisions in the lateral spinal cord , which do not directly injure the ependymal cell processes in the dorsolateral midline . In these animals , ependymal cell progeny were generated and migrated laterally towards the injury ( Figure S8 ) . The ependyma-derived cells migrating to the lesion appeared less numerous after a lateral than after a dorsal incision , suggesting that severance of ependymal cell processes in the midline is not necessary for the activation of ependymal cells , but that it may augment their reaction . The migration of ependyma-derived cells to the site of injury suggests the presence of attractive signals originating in the lesion area . SDF1 , through its receptor CXCR4 , mediates attraction of progeny from neural stem/progenitor cells after some types of injuries [31 , 32] . The majority of ependymal cells as well as their progeny were , however , negative for CXCR4 ( Figure S9 ) , making it unlikely that this receptor mediates the attraction of ependymal cell progeny to a spinal cord lesion . Analysis of the fate of the ependymal cell progeny by molecular markers and electron microscopy after a dorsal funiculus incision revealed that the majority were immunoreactive to Sox9 and vimentin and had an astrocyte-like morphology ( Figures 6E , 6H , 6J , 6M , and S10 ) . A smaller subpopulation of the recombined cells expressed GFAP and nestin , but the vast majority of cells with this phenotype were not recombined ( Figure 6D , 6H , 6I , and 6M ) . Recombined GFAP- and nestin-expressing cells were typically located close to the surface of the spinal cord ( Figure 6D , 6H , 6I , and 6L ) , whereas the Sox9- and vimentin-expressing cells were most abundant in the core of the scar tissue ( Figures 6L and S10 ) . We conclude that the glial scar is comprised of two different populations of astrocyte-like cells , where the majority of the Sox9+/vimentin+ population derives from ependymal cells and the GFAP+/nestin+ cells are mainly reactive resident astrocytes . We further investigated the contribution of ependymal cells to other lineages . None of the recombined cells in the scar tissue had neuronal morphology or were immunoreactive to the neuron-specific epitope NeuN ( unpublished data ) . A population of recombined cells expressed Olig2 ( Figure 6F and 6G ) . The first month after injury , Olig2-expressing recombined cells were scattered throughout the injury site and had an ultrastructural morphology corresponding to immature oligodendrocytes ( Figure 6F , 6K , 6L , and 6M ) . At later time points , Olig2-expressing ependyma-derived cells were excluded from the scar tissue and were restricted to the uninjured tissue that bordered the scar ( Figure 6G , 6L , and 6M ) . Lesions in the lateral funiculus resulted in the generation of ependymal progeny of the same fates as after a dorsal funiculus incision ( Figure S8 ) . The scar tissue that forms at spinal cord injuries is thought to inhibit axonal growth [33 , 34] . Chondroitin sulphate proteoglycans ( CSPG ) appear to be the principal axonal growth inhibiting molecules in glial scars [35] . Ependyma-derived cells at the injury formed a complementary nonoverlapping pattern with areas that were CSPG immunoreactive ( Figure 7A and 7B ) , indicating that ependymal cell progeny do not contribute to the production of axonal growth-inhibiting CSPG . In parallel with the production of axonal growth-inhibiting factors in the glial scar , there is an increase in some axonal growth-promoting molecules , such as the extracellular matrix molecules laminin and fibronectin [36 , 37] . In the injury model employed here , axons send sprouts into the scar tissue , mainly during the first month after an injury , and the axons are preferentially associated with areas in the scar tissue that have high levels of laminin [38 , 39] . Both laminin and fibronectin immunoreactivity were widely distributed throughout the scar tissue , overlapping both with CSPG-immunoreactive domains and areas occupied by ependyma-derived cells ( Figure 7A and 7B ) . Neurofilament-immunoreactive axons were present in the center of the scar tissue and were often wiggly and oriented in all directions ( Figure 7C–7H ) . This is in contrast to the rostrocaudal orientation of axons seen in the uninjured dorsal funiculus , suggesting that many of the axons present in the scar were severed and sprouting into the scar tissue [39] . Neurofilament-immunoreactive axons were present in domains dominated by ependyma-derived cells , as well as in other areas of the scar where these cells were less abundant ( Figure 7C–7H ) . Axons were often present in direct proximity to ependyma-derived cells ( Figure 7D , 7E , 7G , and 7H ) . The finding that ependyma-derived progeny is not associated with the main scar-associated axonal growth-inhibiting factor , CSPG , together with their proximity to axonal sprouts , argues against these cells being a major factor in glial scar-associated axonal growth inhibition . The finding that some ependymal cell progeny displayed a marker profile and ultrastructural morphology suggesting oligodendroglial differentiation ( Figure 6 ) prompted us to characterize these cells further and to address whether they may contribute to axonal remyelination at later time points . Ten months after spinal cord injury , the majority of ependyma-derived progeny are located in the scar tissue that has formed at the injury site , but a substantial number of cells are sparsely distributed in a large area of the intact grey and white matter bordering the lesion ( Figure 8A–8C ) . Most of these cells are Olig2+ and display mature oligodendrocyte morphology with processes that extend along and enwrap myelin basic protein ( MBP ) -immunoreactive myelin ensheathing axons ( Figure 8B–8D ) . Nuclear regions and processes of two ependyma-derived cells were followed in the electron microscope in serial utrathin sections ( Figures 9 and S11 , and unpublished data ) . They both displayed a typical mature oligodendrocyte morphology [25] , such as oval nuclei with clumped chromatin , a cytoplasmic matrix that appeared denser than in surrounding astrocytes , a granular endoplasmic reticulum represented by several short cysternae , and tight junctions with adjacent oligodendrocyte processes ( Figure 9 ) . Few processes emerged from the cell body , and unlike those of astrocytes , they did not form many branches and did not contain evident fibrils ( Figure 9 ) . The processes of the recombined cells could be traced along axons , surrounding their myelin sheaths ( Figure 9D ) . Thus , in addition to the generation of astrocytes , ependymal cells generate myelinating oligodendrocytes . Stem cells are notoriously difficult to identify , and their localization in the adult spinal cord has been controversial [6–9] . We report that ependymal cells constitute the vast majority of cells displaying in vitro neural stem cell properties in the adult spinal cord . Ependymal cells self-renew in vivo , but do not generate appreciable numbers of other cell types under homeostatic conditions . Their normally limited proliferation increases dramatically after spinal cord injury and they then produce oligodendrocytes , and more abundantly , astrocytes that migrate to the site of injury and make a substantial part of the glial scar . The immediate descendants of tissue stem cells , progenitor cells with limited self-renewal capacity and/or lineage potential , can in some situations acquire stem cell properties [40] . For example , spermatogonial progenitor cells can regain stem cell function after injury and during aging and forebrain neurospheres may be derived from committed progenitors [41 , 42] . It appears unlikely that this would explain the neural stem cell properties displayed by ependymal cells in vitro , as they are not replenished by any other cell type in the adult , but are self-renewing . However , although ependymal cells at the population level display cardinal stem cell features in vivo , such as self-renewal and generation of diverse progeny , it is difficult to study these properties at the single cell level in the tissue , and we cannot conclude that they act as stem cells in vivo . In addition to ependymal cells , neural progenitors ( expressing NG2 , Olig2 , and/or Nkx2 . 2 ) reside in the white and gray matter of the adult rodent spinal cord [6 , 43–46] . Different studies have suggested that the parenchymal progenitors represent multipotent neural stem cells or more-restricted glial progenitors [6 , 43 , 47] . Under the standard neurosphere assay conditions employed here , the vast majority of the neural stem cell potential resides within the ependymal population . However , we cannot exclude that other cells contribute , to a comparatively smaller degree , to neurosphere formation under our conditions or that they may display neural stem cell properties under other conditions . The parenchymal progenitors are likely to serve to replace glial cells in the uninjured spinal cord , which we do not find evidence that ependymal cells do . Parenchymal progenitors are rapidly depleted after spinal cord injury , but are later replaced and may participate in the generation of glial cells after injury [6 , 44 , 46] . It is possible that some of the ependyma-derived Olig2+ cells observed shortly after injury represent regenerated parenchymal progenitors . The limited functional recovery typically associated with central nervous system injuries is in part due to the failure of severed axons to regrow and reinnervate their targets . Axonal regeneration is inhibited by scar formation and growth-inhibitory factors associated with myelin and astrocytes [48 , 49] . Modulating the responsiveness to axonal growth-inhibitory factors and glial scar formation are attractive strategies to improve functional recovery after central nervous system injuries [50–53] . The majority of ependyma-derived cells differentiate to astrocyte-like cells after injury and are found in the core of the scar tissue . However , these cells are found in complementary nonoverlapping domains to areas immunoreactive to CSPG , the most important axonal growth inhibitor associated with glial scars [35 , 36] . Moreover , axons in the scar tissue , most likely sprouts from severed axons growing into the scar tissue , were frequently found in direct proximity to ependyma-derived cells . This argues that ependyma-derived cells in the scar tissue do not constitute a major impediment to axonal growth , and may even suggest that they support some local sprouting . Spinal cord injury results in the loss of oligodendrocytes and demyelination of axons even at some distance to the lesion [54 , 55] . Spinal cord injuries are most commonly incomplete in man , leaving spared tissue connecting the spinal cord above and below the lesion , but the function of remaining axons is often compromised due to demyelination . Without insulating sheaths of myelin , spared axons close to , but not directly affected by the injury , become less efficient in their ability to conduct electrical impulses [56] . Moreover , chronically demyelinated axons are vulnerable to degeneration . Axons are remyelinated with time , and this is thought to occur through the generation of new oligodendrocytes by stem or progenitor cells rather than by self-duplication of mature remaining oligodendrocytes [57–59] . We report here that ependymal cells contribute to the regeneration of oligodendrocytes and remyelination after spinal cord injury . The differentiation pattern of ependymal cells after injury is reminiscent to that seen for in vitro expanded adult spinal cord neural stem cells transplanted to the injured spinal cord [5] . Transplanted adult spinal cord-derived neurospheres improve functional recovery , and if they are forced to generate more oligodendrocytes , functional recovery is further improved [5] . Since ependymal cells are the main source of neurospheres from the adult spinal cord ( Figure 3 ) , promoting oligodendrocyte generation from these cells in vivo could potentially improve recovery after spinal cord injury . The development of pharmacological strategies to modulate endogenous stem cells and their progeny may be an attractive alternative to cell transplantation for the treatment of spinal cord injury . For Nestin-CreER , we used the enhancer found in the second intron of the rat nestin gene fused to a minimal hsp68 promoter [18 , 60 , 61] that controls the expression of CreERT2 [62] , as previously described [21] . For FoxJ1-CreER , we used a human FOXJ1 promoter [13] fused to a CreERT2 ires-EGFP construct . Transgenic mice were generated at the Karolinska Center for Transgene Technologies by standard procedures utilizing pronuclear injection of CBA × C57BL/six fertilized eggs . Potential founder animals were screened by Southern blot analysis and PCR analysis using a CreERT2-specific fragment as probe or PCR template . Founder mice were bred to wild-type C57Bl/6 mice . Expression of the transgene was analyzed by confocal microscopy of sections stained with anti-Cre antibodies and cell-specific markers . Recombination was induced by five daily intraperitoneal injections of 2 mg of tamoxifen ( Sigma; 20 mg/ml in corn oil ) . Adult mice were perfused transcardially with PBS followed by 4% formaldehyde in PBS , spinal cords were post-fixed overnight at 4 °C and then cryoprotected in 30% sucrose . Coronal ( 14 or 20 μm ) or sagittal ( 20 μm , from ∼9-mm-long pieces ) sections were collected alternating on ten slides ( 8–10 sections per slide ) . Sections were incubated with blocking solution ( 10% donkey serum in PBS , with 0 . 3% Triton-X100 ) for 1 h at room temperature , then incubated at 4 °C or room temperature in a humidified chamber for 12–48 h with primary antibodies diluted in blocking solution . For MBP staining , sections were first delipidized . For antibodies raised in mouse , the M . O . M . kit ( Vector ) , ABC kit ( Vector ) , and TSA system ( Perkin Elmer ) were used following the manufacturers' instructions . The following primary antibodies were used: β-galactosidase ( 1:5 , 000 , rabbit; ICN Biomedicals , or 1:1 , 000 , goat; Biogenesis ) , BrdU , ( 1:200 , rat; Accurate ) , CD133 ( 1:500 , rat , clone 13A4; eBioscience ) , chondroitin sulfate ( 1:1 , 000 , mouse; Sigma ) , Cre ( 1:2 , 000 , mouse; Nordic BioSite ) , Crocc ( 1:5 , 000 , rabbit , Root6; gift from T . Li ) , CXCR4 ( 1:500 , mouse; BD Pharmingen ) , fibronectin ( 1:1 , 000 , rabbit; Sigma ) , GFAP ( 1:1 , 000 , mouse , clone G-A-5; Sigma ) , Ki67 ( 1:1 , 000 , rabbit; Neomarkers ) , Olig2 ( 1:500 , goat; R&D Systems ) , laminin ( 1:1 , 000 , rabbit; Sigma ) , MBP ( 1:500 , rabbit; Chemicon ) , musashi-1 ( 1:2 , 000 , rat , clone 14H1; gift from H . Okano ) , nestin ( 1:5 , 000 , rabbit [63] or 1:500 , mouse; BD Pharmingen ) , neurofilament heavy ( 1:1 , 000 , chicken; Chemicon ) , PDGFRα ( 1:500; BD Pharmingen ) , RC1 ( 1:200 , mouse; DSHB ) , Sox2 ( 1:500 , mouse; Chemicon , or 1:1 , 000 , goat; gift from J . Muhr ) , Sox3 ( 1:500 , rabbit; gift from T . Edlund ) , Sox9 ( 1:500 , goat; R&D Systems ) , vimentin ( 1:1 , 000 , chicken; Chemicon ) . After washing , antibody staining was revealed using species-specific fluorophore-conjugated ( Cy3 , Cy5 from Jackson , and Alexa 488 from Molecular Probes ) or biotin-conjugated secondary antibodies ( Jackson ) . Biotinylated secondary antibodies were revealed using the ABC kit ( Vector Labs ) with TSA fluorescent amplification kit ( Perkin-Elmer ) . Sections were counterstained with DAPI ( 1 μg/ml; Sigma ) . Control sections were stained with secondary antibody alone . Pictures were taken using a Zeiss Axioplan 2 , Zeiss Axiovert 200M or a LSM510 META confocal microscope with Zeiss and Openlab ( Improvision ) software . Image processing and assembly was performed in ImageJ and Photoshop . Anaesthetized mice were perfused transcardially with 4% paraformaldehyde in PBS . The spinal cord was dissected out and post-fixed for 4 h . Sections ( 60 μm ) were immunolabled with Cre or β-gal antibodies in 0 . 1% Triton X100 and 10% donkey serum in PBS . A secondary antibody conjugated to biotin was used with an ABC kit ( Vector Labs ) . In some sections , a fluorescent secondary antibody was used , and sections were labeled with DAPI to map the location of the nuclei of the surrounding cells ( Figure S11 ) . The sections were post-fixed in 3% glutaraldehyde in 0 . 1 M cacodylate buffer ( pH 7 . 4 ) and in 1% osmium tetraoxide in 0 . 1 M cacodylate buffer , dehydrated in a graded series of ethanol , stained with uranyl acetate , and embedded in Durcupan resin ( Fluka ) . Serial sections ( 200 nm ) of FoxJ1-CreER spinal cord were used to define the ultrastructural morphology of cells lining the central canal . The complete series of sections from individual cells were traced in 90 sections to characterize the morphology in three dimensions . Serial semi- and ultrathin sections ( 2 μm and 70 nm , respectively ) were used in correlative light and electron microscopic evaluation ( CLEM ) of the fate of ependyma-derived cells 4 wk after a spinal cord injury ( Figure S6 ) . The semithin sections were used to identify immunopositive cells , and the ultrathin sections to show the morphology of the identified cells . Sections ( 70–200 nm ) were placed on Formvar-coated copper grids , counterstained with 2% uranyl acetate and Reynold's lead citrate . Sections were examined in a Tecnai 12 electron microscope ( FEI ) equipped with 2kx2k TemCam-F224HD camera ( TVIPS ) . Spinal cords were dissected and cells dissociated using papain ( Worthington ) . Neurospheres were cultured as described [8] in DMEM/F12 medium supplemented with B27 and EGF and bFGF ( both 10 ng/ml ) . Approximately 200 , 000 cells were plated in 10-cm cultures dishes , corresponding to a density of 20 cells/microliter , which allows the generation of clonal neurospheres [64] . For assaying self-renewal and multipotentiality , FoxJ1-CreERxR26R ( n = 6 ) and Nestin-CreERxR26R ( n = 6 ) adult mice were administered with tamoxifen intraperitoneally for 5 d with a washout period of 6 d ( see Figure 3 ) . Single spheres were manually collected and split into two wells . One well was used for continuous passaging and subsequent neural stem cell differentiation , whereas the other well was used for X-gal staining . For assaying self-renewal , four clonal recombined neurospheres per animal were manually isolated after 12 d of primary neurosphere formation . All recombined neurospheres were serially passaged eight times . In vitro differentiation by growth factor withdrawal for 10 d was assessed in passage 3 and passage 6 by staining as described above for βIII-tubulin ( Tuj1 , 1:1 , 000; Covance ) , GFAP ( 1:5 , 000; DAKO ) , and O4 ( 1:200; Chemicon ) . Spinal cords were dissected from FoxJ1-CreER mice and dissociated using papain ( Worthington ) and DNase in 1× HBSS at 37 °C for 1 h . Ovomucoid inhibitor ( Worthington ) was added and cells were collected by centrifugation at 300g for 5 min . Cells were resuspended in Leibovitz-15/B27 with 7AAD , which labels dead cells . Single GFP+ ( based on the ires-GFP signal ) , 7AAD− cells were isolated using a FACSAria ( BD ) . Singlets were identified based on forward scatter width ( FSC-W ) versus forward scatter height ( FCS-H ) [65] . Single cell sorting and GFP fluorescence was confirmed by microscopic examination . Mice were anesthetized with 2 . 5% Avertin , and the dorsal funiculus at mid-thoracic level was cut transversely and was extended rostrally with microsurgical scissors to span one segment [39] . In other animals , the lateral funiculus was cut transversally and the lesion extended rostrally to span one segment . BrdU ( 1 mg/ml and 1% sucrose , exchanged every 3–4 d ) was administered in the drinking water to label dividing cells . In order to correlate recombination frequency in neurosphere cultures to the in vivo recombination frequency of spinal cord tissue , the ratio between CreER+ and β-gal+ cells ( n = 60 ) was quantified in a small postfixed biopsy from the same spinal cords used for neurosphere cultures . The percentage of BrdU+ ependymal cells was obtained from three animals treated for 4 wk with BrdU in the drinking water ( 3–5 coronal sections/animal analyzed ) . The total number of cells per section was obtained counting all nuclei stained with DAPI . The in vivo recombination frequency was assessed by counting the number of recombined cells over the total number of ependymal cells ( Vimentin+ ) from five coronal sections per animals at 2 d ( 4 animals: 2 Nestin-CreERxR26R , 2 FoxJ1-CreERxR26R ) and 10 mo ( 3 animals: 3 FoxJ1-CreERxR26R ) after tamoxifen treatment ( Figure 4E ) . The contribution of recombined cells at the site of injury was established by measuring the relative area occupied by β-gal+ cells within the epicenter of the lesion ( using ImageJ software ) of 3 FoxJ1-CreERxR26R animals ( 2 sagittal sections per animal ) 2 wk after spinal cord injury . The cell fate distribution of ependyma-derived progeny was obtained by scoring recombined cells positive for Olig2 , GFAP or Sox9 in either coronal ( 3–5 sections per animal ) or sagittal ( 1–2 sections per animal ) sections encompassing the lesion site from 1 month ( 3 animals: 1 Nestin-CreERxZ/EG , 2 FoxJ1-CreERxR26R ) and 8–10 mo ( 4 animals: 1 Nestin-CreERxZ/EG and 3 FoxJ1-CreERxR26R ) after spinal cord injury ( Figure 6M ) . The frequency of proliferation of ependymal cells and their progeny was assessed by counting the number of Ki67+ recombined cells over the total number of recombined cells from three segments ( rostral to , caudal to , and at the injury site; average of 15 coronal sections , or 300 recombined cells , per segment analyzed ) from 2 FoxJ1-CreERxR26R animals 4 d after spinal cord injury ( Figure S7C ) .	Spinal cord injuries occur in more than 30 . 000 individuals each year worldwide and result in significant morbidity , with patients requiring long physical and medical care . The recent identification of resident stem cells in the adult spinal cord has opened up for the possibility of pharmacological manipulation of these cells to produce cell types promoting recovery after injury . We have employed genetic tools to specifically address the identity and reaction to injury of a spinal cord subpopulation of cells known as ependymal cell . Genetic labeling of this putative stem cell population allows for the evaluation of stem cell activity in vitro and in vivo . We found that ependymal cells lining the central canal act as neural stem cells in vitro and contribute extensively to the glial scar in vivo . Interestingly , injury induces proliferation of ependymal cells and migration of ependyma-derived progeny towards the site of injury . Moreover , ependymal cell progeny differentiate and give rise to astrocytes as well as myelinating oligodendrocytes . In summary , our results point to ependymal cells as an attractive candidate population for non-invasive manipulation after injury .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "developmental", "biology", "cell", "biology", "neuroscience" ]	2008	Spinal Cord Injury Reveals Multilineage Differentiation of Ependymal Cells
The decreasing cost of sequencing is leading to a growing repertoire of personal genomes . However , we are lagging behind in understanding the functional consequences of the millions of variants obtained from sequencing . Global system-wide effects of variants in coding genes are particularly poorly understood . It is known that while variants in some genes can lead to diseases , complete disruption of other genes , called ‘loss-of-function tolerant’ , is possible with no obvious effect . Here , we build a systems-based classifier to quantitatively estimate the global perturbation caused by deleterious mutations in each gene . We first survey the degree to which gene centrality in various individual networks and a unified ‘Multinet’ correlates with the tolerance to loss-of-function mutations and evolutionary conservation . We find that functionally significant and highly conserved genes tend to be more central in physical protein-protein and regulatory networks . However , this is not the case for metabolic pathways , where the highly central genes have more duplicated copies and are more tolerant to loss-of-function mutations . Integration of three-dimensional protein structures reveals that the correlation with centrality in the protein-protein interaction network is also seen in terms of the number of interaction interfaces used . Finally , combining all the network and evolutionary properties allows us to build a classifier distinguishing functionally essential and loss-of-function tolerant genes with higher accuracy ( AUC = 0 . 91 ) than any individual property . Application of the classifier to the whole genome shows its strong potential for interpretation of variants involved in Mendelian diseases and in complex disorders probed by genome-wide association studies . Advances in next-generation sequencing technologies have considerably reduced the cost of genome sequencing . As a result , there has been an avalanche of personal genomic data with numerous individual genomes sequenced in the last few years [1]–[4] . Variants in protein-coding genes are of special interest due to their stronger likelihood of functional effects . A comprehensive understanding of the functional impact of variants in coding genes requires their integration with various levels of annotations , such as primary sequence of the gene , three-dimensional structures of its protein products and biological networks where genes interact with each other . Functional annotation of single nucleotide variants ( SNVs ) at genomic sequence level results in their classification as nonsynonymous ( which includes missense and nonsense ) , splice site disrupting or synonymous . Similarly , small insertions and deletions ( indels ) in coding genes can be classified as frame-shift or in-frame . Nonsense and splice site disrupting SNVs as well as frame-shift indels are mostly assumed to lead to loss-of-function ( LoF ) of genes [5] . On the other hand , missense SNVs and in-frame indels may or may not be damaging [6] . It is well understood that genes and their protein products rarely act in isolation but rather work closely with other genes and/or their products to form various networks and pathways which accomplish specific goals , for example , signal transduction , metabolism etc . Thus , a comprehensive understanding of the functional impact of variants necessitates the inclusion of these interactions between genes . Network-based approaches are thus often used to study human disease [7] . One feature that has emerged from past studies of disease genes and networks is that protein products of genes associated with similar disorders have a higher likelihood of physical interaction with each other [8] . It has also been noted in many studies that functionally essential genes are more likely to encode for hub ( i . e . highly connected ) proteins in the physical protein-protein interaction ( PPI ) network in both yeast [9] and humans [8] . Moreover , hub proteins are likely to be under stronger negative selection constraints in humans and positive selection tends to occur on network periphery [10] . Similar studies on signaling pathways have revealed that as one goes from extracellular space to the nucleus in the cell , negative selection constraints on genes encoding corresponding proteins tend to increase [11] . Selection studies have also been performed on metabolic pathways where enzyme connectivity signifies the number of other metabolic enzymes that produce the enzyme's reactants or consume its products . For example , in a yeast network of 584 metabolites and comprising about 16% of all yeast genes , Vitkup et al found that highly connected enzymes evolve slower than less connected enzymes [12] . Montanucci et al also reported that genes encoding highly connected enzymes in N-glycosylation metabolic pathway exhibit stronger purifying selection constraints and tend to evolve slowly in primates [13] . In order to obtain a higher resolution understanding of the relationship between selection constraints and networks , some studies have also integrated three-dimensional protein structures with PPI network to obtain structural interaction network ( SIN ) . Kim et al showed that in yeast , hubs in PPI with more than two interaction interfaces are more likely to be essential than those with two or less interfaces [14] . Using structurally resolved human PPI network , Wang et al showed that disease-causing missense SNVs and in-frame insertions and deletions tend to be enriched at the interaction interfaces of proteins associated with corresponding disorders [15] . They also showed that the disease specificity of different mutations of the same gene can be explained by their location on the interaction interfaces . Another important feature that has emerged from studies of genomic variants on protein structure ( without consideration of network interactions ) is that benign missense polymorphisms tend to occur at solvent exposed sites on protein structure , while disease-causing missense SNVs tend to be more buried [16] . Previous studies examining the relationship of functional significance and selection properties of genes with network topology have mostly focused on networks with a singular mode of interactions between genes or their protein products , for example physical protein-protein interactions . However , a gene and its protein products can be involved in various biological networks and its role and consequently its centrality can vary across these networks . For example , SIX5 is a transcription factor gene that targets 360 genes in the human regulatory network but interacts with only one protein in the physical PPI network [17] , [18] . This gene is of high functional significance since its disruption causes Branchio-oto-renal syndrome , a developmental disorder characterized by the association of branchial arch defects , hearing loss and renal anomalies [19] . In this study we examine the relationship of functional essentiality and selection with various biological networks – protein-protein interaction ( PPI ) , phosphorylation , signaling , metabolic , genetic and regulatory . This enables us to understand the functional importance and selection constraints on genes in a global systemic approach . Moreover , although it has been shown that low evolutionary conservation of LoF- tolerant genes and their large distance from recessive disease genes in PPI network can be used to predict disease causation of variants [5] , their unique properties in diverse biological networks have not been exploited before . Here , we use the distinguishing network and evolutionary properties of functionally essential and LoF-tolerant genes to build a predictive model for global damage caused by novel variants . Using this model , we are able to compute functional indispensability scores for all protein-coding genes . The biological networks studied in this work include – PPI , phosphorylation , metabolic , signaling , genetic and regulatory ( Materials and Methods ) . Some of these networks represent direct physical interactions between proteins , for example , PPI . On the other hand , genetic and regulatory networks contain indirect interactions between gene pairs . Additionally , some networks such as phosphorylation , metabolic , signaling and regulatory are directional with an upstream and downstream gene , whereas PPI and genetic interactions are undirected . While a gene can have a vital role in one pathway or network , it might not be as crucial in another network . Therefore , we pool together data from all the above-mentioned biological networks to construct a unified global network , which we term Multinet ( Materials and Methods ) . The Multinet enables the analyses of the genes via their roles in the individual networks and the combined network . We note that some interactions between two different networks can be shared . For example , an interaction in which gene A phosphorylates gene B can occur in both phosphorylation and PPI networks . However , we find that out of ∼110 , 000 interactions in our data set , only 881 interactions occur in more than one network . Thus the vast majority of interactions in our data are unique or non-redundant . This observation reiterates the fact that interactions of genes vary across different networks and it is crucial to include all the networks while analyzing the relationship between functional importance and selection constraints with global network centrality . The distribution of 881 interactions which occur in more than one network is shown in Supporting Figure S1 . The numbers of genes and unique interactions in each network are shown in Supporting Table S1 . In this section we investigate the relationship between functional significance of genes and their properties in various biological networks . All human protein-coding genes are divided into four categories based on their known disease susceptibilities and functional impact . A ‘gene significance score’ ranging from 3 to 0 is assigned to each gene: 3 for essential genes , 2 for all genes with disease-causing mutations in HGMD , 0 for LoF-tolerant genes and 1 for all the remaining genes that do not fit into any of the above categories ( Materials and Methods ) . We then correlate these significance scores with the degree centralities of the genes in all networks . Degree centrality of a gene in any network is defined as the number of its interacting partners in that network . In order to estimate the total number of interacting partners of a gene , we use its connectivity ( number of interactions ) in the Multinet ( Materials and Methods ) . We find that gene significance scores show positive correlation with degree centralities in most networks , though it is statistically significant only in PPI and signaling network and Multinet ( Figures 1 and 2A; Supporting Table S2 ) . Thus , in general , essential genes tend to be more connected in biological systems consistent with previous findings [8] . Surprisingly , we find a small but significant negative correlation between gene significance score and metabolic degree ( Spearman correlation coefficient or SCC = −0 . 07 , pvalue = 0 . 028 ) . We also find that , unlike most other degree centralities , the metabolic degree centrality of genes shows a significant positive correlation with the number of paralogs ( duplicated copies ) ( SCC = 0 . 15; pvalue = 8 . 26e-07 ) ( Supporting Table S3; Materials and Methods ) . Thus , it is possible that in case of a LoF mutation in a participating enzyme , the metabolic pathway can be re-routed to an alternate path , possibly involving a duplicated gene of the disabled enzyme . Our observation in the human metabolic network is in agreement with a previous study by Vitkup et al , in which they found that highly connected enzymes are no more likely to be essential than less connected enzymes in yeast metabolic network [12] . In this study we find that not only are essential genes unlikely to be highly connected in human metabolic network , LoF-tolerant genes ( whenever present in metabolic network ) are indeed more connected than essential genes ( Supporting Table S7 ) . This result demonstrates a major contrast between the structure of the metabolic network and other networks . In most biological networks , highly connected genes tend to have fewer duplicated copies; hence LoF mutations in them can have serious phenotypic consequences . Since this distinct trend of high degeneracy at hub proteins is observed only in the metabolic network , we further posit that this might be an evolutionary mechanism to increase tolerance towards damaging mutations . The uniqueness of such a ‘protective’ effect somewhat suggests an implicit level of greater functional importance of metabolic pathways as compared to other networks of gene interactions . Interestingly , we find that gene significance scores are positively correlated with the number of networks the gene is involved in ( Figures 1 and 2B ) . This indicates that genes involved in many networks can act as information bottlenecks between different systems and thus tend to be more essential . We next examine the relationship between selection constraints on genes and their network properties . We estimate evolutionary constraints over long time-scale by dN/dS ( ratio of missense to synonymous substitution rates ) computed from human-chimp ortholog alignments ( Materials and Methods ) . dN/dS<1 indicates purifying selection while values close to 1 indicate neutral selection and dN/dS>1 indicates positive selection . We find that dN/dS values of genes are negatively correlated with their degree centralities in all networks , though they reach significance in PPI , phosphorylation , regulatory and Multinet networks ( Supporting Table S4 ) . This shows that highly connected genes tend to be under stronger purifying selection constraints over long evolutionary time-scale , in agreement with previous studies [10] . Furthermore , we analyze patterns of genetic variation in modern-day humans in relation to biological networks . We compute average heterozygosity of each gene to estimate its genetic variability using missense SNPs ( single nucleotide polymorphisms ) and their corresponding allele frequencies in three sets of populations from 1000 Genomes Pilot data ( Materials and Methods ) [4] . We find that there is a significant negative correlation between Multinet degree and heterozygosity of missense SNPs for all three populations , indicating more genetic variation at the periphery of networks ( the correlation is also significant for some populations in PPI , phosphorylation and regulatory networks ) ( Supporting Table S5 ) . Interestingly , we do not find a significant correlation of heterozygosity of synonymous SNPs with Multinet degree ( Supporting Table S6; Materials and Methods ) . Putting together , these results suggest that reduced genetic variability of highly connected genes with respect to missense SNPs is indeed due to selection constraints . When network edges between two genes correspond to physical interactions between their protein products , molecular level details of the interaction can be obtained by integrating three-dimensional protein structures with the underlying network data . Therefore , in order to understand the reasons for selection constraints in PPI network at higher resolution , we integrated three-dimensional protein structures with network interaction data to create structural interaction network ( SIN ) ( Figure 3A; Materials and Methods ) [14] , [15] , [20] . SIN is a subset of the larger PPI network and consists of 2 , 102 genes and 11 , 433 interactions . SIN construction allows us to estimate the number of interfaces used by a protein to interact with other proteins ( Figure 3A; Materials and Methods ) . We find that there is a significant positive correlation between gene significance scores and the number of interfaces used by their protein products in SIN ( Figure 1 ) . Thus , protein products of essential genes tend to use more interaction interfaces than those of LoF-tolerant genes . We also find that the number of interfaces used by the protein to interact with other proteins in SIN is positively correlated with their degree centrality in PPI network ( SCC = 0 . 18 , pvalue = 1 . 06e-09 ) . This shows that hub proteins tend to have more interaction interfaces . Thus , it is likely that higher number of interfaces possessed by protein products of essential genes could partly be a result of their higher degree centrality in PPI network . We next examine the impact of missense SNVs on protein structure in relation to SIN . We find that , in general , residues with disease-causing missense SNVs tend to be more buried inside protein structure than polymorphic residues ( Figure 3B ) . Our observation is consistent with previous findings which have reported that missense mutations buried inside protein structure tend to be more deleterious than those on surface [16] . However , these previous studies treated all proteins equally and did not differentiate between hub and non-hub proteins in PPI network . When we treat hub ( degree centrality> = 50 ) and non-hub proteins separately , we find that accessible surface area for residues with missense disease mutations is higher for hub proteins ( Wilcoxon rank sum pvalue = 0 . 014; Supporting Figure S2 ) . We also observe a significant positive correlation between the degree centrality of protein and the accessible surface area of their residues undergoing disease mutations ( SCC = 0 . 028 , pvalue = 3 . 12e-03 ) . These results show that hub proteins tend to have a higher fraction of missense disease mutations on their exposed surface . This result is very reasonable in light of our observation that hub proteins tend to have more interaction interfaces ( see preceding paragraph ) , thereby having a higher fraction of their exposed surface under selection constraints . In order to further examine the close correlation of network and evolutionary properties with gene essentiality we use a logistic regression model to differentiate essential genes from LoF-tolerant genes ( Materials and Methods ) . Network features used to train the logistic regression model include degree centralities in Multinet and all networks separately ( PPI , phosphorylation , signaling , metabolic , genetic and regulatory ) , number of networks the gene is involved in and number of interfaces used in SIN . Selection properties used in the model include human-chimp dN/dS ratios and average heterozygosities of both synonymous and missense SNPs in modern human populations . The average values of these features for LoF-tolerant and essential genes along with corresponding Wilcoxon rank sum pvalues are provided in Supporting Table S7 ( see also Figure 1 ) . Using these features we obtain an excellent classification accuracy for 140 LoF-tolerant and 115 essential genes with AUC = 0 . 914 ( Figure 4A; Materials and Methods ) . Network properties that contribute significantly to the model include degree centralities in regulatory , genetic and metabolic networks as well as number of networks the gene is involved in ( Materials and Methods ) . On further examination of network participation of LoF-tolerant and essential genes , we find that most LoF-tolerant genes are not involved in any network and some of them are involved in a very small number of networks ( Figure 4B ) . On the other hand , most essential genes are involved in many networks ( Figure 4C ) . Genes involved in a variety of networks serve as information bottlenecks between different systems and hence are more likely to be essential . We note that absence in some networks could partially be due to missing network data in our study and/or a bias in existing databases . Essential genes are more likely to have been the focus of previous research studies , for example PPI studies , and hence more likely to be present in our PPI network . They also tend to have more regulatory interactions and thus are more likely to be present in our regulatory network ( which consists of 118 transcription factors and their target genes: the most comprehensive human regulatory network available to our knowledge ) [17] . However , the strength of our model lies in its use of many different network properties to minimize the biases resulting from the use of a single network property or data resource . Furthermore , to test the robustness of our model , we computed the AUC for separation of LoF-tolerant and essential genes multiple times – each time randomly removing 10% of the edges from a network and rebuilding the Multinet . After repeating this for all the networks , we find minimal change in the AUC ( ranging from 0 . 914 to 0 . 912 ) , which shows that our model is quite robust to changing some edges in individual networks . We next perform an independent validation of our model by applying it on all genes that are neither LoF-tolerant nor essential . Interestingly , we find that predicted functional indispensability scores are in the following order: genes with known disease-causing mutations have significantly higher scores than genes identified in genome-wide association ( GWA ) studies ( Wilcoxon rank sum pvalue = 7 . 62e-05 ) , which are in turn significantly higher than all the remaining neutral genes ( Wilcoxon rank sum pvalue<2 . 2e-16 ) ( Figure 4D ) . Genes identified in GWA studies are associated with phenotypic consequences , while they are not necessarily the causal genes . Hence it is reassuring that genes with known disease-causing mutations in HGMD receive significantly higher scores than those identified in GWA studies . This validation exercise demonstrates that our model can help researchers pick candidate disease genes in clinical sequencing studies . We have provided the predicted scores for all the genes in Supporting Table S8 . We note that the predicted functional indispensability scores are continuous scores unlike the discrete gene significance scores used to compute correlations in an earlier section . Genes and their protein products work in collaboration with other genes to form biological systems that perform specific tasks . For a systemic understanding of the role a gene plays , there is a need to integrate different modes of gene interactions . In this work we pool together interaction data from various biological systems ( PPI , phosphorylation , signaling , metabolic , genetic and regulatory ) to create a unified Multinet , enabling the computation of degree centrality of the genes in their individual networks and in the context of the entire Multinet ( Supporting Table S8 ) . Subsequent analysis of functional significance and evolutionary properties of genes allows us to relate genomic sequence variants in individual genes to their functional effects in individual and global networks . We find that highly connected genes in the Multinet and genes that participate in many biological systems tend to be more functionally significant , have fewer paralogs and resist mutations in healthy humans . While we also observe similar trends in most of the constituent networks of the Multinet , the metabolic network seems to be an exception . Highly connected genes in the metabolic network tend to have more paralogs and are more tolerant to LoF mutations . Next , we combine three-dimensional protein structural information with PPI network to create structural interaction network ( SIN ) and understand selection on protein structure at molecular level detail . We find that functionally essential genes ( which are more likely to encode for hub proteins ) tend to use more interfaces to interact with other proteins . We also observe that hub proteins in PPI network contain a higher fraction of disease-causing mutations on their solvent exposed surface , as compared to non-hub proteins . Thus , although generally missense SNVs on exposed protein surface are more likely to be benign , our results show that those on the surface of hub proteins are more likely to be deleterious [21] . Finally , we integrate network and selection properties of genes to build a logistic regression model which can separate LoF-tolerant and essential genes with high accuracy ( AUC = 0 . 91 ) . Application of the model on all genes shows that it predicts higher functional indispensability scores for genes with known disease-causing mutations than genes identified in GWA studies , which themselves have higher scores than remaining neutral genes . The predicted functional indispensability scores for all genes are made publicly available and can be used to predict candidate disease genes in future clinical studies . These scores are indicators of global damage caused by deleterious mutations in coding genes – including nonsense and missense SNVs and in-frame and frame-shift indels . As mentioned above , nonsense SNVs and frame-shift indels are mostly assumed to disable gene function . However , missense SNVs and in-frame indels are more complex since they may or may not have a deleterious impact . Various methods exist to predict the functional effects of missense SNVs , for example , SIFT and PolyPhen [21] , [22] . While these methods examine the tolerance of individual sites in genes to missense mutations , they do not take into account the functional significance of the entire gene . For example , a moderately deleterious missense SNV in a highly significant gene can be equally or more damaging than a strongly deleterious missense SNV in a less significant gene . Our method to compute functional indispensability scores for entire genes can be combined with scores predicted by SIFT and PolyPhen to obtain a more comprehensive view of the functional effects of genomic variation . We note that even though our model is very robust to the removal of some edges in individual networks , the incomplete and biased nature of existing biological networks data may constitute a caveat in our study . However , to our knowledge , this is the first comprehensive genome-wide study linking genetic variants at population scale as well as disease variants with a vast body of available network resources . Models developed and applied in this study can be further expanded as more interaction data is obtained and further population genetics projects are undertaken , particularly with the future phases of the 1000 Genomes project . Human protein-protein interaction and genetic interaction networks were extracted from BIOGRID ( release 3 . 1 . 83 ) [18] containing 43 , 722 and 263 interactions , respectively . Regulatory network ( relationship between transcription factors and target genes ) was from ENCODE data [17] . Metabolic enzyme network contained directed linkages from upstream enzymes to downstream enzymes , based on compound reactions in KEGG [23] . Phosphorylation network in human contains 28 , 637 directed kinase-substrate interactions between 2 , 392 genes [24] . The signaling network in this study is constructed based on 1 , 011 interactions and 527 proteins ( downloaded July 2011 ) from human signaling pathways obtained from the SignaLink database ( http://www . signalink . org/ ) [25] . SignaLink offers an easily-downloadable and well-curated set of interactions from eight major signaling pathways found in humans that are not tissue-specific , namely EGF/MAPK , Ins/IGF , TGF-β , Wnt , Hedgehog , JAK/STAT , Notch and NHR ( Nuclear Hormone Receptors ) . Manual data curation was performed in SignaLink by extensive literature survey of primary experimental evidence of these interactions , resulting in expansion of verified interaction data for the corresponding signaling pathways in protein interaction databases such as the KEGG [26] , Reactome [27] and NetPath [28] , while maintaining substantial overlaps with these databases . A detailed description of the curation process and comparisons between these databases and SignaLink can be found in [25] . Throughout the article , connectivity of the gene in PPI , phosphorylation , signaling and metabolic networks refers to connectivity of the protein product of the gene . Interactions from all the above networks were combined to create Multinet . If a gene pair interacts in multiple networks or shows both upstream and downstream connection in a directional network , the interaction is counted once in Multinet . The list of 140 LoF-tolerant genes was obtained from MacArthur et al [5] . This list contains genes that show homozygous LoF mutations in at least one individual in 1000 Genomes pilot data [4] . The list of 115 essential genes was obtained from Liao et al [29] . These genes exhibit clinical features of death before puberty or infertility when LoF mutations occur . The list of 2 , 451 disease genes was obtained from HGMD ( Human Gene Mutation Database ) [30] . All the genes with any disease-causing mutation ( DM tag in HGMD ) were used . If any gene occurred in more than one category , its category was decided in a hierarchical fashion as follows: essential , followed by disease followed by LoF-tolerant . The remaining 19 , 267 genes were assigned the category of neutral . The list of genes identified in GWA studies was obtained from the NHGRI GWAS catalogue ( https://www . genome . gov/26525384#download ) . Number of paralogs for each gene and dN/dS values for human-chimp orthologs were obtained from Ensembl using BioMart [31] . SNPs in modern-day humans and their allele frequencies were obtained from the low-coverage pilot phase of the 1000 Genomes Project [4] . This phase consisted of 60 individuals of CEU ( Utah residents with Northern and Western European Ancestry ) , 59 individuals of YRI ( Yoruba in Ibadan , Nigeria ) and 60 individuals of CHB+JPT ( Han Chinese in Beijing , China and Japanese in Tokyo , Japan ) populations . Heterozygosity value is calculated as 2pq , where p and q correspond to the frequencies of the two alleles . Average heterozygosity for a gene is defined as the average heterozygosity of the SNPs in that gene , where heterozygosities of missense and synonymous SNPs are computed separately .	The number of personal genomes sequenced has grown rapidly over the last few years and is likely to grow further . In order to use the DNA sequence variants amongst individuals for personalized medicine , we need to understand the functional impact of these variants . Deleterious variants in genes can have a wide spectrum of global effects , ranging from fatal for essential genes to no obvious damaging effect for loss-of-function tolerant genes . The global effect of a gene mutation is largely governed by the diverse biological networks in which the gene participates . Since genes participate in many networks , no singular network captures the global picture of gene interactions . Here we integrate the diverse modes of gene interactions ( regulatory , genetic , phosphorylation , signaling , metabolic and physical protein-protein interactions ) to create a unified biological network . We then exploit the unique properties of loss-of-function tolerant and essential genes in this unified network to build a computational model that can predict global perturbation caused by deleterious mutations in all genes . Our model can distinguish between these two gene sets with high accuracy and we further show that it can be used for interpretation of variants involved in Mendelian diseases and in complex disorders probed by genome-wide association studies .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "systems", "biology", "genomics", "genetics", "biology", "computational", "biology", "genetics", "and", "genomics" ]	2013	Interpretation of Genomic Variants Using a Unified Biological Network Approach
We are flat-faced hominins with an external nose that protrudes from the face . This feature was derived in the genus Homo , along with facial flattening and reorientation to form a high nasal cavity . The nasal passage conditions the inhaled air in terms of temperature and humidity to match the conditions required in the lung , and its anatomical variation is believed to be evolutionarily sensitive to the ambient atmospheric conditions of a given habitat . In this study , we used computational fluid dynamics ( CFD ) with three-dimensional topology models of the nasal passage under the same simulation conditions , to investigate air-conditioning performance in humans , chimpanzees , and macaques . The CFD simulation showed a horizontal straight flow of inhaled air in chimpanzees and macaques , contrasting with the upward and curved flow in humans . The inhaled air is conditioned poorly in humans compared with nonhuman primates . Virtual modifications to the human external nose topology , in which the nasal vestibule and valve are modified to resemble those of chimpanzees , change the airflow to be horizontal , but have little influence on the air-conditioning performance in humans . These findings suggest that morphological variation of the nasal passage topology was only weakly sensitive to the ambient atmosphere conditions; rather , the high nasal cavity in humans was formed simply by evolutionary facial reorganization in the divergence of Homo from the other hominin lineages , impairing the air-conditioning performance . Even though the inhaled air is not adjusted well within the nasal cavity in humans , it can be fully conditioned subsequently in the pharyngeal cavity , which is lengthened in the flat-faced Homo . Thus , the air-conditioning faculty in the nasal passages was probably impaired in early Homo members , although they have survived successfully under the fluctuating climate of the Plio-Pleistocene , and then they moved “Out of Africa” to explore the more severe climates of Eurasia . A flat , short face is one of the legacies of the genus Homo [1 , 2] . The facial component remains short and fully below the expanded forehead in this genus , and this contrasts with earlier and contemporary hominins such as the australopithecines , which possessed a long face that protruded away from the brain case in a manner analogous to nonhuman hominids , e . g . , chimpanzees [1–3] . Consequently , the external nose protrudes from the face [4] , the nasal cavity within the facial cranium is high and quadrangular in a lateral view , and the vertically oriented nasal vestibule is connected close to the floor of the tall nasal cavity in humans [2 , 5] . This pattern contrasts with that found in nonhuman primates , which possess a long and triangular nasal cavity , and a horizontally oriented vestibule that is connected vertically with the middle of the cavity [2 , 5] . However , the subsequent pharyngeal cavity is much longer in humans than in nonhuman primates [6–10] . The nasal passage , including the nasal vestibule and cavity , conditions the inhaled air , as well as performing other functions such as olfactory sensing , dust filtering , and voice resonance [11 , 12] . The pharyngeal cavity also participates in conditioning the air that enters from the nasal cavity [11] . Insufficient conditioning can damage the mucosal tissues in the respiratory system and impair respiratory performance , thereby undermining health and increasing the likelihood of death [11 , 12] . Thus , despite the evolutionary modifications in the nasal anatomy in the phyletic divergence of Homo from the other hominin lineages , adequate air conditioning must have been maintained , particularly to ensure their successful survival in the severely fluctuating climate of the Pleistocene and their subsequent spread from Africa to Eurasia [1 , 13 , 14] . In this study , we compared the principles and performance of air conditioning in humans , chimpanzees , and macaques by using a computational fluid dynamics ( CFD ) model [15] to simulate the airflow and heat and water exchanges over the mucosal surface in the nasal passage . The human CFD model used here simulates the predicted airflow and air-conditioning performance reliable for humans [15] . Three-dimensional models of the nasal passage topology were produced based on tomography scans of the three genera . The models include no paranasal sinuses . We compared the air-conditioning performance in the three genera using the same simulation conditions: heat and water exchange were predicted with a simulation model based on the histological compositions of the mucosal layers and the average surface temperature of the human nasal passage , i . e . , 100% relative humidity ( % RH ) at 34°C ( 3 . 34% of the mass fraction of water; % MF ) [15] . This means that this study examines the differences in performance that are caused by the anatomical differences of the nasal passage in the three genera , but does not simulate a real performance in nonhuman primates . Whereas similar CFD analyses performed in the same subjects of macaques showed a minor contribution of the maxillary sinus to air conditioning [16] , here we examine them again to compare the air-conditioning performance of the three genera using the same simulation conditions . To evaluate the effects of the human external nose , we also produced two virtual topology models: a “no-valve” model where the nasal valve—a narrow slit-like channel between the nasal vestibule and cavity—was removed virtually; and a “horizontal” model where the vertically oriented vestibule was made horizontal , as seen in chimpanzees . We evaluated the performance among the three species in varied ambient atmospheric conditions , and discuss the evolutionary modifications in air-conditioning performance in the divergence of Homo from the other hominin lineages lineage , using nonhuman primates as a model for the latter hominins . This study for animals was performed in strict accordance with the recommendations in the third edition of the Guidelines for the Care and Use of Laboratory Primates at the Primate Research Institute of the Kyoto University ( KUPRI ) , Inuyama , Japan . The protocol was approved by the Animal Welfare and Animal Care Committee at KUPRI ( Permit Numbers: 2009–075 , 2010–027 , 2011–067 , and 2012–075 ) . The chimpanzees were anesthetized intramuscularly with 3 . 5 mg ketamine hydrochloride ( Sankyo-Parke-Davis & Co . , Inc . ) and 0 . 035 mg medetomidine hydrochloride ( Meiji Seika Pharma Co . , Ltd . , Tokyo , Japan ) per kilogram of body weight . The anesthesia was maintained with sevoflurane ( Dainippon Sumitomo Pharma Co . , Ltd . , Osaka , Japan ) delivered in oxygen through a precision vaporizer and a rebreathing circuit . The macaques were anesthetized intramuscularly with 2 . 5 mg ketamine hydrochloride and 0 . 1 mg medetomidine hydrochloride per kilogram of body weight . Every effort was made to minimize suffering . The daily care and housing facilities strictly conformed to the recommendations in the third edition of the Guidelines for the Care and Use of Laboratory Primates at the KUPRI . To ensure the animals' health and welfare , their general appearance was daily monitored and recorded , along with their daily food and fluid intake . This study for humans was performed in strict accordance with the recommendations in the Declaration of Helsinki , Ethical Principle of Medical Research Involving Human Subjects prepared by the World Medical Association . All subjects gave an informed consent . The protocol was approved by the Human Research Ethics Committee of KUPRI ( Permit Number: H2011-06 ) . Ten non-human primates—four chimpanzees , Pan troglodytes [17] , four Japanese macaques , Macaca fuscata , and two Rhesus macaques , M . mulatta—which were reared at KUPRI , Inuyama , Japan , were scanned using a computed tomography scanner ( Asteion Premium 4 , Toshiba Medical Systems Co . , Otawara , Japan ) at the KUPRI ( S1 Table ) . The two species of macaques are here regarded as subjects of a same genus Macaca along with a genus Pan . All of the CT scans obtained in this study came from subjects without any history of surgery and had few abnormal traits in their heads , and few artifacts distorted the images of the nasal region . The scans were registered under PRICT # ( S1 Table ) and are available via the website of the Digital Morphology Museum of KUPRI ( dmm . pri . kyoto-u . ac . jp/archive/ ) . The scans of macaques used here were also used for another CFD study [16] . Six human volunteers were scanned with a magnetic resonance imaging scanner ( Magnetom Verio , Siemens AG , Munich , Germany ) at the Brain Activity Imaging Center , ATR-Promotions , Seika , Japan ( S1 Table ) . Among them , the following CFD simulation analyses used the scans of the five subjects who had no surgery , few abnormal traits , and few artifacts distorting the images in the nasal region . The CFD simulations with heat and water exchange were performed as described [15 , 18] . The physiological model used here , developed based on the previous models [18] , reflects human respiratory physiology and histology , including latent heat , to simulate well real performances of the airflow , heat and water exchanges for humans [15] . The following procedures of the nasal topological models and CFD analyses are the same as those used by Hanida et al . [15] , excluding those for generating the virtual topological models . It is slightly different from the study of Mori et al . [16] using the same macaques , in simulation model and boundary conditions . The voxel data of the nasal passage anatomy were reconstructed from computed tomography scans for nonhuman primates and magnetic resonance imaging scans for human volunteers . The black area representing the air filling the nasal passage was extracted first by using a threshold of brightness with Avizo 7 ( FEI ) , and then the voxel data were reconstructed . After converting the voxel data to STereo Lithography ( STL ) data , these were modified into data representing the smooth surface using Magics 9 . 5 ( Materialize Inc . , Leuven , Belgium ) . Finally , a tetrahedron mesh with the mesh size of Δx = 2 . 10 to 3 . 65×10−4 mm , depending on the size of the subjects , was generated from the modified STL data using Gambit 2 . 4 ( ANSYS Inc . , Canonsburg , PA , USA; S1 Table ) . The computational meshes had 2 . 66 to 3 . 70 million tetrahedral cells ( S1 Table ) . The present solutions are evaluated independently to the mesh size: there were few differences between the solutions by the present mesh size and the minimum mesh size ( Δx = 2 . 00 to 3 . 30×10−4 mm , depending on the size of the subjects ) in each subject; i . e . , up to 1 . 6% of the flow velocity for the same frontal contour . The no-valve and horizontal nasal vestibular topology models were generated from the original smoothed STL data using Rhinoceros ( AppliCraft Co . , Ltd , Tokyo , Japan ) . We defined the basal plane for making the modifications as almost parallel to the narrow channel between the nasal vestibule and the nasal cavity , which approximately corresponded to the nasal valve ( Fig 1A ) . Using the loft function in Rhinoceros , we generated a straight surface between the basal plane to the nostril on each side , thereby removing the effects of the nasal valve ( i . e . , the no-valve model; Fig 1B ) . Next , we tilted the modified vestibule with its straight surface upward , thereby making its lower surface horizontal relative to the floor of the nasal cavity , as seen in the vestibule of chimpanzees ( i . e . , the horizontal model; Fig 1C; tilting angle: vol . 1 , 52 degrees; vol . 2 , 32 degrees; vol . 3 , 40 degrees; vol . 5 , 46 degrees , vol . 6 , 36 degrees ) . We performed steady-state analyses to examine the airflow , where the turbulence model was not employed . A steady simulation is reasonable under a normal breathing frequency and flow rate in the resting stage in humans [15 , 18–24] . The maximum Reynolds numbers ranges from 135 to 1264 at the position of pharyngeal cavity where its cross-sectional area was measured in the subjects use here , by calculated with estimates of the inhaled air velocity in the resting stage ( see the subsection Boundary conditions for calculation ) . These values are lower than the critical Re value of 2300 denoting the transition between laminar to turbulent flow , and much lower than the Re of 5000 which is considered to be completely turbulent [19 , 21 , 24] . The nasal flow is regarded as being mostly laminar in the resting stage of the subject used here . The Strouhal number value for the system is less than 0 . 25 [24] . Moreover , the Womersley number for human breathing is small , thereby indicating that any inertial effects on the flow pattern may be regarded as negligible [24 , 25] . We used the CFD simulation model developed by Hanida et al . [15] to model an incompressible , viscid , laminar airflow in the nasal cavity with heat and water transport . The equations were solved using the fluid simulation software FLUENT 6 . 3 ( ANSYS Inc . , Canonburg , USA ) . The simulation was governed by the Navier–Stokes equation—the conservation of momentum Eq ( 1 ) —by the equation of conservation of mass Eq ( 2 ) , by the transport equation of energy Eq ( 3 ) , and by the transport equation for the mass fraction of water Eq ( 4 ) . Here t , u , p , ρ , v , K , T , Cp , F , and D denote time , velocity , pressure , density , kinematic viscosity , thermal conductivity , temperature , specific heat , mass fraction of water , and mass diffusion coefficient , respectively . We regarded a solution as being of steady-state , after T ( time ) advanced substantially: we repeated the steps of calculation sufficiently until the values of system parameters reach the criterions of convergence , i . e , continuity , X , Y , Z-velocity , energy , and H2O reach 1×10−4 , 1×10−5 , 1×10−7 , and 1×10−4 , respectively , by the FLUENT 6 . 3 . The wall of the nasal passage was modeled by the tissue and epithelial layers ( Fig 2 ) to simulate the exchange of heat and water from the vascular layer to the air via the mucous membrane [15] . The mucous membrane includes the membrane epithelia , nasal glands , blood vessels , and capillary blood vessels [26] , and thus it varies in thickness between 0 . 3 and 5 mm according to location in humans . The present model was composed of a smooth surface and a constant thickness of 0 . 2 mm in the nasal vestibular region and 0 . 5 mm in the other region , which was specified to simulate the actual performance in humans , after Hanida et al . [15] . Thus , the predicted air-conditioning performance may be slightly worse than that actually found in smaller-bodied chimpanzees and macaques having thinner mucosal layers . The optimum values of the temperature and humidity of the inhaled air were calculated by setting a boundary condition representing the heat and water exchange on the surface of the epithelial layer . Fig 2A illustrates the present simulation model for heat exchange [15 , 18] . Heat is transferred between the air and the tissue layer via the epithelial layer . The heat transport of Qep from the tissue side is determined by Eq ( 5 ) . The latent heat of Qlatent is calculated from Eq ( 6 ) , where L and Wbl denote the specific latent heat and water flux from the surface of the epithelial layer , respectively . L is defined by Eq ( 7 ) , which was calculated by cubic fitting to the data reported by Rogers and Yau [27] . The total heat transport , Qtotal , is defined by Eq ( 8 ) as a flux boundary condition for the energy Eq ( 3 ) . TS , TTis , Kep and δep denote the temperature of the surface , the tissue layer temperature , the thermal conductivity of the epithelial layer and the epithelial layer thickness , respectively . TTis is constant , set here at 34°C , the temperature commonly measured in the human nasal region . The predicted air-conditioning performance is slightly inferior compared with that in reality for chimpanzees and macaques having higher body temperature . Nevertheless , this value is commonly used here for simulation both in humans and non-human primates , to examine the differences of air-conditioning performance which are caused by the morphological differences in nasal passage anatomy between them . The thermal conductivity of the mucous membrane Kep is 0 . 6 W/mK that is thermal conductivity of water [28] , because the liquid mucous membrane is assumed in the model used here [15 , 18] . TS is determined by Qtotal , which comprises Qep and Qlatent . Fig 2B illustrates the wall model , implemented with a boundary layer to define the boundary condition of species transport for water exchange [15 , 18] . The model is based on Fick’s law in that the flux diffusion is proportional to the concentration gradient of water . Here , we used the Dirichlet-type boundary condition ( i . e . , fixed transport ) in the FLUENT software . The two-film theory was used here to evaluate the mass of species transport between a liquid phase and a gas phase across a boundary . The thickness of the boundary layer was set at 0 . 5 mm [15] . Wbl is the water flux from the boundary layer , which is determined from Eq ( 9 ) , and this was used to calculate the latent heat of Eqs ( 6 ) and ( 8 ) . Wep is the water flux from the tissue layer , which is determined from Eq ( 10 ) . Here F , FS , FTis , δbl , δep , Dbl , and Dep denote the water fraction in the boundary layer , the water fraction on the epithelial surface , the water fraction on the tissue layer , the boundary layer thickness , the epithelial layer thickness , mass diffusion coefficient of the boundary layer , and mass diffusion coefficient of the epithelial layer , respectively . Dbl and Dep are 3 . 0 × 10−5 m2/s and 2 . 6 × 10−5 m2/s , respectively [29] . FTis is 3 . 34% of the water mass fraction in 100% of the relative humidity at 34°C . It is noted that diffusion in the boundary layer is greater than that of the membrane layers [15] . The water flux is transported from the tissue through the epithelial and boundary layers to the air . Simultaneously solving Eqs ( 9 ) and ( 10 ) for FS gives Eq ( 11 ) . The temperature is not dominant in Eqs ( 9–11 ) , and the water transport is not regarded as being dependent on a temperature in this model . To enable mass flux of species transport , FS was fixed as the boundary condition for water exchange . This boundary condition was implemented as a user-defined function in FLUENT software . Note that the nasal vestibule is covered with epidermis , where water is not exchanged [15] . The external nostril was modeled as a free inlet , and no-slip boundary conditions were applied at the walls , while the outward velocity was assigned at the pharynx [15] . The time-averaged velocity of the inhaled air at the pharynx was calculated based on estimates of the resting tidal volume and the respiratory rate , as well as the measurement of the cross-sectional area of the pharyngeal region at a given position for each subject ( S1 Table ) . The cross-sectional area was calculated at a given location in the pharynx based on the CT scans using Magics software . The resting tidal volume was estimated by Eq ( 12 ) [30] . Here , TV and BW denote the estimate of the resting tidal volume ( ml ) and measured body weight ( kg ) , respectively . The resting respiratory rate was estimated by Eq ( 13 ) [31] . Here , f denotes the estimate of the respiratory rate ( breaths/second , Hz ) . Finally , the time-averaged velocity was calculated by Eq ( 14 ) . Here , FV and CA denote the time-averaged flow velocity ( m/s ) and the measured cross-sectional area at a given location of the pharynx ( mm2 ) , respectively . The CFD simulations were performed in three ambient atmospheric air conditions: cold–dry , 10% RH at 5°C ( 0 . 05% MF ) ; hot–dry , 5% RH at 40°C ( 0 . 23% MF ) ; and warm–wet , 60% RH at 30°C ( 1 . 58% MF ) . The resulting spatial pattern of the vector quantity representing the velocity and direction of the airflow is illustrated using streamlines in different colors that was computed from the points on the plane of external nostrils [15 , 18] . The number of streamlines is decided dependent on the area of the plane of nostrils , and it reflects the relative airflow volume for a given subject , allowing us to examine where the air mainly flows . Those of the scalar quantity representing the temperature and water vapor volume are illustrated using contours in different colors [15 , 18] . Our simulations of CFD showed that the airflow direction in the nasal cavities of chimpanzees and macaques differs in some key regions from that in humans . In humans , the inhaled air flows upward along the ascending vestibular region into the nasal cavity and downward to the oropharynx ( Fig 3A and S1 Fig ) , whereas in chimpanzees and macaques the air flows straight from the horizontal vestibule , through the nasal cavity , and out to the oropharynx ( Fig 3A and S2 Fig ) . Our predictions for humans agree with many studies of different humans with variable external nose and nostril morphologies [19 , 23 , 32–38] . However , irrespective of the differences in the flow direction , the air passes through the mid-medial to the inferior region in greater volume and velocity compared with the peripheral and superior regions of the nasal cavity in all three genera in a similar manner ( Fig 3 and S1 and S2 Figs ) . Although the location differs slightly , this major flow passage was also determined in many previous experimental and CFD simulation studies in humans [19 , 20 , 23 , 32–38] and macaques [39] . Thus , the major flow passage through the nasal cavity in humans is almost the same as in chimpanzees and macaques , although humans have an upward and curved airflow . In our simulations , the inhaled air was conditioned well in chimpanzees and macaques , even in severe ambient conditions ( Figs 4 and 5; Table 1 ) , where the inhaled air was adjusted to approximately 34°C and saturated to almost 100% RH before reaching the nasopharyngeal region , i . e , the most posterior frontal contour ( Figs 4 and 5; Table 1 ) . However , the air-conditioning performance was lower in humans according to our CFD model . Thus , the air was not adjusted fully in the two dry conditions in that air with <80% of the required water content at 34°C remained in the nasopharyngeal region ( Figs 4 and 5; Table 1 ) . The cold and dry air was not well adjusted in terms of temperature so that air at <30°C also remained in the nasopharyngeal region ( see the most posterior contour on Figs 4B , 4D , 5B and 5D; Table 1 ) , whereas the hot and dry air was adjusted well to 34°C ( Figs 4B , 4D , 5B and 5D; Table 1 ) . The warm and wet air was adjusted well to over 32°C at the nasopharyngeal level , but it was not fully adjusted in terms of humidity: air with <90% of the required water content at 34°C remained there ( Figs 4A , 4D , 5A and 5D; Table 1 ) . Nevertheless , such worse performances probably reflect the air-conditioning performance in healthy humans well . In fact , the air is often only conditioned up to 30°C or to 80–90% RH in the nasopharyngeal region of humans , even in comfortable ambient conditions [40–42] . Thus , inhaled air is not well conditioned in the nasal cavity of humans compared with nonhuman primates . The no-valve model resulted in a few changes in the airflow direction and air-conditioning performance for each human subject ( Fig 6A–6C; S3 Fig; Table 2 ) . The normal model exhibited a fast and diffusive flow through the nasal valve in all three genera ( Fig 3 , S3 Fig ) , but this flow was not found in the human no-valve model ( Fig 6A; S3 Fig ) . In the horizontal model , the airflow direction was changed to be slightly horizontal and straight , as seen in chimpanzees , but it had only a minor effect on the air-conditioning performance: the temperature and humidity at the nasopharyngeal level did not differ from those in the normal and no-valve models for each subject ( Fig 6D–6F; S3 Fig; Table 2 ) . Thus , the vertically oriented vestibule contributes to the upward airflow , but the topology of the nasal valve and vestibule makes little contribution to improving the air-conditioning performance in humans . In this CFD study , we used the same simulation conditions to compare the air-conditioning performance in humans and non-human primates . It should be noted that MRI scans were used for reconstructing the nasal passage models in humans , while CT scans were used in nonhuman primates . A same type of tissue is resolved differently between the two modalities . In a precise sense , mixing CT and MRI scans potentially alters the thickness of the airway and flow velocities for a same subject . Nevertheless , the present results using MRI scans also show an airflow pattern , velocities , and air-conditioning performance that are similar to those by a same CFD model using a CT-based model in humans [15] . Such mixing is available for gross comparisons [43] , as like the present gross comparison of humans and nonhuman primates . Thus , our study provides evidence supporting for the view that the air-conditioning performance within the nasal cavity is less effective in humans than in nonhuman primates . The nasal topology is often regarded as being evolutionarily sensitive to the ambient atmospheric conditions of a habitat such as temperature and humidity for a given clade [45–48] . In fact , the inhaled air is already fully conditioned in the anterior region of the nasal cavity in chimpanzees and macaques . This finding does not contradict another CFD study using macaques and savannah monkeys , despite using slightly different CFD methods and conditions [16] . These findings mean that the morphology of the nasal cavity can accept some morphological evolutionary modifications that might impair air-conditioning in nonhuman primates . The earlier hominins other than the genus Homo have a nasal passage in a manner analogous to chimpanzees rather than humans , suggesting that they probably show the effective air-conditioning performance as seen in non-human primates . However , the characteristic facial reorganization in Homo has precluded the developmental elongation of the oral and nasal cavities [7–10] and has impaired their air-conditioning performance . The nasal cavity anatomy is believed to vary with an advantage to the climate conditions of the habitat of a given population in modern humans [47 , 48] , but the present finding supports the idea that morphological modification in the nasal region is only a weak evolutionary response to air-conditioning needs in the divergence of Homo from the other hominin lineages . Rather , the nasal region is regarded just as a buffering module for facial reorganization , in contrast with other modules , including the jaw , eye , and braincase , which have been modified yet have maintained their independent functions . Thus , human nasal topology was probably modified passively by evolutionary facial reorganization in early Homo , and such an evolutionary modification was not prevented by impaired the air-conditioning performance . The unique external nose in Homo is believed to confer some functional advantages during air conditioning , such as retaining the water vapor from expired air [4] or generating a vortex airflow with inhaled air to improve air conditioning [2 , 44] . The nasal vestibule within the external nose is coated with epidermis , including vestibular hairs , and it only exchanges heat with the air , which means that the nasal vestibule itself makes a limited contribution to the air conditioning that occurs within it . Our study confirmed that the nasal valve also has little effect on the air-conditioning performance in humans . In fact , the topography of the nasal valve has specific effects on the local airflow pattern within the nasal cavity such as vortices in the superior meatus [34] , but there are limited effects on the gross airflow pattern , including turbulence [34 , 44] . Further , our study also showed that the vertically oriented vestibule makes a major contribution to the generation of an upward airflow in the nasal cavity in humans , but the inhaled air is still well conditioned mainly in the mid-medial to inferior regions of the nasal cavity as seen in intact humans and non-human primates . Although the location differs slightly , such a major flow passage was confirmed in humans with variable external nose and nostril morphologies in modern humans [38] . Irrespectively of varied nostrils , the vertically oriented vestibule had little effect on improving the air-conditioning performance , although the external nose morphologies including this feature could improve the transport of odorants to the superior olfactory slit in the upper nasal cavity . Thus , the unique external nose has little effect to improving the air-conditioning performance , and the impaired performance is more likely to be a consequence of modifications in the shape of the nasal cavity itself in Homo lineage since it was diversified from the other hominins , including australopithecines , in the beginning of the Early Pleistocene . Inhaled air can be adjusted through the pharyngeal cavity to be fully conditioned in humans , even though it is not fully adjusted in the nasal cavity . In the phyletic divergence of Homo from the other hominin lineages , facial flattening and reorganization has reduced the dimensions of the horizontal oral cavity along with the nasal cavity and pushed the tongue down toward the pharynx , thereby lengthening the vertical pharyngeal cavity [8 , 9] . Although the actual lengths of the pharyngeal cavity are unknown for each previous form of Homo [49 , 50] , the long pharyngeal cavity in extant humans contributes greatly to the sophisticated and flexible modifications of the topology of the supralaryngeal vocal tract from the glottis to the lips , which underlies human speech production [6 , 51] . This feature is also believed to provide a disadvantage for the other physiological functions of the pharynx such as swallowing , increasing the risk of accidental aspiration during deglutition of food and liquid boluses [52–54] . However , the long pharyngeal cavity could in part compensate for the impaired air-conditioning performance within the short nasal cavity . The Late Pliocene to Early Pleistocene periods were characterized by a highly fluctuating climate and a gradual transition from warm and humid to cool and arid environments , especially in the northern hemisphere [13 , 55] . These linked changes in the nasal and pharyngeal regions would in part have contributed to how flat-faced hominins , i . e . , Homo members , must have survived such fluctuations in climate , before they moved “Out of Africa” in the Early Pleistocene to explore the more severe climates and ecological environments of Eurasia .	This is the first investigation of nasal air conditioning in nonhuman hominoids based on computational fluid dynamics with digital topological models of the nasal passage made using medical imaging . Our comparative results of humans , chimpanzees , and macaques show that the inhaled air is conditioned poorly in humans compared with nonhuman primates . We also show that our protruding external nose has little effect on improving air conditioning . The nasal anatomy in Homo was weakly sensitive to the ambient atmosphere conditions in evolution , but was formed passively by facial reorganization in this genus . Even though the inhaled air is not adjusted well within the nasal cavity in humans , it can be fully conditioned subsequently in the pharyngeal cavity , which is lengthened in flat-faced Homo . Thus , despite an impaired air-conditioning conformation in the nasal passages , Homo members must have survived successfully under the fluctuating climate of the Plio-Pleistocene , and then they moved “Out of Africa” in the Early Pleistocene to explore the more severe climates and ecological environments of Eurasia .	[ "Abstract", "Introduction", "Materials", "and", "Methods", "Results", "Discussion" ]	[ "organismal", "evolution", "medicine", "and", "health", "sciences", "nose", "face", "hominin", "evolution", "human", "evolution", "vertebrates", "animals", "mammals", "primates", "respiratory", "system", "animal", "management", "animal", "performance", "old", "world", ...	2016	Impaired Air Conditioning within the Nasal Cavity in Flat-Faced Homo
The manual evaluation , classification and counting of biological objects demands for an enormous expenditure of time and subjective human input may be a source of error . Investigating the shape of red blood cells ( RBCs ) in microcapillary Poiseuille flow , we overcome this drawback by introducing a convolutional neural regression network for an automatic , outlier tolerant shape classification . From our experiments we expect two stable geometries: the so-called ‘slipper’ and ‘croissant’ shapes depending on the prevailing flow conditions and the cell-intrinsic parameters . Whereas croissants mostly occur at low shear rates , slippers evolve at higher flow velocities . With our method , we are able to find the transition point between both ‘phases’ of stable shapes which is of high interest to ensuing theoretical studies and numerical simulations . Using statistically based thresholds , from our data , we obtain so-called phase diagrams which are compared to manual evaluations . Prospectively , our concept allows us to perform objective analyses of measurements for a variety of flow conditions and to receive comparable results . Moreover , the proposed procedure enables unbiased studies on the influence of drugs on flow properties of single RBCs and the resulting macroscopic change of the flow behavior of whole blood . Amongst all human organs , blood is the most delocalized one , delivering oxygen from the respiratory system to the tissues in the body and transporting carbon dioxide back . On a microscopic scale , this is performed by red blood cells ( RBCs ) which form the largest fraction of cells in whole blood ( ≈ 99% ) . At rest , RBCs are biconcave discocytes with an average diameter of 8 μm and a height of 2 μm . Due to their flexible membrane , RBCs alter their shape under external stress prevalent in the microvascular network [1 , 2] . This feature is one of the key properties of RBCs , which allows them to squeeze through geometrical constrictions much smaller than their stress-free shape [3] , which is partly an intrinsic property of RBC morphology [4 , 5] and partly an active adaptation process [6 , 7] . Although data on the mechanical properties of RBC suspensions is widely known from rheological measurements [8] , the linkage to individual cell behavior is limited . Further , the comparison between capillary Poiseuille flow and pure shear flow prevalent in rheometers is difficult . Consequently , mimicking flow under physiological conditions in vitro demands for experimental setups such as PDMS-based microchannels , ubiquitous in lab-on-a-chip devices [9] . In this work , we focus on experiments of individual flowing RBCs , providing a holistic insight to individual cell mechanics . Hereby , the experimental data originate from a previous study on RBC shape geometry [10] and the data is reused for the introduction of a fully automated data analysis approach based on a deep learning convolutional neural network ( CNN ) [11] . As described in the preceding work , two stable RBC shapes are expected from the measurements: The so-called ‘croissant’ and the ‘slipper’ shape [12–14] . Their frequency of occurrence highly depends on the imposed flow conditions . Whereas axisymmetric croissants mostly appear at lower flow velocities , non-axisymmetric slippers are observed at higher shear rates [15 , 16] . However , besides these stable geometries , also a large number of indefinite shapes occur , especially in the so called phase transition range . These outliers or ‘other’ shapes make a considerable amount of the whole statistics . Due to their large shape variance , these cells cannot be assigned to a mutual exclusive class and can therefore not easily be involved in the machine learning processes . Overcoming this drawback , we present a regression based CNN aiming to distinguish between croissants , slippers and others by applying statistically derived thresholds to the net response . As a result , we obtain an automatically generated RBC phase diagram which relinquishes any subjective user input . Prospectively , we will be able to run comparable studies on RBC shape geometries in microchannels of variable size and to generate unbiased phase diagrams . Above all , data evaluation can then be performed in a highly time-effective manner . Human blood withdrawal from healthy donors as well as blood preparation and manipulation were performed according to regulations and protocols that were approved by the ethic commission of the “Ärztekammer des Saarlandes” ( reference No 24/12 ) . We obtained informed consent from the donors after the nature and possible consequences of the studies were explained . CNNs are digital image processing systems [17] with the ability to resolve and evaluate the details of an input image . Usually , they are used to e . g . recognize and classify particular objects or humanoid faces [18] within pictures . Here , a CNN is exploited to distinguish between the shape characteristics of RBCs in flow and to detect undefined outliers which occur due to channel imperfections , membrane damages , cell-cell-interactions , shape transitions or transients , and optical ambiguities . Independent of the particular use case , the architectures of CNNs usually follow the same design rules . They consist of an image input layer followed by a certain number of subsequent convolution stages ( cp . Fig 1 and Table 1 ) , and provide so-called interconnected layers forming an artificial neural network ( ANN ) to combine the convolution data in a final stage before the information is fed to the output layer nodes . The main stages of a CNN usually consist of several sublayers [19] including the actual convolutional layer , a non-linear rectification ‘reLU’ layer [20] , and a pooling layer [21] . Even though more sophisticated designs [22 , 23] exist , for RBC shape recognition we restrict ourselves to these layer types keeping the system as simple as possible . Convolutional layers make use of a number of convolutional kernels ( feature maps ) of particular size and are optimized to find the major characteristics of a set of various input images being subject to a sophisticated training process ( cf . section Training ) . For instance , this can include horizontal or vertical edges but also more detailed characteristics , such as the ‘tail’ of a cell , see Fig 2 . Convolving an image with a set of specially optimized convolution kernels then results in a number of output images showing clear differences in between the RBC shape classes . In a mathematical sense , convolutions are linear operations . Thus , CNNs could easily be contracted to a simple linear signal processing system if there were no non-linearities involved . Indeed , the usage of non-linearities resolving higher order dependencies in between the characteristics of a set of distinct input images is one of the key ideas behind neural networks . For CNNs , it is common practice to use a reLU layer to set the negative values of the subimages to zero . This renders image transformations non-unitary and allows to introduce further non-linear evalution branches within the tree-like structure of the system . After the convolution and rectification stage , the amount of data is fairly increased and usually corresponds to a multiple of the input data . Thus , it is reasonable to reduce data with the aid of pooling layers [21 , 24] . Within such layer , subsets of each intermediate image are pooled according to their mean or maximum values . Results are then assigned to a target image of smaller size . For our CNN , subsets of size 2 × 2px2 with a stride of 2px are evaluated according to the maximum value aiming to halve the sizes of intermediate images at the end of each convolution stage . A further advantage of such pooling strategy is to obtain certain tolerance with respect to spatial translations of objects . As special feature , the here employed CNN makes use of a regression output layer providing a linear transfer function . In contrast to standard classification networks offering a purely binary output , realized e . g . by a logistic transfer function or softmax approach [25] , the output node of our CNN is able to take on a range of floating point values . Taking the number space of input images into account ( 8-bit space ) , we define our output to move in the same range thus the same order of magnitude . Taking values at this scale , we observe a fast convergence of the training process . Consequently , perfect slippers are defined at value −127 whereas croissants are located at 127 . Input images leading to an output value which significantly differs from these targets are assumed to be indefinite and therefore discriminated to be an outlier of type other ( cf . section Training ) . CNNs are not ready to use if neurons are not ‘trained’ to solve certain regression or classification problems . This means that initially randomly chosen convolution kernels , weights and bias terms ( concluded in the vector σ of free parameters ) first need to be reasonably updated . For this purpose , we employ a supervised learning approach which requires three major ingredients: A labeled set of training data , a suitable loss function that has to be minimized , as well as an appropriate optimization strategy . As loss function , the root mean square error ( RMSE ) is chosen ( Eq 1 ) , expressing the cumulated differences between all target and actual output values of the training data set . RMSE = ( 1 N ∑ i = 0 N - 1 ( e i ) 2 ) 1 / 2 = ( 1 N ∑ i = 0 N - 1 ( t i - a i ) 2 ) 1 / 2 ( 1 ) Here , ai expresses the actual output of a given input image i and ti the respective , predefined target value . Finding the global minimum of the RMSE , σ is optimized using a stochastic gradient descent solver with momentum ( SGDM ) , see [26] and [27] . The SGDM algorithm updates the vector σl in discrete steps , where l denotes the actual iteration: σ l + 1 = σ l - α ∇ E ( σ l ) + γ ( σ l + 1 - σ l ) . ( 2 ) E ( σl ) corresponds to the loss function , respectively the RMSE , α = 0 . 001 to the constant learning rate , and γ = 0 . 9 to the momentum term . Achieving a faster convergence , a single optimization iteration takes into account a so-called mini-batch [25 , 28 , 29] consisting of a subset ( 128 images ) of the randomly shuffled training data . In this context , a full pass of the training data is called an ‘epoch’ . A typical temporal evolution of training states is shown in Fig 3 where the loss function is plotted with respect to the number of training epochs . We set a maximum of ten training epochs , since a prolongation to more training epochs rather causes overtraining instead of a gain in performance . This overtraining of the neural network can be monitored in the validation loss: A termination criterion of the training status is implemented by a consecutive increase in validation loss five times , since a divergence between training and validation loss is an indicator for not being in the global optimum . Further , we choose a validation frequency ( update frequency of validation loss ) of 50 mini-batches , to not meet the termination criterion by random fluctuations of the training process . In general , the validation data set consists of data disjoint from the ones for training to guarantee an independent validation of the networks’ performance . For our purposes , 5% of the training data set was chosen randomly for validation , whereas the rest was the true training data . For training , besides slippers and croissants , we additionally define an auxiliary class called ‘sheared croissant’ ( Fig 4 ) . This is due to the fact that the characteristic features of pure croissants and sheared croissants show a larger mutual similarity than pure croissants and slippers . Assuming a linear scale of output values , this subtype is biased towards a value of 64 . Contrasting the nomenclature , they are not any type of croissants , but most probably resemble a slipper flipped by 90° perpendicular to the optical axis . However , due to the lack of information of our 2D images , we cannot make a clear statement on the actual shape . Our training data set consists of 4 , 000 manually classified cells ( 1 , 500 each for slippers , and croissants , resp . and 1 , 000 sheared croissants ) , which is augmented to the doubled number ( 8 , 000 in total ) by mirroring each image along the centerline in flow direction . Each of those 4 , 000 initial training cell images represents one distinct cell , i . e . no cell was considered more than once in the training data set . We intentionally have fewer sheared croissants in the training data set than croissants and slippers since it is only an auxiliary cell class not reflected in the phase diagram but only to increase the precision of croissant classification . The three subsets ( one per cell class ) were taken at different pressure drops . Slippers were recorded at 700–1 , 000mbar , croissants at 100–200mbar and sheared croissants at 300–500mbar . Within these ranges , we applied several different pressure drops to ensure a certain variability of our training data set . All cells of a class are then mixed since they should be identified independent from the applied pressure . Further , we intentionally recorded cells for different lighting conditions by varying the power of the light source randomly ( within a range of 50–80% ) . Together with a bilinear contrast adjustment of the resulting images ( cf . experimental setup ) , this ensures high contrast dynamics of each image and corrects for minor changes in illumination . Due to constant manual observation during the recording process of the input data , we ensure collecting focused cell images . However , to ensure an invariance of our CNN approach with respect to optical misalignment and to reach a robust algorithm in terms of optical imaging , we also train for defocused cells by recording training images at different positions within a slightly tilted channel . The outcome will then be a neural network with a self-contained training set and thus is perfectly suited for a non-biased analysis . For our measurements , dilute suspensions of RBCs are required to observe single cells flowing through microcapillaries . Therefore , capillary blood is drawn with consent from a healthy individual and resuspended in a buffer solution ( phosphate buffered saline , PBS , gibco by life technologies ) . After centrifugation at 1 , 500 g for 5 minutes , RBCs sediment on the bottom of the sample tube , whereas leukocytes and platelets form a buffy coat on top of these ( for further details concerning the blood sampling , see [30] ) . Since the presence of platelets can alter the dynamic properties of RBCs due to shear-induced platelet activation ( SIPA , see [31] ) , this buffy coat is removed together with the supernatant and the residual pellet of RBCs is again resuspended in PBS . This procedure is repeated three times until the final pellet of cells will be adjusted to a hematocrit of Ht ≤ 1% in a base solution of both PBS and bovine albumin ( BSA , Sigma-Aldrich ) at a concentration of 1 mg/ml , to avoid the well-known glass slide effect , turning discocytous RBCs to echinocytes . With the aid of a high-precision pressure device ( Elveflow OB 1 Mk II ) , this highly diluted suspension of RBCs is then driven through microcapillaries for a discrete set of 12 pressure drops Δp ∈ {20 , 50 , 100 , 200 , 300 , 400 , 500 , 600 , 700 , 800 , 900 , 1000}mbar , yielding various flow velocities , and recorded with a high-speed camera at a framerate of 400 Hz . Over the full pressure range , we record in total 3 , 090 RBCs and obtain experimental data , independent of the training data set . The microcapillaries are formed in PDMS ( polydimethylsiloxane ) , have a length of Lx = 4 cm and a rectangular cross-section with dimensions Ly = 11 . 9 ± 0 . 3 μm and Lz = 9 . 7 ± 0 . 3 μm . The y-direction is perpendicular to the camera axis , whereas the z-direction points inward the camera axis . Several channels are branched from a common reservoir to the fluid inlet , which is connected to the pressure device via flexible tubing . Post-processing the recorded media involves single particle tracking of cells , including distinction and sorting out non-isolated RBCs ( i . e . RBCs being too close to each other , such that hydrodynamic interactions are not negligible ) . Due to the used optical setup ( Nikon CFI Plan Fluor 60× oil-immersion objective , NA = 1 . 25 ) camera ( Fastec HiSpec 2G ) and capture settings , typical cell dimensions are in the range of 80 px . However , to allow for minor ( physiological ) variations , we crop the individual cell images to a format of 90 × 90 px2 . Since the channel width is smaller than 90 px , we apply a Tukey window w ( y ) ( 3 ) to the images , causing a smooth fade-out towards the channel walls w ( y ) = { 1 2 [ 1 + cos ( 2 π α ( y - α 2 ) ) ] , 0 ≤ y < α 2 1 , α 2 ≤ y < 1 - α 2 1 2 [ 1 + cos ( 2 π α ( y - 1 + α 2 ) ) ] , 1 - α 2 ≤ y ≤ 1 , ( 3 ) where y denotes the relative position ( 0 ≤ y ≤ 1 ) in vertical direction of an arbitrary cell image . This reduces the influence of markable edges probably influencing the training and output of the CNN ( Fig 2 ) . For image preparation , we additionally map the image intensities to the full 8-bit range , such that the bottom 1% and the top 1% of all pixel values are saturated . This transformation yields higher signal dynamics and equal intensity profiles of all cell pictures and renders our approach more stable regarding slight illumination variations . Using artificial intelligence for classification issues is a well-known , yet rare approach in biological systems . Although a wide variety of accessible , pre-trained , highly sophisticated neural networks ( e . g . AlexNet , ResNet ) exist , they tend to be over-engineered for most purposes in this field . Due to the complexity of their architecture with respect to the degrees of freedom , they require millions of input images . Furthermore , they are not designed for regression problems as required for the present study , but rather have a binary classification output . The presented CNN is a step forward to a fully automated analysis of recorded microscopy images associated with a big gain in evaluation efficiency . It moreover constitutes an unbiased classification system for RBC shapes in flow . Any restriction to cell classes is purely artificial in the sense that we train the CNN solely with RBC shapes in flow . Similarly , one could tailor a set of training data for other distinct cell classes or even cell types . We aim to conduct further experiments with different channel geometries and flow conditions , amending the existing phase diagram . The phase diagram and especially the phase transition point between croissants and slippers can be altered adding drugs to the RBC suspension . Since certain drugs , e . g . acetylsalicylic acid , show a strong effect on the membrane structure [32] , we conjecture to resolve this feature in phase diagrams of future studies , leading to insights towards mechanical properties of individual RBCs in flow . Following this scope , the change in flow behavior of whole blood under drug influence can be predicted due to the influence on its main constituent , the red blood cells . This might play a key role in finding appropriate drugs to avoid pathogenic incidents , e . g . stenosis . Analogously , we clearly see evidence to detect maladies causing an alteration of cell’s intrinsic mechanical properties , e . g . in sickle cell disease , thalassemias and numerous rare anaemias . By slight technical adaptations ( e . g . the usage of high-performance graphics cards ) , it is also possible to classify cells on-the-fly , i . e . in real-time rather than with the presented frame-based approach . Prospectively , the here proposed CNN will be a useful quantitative tool in hematology aiming to investigate cell membrane characteristics .	Artificial neural networks represent a state-of-the art technique in many branches of natural sciences due to their ability to fastly detect and categorize image features with high throughput . We use a special type of neural network , the so-called convolutional neural network ( CNN ) for the classification of human red blood cell shapes in microcapillary Poiseuille flow . Following this approach , phase diagrams of two distinct classes ( slippers , croissants ) are generated and , by comparison with a manually obtained phase diagram , optimized threshold ranges for categorizing the output values are established . This allows us to better understand the complex fluid behavior of blood depending on the intrinsic properties of single red blood cells . For future studies , we aim to predict phase diagrams under the influence of certain drugs .	[ "Abstract", "Introduction", "Materials", "and", "methods", "Results", "and", "discussion" ]	[ "blood", "cells", "medicine", "and", "health", "sciences", "body", "fluids", "neural", "networks", "engineering", "and", "technology", "signal", "processing", "neuroscience", "artificial", "neural", "networks", "red", "blood", "cells", "phase", "diagrams", "artificial...	2018	Classification of red blood cell shapes in flow using outlier tolerant machine learning
Bacterial effector proteins secreted into host plant cells manipulate those cells to the benefit of the pathogen , but effector-triggered immunity ( ETI ) occurs when effectors are recognized by host resistance proteins . The RPS4/RRS1 pair recognizes the Pseudomonas syringae pv . pisi effector AvrRps4 . AvrRps4 is processed in planta into AvrRps4N ( 133 amino acids ) , homologous to the N-termini of other effectors including the native P . syringae pv . tomato strain DC3000 effector HopK1 , and AvrRps4C ( 88 amino acids ) . Previous data suggested that AvrRps4C alone is necessary and sufficient for resistance when overexpressed in heterologous systems . We show that delivering AvrRps4C from DC3000 , but not from a DC3000 hopK1- strain , triggers resistance in the Arabidopsis accession Col-0 . Delivering AvrRps4C in tandem with AvrRps4N , or as a chimera with HopK1N , fully complements AvrRps4-triggered immunity . AvrRps4N in the absence of AvrRps4C enhances virulence in Col-0 . In addition , AvrRps4N triggers a hypersensitive response in lettuce that is attenuated by coexpression of AvrRps4C , further supporting the role of AvrRps4N as a bona fide effector domain . Based on these results we propose that evolutionarily , fusion of AvrRps4C to AvrRps4N may have counteracted recognition of AvrRps4N , and that the plant RPS4/RRS1 resistance gene pair was selected as a countermeasure . We conclude that AvrRps4 represents an unusual chimeric effector , with recognition in Arabidopsis by RPS4/RRS1 requiring the presence of both processed effector moieties . Plants deploy a multilayered immune system to defend against invading pathogens [1–3] . The first layer of defense , pathogen-associated molecular pattern-triggered immunity ( PAMP-triggered immunity ) , relies on membrane-localized pattern recognition receptors which recognize conserved non-self molecules from pathogens [4 , 5] . Detection of these molecules rapidly activates signaling cascades culminating in disease resistance . Effector molecules are secreted from pathogens into the host cell , often blocking PAMP-triggered immune responses or reprograming host cell transcription and physiology , leading to successful colonization of the host plant by the pathogen [6 , 7] . To counteract this , plants have evolved resistance proteins which can either directly or indirectly detect the presence of these effectors and then ramp up a robust immune response called effector-triggered immunity ( ETI ) [8] . The effector AvrRps4 , originally identified in Pseudomonas syringae pv . pisi [9] , has currently uncharacterized biochemical activity in host cells . Full-length AvrRps4 is 221 amino acids in length and is processed within host cells between Gly133 and Gly134 [10] . The processed N-terminal AvrRps4 fragment ( AvrRps4N , amino acids 1–133 ) shares 75% identity with the N-terminus of HopK1 , a native effector of the Pseudomonas syringae pv . tomato strain DC3000 ( DC3000 ) that is predominantly used for bacterial delivery of effectors in Arabidopsis studies . HopK1 is a 338 amino acid protein that contributes significantly to the virulence of DC3000 . Similarly to AvrRps4 , HopK1 is processed in plant cells between amino acids Gly133 and Gly134 [11] . AvrRps4C ( amino acids 134–221 ) is a coiled-coil protein which interacts with WRKY proteins , a plant-specific class of transcription factors of which several have been implicated in defense against pathogens [12–14] . The resistance protein RESISTANCE TO RALSTONIA SOLANACEARUM1 ( RRS1 ) contains an integrated WRKY domain at the C-terminus [15] . RRS1 works as a pair with another resistance protein , RESISTANCE TO PSEUDOMONAS SYRINGAE4 ( RPS4 ) , to trigger defense against pathogens secreting AvrRps4 [16 , 17] . RRS1 and RPS4 belong to the Toll/Interleukin-1 Receptor—Nucleotide-Binding—Leucine-Rich Repeat ( TNL ) class of resistance proteins [15 , 18] and , apart from AvrRps4 , also detect the Ralstonia solanacearum effector PopP2 and an unknown Colletotrichum higginsianum effector [16 , 17] . As a paradigm for the integrated decoy model [19] and TNL signaling , it is of paramount interest to establish how RPS4/RRS1 are activated by the presence of these unrelated effectors . PopP2 has been shown to acetylate host proteins , including RRS1 in its WRKY domain at amino acid positions that mediate contact with DNA . These findings led to the model that PopP2 evolved to target host WRKYs important in resistance but are baited by the decoy WRKY in RRS1 , activating RPS4 and subsequent defense [12 , 13 , 20] . Previously it was reported that AvrRps4C was sufficient to trigger a hypersensitive response ( HR ) in turnip [10] , and that AvrRps4C interacts , directly or indirectly , with the WRKY domain of RRS1 as well [13] . Additional components of RPS4/RRS1 activation by AvrRps4 are its interaction in the cytoplasm and nucleus with ENHANCED DISEASE SUSCEPTIBILITY1 ( EDS1 ) , a positive regulator of both basal resistance and ETI , and the interaction between EDS1 and RPS4 [21 , 22] . EDS1 is required for ETI mediated by the TNL class of resistance proteins [23 , 24] . This may suggest that RPS4/RRS1 guard EDS1 , but the connection between the two models of AvrRps4 recognition is not understood . It was recently confirmed that AvrRps4 does indeed interact with EDS1 even though it was controversial in previous reports [25] , but gaps in knowledge for how this interaction connects to RPS4/RRS1 activation remain . One complication in determining a mechanism is the fact that these resistance protein complexes are dynamic and sensitive to shifts in protein amounts . A careful study demonstrated that the composition and localization of RPS4/RRS1 complexes in transient expression studies vary with the absence or presence of other expressed proteins , specifically EDS1 and PAD4 [25] . The complexity of the system becomes evident when considering that this study did not include other well-known protein partners such as the co-chaperone SGT1 and the immune adaptor protein SRFR1 , and like previous studies from several groups relied on static protein overexpression . Specifically , SRFR1 does not localize to the soluble cytoplasmic fraction in N . benthamiana , or in Arabidopsis when expressed from its native promoter [21 , 26] . This may be important in localizing RPS4 [27] and a sub-pool of EDS1 [21] to the microsomal cytoplasmic fraction , where disruption of protein interactions by AvrRps4 was reported [21] . Here we show that AvrRps4N directly interacts with EDS1 , suggesting that AvrRps4N possesses effector functions outside of it being a signal peptide for secretion and a transit peptide for chloroplast localization . In addition , we endeavored to determine the molecular basis of AvrRps4 triggering RPS4/RRS1 by introducing bacterially delivered AvrRps4 to ascertain function of separate domains of this effector . We provide evidence that HopK1N and AvrRps4N at natural expression levels redundantly contribute to RPS4/RRS1-mediated immunity in the presence of AvrRps4C . Consequently , both AvrRps4C and either HopK1N or AvrRps4N are required for the full ETI-inducing activity of AvrRps4 in Arabidopsis . Additionally , AvrRps4N increases bacterial virulence when overexpressed in Col-0 , and triggers a robust HR when expressed in cultivars of Lactuca sativa ( lettuce ) while AvrRps4C triggers no such response . We propose that separate interactions of AvrRps4N with EDS1 and of AvrRps4C with RRS1 are components of RPS4/RRS1 activation . We previously showed that N-terminally tagged full-length AvrRps4 interacts with EDS1 in vivo and in vitro , indicating that the interaction is direct and possibly mediated by AvrRps4N [21] . To specifically investigate which AvrRps4 fragment binds to EDS1 we repeated these experiments with N-terminally tagged AvrRps4N . Consistent with previous biochemical fractionations of HA-AvrRps4-expressing cells [21] , apart from a nuclear pool AvrRps4N transiently expressed in Nicotiana benthamiana was detected in the cytoplasmic microsomal fraction . In vivo co-immunoprecipitation of AvrRps4N and EDS1 in the microsomal fraction confirmed that AvrRps4N is sufficient for this interaction ( Fig 1A , S1A Fig ) . Association of AvrRps4C and EDS1 was also observed . Additionally , AvrRps4N and AvrRps4C specifically associated with EDS1 in vitro ( Fig 1B , S1B Fig ) . Based on the relative ratios of input and pulldown protein amounts , the interaction of EDS1 with AvrRps4N appeared to be stronger than with AvrRps4C . This interaction implies a previously uncharacterized function of AvrRps4N in bacterial virulence or avirulence in addition to type III secretion and chloroplast localization [11] , since effectors commonly target key hubs of resistance [28] . To identify possible roles of AvrRps4N in virulence or triggering resistance at natural expression levels we utilized the vector pVSP_PsSPdes [29] , which enables secretion from Pseudomonas syringae into host cells of parts of cloned effectors such as AvrRps4C ( S2 Fig ) . This vector includes an AvrRpm1 signal peptide for protein secretion and an HA epitope tag for protein detection . Three additional constructs , each with the AvrRpm1 signal peptide removed , were also constructed . All constructs were driven by the AvrRpm1 promoter . As expected , full-length AvrRps4 with or without the AvrRpm1 signal peptide triggered resistance to DC3000 ( S3 Fig ) . AvrRps4N and AvrRps4C with or without the AvrRpm1 signal peptide failed to restore the resistance phenotype of full-length AvrRps4 . AvrRps4C without the AvrRpm1 signal peptide is not secreted from DC3000 and served as a negative control . In contrast , the negative result with AvrRpm1SP-AvrRps4C superficially did not support the hypothesis based on turnip HR that AvrRps4C is sufficient for resistance [10] . To further investigate why AvrRps4C delivered into cells via the AvrRpm1 signal peptide did not trigger resistance we explored the possibility of protein mislocalization . Previously , it was shown that AvrRpm1 is acylated at its second amino acid residue , Gly , leading to AvrRpm1 localization to the host cell plasma membrane [30] . Host membrane tethering of the protein was disrupted by mutating Gly to Ala . To test for mislocalization of our AvrRps4 proteins we made C-terminal translational GFP fusions of full-length AvrRps4 , AvrRps4N , and AvrRps4C with or without the AvrRpm1 signal peptide . AvrRps4 transiently expressed in N . benthamiana was shown not to be imported into chloroplasts [11] , and AvrRpsN-GFP and AvrRps4C-GFP localized to the nucleus in addition to the cytoplasm ( Fig 2 ) . Therefore , localization of GFP signal within the nucleus and at the plasma membrane can be used as an assay for extrachloroplastic processing of full-length AvrRps4 and mislocalization of AvrRpm1SP-AvrRps4N/C proteins , respectively . Indeed , the AvrRpm1 signal peptide disrupted the localization of AvrRps4 when transiently expressed in N . benthamiana cells ( Fig 2 ) . Whereas nuclear signals were observed with full-length AvrRps4-GFP , AvrRps4N-GFP , and AvrRps4C-GFP , these signals were abolished in the AvrRpm1 signal peptide fusions of AvrRps4N and AvrRps4C . However , a nuclear signal was still detected in AvrRpm1 signal peptide fusion with full-length AvrRps4 , indicating that processed AvrRps4C-GFP was liberated and gained entry into the nucleus . Coexpression of GFP-tagged AvrRps4 proteins with free RFP to mark the cytoplasm and nucleus confirmed the presence ( AvrRps4-GFP , AvrRps4N-GFP , AvrRps4C-GFP , and AvrRpm1SP-AvrRps4-GFP ) and absence ( AvrRpm1SP-AvrRps4N and AvrRpm1SP-AvrRps4C ) of nuclear signal ( S4 Fig ) . Assays with delivery by DC3000 confirmed that AvrRpm1SP-AvrRps4 is still processed in Arabidopsis cells ( see below ) . AvrRps4N is 75% identical to HopK1N , the N-terminus of the native DC3000 effector HopK1 . Both AvrRps4 and HopK1 are processed at the same amino acid residue [11] . This remarkable conservation led us to test whether HopK1N and AvrRps4N could redundantly trigger immunity in the presence of AvrRps4C . Chimeras of HopK1N-AvrRps4C and AvrRps4N-HopK1C were constructed . For this series of experiments , we used the Arabidopsis accession Wassilewskija ( Ws-0 ) , which like Col-0 possesses functional RPS4 and RRS1 for AvrRps4 recognition [9 , 16 , 17] . In contrast to Col-0 , which displays resistance to AvrRps4 uncoupled from HR , Ws-0 responds to AvrRps4 with a strong HR , thus enabling the use of non-pathogenic strains for HR assays in the absence of additional effectors [18 , 31] . When delivered via the Pseudomonas fluorescens Effector-to-Host Analyzer system ( Pfo EtHAn ) [32] the AvrRps4 full-length protein triggered a strong HR in Ws-0 after 24 hours ( Fig 3A ) . HopK1 delivered from Pfo EtHAn did not trigger an HR . Interestingly , a HopK1N-AvrRps4C chimera indeed caused a strong HR , whereas an AvrRps4N-HopK1C chimera did not . To confirm these results , in planta bacterial growth assays were performed using chimeras delivered from a DC3000 hopK1- mutant into Ws-0 . The HopK1N-AvrRps4C chimera triggered resistance similar to full-length AvrRps4 , whereas AvrRps4N-HopK1C supported bacterial growth similar to empty vector and HopK1 ( Fig 3B ) . Additionally , resistance in Ws-0 to chimeric HopK1N-AvrRps4C required wild-type copies of EDS1 and RPS4 to the same extent as resistance to AvrRps4 ( S5 Fig ) . Taken together , these results indicate that AvrRps4N and HopK1N function redundantly in triggering ETI in the presence of AvrRps4C in the Arabidopsis system . Since HopK1 is an effector native to DC3000 , we hypothesized that processed HopK1N may substitute for mislocalized AvrRpm1SP-AvrRps4N delivered from DC3000 in bacterial growth assays . To address this question , we introduced all of our AvrRps4 constructs into DC3000 hopK1- [11] . Surprisingly , whereas the AvrRpm1SP-AvrRps4 full-length protein delivered from DC3000 into Arabidopsis Col-0 triggered resistance similar to wild-type AvrRps4 , the same construct delivered from DC3000 hopK1- failed to reconstitute this strong resistance phenotype ( Fig 4A ) . Protein secretion assays confirmed that AvrRps4 proteins with or without AvrRpm1SP were still being secreted from DC3000 and DC3000 hopK1- ( S6 Fig ) . After harvesting tissue from Col-0 leaves infiltrated with AvrRpm1SP-AvrRps4 from DC3000 hopK1- we could detect both the full-length and processed forms of this fusion protein ( S7 Fig ) , consistent with results obtained by transient expression in N . benthamiana ( Fig 2 ) . Together with our localization data , this shows that although AvrRpm1SP-AvrRps4N mislocalizes to the host membrane , AvrRps4C is still liberated after processing when full-length AvrRps4 is delivered into host cells via the AvrRpm1 signal peptide . Consequently , AvrRpm1SP-AvrRps4 should limit DC3000 hopK1- multiplication in Col-0 leaves if AvrRps4C were sufficient for triggering resistance . In individual replicate experiments AvrRpm1SP-AvrRps4 yielded partial resistance when compared to our empty vector controls , even though in aggregate this difference was not statistically significant ( Fig 4A ) . This suggests that AvrRps4C alone may trigger a partial resistance response , or that the AvrRps4N pool is not completely mislocalized under certain conditions . Taken together , these results show that the addition of HopK1N or AvrRps4N in conjunction with AvrRps4C enhances any resistance triggered by AvrRps4C alone ( see model Fig 4B ) . To test whether this enhanced resistance could be recapitulated in the bacterial system we designed a dual-vector system where DC3000 hopK1- mutants were made to secrete effectors from two different broad host range plasmids , pVSP61 or pML123 , thus ensuring that any cell in contact with this strain would receive both proteins via bacterial delivery . When AvrRps4N was delivered from pML123 in tandem with AvrRpm1SP-AvrRps4 from pVSP61 in DC3000 hopK1- , the resistance phenotype was similar if not stronger to that of full-length AvrRps4 ( Fig 5 ) . These findings further confirm that AvrRps4N and AvrRps4C together function in triggering a full ETI response . AvrRps4N interacted with EDS1 and was required for AvrRps4-triggered immunity . Additionally , in some individual experiments bacterial growth was enhanced by AvrRps4N , even though the difference was not statistically significant in aggregate ( Fig 5 ) . We therefore asked whether AvrRps4N in the absence of AvrRps4C has a detectable virulence activity by generating transgenic Arabidopsis lines overexpressing AvrRps4N from a Dex-inducible promoter . Bacterial growth of DC3000 hopK1- was approximately ten times higher in Dex-treated line 2 ( N2 ) , and five times higher in mock- or Dex-treated line 8 ( N8 ) than in mock- or Dex-treated Col-0 , whereas the growth of wild-type DC3000 was similar in all plants tested ( Fig 6 ) . This indicates that the virulence function of AvrRps4N in DC3000 is redundant to that of native HopK1N . AvrRps4N protein and transcripts were increased in both N2 and N8 within 24 hours after Dex treatment ( S8 Fig ) . Consistent with the bacterial growth curve assays , avrRps4N expression was leakier in N8 than in N2 , and expression of the defense marker gene PR1 was inversely correlated with avrRps4N mRNA levels ( S8A Fig ) . These results strongly suggest that AvrRps4N contributes to bacterial virulence in DC3000 lacking HopK1 , further supporting that AvrRps4 is a bipartite effector . It was previously shown that full-length AvrRps4 and HopK1 expressed in several lettuce cultivars resulted in an HR [33] . Interestingly , the pattern of the variable response between cultivars was identical between AvrRps4 and HopK1 , leading us to hypothesize that the N-terminal moiety of each respective effector was sufficient for such a response . To test this , we first confirmed that AvrRps4 and HopK1 could reproducibly trigger HR on Lactuca sativa cv . Kordaat . We then expressed AvrRps4N , HopK1N , AvrRps4C , and HopK1C and monitored cell death ( Fig 7 ) . Indeed , AvrRps4N and HopK1N triggered a strong HR whereas neither AvrRps4C nor HopK1C induced HR . We consistently observed that full-length AvrRps4 and full-length HopK1 induced weaker cell death phenotypes when compared to AvrRps4N or HopK1N , suggesting the C-terminal peptides may suppress HR induced by the N-terminal peptides . Importantly , AvrRps4C co-expressed with AvrRps4N attenuated the HR caused by AvrRps4N yielding HR phenotypes similar to those caused by AvrRps4 full-length ( Fig 8A ) . To quantify the level of HR we performed electrolyte leakage experiments . These showed that AvrRps4N induced a faster and stronger HR than any other treatment ( Fig 8B ) . Western blot analysis confirmed that AvrRps4N , AvrRps4C as well as full-length AvrRps4 were expressed in L . sativa , even though levels of AvrRps4C were variable when co-expressed with AvrRps4N ( Fig 8C ) . We have demonstrated that AvrRps4 is a bipartite protein , where AvrRps4N functions in three distinct ways , 1 ) triggering of resistance in the presence of AvrRps4C in Arabidopsis Col-0 , 2 ) serving as a virulence factor in the absence of AvrRps4C in Col-0 , and 3 ) triggering defense responses in lettuce . These functions of AvrRps4N are consistent with functions of an effector and with its interaction with the positive regulator of immunity EDS1 shown here . Sequence conservation across the entire N-terminal moieties of AvrRps4 , HopK1 , and the Xanthomonas campestris pv . vesicatoria effector XopO led us to hypothesize that AvrRps4N has functions that extend beyond a T3SS signal and chloroplast transit peptide , which usually are encoded in the first 20–50 amino acids of a peptide [34] . Indeed , the putative effector HopAQ1 is a relatively small 83 amino acid protein from DC3000 with a 44 amino acid signal peptide highly similar to that of AvrRps4 and HopK1 and a dissimilar sequence in its uncleaved C-terminal 39 amino acids [35 , 36] . While bacterial effectors with more than one distinct functional domain , such as AvrPtoB [37] , have been described before , AvrRps4 so far appears unique in that two effector moieties are required to trigger resistance in Arabidopsis when present at native levels . Both AvrRps4 and HopK1 are processed in plant cells between amino acids Gly133 and Gly134 . It was suggested that this processing occurs exclusively in chloroplasts and that the N-terminal 20 amino acids encode a chloroplast transit peptide [11] . HopAQ1 lacks the proposed C-terminal effector domain of AvrRps4N and is predicted to translocate to chloroplasts , although such predictions may be an artifact of prediction models and the similarity of T3SS and chloroplast targeting sequences [38] . We observed processing of AvrRps4 in N . benthamiana and Arabidopsis even with N-terminal fusions such as those to AvrRpm1SP . Since such fusions are predicted to block chloroplast import we propose that a subpool of AvrRps4 is processed in the cytoplasm , and that AvrRps4 has multiple virulence targets in multiple compartments . Based on our results it appears that AvrRps4-triggered ETI is initiated in the cytoplasm or nucleus . This does not exclude virulence functions of chloroplast-localized AvrRps4 [11] . Consistent with results presented here , a recent study also failed to detect chloroplast localization of full-length or processed AvrRps4 in Arabidopsis or N . benthamiana cells [39] . Interestingly , in this study bacterial delivery of AvrRpm1SP-AvrRps4C triggered measurable resistance in Col-0 . To detect bacterial delivery and effector localization in the host cell , effector constructs were tagged with a 13-amino acid split fluorescent protein sequence and were delivered into transgenic plants expressing a split superfolder GFP protein that self-assembles . The strain used for effector delivery , the DC3000 variant Pst CUCPB5500 , expresses HopK1 [40] . Interestingly , capture of tagged AvrRpm1SP-AvrRps4C by the highly soluble superfolder GFP allowed diffusion of this construct into the nucleus [39] . These results are therefore consistent with our model that untethered AvrRps4C in the presence of HopK1N or AvrRps4N triggers resistance . Searches of NCBI’s protein sequence database using the BLAST algorithm [41] failed to identify examples of putative effectors with homology to AvrRps4C fused to only a 50 amino acid signal peptide , but Xanthomonas XopAK is a homolog of HopK1C fused to a signal peptide lacking the 94 amino acid effector domain of AvrRps4N and HopK1N [42] . We therefore propose a model where ancestral AvrRps4 and HopK1 proteins consisted of only AvrRps4N or HopK1N homologs , which presumably targeted hubs important to immunity such as EDS1 . The interactions of these proteins with EDS1 may have triggered resistance on resistant plants , and successful bacteria evaded this resistance by terminal reassortment [36] with either AvrRps4C or HopK1C to suppress resistance . On the plant side , genes such as the RPS4/RRS1 pair arose to recognize the addition of AvrRps4C . Evidence for this include our data which show that AvrRps4N alone is sufficient to enhance bacterial virulence in Arabidopsis Col-0 when overexpressed whereas addition of AvrRps4C again makes Col-0 fully resistant to bacteria expressing AvrRps4N and AvrRps4C . The virulence function of AvrRps4N was occasionally discernible in in planta bacterial growth experiments using DC3000 hopK1- but was not significant overall , as was reported previously [11] . Low levels of bacterial AvrRps4N delivery and functional redundancy , as observed with several groups of sequence-unrelated DC3000 effectors [40] , likely precluded robust detection of AvrRps4N virulence effects at natural protein levels . A prediction of this model is that the capacity to recognize AvrRps4N/HopK1N still exists in some plants . Indeed , we identified lettuce as one plant species which retains the ability to recognize AvrRps4N . Future work includes a large-scale screening of Arabidopsis accessions to identify those which can recognize AvrRps4N . In addition , it would be informative to determine whether AvrRps4C also counteracts the virulence function of AvrRps4N in an rps4/rrs1 background , or only attenuates the avirulence function of AvrRps4N . Our finding that AvrRps4N functions as an effector within plant cells suggests a possible contribution of this moiety when assaying candidate non-bacterial effector molecules by delivery with the Effector Detector Vector 3 ( pEDV3 ) , which enables bacterial secretion of effectors into host cells using the first 136 amino acids of AvrRps4 [43] . However , the functions of AvrRps4N encoded in pEDV3 should not interfere with assays using Pseudomonas pathogens which contain HopK1 or AvrRps4 natively . An outstanding question is how AvrRps4N is involved in resistance-triggering in addition to serving as a virulence factor . We previously proposed that EDS1 has roles both upstream and downstream of resistance protein activation [21] . Interactions of AvrRps4N with EDS1 as a guardee may constitute the first step of RPS4/RRS1 activation but is insufficient to trigger immunity in the absence of concurrent RRS1 activation by AvrRps4C . Alternatively , AvrRps4N through interactions with EDS1 boosts EDS1 immune signaling in the presence of AvrRps4C-activated RPS4/RRS1 . Based on the virulence function of AvrRps4N we favor the first model [44] . While additional work will be required to distinguish between these models , our data presented here confirm the interaction between AvrRps4 and EDS1 and clarify the components required for AvrRps4-triggered immunity in Arabidopsis . This is essential for developing an accurate model of RPS4/RRS1 and EDS1 activity in AvrRps4 recognition . For in vitro pull-down assays , GST- EV , AvrRps4N , AvrRps4C , and AvrRps4FL were cloned into pDEST15 via Gateway recombination as previously described [21] . EDS1 was cloned into pET28a by inserting amplified EDS1 into SalI/XhoI using restriction digestion ( for all cloning primers , see S1 Table ) . Full-length avrRps4 , avrRps4N , and avrRps4C were cloned into pVSP_PsSPdes [29] using Gateway recombination . Constructs with the AvrRpm1 signal peptide deleted were made by amplifying the avrRpm1 promoter with N-terminal EcoRI restriction site and avrRps4 full-length , avrRps4N , and avrRps4C with C-terminal HA tags and HindIII sites , and by joining promoter and avrRps4 fragments with overlap-PCR . The joined clones were then ligated into pVSP_PsSPdes using EcoRI and HindIII . Translational fusions of avrRps4 full-length , avrRps4N , and avrRps4C to GFP were made by Gateway recombination into the pMDC43 plasmid . Constructs with the AvrRpm1 signal peptide added upstream of AvrRps4 were constructed by digesting AvrRps4-GFP pMDC43 with the PacI enzyme and ligating in the AvrRpm1 signal peptide . pTA7002 vectors were used for generating transgenic Arabidopsis lines expressing Dex-inducible AvrRps4N . A single Myc tag was introduced into pTA7002 by amplifying Myc with a 5’ XhoI and 3’ XbaI site . A T4 Polynucleotide Kinase reaction was performed and the insert was ligated into XhoI/SpeI-digested pTA7002 . AvrRps4N was cloned into pTA7002-Myc using restriction cloning . For in planta CoIPs , Agrobacterium tumefaciens strains harboring Myc-EDS1 ( strain GV3101 ) and HA-AvrRps4 N-terminus or HA-AvrRps4 C-terminus ( strain C58C1 ) were infiltrated into N . benthamiana leaves at 0 . 05 OD . After 48 hours 1 g of tissue was collected and ground in 2 ml buffer H ( 50 mM HEPES ph7 . 5 , 250 mM sucrose , 15 mM EDTA , 5% glycerol , 3 mM DTT , 0 . 5% PVPP , 1x Sigma protease inhibitor cocktail ) . Lysate was centrifuged at 1600 x g for 20 minutes . Supernatant was then spun at 100 , 000 x g for 1 hour to pellet microsomes . Pellets were resuspended in 1 ml buffer H with 1% NP-40 . Protein concentration was measured by Bradford assay , and samples were adjusted to 0 . 25 mg/ml protein . 20 μl of HA-conjugated beads ( Sigma ) were added to each sample and incubated at 4°C for 1 hour . Beads were washed three times with buffer H with 0 . 2% NP-40 . For in vitro pull-down assays , BL21 ( DE3 ) competent cells were transformed with GST- EV , AvrRps4N , AvrRps4C , AvrRps4FL or His-T7-EDS1 individually ( GST purification ) or with His-T7-EDS1 and individual AvrRps4 constructs together ( nickel purification ) . Overnight cultures from fresh colonies were used to inoculate 200 mL LB , incubating 3–4 hours at 37°C under selection . Expression was induced with 0 . 2 mM IPTG overnight at 22°C . Cultures were centrifuged 10 min at 3800 x g and resuspended in 20 mL TBS with 0 . 01% NP-40 and 1x cOmplete EDTA-free protease inhibitor ( Roche ) for GST purification , or equilibration buffer ( 50 mM sodium phosphate , 300 mM NaCl , 20 mM imidazole , pH 7 . 4 ) for nickel purification . Working lysate was prepared by French press lysis followed by centrifugation . Nickel ( Fig 1B ) or GST ( S1B Fig ) columns were prepared according to manufacturer’s instructions ( Gold Biotechnology and G-Biosciences , respectively , St . Louis , MO , USA ) . For GST purification , 200 μL bead slurry was washed in each column with 20 mL cold TBS . 10 mL working lysate ( AvrRps4 ) was added to columns and incubated on a rotating shaker for 1 hr at 4°C . Columns were washed with 100x column volume cold TBS . 9 . 5 ml 2x dilution of HIS-T7-EDS1 lysate was added to each column and incubated shaking at 4°C for 1 hr . Columns were again washed with 100x column volume cold TBS . Protein was eluted from beads by adding 250 μL 10 mM glutathione to the columns and shaking for 15 min at room temperature . For nickel column purification , 2 mL of prepared lysate was passed through 1 mL bead slurry and then washed with 10x column volumes 20 mM imidazole wash buffer , and 1 ml each 50 , 100 , 150 , and 200 mM imidazole wash buffers . Protein was eluted from beads in 250 uL fractions of 250 mM imidazole lysis buffer at 4°C . Immunoblot analysis of input and pulldown fractions were performed using 1:10 , 000 αGST-HRP and 1:20 , 000 αT7-HRP ( EMD Millipore , Billerica , MA , USA ) antibody dilutions . AvrRps4-GFP-containing plasmids were moved from E . coli into Agrobacterium tumefaciens GV3101 using electroporation . Plants were infiltrated with Agrobacterium at O . D . 0 . 2 . Cells were visualized using a Leica SP8 confocal microscope ( Leica Microsystems , Wetzlar , Germany ) 3 days post-infiltration . For HR assays in lettuce , N-terminally HA-tagged AvrRps4 full-length , AvrRps4N , AvrRps4C , HopK1 full-length , HopK1N , HopK1C , and HA-pBA empty vector were transiently expressed in Lactuca sativa cv . Kordaat using A . tumefaciens C58C1 infiltrated at an O . D . of 0 . 3 . Cell death phenotypes were visualized three days post-infiltration . Electrolyte leakage experiments were performed as described previously [45] . Lettuce leaves were infiltrated with agrobacterium strains , and 36 hours later , before HR symptoms become visible , 7 mm leaf discs from 6 individual plants per treatment were harvested , submerged in ddH2O , and vacuum-infiltrated . The six leaf disks were then transferred to a 20 mL GC vial containing 10 mL ddH2O . Conductivity was measured using a conductivity meter . Plants were grown in short-day conditions as previously described [26] . Arabidopsis rosette leaves were infiltrated with bacteria at 5 x 104 cfu/mL densities . Leaf discs were recovered in triplicate samples and thoroughly ground in 10 mM MgCl2 and plated on Difco Pseudomonas Agar ( Becton , Dickinson and Company , Sparks , MD ) containing appropriate antibiotics . Unless noted otherwise , statistical analyses of differential bacterial growth between treatments were performed using ANOVA with Holm correction for multiple comparisons . For in vitro secretion assays , bacteria were grown on plates containing appropriate antibiotics overnight . Cells were diluted to an O . D . of 2 x 108 cfu/mL into tubes containing 10 mL minimal media [46] at 19°C overnight . Proteins were precipitated using 100% TCA . Total bacterial pellets or precipitated proteins were subjected to immunoblot analysis using HA antibodies ( Roche ) . Total and secreted proteins were subjected to NPTII antibodies to confirm that effectors were not detected in secreted fraction due to cell lysis . For AvrRps4 in planta cleavage assays , Col-0 plants were infiltrated with 109 cfu/mL of wild-type DC3000 or DC3000 hopK1- containing full-length AvrRps4 with or without the AvrRpm1 signal peptide . Tissues were harvested at 0 . 5 and 7 hpi and ground in 8 M urea for Western blot analysis with HA antibody . Reverse transcription PCR was conducted as described previously [26] . Briefly , total RNA was isolated from indicated plant leaves using TRIzol reagent ( Ambion ) and RNA was reverse transcribed into cDNA using oligo ( dT ) 15 primers and Moloney murine leukemia virus ( MMLV ) reverse transcriptase ( Promega ) according to the manufacturer’s instructions . Equivalent amounts of cDNA were used in each PCR reaction to analyze expression of avrRps4 and PR1 . ACTIN2 was used as an internal control . Oligonucleotide primer sequences used in semi-quantitative RT-PCR are listed in S1 Table .	An important component of the plant immune system relies on the detection of pathogen-derived effectors by immune receptors called resistance proteins . Bacterial pathogens inject effectors into the host cell via a dedicated secretion system to suppress defenses and to manipulate the physiology of the host cell to the pathogen's advantage . Usually , a single resistance protein recognizes a single effector , but an increasing number of exceptions and elaborations on this one-to-one relationship are known . The plant Arabidopsis uses a pair of resistance proteins , RRS1 and RPS4 , to detect the effector AvrRps4 . After injection into the cell , AvrRps4 is cleaved into two protein parts , and it had been assumed that only the C-terminal part needs to be present to trigger RPS4/RRS1 . We show here that both AvrRps4 parts are required for triggering resistance in Arabidopsis , and that the N-terminal part , which previously had been assumed to only function in effector secretion into the host cell , in fact on its own has some functions of an effector . This conclusion is supported by the observation that the N-terminal part of AvrRps4 is sufficient to trigger resistance in lettuce . The fusion of the two AvrRps4 parts may have arisen to counteract plant defenses .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "methods" ]	[ "plant", "anatomy", "viral", "transmission", "and", "infection", "plant", "cell", "biology", "brassica", "microbiology", "chloroplasts", "membrane", "proteins", "developmental", "biology", "plant", "science", "model", "organisms", "experimental", "organism", "systems", ...	2018	The bacterial type III-secreted protein AvrRps4 is a bipartite effector
FERONIA ( FER ) , a plasma membrane receptor-like kinase , is a central regulator of cell growth that integrates environmental and endogenous signals . A peptide ligand rapid alkalinization factor 1 ( RALF1 ) binds to FER and triggers a series of downstream events , including inhibition of Arabidopsis H+-ATPase 2 activity at the cell surface and regulation of gene expression in the nucleus . We report here that , upon RALF1 binding , FER first promotes ErbB3-binding protein 1 ( EBP1 ) mRNA translation and then interacts with and phosphorylates the EBP1 protein , leading to EBP1 accumulation in the nucleus . There , EBP1 associates with the promoters of previously identified RALF1-regulated genes , such as CML38 , and regulates gene transcription in response to RALF1 signaling . EBP1 appears to inhibit the RALF1 peptide response , thus forming a transcription–translation feedback loop ( TTFL ) similar to that found in circadian rhythm control . The plant RALF1-FER-EBP1 axis is reminiscent of animal epidermal growth factor receptor ( EGFR ) signaling , in which EGF peptide induces EGFR to interact with and phosphorylate EBP1 , promoting EBP1 nuclear accumulation to control cell growth . Thus , we suggest that in response to peptide signals , plant FER and animal EGFR use the conserved key regulator EBP1 to control cell growth in the nucleus . FERONIA ( FER ) is a versatile receptor-like kinase ( RLK ) that controls nearly all aspects of plant cellular activity [1 , 2] . FER was originally reported to regulate pollen tube reception for successful double fertilization [3 , 4] . Further studies have revealed its multiple roles in regulating vegetative cell growth . For instance , FER is required for root hair elongation in response to auxin [5] . FER is also essential for the expansion of leaf cells associated with brassinolide ( BR ) response [6] and for hypocotyl cell elongation in relation to ethylene biosynthesis [7] and signal transduction [8] . FER also regulates fruit ripening in strawberry [9] and tomato [10] via ethylene biosynthesis regulation [7] . Our recent work has shown that two FER-like receptor ( FLR ) genes in rice are crucial for cell growth [11] . In addition , FER is involved in biotic and abiotic stress responses . fer mutants are hypersensitive to salt [12 , 13 , 14] , cold , and heat stress [12] and are hypersensitive to nickel ( Ni ) ions but are tolerant to the heavy metals cadmium ( Cd ) , copper ( Cu ) , lead ( Pb ) , and zinc ( Zn ) [15] . The roles of FER in stress response might be particularly attributable to its function in regulating abscisic acid ( ABA ) and rapid alkalinization factors ( RALFs ) signal transduction [12 , 16 , 17] . Studies have shown that fer mutants show a hypersensitive response to exogenous ABA with respect to stomatal closure and primary root growth [17] . FER suppresses ABA signaling by activating an A-type protein phosphatase 2C ( PP2C ) , ABA insensitive 2 ( ABI2 ) , through the guanine nucleotide exchange factor–plant RHO-related GTPases ( GEF-ROP/RAC ) pathway , which in turn inhibits the ABA response mediated by sucrose nonfermenting 1–related protein kinase 2 ( SnRK2 ) [17] . The activated ABI2 phosphatase interacts with and dephosphorylates FER to reduce FER activity , providing a cross-talk node between ABA and the RALF1 ( a ligand of FER ) peptide [12] . RALF1 is a 5-kDa peptide that rapidly induces alkalinization of cell culture medium and inhibits cell growth [18 , 19] . RALF1 binds to FER , increases FER phosphorylation , and further inhibits Arabidopsis H+-ATPase 2 ( AHA2 ) activity , leading to inhibition of root cell elongation [20] . Previous studies have shown that mutations in FER can alter fungal invasion [21 , 22] , implicating FER in immune responses . Along this line , other RALF-family peptides—including RALF17 , RALF23 , and RALF33—may work with FER to regulate 22 amino acid fragment of bacterial flagellin ( flg22 ) -triggered reactive oxygen species ( ROS ) burst [16] . It is further shown that FER exerts its control on immune signaling through a scaffold function to facilitate the complex formation of the immune receptor complexes kinases EF-Tu receptor ( EFR ) and flagellin-sensing 2 ( FLS2 ) with their coreceptor brassinosteroid insensitive 1–associated kinase 1 ( BAK1 ) to initiate immune signaling [16] . FER also works with several other proteins , such as RPM1-induced protein kinase ( RIPK ) [23] and LLG1 [24] , to transmit RALF1 signal . Although a handful of components have been identified in the RALF1 peptide signaling pathway , the mechanism by which FER regulates nuclear events remains unknown . A common theme of signal transduction from the cell surface to the nucleus involves the modification of cytoplasmic proteins by membrane receptors and then their accumulation in the nucleus to alter gene expression [25 , 26] . In animals , this scheme has been well demonstrated by studies of the epidermal growth factor receptor ( EGFR ) family [27] . EGFRs consist of four distinct receptors: erythroblastic leukemia viral oncogene homolog 1 ( ErbB1 ) , ErbB2 , ErbB3 , and ErbB4 [28] . ErbB3 is the receptor of heregulin ( HRG ) /neuregulin ( NRG ) peptides [29 , 30] . When the HRG ligand peptide binds the ErbB3 receptor in breast cancer cells , ErbB3-binding protein 1 ( EBP1 ) is phosphorylated by ErbB3 and accumulates in the nucleus [31] . Thus , in the breast cancer cell , EBP1 functions as a negative regulator of HRG-ErbB3 signal transduction [31] , and overexpression of EBP1 results in reduced cell growth and increased differentiation [32] . EBP1 is a DNA- [33] and RNA-binding [34 , 35] protein , and the proper localization of EBP1 is critical for its growth-suppressive properties [34] . In HeLa cell nuclei , EBP1 interacts with the E2F1 complex to suppress E2F1-regulated gene transcription [33] . Meanwhile , EBP1 also interacts with other transcriptional repressors—such as retinoblastoma gene ( Rb ) [36] , Sin3A [37] , and histone deacetylase 2 ( HDAC2 ) [38]—to suppress downstream gene transcription . EBP1 is also an RNA-binding protein [34 , 35] as a part of ribonucleoprotein ( RNP ) complexes . EBP1 not only associates with mature and precursor rRNA species [34 , 35] but also directly binds to some mRNA species and regulates their translation [35 , 39] . EBP1 is evolutionarily conserved in both animals and plants . Arabidopsis EBP1 has been previously identified as a protein that controls cell size and was named AtCPR [40] . For simplicity , hereafter , we will use the name EBP1 in this report . Overexpression of EBP1 homologs in Solanum tuberosum [41] , Arabidopsis [41] , Zea mays [42] , and Hevea brasiliensis [43] regulate organ size in a dose-dependent manner [41] , suggesting that the abundance of EBP1 is critical for its function in plant growth regulation . Plant EBP1 is also implicated in stress responses . For example , overexpression of AcEBP1 ( Atriplex canescens ) increases plant low temperature , NaCl and ABA sensitivity [44] , and drought stress resistance [43 , 44] . However , despite these overexpression analyses , it remains unknown how EBP1 is regulated by upstream signaling events . In search of FER partner proteins , we identified EBP1 as one of the interacting proteins . We found that RALF1-FER signaling enhanced EBP1 mRNA translation and further promoted its phosphorylation and nuclear accumulation in plants . EBP1 can directly bind to some chromatin loci and regulate their expression , thus connecting RALF1-FER signaling with gene regulation in the nucleus . To identify other components in the RALF1-FER/RIPK signaling pathway , a yeast two-hybrid ( Y2H ) screen was performed using the FER kinase domain ( FER-KD , 469–896 amino acids [aa] ) as a bait against an Arabidopsis cDNA library [12 , 17 , 23] . One truncated version ( 229–401 aa ) of EBP1 was identified as an interacting clone . To confirm this interaction , we cloned the full-length protein of EBP1 into an active domain ( AD ) vector and FER-KD into a binding domain ( BD ) vector . The Y2H assay showed that the full-length protein of EBP1 indeed interacted with the FER-KD in yeast cells ( Fig 1A ) . We also found that EBP1 interacted with other FER relatives that belong to the Catharanthus roseus receptor-like kinase 1-like kinase ( CrRLK1L ) family , such as ANXUR 1 ( ANX1 ) , AT5G24010 , and CURVY 1 ( CVY1 ) , in the yeast system ( S1 Fig ) . A glutathione S-transferase ( GST ) pull-down assay showed that the full-length EBP1 protein tagged with GST ( S2A Fig ) and FER-KD tagged with 6 × His [12] were copurified ( Fig 1B ) . EBP1 also interacted with a kinase-dead version of FER-KD containing the Lys565-to-Arg mutation ( FER-KDK565R-His ) [4] in a GST pull-down assay ( Fig 1B ) . We further confirmed the interaction between EBP1 and FER in planta by a bimolecular fluorescence complementation ( BiFC ) assay in Arabidopsis protoplasts . We simultaneously transferred EBP1–C-terminal cyan fluorescent protein ( cCFP ) and FER-nVenus ( or HERCULES2 [HERK2]-nVenus as negative control ) into Arabidopsis protoplasts to observe the reconstituted fluorescence . We found that EBP1 interacted with FER in the plasma membranes ( indicated by the styryl dye FM4-64 ) of Arabidopsis protoplasts but did not interact with HERK2 ( Fig 1C ) . We performed the western blot to show that the proteins in the BiFC assay were expressed ( S2B Fig ) . We also performed a coimmunoprecipitation ( Co-IP ) assay to examine the interaction between EBP1 and FER in planta using polyclonal antibodies against FER [12 , 23] and EBP1 ( S2C–S2E Fig ) . We used plant materials with or without RALF1 treatment and found that FER and EBP1 interaction was detected in both RALF1-treated and untreated samples ( Fig 1D ) . Using Ubi::FER-FLAG transgenic plants [12] , the interaction between EBP1 and FER was also confirmed via Co-IP assay ( S3 Fig ) . Taken together , these data indicate that FER physically interacts with EBP1 . EBP1 controls cell growth and proliferation in human cancer cell lines [32 , 45] and some higher plants [41 , 42 , 43] . To investigate whether EBP1 is evolutionarily conserved , we assessed 823 EBP1-like protein sequences in Animalia , Plantae , Fungi , and Protista ( S4 Fig ) . In Arabidopsis , we found only one EBP1 or EBP1-like protein ( S4 Fig ) . Plant EBP1s showed high sequence similarity ( S5 Fig ) . Using the UniProt Knowledgebase [46] , we further noticed that 10 α-helixes , 12 β-strands , and one turn might exist in the EBP1 secondary structure ( S5 Fig ) . An additional putative nucleus-localization sequence ( NLS ) and two regions that resemble nuclear localization signals are conserved in Arabidopsis EBP1 ( S5 Fig ) . To investigate the tissues and organs in which EBP1 mRNA may be expressed , we constructed proEBP1::GUS transgenic Arabidopsis and examined the pattern of β-glucuronidase ( GUS ) activity at different growth stages ( S6A–S6J Fig ) . In seedlings that were 7 days after germination ( DAG ) , proEBP1::GUS was expressed mainly in cotyledons and roots ( S6B , S6F and S6H Fig ) . GUS activity was also detected at the radicle tip ( S6D Fig ) . In 4-week-old proEBP1::GUS rosettes , the GUS activity was lower than that in 7-DAG seedlings , and slight GUS activity was detected in the vascular tissues and the mesophyll cells ( S6J Fig ) . No GUS activity was detected in nontransgenic plants ( S6A , S6C , S6E , S6G and S6I Fig ) . We further analyzed the EBP1 mRNA expression pattern using ePlant [47] and obtained similar expression patterns to those in the GUS staining assays ( S6K Fig ) . These data showed that EBP1 was expressed in root and cotyledon at the early stage of plant growth . However , the EBP1 protein level and localization in the next experiment showed that EBP1 mRNA translation is tightly regulated . In the Co-IP assays ( Fig 1D ) , we noticed that RALF1 may enhance the accumulation of EBP1 protein . We investigated this possibility using western blotting . Both Col-0 and the fer-4 mutant were treated with RALF1 , and total protein was collected at the indicated time points ( Fig 2A ) . Although the protein content of EBP1 increased in both Col-0 and the fer-4 mutant after RALF1 treatment ( Fig 2A ) , the RALF1-induced accumulation of EBP1 protein was more significant in Col-0 than in fer-4 ( Fig 2A ) , indicating that RALF1 induces EBP1 protein accumulation in a FER-dependent manner . At three levels ( EBP1 mRNA transcription , EBP1 mRNA stability , and EBP1 mRNA translation ) , we investigated how RALF1 peptide elevates EBP1 protein content ( Fig 2B–2H , S7A and S7B Fig ) . We first investigated whether RALF1 induces EBP1 protein accumulation at the mRNA transcript level . The results of the quantitative reverse transcription PCR ( qRT-PCR ) analysis showed that the EBP1 mRNA level was not significantly affected by RALF1 ( S7A Fig ) , indicating that EBP1 protein levels are subject to tight posttranscriptional control through an unknown mechanism . Using cordycepin ( CRD; an mRNA transcription inhibitor ) to stop the EBP1 mRNA synthesis , we measured the rate of EBP1 mRNA decay and found no significant difference in RNA decay rate with or without RALF1 treatment over 2 hours ( S7B Fig ) . ATHSPRO2 ( a reported gene with highly unstable mRNA [48] ) was used as positive control ( S7B Fig ) . We next investigated whether RALF1 promoted EBP1 mRNA translation by increasing its polysome profiles ( Fig 2B–2G ) . We performed polysome profiling assays and found that the portion of polysome-bound EBP1 mRNA was increased upon RALF1 treatment for 30 minutes in Col-0 ( Fig 2B and 2C ) . However , in the fer-4 mutant , the EBP1 mRNA binding to polysomes was not increased significantly ( Fig 2E and 2F ) , suggesting that RALF1-enhanced EBP1 mRNA translation was dependent on FER . The polysome-bound EIF4A1 mRNA content was measured as a control , showing that RALF1 did not induce polysome binding of EIF4A1 mRNA ( Fig 2D and 2G ) . To gain further evidence for RALF1-induced EBP1 mRNA translation , cycloheximide ( CHX; an mRNA translation inhibitor ) was used to block new protein synthesis . Western blots indicated that RALF1-induced EBP1 protein accumulation was abolished by CHX ( Fig 2H ) . When MG132 ( an inhibitor of the 26S proteasome ) was used , we did not observe significant change in RALF1-induced EBP1 protein accumulation ( Fig 2H ) . These data suggest that RALF1 affected EBP1 protein accumulation through altering the translation of EBP1 mRNA . To investigate the subcellular localization of accumulated EBP1 , Col-0 and fer-4 plants were collected ( with or without RALF1 treatment ) , and a nuclear fractionation assay was performed . Immunoblot assays were performed to measure the EBP1 protein content in both the cytoplasmic and nuclear fractions . In the cytoplasmic fractions of both Col-0 and fer-4 plants ( with or without RALF1 treatment ) , EBP1 was detected at a low level ( Fig 2I ) . RALF1 treatment only slightly increased the content of EBP1 in the cytoplasm ( Fig 2I ) . However , in the nuclear fraction , EBP1 was hardly detected in Col-0 and fer-4 plants without RALF1 treatment ( Fig 2I ) . Accumulation of EBP1 in the nucleus increased remarkably after RALF1 treatment in both Col-0 ( lane 6 ) and fer-4 ( lane 8 ) plants ( Fig 2I ) . However , the nuclear EBP1 level increased more in Col-0 than in fer-4 , suggesting that RALF1-induced nuclear accumulation of EBP1 is partially dependent on functional FER ( Fig 2I ) . We further investigated whether EBP1 accumulation in the nucleus was RALF1 specific ( S7C Fig ) . When treated with PEP1 peptide ( AT5G64900 ) , ABA , and 1-naphthaleneacetic acid ( NAA ) , no distinct EBP1 nuclear accumulation was detected ( S7C Fig ) . EBP1 is localized in the nucleus upon activation of the EGFR pathway in animals [31] . In plants , a previous work reported that EBP1–green fluorescent protein ( GFP ) ( AtCPR-GFP ) can be detected by GFP fluorescence in the guard cells but not in other cell types , even when its expression is driven by the 35S promoter [40] . We made overexpressing transgenic plants harboring a 35S::EBP1-GFP construct ( S7D–S7F Fig ) and further obtained fer-4/EBP1-GFP hybrid plants . Consistent with the previous work [40] , we observed GFP fluorescence in the EBP1-GFP transgenic plant guard cells ( S7G Fig ) . However , we detected only weak GFP fluorescence in roots ( Fig 2J ) , although we detected a high level of EBP1-GFP mRNA expression ( S7E Fig ) . Driven by these results , we further visualized GFP signaling in 7-DAG EBP1-GFP and fer-4/EBP1-GFP seedlings ( treated with or without 1 μM RALF1 peptide in 1/2 Murashige and Skoog [MS] liquid medium ) ( Fig 2J ) . Without the RALF1 peptide treatment , we observed only dim fluorescence in the roots of EBP1-GFP and fer-4/EBP1-GFP hybrid plants ( Fig 2J ) . A strong fluorescent signal was observed in the nuclei of the RALF1-treated EBP1-GFP plants ( Fig 2J ) . The nuclear localization of the EBP1-GFP fluorescent signal was verified by using the nuclear staining dye Hoechst 33258 ( S8A Fig ) . In RALF1-treated fer-4/EBP1-GFP seedlings , the fluorescent signal in the nucleus was observed but was weaker than that in the RALF1-treated EBP1-GFP plants ( Fig 2J ) . The percentage of cells having nuclear fluorescent signals in RALF1-treated EBP1-GFP plants was also higher than that in RALF1-treated fer4/EBP1-GFP plants ( Fig 2K ) . However , RALF1 did not enhance GFP intensity in the roots of 35S::GFP Arabidopsis plants ( S8B Fig ) , and free GFP was not detected in plant sample expressing EBP1-GFP ( S2D Fig ) , suggesting that the RALF1-induced fluorescence enhancement is attributed to EBP1-GFP fusion protein rather than free GFP accumulation . To further prove that RALF1 could promote EBP1 protein accumulation in the nucleus in an FER-dependent manner , an immunofluorescence assay was performed using EBP1 antibody ( S8C Fig ) . Without RALF1 treatment , weak fluorescence was observed both in Col-0 and fer-4 root cells ( S8C Fig ) . Before RALF1 treatment , we noted that EBP1 protein level was slightly higher in the fer-4 background when compared with wild type ( WT ) , which is briefly commented on in the Discussion section . With RALF1 treatment , the EBP1 protein was accumulated in the nucleus in Col-0 ( S8C Fig ) . The nuclear staining dye DAPI was used to confirm the localization of the nucleus ( S8C Fig ) . However , in the RALF1-treated fer-4 mutant , weak fluorescence was observed in the nucleus ( S8C Fig ) . These results suggest that RALF1 peptide promoted EBP1 mRNA translation and protein accumulation in the nucleus in an FER-dependent manner . As a kinase-interacting protein , EBP1 could potentially be phosphorylated by FER . A coexpression system in Escherichia coli was designed to examine the dynamics and the in vivo phosphorylation process . This system has been used successfully to study RLK-mediated phosphorylation reaction in E . coli [49] . Our previous work has shown that PP2C-A-type protein phosphatases ( such as ABI1 and ABI2 ) interact with and dephosphorylate FER , thus inhibiting the kinase activity of FER [12] . It is also known that the ABA receptor pyrabactin resistance 1-like 1 ( PYL1 ) attenuates ABI1’s inhibitory effect on FER in the presence of ABA [12] . Based on these findings , we constructed an ABA-induced in vitro phosphorylation system to investigate whether EBP1 is a phosphorylation substrate of FER ( Fig 3A ) . We coexpressed PYL1 , ABI1 , FER-KD , and EBP1 ( PYL1/ABI1/FER-KD/EBP1 ) in one E . coli strain . In parallel , we coexpressed the same proteins except for a mutant version of FER-KD that contained the Lys565-to-Arg mutation ( FER-KDK565R ) to abolish its kinase activity [4] as negative control . When ABA was not added to the culture , ABI1 would inhibit FER activity to avoid its phosphorylation ability toward to its substrate . After ABA was added to the culture medium , the PYL1-ABA complex interacted with and inhibited ABI1 activity , thus releasing the active FER-KD kinase . In the assay , EBP1 was fused with a His-tag , and FER-KD ( or FER-KDK565R ) was fused with an S-tag so that an immunoblot assay could be used to detect phosphorylation-related band shifts , as in our previous work [12] . First , we detected the phosphorylation status of FER-KD fused to S-tag ( FER-KD-S ) to ensure that this system worked . When no ABA was added , no band shift attributable to FER-KD phosphorylation was detected ( lane 1 ) ( Fig 3A ) . Once ABA was added , the band shift was detected ( lane 2 ) , and alkaline phosphatase ( AP ) dephosphorylated and eliminated band shifts in FER-KD ( lane 7 ) ( Fig 3A ) . This result proved that the addition of ABA activated FER-KD-S phosphorylation activity , and therefore , we investigated whether FER-KD could phosphorylate EBP1 in this system . Indeed , in the absence of ABA , no phosphorylation-dependent shift of EBP1 was detected ( lane 1 ) , whereas the addition of 50 μM ABA induced EBP1 phosphorylation ( p-EBP1 ) ( lane 2 ) ( Fig 3A ) . The ABA-dependent shift of the EBP1 protein band resulted from phosphorylation , because AP dephosphorylated and eliminated this shift in the EBP1 band ( lane 7 ) ( Fig 3A ) . This EBP1 phosphorylation was dependent on active FER kinase , because ABA did not induce an EBP1 band shift when added into the culture medium of the cells containing the kinase-dead FER-KDK565R , PYL1 , ABI1 , and EBP1 ( lane 3 and 4 ) ( Fig 3A ) . To further investigate whether EBP1 was phosphorylated by FER in vivo , we immunoprecipitated the EBP1 protein from Col-0 and fer-4 plants roots using an EBP1 antibody ( Fig 3B ) . Because RALF1-FER pathway alters EBP1 abundance in plants , we had to adjust the total protein levels of EBP1 from different samples to be comparable so that the changes in the phosphorylation levels of EBP1 protein can be easily visible ( Fig 3B ) . Because the mobility shift between EBP1 and p-EBP1 was small , we used phosphoserine ( pSer ) and phosphothreonine ( pThr ) antibodies to monitor the phosphorylation level of EBP1 . Before RALF1 treatment , the EBP1 pSer and pThr phosphorylation levels in Col-0 were both higher than those in fer-4 ( Fig 3B ) . After RALF1 treatment , the EBP1 pSer and pThr phosphorylation levels were increased in Col-0 but not in the fer-4 mutant ( Fig 3B ) , and the FER phosphorylation level was up-regulated , as indicated by the band shift assays . Thus , the phosphorylation assays suggest that EBP1 was phosphorylated by FER kinase . To identify the phosphorylation sites of EBP1 , the phosphorylation assay shown in Fig 3A was performed at a larger scale , and EBP1 was purified and analyzed using mass spectrometry ( S9 Fig ) . Ten EBP1 aa residues ( Ser13 , Thr15 , Ser16 , Thr242 , Thr243 , Tyr245 , Ser378 , Thr379 , Ser387 , and Ser388 ) were identified as phosphorylated in the sample containing ABA-incubated PYL1/ABI1/FER-KD/EBP1 system , but no phosphorylation was observed either in the absence of ABA or in the ABA-incubated PYL1/ABI1/FER-KDK565R/EBP1 system ( S9 Fig ) . To further confirm these 10 phosphorylation sites , we simultaneously mutated all 10 identified phosphorylation sites to alanine residues ( yielding mEBP1-10A ) to constitutively inactivate these phosphorylation sites and tested whether mutated mEBP1-10A-His was phosphorylated by FER-KD ( Fig 3A ) . The results showed that , although ABA activated FER-KD , mEBP1-10A-His was not phosphorylated by FER-KD ( comparing lanes 5 and 6 ) , suggesting that one or multiple of these residues may be phosphorylated by FER . We next further analyzed the physiological roles of these 10 phosphorylation sites . Based on the domain structure of the Arabidopsis EBP1 protein , we analyzed the locations of the 10 phosphorylation sites of EBP1 protein ( S5 Fig ) . We found that 7 phosphorylation sites were located at the N-terminal and C-terminal potential nuclear localization–related regions ( NLRs ) ( S5 Fig ) . To investigate whether these p-EBP1 sites were critical for EBP1 to accumulate in the nucleus , we further tested the functional relevance of these 10 phosphorylated aa residues by examining RALF1-dependent nuclear accumulation of EBP1 ( Fig 3C–3K ) . We cloned the EBP1 or mEBP1-10A sequences into a 35S promoter-driven GFP fusion construct and obtained EBP1-GFP and mEBP1-10A-GFP ( S2D Fig , S7D–S7F Fig ) transgenic Arabidopsis . As in Fig 2J , RALF1 promoted EBP1-GFP protein accumulation in the nucleus ( Fig 3C ) . RALF1 also promoted mEBP1-10A-GFP protein levels ( S7F Fig ) ; however , only weak fluorescence was detected in the nucleus , compared with that in unmutated EBP1-GFP plants ( Fig 3C ) . DAPI staining and plot profile analysis showed that the enhanced fluorescence of RALF1-treated EBP1-GFP was mainly in the nucleus ( Fig 3D–3K ) . We performed a nucleus fractionation assay to detect the accumulation of EBP1-GFP and mEBP1-10A-GFP in the nucleus ( Fig 3L ) . RALF1 treatment increased the content of both EBP1-GFP and mEBP1-10A-GFP in nucleus ( Fig 3L ) . However , a much weaker mEBP1-10A-GFP signal was detected in the nucleus when compared with EBP1-GFP ( Fig 3L ) . Further , we simultaneously mutated the three phosphorylation sites ( Ser13 , Thr15 , and Ser16 , yielding mEBP1-N3A ) located in the N-terminal NLR ( N-NLR ) , the three phosphorylation sites in the middle section ( Thr242 , Thr243 , and Tyr245 , yielding mEBP1-M3A ) , the four sites located in the C-terminal NLR ( C-NLR ) ( Ser378 , Thr379 , Ser387 , and Ser388 , yielding mEBP1-C4A ) , or all 10 phosphorylation sites ( mEBP1-10A ) to alanine residues to constitutively inactivate these phosphorylation sites and incorporated these mutations in the GFP fusion constructs ( S10 Fig ) . We used an Arabidopsis protoplast transfection assay to investigate whether these mutant proteins showed altered localization in response to exogenous RALF1 peptide ( S10 Fig ) . Without RALF1 treatment , the WT EBP1-GFP and all mutated versions showed similar nuclear localization ( S10A and S10B Fig ) . After RALF1 treatment for approximately 1 hour , the percentage of cells with nuclear localization of WT EBP1-GFP was increased to 29 . 2% ( approximately twice more as compared to those before RALF1 treatment , P < 0 . 001 ) ( S10A and S10B Fig ) . However , mEBP1-C4A-GFP and mEBP1-10A-GFP did not show significant changes ( P > 0 . 05 ) in nuclear localization upon RALF1 treatment ( S10B Fig ) . mEBP1-N3-GFP and mEBP1-M3-GFP showed approximately 63 . 5% ( P < 0 . 001 ) and 77 . 7% ( P < 0 . 001 ) increases in cells with nuclear localization after RALF1 treatment , respectively ( S10B Fig ) . These data suggest that these 10 phosphorylation sites might affect RALF1-induced EBP1 nuclear accumulation to different extents . The C-NLR-associated phosphorylation sites ( Ser378 , Thr379 , Ser387 , Ser388 ) seem to be more important for EBP1 nuclear accumulation , consistent with their close proximity to the NLS ( S5 Fig ) . We further examined the effect of mutations in all 10 phosphorylation sites on nuclear accumulation of EBP1 by comparing the average nucleus/cytoplasm ( N/C ) fluorescence ratios of EBP1-GFP and mEBP1-10A-GFP upon RALF1 treatment ( S10C–S10E Fig ) . The N/C fluorescence ratios were 1 . 63 for EBP1-GFP and 1 . 18 for mEBP1-10A-GFP , showing a significant difference ( P < 0 . 001 ) ( S10C–S10E Fig ) . FER plays a critical role in cell growth , and its function may vary in different cell types as well as in distinct hormonal responses [50] . To examine whether EBP1 functions in the FER pathway , we sought to directly compare the phenotypic defects in ebp1 and fer mutants . To this end , we obtained three ebp1 mutant lines and EBP1-overexpression ( EBP1-OE ) lines ( S11 Fig ) , as described in the Methods . One of the phenotypic changes in fer mutants is shorter hypocotyls when grown in the dark [7 , 8] . We compared the hypocotyl lengths of dark-grown Col-0 , fer-4 , and ebp1 mutants . The three ebp1 mutants , similar to fer-4 , all showed shorter hypocotyls than Col-0 ( P < 0 . 001 ) , whereas the EBP1-OE lines showed longer hypocotyls than the WT ( S12A and S12B Fig , P < 0 . 01 ) . We stained the hypocotyls using propidium iodide ( PI ) ( S12C Fig ) and found that the cells in the ebp1 mutant lines were shorter than those in the WT plants ( S12D Fig , P < 0 . 001 ) . Another hallmark of FER function is regulation of ROS-mediated root hair growth through the GEF-ROP/RAC pathways [5] . The fer mutants have shorter root hairs because of a reduced response to auxin . The ebp1 mutants , unlike fer mutants , showed longer root hairs than those of the WT ( Fig 4A and 4B , P < 0 . 001 ) . In contrast , the root hairs in the EBP1-OE lines were shorter than those of the WT ( Fig 4A and 4B , P < 0 . 001 ) . FER plays a negative role in the control of seed size , with the fer-4 mutant producing larger seeds than those of the WT [51] . We measured the seed sizes of ebp1 and EBP1-OE and found that ebp1 had smaller seeds ( P < 0 . 01 ) than those of the WT , whereas EBP1-OE seeds were larger ( P < 0 . 01 ) ( Fig 4C and 4D ) . The positive role of EBP1 in seed size control is also supported by a recent study showing that overexpression of maize EBP1 in transgenic Arabidopsis increases seed size [42] . FER is expressed in guard cells and plays an important role in ABA-induced stomatal closure through the GEF-ROP/RAC signaling network [12 , 17] . Previous research has shown that EBP1 is expressed in guard cells [40] , suggesting that EBP1 may also play a role in guard cells . When seedlings were assayed for greening response to ABA , ebp1 mutant seeds germinated more rapidly than the WT and led to a higher percentage of green seedlings , indicating a lower sensitivity to ABA ( Fig 4E ) . We further tested the stomatal response to ABA in ebp1 and WT plants and found that ebp1 mutants were less sensitive to ABA than Col-0 seedlings ( Fig 4F , P < 0 . 05 ) . The above phenotypic comparisons implicated EBP1 in FER-regulated cellular activities . Next , we performed RALF1 peptide response assays [20] to compare directly if ebp1 mutant and fer mutant displayed related phenotypes . First , we investigated root elongation inhibition in response to RALF1 [23] in plants of various genotypes , including fer-4 , ebp1 , EBP1-GFP and fer-4/EBP1-GFP ( Fig 5A ) . Consistent with previous research [20] , fer-4 was less sensitive to RALF1 than the WT ( Fig 5A , P < 0 . 001 ) . Compared to the mock treatment , Col-0 root length was reduced approximately 52 . 6% in the presence of RALF1 peptide ( P < 0 . 001 ) , whereas fer-4 was reduced approximately 13 . 3% ( Fig 5A , P < 0 . 001 ) . The ebp1 was more sensitive than the WT to RALF1 treatment , as reflected by more severe reduction in root elongation ( 71 . 1% for ebp1-1 , 65 . 6% for ebp1-2 , and 75 . 0% for ebp1-3 ) ( Fig 5A , P < 0 . 001 ) . In contrast , EBP1-GFP was less sensitive to RALF1 than the WT ( 45 . 1% inhibition ) ( Fig 5A , P < 0 . 05 ) . An additive phenotype of RALF1 insensitivity in fer-4/EBP1-GFP ( 13 . 1% inhibition ) compared to EBP1-GFP was observed ( Fig 5A , P < 0 . 001 ) . However , no significant additive RALF1 insensitivity was observed in fer-4/EBP1-GFP relative to fer-4 ( Fig 5A , P > 0 . 05 ) , suggesting that EBP1 played roles downstream of the RALF1-FER signaling pathway , and EBP1 inhibited RALF1 response in roots . The RALF1 responses in the other EBP1-OE lines ( rdr6 background ) and rdr6 control were analyzed ( Fig 5B ) , and we found RALF1 had lower sensitivity in the EBP1-OE lines than that in the rdr6 control ( Fig 5B , P < 0 . 001 ) . The mEBP1-10A-GFP lines also showed lower sensitivity than that of Col-0 ( Fig 5C , P < 0 . 05 ) , albeit to a lesser extent than EBP1-GFP ( Fig 5C , P < 0 . 05 ) . Meanwhile , we crossed the ebp1-1 mutant with plants expressing EBP1-GFP or mEBP1-10A-GFP ( S13 Fig ) and performed root growth response to RALF1 . EBP1-GFP rescued the ebp1-1 mutant in the root growth assay ( S13 Fig , P > 0 . 05 ) , showing the functionality of the EBP1-GFP fusion protein . However , compared with EBP1-GFP , mEBP1-10A-GFP only partially complemented the ebp1-1 mutant phenotype in the same assay ( S13 Fig , P < 0 . 001 ) , again suggesting that the phosphorylation sites are important for EBP1’s role in plants . A study shows that some RALFs ( e . g . , RALF23 , RALF33 , and RALF34 ) inhibit flg22- or 18 amino acid fragment of bacterial elongation factor Tu ( elf18 ) -triggered ROS bursts [16] . Here , we checked the RALF1 response in ebp1 mutant plants using the similar assays . In response to flg22 , the WT and ebp1 mutant plants show a similar level of ROS burst that has been reduced in the fer-4 mutant ( S14A Fig ) . When treated with flg22 and RALF1 combined , the ROS burst in the WT was partially inhibited ( S14A Fig , P < 0 . 001 ) , suggesting that RALF1 suppressed flg22-triggered ROS burst . In ebp1 mutant lines , RALF1 more severely impaired the flg22-triggered ROS burst , again suggesting that ebp1 mutants were more sensitive to the RALF1 peptide than Col-0 ( S14A Fig , P < 0 . 05 ) . The activation of mitogen-activated protein kinase ( MAPK ) cascade is another RALF response indicator [52] . We found that RALF1-induced MAPK activities were higher in ebp1 mutant lines than in Col-0 , further suggesting that ebp1 mutants are more sensitive to RALF1 peptide ( S14B Fig ) . RALF1 induces rapid alkalinization of culture media [18] via the active FER kinase receptor by inhibiting AHA2 activity [20] . Thus , we measured pH variations in culture medium before and after the addition of RALF1 ( S14C and S14D Fig ) . Within 10 minutes , RALF1 significantly increased the medium pH in the WT plants ( S14D Fig ) , similar as reported earlier [18] . The pH values were approximately 7 . 05 , 6 . 27 , 8 . 22 , 7 . 62 , and 7 . 51 in Col-0 , fer-4 , ebp1-1 , ebp1-2 , and ebp1-3 , respectively , after 10 minutes ( S14D Fig ) . Because of the subsequent H+ efflux , the medium pH started to decrease , but the medium pH values of the ebp1 mutants were still higher than that of Col-0 , and the medium pH of the fer-4 plants was the lowest ( S14D Fig ) , suggesting that RALF1 has strong effects on ebp1 mutants , in contrast to Col-0 and fer-4 . As EBP1 affected many aspects of FER function ( especially several nonnuclear RALF1 responses , including RALF1-mediated inhibition of flg22-triggered ROS , MAPK phosphorylation , and proton secretion ) , we tested the possibility that EBP1 might affect FER gene expression . We examined FER mRNA levels in ebp1 mutants and EBP1-OE and found that FER was up-regulated in ebp1 mutants and was down-regulated in EBP1-OE ( Fig 5D ) . Taken together , these assays suggested that EBP1 negatively regulates RALF1 response . EBP1 is a DNA- [33] and RNA-binding [34 , 35] protein . Thus , we wondered whether the RALF1-FER pathway regulates the association of EBP1 with DNA to control gene transcription after its accumulation in the nucleus . First , RNA sequencing ( RNA-seq ) was performed to screen EBP1-regulated genes . We prepared RNA-seq libraries using RNAs isolated from 7-DAG Col-0 ( with or without RALF1 treatment for 2 hours ) , ebp1-1 , and fer-4 seedlings with three biological replicates ( Fig 6A–6C , S15 Fig ) . Comparing Col-0 and ebp1-1 , we found 367 genes with lower transcript levels and 360 genes with higher transcript levels in the ebp1-1 mutant plant ( Fig 6A and 6C ) . These affected genes mainly function in signal transduction ( 142 genes ) , transcription events ( 78 genes ) , and stress responses ( S15A Fig ) . In fer-4 , 3 , 387 genes were affected , with 1 , 634 and 1 , 753 genes showing lower or higher mRNA levels than Col-0 , respectively ( Fig 6A and 6C ) . RALF1 treatment affected approximately 4 , 007 genes ( 1 , 112 genes with lower and 2 , 895 genes with higher mRNA levels than control ) in Col-0 background ( Fig 6B and 6C ) . Using DAVID Functional Annotation tools [53] , we also analyzed the functional annotations ( P < 0 . 001 ) of differentially expressed genes between ebp1-1 and fer-4 ( ebp1-1 versus fer-4 ) and among ebp1-1 , fer-4 , and RALF1-treated Col-0 ( ebp1-1 versus fer-4 versus RALF1-treated Col-0 ) ( S15B and S15C Fig ) . Most of the genes altered in ebp1-1 versus fer-4 were related to signal transduction ( 84 genes ) , transcription events ( 41 genes ) , and stress responses ( S15B Fig ) . Most of the genes altered in ebp1-1 versus fer-4 versus RALF1-treated Col-0 played roles in the transcription process ( 30 genes ) or transcription regulation events ( 31 genes ) ( S15C Fig ) . These results suggested that the EBP1 signaling pathway is related to the RALF1-FER pathway by regulating a subset of overlapping genes . As a nuclear-localized protein , EBP1 might directly bind to and regulate gene expression . Through RNA-seq analysis , we found that 181 overlapping genes were simultaneously affected among ebp1-1 , fer-4 , and RALF1-treated Col-0 ( Fig 6C ) . We speculated that target genes are directly bound by EBP1 and regulated by the RALF1-FER-EBP1 signaling pathway . We tested this idea by examining some of these overlapping genes or related genes that have been reported to be regulated by RALF1 [20 , 54] . Using chromatin immunoprecipitation ( ChIP ) –quantitative PCR ( qPCR ) assay ( with RALF1-treated Col-0 seedlings ) , we screened 17 potential genes ( see the Methods ) in the context of the information from RNA-seq . We found that DNA fragments from four genes ( −213 to +117 of CML38 gene; −1113 to −844 of CKX4 gene; −1593 to −1195 of ERF1B gene; −2675 to −2397 of SAUR9 gene ) were immunoprecipitated by EBP1 antibody , whereas the adjacent DNA regions were not ( S16A–S16H Fig ) . We further found that RALF1 treatment enhanced the EBP1 association with these four target genes ( Fig 6D , S16I–S16K Fig ) . Notably , CML38 was recently revealed as a downstream target gene of the RALF1-FER signaling pathway , as the cml38 mutant becomes insensitive to RALF1 peptide in root elongation assay [54] . Thus , we further analyzed the CML38-detailed motif that was regulated by EBP1 using a candidate DNA screen strategy . Fortunately , we found a CCACGTC motif ( −201 to −194 of CML38 ) that was located in the EBP1-immunoprecipitated DNA fragments of the CML38 gene ( −213 to +117 ) and further confirmed that this motif was bound directly by EBP1 protein in vitro using an electrophoretic mobility shift assay ( EMSA ) ( Fig 6E ) . For performing a transient transcription dual-luciferase assay ( Dual-LUC ) , proCML38::LUC ( combined with the luciferase [LUC] reporter gene fused to the CML38 promoter containing the CCACGTC motif ) and 35S::EBP1 vectors were constructed to express proCML38-derived LUC and EBP1 protein , respectively , as described in Methods . Using this LUC assay , we confirmed that EBP1 suppressed CML38 transcription in the protoplasts isolated from Col-0 plants in a RALF1-dependent manner ( Fig 6F ) . proCML38::LUC and 35S::EBP1 were cotransferred into ebp1-1 mutant , and decreased proCML38::LUC activity was shown after RALF1 treatment ( S17 Fig ) . In the fer-4 mutant , lower suppression levels were detected with and without RALF1 treatment ( Fig 6F ) . We performed qRT-PCR assay to measure CML38 expression levels in ebp1 mutant and in plants overexpressing EBP1-GFP , respectively , in response to RALF1 ( S18A Fig ) . Without RALF1 treatment , CML38 level is higher in ebp1 mutants as compared to the Col-0 ( S18A Fig ) . After RALF1 treatment , CML38 expression was up-regulated in ebp1 but down-regulated in EBP1-GFP when compared with Col-0 ( S18A Fig ) , suggesting that EBP1 inhibits CML38 gene expression . Taken together , these data suggest that EBP1 suppressed CML38 mRNA expression in response to RALF1 . We further detected the relative mRNA expression levels of CKX4 , ERF1B , and SAUR9 ( S18B–S18D Fig ) . Consistent with the RNA-seq data , CKX4 , ERF1B , and SAUR9 showed up-regulated expression in the fer-4 mutant ( S18B–S18D Fig ) . The expression of ERF1B and SAUR9 was also up-regulated in the ebp1 mutants ( S18C and S18D Fig ) , whereas CKX4 was down-regulated in ebp1 mutants ( S18B Fig ) . The mechanisms by which RLK regulates mRNA translation in plants cells are largely unknown . The NIK-RPL10-LIMYB pathway may be the only example in which an RLK ( e . g . , nuclear shuttle protein-interacting receptor-like kinase [NIK] ) may regulate the translation process in response to viral infection in Arabidopsis and tomato . The receptor kinase NIK can interact with the ribosomal protein RPL10 and redirect RPL10 to the nucleus [55] . In the nucleus , RPL10 interacts with L10-interacting MYB domain-containing protein ( LIMYB ) to down-regulate the expression of genes encoding subunits of the translational machinery , leading to global translation suppression [56] . In this study , we identified a new regulatory mechanism in which RALF1-FER promotes translation of EBP1 mRNA . A previous study has shown that EBP1 regulates organ growth in a dose-dependent manner [41] , suggesting that the protein abundance of EBP1 is critical for its function and that EBP1 protein levels must be strictly regulated to maintain an optimal level . Our results have shown that , although EBP1 mRNA was transcribed in the absence of RALF1 treatment ( S6H and S6K Fig ) , it was translated at a very low level in the root tip ( Fig 2J ) . The RALF1-FER signal triggered rapid protein synthesis of EBP1 via increasing EBP1 mRNA translation efficiency ( Fig 2 ) . This phenomenon is reminiscent of a transcription–translation feedback loop ( TTFL ) , a mechanism that has been well defined in circadian rhythm control [57] but has rarely been reported in receptor-mediated signaling pathways . The exact mechanism by which RALF1-FER promotes EBP1 mRNA translation remains unclear . Although we focused on a specific target EBP1 in this report , we cannot exclude the possibility that RALF1-FER may regulate the translation machinery to promote global translation in the cell . In any case , our work provides an example showing that regulation at the level of protein translation may serve as a critical mechanism in signal transduction regulation . In addition to regulation on translation rate by RALF1-FER pathway , EBP1 protein may be subjected to other control mechanisms to maintain the optimal levels . One such mechanism may involve factors that control the protein stability . An intriguing finding in this study was that EBP1 protein content was higher in the fer-4 mutant than the WT plants under normal conditions . This seems to contradict the major results that RALF1-FER promotes EBP1 translation , as discussed above . We interpret this result in two possible ways . First , FER may regulate EBP1 protein stability via an unknown mechanism . We performed an in vitro protein degradation assay ( S19A Fig ) [58] and found that EBP1-His protein was more stable when incubated with total protein extract from fer-4 mutant than with Col-0 extract . Furthermore , mEBP1-10A-His was more stable than EBP1-His when incubated with total protein extract from Col-0 in the protein degradation assay ( S19B Fig ) . These results suggest that FER might reduce EBP1 protein stability . Secondly , in view of multiple roles of EBP1 in the ABA and salt stress response [44] , we suggest that EBP1 protein levels may respond to other environmental cues , such as biotic and abiotic stress conditions . As fer mutants show stress phenotypes ( such as enhanced ABA response ) , altered stress responses in fer mutants may also regulate EBP1 protein levels . The complexity of the control mechanisms for EBP1 protein abundance in plant cells corresponds to the multifunctional nature of EBP1 . We propose here some possible reasons that may explain the observations that EBP1 , like FER , plays different , and in some cases opposite , roles in cell growth in different organs . Firstly , some other CrRLK1L family members , such as ANX1 and CVY1 , can interact with EBP1 but result in different consequences when compared with FER functions . There are known examples showing that CrRLK1L members play opposite roles in specific tissues [59] . The interaction between other CrRLK1Ls with EBP1 may also explain the finding that low levels of EBP1 were still detected in the nuclear fraction in RALF1-treated fer-4 roots ( Fig 2I ) . Secondly , in addition to the response to RALF1 in root growth regulation [20] , FER also responds to other RALF family peptides ( such as RALF17 , RALF23 , and RALF33 ) to regulate different or opposite cellular activities [16] . This assumption is further supported by studies showing that one CrRLK1L can sense distinct RALF members to fulfill an opposite function [60 , 61] . During signal transduction , RLKs often directly interact with proteins in the plasma membrane or cytoplasm [62] . The mechanisms by which RLKs regulate nuclear events often involve a number of steps through the cytoplasm . BR insensitivity 1 ( BRI1 ) /BRI1-associated receptor kinase 1 ( BAK1 ) RLKs regulate Brassinazole resistant 1 ( BZR1 ) /bri1-EMS-suppressor 1 ( BES1 ) phosphorylation status and their accumulation in the nucleus through the bri1-suppressor 1 ( BSU1 ) -BR insensitive 2 ( BIN2 ) phosphorylation cascade , serving as a good paradigm [62] . In contrast to this paradigm , however , we found that FER may directly phosphorylate and enhance the nuclear accumulation of EBP1 , a transcriptional regulator , in response to RALF1 peptide ( Fig 3 ) . We suggest that this novel mechanism of RLKs directly interacting with and phosphorylating a nuclear-cytoplasmic shuttling protein might represent a rapid and effective strategy in response to extracellular signals to control nuclear events , although similar mechanisms have rarely been identified in RLK regulation . In the future , experimental work needs to be done at structural and biochemical levels to address the question of how FER in plants and EGFR in animals share a similar substrate , EBP1 , to transmit the signal from the cell surface to the nucleus . As a nuclear-localized protein , EBP1 may bind to and thus regulate gene transcription in the nucleus . In this study , we showed that EBP1 binds to the promoter region and inhibited the expression of CML38 , which has been shown to play a role in RALF1-induced inhibition of root elongation [54] . When CML38 function is disrupted in Arabidopsis plants , RALF1 peptide fails to inhibit root growth [54] . However , CML38 may not be sufficient to count for all RALF-regulated cellular activities . Indeed , except for CML38 , EBP1 has many other targets ( e . g . , CKX4 ) that may be relevant in RALF1-regulated processes such as extracellular alkalinization . Additionally , mutations in EBP1 may affect the expression of some signaling components in the RALF1-FER pathway . In support of this assumption , we found that FER was up-regulated in ebp1 mutants but was down-regulated in EBP1-OE ( Fig 5D ) . The altered level of FER expression may thus , at least in part , explain why ebp1 plants showed altered responses to RALF1 in nonnuclear events , such as the RALF1-mediated inhibition of flg22-triggered ROS , MAPK phosphorylation , and proton secretion ( S14 Fig ) . Further work is needed to identify EBP1 target genes at the whole-genome level by , for example , ChIP sequencing ( ChIP-seq ) assays to further address the broad function of EBP1 . At the posttranscriptional level , EBP1 has been identified as a part of RNP complexes and can bind to RNA directly in previous studies [34 , 35] . One example showed that EBP1 binds to mRNA and regulates its translation [39] . In the plant kingdom , whether EBP1 also regulates RNA-related events such as mRNA translation , mRNA alternative splicing , or mRNA decay remains an open question . The ebp1-1 ( SALK_030408 ) , ebp1-2 ( SALK_052695 ) , and ebp1-3 ( CS854731 ) T-DNA insertion mutants were obtained from the Salk Institute ( http://signal . salk . edu ) . All three ebp1 mutants were confirmed regarding their T-DNA insertion locations using genomic DNA PCR amplification ( S11A Fig ) , and the exact insertion sites of ebp1 lines were identified by sequencing ( S11B Fig ) . The primer sequences used to identify the ebp1 mutants are shown in S1 Table . The T-DNA was located after the 976th bp in EBP1 CDS ( inside the eighth exon ) in ebp1-1 , behind the 915th bp in EBP1 CDS ( inside the seventh exon ) in ebp1-2 ( S11B Fig ) , and after the 2 , 293th bp in EBP1 genomic sequence ( inside the eighth intron ) in ebp1-3 ( S11B Fig ) . Immunoblot assay was performed to confirm that all three ebp1 mutants lacked a detectable level of EBP1 protein ( S11C Fig ) . The 35S::EBP1-GFP and 35S::mEBP1-10A-GFP were constructed using pMD1-GFP vector for obtaining transgenic Arabidopsis in Col-0 background . The 35S::EBP1 ( EBP1-OE ) was constructed using pBI121 vector for obtaining EBP1-OE in rdr6-11 background to reduce the transgene-induced gene silencing [63] . We confirmed the EBP1 was overexpressed in two transgenic lines , EBP1-OE-1 ( about 7 . 5 times to WT ) and EBP1-OE-2 ( about 10 times to WT ) , using real-time RT-PCR ( S11D Fig ) . The primer sequences used in this section are provided in S1 Table . For plant culture on the agar plates , Arabidopsis thaliana seeds were sterilized and then vernalized at 4°C for 3 days before being grown on 1/2 MS with 0 . 8% ( w/v ) sucrose solidified with 1% ( w/v ) agar ( A7002 , Sigma-Aldrich ) . For hypocotyl elongation assay , Arabidopsis was grown in the agar plates in vertical position in complete darkness in 23°C for 5 days . For ABA treatment assay , Arabidopsis was grown on 1/2 MS medium supplemented with ABA in indicated concentrations . The GST-tagged EBP1 protein was purified as described in the manufacturer’s manual using Pierce Glutathione Agarose ( 16102 , Thermo Fisher Scientific , USA ) . The 6×His-tagged EBP1 and 6×His-tagged RALF1 protein [23] were purified as described in the manufacturer’s manual , using Ni-NTA Purification System ( R901-15 , Invitrogen , USA ) . For EBP1 antibody production , purified EBP1-His protein was used as the antigen to inject ( S2C Fig ) . A 1-month-old ICR mouse ( SLAC laboratory animal ) was injected with 50 μg EBP1-His protein emulsified with Complete Freund’s adjuvant ( F5881 , Sigma-Aldrich ) . Two weeks later , 50 µg EBP1-His protein emulsified with Incomplete Freund’s adjuvant ( F5506 , Sigma-Aldrich ) was injected into the ICR mouse and then once again in the next week . The serum of the immunized mouse was obtained as EBP1 antibody for immunoblot detection . We tested and ensured that the EBP1 antibody was specific by an immunoblot , using protein extracts from ebp1 mutant lines ( S11B Fig ) , EBP1-GFP ( S2D Fig ) , and EBP1-FLAG transgenic plants ( S2E Fig ) . To separate the native EBP1 and EBP1-FLAG protein clearly , 20% ( V/V ) glycerol was added into the PAGE gel , and the electrophoresis was performed for 15 hours under low voltage ( 60 V ) . The Y2H assay was performed as described [64] . The coding sequences of FER-KD were fused in-frame with the GAL4 DNA-binding domain of the bait vector pGBKT7 . The other CrRLK1Ls subfamily members were constructed into pGBKT7 as described in our previous work [23] . The coding sequence of EBP1 was fused in-frame with the GAL4 DNA-activating domain of the prey vector pGADT7 . The bait plasmid FER-BD and the prey plasmids or cDNA library were cotransformed into the yeast strain AH109 . The ORF sequences of FER ( or HERK2 as negative control ) and EBP1 were amplified by PCR and cloned into plasmid pE3308 and pE3449 , respectively [17] . Protoplasts were isolated from 5-week-old Arabidopsis rosette essentially as described [65] . Leaf strips were incubated in the cell wall–degrading enzyme solution in the dark for 3 hours . Protoplasts were purified [65] and transfected with 20 μg of plasmid DNA and an equal volume of PEG solution . The transfected protoplasts were incubated in the dark at 23°C for 16 hours to allow expression of the BiFC proteins . Recombinant FER-KD-His protein was incubated overnight at 4°C with GST beads coupled with GST-EBP1 in the binding buffer ( 20 mM HEPES [pH 7 . 5] , 40 mM KCl , 5 mM MgCl2 ) . The beads were washed five times with the TBS buffer ( 50 mM Tris [pH 7 . 5] , 150 mM NaCl ) and boiled in SDS-PAGE sample buffer , and eluted proteins were analyzed by immunoblot with anti-His ( M20001 , Abmart ) or anti-GST ( SC-80998 , CMC ) antibody . For performing a Co-IP assay using A/G agarose and FER-antibody , 30 μL A/G beads ( 20421 , Thermo Fisher Scientific ) were resuspended and washed three times using NEB buffer ( 20 mM HEPES [pH 7 . 5] , 40 mM KCl , 5 mM MgCl2 ) before adding 8 µL anti-FER antibody [12 , 23] ( or preimmune serum as negative control ) in a total volume of 500 µL NEB buffer and incubating for 3 hours at 4°C . The antibody-beads mixture was centrifuged at 200g for 1 minute to remove the supernatant . For protein extraction from plants , 7-DAG seedlings were transferred from the solid medium into liquid 1/2 MS medium and preincubated for 12 hours before RALF1 treatment to avoid manipulation-related effect during plant transfer from solid to liquid medium . Then , seedlings were soaked with 1 μM RALF1 ( included in the 1/2 MS liquid medium ) or mock control ( 1/2 MS medium containing RALF1 buffer only ) for 30 minutes , and then these seedlings were ground to a fine powder in liquid nitrogen and solubilized with NEBT buffer ( 20 mM HEPES [pH 7 . 5] , 40 mM KCl , 5 mM MgCl2 , 0 . 5% Triton X-100 ) containing 1 × protease inhibitor cocktail ( 78430 , Thermo Fisher Scientific ) and 1 × phosphatase inhibitor ( 78420 , Thermo Fisher Scientific ) and incubated for 2 hours at 4°C . The extracts were centrifuged at 16 , 000g at 4°C for 15 minutes , and the resultant supernatant was incubated with prepared antibody beads from the above step . The tube was rotated overnight at 4°C . Then , the agarose gel was washed five times with the NEBT buffer and eluted with elution buffer ( 0 . 2 M Glycine , 0 . 5% Triton X-100 [pH 2 . 5] ) . Anti-FER and anti-EBP1 antibody were used for immunoblot assay to detect the immunoprecipitates . For performing a Co-IP assay using FLAG agarose , FER-FLAG protein extract was prepared in a similar manner as described above . The protein extract was incubated with prewashed anti-FLAG M2 agarose gel ( A2220 , Sigma-Aldrich ) overnight at 4°C . Then , the agarose gel was washed five times with the NEBT buffer and eluted with 3 × FLAG peptide ( F4799 , Sigma-Aldrich ) . Anti-FLAG ( M20008 , Abmart ) and anti-EBP1 antibody were used for immunoblot assay to detect the immunoprecipitates . The ABA-induced phosphorylation coexpression system was established similarly as described previously [49] . Vectors of pACYCDuet-1 ( pACYC for short ) ( 71147 , Novagen ) and pRSFDuet-1 ( pRSF for short ) ( 71341 , Novagen ) were used for this system . For phosphorylation assay , pACYC-PYL1-FER , pACYC-PYL1-FERK565R , and pRSF-ABI1-EBP1 were constructed . pRSF-ABI1-EBP1 together with pACYC-PYL1-FER ( or pACYC-PYL1-FERK565R ) were transformed into BL21 E . coli . The transformed E . coli were inoculated into LB medium ( containing kanamycin and chloromycetin ) and cultured at 37°C until OD600 reached 0 . 6 . Then , 500 μM isopropyl-β-d-thiogalactoside ( IPTG ) was added to induce the protein expression for 2 hours before 50 µM ABA was added into the bacterial culture to release the FER phosphorylation activity for 10 minutes . The dephosphorylation assay was performed as described in the manufacturer’s manual using FastAP Thermosensitive Alkaline Phosphatase ( EF0651 , Thermo Fisher Scientific , USA ) . Immunoblot assay was performed to detect the phosphorylation band shift using anti-His antibody . To clearly separate the phosphorylated and dephosphorylated protein , 20% ( V/V ) glycerol was added into the PAGE gel , and the electrophoresis was performed for about 15 hours under 60 V voltage . For immunoprecipitation-phosphorylation ( IP-phosphorylation ) assay , the native EBP1 protein was immunoprecipitated using EBP1 antibody and A/G agarose , as described earlier . Four-week-old Col-0 and fer-4 plants were soaked in 1/2 MS liquid medium with or without 1 μM RALF1 for 30 minutes . Then , roots were ground to a fine powder in liquid nitrogen for immunoprecipitation assay , as described earlier . FER antibody [12] was used to detect the phosphorylation status of FER . EBP1 antibody was used for analyzing loading control . Antibodies against pSer ( ab9332 , Abcam ) and pThr ( 9381 , Cell Signaling Technology ) were used to detect the phosphorylation of EBP1 protein . We adjusted the level of total EBP1 protein to be the same among different plant samples so that the changes in phosphorylation levels of EBP1 can be easily visible . SDS-PAGE was performed , and the gel was stained by Coomassie G-250 , as described [66] . The interest bands were isolated and placed into tubes . Mass spectrometry was performed as described by Du and colleagues [23] . The band strips were first managed by destaining and dehydration . Then , after reducing and alkylating , protein was enzymolyzed by trypsase ( V511A Promega ) . Then , mass spectrometry was performed using Orbitrap , followed by LTQ . Raw data were analyzed by Xcalibur v . 2 . 1 ( Thermo Scientific , Waltham , MA , USA ) and Proteome Discoverer v . 1 . 3 beta ( Thermo Scientific , Waltham , MA , USA ) against the Arabidopsis database ( in early 2016 . 10 . 1 , found in UniProt/Swiss-Prot and UniProt/TrEMBL ) . Using an aa sequence of Arabidopsis EBP1 , genome sequences of different species were searched with the BLAST tool in the National Center for Biotechnology Information ( NCBI ) . Sequences of EBP1-like proteins from different species were downloaded from NCBI and aligned using ClustalX2 . 1 and Bioedit with default settings . A phylogenetic tree was built with MEGA5 software using the Neighbor-Joining method with the following parameters: Kimura 2-parameter; Bootstrap replications 1000; Random seed 64238 . The iTOL [67] software was used to display the phylogenetic tree . Plant materials of indicated growth stages were collected and incubated in GUS staining solution ( 50 mM sodium phosphate [pH 7 . 2] , 0 . 1% Triton X-100 , 2 mM potassium ferrocyanide , 2 mM potassium ferricyanide , 10 mM EDTA , 1 mM X-Gluc ) . The incubated samples were infiltrated under vacuum for 10 minutes . The vacuum was released slowly . Then , the samples were incubated in the GUS staining solution in the dark for 1 hour at 37°C . After removing the GUS staining solution , the materials were incubated in 70% ethanol at room temperature for about 6 hours ( the 70% ethanol was renewed several times when the solution turned chromatic ) . Then , the photos were taken under a dissecting microscope . Seedlings were treated with RALF1 in liquid medium . To avoid manipulation-related effect during plant transfer from solid to liquid medium , we transferred the seedlings into liquid 1/2 MS medium and preincubated them for 12 hours before RALF1 treatment . Then , the RALF1 peptide ( or same volume of buffer without RALF1 as mock control ) was added to the medium for the indicated time period . Root growth inhibition in response to RALF1 was recorded as described [20 , 23] , with some modifications . Seeds were germinated on 1/2 MS agar plates vertically positioned for 5 days under constant light at 23°C . Then , seedlings were aseptically transferred to 500 μL liquid medium containing 1/2 MS and 1 μM RALF1 peptides in a 24-well cluster plate ( 3524 , Costar ) . Then , the seedlings were incubated at 23°C for 48 hours with mild shaking ( 100 rpm ) before measuring . For RALF1-induced mRNA and protein accumulation assay , 7-DAG seedlings were collected and soaked in 1/2 MS liquid medium with or without 1 µM RALF1 for 2 hours or as otherwise indicated for time course . Then , the seedlings were collected for mRNA or protein extraction . Total RNA was extracted using RNAiso Plus ( 9109 , Takara ) with the method described in the manufacturer’s manual . The first-strand cDNA was synthesized using a Takara PrimeScript RT reagent Kit with gDNA Eraser ( RR047A , Takara ) . For total protein extraction , the collected seedlings were ground to a fine powder in liquid nitrogen and then resuspended and extracted in RIPA buffer ( 50 mM Tris , 150 mM NaCl , 1% NP-40 , 0 . 5% Sodium deoxycholate , 0 . 1% SDS , 1 mM PMSF , 1 × protease inhibitor cocktail , 1 × phosphatase inhibitor [pH 8 . 0] ) for 1 hour . Generally , to avoid manipulation-related effect during plant transfer from solid to liquid medium , we transferred the seedlings into liquid 1/2 MS medium and preincubated them for 12 hours before treatment . The procedure for CRD treatment was as described [68] . The 7-DAG seedlings were collected and placed in incubation buffer ( 1 mM pipes [pH 6 . 25] , 1 mM sodium citrate , 1 mM KCl , 15 mM sucrose ) , 200 μM CRD ( C3394 , Sigma-Aldrich ) , with or without 1 µM RALF1 . The samples were infiltrated under vacuum for 3 minutes in incubation buffer , followed by incubation in 23°C for indicated times . Then , plant samples were harvested for RNA extraction . Treatment of plants with MG132 and CHX was conducted as described [69] . Seven-day-old Col-0 and fer-4 seedlings grown on 1/2 MS agar plates were collected and transferred to liquid MS medium ( with or without 1 µM RALF1 ) in the presence or absence of 50 μM MG132 ( S2619 , Selleckchem ) or 100 μM CHX ( 01810 , Sigma-Aldrich ) for 2 hours . Whole-plant samples were harvested and frozen in liquid nitrogen for protein extraction and immunoblot assay . The method of polysome profiling was as described [70] . To avoid manipulation-related effect during plant transfer from solid to liquid medium , we transferred the seedlings into liquid 1/2 MS medium and preincubated them for 12 hours . Then , 7-DAG seedlings were treated with or without 1 µM RALF1 for 30 minutes and ground in liquid nitrogen , followed by resuspension in polysome extraction buffer . The extract was loaded onto a 15%–60% sucrose gradient and spun in a Beckman Optima L-100XP centrifuge at 30 , 000 rpm for 4 hours at 4°C . Sixteen fractions were collected by a gradient fractionator for RNA extraction and reverse transcription . The 40S and 80S of ribosome and polysomes were quantified by OD 260 absorbance profile . RNA was extracted using RNAiso Plus ( 9109 , Takara ) with the method described in the manufacturer’s manual . The first-strand cDNA was synthesized using a Takara PrimeScript RT reagent Kit with gDNA Eraser ( RR047A , Takara ) . Actin was used as reference gene . The procedure was as described [71] , with some modifications . Seven-DAG plant tissues ( with or without 1 µM RALF1 treatment for 2 hours ) were fixed by paraformaldehyde-based fixative . Fixed plant tissues were incubated with anti-EBP1 antibody at 4 µg/mL overnight at 4°C . The labeled tissues were probed with an IF555 Goat Anti-Mouse IgG ( GM200G-37C , Sungene ) secondary antibody for confocal observation . DAPI ( D9542 , Sigma-Aldrich ) was used for nucleus staining . The excitation wavelengths for imaging IF555 and DAPI were 561 and 405 nm , respectively . Briefly , 4-week-old rosettes of Col-0 , fer-4 were collected for nucleus fractionation . Rosettes ( with or without 1 µM RALF1 treatment for 2 hours ) were ground to a fine powder in liquid nitrogen and resuspended in 100 μL fractionation buffer as described [72] . The suspension mixture was centrifuged at 3 , 000g for 10 minutes , and the supernatant was used as cytosolic fraction . The nuclei-containing pellets were resuspended in 500 μL fractionation buffer and centrifuged at 3 , 000g for 5 minutes to wash out the cytosolic residues in the supernatant . This step was repeated five times to obtain the nuclei-containing pellet fraction . The nuclear fractions were resuspended with 100 μL RIPA buffer ( containing 1 mM PMSF , 1 × protease inhibitor cocktail , 1 × phosphatase inhibitor ) on ice for 1 hour for nucleus protein extraction . Then , aliquots of the cytosolic fraction and nucleus fraction were boiled in SDS-PAGE sample buffer for immunoblot assay . Subcellular localization of fusion proteins in transgenic Arabidopsis was performed in roots from 7-DAG seedlings grown in a vertical position on 1/2 MS medium supplemented with 0 . 8% sucrose and solidified with 1% ( w/v ) agar . RALF1 treatment was performed by transferring seedlings to liquid 1/2 MS medium containing 1 μM RALF1 peptide in 23°C for 2 hours . The excitation lines for imaging Hoechst 33258 and GFP were 405 and 488 nm , respectively . Five-DAG etiolated seedlings were collected and immersed into PI staining solution ( 0 . 01 M PBS [pH 7 . 4] , 0 . 01 μM PI ) for 2 hours . The excitation lines for imaging GFP , chlorophyll red auto fluorescence , and PI staining were 488 , 561 , and 488 nm , respectively . The quantification of fluorescent protein signal is performed as described [73] . The method of ROS burst measurement assay was as described [16] . Four leaf discs ( 4 mm in diameter ) per individual genotype were collected into 96-well plates containing sterile water and recovered overnight . The next day , the sterile water was replaced by a solution containing 20 μm/L luminol L-012 ( Sigma-Aldrich ) , 1 μg/mL horseradish peroxidase ( HRP , Sigma-Aldrich ) , and 100 nm/L flg22 , 1 μM RALF1 ( or without RALF1 in control ) . Luminescence was measured for the indicated time period using Fluoroskan Ascent FL ( Thermo Scientific ) . The proton secretion assay was performed as described [20] . Seedlings were vertically grown for 5 days on 1/2 MS solid plate . To avoid manipulation-related effect during plant transfer from solid to liquid medium , we transferred the seedlings into liquid 1/4 MS medium and preincubated them for 12 hours before reaction . Then , different genotype lines were transferred to a microwell containing 1 μM RALF1 , 200 μL 1/4 MS , 1% sucrose ( pH 5 . 8 ) and a 30 μg/mL pH indicator fluorescein-Dextran conjugate ( Sigma-Aldrich ) . Reaction was terminated at the time points shown in S14E Fig by removing seedlings from wells , and fluorescent intensity ( excitation at 485 nm wavelength , emission at 535 nm wavelength ) was recorded with Fluoroskan Ascent FL ( Thermo Scientific ) . A standard curve for pH was obtained for each time point and calculated using the 1/4 MS adjusted to pH 5 . 6 , 5 . 8 , and 6 . 2 . This procedure was performed by OE Biotech ( Shanghai , People’s Republic of China ) . Total RNA was extracted using mirVana miRNA Isolation Kit ( Ambion ) following the manufacturer protocol . RNA integrity was evaluated using the Agilent 2100 Bioanalyzer ( Agilent Technologies , Santa Clara , CA , USA ) . The samples with RNA integrity number ( RIN ) ≥ 7 were subjected to the subsequent analysis . The libraries were constructed using TruSeq Stranded mRNA LTSample Prep Kit ( Illumina , San Diego , CA , USA ) according to the manufacturer instructions . The libraries were sequenced on the Illumina sequencing platform ( HiSeqTM 2500 or Illumina HiSeq X Ten ) , and 125 bp/150 bp paired-end reads were generated . Raw data ( raw reads ) were processed using NGS QC Toolkit [74] . The reads containing poly-N and the low-quality reads were removed to obtain the clean reads . Then , the clean reads were mapped to reference genome using HISAT2 [75] . The FPKM value of each gene was calculated using cufflinks , and the read counts of each gene were obtained by htseq-count [76] . Differentially expressed genes were identified using the DESeq [77] R package functions estimateSizeFactors and nbinomTest . P value < 0 . 05 and foldChange > 2 or foldChange < 0 . 5 were set as the threshold for significantly differential expression . Genes with more than 2-fold change and P value < 0 . 05 were defined as significantly regulated genes . DAVID Functional Annotation tools [53] was used to analyze the functional annotations ( P value < 0 . 001 ) of differentially expressed genes . The raw RNA-seq data were uploaded to the NCBI database with the access number SRP151541 . The ChIP assay was performed as described [78] . To avoid manipulation-related effect during plant transfer from solid to liquid medium , we transferred the seedlings into liquid 1/2 MS medium and preincubated them for 12 hours . Then , seedlings were soaked with 1 μM RALF1 ( included in the 1/2 MS liquid medium ) or mock control ( 1/2 MS medium containing protein elution buffer , which is used for RALF1 purification ) for 2 hours and then treated with 1% formaldehyde under vacuum for 15 minutes at room temperature . Glycine was added to a final concentration of 0 . 125 M to stop cross-linking . The seedlings were washed twice with sterile water , frozen in liquid nitrogen , ground to a fine powder , and homogenized in the nuclear extraction buffer 1 ( 10 mM Tris-HCL [pH 8 . 0] , 0 . 4 M sucrose , 10 mM MgCl2 , 0 . 1 mM PMSF , and protease inhibitor [78430 , Thermo Fisher Scientific] ) . Nuclei were precipitated by centrifugation in a centrifuge at 4 , 000g for 20 minutes , washed with the nuclear extraction buffer 2 ( 10 mM Tris-HCl [pH 8 . 0] , 0 . 25 M sucrose , 10 mM MgCl2 , 1% Triton X-100 , 0 . 1 mM PMSF , and protease inhibitor [78430 , Thermo Fisher Scientific] ) , and lysed in the nuclei lysis buffer ( 50 mM Tris-HCl [pH 8 . 0] , 10 mM EDTA , 1% SDS , 0 . 1 mM PMSF , and protease inhibitor [78430 , Thermo Fisher Scientific] ) . Chromatins were sheared by sonication to approximately 500 bp . The chromatin solution was diluted 10-fold with ChIP dilution buffer ( 16 . 7 mM Tris-HCl [pH 8 . 0] , 167 mM NaCl , 1 . 1% Triton X-100 , 1 . 2 mM EDTA , 0 . 1 mM PMSF , and protease inhibitor [78430 , Thermo Fisher Scientific] ) . Anti-EBP1 antibody ( or preimmune serum as negative control ) prebound to the A/G agarose ( 20421 , Thermo Fisher Scientific ) was mixed with the chromatin solution and incubated at 4°C overnight . Immunocomplexes were precipitated and washed with four different buffers: low-salt buffer ( 20 mM Tris-HCl [pH 8 . 0] , 150 mM NaCl , 0 . 2% SDS , 0 . 5% Triton X-100 , 2 mM EDTA ) , high-salt buffer ( 20 mM Tris-HCl [pH 8 . 0] , 500 mM NaCl , 0 . 2% SDS , 0 . 5% Triton X-100 , 2 mM EDTA ) , LiCl washing buffer ( 20 mM Tris-HCl [pH 8 . 0] , 0 . 25 M LiCl , 1% NP40 , 1% sodium deoxycholate , 1 mM EDTA ) , and TE washing buffer ( 10 mM Tris-HCl [pH 8 . 0] , 1 mM EDTA ) . The bound chromatin fragments were eluted with the elution buffer ( 50 mM Tris-HCl [pH 8 . 0] , 10 mM EDTA , 1% SDS ) , and the cross-links were reversed by incubating at 65°C overnight . The mixture was treated with protease-K for 1 hour at 45°C to remove proteins . DNA was extracted by phenol/chloroform/isoamyl alcohol and precipitated with 2-fold volume of 100% ethanol at −80°C for 4 hours . To recover the DNA , it was spun at 16 , 000 rpm for 20 minutes at 4°C . The pellet was dried briefly and resuspended in 25 μL TE buffer for further real-time PCR analysis . We have screened promoters of the following 17 genes: CML38 , AT1G76650; CML40 , AT3G01830; CKX4 , AT4G29740; ERF1B , AT3G23240; ERF2 , AT5G47220; ERF098 , AT3G23230; SAUR9 , AT4G36110; SAUR41 , AT1G16510; RVE1 , AT5G17300; RVE2 , AT5G37260; RVE3 , AT1G01520; WAK2 , AT1G21270; GA2OX1 , AT1G78440; ANAC036 , AT2G17040; SIGE , AT5G24120; IPS1 , AT3G09922; EXPA16 , AT3G55500 . The EMSA was performed as described [78] , with some modified steps . EBP1-GST protein and GST protein were used for EMSA assay . The primer sequences were synthesized and labeled with FITC fluorescence probe ( TsingKe Biological Technology ) . The DNA-EBP1 binding reaction contained 100 pg probe , 100 ng EBP1-GST protein , 10 mM Tris ( pH 7 . 5 ) , 5% glycerol , 1 mM MgCl2 , 50 mM KCl , 0 . 2 mg/mL bovine serum albumin ( BSA ) , 0 . 5 mM DTT , 0 . 5 mg/mL polyglutamate , and the indicated amount of unlabeled competitor . The reactions were incubated at room temperature for 20 minutes and fractioned by electrophoresis in a 6% native polyacrylaminde gel ( acrylaminde:bisacrylamide , 29:1 ) containing 10% glycerol , 89 mM Tris ( pH 8 . 0 ) , 89 mM boric acid , and 2 mM EDTA . The FITC signal was detected after electrophoresis using KODAK 4000MM Image Station . The Dual-LUC was performed as described [78] , with modified steps . The putative EBP1-bound sequence of CML38 promoter was cloned and constructed into pGreen-0800-LUC vector as the reporter plasmid ( proCML38::LUC ) . The effector plasmid 35S::EBP1 was constructed using pBI121 vector as described above . The reporter plasmid and effector plasmid were transferred into Arabidopsis protoplast simultaneously as described above in BiFC assay . Samples were collected for the Dual-LUC using Dual Luciferase Reporter Gene Assay Kit ( RG027 , Beyotime ) . The LUC and REN signals were detected using Modulus Microplate Multimode Reader ( Turner Biosystem ) . Three biological repeats were measured for each sample , and similar results were obtained . For RALF1-treatment , 0 . 1 μM RALF1 was added into the transfected protoplasts . Then , the cells were incubated in the dark at 23°C for 16 hours for further experimentation . The in vitro protein degradation assay was performed as described [58] . Leaves from 4-week-old Arabidopsis were ground in liquid N2 and resuspended in the proteolysis buffer ( 20 mM Tris [pH 7 . 5] , 10 mM MgCl2 , 10 mM NaCl , 10 mM ATP , and 5 mM DTT ) . After centrifugation , the supernatants from Arabidopsis were mixed with EBP1-His protein ( S2C Fig ) and incubated at room temperature for 30 minutes . The reactions were stopped by boiling in SDS-PAGE sample buffer . Immunoblot assay was performed to detect the protein . Anti-His ( M20001 , Abmart ) or anti-Actin ( M20009 , Abmart ) were used to detect EBP1-His or Actin , respectively . Statistical significance was determined based on one-way ANOVA analysis using SPSS 23 . 0 software ( SPSS , USA ) .	Receptor-like kinase FERONIA ( FER ) is an important regulator of plant growth and stress response and is activated by binding its peptide ligand , rapid alkalinization factor 1 ( RALF1 ) . However , how FER , a plasma membrane–localized receptor protein , regulates gene expression in the nucleus remains unclear . Here , we show that RALF1-FER signaling increases the abundance of ErbB3-binding protein 1 ( EBP1 ) protein , which then accumulates in the nucleus and controls gene expression . The receptor kinase FER also directly interacts with and phosphorylates EBP1 , a required step for EBP1 accumulation in the nucleus . Ultimately , EBP1 protein binds to the promoters of some RALF1-FER-regulated genes and inhibits their expression , leading to a negative regulation of RALF1-FER response . This study thus reveals a link between a plasma membrane receptor and the control of gene expression in the nucleus and establishes a similar mode of action for EBP1 in both animals and plants .	[ "Abstract", "Introduction", "Results", "Discussion", "Methods" ]	[ "phosphorylation", "messenger", "rna", "brassica", "green", "fluorescent", "protein", "luminescent", "proteins", "model", "organisms", "experimental", "organism", "systems", "seedlings", "plants", "arabidopsis", "thaliana", "research", "and", "analysis", "methods", "anima...	2018	EBP1 nuclear accumulation negatively feeds back on FERONIA-mediated RALF1 signaling
When a flashed stimulus is followed by a backward mask , subjects fail to perceive it unless the target-mask interval exceeds a threshold duration of about 50 ms . Models of conscious access postulate that this threshold is associated with the time needed to establish sustained activity in recurrent cortical loops , but the brain areas involved and their timing remain debated . We used high-density recordings of event-related potentials ( ERPs ) and cortical source reconstruction to assess the time course of human brain activity evoked by masked stimuli and to determine neural events during which brain activity correlates with conscious reports . Target-mask stimulus onset asynchrony ( SOA ) was varied in small steps , allowing us to ask which ERP events show the characteristic nonlinear dependence with SOA seen in subjective and objective reports . The results separate distinct stages in mask-target interactions , indicating that a considerable amount of subliminal processing can occur early on in the occipito-temporal pathway ( <250 ms ) and pointing to a late ( >270 ms ) and highly distributed fronto-parieto-temporal activation as a correlate of conscious reportability . One of the most obvious and yet unexplained properties of conscious perception is the existence of a threshold for conscious access: when a stimulus is flashed and followed by a backward mask , subjects do not report perceiving it until the target-mask interval exceeds a threshold duration [1 , 2] . Below-threshold—or “subliminal”—stimuli receive complex perceptual and even semantic processing [3–6] , but for an unknown reason , these processes remain inaccessible to consciousness . Understanding the neural mechanisms that distinguish such conscious and nonconscious processes remains a crucial issue in cognitive neuroscience . We used high-density recordings of event-related potentials ( ERPs ) to ask several questions: ( 1 ) What sequence of activations is evoked by subliminal masked stimuli ? ( 2 ) What additional sequence of brain events leads a stimulus to cross the threshold for conscious reportability ? ( 3 ) At what time does this access to conscious report occur ? Existing models of conscious access differ markedly with respect to the brain areas involved ( posterior versus anterior ) and the timing of their activation ( early versus late ) . A first category of model views conscious visual perception as a phenomenon localized to posterior brain areas , and whose contents are determined by the pattern of neuronal activity in early visual and/or occipito-temporal areas [7–13] . According to these proposals , the threshold for conscious perception during masking should be determined solely within the visual cortex , either within a single area or due to short-range recurrent interactions among posterior occipito-temporal regions [10] . As far as timing is concerned , some authors have proposed that conscious perception is already detectable in the ascending , feedforward activation evoked as early as ∼100 ms after stimulus presentation [7] . For others , visual consciousness is not associated with feedforward visual activation , but requires a subsequent period of “localized recurrent processing” [10] , still relatively early and confined to posterior occipito-temporal brain systems . Electrophysiological recording in macaque monkeys during masking suggest a peak effect of recurrent interactions in area V1 around 100–140 ms after stimulus presentation [9] , and intracranial human recordings suggest that category-specific ventral occipito-temporal cortices are already strongly activated by 150–200 ms , although extension to more anterior regions and reverberation effects can extend for much longer ( 290–700 ms ) [14] . Thus , although predictions concerning timing are less precise , this category of models would predict that subjective reports of conscious perception should correlate with posterior and relatively early brain events . At odds with this view , a second category of models views conscious access as the formation of a late brain-scale neuronal assembly involving recurrent long-distance interactions among distributed thalamo-cortical regions , particularly the prefrontal cortex and higher cortical association areas [15–28 , 66] . In a detailed neural network simulation , those areas , when linked by reciprocal top-down and bottom-up connections , exhibit a threshold for “global ignition” [19 , 20] . When this threshold is exceeded , even a brief external stimulation can simultaneously activate many distant areas and yield a long-lasting pattern of reverberating activity . It is claimed that such a distributed pattern corresponds to a consciously reportable state , because its active contents are broadcast to many specialized processors , including those for verbal or motor report . When an incoming activation fails to exceed the ignition threshold , it can still briefly propagate through the processors but quickly vanishes , because it is not supported by recurrent self-amplifying loops . This state may correspond to a situation of subliminal processing [21] . Figure 1 illustrates the schematic predictions that may be expected from this “global neuronal workspace” model as a stimulus is made increasingly more visible by lowering its masking strength ( for detailed simulations , see [19–21] ) . First , initial visual activation should be nearly identical for all stimuli , even heavily masked ones , irrespective of whether they are perceived or not . Second , as masking strength decreases , activation should propagate to increasingly deeper levels of processing in occipito-temporal and parietal cortices . Third , when reaching higher cortical areas , particularly the prefrontal cortex , activation should diverge in a nonlinear manner , either quickly building up to a high level ( supra-threshold stimulus ) , or decaying back to baseline ( subliminal stimulus ) . Fourth , the surge of activation to supra-threshold stimuli should occur simultaneously in a global network , including prefrontal , parietal , as well as posterior occipito-temporal regions , thus creating a second , late peak of activation in early visual areas . To test these predictions , we collected high-density recordings of ERPs during a backward masking paradigm [2] in which a single brief ( 16 ms ) parafoveal target digit was followed by a mask consisting of letters surrounding the target location ( see Figure 2 ) . We varied quasi-continuously the target-mask stimulus onset asynchrony ( SOA ) so that , with increasing SOA , the same target stimulus crossed the threshold from subliminal processing to conscious perception . We documented this transition behaviorally using objective ( forced-choice comparison of the digit to five ) and subjective ( continuous scale of visibility ) measures of conscious access . Both the proportion of “seen trials” and the objective performance increased nonlinearly as a function of SOA , thus tracing a characteristic sigmoidal curve with a well-defined threshold ( Figure 2 ) . Recordings of ERPs evoked by these stimuli then allowed us to examine how brain activity changed around this transition point . By subtracting out the brain activity evoked by the mask only , we isolated the entire sequence of target-evoked ERPs , thus allowing us to visualize both early visual activation and later brain events . To determine which of these components were associated with conscious-level processing , we used two independent criteria . First , we searched for ERPs whose profile of amplitude as a function of SOA traced a nonlinear curve parallel to the proportion of “seen” trials . In other words , we used this nonlinear profile of subjective visibility ratings as a “signature” of conscious processing in the brain ( as done in functional MRI ( fMRI ) by Haynes et al . [29] for metacontrast masking ) . As a second criterion , we also examined , at a fixed intermediate SOA value , the difference between trials reported as “seen” and trials reported as “not seen . ” The central issue was whether those two criteria ( nonlinear signature and difference between seen and not-seen reports ) pointed to an early activation localized to occipito-temporal areas or to a late global ignition in prefrontal and parietal areas . We first evaluated the subjects' objective and subjective perception of masked digits ( Figure 2B–2E ) . Figure 2E illustrates how subjective visibility ratings neatly clustered into two well-defined “seen” and “not-seen” states . When the target was absent , subjects hardly ever gave subjective ratings exceeding 25% of the scale ( 0 . 3% of trials ) . Conversely , at the longest SOA where the target was highly visible , the ratings almost always exceeded 25% of the scale ( 97 . 4% of trials ) . At intermediate SOAs , the distribution of cursor positions reached a minimum around 25% , separating the data into two sets , one peaking at visibility zero ( “not-seen” trials ) , and the other at maximal visibility ( “seen” trials ) , replicating our earlier observations [2 , 30 , 31] . Thus , the value of 25% visibility was used as a cut-off between seen and not-seen trials . We then calculated for each SOA and each subject the percentage of seen trials ( Figure 2C ) . The percentage of seen trials increased significantly as a function of SOA ( F ( 5 , 55 ) = 154 . 17 , p < 0 . 001 ) . The increase was nonlinear and showed a quick transition between 33 and 66 ms from mostly not-seen to mostly seen responses . Indeed , there was a significant difference ( p < 0 . 01 ) between the rate of increase of subjective visibility ratings in the SOA interval from 33 to 66 ms and that in the 16–33 ms and 66–83 ms intervals , indicating a significant nonlinearity . For each subject , the curve relating the percentage of seen trials to SOA was well fitted by a sigmoid . The subjective visibility threshold was defined as the SOA where the sigmoid reached its inflexion point ( for mathematical details , see Materials and Methods and [2] ) . The mean subjective threshold was 43 . 9 ms ( standard deviation [SD]= 10 . 5 ms ) . Similarly , in the objective task , performance in comparing the target with the numeral 5 increased with SOA ( F ( 5 , 50 ) = 82 . 02 , p < 0 . 001; see Figure 2B ) . This performance was at chance level only for the shortest SOA of 16 ms . Again , the rate of increase was larger in the SOA interval from 33 to 66 ms than in the 16–33 ms and 66–83 ms intervals ( p < 0 . 05 ) . A sigmoid provided a good fit and allowed us to define an objective visibility threshold for each subject ( mean = 40 . 8 ms; SD = . 12 . 4 ms ) . As in our previous work [2] , a strong linear correlation was found between objective and subjective thresholds ( r2 = 0 . 747 , slope = 1 . 03 ) , once a single outlier subject was excluded . This indicated that subjective and objective measures followed a similar nonlinear transition with a single threshold parameter which varied across subjects . To further study the relation between subjective and objective measures of conscious access , we subdivided the subjective visibility scale into four bins according to the position of the cursor ( 1st , 2nd , 3rd , and 4th quarters of the scale , with the 1st quarter corresponding to the above “not-seen” category ) . We then computed the objective performance for the different SOA within each of these four subjective bins ( Figure 2D ) . There was a good cross-validation of the objective and subjective measures inasmuch as objective performance was always way above chance at all SOAs whenever trials were classified as “seen” ( bins 2 , 3 , and 4 ) . Conversely , in bin 1 corresponding to not-seen trials , objective performance was at chance level for SOAs 16 and 33 ms ( p = 0 . 11 and p = 0 . 059 , respectively ) , but above chance for longer SOAs 50 and 66 ms ( p < 0 . 001 ) . Those results indicated that on some trials , subjects could compare accurately the target to the numeral 5 even if they reported that they could not see it . Such above-chance performance in the absence of conscious reportability is reminiscent of “blindsight” and is indicative of subliminal processing , as previously reported in subliminal priming tasks [5 , 32 , 33] . In summary , both objective and subjective data followed a sigmoidal curve as a function of SOA , with a similar nonlinear threshold ( defined as the inflection point of the sigmoid ) . We take this parallelism between objective and subjective measures of conscious access as an indication that a major transition in processing occurs around SOA = 50 ms , and this transition affects a broad range of cognitive processes , including those leading to introspective reports as well as those leading to simple numerical decisions . We are mostly concerned with the neural correlates of this major transition , which is captured by both objective and subjective measures . Note that our definition of the conscious threshold departs from the classical psychological definition , which is based on the SOA at which objective performance first departs from chance . Indeed , a partial dissociation was found at short SOAs , where we observed a proportion of subjectively not-seen trials accompanied by above-chance objective performance . This dissociation is compatible with the notion of a distinct “objective threshold” within the zone associated with an absence of subjective perception [34] . Unfortunately , however , there were too few trials in this “gray zone” between objective and subjective definitions of conscious perception to address the issue of its underlying brain mechanisms . We focused our analysis on the correlates of the subjective threshold and , more precisely , on the issue of which ERP components exhibited a nonlinear curve parallel to the subjective reports . As noted above , behavioral visibility ratings increased suddenly from SOA = 33 ms to SOA = 66 ms , whereas they varied little either from 16–33 ms or from 66–83 ms . This specific shape was used as a litmus test to identify which ERP components correlated with conscious access . As a second criterion , we also examined which ERP components distinguished seen and not-seen subjective reports at the threshold SOA of 50 ms . Our procedure isolated the brain activity evoked by a masked visual stimulus and examined how it was affected by backward masking at variable target-mask SOAs . Although several ERP components were affected by backward masking , two criteria suggested that the P3 component was most tightly associated with subjective perception: ( 1 ) its amplitude as a function of SOA exhibited the same sigmoidal shape as the fraction of seen trials; and ( 2 ) it showed a significant difference between seen and not-seen trials at a fixed SOA . Source analysis suggested that the underlying cerebral mechanism was a sudden activation of a distributed bilateral fronto-parieto-temporal network starting about 270 ms after stimulus onset . In the preceding period ( 140–270 ms ) , the results revealed the progressive build-up , in posterior occipito-temporal and parietal areas , of a nonlinear divergence of activation as a function of SOA , which points to a dynamic nonlinear amplification as the neural correlate of the masking threshold . A first important outcome of our experiment is to characterize further the fate of subliminal masked stimuli . Our results confirm that nonreportable visual stimuli can propagate through a series of cortical processing stages [1 , 3–6] . As suggested by several previous behavioral experiments [38 , 39] , they indicate that the depth of processing depends on the target-mask interval ( SOA ) . At the shortest SOA ( 16 ms ) , there was a strong reduction in all ERP components ( though not reaching significance for the P1a ) . This observation suggests that masking at such short target-mask interval may already partially occur at a peripheral or subcortical level , thus leaving little target-induced activation except in early occipital cortex . As soon as SOA reached 33 ms or more , however , we observed strong subliminal activation in the contralateral occipito-temporal pathway ( particularly the mid-ventral temporal cortex , where the activation corresponding to scalp N1 had the same intensity for all SOAs > 16 ms ) , with a small spread to ipsilateral temporal and bilateral parietal cortices . It could be argued that at SOA = 33 ms , the averaged ERPs were contaminated by a non-negligible proportion of seen trials , because the sigmoidal curves for objective and subjective reports had already taken off at this SOA value . Additional analyses , which are presented in Figure S3 , however , demonstrated that the results were essentially unchanged when excluding the minority of seen trials and restricting the analysis to not-seen trials at SOA = 33 ms . Thus , early visual processing indexed by the P1a and N1 is essentially unaffected even on trials subjectively rated as invisible . This conclusion is compatible with previous single-cell recordings during backward masking paradigms , which revealed preserved selective firing in occipital [9 , 40 , 41] and inferior temporal cortices in response to heavily masked stimuli [42 , 43] . They also corroborate previous human ERP data , demonstrating preservation of P1 and N1 waveforms during backward masking and attentional blink paradigms [30 , 35–37] , Similarly , recent fMRI studies of backward masking paradigm have shown that early visual activity need not be reduced under conditions of stimulus invisibility [29 , 44] . Even under conditions of strong metacontrast masking , visual activation is maintained in areas V1 and V2 [12] and contains sufficient information to allow partial stimulus decoding [6] . By sorting trials using subjective reports of “seen” versus “not seen , ” we also examined the cerebral processing of stimuli presented for an even longer target-mask interval ( 50 ms ) and yet still reported as being invisible . Those result ( Figure 7 ) indicated that at these lags , an even longer-lasting subliminal activation , extending beyond early visual processing , can be induced . Indeed , all ERP components , including the P3 , showed a significant subliminal activation relative to target-absent trials . Nevertheless , although activation amplitudes were initially identical between seen and not seen stimuli before 270 ms , this activation then decreased in amplitude relative to seen trials ( Figure 7 ) , and cortical source models indicated that it dropped toward zero within about 500 ms ( Figure S2 , green curve , lower panel ) . We thus conclude that a brief subliminal stimulus can produce a transient activation in many areas , even those associated with the scalp P3 waveform , but that this activation is small and brief . This conclusion fits with several previous observations of anterior brain activation under subliminal conditions , including frontal eye field activation by masked stimuli in monkeys [45] , anterior negativity evoked by undetected errors [46] , and frontal and cingulate P300 responses evoked by subthreshold visual oddballs [47 , 48] . In these reports as in the present data , the subliminal activation was always strongly reduced compared to the supraliminal case . As discussed further below , a minimal threshold level of activation seems to be needed to cause a global and sustained ignition associated with conscious perception . A second contribution of the present results is to visualize some of the cortical interactions by which a retrograde mask can interfere with a preceding visual target and prevent its conscious perception . Masking must be caused by an interference of the mask on the brain activity induced by the target , but its locus remains debated [1 , 9 , 42 , 43 , 49] . Our source localization data provide a fine characterization of the interactions between target and mask . We first identified an early local competition reflected in the target- and mask-evoked N1 components: as SOA increases , the target N1 increases in amplitude whereas the mask N1 decreases . Source analysis indicates that this antagonistic relation occurs in contralateral occipital , posterior parietal , and posterior ventral temporal cortices . This result is compatible with the hypothesis of a competition between the early visual events evoked respectively by the target and mask , as postulated in many models of masking [1 , 49] . At a later stage , a similar antagonistic relation occurs at the level of the N2 waveforms evoked by the target and the mask , and this is associated with a distributed bilateral activation of ventral occipito-temporal and posterior parietal cortex . We found that the target-evoked N2 starts at a fixed latency relative to target presentation and continues until it is interrupted by mask-induced activity . Interestingly , neurophysiological recordings in macaque inferior temporal cortex ( IT ) during a picture masking paradigm have identified single neurons with a similar response profile [42] . Exactly as in the present N2 component , these neurons start firing selectively at a fixed latency after picture onset , but they suddenly cease firing at a fixed latency after mask onset . The N2 might thus represent an ERP signature of this neuronal activation of IT cortex . Both findings indicate that in spite of their brief initial presentation ( 16 ms ) , visual stimuli can evoke durable activation in a posterior network of areas . The mask presentation has the effect of terminating this activation . This effect of the mask is compatible with interruption models of masking , according to which the presentation of the mask erases the cortical representation of the target by suddenly interrupting target-induced ongoing activity in a visual buffer [1 , 49 , 50] . The N2 may thus index a stage of cortical processing at which the representation of a brief visual target progressively gains in strength within a distributed set of posterior occipital , temporal , and parietal areas as masking strength decreases . Yet the N2 still fails to meet our two criteria for a genuine correlate of conscious access: ( 1 ) its activation increases linearly rather than sigmoidally with SOA , and ( 2 ) at at fixed SOA of 50 ms , it does not differ for seen versus not-seen trials . Thus , we conclude that this waveform does not yet constitute a correlate of conscious access stricto sensu . The fact that it shows a relatively high level of activation at short SOAs of 33 and 50 ms ( Figure 4 ) indicates that its activation strength is not always predictive of conscious access . Our two criteria for conscious access were only met at a later time period , ranging from 270 ms to about 400 ms , and were associated with a broad fronto-parieto-temporal network . During this late time period , which coincides with the scalp-evoked P3 waveform , ( 1 ) the neural activity evoked by masked stimuli varies nonlinearly with SOA in parallel to the sigmoidal behavioral curve and ( 2 ) its amplitude differs between seen and not-seen trials . This conclusion is nicely corroborated by a recent study of metacontrast masking using magnetoencephalography [51] . Van Aalderen et al . compared two masking procedures , one of which lead to an absence of subjective conscious perception ( effective mask ) whereas the other did not ( pseudomask ) . In both cases , the target-mask SOA was varied in small steps , as in the present study . While several MEG components varied with SOA , it was not until a late time window ( 290–390 ms ) that a significant correlate of conscious access was observed , in the form of a U-shaped activation curve only found in the effective mask condition and tightly paralleling subjective reports . The hypothesis of a late correlate of conscious access also meshes well with a previous experiment from our laboratory using a different paradigm for rendering stimuli invisible , the attentional blink [30] . In this study , Sergent et al . contrasted the neural activity evoked by words which were either consciously perceived or attentionally “blinked” by a concurrent task . When comparing seen and not-seen trials , a critical transition in neural activity was also found from 270 ms after stimulus presentation . As in the present study , the transition was also preceded by an intermediate period of divergence between 170 and 300 ms , and was followed by an all-or-none spreading of activity to a distributed fronto-parieto-temporal network similar to the present one . Sergent et al . only contrasted ERPs in two quite distinct states: during the attentional blink ( T1-T2 lag = 258 ms ) , or outside of it ( T1-T2 lag = 688 ms ) . By contrast , the present masking experiment varied the proportion of seen trials quasi-continuously by manipulating the target-mask interval , thus allowing us to probe the transition to conscious access in a parametric manner . This design revealed that several ERP components such as P1b and N2 , occurring before the late global P3 , are already partially correlated with subjective reports , although their shape does not exactly match the nonlinearity of conscious access . Those observations suggest that the processing leading to conscious access for masked stimuli is not instantaneous , but follows a progressive dynamics of target-mask competition extending over a large duration ( about 140–270 ms post-stimulus ) and ultimately resulting in the P3 . One interpretation of these data could be that all components prior to the P3 reflect a progressive process of conscious access , whereas the P3 characterizes the final dichotomous result of this process ( “seen” or “not seen” ) . This hypothesis , while clearly speculative , may reconcile the apparent all-or-none character of conscious reports [31] with the continuous and cumulative character of brain dynamics . As demonstrated in computer simulations of a global neuronal workspace [19 , 20] , highly interconnected thalamo-cortical networks , although evolving continuously over time , may present a dynamical phase transition leading them , over a brief divergence period , into one of two radically distinct states ( either global ignited or quickly decaying ) . Although the transition seems abrupt and its end result is sharply defined ( essentially all or none ) , studies at a high temporal resolution , as performed here , reveal that the transition is characterized by a quick succession of transient intermediate states . The localizations that we obtained from ERP source reconstructions suggest that the P3 relates to the activation of a highly distributed network with key nodes in inferior frontal , posterior parietal , and anterior temporal regions , as well as a joint amplification of activation in posterior occipito-temporal regions . Importantly , this late global activation occurs simultaneously in both hemispheres , regardless of the initial hemifield of target presentation . This observation fits with the hypothesis that conscious access is associated with a breakdown of local “modular” processing and the broadcasting of accessed information to many bilateral cortical regions through long-distance cortico-cortical connections including those of the corpus callosum [17] . The inferred localizations must be taken cautiously , because they represent an indirect inference based on the choice of one out of many possible cortical activity patterns compatible with scalp recordings . Thus , it will be important to cross-validate the present findings with anatomically more accurate methods . In that respect , it is noteworthy that our cortical source modeling solutions are highly compatible with intracranial recordings which indicate a distributed pattern of sources of the scalp P3 and , crucially , its bilateral origin irrespective of the stimulated hemifield [52] . With fMRI , Haynes , Driver , and Rees [29] used a parametric method similar to ours in order to search for brain activity patterns that covaried with stimulus visibility ( which , in their case , traced a U-shaped curve ) . They found a correlation of conscious reports with activity in extrastriate visual cortex and in distant fronto-parietal regions . A similar fronto-parietal engagement during conscious access to visual stimuli has been described in several other fMRI studies [18 , 27 , 53–55] . At the single-cell level , neurophysiological recordings in the macaque monkey during a threshold tactile detection task have also reported a progressive increase in the correlation of neural activity with perceptual judgments [56] and a tight correlation of trial-by-trial subjective reports with late frontal activity [57] . In the present backward masking paradigm , conscious access was associated with the late activation of a broadly distributed cortical network , starting at a latency of ∼270 ms . Those findings are incompatible with models postulating that local amplitude modulations—confined to early striate , extrastriate , or occipito-temporal cortices—constitute the necessary and sufficient conditions for consciousness [7 , 11 , 12] . We do not dispute the fact that early extrastriate differences between masked and unmasked stimuli can be found , both in event-related potentials [7 , 17 , 58 , 59] and in fMRI [7 , 12 , 13 , 17 , 60 , 61] . Indeed , the present results confirm that early posterior components such as P1b and N2 are modulated by masking . However , by measuring visibility on a trial-by-trial basis over an entire range of SOAs , we found that these early ERPs ( 1 ) can occur without conscious perception , ( 2 ) do not exhibit the signature sigmoidal shape of subjective reports , and ( 3 ) do not differ for seen and not-seen targets . Note that these conclusions might have been missed , and we might have concluded erroneously that these early components correlate with conscious perception , if we had only compared two extreme states ( heavily masked versus lightly masked ) , as was done in many studies . We therefore suggest that in brain-imaging studies of masking and conscious access , it is essential to manipulate masking strength quasi-continuously and to base inferences on the entire activation profile [29 , 51] . Lamme and colleagues [8–10] have proposed that conscious perception is not associated with the first feedforward pass of activation in visual cortex , but with a later feedback reverberation . They hypothesize that “localized recurrent processing” [10] associated with short-range interactions among posterior occipito-temporal brain systems is the primary mechanism of perceptual awareness . Our results are only partially compatible with this notion . We did observe a progressive build-up of activation in posterior areas as the strength of masking decreased , as indexed by the P1b and N2 . Our source reconstructions suggest that the ipsilateral P1b , which correlated with subjective report , reflects a reverberation of visual activity in bilateral extrastriate cortices , while the later N2 component may be associated with recurrent processing in a broader though nonglobal occipito-temporo-parietal network . These results are therefore compatible with a reverberation of activation within an increasingly global network that starts within posterior visual areas [8 , 10] , However , our results also indicate that , while those early components may contribute to the subsequent transition toward conscious access or to its failure , they do not yet correspond to a full-blown conscious state . Our observations are most consistent with the theory of a global workspace formed by multiple distant associative areas including prefrontal , parietal , and temporal cortices [16 , 19 , 20] . In Figure 1 , we outlined the predictions derived from this model concerning the effect of decreasing masking strength . Those main predictions were: ( 1 ) initial visual activation unaffected by masking; ( 2 ) progressive increase in activation depth as masking strength decreases; ( 3 ) non-linear divergence at a late stage; and ( 4 ) global reverberation simultaneously engaging a distributed set of prefrontal , parietal , and posterior occipito-temporal regions . All of these predictions were supported by the data . Most importantly , cortical source modeling clearly indicated a strong contribution of inferior frontal and posterior parietal cortices to the observed correlate of conscious access . This is the single observation that most clearly supports the proposed global workspace model relative to Lamme et al . 's proposal of “localized recurrent processing . ” We conclude with a brief consideration of the limits of the present work . First , our conclusion rests on the absence of early differences in brain activity correlating with subjective reports . We therefore cannot exclude that a more sensitive method such as single-cell recording or phase synchrony analysis would reveal consciousness-related events prior to the present suggested onset time of 270 ms [58] . Second , we used multiple statistical comparisons to compare various ERP components and isolate the P3 as the sole correlate of subjective reports . Since the corresponding statistics were weakly significant ( see Table 1 ) , our conclusions may be affected by a type I error ( i . e . , the P3 effect may be a false positive ) . However , the P3 effect was found statistically significant by two independent analyses ( difference between seen and not-seen trials at SOA = 50 ms and sigmoidal shape as a function of SOA ) . Furthermore , the effect was predicted by our model as well as by several previous studies [30 , 51] . For these reasons , a type I error seems unlikely , though not fully ruled out . A third limitation is that we cannot yet ascertain whether the late ERP events that we have identified are all necessary for conscious perception—it remains possible that another paradigm , perhaps side-stepping the need for overt report , or using only an objective criterion for perceptual awareness , would reveal that conscious perception can occur without such global activity or with only a subset of the distributed cortical regions observed here [10 , 62] . Fourth , correlation is not causation . We only report a correlation between subjective perception and late ERPs , but another method capable of interfering with brain activity patterns , such as transcranial magnetic stimulation [63] , would be needed to establish whether these late events play a causal role in the conscious state . Finally , the global neuronal workspace model , while providing a good fit to the observed dynamics of conscious access , remains underspecified at both anatomical and functional levels . In the future , it will be essential to provide a more precise specification of the brain areas associated with conscious-level processing and their respective computational roles . The present research only begins to narrow down the search for the mechanisms of conscious access .	Understanding the neural mechanisms that distinguish between conscious and nonconscious processes is a crucial issue in cognitive neuroscience . In this study , we focused on the transition that causes a visual stimulus to cross the threshold to consciousness , i . e . , visibility . We used a backward masking paradigm in which the visibility of a briefly presented stimulus ( the “target” ) is reduced by a second stimulus ( the “mask” ) presented shortly after this first stimulus . ( Human participants report the visibility of the target . ) When the delay between target and mask stimuli exceeds a threshold value , the masked stimulus becomes visible . Below this threshold , it remains nonvisible . During the task , we recorded electric brain activity from the scalp and reconstructed the cortical sources corresponding to this activity . Conscious perception of masked stimuli corresponded to activity in a broadly distributed fronto-parieto-temporal network , occurring from about 300 ms after stimulus presentation . We conclude that this late stage , which could be clearly separated from earlier neural events associated with subliminal processing and mask-target interactions , can be regarded as a marker of consciousness .	[ "Abstract", "Introduction", "Results", "Discussion" ]	[ "neuroscience", "homo", "(human)" ]	2007	Brain Dynamics Underlying the Nonlinear Threshold for Access to Consciousness
Infections by Neisseria meningitidis show duality between frequent asymptomatic carriage and occasional life-threatening disease . Bacterial and host factors involved in this balance are not fully understood . Cytopathic effects and cell damage may prelude to pathogenesis of isolates belonging to hyper-invasive lineages . We aimed to analyze cell–bacteria interactions using both pathogenic and carriage meningococcal isolates . Several pathogenic isolates of the ST-11 clonal complex and carriage isolates were used to infect human epithelial cells . Cytopathic effect was determined and apoptosis was scored using several methods ( FITC-Annexin V staining followed by FACS analysis , caspase assays and DNA fragmentation ) . Only pathogenic isolates were able to induce apoptosis in human epithelial cells , mainly by lipooligosaccharide ( endotoxin ) . Bioactive TNF-α is only detected when cells were infected by pathogenic isolates . At the opposite , carriage isolates seem to provoke shedding of the TNF-α receptor I ( TNF-RI ) from the surface that protect cells from apoptosis by chelating TNF-α . Ability to induce apoptosis and inflammation may represent major traits in the pathogenesis of N . meningitidis . However , our data strongly suggest that carriage isolates of meningococci reduce inflammatory response and apoptosis induction , resulting in the protection of their ecological niche at the human nasopharynx . Neisseria meningitidis is the causative agent of meningococcal meningitis and fulminant sepsis . It is a common inhabitant of the human nasopharynx , being asymptomatically carried by approximately 10% of the population , worldwide . The incidence of meningococcal disease varies from 1 to 50 cases per 100 , 000 . The reported fatality rate in meningococcal meningitis is about 10% and up to 20% of survivors suffer from sequelae [1] . While in industrialized countries , meningococcal disease occurs usually as sporadic cases , large epidemics occur periodically in the “meningitis belt” of sub-Saharan Africa ( from Senegal to Ethiopia ) . There is increasing evidence that invasive meningococcal infections lead to cytopathic effects that often result in marked tissue and cell damage mainly characterized by the breakdown of cell tight junctions , the deterioration of the cell layers , changes in normal cell morphology , and loss of cells [2]–[8] . These observations are consistent with the extensive tissue damage and cell death seen in autopsy material from fatal human cases [9] . Tissue damage may occur through apoptosis and/or necrosis . Various components of the bacterial cell wall are capable of activating proinflammatory response , notably the meningococcal lipooligosaccharide ( LOS ) , the major structural component of the outer membrane . Apoptosis plays an important role in the pathogenesis of several infectious agents that induce or block this process [10]–[14] . Pathogenic Neisseriae have been shown to induce the expression of apoptosis-related genes and to trigger apoptosis in different cell types [15]–[19] . However , the mechanisms leading to this programmed cell death are still poorly understood . Interestingly , conflicting evidences exist on the ability of the neisserial porin PorB to induce or to protect cells from apoptosis [16] , [17] , [20] . Moreover , meningococci-cell interaction is complicated by the high variability of meningococcal isolates . Multilocus sequence typing ( MLST ) analysis classifies meningococcal isolates according to polymorphisms in seven housekeeping genes . Meningococcal isolates can be clustered into clonal complexes comprising closely related isolates varying by no more than two loci [21] . Molecular epidemiology studies comparing the overall diversity between the pathogenic and carriage population , show that isolates from asymptomatic carriage are more diverse than those from invasive disease [22] , [23] . Despite the diversity of carried meningococci , only a limited number of clonal complexes , termed the hyper-invasive lineages , are responsible for most reported disease [22]–[24] . Among these clonal complexes , isolates belonging to the clonal complex ST-11 seem to be significantly associated with the disease and rarely found in carriers [22] . Moreover , we have recently shown that isolates of the ST-11 clonal complex positively correlated with fatal outcome , a higher virulence for mice and a higher damage to human epithelial cells , disregarding their serogroups [25] . The aim of this work was to explore the apoptosis pathway induced by hyper-invasive ST-11 isolates of N . meningitidis in comparison to non invasive carriage isolates . We selected ten pathogenic meningococcal isolates of the sequence type ST-11 ( ST-11/ET-37 complex ) that belong to different serogroups ( B , C , and W135 ) . Eight carriage isolates of different serogroups belonging to different clonal complexes were also included ( Table 1 ) . We first explored the cytopathic effect of these isolates to Hec-1-B epithelial cells . All disease isolates were cytopathic to Hec-1-B cells particularly after 9 h of infection , regardless their serogroups . In contrast , none of the carriage isolates tested was cytopathic ( Table S1 ) . This cytopathic effect was mainly due to induction of apoptosis in infected cells . Indeed , all the ST-11 pathogenic isolates significantly induced apoptosis in infected cells as estimated by FITC-conjugated Annexin V staining and flow cytometry ( FACS ) analysis ( Figure 1 and Table S1 ) . However , FITC-conjugated Annexin V staining in cells infected with carriage isolates was comparable to that of uninfected control cells . All the eighteen tested isolates grew similarly in the medium of infection ( data not shown ) . This excludes a possible effect of differential bacterial growth on apoptosis . The induction of apoptosis was further confirmed by fluorescence microscopy . Hec-1-B cells were infected with the pathogenic isolate LNP19995 or the carriage isolate LNP21019 , both expressing the red fluorescent protein DsRed , and were stained with FITC-Annexin V 9 h post-infection . As a positive control , cells were treated with 1 µM staurosporine ( STRP ) . We found that similarly to STRP , LNP19995 efficiently induced cells to become apoptotic as revealed by FITC-Annexin V staining ( Figure 2A ) . In contrast , Hec-1-B cells infected with the carriage isolate LNP21019 were resistant to staining with Annexin V after the same period of infection ( Figure 2A ) . To further confirm that this alteration was due to apoptotic cell death , a number of apoptosis-specific assays were performed . Caspase-3 activity was detected in cell lysates of Hec-1-B infected with the ST-11 pathogenic isolates . This activity was blocked by the caspase-3-specific inhibitor DEVD ( N-acetyl-Asp-Glu-Val-Asp ) . In contrast , a basal level of caspase-3 activity similar to uninfected cells was observed in cells infected with the non-cytopathic isolates ( Figure 2B and Table S1 ) . Further evidence of apoptosis was provided by the distinct presence of DNA laddering effect when genomic DNA was resolved by agarose gel electrophoresis . This laddering effect was absent in cells infected with carriage isolates as in untreated control cells ( Figure S1 ) . Other epithelial cell lines behave similar to Hec-1-B cells . Indeed , the pathogenic isolate LNP19995 but not the carriage isolate LNP21019 , was able to induce apoptosis in A549 and HEp-2 human epithelial cell lines as revealed by FITC-Annexin V staining ( Figure S2 ) . These results suggest that the apoptotic effect of the pathogenic isolates towards infected cells is not only restricted to Hec-1-B epithelial cell line . Taken together , our data suggest that pathogenic meningococcal ST-11 isolates induced apoptosis as a major cytopathic mechanism to infected epithelial cells . Isolates used in this study ( both pathogenic and carriage isolates ) are piliated and capsulated as evaluated by specific sera against pili and capsule ( data not shown ) . We therefore explored the impact of these structures on the induction of apoptosis . All tested pathogenic isolates were adherent to Hec-1-B cells ( Table 1 ) . However , carriage isolates were heterogeneous in their adhesiveness to Hec-1-B epithelial cells . These observations suggest that adhesion could be necessary but insufficient to establish a cytopathic effect . We next compared induction of apoptosis in human Hec-1-B epithelial cells upon infection with the wild-type pro-apoptotic isolate LNP19995 as well as its isogenic pilus-deficient mutant NM0706 , using flow cytometry following Annexin V staining . While the pilus-deficient mutant failed to induce significant apoptosis , wild-type LNP19995 triggered 5 fold increase of phosphatidylserine asymmetry breakdown as detected by an increase in FITC-Annexin V binding within 9 h of infection . Centrifugation of bacteria to infected cells conferred apoptotic properties to NM0706 strain ( Figure 3A and 3B ) , suggesting that cell contact rather than pili expression is necessary to induce apoptosis . Moreover , no difference in the induction of apoptosis was observed in the presence of 10 µM cytochalasin D that blocks bacterial invasion to Hec-1-B cells , suggesting that invasion is not involved in this process . An isogenic capsule deficient mutant ( NM0707 ) was also compared to its parental strain LNP19995 and showed a higher level of induction of apoptosis in Hec-1-B cells than the parental strain ( Figure 3A and 3B ) . In contrast , the capsule-deficient isogenic mutant of the carriage isolate LNP21019 did not show any significant change in the apoptotic level compared to the parental strain ( Figure 3A ) . None of these tested mutants showed growth defect when compared to their parental strains , excluding the possibility of differential apoptosis due to differential growth rate ( data not shown ) . These results strongly suggest that capsule and pili expression is not directly correlated with induction of apoptosis . We next showed that sonicated extracts of the pathogenic isolate LNP19995 was able to induce apoptosis to a similar level to that observed with live LNP19995 bacteria . In addition , heat-killed bacteria ( 60°C for 30 min ) of the isolate LNP19995 were still able to induce apoptosis in Hec-1-B cells although to a significant lesser extent ( Figure 4A ) . These results suggest that bacterial growth rate does not affect the induction of apoptosis . Moreover , both proteinaceous and non proteinaceous bacterial factors are involved in inducing apoptosis . Indeed , both the isogenic LOS-defective mutant Z0305 and the isogenic PorB-defective mutant NM0401 , showed reduced levels of induction of apoptosis that were similar to that induced by heat-killed parental wild-type bacteria ( Figure 4A ) . Infection with the double knock out mutant NM0705 , defective in both LOS and PorB production , abrogated apoptosis to a level comparable to uninfected control cells ( Figure 4A ) . Both purified LOS and PorB from the pathogenic isolate LNP19995 were able to induce apoptosis in Hec-1-B cells in a dose-dependent manner . The effect of LOS was abrogated by treatment of cells with the endotoxin drug inhibitor polymixin B ( Figure 4B and data not shown ) . Surprisingly , purified LOS from the carriage isolate LNP21019 induced apoptosis to a similar extent to that purified from the pathogenic isolate LNP19995 ( data not shown ) . Moreover , heat-killed bacteria of the carriage isolate also induced apoptosis ( Figure 4A ) . Taken together , these results suggest that the pathogenic ST-11 isolate LNP19995 induces apoptosis , at least in part , in LOS and PorB-dependent manner . This process was actively suppressed by the carriage isolate . To further confirm this hypothesis , we designed a “functional” complementation using co-infection experiments . Hec-1-B cells were infected with the pathogenic isolate LNP19995 or treated with purified LOS from the isolate LNP21019 in presence of increasing numbers of the non-pathogenic isolate LNP21019 . Figure 4C clearly shows that apoptosis induced by the pathogenic isolate LNP19995 or LOS treatment decreased gradually by the co-infection of Hec-1-B cells with the non-pathogenic isolate LNP21019 in dose-dependent manner . Meningococcal LOS is a potent inducer of TNF-α [26] . TNF-α is also known to trigger apoptosis of a diverse range of cells in vitro [27] . Surprisingly , both the pathogenic isolate LNP19995 and the carriage isolate LNP21019 as well as purified LOS , were shown to induce secretion of TNF-α ( Figure 5A ) . However , only the pathogenic isolate LNP19995 and purified LOS were able to induce apoptosis in Hec-1-B cells . This induction was dramatically reduced using anti–TNF-α neutralizing antibody ( Figure 5B , left panel ) . Similar results were obtained using A549 and HEp-2 epithelial cell lines ( Figure S2 ) . However , induction of apoptosis by TNF-α in these cell lines required cyclohexemide ( CHX ) treatment , an inhibitor of eukaryotic protein synthesis ( Figure S2 ) . The decrease of apoptosis in Hec-1-B cells challenged with the LOS-devoid mutant Z0305 correlated with the decrease of released TNF-α ( Figure 5A ) . The remaining apoptotic activity of this mutant was insensitive to anti–TNF-α neutralizing antibody ( Figure 5B , left panel ) . In contrast , the residual induction of apoptosis in Hec-1-B cells when infected by the PorB-deficient mutant NM0401 ( Figure 5A ) was abolished in presence of anti–TNF-α neutralizing antibody compared to cells treated with an irrelevant antibody ( Figure 5B , left panel ) . We further showed that infection of Hec-1-B cells with the wild-type isolate LNP19995 or the LOS-deficient mutant Z0305 was associated with an alteration of the mitochondrial membrane potential ( MMP ) . This alteration was not affected by an anti–TNF-α antibody . Moreover , this alteration was neither observed when infection was performed using PorB-deficient mutants ( NM0401 and NM0705 ) nor upon TNF-α treatment . ( Figure 5B , right panel ) . Taken together , these results suggest that ST-11 pathogenic isolates induce apoptosis through both an extrinsic pathway involving LOS induction of TNF-α and an intrinsic pathway involving PorB . To understand why the non-pathogenic isolate LNP21019 did not induce apoptosis despite the similar release of TNF-α as the pathogenic isolate LNP19995 , the bioactivity of secreted TNF-α was evaluated using the cell line L929 as described in Materials and Methods . Bioactive TNF-α was only detected when Hec-1-B cells were infected by the pathogenic isolate LNP19995 and purified LOS and was specifically reduced by anti–TNF-α neutralizing antibody ( Figure 5C ) . We therefore explored the mechanism leading to alteration of TNF-α bioactivity upon infection with the carriage isolate . TNF-α induces apoptosis through signaling with its death domain-containing receptor TNF-RI [28] . We next showed , using FACS analysis , that after 9 h of bacterial infection a sustained increase in TNF-RI staining was only observed at the surface of Hec-1-B cells infected with LNP19995 but not at the surface of Hec-1-B cells infected with LNP21019 ( Figure 6A ) . Surface expression of TNF-RII , another receptor belonging to TNF-α receptor superfamily but lacking the death domain , remained unaffected ( Figure 6A bottom panel ) . We complemented our flow cytometry studies by immunofluorescence microscopy to visualize the altered surface expression of TNF-RI in Hec-1-B cells infected with LNP21019 . Monolayers of Hec-1-B cells were infected with GFP-expressing LNP21019 or LNP19995 and were stained for TNF-RI at 9 h post-infection . As TNF-α treatment , cells infected with LNP19995 exhibited increased surface staining of TNF-RI compared to uninfected cells . In contrast , cells infected with LNP21019 showed low level surface staining pattern similar to uninfected cells ( Figure 6B ) . Pre-incubation of Hec-1-B cells with anti-TNF-RI monoclonal antibody ( mAb ) , but not with anti-TNF-RII , significantly reduced apoptosis induced by the isolate LNP19995 and TNF-α compared to an irrelevant antibody ( ∼23% decrease ) . No changes were observed after infection with the carriage isolate LNP21019 ( Figure 6C ) . Under all tested conditions , antibody treatment did not affect the adhesion levels of meningococcal isolates ( data not shown ) . Taken together , these data suggest that TNF-RI may be involved in epithelial cell death induced by pathogenic isolates . The reduction of TNF-RI cell surface expression by the carriage isolate did not seem to result from transcriptional down-regulation of its encoding gene TNFRI . RT-PCR analysis showed that TNFRI was up-regulated upon TNF-α stimulation compared to uninfected cells . Surprisingly , bacterial challenge brought about marked and comparable up-regulation in mRNA levels ( Figure 7A ) . Neither the expression of TNF-RII encoding gene ( TNFRII ) nor the level of expression of β-actin encoding gene ( internal control ) showed any alteration under the tested conditions ( Figure 7A ) . Moreover , FACS and Western blot analysis in permeabilized cells after 9 h of infection further showed that intracellular amounts of TNF-RI increased similarly in cells infected with each isolate compared to uninfected cells ( Figure S3 ) . To explain the low level of surface expression of TNF-RI in cells infected with the carriage isolate , we monitored the level of soluble receptor ( sTNF-RI ) formation . Indeed , TNF-RI is known to decrease from the cell surface by shedding of the extracellular domain into surrounding environment leading to formation of sTNF-RI [29]–[31] . Supernatants from Hec-1-B infected cell cultures were collected 2 , 4 , 6 , and 9 h post-infection and assayed for sTNF-RI by Enzyme-Linked ImmunoSorbent Assay ( ELISA ) . While low levels of sTNF-RI were detected in uninfected cells over time , the level of sTNF-RI strongly increased in a time dependent manner in cells infected with LNP21019 and was 3 times higher than in cells infected with LNP19995 at 9 h post-infection ( Figure 7B ) . The difference in the sTNF-RI levels was statistically significant at 6 and 9 h after infection ( n = 3 , P<0 . 05 ) . Levels of sTNF-RII measured by ELISA did not significantly change following infection ( data not shown ) . The presence of TNF-α bound to sTNF-RI in supernatants of cell cultures infected with LNP21019 was examined by a mixed ELISA . The TNF-α/sTNF-RI complexes were captured by immobilized antibodies to sTNF-RI and quantified by using antibodies to TNF-α . Detectable levels of TNF-α/sTNF-RI complexes were found in cultures of Hec-1-B infected with both N . meningitidis isolates and were not found in uninfected cells . The level of complexes in the supernatant of Hec-1-B cells infected with the carriage isolate LNP21019 was about 4 times higher than that in the supernatant of Hec-1-B cells infected with the pathogenic isolate LNP19995 ( Figure 7C ) . Taken together , these results strongly suggest that the carriage isolate LNP21019 in contrast to the ST-11 pathogenic isolate LNP19995 provoked significant shedding of TNF-RI that seems to chelate TNF-α in the cellular environment . To further establish a causal relationship between the shedding of TNF-RI and the protection from TNF-α-induced apoptosis , Hec-1-B cells were infected with the pathogenic isolate LNP19995 in presence of 250 pg/ml of sTNF-RI ( which is similar to the level of sTNF-RI shedded upon infection with the non-pathogenic isolate LNP21019 ) . Figure 7D showed that treatment of cells with sTNF-RI led to decrease of apoptosis induced by the pathogenic isolate LNP19995 . These results strongly suggest that the release of sTNF-RI is a potential mechanism of protection against TNF-α-induced cell death . Asymptomatic carriage of N . meningitidis contrasts sharply with invasive infections that are characterized by severe symptoms and a major inflammatory response particularly cytokine production and coagulopathy [32] , [33] . Host susceptibility is known to impact on meningococcal disease ( for review see [34] , [35] ) . Most likely , the co-evolution between this exclusive human bacterium and its host derived diversity in both host and bacteria [36] . Moreover , MLST analysis revealed that a limited number of “hyper-invasive lineages” are most frequently associated with meningococcal disease [37] , [38] and that isolates belonging to these lineages , particularly ST-11 isolates , are under-represented in healthy carriers [22]–[24] . In a previous study we showed a positive correlation between virulence in mice and proapoptotic effects to epithelial cell line and isolates of the hyper-invasive lineage ST-11 [25] . We further confirmed here that ST-11 isolates from invasive meningococcal infection are responsible for constant induction of apoptosis in Hec-1-B epithelial cell model , in contrast to isolates from healthy carriers . Moreover , we have previously reported that pathogenic isolates of other clonal complexes ( such as the clonal complex ST-32 ) variably induced apoptosis and that correlated with heterogeneous virulence in mice and higher diversity of isolates compared to ST-11 isolates [25] . Indeed , well established laboratory isolates such as the isolates MC58 and H44/76 ( both belonging to the clonal complex ST-32 ) were unable to induce apoptosis in Hec-1-B cells ( data not shown ) . A growing number of studies have shown that apoptosis can be modulated ( inhibited or promoted ) by bacteria and protozoan parasites ( for a review see [10]–[13] ) . Recently , Schubert et al . [15] reported up-regulation of apoptosis-related genes ( bad , bak , asp , and immediate-early response gene 1 ) in meningococcal infected cells . Further analyses confirmed that cells displayed several hallmarks of apoptosis in response to meningococcal infection , namely , phosphatidylserine translocation and activation of caspase-3 and AMP-activated protein kinase . Meningococcal adhesion but not invasion seems to be necessary for the induction of apoptosis . The initial contact with the host cells is brought by pili , where a strong association of the pathogen to the cell takes place [39] . Our results indicate that adhesion contributes to induce apoptosis of cells infected with pathogenic isolates . This could occur through increasing local delivery of bacterial factors ( such as LOS and PorB ) leading to apoptosis of infected cells . In favor of this explanation is the fact that greater and earlier apoptosis was observed in response to the capsule-deficient mutant than in response to the wild-type strain , most likely through unmasking bacterial surface . Moreover , the abrogation of apoptosis in cells infected with the mutant strain deficient in both PorB and LOS production is also in favor of this hypothesis . In Pseudomonas aeruginosa , pili-producing strains were essential in initiating apoptosis of human Chang conjunctiva cells [40] and piliated enteropathogenic Escherichia coli significantly enhanced apoptotic cell death in HeLa and Caco-2 cell lines , compared to non-piliated mutants [41] . In the current study , we have shown that both meningococcal LOS and PorB act independently but synergistically to induce apoptosis in several human cell lines . Conflicting results on the role of the major outer membrane porin , PorB , in the induction of apoptosis in pathogenic Neisseria ( N . gonorrhoeae and N . meningitidis ) have been reported [16] , [17] , [20] . Similar to our experimental conditions , PorB-related apoptosis has been reported in absence of serum [17] . Absence of serum did not modify the LOS effect as all cell lines tested in this study express membrane CD14 and TLR4 ( [42] , [43] and data not shown ) . LOS signaling is dependent on TLR4 [44]–[46] , while meningococcal PorB has been described as the ligand of TLR2 [47] , [48] . Although signaling through both receptors lead to TNF-α release , our data suggest that additional PorB-induced apoptosis to be independent from secretion of TNF-α and may occur through the intrinsic pathway by alteration of the mitochondrial membrane permeability [16] , [17] . The role of endotoxin of Gram negative bacteria ( including Neisserial LOS ) in cytotopathic effects has been documented [49] , [50] . The severity of meningococcal disease is thought to be linked to the degree of the inflammatory response induced during invasive infection [1] , [51] . As a potent inducer of pro-inflammatory cytokine response , meningococcal LOS seems to play a key role in inducing apoptosis through TNF-α [52] , [53] . In the current study , we have shown that invasive isolates of N . meningitidis induced the secretion of TNF-α and apoptosis in Hec-1-B epithelial cells that could be abrogated by the addition of anti–TNF-α neutralizing antibodies . A relationship between TNF-α secretion and apoptosis in the fallopian tube has also been suggested by studies using a mouse model of infection with a Chlamydia trachomatis mouse-specific pneumonitis strain . Infection with this pathogen leads to a large increase in apoptotic cells in murine oviducts , but treatment with anti–TNF-α antibodies leads to a significant decrease in the level of apoptosis in the upper genital tract [54] . Induction of TNF-α and subsequent apoptosis at the portal of entry of N . meningitidis ( the nasopharynx ) could enhance the invasiveness of pathogenic isolates and hence promote the hyper-invasiveness of isolates belonging to the clonal complex ST-11 . However , the origin of locally induced TNF-α remains to be determined . It might be produced by resident macrophages and/or airway epithelial cells . It has been shown that TNF-α levels were significantly higher in children with demonstrable H . influenzae growth in nasopharyngeal cultures than in culture-negative children [55] . An important mechanism for inducing apoptosis is the activation of the death receptor pathway by extracellular death-inducing ligand TNF-α , which binds to the cognate cell surface receptor TNF-RI [56] . A major response to infectious disease is a cytotopathic effect resulting from activation of this pathway , and the production of TNF-α has been shown to correlate with a cytopathic effect in infected cells [19] , [57] . Indeed , we were able to demonstrate that neutralization of TNF-RI with a specific antibody as well as pretreatment of cells with sTNF-RI provoked a decrease of apoptosis generated by the invasive meningococcal isolate . Moreover , we further showed that meningococcal carriage isolate ( but not pathogenic isolate ) can manipulate this pathway by increasing the shedding of TNF-RI , resulting in inactivation of TNF-α by soluble receptor-ligand complex formation . The mechanism of this enhancement is not yet clear . Shedding of TNF-RI can be mediated by TNF-α converting enzyme ( TACE/ADAM 17 ) , a metalloproteinase localized in the cytoplasmic membrane [58] , [59] . A plausible explanation of the resistance to TNF-α-induced apoptosis by the carriage non-pathogenic isolates is the activation of TACE leading to generation of soluble TNF-α-binding protein that binds and inhibits TNF-α bioactivity . Chlamydia trachomatis and Streptococcus pneumoniae have been shown to reduce the display of TNF-RI at the surface of infected HEp-2 and the airway epithelial cells , respectively through activation of TACE [60]–[63] . Alternatively , an unknown TNF-RI “sheddase” could be expressed from the non-pathogenic isolates , to undergo an interaction with and cleavage of TNF-RI . The molecular basis of the pathogenicity of N . meningitidis remains poorly understood . Full genome sequencing failed to reveal specific components that could determine transmissibility and colonization by carriage isolates or invasive infection by hyper-invasive isolates . The present study provides further evidence for the ability of pathogenic isolates of the ST-11 complex to exert cytopathic effects on epithelial cells by inducing apoptosis [25] . Our data strongly suggest that invasiveness of isolates of the clonal complex ST-11 positively correlates with the induction of apoptosis by these isolates . At the opposite , carriage isolates have co-evolved with human in a manner to reduce inflammatory response and apoptosis induction that may correlate with the state of asymptomatic carriage . RPMI 1640 , HBSS , and trypsin-EDTA were purchased from Invitrogen ( Cergy , France ) . Anti-human TNF-α polyclonal antibody ( Anti hTNF-α ) was from MBL ( Montrouge , France ) . Anti-TNF-RI monoclonal ( mAb , Clone 16803 ) and anti-TNF-RII polyclonal antibodies were from R & D system ( Lille , France ) . FITC- , Rhodamine- and HRP-conjugated secondary antibodies were from Invitrogen . Caspase 3 Assay Kit , Polymixin B and Staurosporine ( STRP ) were purchased from Sigma Aldrich ( Lyon , France ) . FITC-Annexin V kit was from Immunotech ( Marseille , France ) . ApoPercentage Apoptosis Kit was purchased from Biocolor Ltd . ( Newtownabbey , Northern Ireland , UK ) . DiIC1 ( 5 ) assay kit was purchased from Molecular Probes ( France ) . Meningococcal clinical isolates in France are sent to the National Reference Centre for Meningococci ( NRCM ) for full determination and typing . Bacteria were grown in GCB medium with Kellogg supplements [64] . Phenotypes ( serogroup: serotype: serosubtype ) and MLST genotypes were determined as previously described [21] . Sequence types ( ST ) and clonal complexes were assigned using the Neisseria MLST database ( http://pubmlst . org/neisseria ) . All N . meningitidis strains used in this study and their characteristics are listed in Table 1 . The Pilin-deficient mutant strain NM0706 was generated by transforming the wild type strain LNP19995 with the genomic DNA of the previously described strain pilE::aph3' [65] . The isogenic mutants Z0305 devoid of LOS ( lpxA::aph3' ) was obtained by transforming the wild type strain LNP19995 with genomic DNA isolated from the previously described mutant strain Z0204 [38] . Isolates were selected on GCB plates supplemented with 100 µg/ml of kanamycin . To construct the non-capsulated mutant NM0707 and PorB-deficient mutant strain NM0401 , open reading frame of either ctrA or porB genes were PCR-amplified from the pathogenic isolate LNP19995 genomic DNA using the primer sets CtrA-1F/CtrA-100R and PorB0-F/PorB-100R , respectively ( Table 2 ) . PCR fragments were cloned into pGEM®-T Easy ( Promega ) , generating pGEM-CtrA and pGEM-PorB recombinant vectors , respectively . Blunt-ended PCR-generated cassettes aph3' ( conferring resistance to kanamycin ) and erm ( conferring resistance to erythromycin ) were inserted into the blunt-ended unique sites pmlI and KpnI which cut within ctrA and porB respectively , resulting into the recombinant vectors pGEM-ctrA::aph3' and pGEM-porB::erm , respectively . These recombinant vectors were separately introduced into LNP19995 and transformants were selected on kanamycin- or erythromycin-supplemented GCB agar plates , respectively . The strain NM0705 inactivated in both porB and lpxA genes was generated by transforming the genomic DNA of Z0305 strain ( lpxA::aph3' ) in the strain NM0401 ( porB::erm ) . Positive clones were selected on GCB agar plates supplemented with 100 µg/ml kanamycin and 15 µg/ml erythromycin . To construct the non-capsulated mutant strain NM0813 , ctrA was PCR-amplified from the genomic DNA of LNP21019 using the primers 0801-Fw/0801-Rev ( Table 2 ) and cloned into pGEM-T easy plasmid . The blunt-ended aph3' cassette was then inserted into the pmlI restriction site . The resulting recombinant vector was transformed into the parental strain LNP21019 and transformants were selected onto GCB agar plates supplemented with 100 µg/ml kanamycin . All knock-out mutants were confirmed by PCR and Southern blotting . To rule out the possibility that inserted cassettes may have a polar effect , RT-PCR analysis were performed to monitor the expression of genes downstream the inactivated target . The absence of pilin expression from the mutant NM0706 and the absence of capsule expression from the mutants NM0707 and NM0813 were further confirmed by immunolotting using specific polyclonal sera . Silver stained sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) and ELISA were further used to confirm the absence of LOS and PorB expression from corresponding mutants . The EGFP and DsRed-N1 encoding genes ( ClonTech ) were amplified using the primers EGFP-Fw/EGFP-Rev and Red-Fw/Red-Rev , respectively . Both forward primers harbor BamHI restriction site and reverse primers harbor HindIII restriction site as adapters to facilitate the cloning ( Table 2 ) . EGFP and DsRed fragments were subcloned into the previously described pAD3 vector [66] restricted with BamHI and HindIII , resulting in recombinant vectors pCRG-GFP and pCRG-Red , respectively . The 400 bp blunt-ended promoter region of porB amplified using the primers PorB3/PorB4 ( Table 2 ) was inserted into BamHI site of pCRG-GFP and pCRG-Red placing both EGFP and DsRed under control of the constitutive meningococcal promoter porB . The resulting vectors were called pCV1 and pCV3 , respectively . Blunt-ended aph3' cassette was then inserted into the sites ClaI of pCV1 and pCV3 downstream EGFP and DsRed to produce the vectors pCV2 ( PporB-EGFP ) and pCV4 ( PporB-DsRed ) , respectively . Plasmids pCV2 and pCV4 were separately transformed into the isolates LNP19995 and LNP21019 . Fluorescence of transformants was confirmed using immunofluorescence microscopy and flow cytometry . One isolate of each transformation was selected and designated LNP19995-GFP ( /Red ) and 21019-GFP ( /Red ) . These transformants were used in indicated experiments . The human endometrial carcinoma Hec-1-B , the respiratory A549 and the laryngeal carcinoma HEp-2 epithelial cell lines and the mouse fibroblasts L929 cell lines were from American Type Culture Collection ( Manassas , VA ) . The human cell lines ( Hec-1-B , A549 and HEp-2 cells ) and the L929 cells were maintained at 37°C under 5% CO2 humidified atmosphere in RPMI 1640 and DMEM , respectively , supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) . Cells were seeded in 24- or 96-well culture plates , at a density of 5×105 or 5×104 cells/well . Infection was carried out at bacteria to cell ratio of 10: 1 unless otherwise specified . All infection experiments and treatments were carried out in absence of serum to avoid the interference of the serum with PorB [17] . When indicated , cells were incubated with 1 µM STRP ( used as positive control for apoptosis ) or purified LOS equivalent to MOI 10 . TNF-α was used at a final concentration of 5 ng/ml alone or in combination with 20 µg/ml cycloheximide . For antibody neutralizing experiments , cells were co-incubated for the indicated time with 100 µg/ml of the specific neutralizing antibody or isotype-matched irrelevant IgGs , or with 250 pg/ml of sTNF-RI ( Monosan , France ) . Cells were then carefully washed several times with PBS before further analysis Adhesion assay was performed as described previously [66] . Cell viability was measured by Naphthol blue black ( NBB ) staining assay [67] . Cells infected in 96-well plates for the indicated time points as described previously , were stained with NBB solution ( 0 . 05% NBB in 9% acetic acid with 0 . 1 M sodium acetate ) for 30 min at room temperature . The wells were washed three times with PBS to remove the free dye and fixed in 10% formalin for 5–10 min . The attached dye was resolved with 150 µl of 50 mM NaOH . The absorbance at 620 nm was measured by using a Multiskan Ascent Autoplate reader ( Thermo Scientific , France ) . The percent of cytopathic effect was calculated as ( 1- A620 nm of infected cells/A620 nm of uninfected cells ) ×100 . Annexin V assay . Annexin V specifically binds to phosphatidylserine , a plasma membrane lipid that rapidly relocalizes from the inner leaflet to the outer leaflet in cells that are undergoing programmed cell death ( apoptosis ) . Concomitantly , the extent of overall cytotopathogenicity was measured by the standard PI staining . Co-staining with Annexin V and PI allows differentiation of viable cells ( Annexin V− , PI− ) from early apoptotic cells ( Annexin V+ , PI− ) and late apoptotic cells ( Annexin V+ , PI+ ) . Cells infected in 24-well tissue culture plates as described earlier , were harvested using cold PBS/0 . 02% EDTA and washed twice in PBS . Double staining with FITC-Annexin V and Propidium iodide ( PI ) was carried out using the FITC-Annexin V kit , according to the manufacturer's recommendations and then analyzed by FACS . When indicated , apoptosis was also quantified using the ApoPercentage Apoptosis Kit according to the manufacturer's instructions . Assay for mitochondrial permeability transition ( MPT ) : To determine variations of mitochondrial permeability transitions ( MPT ) , as indicative of apoptosis , cells were stained with the potential-sensitive dye 1 , 1′ , 3 , 3 , 3′ , 3′-hexamethylindodicarbo-cyanine iodide ( DiIC1 ( 5 ) ) for flow cytometric analysis , as recommended by the manufacturer ( Molecular Probes ) . This dye is reportedly only incorporated into mitochondria with intact membrane potential . After 20 min incubation with the dye at room temperature and darkness , cells were washed and analyzed by flow cytometry . Cells with high fluorescence intensity for DiIC1 ( 5 ) were considered as living cells . Cells with low fluorescence intensity for DiIC1 ( 5 ) were considered apoptotic cells . Fluorometric analysis of caspase-3 activity . Following infection or STRP treatment , cells were harvested and processed for caspase-3 assay using caspase-3 assay kit , as recommended by the manufacturer . After incubation at 37°C for 70 min , fluorescence was measured at 405 nm by a microplate reader Multiskan Ascent , ( Thermo Scientific , France ) . Standard p-Nitroaniline ( pNA ) solution was used for calculating caspase activity . Extraction of cellular DNA and gel electrophoresis . Harvested cells were lysed for 1 h at 50°C in lysis buffer ( 0 . 1% Triton X-100 , 5 mM Tris-HCl , pH 7 . 5; 0 . 5 mM EDTA , pH 7 . 2 ) supplemented with proteinase K ( 5 mg/ml ) and digested for 1 h with RNase A ( 0 . 5 mg/ml ) . After two phenol/chloroform ( 1∶1 ) extractions , the DNA was precipitated with ethanol , dried , resolved in TE ( 10 mM Tris-Cl , pH 7 . 5 , 1 mM EDTA pH 8 ) , separated on a 1 . 5% agarose gel and visualized under ultra-violet light after ethidium bromide staining . To measure cell surface expression of TNF-RI , cells were collected in HBSS containing 0 . 1% sodium azide ( NaN3 ) and 1% FBS ( staining buffer ) . Cells were labeled with anti TNF-RI mAb , anti-TNF-RII or irrelevant isotype-matched IgGs for 20 min . Cells were then washed twice with staining buffer and labeled with FITC-conjugated secondary antibody for 20 min . Cells were then fixed in 2% paraformaldehyde in staining buffer . Samples were analyzed using a FACSCalibur flow cytometer ( BD Biosciences , France ) equipped with FITC signal detector FL1-H ( excitation = 488 nm , green ) . Relative fluorescence intensities were recorded from a total of 10 , 000 events . Results were analyzed using WinMDI 2 . 8 software . The immunofluorescence staining was performed as described previously [68] . Briefly , cells adherent to glass coverslips were infected for the indicated time , then fixed in 2 . 5% paraformaldehyde in PBS and blocked with 1% normal goat serum in PBS . Anti-TNF-RI mAb was used at 10 µg/ml and rhodamine-conjugated goat anti-mouse IgG was used at a dilution of 1∶1000 . Coverslips were mounted on microscope slides in ProLong Gold antifade reagent ( Invitrogen ) to minimize photobleaching . Slides were then examined by digital confocal microscopy using Zeiss Axio Imager . D1 fluorescent microscope coupled to AxioCam MRm vers . 3 ( Carl Zeiss , Germany ) . Digital images were acquired using appropriate filters and combined using the Axiovision Rel . 4 . 6 software ( Carl Zeiss ) . Supernatants of infected cells were collected , cleared of bacteria and cells by centrifugation at 14 , 000×g and kept frozen at −80°C until use . The amounts of TNF-α secreted in the supernatants were determined using the TNF-α enzyme-linked immunosorbent assay ( ELISA ) kit ( R&D Systems , Abingdon , UK ) in accordance with the manufacturer's instructions . TNF-α bioactivity was assessed by measuring cytopathic effect on L929 murine fibroblasts cells . In brief , L929 cell monolayers cultured for 24 h in 96-well plates ( 5×105 cells/well ) , were overlaid with twofold serial dilutions of culture supernatants from Hec-1-B infected cells in RPMI 1640 supplemented with polymixin B to a final concentration of 1 µM to avoid the effect of free circulating LOS . After incubation at 37°C for 24 h the cells were carefully washed with PBS , and processed for NBB staining as described above . Estimates of the concentrations of bioactive TNF-α in the supernatants were obtained by comparison with calibration curves established with an rhTNF-α standard . TNF-α bioactivity in Hec-1-B supernatant samples was inhibited by anti–TNF-α Ab , but not by control rabbit IgG , indicating that the cytopathic activity in Hec-1-B supernatants represents TNF-α . TNF-α bound to TNF-RI was detected in supernatants of N . meningitidis-infected Hec-1-B cells by a mixed Ab ELISA . The supernatants were added to 96-well microtiter plates coated with an anti TNF-RI mAb , and , after 2 h at room temperature , the plates were washed and incubated for 1 h at room temperature with a rabbit anti-human TNF-α Ab ( 0 . 2 µg/ml ) . After washing , HRPO-conjugated goat anti rabbit IgG ( Caltag ) was determined . Peroxidase activity was determined by addition of citrate buffer and A450 was performed using a microplate reader . Calibrated dilutions of rhTNF-α captured by immobilized anti–TNF-α Ab were used as an internal standard for the comparative measurements of TNF-α complexed with TNF-RI . Total RNA was prepared using Total RNA isolation kit ( Biolabs , France ) according to manufacturer's recommendations . cDNA synthesis and PCR conditions were previously described ( Deghmane et al . , 2002 ) . To assess the expression analysis of TNFRI and TNFRII the primer pairs TNFR1-Fw/TNFR1-Rev , TNFR2-Fw/TNFR2-Rev were used ( Table 2 ) . β-actin-Fw/β-actin-Rev ( Table 2 ) were used to amplify β-actin as internal control . Controls without RNA and/or reverse transcriptase were included in the assay . Meningococcal LOS was prepared with similar MOIs as previously described [69] . PorB was semi-purified as described by [16] and used in combination with 1 µM polymixin B to prevent any effect of LOS . For Immunoblot , cells were lysed 10 min on ice by adding RIPA lysis buffer [50 mm Tris-HCl ( pH 7 . 4 ) , 150 mm NaCl , 0 . 25% sodium deoxycholate , 1% NP40 , 0 . 1% sodium dodecyl sulfate and freshly added protease Inhibitor Cocktail] . Whole cell lysates were then centrifuged at 13 , 000×g for 10 min at 4°C; and supernatants were collected to obtain protein extracts . Protein concentrations were determined by the Bradford assay ( Bio-Rad , Hercules , CA ) . Fifty micrograms of extracted protein was run on SDS-PAGE and transferred to a nitrocellulose membrane . The blots were then probed overnight at 4°C with the relevant antibodies , washed and probed again with species-specific secondary antibodies coupled to HRPO and signal was detected by chemiluminescent reagents . Data are expressed as the mean±SD of triplicate samples , and the reproducibility was confirmed at least in three separate experiments . Statistical analysis were performed using student t test ( two-way Annova ) test , and considered significant if P<0 . 05 .	Acquisition of Neisseria meningitidis often leads to asymptomatic colonization ( carriage ) and rarely results in invasive disease associated with tissue injury . The reasons that make disease-associated isolates ( pathogenic isolates ) but not asymptomatic carriage isolates able to invade the host to establish disease are not understood . Isolates belonging to the ST-11 clonal complex are most frequently associated with the disease and rarely found in carriers . These hyper-invasive isolates may be able to induce cytopathic effects in target cells . We aimed to investigate the cytopathic effect of meningococcal isolates on epithelial cells using both ST-11 pathogenic isolates and carriage isolates . We showed that cytopathic effects were strongly associated with pathogenic isolates and infected cells exhibited features of apoptosis . This effect is mainly mediated by bacterial endotoxin ( lipooligosaccharide ) and involved an autocrine signaling mechanism of secreted TNF-α through its receptor TNF-RI . In contrast , carriage isolates down-regulate TNF-RI on the surface of infected cells by increasing TNF-RI shedding into the medium . We suggest that chelating secreted TNF-α protects cells from apoptosis . Our results unravel a differential modulation of TNF-α signaling by meningococcal isolates leading to cell survival or death and would therefore contribute to better understanding of the duality between carriage and invasiveness .	[ "Abstract", "Introduction", "Results", "Discussion", "Materials", "and", "Methods" ]	[ "microbiology/cellular", "microbiology", "and", "pathogenesis", "microbiology/immunity", "to", "infections" ]	2009	Differential Modulation of TNF-α–Induced Apoptosis by Neisseria meningitidis
The human filarial parasite Onchocerca volvulus is the causative agent of onchocerciasis ( river blindness ) . It causes blindness in 270 , 000 individuals with an additional 6 . 5 million suffering from severe skin pathologies . Current international control programs focus on the reduction of microfilaridermia by annually administering ivermectin for more than 20 years with the ultimate goal of blocking of transmission . The adult worms of O . volvulus can live within nodules for over 15 years and actively release microfilariae for the majority of their lifespan . Therefore , protracted treatment courses of ivermectin are required to block transmission and eventually eliminate the disease . To shorten the time to elimination of this disease , drugs that successfully target macrofilariae ( adult parasites ) are needed . Unfortunately , there is no small animal model for the infection that could be used for discovery and screening of drugs against adult O . volvulus parasites . Here , we present an in vitro culturing system that supports the growth and development of O . volvulus young adult worms from the third-stage ( L3 ) infective stage . In this study we optimized the culturing system by testing several monolayer cell lines to support worm growth and development . We have shown that the optimized culturing system allows for the growth of the L3 worms to L5 and that the L5 mature into young adult worms . Moreover , these young O . volvulus worms were used in preliminary assays to test putative macrofilaricidal drugs and FDA-approved repurposed drugs . The culture system we have established for O . volvulus young adult worms offers a promising new platform to advance drug discovery against the human filarial parasite , O . volvulus and thus supports the continuous pursuit for effective macrofilaricidal drugs . However , this in vitro culturing system will have to be further validated for reproducibility before it can be rolled out as a drug screen for decision making in macrofilaricide drug development programs . The human parasitic filarial nematode , Onchocerca volvulus , is the causative agent of onchocerciasis ( river blindness ) . Approximately 18 million people are infected with the parasite , mostly in sub-Saharan Africa , and of these individuals an estimated 270 , 000 people are completely blind , with an additional 6 . 5 million suffering from severe skin pathologies [1] . Onchocerciasis is a major cause of global disability and is the second leading cause of blindness due to infection , after Trachoma . Since 2010 , over 3 million disability-adjusted life years have been lost due to O . volvulus infections [2] . As with all other filarial parasites , O . volvulus is a vector-borne infection . During a blood meal , the blackfly vector ( genus Simulium ) ingests microfilariae ( mf , L1 ) , which develop over 6–10 days and undergo two molts to the infective stage larvae ( third-stage , L3 ) . The O . volvulus L3 ( OvL3 ) migrate to the head and proboscis of the blackfly and are introduced into the bite wound during the blackfly’s subsequent blood meal . These larvae penetrate the wound , migrate deeper into the dermis and subcutaneous tissues of the human host , molt twice from L3 to fourth stage-larvae ( L4 ) and from L4 to pre-adult fifth-stage larvae or L5 . Little is known about the time course of this developmental process in vivo , but the molt from L3 to L4 occurs in the first 3–7 days [3] , while the L4 to L5 molt is estimated to occur after 2 months [4] . Early L5 are considered young adults , and at this stage the worms have partially developed gonads [5] . It takes 279–532 days post infection ( dpi ) for the closely related O . ochengi parasite of cattle to develop into fully mature and fertile adult worms capable of releasing mf [6] , and more than 400 days post infection ( dpi ) to do the same in a chimpanzee model for O . volvulus [4 , 7] . The majority of the microfilariae migrate to the skin where they are transmitted to the blackfly during a blood meal , thus continuing the life-cycle of the parasite . Both the adult female ( 33–50 cm x 130–210 μm ) and adult male ( 19–42 cm x 130–210 μm ) worms are encapsulated in collagenous subcutaneous nodules known as onchocercomas [8] . Adult males , however , can migrate within the onchocercomas to fertilize resident females [9] . Fecund females release 1 , 000–3 , 000 mf per day , and these first stage larvae called microfilariae ( 220–360 μm x 5–9 μm ) can live for 12–18 months in the human host . Mf are typically found in the skin , and therefore the diagnosis of infection is traditionally made via the microscopic detection of mf released from skin snips , or less commonly by molecular methods such as PCR [10 , 11] . Current international control programs focus on reducing transmission with the ultimate goal of eliminating onchocerciasis by 2025 in both the Americas and Africa [12] . Currently only ivermectin ( IVM ) is used for mass drug administration ( MDA ) , but MDA with IVM alone has several limitations . Though IVM does successfully target mf ( microfilaricidal ) , the drug has no effect on adult parasites . Because O . volvulus can live in the nodules of infected humans for over 15 years and actively release mf for the majority of their lifespan , protracted treatment courses ( >20 years ) are required to curb transmission and eventually eliminate the disease [13 , 14] . Additionally , the emergence of IVM-resistant O . volvulus could greatly hinder this effort [15] . Therefore , despite the challenge of studying a parasite with no small animal model for infection , a critical need still exists for further research to identify novel drug targets , and to develop a new generation of macrofilaricidal drugs against O . volvulus . One such approach is to target the endosymbiotic bacteria , Wolbachia , that reside intracellularly in the parasite . Clinical trials using antibiotic drugs , such as doxycycline or rifampicin , have demonstrated that clearance of Wolbachia from the filarial worms results in the slow death of adult worms ( macrofilaricidal effect ) , preceded by sterilization and reduction in transmission frequency [16–19] . Doxycycline treatment is indicated in certain situations , though large scale MDA is limited by both the length of treatment required and contraindications for children and pregnant woman [16 , 20 , 21] . Currently , no feasible method exists to permanently sterilize adult female parasites or to kill adult worms outright , highlighting the need for additional research to discover novel macrofilaricidal drug candidates . In the 1990s , O . volvulus adult male and female worms , which can only be obtained from infected individuals and after nodulectomy , were used ex vivo to screen promising new compounds [9] . Unfortunately , this is not feasible anymore and therefore most of the present screening funnels for novel macrofilaricidal drugs depend on surrogate in vitro and in vivo filarial models such as Brugia spp , Litomosoides sigmodontis , and cattle Onchocerca spp [21–26] . In the present study we describe a newly developed in vitro culturing system that supports the growth and development of O . volvulus young adult worms ( OvL5 ) from the L3 stage . We show that the optimized culturing system allows the OvL5 to mature into young adult worms that have distinguishable gonads and express adult-specific transcripts . Importantly , we show in preliminary assays that these in vitro developed worms can be used as an auxiliary for screening putative macrofilaricidal drugs against the ultimate target organism , O . volvulus . We also show that the preliminary assays can clearly differentiate inhibitory activities of various FDA repurposed drugs . The procedures used for the production of O . volvulus third-stage-larvae ( L3 ) were approved by an NIH accredited Institutional Review Board of the Medical Research Council Kumba , Cameroon ( Protocol 001 ) , and by the Le Comité National d’Ethique de la Recherche pour la Santé Humaine , Yaoundé , Cameroon ( Protocol 677 ) . L3 were collected from black flies ( Simulium damnosum ) that were fed on consenting infected donors . The consenting donors were offered and provided with ivermectin treatment by the end of their participation . After seven days the infected flies were dissected and the developed L3 were collected , cleaned and cryopreserved . The cryopreserved L3 were shipped to the New York Blood Center in liquid nitrogen and upon arrival in New York were stored in liquid nitrogen . All protocols using the L3 cryopreserved samples in this study were approved by the New York Blood Center’s IRB ( Protocol 321 and Protocol 603–09 ) . All L3 samples were anonymized . The peripheral blood mononuclear cells ( PBMCs ) used to culture O . volvulus L3 were isolated from human leukopaks collected from healthy donors following the New York Blood Center’s approved IRB protocol ( Protocol 420 ) . The de-identified human leukopaks were obtained from the New York Blood Center Component Laboratory . The New York City Blood Center obtained written informed consent from all participants involved in the study . All protocols were conducted in accordance with National Institutes of Health guidelines for the care and use of human subjects . Cryopreserved O . volvulus L3 stage larvae ( OvL3 ) were thawed and washed in “larvae” medium that contains 1:1 NCTC-109 and IMDM supplemented with Glutamax ( 1x ) and 2x Antibiotic-Antimycotic ( Life Technologies ) . After washing , the larvae were dispersed in a 96-well plate ( 10 larvae/well ) containing 1 . 5x105 human Peripheral Mononuclear Blood Cells ( PBMC ) per well in complete “larvae” medium , which was supplemented with heat inactivated ( HI ) 20% fetal bovine serum ( FBS , Sigma ) . Worms were cultured at 37 °C with 5% CO2 until day 6 when the molting rate was estimated under an inverted microscope based on the presence of the highly motile L4 ( OvL4 ) and the empty L3 cuticle . To culture OvL4 long-term , we first tested several monolayer cell lines: Human Umbilical Vein Endothelial Cells ( HUVEC ) , Human Dermal Fibroblasts ( HDF ) , Human Skeletal Muscle Cells ( HSkMC ) , Human Dermal Microvascular Endothelial Cells ( HMVEC-D ) , Human Dermal Lymphatic Microvascular Endothelial Cells ( HDLMVEC ) , and Keratinocytes , all purchased from Lonza Inc . ( Allendale , NJ ) . Each of the cell lines were maintained in the laboratory in their respective recommended base media . The motile OvL4 larvae were transferred into new 96-well plates that had been seeded with the various cell lines . The culturing medium consisted of 50:50 “larvae” media to cell media , supplemented with 20% HI FBS and 1x Antibiotic-Antimycotic . The growth of the larvae was measured over 70 days; the two cell lines that supported the best growth of the OvL5 were HUVEC and HDF ( Fig 1 ) . Significance was determined using ANOVA ( day 56 ) and a t-test ( day 70 ) . To prevent OvL4 from becoming entangled within the cell monolayer in the 96-well plates , we devised an alternative culturing system where ~10 OvL4 were placed inside a transwell ( Corning Transwell , 6 . 5 mm Transwell with 3 . 0 μm pore polyester membrane insert , Sigma CLS3472 ) in 24-well plates seeded with HDF or HUVEC monolayers at 1x104 cells/well in 1 ml of OvL4-media . This media was comprised of 1:1 NCTC-109: MEM-alpha with either DMEM/F12 ( Life Technologies ) or EBM-2 ( Clonetics EBM-2 Lonza CC-3156 ) , the recommended base media for each cell line at 60:40 ratios . The OvL4-media was further supplemented with 20% HI FBS , 1x Antibiotic-Antimycotic , 1% Glucose ( Sigma G8769 ) and 1% Sodium Pyruvate and 1% ITS ( Life Technologies 11360–070 and 51500–056 ) . Unfortunately , some of the HDF cells present in suspension tended to migrate over the outer membrane of the transwell during the media change and to overgrow with time , causing the worms to become entangled . Therefore , to maintain worms free of any cell entanglements , the subsequent and final cultures , including the screening assays , were set up using only the HUVEC monolayer . To further optimize culturing conditions , we first tested the outcomes of culturing the worms after adding 0 . 1% Lipid Mixture-1 ( Sigma L0288 ) to the OvL4-media . This medium was named complete OvL4 ( OvL4-CM ) medium . OvL4s were maintained for at least 104–120 days within the transwells with 3 media exchanges per week: 0 . 5 ml of old medium was removed , and 0 . 5 ml of fresh media was added . Fresh HUVEC monolayers were prepared weekly and the transwells containing the developing OvL4 were transferred once a week into freshly seeded monolayers with 1 ml of OvL4-CM . To further augment the growth of the developing worms , the OvL4-CM medium was also supplemented with 1% Non-Essential Amino Acids ( Life Technologies 11140–050 ) , and 25% HI FBS ( instead of 20% ) . This newly devised media composition ( OvL4-CMS ) produced worms with the longest length thus far , with the longest worm measuring > 3 , 000 μm ( See Results ) . Significance was determined using a t-test . The measurements , images , and videos of live larvae were taken on a Nikon Eclipse TS100 inverted microscope equipped with a Nikon DS-Fi2 camera controlled by the NIS Elements version 4 . 3 Windows based imaging program . For transmission electron microscopic analyses of the developing worms , samples of OvL4 ( n = 10 ) on days 48–50 , molting OvL4 ( n = 10 ) on days 50–60 , OvL5 ( n = 10 ) on days 60–75 or 120 were fixed with 2 . 5% glutaraldehyde and 2% paraformaldehyde in sodium cacodylate buffer ( 0 . 1M ) for 2 h . After fixation , worms were washed three times in sodium cacodylate buffer ( 0 . 1M ) and post-fixed with 1 . 5% OsO4 for 1 h . Worms were then washed and dehydrated using a series of increasing ethanol concentrations ( 50–100% ) with a final wash with propylene oxide . Samples were then embedded in plastic resin ( Epon 812 , EMS USA ) and prepared for sectioning . Ultrathin sections were contrasted with UranyLess ( EMS , USA ) and lead citrate and then were analyzed under the Tecnai G2 Spirit transmission electron microscope . Ten ( n = 10 ) developing OvL4 D21 , 10 OvL5 D76 and 10 OvL5 D96 were collected from in vitro cultures . Worms were washed with PBS and stored at -80°C . OvL4 and OvL5 were collected four times for separate RNA extractions and analysis . Two aliquots , 50 OvL3 each , were prepared from thawed and washed cryopreserved larvae collected from black flies . One pool of adult RNA was prepared from 1 female and 10 male worms that were frozen . Total RNA was extracted from the collected worms by a TRIzol-based method followed by further purification using a column based mini RNA extraction kit ( Invitrogen ) . During purification RNA was treated with DNase I ( Invitrogen ) on columns according to the manufacturer’s instructions . Purified RNA was used as a template for cDNA synthesis performed using the SuperScript III first-strand synthesis system ( Invitrogen ) . cDNA synthesized from adult worms was diluted 100 times with water . The cDNA was amplified by PCR using primers for O . volvulus transcripts corresponding to OVOC11951 , OVOC5433 , OVOC9683 , OVOC12838 , OVOC2456 ( S1 Table ) , which were selected based on their overexpression in O . volvulus adult worms vs . mf and OvL3 [27] . For the positive control we used primers for Ov-Tubulin . Primers for an intron were used as a negative control to test that there was no contamination of genomic DNA . PCR products were run using 1 . 5% agarose gel electrophoresis . A week prior to performing an assay , cultured OvL5s were transferred from the transwells into a small petri dish containing fresh OvL4-CMS . Aliquots of 8–10 worms were retrieved from the petri dish and placed inside new transwells over newly seeded HUVEC monolayers in OvL4-CMS . The HUVEC cell monolayers ( 1x 104/well of a 24-well plate , with 0 . 5 ml of OvL4-CMS ) were prepared as described above one to two days prior to the drug screening assay . After transferring OvL5 to these plates , test drugs ( flubendazole or oxfendazole ) were added to each well as needed . Tested compounds were dissolved in DMSO at a concentration of 10 mM and added to the OvL5 wells in 0 . 5 ml OvL4-CMS containing 2x the desired final concentration . Plates were incubated at 37°C in a 5% CO2 incubator . Media containing freshly prepared compound was replaced every 2–3 days for the 14 days of treatment ( or less depending on the assay ) by removing 0 . 5 ml of the media and adding 0 . 5 ml of the appropriate treatment media to maintain the concentration of a drug . For the normal growth control , we used complete media with a final concentration of 0 . 05% DMSO . After the treatment period ( 14 days or less ) , the media was replaced completely with 1 ml of OvL4-CMS , and then every 2–3 days with fresh media as described above for 14 additional days . Motility was recorded every 2–3 days over the full period of the assay according to the following scale: 100% motility , constant coiling movement; 75% motility , slower coiling; 50% motility , slow and intermittent movement; 25% motility , very slow movement or twitching; and 0% motility , no movement ( S1 Video ) . Observations were done blinded by two individuals . On Day 28 , viability was assessed by MTT staining as previously described [28] . Untreated control and treated OvL5 within the transwells were washed with PBS and then incubated with MTT ( 3- ( 4 , 5-Dimethylthiazol-2-yl ) -2 , 5-Diphenyltetrazolium Bromide ) ( 0 . 1% ) in PBS at 37°C under 5% CO2 for 1 hour , followed by an additional wash of the parasites with PBS . The product of MTT reduction ( formazan ) is blue . Worms were considered dead if no staining or < 50% within the worm was observed using an inverted microscope ( S1 Fig ) . Worms stained blue or > 50% stained were considered alive . All test treatments were performed in duplicates . Significance was determined using a t-test . To determine the IC50 for drug activity on OvL5 , serial 2X concentrations of the test compounds ( 60 , 20 , 6 , 2 , 0 . 6 , and 0 . 2 μM ) were prepared in OvL4-CMS medium . The 0 . 5 ml of media containing the 2x compound was added to transwells containing OvL5 worms ( n = 10 , 75–80 days or less based on the assay ) in 0 . 5 ml of media to yield final test concentrations of 30 , 10 , 3 , 1 , 0 . 3 , and 0 . 1 μM . The doses of flubendazole used were in the same range as previously done with adult B . malayi worms in vitro [29] . Moreover , in previous in vivo studies of flubendazole in rats , doses of 65–402 mg/kg that corresponded to Cmax = 0 . 90–1 . 33 μg/ml were used [30] . In our studies we used flubendazole at the concentration used in vivo ( 2 μM ) , which is equivalent to 0 . 62 μg/ml . OvL5 ( n = 20 ) cultured in the presence of a final concentration of 0 . 05% DMSO in media served as a control for normal development . All test treatments were performed in duplicate . Plates were incubated at 37 °C in a 5% CO2 incubator . Media replacement with or without compounds followed the same procedure described above . Motility was monitored over the full course of treatment and thereafter , with IC50 for motility determined on the last day of culturing . Viability was determined the following day . The percent inhibition of motility and viability for treated OvL5 worms were calculated with respect to the movement and viability of OvL5 in wells containing the DMSO in normal complete media . IC50 values were calculated using Graph Pad Prism v6 . 0 ( http://www . graphpad . com ) . The probability of assumption ( R2 ) was calculated by the software and is reported in each graph . Efficacy of drugs on O . volvulus L3 larvae using the molting assay was performed as described previously [31] . Briefly , 5–10 L3 per 50 μL in complete medium ( CM ) containing 20% heat inactivated FBS were distributed to 10 wells of a 96-well plate and 1 . 5 × 105 normal PBMCs were added per well in 50 μL . 2X dilutions of a compound ( final concentration of 30 , 10 , 3 , 1 , 0 . 3 , 0 . 1 and 0 . 03 μM ) were then added to each well , 100 μL per well . The total number of worms tested for each treatment condition was 30 and 60 for controls . Controls of DMSO in complete medium and complete medium without DMSO were included in each assay . The concentration of DMSO within the cultures was never higher than 0 . 05% . The 96-well plates were then incubated at 37 °C in a 5% CO2 incubator until day six when molting was recorded using an inverted microscope and observed for the presence of the fourth-stage larvae ( L4 ) and the empty casts of the L3 . The probability of assumption ( R2 ) was calculated by the software and is reported in each graph . Adult female B . pahangi were assayed using methods described by Marcellino et al . ( 2012 ) and Bulman et al . ( 2015 ) [23 , 26] . Individual females were placed in each well of a 24-well plate with complete media Extra media was removed from plates using a Biomek FxP , leaving 500 μL of media per well . Auranofin , niclosamide , and nitazoxanide were dissolved in DMSO ( Sigma ) and 10 mM stock solutions were stored at -20°C . Drugs were tested using 4 worms per compound per concentration . Plates were kept in a 37°C , 5% CO2 incubator for 3 days . 6-point serial dilutions were used to determine IC50s: for auranofin the concentrations were 10 μM , 3 μM , 1 μM , 0 . 3 μM , 0 . 1 μM and 0 . 03 μM; for niclosamide and nitazoxanide the concentrations were from 30 μM to 0 . 1 μM; 1% DMSO was used as a control . These concentrations were selected based on published data [26] . Non-linear regression curves were used to calculate IC50 values using Graph Pad Prism v6 . 0 . IC50 assays were repeated 2–3 times and the means with R2 values greater than or equal to 0 . 7 are reported . The preparation of OvL4 followed our standard protocol using cryopreserved L3 cultured in complete ‘larvae’ medium in the presence of human PBMCs [26 , 32] . As expected , O . volvulus L3 molted to L4 within the first 6 days of culture; molting of L3 was confirmed by detecting empty L3 casts within the well and observing highly motile OvL4 ( S2 Fig ) . In general , 30–60% of the OvL3 molted to OvL4 . The varied molting rate was associated with the numerous lots of the cryopreserved , thawed O . volvulus L3 . The L3 stage larvae had a median length of 647 μm ( ranged from 558 μm– 708 μm ) , while the newly molted L4 had a median length of 703 μm ( 583 μm– 810 μm ) . In our initial attempts to keep OvL4 for long term culturing , we kept the newly molted OvL4 in the presence of freshly cultured PBMCs but this proved unsuccessful , as larvae did not thrive more than a week later in the culture . We then decided to investigate which human cell lines would best support OvL4 growth . We cultured L4 ( n = 10 per cell line ) in 96-well plates seeded with six different cell line monolayers; HMVEC-D , HUVEC , HDF , HSkMC , BESM , and LMVE . L4 cultured in HMVEC-D , BESM or HDLMVEC had the lowest growth ( p = 0 . 016 ( n = 9 ) , d56 ) and highest mortality rate ( Fig 1 ) . In comparison , the HUVEC and HDF monolayers showed promise , with growth continuing after day 56 ( Fig 1 ) . To further optimize media composition all experiments were performed with the HUVEC monolayers only using 24-well plates containing transwells as described above ( Fig 2 ) . This was based on two observations . The median length of OvL5s maintained on HUVEC monolayers was slightly longer on day 70 ( range of 804–1504 μm , median 1040 μm ) , even if not statistically longer than those grown on the HDF monolayers ( range of 803–1374 μm , median 1013 μm ) ( Fig 1 ) . Additionally , it prevented the logistical challenges with HDF cells , which overgrew within the cultures , causing the growing OvL5 to become entangled within the cellular monolayer , and subsequently negatively affecting the fitness of developing worms . This culturing system , the HUVEC monolayers in a 24-well plate containing the larvae in the transwells , was used to further refine our media composition and establish whether supplements c improve the growth of OvL5 ( Fig 2A ) . The first supplement tested was the addition of 0 . 1% lipid mixture ( Fig 2B ) . This was followed by the addition of 1% Non-Essential amino acids to the 0 . 1% lipid mixture and an increase in the final concentration of FBS from 20% to 25% ( Fig 2C ) ; this modified media is referred as OvL4-CMS . As observed before in 96-well cultures ( Fig 1 ) , the OvL4 in OvL4-CMS ( Fig 2C ) as well as those in OvL4-media and OvL4-CM media ( Fig 2A and 2B ) grew slowly until day 53 ( range 678–1376 μm , median 800 μm ) after which time , there was an increase in worm length between day 61 ( 680–1743 μm , median 871 μm ) and day 75 ( range of 776–2254 μm , median 1046 μm ) that coincided with molting of L4 to L5 ( Fig 3C ) . By day 104 the worms that were cultured in OvL4-CMS were much longer ( range of 857–3329 μm , median 1358 μm ) than in the other two media systems ( p<0 . 01 ( n = 26 ) ) . As seen clearly in Fig 2C , two clusters of worms were observed , those larger than 1358 μm and those that were smaller than 1358 μm ( median ) . Sexual dimorphism could explain this growth difference . Notably , starting on day 75 we found that it easier to distinguish between male and female worms in the same culture well ( male worms were generally smaller than female worms ) . The sharp increase in growth on day 75 ( Fig 2C ) and thereafter might be also attributed to completion of the molting process from OvL4 to OvL5 as evidenced by structural analysis using transmission electron microscopy ( Fig 3 ) . The L4 to L5 molting process is clearly evidenced by the presence of a separation between the cuticles of L4 and the developing L5 worms on days 48–60; the molting process was completed by days 60–75 ( Fig 3 ) . Notably , the developing L5 had a highly annulated cuticle underneath the L4 cuticle ( Fig 3H–3J ) , and the OvL5 ( day 120 ) had a distinguishable epicuticle , gut , and developing gonads ( Fig 4 ) . The stage-specific transcriptomic and proteomic datasets for O . volvulus L3 , L4 , and adult males and females [27] were analyzed to select transcripts that appear specifically and/or exclusively expressed in adult worms . Five genes expressed predominately in adult worms were selected for analysis of their expression in OvL3 collected from blackflies ( D0 ) , OvL4 ( D21 ) , OvL5 after molting ( D76 ) , and pre-adult OvL5 ( D96 ) in comparison to their expression in female and male adult worms recovered from nodules collected from humans after nodulectomy . The expression of tubulin in each stage was used as a positive control . As seen in Fig 5 , the transcript of OVOC11951 was specifically expressed on day 76 and day 96 OvL5 and in adult worms , while the OVOC5433 and OVOC9683 genes were expressed only in the pre-adult OvL5s ( D96 ) and in adult female and male worms . The OVOC2456 gene , by comparison , was already expressed in L4 day 21 and its expression was continuously present in the developing L5 and in adult worms . The OVOC12838 gene was , however , expressed only in the adult female and male worms . None of the transcripts , except tubulin ( control of the quality of cDNA ) , were expressed in OvL3 as predicted by the transcriptomic dataset . Once our in vitro culturing system for OvL5 was optimized , we tested whether this newly developed system could be applied to macrofilaricidal drug screening assays . We first tested flubendazole , a known macrofilaricidal drug for O . volvulus infection [33 , 34] . Since we were able to maintain OvL5 for up to 104 days in culture , we decided to test the OvL5 worms ( day 73 ) with various concentrations of flubendazole ( 0 . 03–10 μM ) until we first observed effects on motility . The drug was then removed ( media replacement ) and observations of worm motility continued over an additional period of culturing . By day 14 , 1 μM flubendazole started to have a noticeable effect on motility ( ~40% inhibition , p<0 . 01 , n = 11 ) . We therefore decided to stop the treatment on day 14 for all treatment wells and followed worm motility in the presence of normal media for an additional 14 days ( Fig 6 ) . The IC50 for motility was determined on day 28 and for viability ( MTT assay ) on day 29 ( S3 Fig ) . The IC50 was 0 . 13 μM for both the inhibition of motility and for the inhibition of viability ( S3A and S3B Fig ) . When oxfendazole , another benzimidazole drug previously shown to have an in vivo macrofilaricidal activity against several other filarial species in animal models [35] , was tested on OvL5 ( day 80 ) , we found some differences based on the phenotype analyzed . The IC50 for inhibition of motility was similar to that of flubendazole ( 0 . 12 μM , ( S3C Fig ) ) , however , the IC50 for inhibition of worm viability was 0 . 54 μM ( S3E Fig ) —4 times higher than after the treatment with flubendazole . Second , we tested the effects of three FDA repurposed drugs ( auranofin , niclosamide and nitazoxanide ) on the molting of OvL4 ( D57 ) and their development into OvL5 ( D63 ) ; OvL4 on D57 in culture were used in this drug screening experiment . Auranofin , a gold-containing drug used for rheumatoid arthritis , has been shown in vitro to inhibit the molting of L3 of O . volvulus and inhibit the motility of adult female Brugia pahangi , with IC50 values in the low μM range of 0 . 3 and 0 . 5 , respectively [26] . Niclosamide is an orally bioavailable chlorinated salicylanilide used as an anthelmintic drug against tapeworm infection . In our in vitro screens with O . volvulus L3 and B . pahangi female worms , niclosamide inhibited molting of O . volvulus L3 with an IC50 of 0 . 08 μM ( S4 Fig ) , while it inhibited the motility of the B . pahangi female adult worms with an IC50 of 1 . 2 μM ( Table 1 ) . Nitazoxanide , a broad-spectrum antiparasitic drug , was less effective on O . volvulus L3 ( S4 Fig ) and B . pahangi female adult worms; IC50 of 4 . 71 and 5 . 8 μM , respectively ( Table 1 ) . Using our in vitro OvL5 assay , two of these three drugs ( auranofin , niclosamide ) were shown to be effective in killing the pre-adult worms after a 6-day treatment in vitro: 3 μM auranofin killed 100% of the O . volvulus worms and 3 μM niclosamide killed 88% of the O . volvulus worms . In comparison , 30 μM nitazoxanide killed only 65% of the O . volvulus worms . The corresponding IC50 values for viability were 1 . 0 μM for auranofin , 1 . 5 μM for niclosamide , and 24 . 0 μM for nitazoxanide . The range of IC50 values for the OvL5 appear to be somewhat similar to the μM range observed using adult Brugia assays , thus supporting the validity of this in vitro model for future screening of novel macrofilaricidal drugs against O . volvulus . We have developed , optimized , and tested a novel in vitro culturing system that supports the growth and development of O . volvulus young adult worms from the L3 stage . It shows that the OvL4 worms can be optimally maintained in OvL4-CMS medium containing supplements ( a mix of amino acids , lipids and 25% FBS ) with a supportive monolayer of HUVEC ( Fig 2C ) . Using this culturing system , not only the molting of OvL4 to L5 is supported , but also the subsequent morphogenesis into young pre-adult stages . Filarial nematodes molt their cuticles ( an extracellular hydroskeleton that overlays the hypodermis of worms ) four times during pre-adult development . Here , we observed that worms developing from OvL4 to OvL5 in the in vitro setting shed their old cuticle by first building a new one beneath ( Fig 4 ) . Development of O . volvulus beyond the fourth-stage larval stage has previously been attempted in a few animal models , including mice , non-human primates , and chimpanzees , using the implantation of diffusion chambers containing the larvae . The greatest increase in the length of the recovered worms was after 4 weeks—from 350 μm to 424–601 μm in DBA/2J mice and up to 559–759 μm in chimpanzees [5] . These in vivo models did not support further worm development past 63 days [5] . In our optimized in vitro culturing system , we were able to culture worms for more than 120 days and the worms showed continuous growth up to 3329 μm ( range of 857–3329 μm , day 104 ) ( Fig 2C ) . According to a recently published datasets of the transcriptome of O . volvulus and the transcript levels in the worms during different stages of development , a few candidate transcripts were identified that could be used as potential biomarkers of adult stages of development [27] . We assumed that these biomarkers could be also employed to monitor the progress of worm development in our in vitro system . We found that three ( OVOC11951 , OVOC5433 , and OVOC9683 ) of the five selected adult–stage-specific transcripts were also expressed by young adults developed in our culturing system ( Fig 5 ) . These three transcripts could be therefore used as potential markers to monitor the efficacy of treatments performed using the in vitro system for O . volvulus drug screening . Moreover , TEM analysis of OvL5 ( D120 ) showed the presence of an epicuticle that appears to be characteristic of adult worms ( Fig 5 ) , this further confirming that the OvL5s developed in vitro are pre-adult worms . The most important application of our ability to culture OvL4 and OvL5 in vitro is the promising novel capacity to screen compounds ( including repurposed drugs ) against the target filarial parasite O . volvulus . In preliminary assays we tested the effects of few putative macrofilaricidal drugs on the motility and viability of the in vitro developed young O . volvulus worms or on the OvL4-OvL5 molting process . Both effects ( motility/ viability of OvL5 and molting to OvL5 ) are important , as an ideal drug candidate would arrest the development of growing worms in addition to their macrofilaricidal activity on adult worms within the nodule . By employing our novel long term in vitro culturing system we were able to estimate initial IC50 values for several repurposed drugs including flubendazole , a known macrofilaricidal drug candidate for O . volvulus infection [33 , 34] . In comparison , when other in vitro screening models were used such as that with B . malayi adult worms , flubendazole did not inhibit worm motility after 5 days , although it caused some damage [29] . With the other in vitro screening models , worm motility is limited to only a few days ( 7 days at the most using the adult female O . ochengi worms ) . In vivo it takes 42 days until the effects of flubendazole can be observed on implanted B . malayi adult worms [36] . Because we were able to optimize the long-term culturing of OvL4 and OvL5 , we were able to discover that only after the OvL5 were treated with flubendazole for 14 days a significant reduction in their motility ( 0 . 3 μM and higher ) could be observed . As our in vitro model contains also human cells ( feeder layer ) , it is possible that flubendazole can be metabolized to hydrolyzed flubendazole [30 , 37] . It is promising that a flubendazole metabolite might also has an intrinsic activity against filariae [34] . The initial IC50 value ( S3 Fig ) of oxfendazole , a sulfoxide metabolite of fenbendazole , and a macrofilaricidal drug used against several filarial species in animal models [35] , was also obtained for OvL5 and is similar to that of flubendazole . Additionally , this in vitro culturing system allowed us to analyze the activity of a set of other FDA repurposed drugs ( auranofin , niclosamide and nitazoxanide ) against O . volvulus . It had been previously shown that auranofin was effective at killing adult Brugia and O . ochengi worms in vitro . Auranofin is of particular interest as a macrofilaricidal drug , as it has a 43-fold higher IC50 against the microfilariae of L . loa compared with the IC50 for adult female O . ochengi [26] . This may help overcome a major contraindication for MDA , as severe adverse reactions to the drug-induced death of L . loa microfilariae is likely in areas where O . volvulus and Wuchereria bancrofti are highly co-endemic with L . loa . Our preliminary assays with several anti-helminthic drugs , have already demonstrated that this system , once further validated for reproducibility , could be rolled out as a drug screen for decision making in macrofilaricide drug development programs . Furthermore , the in vitro model could support additional studies focused on parasite’s development . Overall , we are confident that this novel culturing system could offer a promising and much needed platform to study the activity of novel or repurposed drugs against the human filarial parasite O . volvulus , testing their effects on the growing O . volvulus worms in vitro will support their validation as promising novel and/or repurposed macrofilaricidal drugs .	The human filarial parasite Onchocerca volvulus is the causative agent of onchocerciasis ( river blindness ) . To curb transmission and eventually eliminate the disease , current mass drug administration programs rely on ivermectin that primarily targets the microfilariae , which are released from adult female worms . The female worms , which can live for more than 10–15 years , release millions of microfilariae throughout the course of infection . Thus , to stop transmission of infection and shorten the time to elimination of this disease , a safe and effective drug that kills the adult stage is needed . However , there is no small animal model for an O . volvulus infection that enables investigators to validate macrofilaricidal drugs . In the present study we describe a newly developed in vitro culturing system that supports the growth and development of O . volvulus young adult worms ( OvL5 ) from the L3 stage and show in preliminary assays that these young worms can be used as an auxiliary for screening putative macrofilaricidal drugs against the ultimate target organism , O . volvulus .	[ "Abstract", "Introduction", "Materials", "and", "methods", "Results", "Discussion" ]	[ "invertebrates", "medicine", "and", "health", "sciences", "onchocerca", "volvulus", "helminths", "tropical", "diseases", "parasitic", "diseases", "animals", "onchocerca", "drug", "screening", "physiological", "processes", "developmental", "biology", "age", "groups", "negl...	2019	Development of a preliminary in vitro drug screening assay based on a newly established culturing system for pre-adult fifth-stage Onchocerca volvulus worms

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.